Ask the Scholar
Document scope · 1 page
Scholar
Ask about this object, its catalog metadata, its source description, or the page inventory.
For page-specific OCR and visual context, open one of the page chats.
Scholar Source Context
Document identity
localId
158698194
label
Administration Documents on Climate Change [binder 1] [2]
core
doc
dtoType
document
citationUrl
pageCount
1
Source metadata
id
158698194
contentType
document
title
Administration Documents on Climate Change [binder 1] [2]
citationUrl
collections
Records of the White House Task Force on Climate Change
John Gibson's Files
imageCount
1
hasImages
yes
source
import
hasTranscription
no
Source extras
naId
158698194
levelOfDescription
fileUnit
otherTitles
42-t-146925348-20171094F-001-002-2019
recordType
description
ocrSource
nara-archive
Single page context
seq
1
pageIndex
0
type
document
mediaId
b0307feb39a5f0d7
ocrText
FOIA Number: 2017-1094-F
FOIA
MARKER
This is not a textual record. This is used as an
administrative marker by the William J. Clinton
Presidential Library Staff.
Collection/Record Group:
Clinton Presidential Records
Subgroup/Office of Origin:
WH Task Force on Climate Change
Series/Staff Member:
John Gibson
Subseries:
OA/ID Number:
41017
FolderID:
Folder Title:
Administration Documents on Climate Change [binder 1] [2]
Stack:
Row:
Section:
Shelf:
Position:
S
100
3
9
2
Clinton Presidential Records
Digital Records Marker
This is not a presidential record. This is used as an administrative
marker by the William J. Clinton Presidential Library Staff.
This marker identifies the place of a tabbed divider. Given our
digitization capabilities, we are sometimes unable to adequately
scan such dividers. The title from the original document is
indicated below.
G
Divider Title:
DOE/EIA-0573(96)
Distribution Category UC-950
Emissions of Greenhouse Gases
in the United States 1996
October 1997
Energy Information Administration
Office of Integrated Analysis and Forecasting
U.S. Department of Energy
Washington, DC 20585
This report was prepared by the Energy Information Administration, the independent statistical and analytical
agency within the Department of Energy. The information contained herein should not be construed as advocating
or reflecting any policy position of the Department of Energy or of any other organization.
Preface
Title XVI, Section 1605(a) of the Energy Policy Act of
Information Administration shall annually update
1992 (enacted October 24, 1992) provides:
and analyze such inventory using available data.
This subsection does not provide. any new data
Not later than one year after the date of the enact
collection authority.
ment of this Act, the Secretary, through the Energy
Information Administration, shall develop, based on
The first report in this series, Emissions of Greenhouse
data available to, and obtained by, the Energy In-
Gases 1985-1990, was published in September 1993. This
formation Administration, an inventory of the na-
report-the fifth annual report, as required by law-
tional aggregate emissions of each greenhouse gas for
presents the Energy Information Administration's latest
each calendar year of the baseline period of 1987
estimates of emissions for carbon dioxide, methane,
through 1990. The Administrator of the Energy
nitrous oxide, and other greenhouse gases.
Energy Information Administration Emissions of Greenhouse Gases in the United States 1996
Executive Summary
In 1996, U.S. emissions of greenhouse gases increased
warming, with the effect of carbon dioxide being equal
by 3.4 percent over 1995 emissions, the highest rate of
to 1 (see "Units for Measuring Greenhouse Gases" on
increase in recent years. Although U.S. emissions have
page 5). The GWPs for other greenhouse gases are
been growing since 1991, their growth accelerated in
considerably higher (see discussion in Chapter 1)-
1996. Greenhouse gas emissions expanded more rapidly
Overall, GWP-weighted emissions rose by 8.3 percent
than U.S. energy consumption in 1996, and the growth
between 1990 and 1996 and by 3.4 percent between
of energy consumption (up 3.2 percent) exceeded the
1995 and 1996. On a GWP-weighted basis, carbon di-
growth of the U.S. economy (up 24 percent).¹ Three
oxide emissions account for 85 percent of U.S. green-
principal sources contributed to the growth in U.S.
house gas emissions (Figure ES1). While carbon dioxide
greenhouse gas emissions:
emissions are growing, methane and nitrous oxide
emissions have been roughly stable.
Energy consumption increased more rapidly in 1996
than in recent years, buoyed by strong economic
growth and unusually severe weather. Residential
Figure ES1. U.S. Greenhouse Gas Emissions
and commercial carbon dioxide emissions (includ-
by Gas, 1996
ing their prorated share of electric utility emissions)
expanded by 6.3 and 5.5 percent, respectively.
Million Metric Tons Carbon Equivalent
The rapid growth of relatively low-carbon natural
Energy-Related Carbon
gas consumption, which has tended to moderate the-
1,473.8 (84.3%)
growth of total carbon dioxide emissions in recent
years by "capping" high-carbon coal use, slowed as
natural gas prices increased. Consequently, electric
HFCs, PFCs, and SF,
utilities met the demand for increased electricity
42.4 (2.4%)
largely with coal-fired power generation. Electric
utility carbon dioxide emissions increased by 4.7
Nitrous Oxide
percent, divided between a 2.4-percent rise resulting
32.8 (1.9%)
from increased electricity sales and a 2.3-percent
Methane
increase resulting from the use of fuels with higher
177.3 (10.1%)
carbon content.
Other Carbon Dioxide
22.1 (1.3%)
Estimated emissions of exotic gases, such as hydro-
fluorocarbons (HFCs), perfluorocarbons (PFCs), and
sulfur hexafluoride-paced by increased emissions
Source: EIA estimates documented in this report.
of HFC-134a, the widely accepted substitute for
chlorofluorocarbons (CFCs)-grew by more than 10
percent in 1996, though from very low levels.
Table ES2 excludes several radiatively important gases:
the criteria pollutants carbon monoxide, nitrogen
Table ESI shows trends in emissions of the principal
oxides, and particulates, as well as CFCs and hydro-
greenhouse gases, measured in million metric tons of
chlorofluorocarbons (HCFCs). These gases have am-
gas. In Table ES2; the value shown for each gas is
biguous effects on climate, which are difficult to
weighted by its global warming potential (GWP), which
quantify. In addition, CFCs and HCFCs are specifically
is as a measure of radiative forcing. This concept,
excluded from coverage under the international climate
developed by the Intergovernmental Panel on Climate
treaty, the Framework Convention on Climate Change.
Change (IPCC),2 provides A comparative measure of
(See Chapters 1, 5, and 6 for discussion related to the
the impacts of different greenhouse gases on global
effects and emissions of these gases.)
"Energy Information Administration, Annual Energy Review 1996, DOE/EIA-0384(96) (Washington, DC, July 1997), PP- 5 and 367.
Intergovernmental Panel on Climate Change, Climate Change 1995: The Science of Climate Change (Cambridge, UK: Cambridge University
255, 1996).
Energy information Administration/ Emissions of Greenhouse Gases in the United States 1996
is
EXECUTIVE Summary
The historical emissions estimates for the years 1989
revised from those in last year's report (see the box on
through 1995 presented in this report are only slightly
page 2 "What's New in This Report")
Table ES1. Estimated U.S. Emissions of Greenhouse Gases by Gas, 1969-1996
(Million Metric Tons of Gas)
Gas
1989
1990
1991
1992
1993
1994
1995
P1996
Carbon Dioxide
5,091.8
5,037.1
4,987.3
5,059.8
5,175.9
5,256.1
5,296.9
5,484.9
Methane
31.3
31.6
31.6
31.7
30.8
31.4
30.9
30.9
Nitrous Oxide
0.4
0.4
0.5
0.5
0.5
0.5
0.5
0.4
Halocarbons and Minor Gases
CFC-11, CFC-12, CFC-113
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0.1
HCFC-22
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
*
*
.
*
HFCs, PFCs, and SF 6
Methyl Chloroform
0.3
0.3
0.2
0.2
0.1
0.1
.
-
Criteria Pollutants
Carbon Monoxide
93.5
91.3
68.3
85.3
85.4
89.6
83.5
NA
Nitrogen Oxides
21.1
20.9
20.6
20.7
21.1
21.5
19.7
NA
Nonmethane VOCs
21.7
21.4
20.8
20.3
20.5
21.1
20.7
NA
"Less than 50,000 tons of gas. Estimated hydrofluorocarbon, perfluorocarbon, and sutfur hexafluoride emissions combined totaled
0.008 million metric tons in 1989, rising to 0.034 million metric tons in 1996.
P = preliminary data. NA = not available.
Note: Data in this table are revised from the data contained in the previous EIA report, Emissions of Greenhouse Gases in the
United States 1995, DOE/EIA-0573(95) (Washington, DC, October 1996).
Sources: Carbon dioxide, methane, nitrous oxide emissions: EIA estimates described in Chapters 2, 3, and 4 of this report.
Halocarbons and minor gases: 1990 and 1994 estimates from the U.S. Environmental Protection Agency, Inventory of U.S.
Greenhouse Gas Emissions and Sinks, 1990-1994, EPA-230-R-96-006 (Washington, DC, November 1995). pp. 46-50. Other years
from ETA estimates described in Chapter 5 of this report. Criteria pollutants: U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, National Air Pollutant Emission Trends, 1900-1995, EPA-454/R-96-007 (Research Triangle Park,
NC, October 1996), Tables A-1-A-3, PP: A-2-A-16.
Table ES2. U.S. Emissions of Greenhouse Gases, Based on Global Warming Potential, 1989-1996
(Million Metric Tons of Carbon or Carbon Equivalent)
Gas
1989
1990
1991
1992
1993
1994
1995
P1996
Carbon
1,389
1,374
1,360
1,380
1,412
1,433
1,445
1,496
Methane
179
181
181
162
177
180
177
177
Nitrous Oxide
38
38
38
38
39
40
38
38
HFCs, PFCs, and SF₆
26
25
26
28
27
31
36
42
Total
1,632
1,618
1,606
1,628
1,654
1,684
1,696
1,753
P = preliminary data.
Note: Data in this table are revised from the data contained in the previous EIA report, Emissions of Greenhouse Gases in the
United States 1995, DOE/EIA-0573(85) (Washington, DC, October 1096).
Sources: EIA estimates documented in this report.
Energy Information Administration/ Emissions of Greenhouse Gases in the United States 1996
Executive Summary
Carbon Dioxide
tion of water to accommodate salmon, 1996 hydro-
electric generation was the second highest on
98.5 percent of U.S. anthropogenic carbon dioxide
record.
sions come from the combustion of fossil fuels.
Changes in carbon dioxide emissions can be traced to
Currently, however, the growth in nuclear power gen-
energy consumption trends and changes in the compo-
eration has leveled off, and it is unlikely that future
sition of fossil fuels burned to provide energy services.
hydroelectric generation will often match 1996 levels.
During the 1980s and early 1990s, the energy intensity
World oil prices remain relatively low, and the U.S.
of the U.S. economy and the carbon intensity of U.S.
economy is growing rapidly.
energy consumption steadily declined (Figure ES2).
Severe weather conditions in 1996 produced a series of
anomalous results: residential and commercial natural
Figure ES2. Emissions Intensity of U.S. Gross
gas consumers used 7.8 percent more natural gas and
Domestic Product, Population, Energy
3-5 percent more electricity than in 1995, and natural
Use, and Electricity Production,
gas prices increased sharply. In response to the price
1980-1996
signals, electric utilities reduced their gas consumption
by 15 percent and substituted coal. The result was a
130
Greenhouse Gas Emissions
per Unit of Gross Domestic Product
sharp increase in both total carbon emissions and
emissions per kilowatthour for the electric utility sector,
CO, Emissions
120
por Unit of End-Use Energy
accompanied by rapid increases in both direct (from
natural gas and heating oil) and indirect (from elec-
Greenhouse Gas Emissions
tricity) emissions from the residential and commercial
Index (1990=100)
110
per Capita
sectors. Emissions from the industrial and transporta-
tion sectors increased by a "more normal" 26 percent
and 23 percent, respectively, in 1996 (Figure ES3).
100
Figure ES3. U.S. Carbon Dioxide Emissions
Electric Utility CO, Emissions
90
per-Unit of Electricity Produced
by Sector, 1960-1996
115
0
1980
1985
1990
1996
110
Industrial
Sources: EIA estimates documented in this report.
105
Index (1990 100)
Residential
Several unrelated factors caused the decline:
100
The deregulation of the natural gas industry bore
95
fruit in the form of greatly increasing gas supplies
at low prices. Natural gas use expanded rapidly in
90
the residential, commercial, and industrial sectors,
Transportation
accounting for much of the growth of energy
85
Electric Utility.
consumption.
Commercial
0
Many events of the period-including the Gulf War,
1980
1985
1990
1996
the oil price spike of 1990, the recession of 1991,
and the vogue for utility demand-side management
Source: EIA estimates documented in Chapter 2 of this
programs-tended to restrain the growth of energy
report.
consumption.
Utility operators began to solve nuclear power plant
operating problems and, by 1995, were able to pro-
Methane
duce 17 percent more electricity from nuclear plants
Methane emissions estimates are more uncertain than
than in 1990.
those for carbon dioxide. U.S. anthropogenic methane
With more snowfall in the Pacific Northwest,
emissions have three principal sources: production and
hydroelectric power generation has returned to the
transportation of coal, natural gas, and oil; anaerobic
levels of the early 1980s. Despite widespread alloca-
decomposition of municipal waste in landfills; and
Energy Information Administration/ Emissions of Greenhouse Gases in the United States 1996
xi
raising livestock. Smaller sources include combustion of
coal they produce. Output from these mines does
fossil fuels, rice cultivation, and industrial processes.
not necessarily track with national trends. Further,
Estimated 1996 methane emissions are unchanged from
a number of them have introduced methane capture
emissions in 1995 (Figure ES4).
programs, which have also reduced emissions.
Figure ES4. U.S. Methane Emissions by Source,
1980-1996
Nitrous Oxide
80
Nitrous oxide emissions estimates are more uncertain
Energy Use
than estimates of methane emissions, and the uncertain-
70
ty of the estimation methods makes it difficult to be
Million Metric Tons Carbon Equivalent
60
confident of apparent trends. The principal sources are
Waste Management
believed to be "excess". emissions from agricultural soils
50
associated with fertilizer use, industrial process emis-
Agriculture
sions, and emissions from combustion of fossil fuels.
40
Nitrous oxide emissions, estimated at 0.45 million
30
metric tons in 1996, are essentially unchanged since
1990. Declining agricultural emissions (due to reduced
20
use of nitrogen fertilizers) have offset slowly increasing
energy-related and industrial emissions (Figure ES5).
10
0
Figure ES5. U.S. Nitrous Oxide Emissions
1980
1985
1990
1996
by Source, 1980-1996
Source: EIA estimates documented in Chapter 3 of. this
16
report.
Energy
Note: Methane emissions from industrial sources were
14
approximately 750,000 metric tons in 1996.
Million Metric Tons Carbon Equivalent
12
Agriculture
Methane emissions rose during the late 1980s. The
10
principal cause appears to have been increased pro-
8
duction from a group of underground coal mines with
very high rates of methane emissions. Meanwhile,
Industry
6
emissions from municipal landfills appear to have been
stable, because growth in the volume of solid waste
4
generated was offset by a growing volume of waste
2
burned for energy recovery and by increased recovery
of Lethane at landfill sites.
U
1980
1985
1990
1996
In the 1990s, several factors have tended to limit or
reduce estimated methane emissions:
Source: EIA estimates documented in Chapter 4 of this
The rapid expansion of recycling (particularly of
report.
paper and garden cuttings) and "waste-to-energy"
plants has reduced the volume of material going
Halocarbons and
into landfills. The expansion of landfill methane
capture projects (both for energy and emissions
Related Compounds
control), although much slower, has reduced meth-
ane emissions from existing sources.
Halocarbons and related compounds include chloro-
fluorocarbons (CFCs), hydrochlorofluorocarbons
Coal mine methane emissions are dominated by a
(HCFCs), hydrofluorocarbons (HPCs), perfluorocarbons
relatively small group of underground coal mines.
(PFCs), and other compounds that act as greenhouse
While the coal industry as a whole has been in-
gases. Halocarbons have many uses, but most emissions
creasing output in recent years, these mines have
come from their use as refrigerants in cooling equip-
suffered from vicissitudes particular to their
ment, as solvents, or as blowing agents and from fugi-
locations, production economics, and the grades of
tive emissions in industrial processes.
xii
Energy Information Administration/ Emissions of Greenhouse Gases in the United States 1996
Executive Summary
CFCs are currently being phased out because they
Figure ES6. U.S. Emissions HFCs, PFCs, and
nage the stratospheric ozone layer. The warming
Sulfur Hexafluoride, 1980-1996
its of CFCs and HCFCs are offset to some extent
fuse they also destroy ozone, which is a potent
16
HFC-23
greenhouse gas. Compounds that contain no chlorine
(such as HFCs and PFCs) do not affect ozone, and their
14
effects on climate are therefore easier to measure.
about emissions of "new" HCFCs, such as HCFC-141b
Million Metrio Tons Carbon Equivalent
12
HFC-134a
The available data suggest that emissions of CFCs-
10
about 0.2 million metric tons in 1990-are declining.
Sulfur Hexafluoride
Estimated HCFC emissions (almost entirely HCFC-22,
8
a popular refrigement for home air conditioners) have
6
Perfluorocarbons
been largely stable since 1993. There is little information
4
and HCFC-142b, which are CFC substitutes.
Other
2
HFCs
HFC emissions were very low-perhaps 0.006 million
metric tons-in 1990. Emissions of HFC-23, a byproduct
0
of HCFC-22 production, have also been roughly stable
1980
1985
1990
1996
since 1993. Emissions of the CFC substitutes HFC-134a
Source: EIA estimates documented in Chapter 5 of this
and HFC-152 have risen substantially in the past
report.
3 years, from a base total of less than 0.001 million
metric tons in 1990 (Figure ES6). HFC-134a became the
these gases varies with local atmospheric conditions, it
standard automobile air conditioner refrigerant in 1994,
is not possible to compute their effects directly. As pre-
and emissions will grow rapidly as CFCs are replaced
cursors to urban "smog," their emissions are regulated
throughout the automobile fleet. Consumption of
under the Clean Air Act. The principal source of emis-
HFC-152 is growing rapidly, but it has a relatively low
sions of criteria pollutants is the combustion of fossil
llobal warming potential of 140.
fuels, particularly in motor vehicles.
principal quantifiable source of PFCs is as a
According to estimates from the U.S. Environmental
tive emission from aluminum smelting. Aluminum
Protection Agency, national-level emissions of carbon
smelting rebounded in 1996, after declining in the early
monoxide have been declining since the late 1970s
1990s, increasing the estimated emissions of PFCs. In
(Figure ES7). Emissions of nonmethane volatile organic
recent years, several PFCs have found markets in the
semiconductor industry, but it has proved difficult to
Figure ES7. U.S. Emissions of Selected
obtain reliable information about sales and consump-
Criteria Pollutants, 1980-1995
tion, other than that the numbers are relatively small.
120
Another compound included in this category is sulfur
Cartion Monoxide
hexafluoride, which is used primarily as an insulating
100
gas in electrical switchgear. The amounts used are
uncertain but appear to be quite small (around 1,000
metric tons per year); however, sulfur hexafluoride has
Million Metrio Tons of Gas
80
an extremely high global warming potential (around
25,000), and, hence, even small emissions have dis-
60
proportionate consequences.
40
Nonmethane Volatile Organic Compounds
Criteria Pollutants
20
Nitrogen Oxides
Criteria pollutants (carbon monoxide, nitrogen oxides,
0
and nonmethane volatile organic compounds) are reac-
1980
1985
1990
1995
Live gases that usually decay quickly in the atmosphere.
they are not necessarily greenhouse gases themselves,
Source: U.S. Environmental Protection Agency, Office of Air
but they can promote atmospheric chemical reactions
Quality Planning and Standards, National Air Pollutant Emis-
create tropospheric ozone, which is a potent
sion Trends, 1900-1995, EPA-454/96-007 (Research Triangle
house gas. Because the ozone-creating effect of
Park, NC. October 1996). Tables A-1-A-11, pp. A-2-A-16.
Energy Information Administration/ Emissions of Greenhouse Gases in the United States 1996
xiii
E
EXPENTIVE summary
compounds and nitrogen oxides have been essentially
The IPCC recommends including emissions and seques-
unchanged in recent years.
tration from land use changes in national inventories,
and the U.S. "National Communication" for the Frame-
work Convention follows this practice. 4. The EIA,
Land Use Issues
however, has elected not to include carbon seques-
tration from forestry in. its "total" estimate of U.S.
emissions, for the following reasons:
Changes in land use can also have large, though diffi-
cult to quantify, effects on atmospheric concentrations
There is insufficient information to determine the
of greenhouse gases. In the United States, the expansion
extent, if any, of year-to-year changes in anthro-
of forest land and the growth of existing forests are
pogenic sequestration of carbon. Changes in nation-
responsible for removing large amounts of carbon from
al sequestration rates can be estimated only at 5-
the atmosphere. Several studies of carbon sequestration
year intervals. The current estimate is an annual
in U.S. forests suggest that in the late 1980s and early
average for the period 1987-1992; new data will
1990s, some 111 to 238 million metric tons of carbon
become available in 1998, covering the 1992-1997
was sequestered annually, equivalent to about 8 to 17
period.
percent of U.S. anthropogenic carbon emissions. 3 How-
The magnitude of the sequestration estimate is sub-
ever, considerable uncertainty is associated with this
ject to considerable uncertainty on several counts.
estimate-particularly with the amount of carbon
The largest source of uncertainty is the amount of
sequestered in forest soils.
carbon sequestered annually in forest soils.
R.A. Birdsey and L.S. Heath, "Carbon Changes in U.S. Forests." in LA. Joyce (ed.), Productivity of America's Forests and Climate Change
(Fort Collins, CO: USDA Forest Service, General Technical Report RM-GTR-271, 1995).
U.S. Environmental Protection Agency, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-1993, EPA-230-R-94-014 (Washington
DC, September 1994), P. ES-7.
XIV
Energy Information Administration/ Emissions of Greenhouse Gases in the United States 1996
Clinton Presidential Records
Digital Records Marker
This is not a presidential record. This is used as an administrative
marker by the William J. Clinton Presidential Library Staff.
This marker identifies the place of a tabbed divider. Given our
digitization capabilities, we are sometimes unable to adequately
scan such dividers. The title from the original document is
indicated below.
H
Divider Title:
PRESIDENT
$ 07 OFFICE THE UNITED ADDIONAL PRESIDENTS STATES an COMMITTEE STATES & Z ADVASION 0
FEDERAL ENERGY RESEARCH AND DEVELOPMENT
FOR THE
CHALLENGES OF THE TWENTY-FIRST CENTURY
REPORT OF THE ENERGY RESEARCH AND DEVELOPMENT PANEL
THE PRESIDENT'S COMMITTEE OF ADVISORS
ON SCIENCE AND TECHNOLOGY (PCAST)
SEPTEMBER 30, 1997
EXECUTIVE OFFICE OF THE PRESIDENT
PRESIDENT'S COMMITTEE OF ADVISORS ON SCIENCE AND TECHNOLOGY
WASHINGTON, D.C. 20500
September 30, 1997
President William J. Clinton
The White House
1600 Pennsylvania Avenue N.W.
Washington, DC 20500
Dear Mr. President:
I am pleased to transmit on behalf of the President's Committee of Advisors on Science and
Technology (PCAST) the Executive Summary of our report Federal Energy Research and
Development for the Challenges of the 21st Century. This Executive Summary, unanimously
approved by the PCAST, is in response to your January 14, 1997, letter to me as PCAST Co-Chair
requesting a review of the current national energy R&D portfolio. You asked for recommendations
by October 1 on how to ensure that the United States has a program that addresses its energy and
environmental needs for the next century.
PCAST endorses the report's findings that this country's economic prosperity,
environmental quality, national security, and world leadership in science and technology all
require improving our energy technologies, and that an enhanced national R&D effort is needed
to provide these improvements. The inadequacy of current energy R&D is especially acute in
relation to the challenge of responding responsibly and cost-effectively to the risk of global
climatic change from society's greenhouse gas emissions, in particular, carbon dioxide from
combustion of fossil fuels.
We recommend focusing the government's energy R&D on projects where high potential
payoffs for society as a whole justify bigger R&D investments than industry would be likely to
make on the basis of expected private returns and where modest government investments can
effectively complement, leverage, or catalyze work in the private sector.
The report recommends an increase, over a five-year period, of $1 billion in the
Department of Energy's annual budget for applied energy-technology R&D. The largest shares
of such an increase would go to R&D in energy efficiency and renewable energy technologies,
but nuclear fusion and fission would also receive increases. The composition of the R&D
supported on advanced fossil-fuel technologies would change in favor of longer-term
opportunities, including fuel cells and carbon-sequestration technologies, but the overall spending
level for fossil-fuel technologies would stay roughly constant in real terms.
The proposed total for FY 2003 would return the DOE's real level of effort in applied
energy-technology R&D in that year to about where it was in FY 1991 and FY 1992. In constant
dollars, the average real growth rate would be 8.3 percent per year.
President William J. Clinton
Page 2
September 30, 1997
We respectfully urge, further, that you increase your efforts to communicate clearly to the
public the importance of energy and energy R&D to the nation's future, and we recommend that
you clearly designate the Secretary of Energy as the national leader and coordinator for developing
and carrying out the national energy strategy.
The report also makes recommendations for improving the Department of Energy's
management of its energy R&D portfolio, including the naming of a single individual with
responsibility for the whole portfolio and reporting directly to the Secretary.
PCAST hopes that these recommendations will be helpful to you as you consider how the
United States can best face major energy related challenges as it enters the 21st century. Of
particular importance, prudence requires having in place an adequate energy R&D effort designed
to expand the array of technological options to enable significant reductions in greenhouse gases at
the lowest possible economic, environmental, and social cost.
The energy R&D portfolio we propose will be of crucial importance in meeting that
challenge. Many of the energy-technologies that will help with the problem of climate change,
moreover, will also help address other energy-related challenges, including reducing dependence on
imported oil, diversifying the U.S. domestic fuel- and electricity-supply systems, expanding U.S.
exports of energy technologies, reducing air and water pollution, and reducing the cost, safety and
security risks of nuclear energy systems around the world.
Sincerely,
John A. Young
Co-Chairman
President's Committee of Advisors on Science and Technology
cc: Vice President Al Gore
THE PRESIDENTS COMMITTEE OF ADVISORS ON SCIENCE AND TECHNOLOGY
ENERGY RESEARCH AND DEVELOPMENT PANEL
John P. Holdren (Chairman)
Teresa and John Heinz Professor of Environmental Policy
Harvard University
John Ahearne
Hal Harvey
Virginia V. Weldon, M.D.
Adjunct Professor of Civil and
Executive Director
Senior Vice President,
Environmental Engineering, and
The Energy Foundation
Public Policy
Lecturer in Public Policy
Monsanto Company
Duke University, and
Daniel A. Lashof
Director, Sigma Xi Center
Senior Scientist
Robert H. Williams
Natural Resources Defense
Senior Research Scientist
Richard Balzhiser
Council
Center for Energy and
President Emeritus
Environmental Studies
Electric Power Research Institute
Diana MacArthur
Princeton University
Chair and CEO
Joan T. Bok
Dynamac Corporation
Lilian Shiao-Yen Wu
Chairman of the Board
Research Scientist
New England Electric System
Lawrence T. Papay
IBM Watson Research Center
Senior Vice President and
Robert W. Conn
General Manager
John Young
Dean, School of Engineering
Technology and Consulting
Former President and CEO
University of California.
Bechtel Corporation
Hewlett-Packard Co.
San Diego
Donald L. Paul
fiam L. Fisher
Vice President for Technology
Panel Associates:
TOW Chair
and Environmental Affairs
Department of Geological Sciences
Chevron Corporation
Brian C. Elliott¹
University of Texas at Austin
Vice President
Maxine Savitz
Northern Illinois Gas
Thomas L. Fisher
General Manager
Chairman. President and CEO
AlliedSignal Ceramic
Jefferson W. Tester
NICOR Inc. and
Components
HP. Meissner Professor of Chemical
Northern Illinois Gas
Engineering, and Director. Energy
Laura Tyson
Laboratory
Robert A. Frosch
Class of 1939 Professor of
Massachusetts Institute of
BCSIA
Economics and Business
Technology
John F. Kennedy School of
Administration
Government
University of California.
Study Executive Director
Harvard University
Berkeley
Samuel F. Baldwin
William Fulkerson
Charles M. Vest
National Science and
Senior Fellow. Joint Institute for
President
Technology Council
Energy and Environment
Massachusetts Institute
Agency Representative
University of Tennessee
of Technology
Former Associate Director
Oak Ridge National Laboratory
i
Deputy to Thomas L Fisher
:
Deputy to Charles Vest
STAFF:
Jim Bartis
Rod Judkins
RAND Corporation
Oak Ridge National Laboratory
Mark Bernstein
Nikki Kelly
Office of Science and Technology Policy
National Renewable Energy Lab
Ellison Burton
Cluster Myers
Dynamac Corporation
National Renewable Energy Lab
Paul de Sa
David Pelly
Harvard University
Harvard University
Miriam Forman
Susan Resetar
National Science and Technology Council
RAND Corporation
Agency Representative
Ambuj Sagar
Beverly Hartline
Harvard University
Office of Science and Technology Policy...
This was a Panel of twenty-one persons of diverse backgrounds and viewpoints. tackling an immensely complex
subject. Inevitably. not every member of the Panel is entirely happy with every formulation in the report. But we
are unanimous that the main messages and overall balance in this joint product are correct and appropriate.
THE PRESIDENT'S COMMITTEE OF ADVISORS
ON SCIENCE AND TECHNOLOGY
Chairs
John H. Gibbons
John A. Young
Assistant to the President for Science and Technology
Former President and CEO
Director, Office of Science and Technology Policy
Hewlett-Packard Co.
Members
Norman R. Augustine
Peter H. Raven
Chairman of the Board
Director, Missouri Botanical Garden
Lockheed Martin Corporation
Engelmann Professor of Botany
Washington University in Saint Louis
Francisco J. Ayala
Donald Bren Professor of Biological Sciences
Sally K. Ride
Professor of Philosophy
Director, California Space Institute
University of California, Irvine
Professor of Physics
University of California, San Diego
Murray Gell-Mann
Professor, Santa Fe Institute
Judith Rodin
R.A. Millikan Professor Emeritus of
President
Theoretical Physics
University of Pennsylvania
California Institute of Technology
Charles A. Sanders
David A. Hamburg
Former Chairman
President Emeritus
Glaxo-Wellcome Inc.
Carnegie Corporation of New York
David E. Shaw
John P. Holdren
Chairman
Teresa and John Heinz Professor
D.E. Shaw & Co. and Juno Online Services
of Environmental Policy
John F. Kennedy School of Government
Charles M. Vest
Harvard University
President
Massachusetts Institute of Technology
Diana MacArthur
Chair and CEO
Virginia V. Weldon
Dynamac Corporation
Senior Vice President for Public Policy
Monsanto Company
Shirley M. Malcom
Head
Lilian Shiao-Yen Wu
Directorate for Education and
Member, Research Staff
Human Resources Programs
Thomas J. Watson Research Center
American Association for the Advancement of
IBM
Science
Mario J. Molina
Department of Earth, Atmospheric and Planetary
Sciences
Massachusetts Institute of Technology
Executive Secretary
Angela Phillips Diaz
Federal Energy Research and Development
for the Challenges of the Twenty-First Century
Panel on Federal Energy R&D,
President's Committee of Advisors on Science and Technology
Executive Office of the President of the United States
30 September 1997
EXECUTIVE SUMMARY
The United States faces major energy-related challenges as it enters the twenty-first
century. Our economic well-being depends on reliable, affordable supplies of energy. Our
environmental well-being - from improving urban air quality to abating the risk of global
warming - requires a mix of energy sources that emits less carbon dioxide and other pollutants
than today's mix does. Our national security requires secure supplies of oil or alternatives to it, as
well as prevention of nuclear proliferation. And for reasons of economy, environment, security,
and stature as a world power alike, the United States must maintain its leadership in the science
and technology of energy supply and use.
All of these energy-related challenges to the well-being of this country are made more
acute by what is happening elsewhere in the world. The combination of population growth and
economic development in Asia, Africa, and Latin America is driving a rapid expansion of world
energy use, which is beginning to augment significantly the worldwide emissions of carbon
dioxide from fossil-fuel combustion, increasing pressures on world oil supplies, and exacerbating
nuclear proliferation concerns. Means must be found to meet the economic aspirations and
associated energy needs of all the world's people while protecting the environment and preserving
peace, stability, and opportunity.
Improvements in energy technologies, attainable through energy research and
development, are the key to the capacity of the United States to address - and to help the rest
of the world address - these challenges.
Many of the energy R&D programs of the Federal government, which are primarily
conducted by the Department of Energy (DOE), have been well focused and effective within the
limits of available funding. But these programs, taken as a whole, are not commensurate in scope
and scale with the energy challenges and opportunities the twenty-first century will present.
(This judgment takes into account the contributions to energy R&D that can reasonably be
expected to be made by the private sector under market conditions similar to today's.) The
inadequacy of current energy R&D is especially acute in relation to the challenge of responding
prudently and cost-effectively to the risk of global climatic change from society's greenhouse-gas
emissions, of which the most important is carbon dioxide from combustion of fossil fuels. Much
of the new R&D needed to respond to this challenge would also be responsive to the other
challenges.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 2
SYNOPSIS OF MAIN RECOMMENDATIONS
To close the gap between the current energy R&D program and the one that the
challenges require, the Panel recommends strengthening the DOE applied energy-technology
R&D portfolio by increasing funding for four of its major elements (energy end-use efficiency,
nuclear fission, nuclear fusion, and renewable energy technologies) and restructuring part of the
fifth (fossil-fuel technologies). We also recommend better coordination between the
Department's applied energy-technology programs and the fundamental research carried out in
the program on Basic Energy Sciences; increased Department efforts in integrated analysis of its
entire energy R&D portfolio and the leverage the portfolio offers against the energy challenges of
the next century; targeted efforts to improve the prospects of commercialization of the fruits of
publicly funded energy R&D in specific areas; increased attention to certain international aspects
of energy R&D; and changes in the prominence given to energy R&D in relation to the
Department's other missions, coupled with changes in how this R&D is managed.
Applied Energy-Technology R&D Recommendations
The overall budgets we propose for applied energy-technology R&D to the year 2003,
based on analyses summarized in our main report and set out in more detail in its appendices, are
summarized in Table ES-1. (The table provides these figures both in as-spent dollars, which are
the usual currency of official budget planning, and in constant 1997 dollars, which are more
informative about what is really happening to the size of the effort.)
The applied energy-technology R&D programs, which have been the main focus of the
Panel's study and which are shown in Table ES-1, contain only part of the activities constituting
DOE's Congressional budget lines for "Energy R&D". Table ES-2 shows the relation, under the
FY 1997 Congressional appropriation and the FY 1998 DOE request, between the amount
budgeted for the activities included in our "applied energy-technology R&D" category and the
amounts budgeted for the other activities included under "Energy R&D" in the Congressional
budget lines.
The Panel was not able to review in detail the Basic Energy Sciences budget line (which
includes research in materials science, chemistry, applied mathematics, biosciences, geosciences,
and engineering that is not directed at the development of a particular class of energy sources),
and it did not review at all the other "Energy R&D" budget lines shown in Table ES-2 (which
contain mostly items that are either not very closely linked to advances in civilian energy
technology or are not really R&D at all). Accordingly, we do not offer any recommendations
about the future sizes of these budgets. We note, however, that because advances produced by
research in the Basic Energy Sciences (BES) category provide an important part of the expanding
knowledge base on which progress in applied energy-technology R&D in the public and private
sectors alike depends, the Department may want to consider expanding its support for BES as the
applied energy-technology R&D areas grow.
As indicated in Table ES-1, our proposals for the applied energy-technology R&D
programs would increase spending in that category from $1.3 billion in 1997 to $2.4 billion in
2003, in as-spent dollars. In constant-dollar terms, the increase from 1997 through 2003 is 61
percent, amounting to an average real growth rate of 8.3 percent per year. The proposed figure
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 3
for 2003 would return the DOE's real level of effort in applied energy-technology R&D in that
year to about where it was in FY 1991 and FY 1992.
Table ES-1. Recommended DOE Budget Authority for Applied Energy-Technology R&D
In millions of as-spent dollars
1997
1998
1999
2000
2001
2002
2003
actual
request
Efficiency
373
454
615
690
770
820
880
Fission
42
46
66
86
101
116
119
Fossil
365
346
379
406
433
437
433
Fusion
232
225
250
270
290
320
328
Renewables
270
345
475
585
620
636
652
TOTAL AET
1282
1416
1785
2037
2214
2329
2412
In millions of constant 1997 dollars
1997
1998
1999
2000
2001
2002
2003
actual
request
Efficiency
373
442
584
638
695
721
755
Fission
42
45
63
80
91
102
102
Fossil
365
337
360
376
391
384
371
Fusion
232
219
237
250
262
281
281
Renewables
270
336
451
541
559
559
559
TOTAL AET
1282
1379
1695
1885
1998
2047
2068
Note: What is called "energy end-use efficiency" in this report and is abbreviated as "efficiency" in these tables
appears as "conservation" in many budget documents.
Of the Panel's proposed increases in DOE's applied energy-technology R&D accounts,
the largest in dollar magnitude is in the end-use-efficiency programs, in which annual spending in
FY 2003 would reach $880 million, about $500 million more than in 1997 (as-spent dollars). This
large increase is appropriate because of the high promise of advanced efficiency technologies for
relatively quick-starting and rapidly expanding contributions to several important societal goals,
including cost-effective reductions in local air pollution and carbon-dioxide emissions, diminished
dependence on imported oil, and reductions in energy costs to households and firms.
Improvements in energy efficiency reduced the energy intensity of economic activity in the
United States by nearly one-third between 1975 and 1995, an improvement that is now saving
U.S. consumers about $170 billion per year in energy expenditures and is keeping U.S. emissions
of air pollutants and carbon dioxide about a third lower than they would otherwise be
PCAST Energy R&D Panel Executive Summary
30 September 1997
page J
Table ES-2. Relation of Applied Energy Technology R&D to "Total Energy R&D"
In millions of as-spent dollars.
1997
1998
actual
request
APPLIED ENERGY TECHNOLOGY R&D
1282
1416
"Energy Research": Basic Energy Sciences
641
661
"Energy Research": Other Non-Fusion
539
585
"Other Nuclear R&D"
216
255
"Other Conservation R&D"
177
234
TOTAL "ENERGY R&D" BUDGET LINES
2855
3151
Notes: DOE's Office of Energy Research includes the Department's R&D on fusion energy, as
well as Basic Energy Sciences and some other science and technology programs including
biomedical and environmental research. research in computing. and science education. "Other
Conservation R&D" includes the State and Local Partnership Programs and the Federal Energy
Management Program (which are not really R&D at all). among other items. "Other Nuclear
R&D" includes radioisotope power sources for spacecraft and isotopes for medical applications.
among other items. The Panel included fusion in its analysis of applied energy-technology R&D
(although, as noted in that analysis, much fusion R&D is in fact basic science).
Further major increases in efficiency can be achieved in every energy end-use sector: in
transportation, for example, through much more fuel-efficient cars and trucks; in industry
through improved electric motors, materials-processing technologies, and manufacturing
processes; in residential and commercial buildings through high-technology windows, super-
insulation, more efficient lighting, and advanced heating and cooling systems.
The second largest of the Panel's proposed increases is for renewable energy technologies,
in which annual spending in FY 2003 would reach $650 million, nearly $400 million more than in
1997 (as-spent dollars). This increase makes sense in light of the rapid rate of cost reduction
achieved in recent years for a number of renewable energy technologies, the good prospects for
further gains, and the substantial positive contributions these technologies could make to
improving environmental quality, reducing the risk of climate change, controlling oil-import
growth, and promoting sustainable economic development in Africa, Asia, and Latin America.
Opportunities exist for important advances in wind-electric systems, photovoltaics, solar-
thermal energy systems, biomass-energy technologies for fuel and electricity, geothermal energy,
and a range of hydrogen-producing and hydrogen-using technologies including fuel cells. As in
the case of the proposed increases in energy-efficiency R&D we propose, the increased support
for these renewable-energy technologies would focus on areas where the expected short-term
returns to industry are insufficient to stimulate as much R&D as the public benefits warrant.
Fusion R&D is proposed for the third largest increase annual spending for it in FY 2003
would reach about $100 million more than the 1997 figure in as-spent dollars. In this scenario,
fusion funding would reach by 2002 the $320 million figure recommended in the 1995 PCAST
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 5
study of fusion-energy R&D as a constant level of spending in as-spent dollars to be maintained
from FY 1996 onward. (This earlier PCAST recommendation did not prevail. and fusion funding
fell instead from $369 million in FY 1995 to $232 million in FY 1997.)
The Panel judges this amount warranted for two reasons: (1) About $200 million per year
of it would continue a very productive element of the country's basic science portfolio
(comparing favorably in cutting-edge contributions and valuable spinoffs with other fields in that
category); and (2) the rest is easily justified as the sort of investment the government should be
making in a high-risk but potentially very-high-yield energy option for society, in which the size
and time horizon of the program essentially rule out private funding.
DOE's R&D in nuclear-fission energy systems, which fell 12-fold in real terms between
1986 and 1997, would increase under our proposal from about $40 million per year in FY 1997 to
about $120 million per year in 2003 (as-spent dollars), thereby returning in real level of effort to
that of 1995. Nuclear fission currently generates about 17 percent of the world's electricity; if
this electricity were generated instead by coal, world carbon dioxide emissions from fossil-fuel
consumption would be almost 10 percent larger than they currently are.
Fission's future expandability is in doubt in the United States and many other regions of
the world because of concerns about high costs, reactor-accident risks, radioactive-waste
management, and potential links to the spread of nuclear weapons. We believe that the potential
benefits of an expanded contribution from fission in helping address the carbon-dioxide challenge
warrant the modest research initiative proposed here, in order to find out whether and how
improved technology could alleviate the concerns that cloud this energy option's future. To write
off fission now as some have suggested, instead of trying to fix it where it is impaired, would be
imprudent in energy terms and would risk losing much U.S. influence over the safety and
proliferation resistance of nuclear-energy activities in other countries. Fission belongs in the
R&D portfolio.
Energy from fossil fuels currently contributes 85 percent of U.S. annual energy use and 75
percent of the world's These fuels will continue to provide immense amounts of energy through
the middle of the next century and beyond, under any plausible scenario. We judge that DOE's
current fossil-energy R&D program is about the appropriate size in relation to the array of
relevant needs, opportunities, and likely continuing private-sector fossil-energy R&D activities.
Our proposed budget for DOE's fossil-energy R&D, which increases funding in as-spent dollars
by about $70 million per year between 1997 and 2003, actually holds the real level of effort
approximately level near its FY 1997 value of $365 million per year.
We do, however, recommend some changes in emphasis within this program. Specifically,
we propose phasing out DOE's R&D on near-term coal-power technologies and promptly ending
the funding for direct coal liquefaction, while increasing the Department's R&D on advanced
coal-power programs. carbon capture and sequestration, fuel cells and other hydrogen
technology, and advanced oil and gas production and processing. These changes are designed to
increase the responsiveness of DOE's fossil-energy R&D to the carbon-dioxide and oil-import
challenges (including technology export opportunities that could favorably affect other countries'
carbon emissions and oil imports while improving the U.S. balance of payments). and to improve
the program's complementarity with (or help to stimulate) R&D efforts in the private sector.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 6
Our recommendations for R&D initiatives in the efficiency, renewables, fusion, fission.
and fossil-fuel components of DOE's applied energy-technology portfolio are described in more
detail later in this Executive Summary and are summarized, together with the budgets we propose
for these efforts, in Table ES-3.
Recommendations on Cross-Cutting Issues
The Panel recommends that coordination between the Basic Energy Sciences program
and the applied energy-technology programs be improved using mechanisms such as co-
management and co-funding.
We recommend that the Department make a much more systematic effort in R&D
portfolio analysis: portraying the diverse characteristics of different energy options in a way
that facilitates comparisons and the development of appropriate portfolio balance, in light of
the challenges facing energy R&D and in light of the nature of private-sector and international
efforts and the interaction of U.S. government R&D with them.
After consideration of the market circumstances and public benefits associated with the
energy-technology options for which we have recommended increased R&D, the Panel
recommends that the nation adopt a commercialization strategy in specific areas
complementing its public investments in R&D. This strategy should be designed to reduce the
prices of the targeted technologies to competitive levels, and it should be limited in cost and
duration.
The Panel recommends that the government and government/national-
laboratory/industry/university consortia should engage strongly in international energy
technology R&D and, where appropriate, development and commercialization efforts to
regain and/or maintain the scientific, technical, and market leadership of the United States in
energy technology.
We recommend that overall responsibility for DOE energy R&D portfolio should be
assigned to a single person reporting directly to the Secretary of Energy, and that, similarly, a
single individual should be given the responsibility and authority for coordination of cross-cutting
programs between the applied-technology programs, reporting to the single person responsible for
the overall R&D portfolio.
The Panel recommends that industry/national-laboratory-university oversight committees
should work with DOE to provide overall direction to energy R&D programs, with DOE
facilitating and administering the process; and we recommend that all DOE energy R&D
programs undergo outside technical peer review every 1-2 years, while interim internal process-
oriented reviews are reduced to a minimum.
Additional recommendations and discussion on cross-cutting issues appear later in this
Executive Summary.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page
7
RATIONALE FOR THE RECOMMENDATIONS
The rationale for the recommendations summarized above - and for others to be found in
the more detailed treatment later in this Executive Summary - is presented in what follows in
terms of the importance of energy to our national well-being, the evolution of U.S. and world
energy supply and demand, the challenges this evolution poses to energy R&D, recent trends in
public and private funding for energy R&D, and the implications of those trends (and the energy
R&D status quo) for the prospects of meeting the energy and environmental challenges of the
next century.
The Importance of Energy
The characteristics of the technologies available to this nation and others for energy supply
and energy end-use are critical to our country's economic well-being, environmental quality, and
national security:
Economically, expenditures on energy account for 7 to 8 percent of gross economic
product in the United States and worldwide and a similar fraction of the value of U.S. and
world trade. Experience has shown that periods of excessive energy costs are associated
with inflation, recession, and frustrated economic aspirations. Sales of new energy-supply
technologies globally run in the multi-hundreds of billions of dollars per year.
Environmentally, energy supply accounts for a large share of the most worrisome
environmental problems at every geographic scale - from wood-smoke in Third World
village huts, to regional smogs and acid precipitation in industrialized and developing
countries alike, to the risk of widespread radioactive contamination from accidents at
nuclear-energy facilities, to the build-up of carbon dioxide and other heat-trapping gases in
the global atmosphere.
National security is linked to energy through the increasing dependence of this country
and many others on imported oil, much of it from the politically troubled Middle East:
through the danger that nuclear-weapons-relevant knowledge and materials will be
transferred from civilian nuclear-energy programs into national nuclear arsenals or terrorist
bombs; and through the potential for large-scale failures of energy strategy with economic
or environmental consequences serious enough to generate or. aggravate social and
political instability.
Scientific and technological progress, achieved through R&D, is crucial to minimizing
current and future difficulties associated with these interactions between energy and well-being,
and crucial to maximizing the opportunities. If the pace of such progress is not sufficient, the
future will be less prosperous economically, more afflicted environmentally, and more burdened
with conflict than most people expect. And if the pace of progress is sufficient elsewhere but not
in the United States, this country's position of scientific and technological leadership - and with
it much of the basis of our economic competitiveness, our military security, and our leadership in
world affairs - will be compromised.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 8
Past, Present. and Projected Patterns of Energy Supply
The challenges and opportunities associated with the economic, environmental, and
national-security dimensions of energy have become what they are primarily as a consequence of
the tremendous increase in energy use, and especially fossil-fuel use, over the past century and a
half. This increase, in which world energy use grew 20-fold between 1850 and 1995 and fossil-
fuel use increased more than 100-fold, arose principally from the combination of population
growth and rapid economic development in the industrialized countries.
In contrast, by far the largest part of the future growth of world energy use is expected to
take place in the currently less developed countries of Asia, Africa, and Latin America. Today,
with nearly 80 percent of the world's population, these countries still account for only a third of
the energy use. But if recent trends continue (the "business as usual" energy future), they will
pass the industrialized countries in total energy use (and in carbon dioxide emissions) between
2020 and 2030, and their growth will be the primary driver of a doubling in global energy use
between 1995 and 2030 and a quadrupling between 1995 and 2100.
Energy use in industrialized countries would continue to increase in a business-as-usual
future, but not as rapidly as in the less developed countries and not as rapidly as in the past. A
business-as-usual energy trajectory for the United States would entail increases in energy use,
above the 1995 level, of about 40 percent by 2030 and nearly 75 percent by 2100.
The fossil fuels - oil, natural gas, and coal - accounted for 75 percent of energy supply
worldwide in 1995. The remainder was nuclear energy (6 percent), hydropower (6 percent), and
biomass fuels (13 percent, mostly fuel-wood in developing countries), with wind, solar, and
geothermal energy together contributing less than half a percent. The dominance of the fossil
fuels would decline only slowly in a business-as-usual future: the world as a whole would still be
obtaining perhaps two-thirds of all its energy needs from fossil fuels in 2030 and half or more in
2100. Fossil-fuel resources are adequate to support such an outcome, albeit perhaps with higher
dependence on coal than today, relative to oil and gas.
The United States obtained 85 percent of its energy from fossil fuels in 1995, nearly 40
percent from oil alone (of which half was imported). U.S. fossil-fuel dependence, like that of the
rest of the world. would decline only slowly in a business-as-usual future. U.S. oil imports.
according to the "reference" forecast of the Department of Energy, would grow from 9 million
barrels per day in 1995 to 14 million barrels per day in 2015 and continue to increase for some
time thereafter.
The Challenge to Energy R&D
Improvements in energy technology can and must play a major role in reducing the costs,
increasing the benefits, and alleviating the periis that a business-as-usual energy future without
such improvements would be likely to entail.
Energy-technology improvements. achieved in the United States and then deployed here
and elsewhere, could
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 9
lower the monetary costs of supplying energy,
lower its effective costs still further by increasing the efficiency of its end uses,
increase the productivity of U.S. manufacturing,
increase U.S. exports of high-technology energy-supply and energy-end-use products and
know-how,
reduce over-dependence on oil imports here and in other countries, thus reducing the risk
of oil-price shocks and alleviating a potential source of conflict,
diversify the domestic fuel-supply and electricity-supply portfolios to build resilience
against the shocks and surprises that an uncertain future is likely to deliver,
reduce the emissions of air pollutants hazardous to human health and to ecosystems,
improve the safety and proliferation resistance of nuclear energy operations around the
world,
slow the build-up of heat-trapping gases in the global atmosphere, and
enhance the prospects for environmentally sustainable and politically stabilizing economic
development in the many of the world's potential trouble spots.
The direct and indirect effects of the pursuit of improved energy technologies for these
purposes through appropriately sized, tailored, and publicized R&D programs, moreover, will
strengthen this country's science and technology base, bolster our research universities, build
effective industry-government-university partnerships, help to stem the decline in enrollments of
our most talented young people in science and engineering disciplines, and contribute to
maintaining the global leadership and influence of the United States in relation to scientific and
technological developments worldwide and their application to the betterment of the human
condition.
Among all of these good reasons for adequately funded, suitably focused. effectively
managed energy R&D, one is particularly demanding in what it requires of the R&D effort: the
need to expand the array of energy technologies available for responding cost-effectively to the
risk of global climatic change from greenhouse gases. most importantly carbon dioxide from
fossil-fuel combustion.
Many of the characteristics of this risk and of society's potential responses to it are
subject to considerable uncertainty and controversy. These characteristics aspects include the
pace at which climatic change may become more obvious as greenhouse-gas concentrations grow,
the magnitude and geographic distribution of the ecological and human consequences of such
change, and the impacts on the U.S. and world economies of various measures that might be
undertaken to constrain carbon-dioxide emissions.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 10
If greenhouse-gas-induced climate change were to develop along the path deemed most
likely in the latest assessment by the Intergovernmental Panel on Climate Change (IPCC), there
would be a significant chance that changes in patterns of temperature, humidity, rainfall, soil
moisture, and ocean circulation, plus increases in sea level, would be adversely impacting human
well-being over substantial areas of the planet by some time in the twenty-first century. The IPCC
assessment also indicates that slowing the build-up of carbon dioxide in the atmosphere will be
very difficult to achieve, because of the upward pressure of population growth and economic
aspirations on energy demand, the large energy contribution and long turnover time of the fossil-
fuel technologies that are the primary source of CO₂ emissions, and the long residence time of this
gas in the atmosphere.
Of course, the work of the IPCC to date will not be the last word on the issue of
greenhouse-gas-induced climate change. Some members of the research community think the
IPCC's projections of future climate change and its consequences are too pessimistic. Others
think they are too optimistic. And some contend that adaptation to climate change would be less
difficult and less costly than trying to prevent the change, whereas others argue that a strategy
combining prevention and adaptation is likely to be both cheaper and safer than one relying on
adaptation alone. Within our own Panel there are significant differences of view on some of these
questions.
Notwithstanding these differences, however, the Panel members are in complete
agreement about the implications of the climate-change issue for energy R&D strategy:
First, there is a significant possibility that governments will decide, in light of the
perceived risks of greenhouse-gas-induced climate change and the perceived benefits of a
mixed prevention/adaptation strategy, that emissions of greenhouse gases from energy
systems should be reduced substantially and soon. Prudence therefore requires having in
place an adequate energy R&D effort designed to expand the array of technological
options available for accomplishing this at the lowest possible economic, environmental,
and social cost;
Second, because of the large role of fossil-fuel technologies in the current U.S. and world
energy systems, the technical difficulty and cost of modifying these technologies reduce
their carbon dioxide emissions, their long turnover times, their economic attractiveness
compared to most of the currently available alternatives, and the long times typically
required to develop new alternatives to the point of commercialization, the possibility of a
mandate to significantly constrain greenhouse-gas emissions is the most demanding of all
of the looming energy challenges in what it requires of national and international energy
R&D efforts.
Third (and this finally is the good news about the greenhouse-gas issue), many of the
energy-technology improvements that would be attractive for this purpose also could
contribute importantly to addressing some of the other energy-related challenges that lie
ahead, including reducing dependence on imported oil, diversifying the U.S. domestic
fuel- and electricity-supply systems, expanding U.S. exports of energy-supply and energy-
end-use technologies and know-how. reducing air and water pollution from fossil-fuel
technologies, reducing the cost and safety and security risks of nuclear energy systems
PCAST Energy R&D Panel Executive Summary
30 September 1997
page
11
around the world. fostering sustainable and stabilizing economic development, and
strengthening U.S. leadership in science and technology.
Energy R&D Spending in Decline
Society's capacity to respond effectively to the challenges described above will be
determined in large measure by the output of its energy R&D efforts (as well as to the success of
measures undertaken to ensure that the output is effectively deployed); and the output of R&D
efforts will be substantially affected (with variations depending on the efficiency with which the
R&D is managed and conducted) by the input, that is, by R&D spending.
Nonetheless, while the challenges looming in the energy futures of the United States and
the world have been growing in recent years - or at least growing more apparent -
expenditures on R&D have been declining. In the United States, this has been the case in both the
public and the private sectors, although the decline in funding from the public sector has been
considerably steeper than the decline in funding from industry. Government funding for energy
R&D has also been falling in most other industrialized countries, with the conspicuous exception
of Japan. (The Panel was not able to compile plausible estimates of trends in private-sector R&D
funding in other countries.)
By far the largest part of Federal funding for energy R&D (about 90 percent) comes from
DOE. The Department's FY 1997 budget for applied energy-technology R&D was $1.28 billion,
compared to $2.18 billion five years earlier, in FY 1992, and $6.15 billion twenty years earlier, in
FY 1978 (all figures in constant 1997 dollars).
If one includes DOE's funding for Basic Energy Sciences, the energy R&D decline was
from $6.55 billion in FY 1978 to $3.04 billion in FY 1992 to $1.92 billion in FY 1997. Thus the
decrease in the past 5 years was between 37 and 42 percent, depending on whether BES is
included in the totals, and the decrease between 1978 and 1997 was between 3.4- and 4.8-fold.
As a fraction of real GDP, DOE's 1997 spending for energy-technology was less than half that of
DOE's predecessor agencies 30 years earlier, in 1967, at the height of pre-oil-shock American
complacency about energy supply and energy prices.
Although data for energy R&D in the U.S. private sector are less comprehensive than
those for government spending, the most recent systematic study of energy-industry R&D trends
found that the industry's spending for R&D fell 40 percent in real terms between 1985 and 1994,
from $4.4 billion to $2.6 billion. The R&D spending of the 112 largest U.S. operating electric
utilities fell 38 percent between 1993 and 1996 alone, and the R&D of the four U.S. oil firms with
the largest research efforts approximately halved between 1990 and 1996.
There is evidence that Federal and private investments in R&D in general (that is, not for
energy alone) tend to rise and fall together, rather than one's rising in compensation when the
other goes down. State-government funding of energy R&D in the United States, which was
probably under $200 million in 1995, may follow electric-utility funding downward.
In the G-7 countries other than the United States and Japan, public-sector energy R&D
has fallen sharply, decreasing between 1984 and 1994 by more than 4-fold in both Germany and
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 12
Italy. by about 6-fold in the United Kingdom, and by 2-fold in Canada. Public spending on energy
R&D in France, for which 1984 figures were not available, was declining slowly between 1990
and 1995. Japan, however, increased its public-sector energy R&D spending from about $1.5
billion in 1974 to $4.2 billion in 1980; by 1995 the figure was $4.9 billion, about twice as high as
DOE's energy R&D spending (BES included) in that year.
Explanations and Implications of the Declines in Public and Private R&D
Many explanations for the overall downward trends in energy R&D in recent years
suggest themselves. One important factor is surely low energy prices. World oil prices fell
sharply after 1980, and in the 1990s they have been about where they were in the 1920s and in the
1950s (in inflation-corrected dollars); and natural gas prices in the United States are so low that
no other means of electricity generation can compete with gas-fired combined-cycle power plants
where gas is available. This situation discourages investment in the development of new energy
technologies. The very large demonstration projects in fossil, nuclear, and renewable energy that
accounted for much of the post-oil-shock increase in U.S. Federal energy R&D spending came to
be regarded as costly anachronisms after prices fell again, and their cancellation was, for the most
part, appropriate.
In addition, public-sector spending on energy R&D has experienced downward pressure
from overall budgetary stringency in government and from public and policy-maker complacency
attributable to low prices, no gasoline lines, and high confidence in the capacity of the United
States and allied military forces to preserve access to Middle East oil. DOE has experienced
particular budget-inhibiting scrutiny by critics of "big government", and its energy R&D spending
has been further constrained from within by pressure from larger parts of the Department's budget
(notably environmental cleanup-and nuclear-weapons programs).
In the competitive environment of declining government spending on energy R&D,
moreover, advocates of each energy option have tended to disparage the prospects of the other
options, in hopes of gaining a greater share of the budget for their favorite. Thus the energy
community itself has formulated the arguments that budget-cutters have used to downsize energy
R&D programs one at a time ("renewables are too costly", "fossil fuels are too dirty", "nuclear
fission is too risky", "fusion will never work", "conservation means sacrifice"), with no coherent
energy-community voice calling for a responsible portfolio approach to energy R&D - that is, an
approach that seeks to address and ameliorate the shortcomings of all of the options.
The private sector, meanwhile, has been experiencing a paradigm shift driven by the
increasing globalization of the economy, the revolution in information technology, the increasing
power of shareholders and financial markets over corporate decisions, and deregulation and
restructuring in important parts of the energy business. These factors have combined with low
energy prices and the inherently low profit margins of commodity-based businesses to cause
energy companies to place more emphasis on the short-term bottom line, to decrease risk taking
on broad-based or long-range R&D projects, and to outsource their R&D to specialized R&D
contractors (which may represent a part of private-sector energy R&D that is not shrinking).
It is also possible, finally, that energy R&D in the private sector, the public sector. or both
has become more efficient, in which case declining inputs (funding) need not mean
PCAST Energy R&D Panel Executive Summary
30 September 1997
page
13
correspondingly declining outputs (innovations that can be successfully marketed or that
otherwise improve the human condition). The Panel hopes that this is so, although it is difficult
to verify (partly because there are often significant time lags between the conduct of research and
its effects on the actual flow of innovations. so that if outputs remained high while inputs fell this
might be a temporary condition).
In any case, that the overall declines in both public-sector and private-sector funding for
R&D are largely explainable, and that some of what has disappeared was not needed or effective,
does not establish whether what remains is adequate in relation to current and future needs.
In the private sector, energy R&D has been an important engine of progress, enabling
firms to improve their products and invent new ones, so as to increase their shares of existing
markets, establish and penetrate new ones, and maintain or increase performance while reducing
costs. Perhaps these benefits will flow in adequate measure from the new paradigm; but it is also
possible that important parts of an industrial R&D system that has served our society extremely
well for many decades are now being sacrificed for short-term gain. Concerns have been
expressed that the trend toward decentralization of industrial R&D, for example, could erode the
interconnectedness among people and among different bodies of knowledge that contributes much
to technological innovation in the long term.
Public-sector R&D funding has the responsibility for addressing needs and opportunities
where the potential benefits to society warrant a greater investment than the prospective returns
to the private sector can elicit. Such needs and opportunities relate to public goods (such as the
national-security benefits of limiting dependence on foreign oil), externalities (such as unpenalized
and unregulated environmental impacts), and situations where lack of appropriability of the
research results, or the structure of the market, or the size of the risk. or the scale of the
investment, or the length of the time horizon before potential gains can be realized dilute
incentives for firms to conduct R&D that would greatly benefit society as a whole.
Needs for public-sector R&D can increase over time if the public-goods and externality
challenges grow or if changing conditions shrink the incentives of firms to conduct some kinds of
R&D that promise high returns to society. We have said enough already to suggest that both
things might recently have been happening. But the real test of whether the current portfolio of
public energy R&D is adequate comes from asking whether the R&D programs in the portfolio
are addressing, effectively and efficiently, all of the needs and opportunities where the prospects
of substantial societal benefits are good and the prospective returns to the private sector are
insufficient to elicit the needed R&D. The Panel has analyzed DOE's energy R&D portfolio in
these terms.
ELABORATION OF FINDINGS AND RECOMMENDATIONS
We turn now to what we found, first in relation to the content of the portfolio's major
energy-technology compartments - end-use efficiency, fossil-fuel technologies. nuclear
technologies (fission and fusion), and renewable-energy technologies - and then in relation to
cross-cutting issues including the role of Basic Energy Sciences. portfolio analysis,
commercialization considerations. international dimensions. and DOE management of its energy
R&D programs.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 14
End-Use-Efficiency Technology
Between 1975 and 1986, the United States increased its energy efficiency by almost a
third. This extraordinary achievement helped pull the country out of its two oil shocks and
their attendant stagflation. Efficiency improvements now save U.S. consumers some $170
billion per year; and U.S. emissions of air pollution and CO₂ have been reduced by a third or
more from their expected values.
Challenges and Opportunities
Those achievements are instructive as we look at future energy, economic, and
environmental issues. Technological advances and investments in energy efficiency helped
rescue the U.S. economy once, and gave the country decades of breathing room to deal with
the energy problem. Many of these advances were made possible by DOE-sponsored R&D.
Can a similar improvement be achieved in the years ahead?
The Panel believes it can. We find that investments in energy efficiency are generally
the most cost-effective way to simultaneously reduce the risks of climate change, world oil-
supply interruptions, and local air pollution, and to improve the productivity of the economy.
We have reviewed technologies that can reduce energy use in automobiles by half or more; in
motors and drive systems by half; and in buildings by over 70 percent. Many of these
technologies are in their infancy, and require a serious R&D effort to make them commercially
viable. Others are near market readiness, but need standards and incentives to ensure they
spread rapidly.
Budget, Goals, and Initiatives
The Panel recommends that the R&D components of the DOE's energy efficiency
budget grow steadily over the next 5 years, from $373 million to $755 million (constant
1997$). The Panel has identified the following goals (some pre-existing, and some newly
proposed hee) for each of the sectors:
Buildings. To fund and carry-out research on equipment, materials, electronic and
other related technologies and work in partnership with industry, universities, and state and
local governments to enable by 2010: (1) the constructing of 1 million zero-net-energy
buildings; and (2) the construction of all new buildings with an average 25-percent increase in
energy efficiency as compared to a new building in 1996. Additional longer term research in
advanced energy systems and components will enable all new construction to average 70
percent reductions and all renovations to average 50 percent reductions in greenhouse-gas
emissions by 2030.
Industry.
By 2005, develop with industry a more than 40-percent efficient
microturbine (40 to 300 kW), and introduce a 50-percent efficient microturbine by 2010. By
2005, develop with industry and commercially introduce advanced materials for combustion
systems to reduce emissions of nitrogen oxides by 30 to 50 percent while increasing efficiency
PCAST Energy R&D Panel Executive Summary
30 September 1997
page
15
5 to 10 percent. By 2010, achieve a more than one-fourth improvement in energy intensity of
the major energy- consuming industries (forest products. steel. aluminum. metal casting,
chemicals, petroluem refining, and glass) and by 2020 a 20 percent improvement in energy
efficiency and emissions of the next generation of these industries.
Transportation. By 2004, develop with industry an 80-mile-per-gallon production
prototype passenger car (existing goal of the Partnership for a New Generation of Vehicles -
PNGV). By 2005, introduce a 10-mpg heavy truck (Classes 7 and 8) with ultra low emissions
and the ability to use different fuels (existing goal); and achieve 13 mpg by 2010. By 2010,
have a production prototype of a 100-mpg passenger car with zero equivalent emissions. By
2010, achieve at least a tripling in the fuel economy of Class 1-2 trucks, and double the fuel
economy of Class 3-6 trucks.
The R&D areas requiring increased funding to meet these goals have been identified.
The Department has a sufficiently rich agenda, management expertise, history of success, and
most important, potential for future contribution, to justify these increases.
Further Findings and Recommendations
The buildings program needs high-profile leadership from within the Administration,
closer links with industry, and better mechanisms to distribute its research results. These
elements could be brought together in the Buildings for the 21st Century Initiative." The
codes and standards program needs to be expanded to give greater technical assistance to states
and to speed internal progress.
The industries program is effective. It should be expanded to include more industries,
and the crosscutting research - which develops technologies for use in many industries -
should grow significantly.
Transportation research, most notably the PNGV, is extremely valuable. The PNGV
program is insufficiently funded and cannot meet all its goals at current levels. It should be
complemented by a "PNGV II" to augment efforts on long-term technologies, such as fuel
cells, with extraordinary potential after 2005. PNGV also needs to give greater attention to
air-quality issues, to ensure that technologies selected do not undermine national and state
clean-air programs. The Administration must also develop new transportation policies that
shift the auto fleet, over time, toward higher efficiency. And advanced vehicle development
programs should be coordinated with alternative fuels programs to ensure they are
complementary for transportation systems of the future.
R&D in the Department of Transportation should be reorganized around clear public
interest goals, and Transportation's energy and environmental pursuits should be consonant
with DOE's goals. DOT should pursue more multimodal research and system optimization
and should increase its focus on developing integrated transit systems with improved
efficiency, to reduce urban congestion and enhance air quality. The Automated Highway
System research needs to be thoroughly evaluated, key technical assumptions must be
PCAST Energy R&D Panei Executive Summary
30 September 1997
page 16
documented and peer-reviewed, and then the program should be reorganized around the public
interest goals mentioned above.
Increasing energy efficiency has an extraordinary payoff. It simultaneously saves
billions of dollars, reduces oil imports and trade deficits, cuts local and regional air pollution,
and cuts emissions of carbon dioxide. DOE research, complemented by sound policy, can
help the country increase energy efficiency by a third or more in the next 15 to 20 years.
Fossil-Energy Technology
Fossil fuels supply 85 percent of U.S. energy and 75 percent of all energy globally.
They will continue to be essential to the energy economies of the United States and the world
well into the twenty-first century. R&D on fossil-fuel technologies is warranted to minimize
the costs, impacts, and risks of this continuing reliance on fossil fuels and to exploit the
opportunities it represents for U.S. industry and the U.S. economy.
Challenges and Opportunities
DOE Fossil Energy R&D programs are directed - appropriately in the Panel's
judgment - at two important challenges: (1) reducing the environmental impacts (including
CO₂ emissions) that constrain fossil-fuel use; and (2) reducing the vulnerability of the
economy to oil price shocks (caused by excessive dependence on imported oil and potential
instabilities in the Middle East) by helping ensure the availability of secure and affordable
transportation fuels. In the process, the Department aims to maintain U.S. science and
technology leadership in fossil-fuel related fields.
Over the past two decades, enormous progress has been made in reducing the
environmental impacts of fossil-fuel use - particularly of coal use in electric power
production - in cost-effective ways. This progress has partly been the result of DOE-
industry collaborative R&D and the Clean Coal Technology Demonstration Program. DOE
seeks to maintain this progress through pursuit of an idea called Vision 21, with the objective
of economical coal and gas power and fuels technology with zero-to-small CO₂ emissions and
very low emissions of other air pollutants. This is a most ambitious goal, requiring significant
breakthroughs to achieve very high efficiencies of conversion to electricity (and fuels) and
cost-effective methods for separating and sequestering CO₂.
In the United States, natural gas has become the fuel of choice for new electric
generation because of its low cost, small environmental impacts, relatively small scale
(yielding versatile siting and quick installation), and rapidly advancing turbine technology, and
because of the competitive pressures of electric industry restructuring. This trend to natural
gas is likely to continue for several decades and contributes positively to DOE's environmental
objective, particularly by reducing CO₂ emissions to the extent that gas replaces coal.
As a consequence, the major markets for advanced coal power and fuels technologies
will not be in the United States but in coal-intensive developing countries such as China and
India where gas is not widely available for these purposes. Providing attractive coal
PCAST Energy R&D Panel Executive Summary
30 September 1997
page
17
technologies that are much more efficient with greatly reduced CO₂ and other emissions
contributes to DOE environmental objective. For the United States to take advantage of this
environmental opportunity, it must maintain technological leadership in coal-power
technologies and develop a strong international program including collaborative R&D,
development, and commercialization activities. This will require a paradigm shift away from
the current focus on the U.S. market and toward a focus on coal-intensive developing
countries.
Relative to the challenge of ensuring secure and affordable transportation fuels, DOE
R&D is developing and demonstrating technologies that can enhance domestic oil and gas
production, diversify supply, and reduce the cost of converting natural gas (and coal, biomass
and waste) to clean fuels for transportation. Activities to enhance production include
technology transfer to independent oil and gas producers to help bolster production from
mature resources and high risk R&D investments at the front end of the resource cycle for
frontier provinces. The potential return to the government from taxes and royalties alone
justifies the investment, not to mention reducing balance of payment imbalance and losses to
the economy in event of a future oil price shock. It is good insurance both from the point of
view of oil dependence and for the climate change issue because of the importance of natural
gas as a transition fuel during the next century.
Budget, Goals, and Initiatives
The Panel's analysis of these challenges and opportunities leads us to recommend that
the that the Fossil Energy budget remain at about the current level in constant dollars but with
a significant reorientation and new initiatives aimed at Vision 21, gas as a transition fuel, and a
comprehensive transportation fuel R&D strategy.
Coal and Gas Power and Fuels. The Panel endorses Vision 21 as the long-term
objective and recommends reorientation of DOE R&D priorities toward it. This should
include continued emphasis to improve efficiency of the combined cycle using high
temperature fuel cells, development of advanced gasification technologies (for coal, biomass,
or waste) for the flexible production of power and clean transportation liquid fuels (ultimately
hydrogen and separated CO₂). It should also include initiating a science-based CO2.
sequestration program in cooperation with the US Geological Survey, industry, and
universities, with an annual budget rising to $20 million dollars or more in 2003. Hydrogen
may prove to be the transportation fuel of the future if fuel cells become the power source of
choice for vehicles. and fossil fuels are the likely least expensive route to hydrogen assuming
sequestration is practical.
Phase-Outs. As part of this reorientation, the Panel recommends that the Department
terminate: (1) direct liquefaction of coal, because it doesn't fit Vision 21; (2) the solid fuels
and feedstocks program, directing the funding instead toward a comprehensive, science-based
program to reduce hazardous air emissions from existing and future coal power plants; and (3)
the Low Emissions Boiler System program. It should phase out near-term clean-coal programs
that do not contribute to Vision 21 or to providing much better low-CO₂-emissions technology
choices for developing countries.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 18
Oil and Gas Production and Processing. Because of its importance as a transition fuel
for the United States in controlling CO₂ emissions, the Panel recommends more intense effort
on natural gas production and processing, including a major initiative for DOE to work with
USGS, the Naval Research Lab. Mineral Management Services, and the industry to evaluate
the production potential of methane hydrates in US coastal waters and world wide. The
resource is very large indeed, in the range of 100,000 to 1,000,000 Tcf (trillion cubic feet).
This research might well interface with hydrogen-production and CO₂-sequestration efforts
with CO₂ hydrates as the sequestered state of the gas.
Transportation Fuels Strategy. The Panel recommends that DOE develop a
comprehensive transportation fuels strategy, beginning with an analysis of the potential for
technologies to increase the price elasticity of oil supply and demand including the impact of
substitutes. This effort should include, for example, R&D focused on reducing the cost of
producing transportation fuels from natural gas and work on indirect liquefaction of coal and
biomass. Such an effort is supportive of Vision 21 and may improve its flexibility for
combined fuel and power generation, including eventually producing hydrogen for central or
distributed use with CO2 sequestering.
Nuclear-Energy Technology
Nuclear energy can be generated by fission (the splitting of a nucleus) or by fusion (the
joining of two nuclei). Neither fission nor fusion reactions generate greenhouse gases or the air
pollutants that produce urban smogs and regional acid precipitation. Fission power currently
provides about 17 percent of the world's electric power, with 442 nuclear power reactors
operating in 30 countries and 36 more plants under construction. Fusion power requires much
additional work in the quest to make the fusion reaction self-sustaining and to design and build
practical fusion power plants; the most optimistic timetable for fusion to reach
commercialization is another half century. But the potential benefits of fusion are so large that
fusion R&D is an important component of current energy R&D portfolios in the United States
and internationally.
Challenges and Opportunities: Fission
Several problems compromise fission's potential as an expandable energy source today
and into the future: disposal of spent nuclear fuel; concerns about nuclear weapons
proliferation; concerns about the safe operation of plants; and uncompetitive economics. But
given the projected growth in global energy demand as developing nations industrialize, and
given the desirability of stabilizing and reducing GHG emissions, it is important to establish
fission energy as a widely viable and expandable option if this is at all possible. A properly
focused R&D effort to address the problems of nuclear-fission power - economics, safety,
waste, proliferation - is therefore appropriate. World leadership in nuclear-energy
technologies and the underlying science is also vital to the United States from the perspective
of national security, international influence, and global stability.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 19
Although the United States has the largest number of operating reactors of any country
in the world, the outlook is that no new nuclear plant will be built in this country in the next
10 to 20 years. The decline of nuclear power in the United States has resulted from many
factors: a sharp drop in annual electricity consumption growth rates, low gas prices and
improved efficiency of gas-fired combined cycle plants, rapid escalation of nuclear plant
construction costs, the unresolved problems of waste disposal and storage, and concerns about
proliferation and safety. These factors, combined with the upcoming deregulation of the
electric utilities, may lead to early shutdown of operating nuclear plants in the United States.
Budget, Goals, and Initiatives: Fission
Based on its analysis of the potential and problems of fission power, the PCAST
Energy R&D Panel recommends that nuclear fission R&D be increased from $42 million in
FY 1997 to $119 million in FY2003 (as-spent dollars). Included in these totals throughout the
period is about $6 million per year for university programs, including fellowships and fuel
support for university reactors. The Panel makes the following further observations and
recommendations about the fission R&D effort:
Operating Reactors. Extending the operation of nuclear plants will make it easier to
meet GHG emission goals. The panel recommends that DOE work with its laboratories and
the utility industry to develop a program to address the problems that may prevent continued
operation of current plants. We recommend such a program be funded at $10 million per year,
to be matched by industry.
Nuclear Energy Research Initiative. DOE should establish a new program - the
Nuclear Energy Research Initiative - funded initially at $50 million per year and increasing
by FY 2002 to $100 million per year (as-spent dollars), which would competitively select
among proposals by researchers from universities, national laboratories, and industry to
address key issues affecting the future of fission energy including: proliferation-resistant
reactors or fuel cycles; new reactor designs with higher efficiency, lower-cost, and improved
safety to compete in the global market; lower-output reactors for use in settings where large
reactors are not attractive; and new techniques for on-site and surface storage and for
permanent disposal of nuclear waste. This approach is in contrast to the traditional style of
directed research by DOE Nuclear Energy program (in which the program office defines the
topics, milestones, and scope) and follows instead a model along the lines of the
Environmental Management Science Program (EMSP).
Coordination: DOE should improve coordination and integration among the eight DOE
program offices sponsoring R&D applicable to fission energy.
Challenges and Opportunities: Fusion
The objective of DOE's fusion energy sciences program is to develop the scientific and
technological basis for fusion as a long-term energy option for the United States and the world.
The fusion R&D program is strongly centered in basic research and supports the important
field of plasma science. Results and techniques from fusion plasma science have had
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 20
fundamental and pervasive impact in many other scientific fields, and they have made
substantial contributions to industry and manufacturing. Since 1970, fusion power in
experiments has increased from less than 0.1 watt to more than 10 megawatts.
The nation's fusion energy research program has received three major reviews since
1990, the most comprehensive being the 1995 study by the PCAST Panel on the U.S. Program
of Fusion Energy Research and Development.(PCAST-95). PCAST-95 recommended an
annual budget of $320 million. In FY 1996, Congress reduced the fusion budget by about
one-third and directed DOE to restructure its fusion energy program. The present funding
level of $230 million is too low in the view of the PCAST Energy R&D panel; it allows no
significant U.S. activity relating to participation in an international program to develop
practical low-activation materials; reduces the level of funding for the design of the
International Thermonuclear Experimental Reactor (ITER); forced an early shutdown for the
largest U.S. fusion experiment; and canceled the next major U.S. plasma science and fusion
experiment. It also limited the resources available to explore alternative fusion concepts.
Budgets, Goals, and Initiatives: Fusion
Based on its analysis of the potential of fusion power and the challenges and
opportunities in this field, as just described, the PCAST Energy R&D Panel recommends that
fusion R&D funding be increased from its annual level of $232 million in the FY 1997
appropriation to reach $320 million per year by FY2002 (as-spent dollars). This would restore
fusion R&D funding to the level
which the 1995 PCAST study of fusion-energy R&D recommended be maintained from FY
1996 onward.
Our Panel reaffirms support also for the specific elements of the 1995 PCAST
recommendation that the program's budget-constrained strategy be around three key principles:
(1) a strong domestic core program in plasma science and fusion technology; (2) a
collaboratively funded international fusion experiment focused on the key next-step scientific
issue of ignition and moderately sustained burn; and (3) participation in an international
program to develop practical low-activation materials for fusion energy systems. The Panel
makes the following further observations about the fusion R&D effort:
International Collaborations. The U.S. program should establish significant
collaborations with both the JET program in Europe and the JT-60 program in Japan. Such
collaboration should provide experience in experiments that are prototypes for a burning
plasma machine, such as ITER, and that can explore driven burning plasma discharges
ITER. The Panel judges that the proposed 3-year transition between completion of the
EDA and an international decision to construct is reasonable and that the ITER effort merits
continued U.S. involvement. It would be helpful to all parties in the ITER enterprise if at
least one of the parties would express, within the next year or two, its intention to offer a
specific site for ITER construction by the end of the 3-year period. Clearly, one major hurdle
to ITER construction is its total project cost, most recently estimated to be $11.4 billion, with
the host party expected to fund a substantial share. If the parties agree to move forward to
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 21
construction, the U.S. should be prepared to determine. with stakeholder input, what the level
and nature of its involvement should be. The PCAST Energy R&D Panel believes that if no
party offers to host ITER in the next three years, it will nonetheless be vital to continue
without delay the international pursuit of fusion energy. A more modestly scaled and priced
device aimed at a mutually agreed upon set of scientific objectives focused on the key next-step
issue of burning plasma physics may make it easier for all parties to come to agreement.
Renewable-Energy Technology
Renewable energy technologies (RETs) can provide electricity, fuels for transport, heat
and light for buildings, and power and process heat for industry. These technologies generally
have little or no emissions of greenhouse gases (GHGs), air pollutants, or other environmental
impacts. RETs can also offset imports of foreign oil and offer important economic benefits;
for example growing biomass energy crops on excess agricultural lands would increase farm
income while potentially allowing a reduction in Federal farm income support programs.
Challenges and Opportunities
The primary challenge facing RETs today is relatively high unit costs, but remarkable
progress has been over the past two decades. Costs of energy from RETs such as wind
turbines and photovoltaics (PVs) have come down by as much as 10 times. Much further
progress is expected, to the extent that RETs could become major contributors to U.S. and
global energy needs over the next several decades. The Shell International Petroleum
Company, for example, projects that by 2025 renewable energy sources could contribute to
lobal energy one-half to two-thirds as much as fossil fuels do at present, with new renewable
sources (excluding hydropower and traditional biomass) accounting for one-third to one-half of
total renewables.
Much of the global market growth for RETs, as well as for total energy, will take place
in developing countries. The small scales and modularity of most RETs are well matched to
energy technology needs in developing countries. Also, the inherent cleanliness of most RETS
will have a special appeal, making it possible to reduce environmental problems without
resorting to complex regulatory controls as is done for conventional energy systems.
Budget, Goals, and Initiatives
In light of the remarkable progress already made in many areas of DOE Renewable
Energy program, the good prospects for further gains, and the substantial potential impacts
renewables could have in addressing the multiple challenges posed by energy system in the
United States and worldwide, the PCAST Energy R&D Panel believes that the Renewable
Energy R&D Program should be substantially expanded. from annual spending of $270 million
in FY 1997 to a level of about $650 million in 2003 (as-spent dollars), with goals that include
the following:
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 22
Wind. Reduce by 2005 wind electricity costs to half of today's costs. so that wind
power can be widely competitive with fossil-fuel-based electricity in a restructured electric
industry, through R&D on a variety of advanced wind turbine concepts and manufacturing
technologies.
Photovoltaics (PV): Pursue R&D that would lead to PV systems prices falling from
the present price of $6,000/kW to $3,000/kW in 5 years, to $1500/kW by 2010, and to
$1,000/kW by 2020. R&D activities should include assisting industry in developing
manufacturing technologies, giving greater attention to balance of system issues, and
expanding fundamental research on advanced materials.
Solar Thermal Electric Systems. Strengthen ongoing R&D for parabolic dish and
heliostat/central receiver technology with high temperature thermal storage, and develop high
temperature receivers combined with gas-turbine based power cycles; goals should be to make
solar-only power (including baseload solar power) widely competitive with fossil fuel power
by 2015.
Biopower. Enable commercialization, within ten years, of advanced energy-efficient
power-generating technologies that employ gas turbines and fuel cells integrated with biomass
gasifiers, building on past and ongoing R&D for coal in such configurations, and exploiting
the advantages of biomass over coal as a feedstock for gasification. These technologies could
be widely competitive in many developing country markets and in U.S. markets that use
biomass residues or use energy crops in systems that derive coproducts from biomass.
Geothermal Energy. Continue work on hydrothermal systems, and reactivate R&D on
advanced concepts, giving top priority to high-grade hot dry-rock geothermal; this technology
offers the long-term potential, with advanced drilling and reservoir exploitation technology, of
providing heat and baseload electricity in most areas.
Biofuels. Accelerate core R&D on advanced enzymatic hydrolysis technology for
making ethanol from cellulosic feedstocks, with the goal that between 2010 and 2015 ethanol
produced from energy crops would be fully competitive with gasoline as a neat fuel, in either
internal combustion engine or fuel cell vehicles; coordinate this development with the
biopower program so as to co-optimize the production of ethanol from the carbohydrate
fractions of the biomass and electricity from the lignin using advanced biopower technology.
Hydrogen. Carry out R&D on hydrogen using and producing technologies; coordinate
hydrogen-using technology development with proton-exchange-membrane fuel-cell vehicle
development activities in the Department's Energy Efficiency program. Give priority in
hydrogen-production R&D to co-optimizing the production of hydrogen from fossil fuels and
sequestration of the CO₂ separated out during the production process, in collaboration with the
Fossil Energy program.
Hydropower. To sustain and increase over 92,000 MWe of hydro capacity, additional
R&D is needed to provide a new generation of turbine technologies that are less damaging to
fish and aquatic ecosystems. By deploying such technologies at existing dams and in new low-
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 23
head. run-of-river applications, as much as an additional 50,000 MWe could be possible by
2030.
Crosscutting and other programs. Crosscutting programs that should be strongly
supported include Resource Assessment, International Programs, and Analysis. In addition,
R&D is needed energy storage, electric sytems, and systems integration.
Further Findings and Recommendations
The Panel believes that there are good prospects that these goals can be realized with
the combination of an expanded R&D effort and appropriate demonstration and
commercialization initiatives. DOE program has demonstrated remarkable gains in technology
performance and cost reductions and has laid the foundation for large further gains. The R&D
effort should be intense over the course of the next decade, with much more emphasis than at
present in DOE program on both core applied research and development and fundamental
research directed to serving needs identified in the programs.
For technologies that continue to show promise, R&D budgets should be sustained at
the elevated levels for several years (the number varying with the technology) until the
technologies become established in the market, the industry has sufficient revenues from these
RET markets to shoulder a greater share of needed continuing R&D, and government's role
can be reduced to supporting mainly long-term R&D. For both wind power and biopower
most of the principal R&D goals could be met in a decade or less; for these technologies
Federal R&D budget support could thereafter begin to decline. For other technologies it will
ake longer, but in nearly all cases principal program goals should be achievable in less than 20
cars.
Cross-Cutting Issues
In what follows we elaborate briefly our findings and recommendations relating to four
sets of issues that cut-across the applied energy-technology R&D programs discussed above:
the relation of DOE's Basic Energy Sciences program to applied energy-technology R&D,
analysis of the portfolio as a whole and the leverage it offers against the energy challenges
faced by the nation and the world; considerations related to commercialization of the fruits of
R&D; and certain international aspects of R&D.
Links Between Applied Energy Technology R&D and Basic Energy Sciences
The Panel's review of DOE energy R&D activities identified many areas where
technological advance could be accelerated if more attention were given to fundamental questions
identified in these programs. Examples include better understanding of reactions at the interface
of electrodes and electrolytes in fuel cells, the capacity of carbon nanostructures for hydrogen
storage, the chemistry and fluid dynamics of CO₂ storage in saline aquifers. the physics of thin-
film photovoltaic materials, and many others. The Panel found that linkages between the Basic
Energy Sciences (BES) programs (where such issues are investigated) and the applied energy-
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 24
technology programs (where the findings could be put to use) need to be strengthened in many
cases.
While the technology programs do benefit today from the growing body of fundamental
knowledge being generated under BES programs. they would benefit much more if BES were to
address specific questions identified as important in these programs. The Panel recommends that
BES allocate additional resources to support fundamental research activities addressing needs of
the technology programs. This could be facilitated by mechanisms such as co-management and
co-funding with - or budget sign-off by, or re-routing budgets through the applied energy-
technology programs.
Our recommendation that BES direct some of its resources to serving these needs might
raise concerns that the creativity of basic science will be lost if it is constrained by premature
thought of practical use, and that applied research invariably drives out pure, if the two are mixed.
What is being sought here, however, is not to redirect BES resources to applied research. The
technology programs support applied research but give little attention to addressing fundamental
questions such as the above. The net effect of this recommendation should be to expand, not
diminish, the portfolio of fundamental research activities within the limits of overall budget
constraints. In light of the growing interest among policy planners in harnessing science for the
technological race in the global economy, the allocation of some BES resources to the
development of fundamental research programs that would serve the energy technology programs
should add to the political appeal of supporting basic research generally.
Portfolio Analysis and Leverage
Developing the appropriate degree of diversity and balance in the Department's overall
energy R&D portfolio is difficult. Technologies have many different attributes - cost (of the
R&D to develop them and of the technologies themselves, once they are developed),
performance, risk, return, potential contributions over time to energy and environmental goals,
and others. How can one fairly evaluate the many R&D alternatives and select an R&D
portfolio that best meets our national goals and needs? No single quantitative measure can
encompass the range of relevant attributes. One technology may have substantial
environmental benefits, a second may contribute more to national security, a third may have
only modest benefits but have low risks and costs to develop.
The Panel has worked hard at exploiting and refining various ways to portray the
diverse characteristics of different energy options in a way that facilitates comparisons and the
development of an appropriate portfolio balance in light of the challenges facing energy R&D
and in light of the nature of private-sector and international efforts and the interaction of U.S.
government R&D with them. We have made some progress, but a much larger and
continuing effort in this direction by the Department of Energy itself is called for. (In saying
this we echo one of the strongest recommendations of the 1995 Secretary of Energy Advisory
Board report on Strategic Energy R&D - a recommendation that alas has so far borne little
fruit.) Such analyses should be done on a regular basis as national needs and R&D options
and opportunities change. We recommend that DOE regularly and systematically conduct -
with external peer review - a portfolio analysis across the breadth of R&D options and to use
this as an input to overall program planning.
PCAST Energy R&D Panei Executive Summary
30 September 1997
page 25
The potential overall impact of the sector-by-sector energy R&D portfolio developed by
the Panel can be illustrated by some simple "back-of-the-envelope" analyses. Examples for
oil imports and carbon emissions are schematically shown in figures ES-1 and ES-2; details of
these highly simplified projections are provided in Chapter 7. For clarity, only a few. highly
aggregated sets of technologies are shown.
Consider oil imports. Under "business-as-usual conditions, U.S. oil imports could
increase from 8.5 million barrels per day at a cost of $64 billion dollars in 1996 to nearly 16
million barrels per day at a cost of $120 billion (assuming $20 dollars per barrel) in 2030.
With continued R&D to increase domestic production from marginal oil supplies, an
aggressive ethanol program (based on cellulosic biomass, not corn), and rapid development
and penetration of the market by PNGV and light- and heavy-duty truck technologies, we
estimate that this import could be reduced to something on the order of 6 million barrels per
day oil import demand in 2030, as illustrated in figure ES-1. Estimates of this sort are
necessarily highly approximate, since they depend not only on the somewhat unpredictable
pace of R&D successes but also future market conditions and measures taken to speed market
penetration under whatever those conditions are; nonetheless, such "ballpark" estimates give
at least a rough indication of the magnitude of the challenge the nation faces and and size of
the opportunity to address it with the stronger R&D program outlined here.
Potential impact on carbon dioxide emissions (customarily measured in tons of carbon
contained in the emitted CO₂) is clearly also a crucial element of a portfolio's leverage against
the energy-related challenges of the next century. Figure ES-2 illustrates. in a highly stylized
and schematic way, how the factors most germane to an analysis of leverage against CO₂
emissions can be portrayed in a single diagram: the length of time until a new technology is
ready to begin penetrating the market, the cost of the R&D effort needed to get to that point,
and the rate at which the technology could penetrate the market (reflected in the diagram as the
rate of increase in avoided CO₂ emissions) after that time. (With some modification such a
diagram could also show the effect, on the potential for emissions avoidance. of the different
sizes of the various energy-supply or end-use markets being penetrated.)
The Panel has not been able, in the time available for this study, to complete the sorts
of analyses that would be necessary to specify the relevant market-entry points, associated
research investments, and plausible penetration rates - and the uncertainty ranges associated
with all of these - with any confidence. Figure ES-2 is based on very approximate
understandings of needed research investments and market-entry points developed in the course
of our study, and on crude guesses about penetration rates (which were uniform across the
technologies shown, in the absence of the sort of analysis that would be required to do this in a
differentiated way). What can be said in favor of this very rough and preliminary depiction of
potential leverage is that (a) it illustrates what we believe DOE should be doing in the way of
portfolio analysis, with a much larger analytical effort behind it than they or we have mustered
until now, and (b) the timing and magnitudes of the conceivably achievable avoided carbon
emissions shown in the diagram are roughly consistent with what other major recent studies of
the potential of new technologies for this purpose have found.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 26
Our Figure ES-2 shows only technologies that would not begin penetrating markets
until after 2010. They offer large emissions-avoidance potential, but only well into the next
century. (Of course, the point of increasing R&D investments in appropriately targeted areas
is to move forward the date at which such technologies can begin penetrating their markets.)
Options that could have an impact by 2010 are not shown here but have been separately
examined by DOE in a recently released report; these earlier-impacting options necessarily
depend largely on R&D that has already been done.
Commercialization Considerations
To achieve the sorts of impacts illustrated schematically in Figures ES-1 and ES-2
would require more than R&D in many cases.. New technologies face the chicken-and-egg
problem of generally having high costs, and thus being limited to low market volumes, but
needing large market volumes to drive costs down. Making this transition is difficult given the
low costs of energy today and given that the public benefits of new energy technologies -
notably environmental quality and national security - - are generally not valued in the market.
Industry-led, public-private collaborations in demonstration and commercialization of new
energy technologies can be an appropriate way to address this difficulty in ways that ensure
that R&D programs are appropriately targeted and market relevant and that the benefits of the
public investment in R&D are realized in market penetration rates commensurate with the sum
of the private and public benefits of such penetration.
After consideration of the market circumstances and public benefits associated with the
energy-technology options for which we have recommended increased R&D, the Panel
recommends that the nation adopt a commercialization strategy in specific areas
complementing its public investments in R&D. This strategy should be designed to reduce the
prices of the targetted technologies to competitive levels, and it should be limited in cost and
duration. The Panel does not make a recommendation as to the source of funds for such an
initiative. We do believe, however, that such a commercialization effort should be designed to
be very efficient in allocating funds to drive prices down, minimally disruptive of
energy/financial systems, and temporary.
International Aspects
Markets for many new energy supply technologies will be very limited in the United
States for the next decade or two due to slow growth in demand and the availability of low
cost natural gas; most of the growth in world energy production and use and in carbon
emissions will take place in developing countries. For the United States to maintain scientific,
technological, and market leadership in these critical energy technologies, it will be essential
for public R&D and development and commercialization programs to broaden their scope to
directly address international energy issues, including both collaborative R&D and market
competition. This can provide us as well as our partners substantial economic and
environmental benefits.
The Panel recommends that the government and government/national-
laboratory/industry/university consortia should engage strongly in international energy
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 27
technology R&D and development and commercialization efforts to regain and/or maintain the
scientific, technical, and market leadership of the United States in energy technology. This
should include increased R&D - particularly in collaboration with developing countries,
temporary support for D&C activities where appropriate - and responses to foreign export
promotion activities where necessary.
DOE Management of Its Energy R&D Programs
The necessity of linking fundamental research with applied R&D and with development
and commercialization, the increasing complexity of R&D efforts, globalization of R&D and
technology markets, heightened global market competition, and other evolving factors in the
energy field have several important implications for energy R&D management. The
complexity and technical demands of R&D require increased industry/national-lab/university
peer review and technical oversight and direction of R&D programs. Linkages require
improved coordination.
Better communications can enable reduced administrative procedures and management
overheads, and can improve coordination by pushing these responsibilities down to the
operational level. Efficient use of resources requires careful establishment of R&D targets and
timelines, and ongoing measurement of progress. Although DOE has been making some
efforts in these areas and some programs are beginning to establish effective models that can
be applied more broadly, in general these factors need to be better addressed in DOE energy
R&D management.
To address these management issues, and above all to increase the efficiency with
which public dollars invested in energy R&D yield the results that the national interest
requires, the Panel offers the following specific recommendations:
(1) Overall responsibility for DOE energy R&D portfolio should be assigned to a single
person reporting directly to the Secretary of Energy; similarly, a single individual
should be given the responsibility and authority for coordination of cross-cutting
programs between the applied-technology programs, reporting to the single person
responsible for the overall R&D portfolio.
(2) Industry/national-laboratory/university technical oversight committees should work
with DOE to provide overall direction to energy R&D programs, with DOE facilitating
and administering the process;
(3) All R&D programs should undergo outside technical peer review every 1-2 years, but
interim internal process-oriented reviews should be reduced to a minimum.
(4) DOE Staff technical skills should be strengthened by training, targeted hiring, and by
systematically rotating external technical (and managerial) staff through DOE as senior
professionals with significant responsibilities for all aspects of program management.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 28
(5) Lead laboratories should be named and laboratories should be treated by DOE as
integrated entities, not as collections of projects independently controlled from DOE
headquarters.
(6) Industry/laboratory/university partnerships should conduct the energy R&D that is
funded by DOE, in most cases.
(7) The national laboratories should be encouraged to perform work for clients other than
DOE, inside and outside the government, as appropriate, and processes for doing this
should be streamlined.
(8) DOE staff procedures for energy technology programs should be reviewed in detail,
and staff levels adjusted accordingly.
CONCLUDING OBSERVATIONS AND ONE MORE RECOMMENDATION
Funding and managing the energy R&D needed to help address the energy challenges and
opportunities of the next century are tasks not for the Federal government alone but for all levels
of government, for industry, for universities, for the nonprofit sector, and for a wide variety of
kinds of partnerships among entities in these different categories. The Panel's charge was to
review Federal energy R&D, but we have been attentive to the ways in which the role of the
government relates to and interacts with the roles of the other sectors. Our recommendations
aim to focus the government's resources on R&D where high potential payoffs for society as a
whole justify bigger R&D investments than industry would be likely to make on the basis of its
expected private returns, and where modest government investments can effectively complement,
leverage, or catalyze work in the private sector.
The funding increases we are proposing for Federal energy R&D, in order to better match
the energy R&D portfolio of the public and private sectors combined to the energy-related
challenges and opportunities facing the nation, appear quite large when expressed as percentage
increases in some of the particular DOE programs that would be affected. But the increase in
annual spending - amounting altogether to an extra billion dollars in 2003, compared to that in
1997, for R&D on all the applied energy-technology programs together - is equal to less than a
fifth of one percent of the sum that U.S. firms and consumers spent on energy in 1996; and it
would only bring the Department of Energy's spending on applied energy-technology R&D back
to where it was in 1992, in real terms. The potential returns to society from this modest
investment are very large. They can be measured in energy costs lower than they would
otherwise be, oil imports smaller than they would otherwise be, air cleaner than it would
otherwise be, more diverse and more cost-effective options for reducing the risk of global climate
change than we would otherwise have, and much more.
If this is such a good case, why hasn't it been made and accepted before now? Actually
the case has been made often before, by energy experts and by studies like this one. It has not
been entirely heeded for a variety of reasons, most of them discussed above and many of them
perfectly understandable. But perhaps the most important reason that the government today is
not doing all that it should in energy R&D is that the public has been lulled into a sense of
complacency by a combination of low energy prices and little sense of the connection between
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 29
energy and the larger economic, environmental. and security issues that people do care very much
about. In a way the low priority given to energy matters is reflected even in the Department of
Energy itself, where energy is only a modest part of the Department's array of missions and there
is no official responsible for all of the Department's energy activities and those alone.
What we have here is therefore, in part. an education problem. There needs to be more
public discussion and a growing public understanding of why energy itself - - and therefore energy
R&D - is important to the well-being of our nation and the world. In this the scientific and
technological community has an obvious role to play, and we hope this report will be seen as a
positive contribution to that. But the Federal government, led by the President, also has an
important educational role to play, reflected in what is said and in what is done. As the last of the
recommendations in this report, which was commissioned by the President, we therefore offer the
following:
We believe the President should increase his efforts to communicate clearly to the public
the importance of energy and of energy R&D to the nation's future, and that he should
clearly designate the Secretary of Energy as the national leader and coordinator for
developing and carrying out a sensible national energy strategy, which of course includes
not only energy R&D but much else.
Table ES-3. Recommended DOE Applied Energy-Technology (ET) R&D Initiatives and Budget Authority (in millions of as-spent dollars)
PROGRAM
R&D Activities, Initiatives, and Budget Changes
FY
FY
FY
FY
FY
FY
FY
1997
1998
1999
2000
2001
2002
2003
Efficiency:
Building System Design and Operation: advanced sensors; smart controls; automated
24
33
38
48
60
72
84
Buildings
diagnostics; and whole-building optimization and design tools.
Building Equipment and Materials: advanced materials; advanced energy-efficient HVAC,
27
37
57
72
85
98
111
lighting, windows, appliances, office equipment, etc.; and insulation initiative.
Codes and Standards: for efficient appliances and buildings; technical assistance.
12
21
25
25
25
25
25
Crosscutting Activities: technology roadmapping and partnership development with industry
:
--
20
25
30
35
35
following the model of the DOE Industries of the Future program.
Other: management and planning, and other activities.
19
20
20
20
20
20
20
Subtotal
81
111
160
190
220
250
275
Efficiency:
Industries of the Future: advanced technologies for energy intensive industries-aluminum,
46
56
65
75
85
95
110
Industry
cement, chemicals, forest products, glass, metal casting, refining, steel, agriculture-and for
emerging energy-intensive industries following technology roadmaps.
Crosscutting Activities: advanced microturbines (40-200 kW), sensors, motor drive systems, and
38
38
70
80
90
95
100
materials; work with DOE/OUT on biomass Integrated Gasification Combined Cycle.
Technology Access: innovation grants; industrial assessments, "Climate Wise", and motors.
25
37
40
40
45
45
50
Other: management and planning, and other activities
7
8
10
10
10
10
10
Subtotal
116
139
185
205
230
245
270
Efficiency:
PNGV: better emissions controls for light diesels; hybrid vehicles; and system integration.
105
129
100
100
100
100
75
Transport
PNGV II: fuel cells, microturbines, advanced energy storage, system integration.
--
--
75
85
100
100
125
Advanced Heavy Vehicles: efficient diesels, diesel pollution reduction, and hybrids.
20
18
30
40
50
55
60
Advanced Materials: high-temperature/high-strength materials to reduce weight 25%.
33
31
35
40
40
40
45
Technology Deployment: clean cities program, alternative fuel vehicles, and other activities.
11
17
20
20
20
20
20
Other: management and planning, and other activities
7
9
10
10
10
10
10
Subtotal
176
204
270
295
320
325
335
Fossil Energy
Coal Power: end Low Emission Boiler System, phase out near-term clean coal activities,
86
84
79
90
87
88
82
accelerate R&D on advanced power systems.
Coal Fuels: end direct liquefaction and solid fuels and feedstocks R&D; develop science-based
16
16
9
12
15
16
16
hazardous air emissions program.
Gas Power: strengthen solid-oxide fuel-cell R&D and other advanced research.
97
78
92
92
83
74
70
Oil and Gas Production and Processing: maintain oil programs for marginal resources;
70
77
86
94
107
110
113
strengthen gas production and processing R&D; and increase advanced research.
Carbon Sequestration: strengthen science-based carbon sequestration program.
I
2
10
11
17
23
22
Methane Hydrates: develop science-based program with industry, Federal agencies, and the Navy
0
0
5
5
11
11
12
to understand the potential of methane hydrates worldwide
Hydrogen Manufacture/Infrastructure conduct R&D on hydrogen production from fossil fuels
0
0
I
2
6
6
7
Technology/Oil Price Elasticities: analyze technologies to reduce cost of oil shocks.
0
0
I
I
I
I
0
Developing-Country Technologies: conduct collaborative R&D with other countries.
0
0
I
2
6
6
6
Other: management and planning: environmental restoration; cooperative R&D, etc.
95
89
95
97
100
102
105
Subtotal
365
346
379
406
433
437
433
N
Fission
Operating Reactors: R&D to address problems that
ent continued operation of existing
4
25
10
10
)
10
reactors.
Nuclear Energy Research Initiative: competitively select among proposals by researchers from
0
0
50
70
85
100
103
universities, national laboratories, and industry that address issues including proliferation-
resistant reactors or fuel cycles, new reactor designs with higher efficiency, lower cost, and
improved safety; low-power reactors; and new techniques for on-site and surface storage and
for permanent disposal of nuclear waste
Education: university research reactors and other support
4
6
6
6
6
6
6
Other: advanced light water reactor and reactor concepts
34
15
0
0
0
0
0
Subtotal
42
46
66
86
101
116
119
Nuclear Fusion
Plasma Science: conduct research on fundamental plasma science; develop fusion science and
technology and plasma confinement innovations; and pursue fusion energy science and
technology as a partner in international efforts.
Subtotal
232
225
250
270
290
320
328
Renewable Energy
Biomass Fuels: strengthen feedstock development; advance enzymatic hydrolysis and other
28
38
58
76
94
97
99
conversion technologies in integrated power and fuel systems.
Biomass Power: develop biomass materials handling equipment; integrated gasification
28
38
63
86
89
91
93
combined cycles; biogasification-fucl cell systems; and small gasification-engine systems.
Geothermal: strengthen hydrothermal research; reactivate R&D on advanced resources; expand
30
30
42
49
50
51
52
advanced drilling R&D; and increase R&D on reservoir testing and modeling.
Hydrogen reorient near-term demonstrations and launch initiative with DOE Fossil Energy on
15
15
16
16
17
17
17
innovative hydrogen production from fossil fuels with sequestration.
Hydropower: develop "fish-friendly" turbines and low-head run-of-river turbines; analyze
I
I
4
8
11
11
12
coupling of hydropower to intermittent renewables.
Photovoltaics: accelerate basic PV science; develop laboratory scaleup to first-time
60
77
105
130
133
137
140
manufacturing; and support engineering science for large-volume, low-cost production.
Solar-Thermal: strengthen power tower and dish-stirling, esp. optical materials and solar
22
20
32
43
44
46
47
manufacturing initiative; launch initiative on advanced high-temperature receivers.
Wind: accelerate R&D on lightweight adaptive systems, direct-drive variable speed generators,
29
43
53
65
66
68
70
hybrid systems, system integration-including with storage; wind technology manufacturing
initiative; fundamental work on materials, and computational acrodynamics.
Systems and Storage: energy storage, esp. for system integration with intermittents.
32
46
51
54
55
57
58
Solar Buildings: R&D in efficient and passive whole building design and design tools; building
3
4
6
9
9
9
9
integrated PVs and thermal systems; and initiative on low cost solar water heaters and others.
International: applications-specific systems integration and development, and field studies;
I
7
11
13
13
14
14
collaborative R&D and training: technical assistance; technical/policy analysis.
Resource Assessment: integrated assessments across all resources; further development of
0
0
5
5
6
6
6
geographic information systems; and collaborative R&D with developing nations.
Analysis: systematic analyses of technologies. system integration, markets, and policies.
0
0
4
5
6
6
6
Other: management and planning: renewable energy production incentive, other.
21
26
25
26
27
26
29
Subtotal
270
345
475
585
620
636
652
SUBTOTAL
1282
1416
1785
2037
2214
2329
2412
"Activities should be done through various partnerships between industry, national laboratories, universities, and Federal/state agencies, as appropriate.
PCAST Energy R&D Panei Executive Summary
30 September 1997
page 32
25
PNGV
20
Demand
ght Trucks
Heavy Trucks
MMBbl/Day
15
Imports
10
Biomass liquids
Marginal Rsrce Oil R&D
5
Domestic Supply
0
1950
1960
1970
1980
1990
2000
2010
2020
2030
Figure ES-1. Potential Reduction of U.S. Oil Imports by Selected Advanced Technologies.
Historical data and baseline projections from Energy Information Administration (EIA). Vehicle efficiency improvements
assume R&D completed by 2004 and commercial production is underway by 2010. with straight-line penetration to 100%
of the market by 2030. Improvements entail roughly 60% reductions in fuel intensity for cars and light trucks. 40% for
heavy trucks. Contributions from R&D to exploit marginal domestic resources are based on DOE projections. Biomass
liquids estimate is based on an aggressive program to produce ethanol from cellulosic biomass. Many other technological
possibilities are not shown.
PCAST Energy R&D Panel Executive Summary
30 September 1997
page 33
Schematic Portrayal of R&D Portfolio Analysis of
Carbon-Reduction Potential
600
500
Carbon, MMTC/year
400
Wind+
Adv
Fusion
300
Fission
Storage
200
PNGV
Zem-net energy
100
Residences
AIGCC
0
0.1
W
A.F.
Billion
0.4
0.9
1.6
P
2.5
L
3.6
49
2000
2020
2040
2060
2080
2100
2120
Figure ES-2. Schematic Portrayal of R&D Portfolio Analysis of Carbon-Reduction Potential.
This drawing depicts an approximate range of times when a technology might be available for commercial use-where
the shaded wedges touch the time-axis; the potential carbon savings as the technology penetrates the market-depicted by
the shaded wedges indicating a range of penetration rates: and the approximate cost of the R&D to develop these
technologies to commercialization-depicted by the squares at the bottom of the drawing. which have areas proportional
to the discounted present value of the R&D costs. The width of the wedges and shading in the boxes depict uncertainty in
these estimates. Maximum slopes of penetration-rate wedges are based on 100% capture of the market for new units and
specified turnover times for old units: 15 years for cars. 40 years for electric power plants. 80 years for residential
buildings. For simplicity. carbon intensities of the various sectors are assumed to be frozen at 1995 levels. Funding
estimates are for applied technology development only: they do not include fundamental science research. Funding for
buildings includes commercial buildings. for which carbon savings are not shown. Large. long-term R&D programs
assume international collaborations. With refinement and more nuanced analysis behind it. such an approach to
illustrating the leverage of an R&D portfolio versus time and investment could be very informative.
Clinton Presidential Records
Digital Records Marker
This is not a presidential record. This is used as an administrative
marker by the William J. Clinton Presidential Library Staff.
This marker identifies the place of a tabbed divider. Given our
digitization capabilities, we are sometimes unable to adequately
scan such dividers. The title from the original document is
indicated below.
I
Divider Title:
SCENARIOS OF U.S. CARBON REDUCTIONS
Potential Impacts of Energy Technologies by 2010 and Beyond
Prepared by the
Interlaboratory Working Group on
Energy-Efficient and Low-Carbon Technologies
Oak Ridge National Laboratory*
Lawrence Berkeley National Laboratory*
Pacific Northwest National Laboratory
National Renewable Energy Laboratory
Argonne National Laboratory
Prepared for
Office of Energy Efficiency and Renewable Energy
U.S. Department of Energy
*Coordinating laboratories for this study
EXECUTIVE SUMMARY
This report presents the results of a study conducted by five U.S. Department of Energy national
laboratories that quantifies the potential for energy-efficient and low-carbon technologies to reduce
carbon emissions in the United States.¹ The study documents in detail how four key sectors of the economy
- buildings, transportation, industry, and electric utilities - could respond to directed programs and
policies to expand adoption of energy-efficiency and low-carbon technologies, an increase in the relative
price of carbon-based fuels by $25 or $50/tonne (e.g., as a result of a cap on domestic carbon emissions and a
market for carbon "permits"), and an aggressive program of targeted research and development. Current
projections suggest that a carbon emissions reduction of 390 million metric tons per year (MtC/year) is
required to stabilize U.S. emissions in 2010 at 1990 levels.
The study, which has been peer-reviewed by industry and academic experts, uses a technology-by-
technology assessment as well as an engineering-economic modeling approach. It draws upon a wide
variety of technology cost and performance information to assess potential impacts. Analysis of the
buildings, industry, and transportation sectors quantifies the impacts of end-use energy-efficiency
improvements on carbon emissions. The utility sector analysis estimates the impacts of those
improvements on utility carbon emissions, and quantifies additional emissions reductions through
conversion of a number of coal power plants to natural gas, dispatching of the utility grid with $25 and
$50/tonne carbon permit prices, the accelerated use of biomass cofiring and wind energy, and other low-
carbon electricity supply options. Finally, a number of other promising low-carbon technologies are
examined to determine their potential for reducing emissions in the end-use sectors, including advanced gas
turbines in industry, transportation biofuels, and fuel cells in buildings.
Three overarching conclusions emerge from the analysis of alternative carbon scenarios. First, a vigorous
national commitment to develop and deploy energy-efficient and low-carbon technologies has the
potential to restrain the growth in U.S. energy consumption and carbon emissions such that levels in 2010
are close to those in 1997 (for energy) and 1990 (for carbon). We analyze a case in which energy efficiency
can reduce carbon emissions by 120 MtC/year by 2010. We analyze a second case, with policies that
promote adoption of energy-efficient and low carbon technologies and a $25/tonne carbon permit price,
with emission reductions of 230 MtC/year in 2010. Under a $50/tonne carbon permit price and aggresive
policies, 2010 emissions could be cut by about 390 MtC/year. The analysis also suggests that substantial
additional savings are available if permit prices were to begin to rise above the $50/tonne level.
The second conclusion is that, if feasible ways are found to implement the carbon reductions as described
above, all the cases (with reductions varying between 120 and 390 MtC/year by 2010) can produce energy
savings that are roughly equal to or exceed costs.² The analysis includes only technologies estimated to be
cost-effective under 2010 energy prices (with a $25/tonne and $50/tonne carbon permit price for the
respective cases); it has not, however, analyzed specific policies to achieve the cases, identified the
political feasibility of policies, or described a pathway to achieve the cases.
The third conclusion is that a next generation of energy-efficient and low-carbon technologies promises to
enable the continuation of an aggressive pace of carbon reductions over the next quarter century. This
report documents a wide array of advanced technology options that could be cost-competitive by the year
2020, assuming a vigorous and sustained program of energy R&D beginning now and extending beyond 2010.
1
The five national laboratories participating in the study were: Argonne National Laboratory (ANL), Lawrence Berkeley
National Laboratory (LBNL), National Renewable Energy Laboratory (NREL), Oak Ridge National Laboratory (ORNL),
and Pacific Northwest National Laboratory (PNNL). LBNL and ORNL were the co-leaders of the effort.
2 Here we count as benefits only the energy savings to the nation. We have not credited reduced CO₂ emissions or other
external benefits. Costs include the increased technology cost plus an approximate estimate of the costs of program and
policy implementation.
September 22, 1997
AUTHORSHIP
Marilyn A. Brown (Oak Ridge National Laboratory) and Mark D. Levine (Lawrence Berkeley
National Laboratory) were responsible for the overall leadership of the project. They jointly
authored the Executive Summary, Chapter 1 (Analysis Results), and Chapter 2 (Introduction and
Background).
Chapter 3 (Buildings) authorship is best described in terms of the analysis for 2010 (Sections 3.2 and
3.3 and associated appendices) and R&D potential in 2020 (Section 3.4). Jonathan Koomey (Lawrence
Berkeley National Laboratory) was the lead author for the 2010 analysis. Nathan Martin, Lynn
Price, and Mark Levine (LBNL) were co-authors. Marilyn Brown was the lead author for the R&D
section, with support from staff at Oak Ridge National Laboratory and Lawrence Berkeley
National Laboratory.
Gale Boyd (Argonne National Laboratory), Joseph M. Roop, and Madeline G. Woodruff (Pacific
Northwest National Laboratory) were the lead authors of Chapter 4 (Industry) with the exception
of Section 4.3 (Low-Carbon Technologies), which was prepared by Helena Chum and Ralph P.
Overend (National Renewable Energy Laboratory), Tony Schaffhauser and Marilyn Brown (Oak
Ridge National Laboratory), and Tina Kaarsberg (Vista Technologies).
David Greene (Oak Ridge National Laboratory) and Steve Plotkin (Argonne National Laboratory)
were responsible for Chapter 5 (Transportation).
Stanton Hadley and Eric Hirst (Oak Ridge National Laboratory) were the authors of Chapter 6
(The Electricity Sector's Response to End-Use Efficiency Changes).
Chapter 7, Electricity Supply Technologies, consists of a variety of topics: Conversion of Coal-
Based Power Plants to Natural Gas (Section 7.2) was prepared by David South and Jack Siegel
(Energy Resources, Inc.); Renewable Electricity Technologies (Section 7.3) was written by Eldon Boes
and Erik Ness with contributions from National Renewable Energy Laboratory staff; the section on
Advanced Coal Technologies (Section 7.6) was written by Stanton Hadley (Oak Ridge National
Laboratory); and the sections on Efficiency Improvements in Generation and T&D (Section 7.4) and
Nuclear Plant Life Extension (Section 7.5) were written by Marilyn Brown (Oak Ridge National
Laboratory). Other sections of this chapter are summaries of published materials.
ACKNOWLEDGMENTS
Funding for this report was provided by the U.S. Department of Energy's Office of Energy Efficiency
and Renewable Energy (EERE). Overall guidance and advice on the report was provided by Joe
Romm, Eric Petersen, and Art Rosenfeld. Other EERE staff members provided input and feedback on
individual chapters, including Bill Raup and and John Ryan (Office of Building Technologies, State
and Community Programs), Lou Divone and Jim Quinn (Office of Industrial Technologies), Phil
Patterson (Office of Transportation Technologies), and Joe Galdo (Office of Utility Technologies).
Staff at Lawrence Berkeley National Laboratory (LBNL) and Oak Ridge National Laboratory
(ORNL) put enormous efforts into producing this report, especially as contributions came from many
different places. We gratefully acknowledge Barbara Maximovich, Leslie Shown, Sam Webster,
and Nathan Martin (LBNL) and Tonia Edwards and Kathi Vaughan (ORNL) for word processing
(BM and TE), copy editing (LS and NM), and cleanup of figures (KV and SW).
The completion of this study was guided by a committee of experts from industry, universities, and
utility research organizations. The committee was chaired by Bill Fulkerson (University of
Tennessee) and included: Morton H. Blatt (Electric Power Research Institute), Daniel E. Steinmeyer
(Monsanto Chemical Company), Robert A. Frosch (Kennedy School, Harvard University), Douglas
C. Bauer (National Academy of Sciences), Hillard G. Huntington (Energy Modeling Forum, Stanford
University) and Thomas Roose (Gas Research Institute).
This report benefited from the contributions and assistance of numerous experts on energy efficiency
and electricity production. The authors would like to acknowledge contributors by chapter:
Chapter 1 (Summary): Jonathan Koomey.
Chapter 3 (Buildings): George Courville, Mike MacDonald, Jeff Muhs, John Tomlinson, Jim Van
Coevering, Robert Wendt (ORNL); Steve Selkowitz, Joe Huang and Steve Johnson (LBNL).
Chapter 4 (Industry): Jim Chang, Hann Huang, Zhuoxiong Mao, John Molburg, Ken Natesan, Leslie
Nieves, and Mike Petrick (ANL), Scott L. Freeman, Gary B. Josephson, and Mark J. Niefer (PNNL),
Wayne Hayden (ORNL), Keith Davidson and Bill Major (OnSite Energy, Inc.), and Nancy Margolis
(Energetics, Inc.).
Chapter 5 (Transportation): K. G. Duleep (Energy and Environmental Analysis, Inc.).
Chapter 7 (Electricity Supply Technologies): Helena Chum, David Kline and Ralph P. Overend
(NREL), Jack Siegel (Energy Resources International, Inc.), Claud Pugh and Mike Sale (ORNL).
Ronald Wolk contributed to Appendix G-1 and Ronald Fisher contributed to Appendix G-4.
Staff members of DOE's Energy Information Administration (ELA) participated in the planning
process for this report, provided advice and assistance with the modelling described in the report,
and offered insightful comments on previous drafts. Leading this group were Mary Hutzler, Andy
Kydes, and Barry Cohen. Sector-specific assistance and feedback was provided by ELA's Erin
Boedecker and John Cymbalsky (buildings), Crawford Honevcutt (industry); David Chien and Mark
Friedman (transportation); and Art Holland and Dave Schoeberlein (electricity).
ii
Non-participating DOE laboratories were asked to comment on various draft materials. This group
of reviewers included Jerome LaMontagne (Brookhaven National Laboratory) and Dan Arvizu
(Sandia National Laboratory). Additional valuable comments on earlier drafts of this report were
received from John Sheffield (ORNL), Jay Braitsch and Doug Carter (DOE's Office of Fossil
Energy), and several staff of the American Council for an Energy-Efficient Economy. Economists Al
Link (University of North Carolina at Greensboro) and Stephen DeCanio (University of California
at Santa Barbara) also provided review comments on the report.
By acknowledging the involvement of the above individuals and the extensive review process in
which they participated, we do not mean to imply their endorsement of the report. Final
responsibility for the content of this report lies solely with the authors.
iii
Table of Contents
Chapter 1
ANALYSIS RESULTS
1.1
1.1 OBJECTIVES OF THE REPORT
1.1
1.2 METHODOLOGY
1.2
1.3
BACKGROUND
1.3
1.4 RESULTS
1.3
1.4.1
Prospects for Improved Efficiencies by the Year 2010
1.3
1.4.2
R&D's Potential for Further Benefits by 2020
1.8
1.5 ASSESSMENT OF COSTS, ENERGY SAVINGS, AND SOURCES OF CARBON
REDUCTIONS
1.11
1.6 REFERENCES
1.15
Chapter 2
INTRODUCTION AND BACKGROUND
2.1
2.1
OBJECTIVES OF THE STUDY
2.1
2.2
METHODOLOGY
2.2
2.2.1
Overview
2.2
2.2.2
Time Frame
2.3
2.2.3
End-Use Efficiency Scenarios
2.3
2.2.4
Methodological Differences Across Sectors
2.6
2.2.5
What the Study Does Not Do
2.8
2.3
OVERVIEW OF THE REPORT
2.8
2.4
HISTORICAL ENERGY TRENDS
2.9
2.4.1
National Trends
2.9
2.4.2
Sectoral Trends
2.10
2.5
THE GOVERNMENT'S ROLE IN ENERGY R&D
2.13
2.5.1
Rationale for Government Support
2.13
2.5.2
Past R&D Successes
2.14
2.6
REFERENCES
2.15
Chapter 3
THE BUILDINGS SECTOR
3.1
3.1
INTRODUCTION
3.1
3.2
PROVEN AND NEAR-TERM TECHNOLOGIES
3.1
3.2.1
Generic Assumptions
3.1
3.2.2
Scenario Definitions
3.2
3.3
SCENARIOS For THE YEAR 2010
3.3
iv
: 3.1
Business-as-Usual Scenario
3.7
1.2
Maximum Cost-Effective Energy-Efficiency Potential
3.9
3
Efficiency Scenario Results
3.11
4
High-Efficiency/Low-Carbon Scenario Results
3.11
3.4
POTENTIAL FOR ADVANCED TECHNOLOGIES IN 2020
3.14
3.4.1
New Technologies and Practices
3.14
3.4.1.1
Advanced Construction Methods and Materials
3.15
3.4.1.2
Environmental Integration and Adaptive Envelopes
3.16
3.4.1.3
Multi-Functional Equipment and Integrated System Design
3.18
3.4.1.4
Advanced Lighting Systems
3.20
3.4.1.5
Controls, Communications, and Measurement
3.21
3.4.1.6
Self-Powered Buildings
3.22
3.4.2
Best Practice Buildings in the Year 2020
3.24
3.4.2.1
"Best Practice" Housing in 2020
3.24
3.4.2.2
"Best Practice" Commercial Buildings in 2020
3.26
3.5
Improvements to this analysis
3.28
3.6
SUMMARY AND CONCLUSIONS
3.29
3.7
REFERENCES
3.31
Chapter 4
THE INDUSTRIAL SECTOR
4.1
INTRODUCTION
4.1
1
Approach
4.1
1.1.1
Scenario Analysis
4.1
4.1.1.2
Technology Examples
4.2
4.1.1.3
A Continuing Stream of New Technologies
4.4
4.2
ENERGY EFFICIENCY EMISSIONS REDUCTIONS
4.4
4.2.1
Business-as-Usual Case
4.5
4.2.2
Efficiency and High-Efficiency/Low-Carbon Cases
4.6
4.2.3
Comparison with the NEMS model
4.11
4.2.4
The Historical Context of Energy Efficiency in Industry
4.11
4.2.5
The Costs of Achieving the Efficiency and HE/LC Cases
4.12
4.3
ADDITIONAL EMISSIONS REDUCTIONS FROM INDUSTRIAL LOW-CARBON
TECHNOLOGIES
4.14
4.3.1
Introduction and Summary
4.14
4.3.2
Power System Efficiency Maximization Technologies
4.16
4.3.2.1
Advanced Turbine Systems (ATS) for Industrial Applications
4.17
4.3.3
Fuel-Switching Technologies
4.21
4.3.3.1
Integrated Gasification Combined Cycle Technology for the Forest Products Industry 4.22
4.3.4
Low Process Carbon Technologies
4.25
4.3.4.1
Industrial Sources of Non-CO2 Greenhouse Gasses
4.25
4.3.4.2
Process CO2 Emissions
4.27
4.3.4.3
Low-Carbon Technologies in Primary Aluminum Production
4.28
4.3.4.4
Replacing Cement Clinker with Solid Wastes
4.30
4.4
PROVEN INDUSTRIAL TECHNOLOGIES
4.31
4.1
Cross-Cutting Technologies
4.33
1.1
Combined Heat and Power
4.33
V
4.4.1.2
Motor Systems
4.33
4.4.2
Pulp and Paper
4.34
4.4.3
Chemicals
4.35
4.4.4
Petroleum Refining
4.36
4.4.4.1
Monitoring Overall Energy Performance
4.36
4.4.4.2
Utility System Improvements
4.36
4.4.4.3
Process/Equipment Modifications
4.36
4.4.4.4
Fluid Catalytic Cracking
4.36
4.4.4.5
Fouling Mitigation in Heat Exchangers
4.37
4.4.5
Glass
4.37
4.4.5.1
Oxy-Fuel Process
4.37
4.4.5.2
Advanced Burner Technology
4.37
4.4.5.3
Glass Batch/Cullet Preheater Technology
4.37
4.4.6
Aluminum
4.38
4.4.6.1
Improving Hall-Heroult Cell Efficiency
4.38
4.4.6.2
Materials Recycling
4.38
4.4.6.3
Improve Furnace Efficiency
4.38
4.4.7
Iron and Steel
4.39
4.4.7.1
Direct Smelting / Direct Reduction
4.39
4.4.7.2
Scrap Preheating
4.39
4.4.7.3
Hot Connection
4.39
4.4.7.4
Near Net Shape Casting
4.40
4.4.8
Metal Casting
4.40
4.4.8.1
Computer-Aided Casting Design
4.40
4.4.8.2
Optimized Coreless Induction Melting
4.40
4.5
THE LONGER TERM
4.40
4.5.1
Pulp and Paper
4.41
4.5.1.1
Polyoxometalate Bleaching
4.41
4.5.2
Chemicals
4.41
4.5.2.1
Biological/Chemical Caprolactam Process
4.41
4.5.2.2
Flexible Chemical Processing of Polymeric Materials
4.42
4.5.2.3
Genetic Engineering
4.42
4.5.3
Petroleum Refining
4.42
4.5.3.1
Development of Improved Catalysts
4.42
4.5.4
Glass
4.43
4.5.4.1
Optimizing Electric Boost to Reduce Total Energy Consumption
4.43
4.5.4.2
Recovering and Reusing Waste Heat from Oxy-Fired Furnaces
4.43
4.5.5
Iron and Steel
4.44
4.5.6
Metal Casting
4.44
4.6
CONCLUSIONS
4.45
4.7
REFERENCES
4.48
Chapter 5
TRANSPORTATION SECTOR
5.1
5.1
INTRODUCTION
5.1
5.2
PROVEN AND ADVANCED TECHNOLOGIES
5.7
5.2.1
Material Substitution
5.7
5.2.2
Aerodynamic Drag Reduction
5.8
5.2.3
Improved Automatic Transmissions
5.8
vi
5.2.4
Engine Friction Reduction
5.9
5.2.5
Variable Valve Timing
5.9
5.2.6
Lean-Burn Engines
5.10
5.2.7
Advanced Tires
5.10
5.2.8
Advanced Drag Reduction
5.11
5.2.9
Hybrid-Electric Power Trains
5.11
5.2.10
Direct Injection Stratified Charge (DISC) Gasoline Engines
5.12
5.2.11
Turbocharged Direct Injection (TDI) Diesel Engines
5.13
5.2.12
Proton Exchange Membrane (PEM) Fuel Cell Powertrains
5.14
5.2.13
Fuel Cells in Heavy Trucks and Locomotives
5.15
5.2.14
Costs and Timing of Technology
5.16
5.2.15
Alternative Transportation Fuels
5.17
5.3 -SCENARIOS FOR 2010
5.20
5.3.1
The Business-as-Usual Scenario for Transportation
5.20
5.3.2
The Efficiency Scenario For Transportation
5.22
5.3.2.1
Changes to the Modal Models
5.22
5.3.2.2
New Technologies
5.24
5.3.2.3
Valuing Energy Savings
5.24
5.3.2.4
Trends in Vehicle Performance
5.25
5.3.2.5
NEMS New Light-Duty Vehicle Fuel Economy Estimates
5.25
5.3.2.6
Changes to the Heavy Truck Model
5.27
5.3.2.7
Changes to the Rail Model
5.28
5.3.2.8
Changes to the Air Model
5.29
5.3.2.9
Introduction of Cellulosic Ethanol
5.29
5.3.2.10
Adjustment of NEMS Gasoline Forecast
5.32
5.3.2.11
Adjustments for Increased Light-Duty Vehicle Diesel Use
5.33
5.3.3
The High-Efficiency/Low-Carbon Scenario for Transportation
5.33
5.3.3.1
Light-Duty Vehicles
5.34
5.3.3.2
Changes to the Medium and Heavy Truck Model
5.36
5.3.3.3
Changes to Other Modes
5.36
5.3.4
Comparison of Forecasts
5.36
5.3.5
Cost-Effectiveness of Light-Duty Vehicle Fuel Economy Improvement
5.44
5.3.6
Oil Imports and Oil Market Benefits
5.48
5.4
R&D POTENTIAL FOR ADVANCED TECHNOLOGIES IN 2020
5.49
5.4.1
Light-Duty Vehicles
5.49
5.4.2
Freight Trucks and Locomotives
5.51
5.5
SUMMARY
5.51
5.6
REFERENCES
5.54
Chapter 6
THE ELECTRICITY SECTOR'S RESPONSE TO END-USE EFFICIENCY CHANGES
6.1
6.1
INTRODUCTION
6.1
6.2
BACKGROUND
6.1
6.3
MODEL DESCRIPTION
6.2
6.4
SCENARIOS FOR 2010
6.6
6.4.1
Calibration to ELA AEO97
6.6
6.4.2
The Base Case for a Competitive Market
6.8
vii
6.4.3
Efficiency and High-Efficiency/Low-Carbor Cases
6.9
6.5
SUMMARY
6.16
6.6 REFERENCES
6.17
Chapter 7
ELECTRICITY SUPPLY TECHNOLOGIES
7.1
7.1 INTRODUCTION
7.1
7.2 REPOWERING COAL-BASED POWER PLANTS WITH NATURAL GAS
7.1
7.2.1
Repowering Approachs
7.2
7.2.2
Repowering Issues
7.2
7.2.2.1
Increase in Natural Gas Demand
7.3
7.2.2.2
Gas Deliverability
7.5
7.2.3
Emissions Reductions
7.6
7.2.4
Cost-Effectiveness
7.8
7.3 RENEWABLE ELECTRICITY TECHNOLOGIES
7.12
7.3.1
Renewable Electricity in 2010
7.14
7.3.1.1
Cofiring Coal with Biomass
7.14
7.3.1.2
Wind Power
7.17
7 1.3
Increasing Generation and Capacity at Existing Hydropower Plants
7.20
7. 1.4
Landfill Gas
7.21
7.3.1.5
Other Renewable Power Technologies
7.22
7.3.2
The Long-Term Role of Renewables
7.24
7.4 EFFICIENCY IMPROVEMENTS IN GENERATION AND TRANSMISSION &
DISTRIBUTION
7.28
7.5 NUCLEAR PLANT LIFE EXTENSION
7.29
7.6 ADVANCED COAL TECHNOLOGIES
7.31
7.7
SUMMARY
7.32
7.8 REFERENCES
7.33
viii
Tables
Chapter 1
Table 1.1
Primary Energy Use in Quads
1.4
Table 1.2
Carbon Emissions (MtC)
1.5
Table 1.3
Average Annual Energy and Carbon Growth Rates, 1997 to 2010, for Four Cases
1.8
Table 1.4
Potential Annual Reductions in Carbon Emissions in 2010, Compared to the Business-
As-Usual Forecast for 2010 (MtC)
1.12
Table 1.5
Estimated Costs and Energy Savings of the Efficiency and High-Efficiency/Low-
Carbon Scenarios
1.14
Chapter 2
Table 2.1
Conceptual and Operational Definitions of Scenarios for 2010
2.7
Table 2.2
Primary Energy Use in Quads
2.11
Table 2.3
Historical Energy Growth Rates
2.12
Table 2.4
Carbon Emissions from Fossil Energy Consumption
2.12
Table 2.5
Cumulative Net Savings and Carbon Reductions from Five Energy-Efficient
Technologies Developed with DOE Funding
2.15
Chapter 3
Table 3.1
Primary Energy Use in the Buildings Sector (quads)
3.4
Table 3.2
Carbon Emissions in the Buildings Sector (MtC)
3.4
Table 3.3
Annual Total Cost of Energy Services in the Buildings Sector (billions of 1995$)
3.5
Table 3.4
Cost-Effective Energy Savings Potentials for Selected End-Uses in the Residential
and Commercial Buildings Sector*
3.10
Chapter 4
Table 4.1
Industrial Energy Use
4.7
Table 4.2
LIEF Results
4.8
Table 4.3
LIEF Results
4.9
Table 4.4
Change in Industrial Energy Use by Fuel Type
4.9
Table 4.5
Carbon Emissions Estimates (MtC per year)
4.10
Table 4.6
Industry-Specific Reductions in Carbon Emissions (MtC per year in 2010)
4.10
Table 4.7
Comparison of Year 2010 Total Energy Savings Relative to BAU in the NEMS and
LIEF Models
4.11
Table 4.8
Cumulative Incremental Investment (1998-2010) for Energy Efficiency Implied by the
LIEF Model to Achieve the Forecast Energy Reductions (billions of 1995$)
4.12
Table 4.9
Net Costs of Private Investment for Energy Savings in the Efficiency and High-
Efficiency/Low-Carbon Cases (millions of 1995$)
4.13
ix
Table 4.10 Effect of Different Carbon Shadow Price Simulations on Electricity and Fossil Fuel
Reductions
4.14
Table 4.11
Examples of Additional Carbon Equivalent Reductions by 2010 Resulting From Low-
Carbon Technologies* (MtC equivalent)
4.16
Table 4.12 Calculation of 2010 ATS Carbon Savings (MtC) and Corresponding ATS Electricity
Generation (TWh)**
4.20
Table 4.13
Process Carbon Emissions and Energy Use by Sector
4.26
Table 4.14
Carbon Reductions from Advanced Aluminum Production Cells, in 2010 (MtC)
4.30
Table 4.15
Summary of Technology Examples
4.47
Chapter 5
Table 5.1
Comparison of Three Transportation Energy Scenarios to the AEO97 Reference Case 5.6
Table 5.2
New Light-Duty Vehicle Technologies Added to the Efficiency and High-
Efficiency/Low-Carbon Scenarios+
5.24
Table 5.3
Maximum Technological Fuel Economy Potential Versus NEMS New Car Average
Estimates
5.26
Table 5.4
Key Heavy Truck Fuel Economy Technologies for the Efficiency Scenario in 2010
5.28
Table 5.5
Greenhouse Gas Emissions Factors for Transportation Fuels
5.31
Table 5.6
Impact of Cellulosic Ethanol on Greenhouse Gas Emissions from Light-Duty Vehicles
in 2010
5.33
Table 5.7
Transportation Sector Projections to 2010 and 2015 Efficiency Scenario (cont. next
page)
5.38
Table 5.8
Transportation Sector Projections to 2010 and 2015 High-Efficiency/Low-Carbon
Scenario (cont. next page)
5.40
Table 5.9
Simple, Total Cost-Effectiveness Estimates for Light-Duty Vehicle Fuel Economy
Technology
5.47
Table 5.10
Transportation Energy Use by Fuel Type
5.52
Table 5.11
Carbon Emissions in 2010 (MtC)
5.53
Chapter 6
Table 6.1
Comparison of Year 2010 AEO97 and ORCED Estimates of U.S. Generating Capacity
and Generation
6.7
Table 6.2
Comparison of EIA and ORCED Estimates of Generation Costs (1995c/kWh)
6.8
Table 6.3
Comparison of Year 2010 Forecasts
6.9
Table 6.4
Comparison of Year 2010 Forecasts
6.11
Table 6.5
Comparison of Year 2010 Forecasts
6.12
Table 6.6
Carbon Reductions from Electricity Savings by Sector under the Efficiency and
High-Efficiency/Low-Carbon Cases (MtC)
6.14
Table 6.7
Allocation of Carbon Reductions from the Electricity Saved by the High-
Efficiency/Low-Carbon Case (MtC)
6.15
X
Analysis Results
Chapter 1
Chapter 1
ANALYSIS RESULTS
This report presents the results of a study conducted by five U.S. Department of Energy national
laboratories that quantifies the potential for energy-efficient and low-carbon technologies to reduce
carbon emissions in the United States.¹ The stimulus for this study derives from a growing
recognition that any national effort to reduce the growth of greenhouse gas emissions must consider
ways of increasing the productivity of energy use. To add greater definition to this view, we
quantify the reductions in carbon emissions that can be attained through the improved performance
and increased penetration of efficient and low-carbon technologies by the year 2010. We also take a
longer-term perspective by characterizing the potential for future research and development to
produce further carbon reductions over the next quarter century. As such, this report makes a strong
case for the value of energy technology research, development, demonstration, and diffusion as a
public response to global climate change.
Three overarching conclusions emerge from our analysis of alternative carbon reduction scenarios.
First, a vigorous national commitment to develop and deploy cost-effective energy-efficient and
low-carbon technologies could reverse the trend toward increasing carbon emissions. Along with
utility sector investments, such a commitment could halt the growth in U.S. energy consumption and
carbon emissions so that levels in 2010 are close to those in 1997 (for energy) and in 1990 (for carbon).
It must be noted that such a vigorous national commitment would have to go far beyond current
efforts. Second, if feasible ways are found to implement the carbon reductions, the cases analyzed in
the study are judged to yield energy savings that are roughly equal to or greater than costs. Third, a
next generation of energy-efficient and low-carbon technologies promises to enable the continuation
of an aggressive pace of carbon reductions over the next quarter century.
1.1 OBJECTIVES OF THE REPORT
The purposes of this study are threefold:
1. To provide a quantitative assessment of the reduction in energy consumption and carbon
emissions that could result by the year 2010 from a vigorous national commitment to accelerate
the development and deployment of cost-effective energy-efficient and low-carbon
technologies;
2. To document the costs and performance of the technologies that underpin a year 2010 scenario
in which substantial energy savings and carbon emissions reductions are achieved;
3. To illustrate the potential for energy-efficiency and renewable energy R&D to produce further
reductions in energy use and carbon emissions by the year 2020.
1.2 METHODOLOGY
To achieve these objectives, we started with the Annual Energy Outlook 1997 (AEO97) reference case
forecasts for the year 2010 (Energy Information Administration, 1996). After thoroughly reviewing
these forecasts on a sector-by-sector basis, and working with ELA staff, we chose to accept the EIA
"business-as-usual" (BAU) scenario as is for buildings and industry. We modified some of the
September 22, 1997
1.1
Chapter 1
Analysis Results
assumptions and data to produce a new BAU case - not greatly different from the ELA case for the
transportation and the electric utility sectors.²
We then assembled existing information on the performance and costs of technologies to increase
energy efficiency or, for selected end-uses, to switch from one fuel to another (e.g., from electricity to
natural gas for residential end-uses or from gasoline to biofuels for transportation). For the buildings
sector, the technology performance and cost data base are extensive. For transportation, the data
base although less fully developed than for buildings - is sufficient for our purposes. For industry,
only partial information on technologies and costs is presently available. As a result, the analysis
for industry relies primarily on historical relations between energy use and economic activity and
much less on explicit technological opportunities. The industrial analysis also includes some
examples of industrial low-carbon technologies. The analysis of low-carbon supply technologies in
the electricity sector is based on.a review of the literature including detailed technology
characterizations prepared by DOE in conjunction with its national laboratories and industry.
Next we created scenarios of increased energy efficiency and lower carbon emissions using the
technology data (or, in the industrial sector, historical relations) as key inputs. We chose to run
three scenarios other than the BAU case. We have termed the first the "efficiency" (EFF) case. It
assumes that the United States increases its emphasis on energy efficiency through enhanced public-
and private-sector efforts. The general philosophy of the efficiency case is that it reduces, but does
not eliminate, various market barriers and lags to the adoption of cost-effective energy efficiency
technology.³
The other two cases, dubbed the $25 permit and the $50 permit "high-efficiency/low-carbon"
(HE/LC) cases, describe a world in which, as a result of commitments made on a climate treaty or
other factors, the nation has embarked on a path to reduce carbon emissions. Both of these cases
assume a major effort to reduce carbon emissions through federal policies and programs (including
environmental regulatory reform), strengthened state programs, and very active private sector
involvement. Both also include a focused national R&D effort to develop and transform markets for
low-carbon energy options (e.g., fuel cells for microcogeneration in buildings and advanced turbine
systems for combined heat and power in industry). The difference between the two HE/LC cases is in
the assumption of a carbon permit price resulting from a domestic trading scheme for carbon emissions
with a cap on U.S. emissions (or from equivalent policy measures that increase the price of carbon-
based fuels relative to those with less carbon). We assume a domestic permit price of $25 and $50
per tonne of carbon for the two cases. Both of these HE/LC cases include a program of research,
development, demonstration and diffusion that is more vigorous than in the efficiency case. In the
buildings and industry sectors, the carbon price signal, combined with policies promoting energy
efficiency, is believed to trigger most of the additional carbon reductions. In the transportation
sector, it is the R&D-driven technology breakthroughs that generate the bulk of the carbon
reductions beyond the efficiency case. For the electricity sector, higher prices for carbon-based fuels
cause larger shifts from coal to natural gas; for this sector, these same higher relative prices
combined with federal and private research, development, and demonstration can bring advanced
low-carbon technologies to market.
Although most of the analysis focuses on 2010, we also look beyond this date. Here we describe new
technologies, materials, processes, manufacturing methods, and other R&D advances that promise
to offer significant energy benefits by the year 2020; for this time period, we make no effort to
forecast specific levels of market penetration, energy savings, or carbon reductions. Thus, instead of
creating scenarios we describe the technological innovations that could enable the continuation of an
aggressive pace of decarbonization well into the next quarter century, if appropriate investments in
R&D were made.
1.2
September 22, 1997
Analys "esults
Chapter 1
1.3 BACKGROUND
The are of gains in energy productivity achieved by the U.S. following the 1973-74 Arab oil
"mbary") represents a period of economic growth that was decoupled from increases in energy
consumerion, resulting in substantial economic benefits. Between 1973 and 1986, the nation's
consumpAion of primary energy froze at about 74 quads - while the GNP grew by 35%. Starting in
1986, " prices began a descent in real terms that has continued to the present. As a result,
"nergy "mand grew from 74 quads in 1986 to 91 quads in 1995, and carbon emissions have been
increasing at a similar pace.
Despite the growth in energy consumption since 1986, the U.S. economy today remains more energy
productive than it was 25 years ago. In 1970, 19.6 thousand Btu of energy were consumed for each
(1992) dollar of GDP. By 1995, the energy intensity of the economy had dropped to 13.4 thousand Btu
of energy per (1992) dollar of GDP. The U.S. Department of Energy (DOE) estimates that the
country 15 saving $150 to $200 billion annually as a result of these improvements.
Nevertheless, many cost-effective energy-efficient technologies remain underutilized, as discussed
in Chapter 2. A host of market barriers account for these lost opportunities. And declining energy
R&D expenditures may cause promising technology options to be foregone.
The rationale for government support of energy-efficiency R&D is strong. Much energy-efficiency
research is both long-term and high-risk and therefore is not adequately funded by the private
sector - despite the possibility of sizable gains in the long run. Furthermore, advances in energy
efficiency offer substantial public benefits (such as carbon reductions and improved national security
through greater oil independence) that cannot be fully captured in the private marketplace.
The benefits of past public investments in energy-efficiency R&D have been well documented.
Between 1978 and 1996, DOE spent approximately $8 billion on energy-efficiency research,
development and demonstration (RD&D). Just five of the technologies that were developed or
demonstrated with a fraction of this DOE support have resulted in net benefits of $28 billion
through 1996. Many other R&D successes have produced technologies yielding substantial energy
and cost Havings in the market. The DOE RD&D portfolio has also led to significant environmental,
health, productivity, and economic competitiveness benefits.
1.4 RESULTS
1.4.1 Prospects for Improved Efficiencies by the Year 2010
l'able 1.1 and Figure 1.1 compare the nation's primary energy use in quads for the years 1990 and 1997
(projected) with the results of three scenarios for 2010. (We have included only the high-
efficiency/low-carbon case at $50/tonne in the table and figure for simplicity.) The $50/tonne
HE/LC came shown below does not reflect the energy impacts of the selected low-carbon technologies
described later in this summary (e.g., stationary fuel cells for buildings, advanced turbine systems
and biomass gasification in industry) or the supply-side options shown in Table 1.4.
September 22, 1997
1.3
Chapter 1
Analysis Results
Table 1.1 Primary Energy Use in Quads: 1990-2010
2010
Business-as-
High-Efficiency/
1990
1997
Usual
Efficiency
Low-Carbon
Case
Case
Case ($50/tonne C)
Buildings
29.4
33.7
36.0
34.1
32.0
Industry
32.1
32.6
37.4
35.4
33.6
Transportation
22.6
25.5
32.3
29.2
27.8
Total
84.2
91.8
105.7
98.7
93.4
Source: Energy use estimates for 1990 come from ELA (1996a, Table 2.1, p. 39). Energy use estimates for 1997 come
from forecasts conducted for ELA (1996b). Numbers may not add to the totals due to rounding.
The major observations are as follows:
In the business-as-usual case, energy use increases by 22 quads (26%) between 1990 and 2010; 8
quads of this increase have occurred during the first seven years of this 20-year period. The
fastest growing sector during these initial seven years has been buildings (4.3 quads) followed
by transportation (2.9 quads) and industry (0.5 quads). In the BAU case, the fastest growing
sector during the remaining 13 years is transportation (6.8 quads). This is followed by industry
(4.8 quads) and then buildings (2.3 quads). The rapid projected growth in the energy consumed
for transportation is driven by estimates of increased per capita travel and minimal fuel
efficiency gains.
The efficiency scenario cuts the overall growth between 1990 and 2010 from 22 to 15 quads. This
is a 17% increase over the level of energy consumption in 1990, down from a 26% increase in the
BAU case. Relative to the BAU case, the efficiency scenario for transportation delivers
slightly more energy savings (3.1 quads) than do the same scenarios for the industrial (2.0) or
buildings (1.9) sectors. Compared with 1997 levels, the smallest increase in energy growth for
this case is in buildings (0.4 quads), followed by industry (2.8 quads), and transportation (3.7
quads).
The high-efficiency/low-carbon scenario with a $50/tonne carbon charge further decreases the
overall growth between 1990 and 2010, reducing it from 22 to 9 quads. This is an 11% increase
over the level of energy consumption in 1990. Relative to the BAU case, the high-
efficiency/low-carbon scenario for buildings, industry, and transportation delivers energy
savings ranging from 3.8 to 4.5 quads for each sector. Compared with 1997 levels, the buildings
sector is down about 2 quads and industry and transportation are up 1 and 2 quads, respectively.
Table 1.2 documents the impact of these projected energy savings in 2010 on carbon emissions in that
same year. It also presents the results of the HE/LC scenarios with both $25 and $50 per tonne
carbon charges. These scenarios show significant carbon reductions from the combination of greater
efficiency improvements and increased use of advanced low-carbon technologies. In these cases, a
number of low-carbon technologies have high rates of adoption (e.g., advanced turbine systems and
biomass gasification in industry), the utility grid is dispatched to reduce carbon emissions (by using
many coal plants for intermediate power and by running more natural gas plants as base load), a set
of coal-based power plants are repowered, nuclear plant lifetimes are extended, and key renewable
energy technologies are deployed. In all cases, these technologies and measures are estimated to be
cost-effective with a differential carbon fee of $50/tonne.
1.4
September 22, 1997
Analysis Results
Chapter 1
Figure 1.1 Primary Energy Use in Quads: 1990-2010
120
100
80
Buildings
Energy
60
(Quads/year)
Industry
40
20
Transportation
0
1973
1986
1990
1995
1997
Efficiency
Case
Business
High
as
Efficiency/
Usual
Low
Carbon
2010 Scenarios
Note: The high efficiency/low carbon scenario values represent the $50 per tonne carbon charge.
Table 1.2 Carbon Emissions (MtC): 1990-2010
2010
Business-as-
High-Efficiency/
Usual (BAU)
Efficiency Case
Low-Carbonb
1990
1997
Case
$25/tonne
$50/tonne
Buildings
460
511
571
546
527
509
Industry
452
482
548
520
494
455
Transportation
432
486
616
543
528
513
Utilitiesc
-
-
I
-
-48
-136
Total (rounded)
1340
1480
1730
1610
1500
1340
Change from 1990
140
390
270
160
0
Change from BAU
-
-
-
-120
-230
-390
a Two of these numbers differ from the AEO97 BAU case. The estimate for buildings (571 MtC) is slightly lower
than the AEO97 estimate (576 MtC) due to the use of different ratios for converting "other" fuels (i.e., liquid
propane gas, kerosene, and coal) to carbon. The estimate for transportation (616 MtC) is higher than the AEO97
estimate (598 MtC) due to the assumption that auto fuel economy does not increase.
This scenario includes the carbon emission reductions resulting from a carbon permit price of $25 or $50/tonne:
(1) dispatch of power plants in which natural gas is favored relative to coal, (2) repowering and partial
repowering of coal-based power plants to convert to natural gas, and (3) introduction of selected low-carbon
technologies to replace conventional ones, primarily in the industrial and utility sectors.
The entries in the last two columns are negative as they correspond to reductions in carbon emissions resulting
from the increased use of natural gas and low-carbon technology for electricity generation as a result of the
$50/tonne carbon permit price in this scenario.
September 22, 1997
1.5
Chapter 1
Analysis Results
Table 1.2 presents results for the business as usual and three efficiency and/or low carbon cases in
2010 as point estimates, because they are meant to be scenarios. When we use these scenarios for
analysis, in section 1.5, we describe sources of uncertainty and the effects of uncertainty on our
understanding of the implications of these cases. For now, we only describe the different cases.
Figures 1.2 and 1.3 complement the above table by illustrating the carbon emissions reductions from
each scenario. The major observations are:
In the BAU case, carbon emissions are forecast to increase by approximately 390 million tonnes
from 1990 levels.
The energy-efficiency gains incorporated in the efficiency case cut overall growth between 1990
and 2010 by one-third (from 390 to 270 million tonnes). This represents a carbon increase of 20%
above 1990 emissions.
The HE/LC scenario with $25/tonne carbon charge has the potential to reduce carbon emissions
by 230 million tonnes from the BAU case in 2010. The largest part of these carbon reductions are
from increased efficiency, but major changes in electricity supply (retirements of coal plants,
repowering, and carbon-based dispatching) contribute 34 million tonnes, and other low-carbon
technology, particularly renewables and advanced turbine systems, produce another 14 million
tonnes.
The HE/LC scenario with $50/tonne carbon charge has the potential to reduce carbon emissions
by approximately 390 million tonnes, thereby achieving 1990 carbon emission levels in 2010. Of
this 390 million tonne carbon reduction, 205 million tonnes are from increased energy efficiency,
135 million tonnes results from increases in the use of low-carbon fuels and technologies in the
utility sector, and 50 million tonnes results from the use of low-carbon technology in industry
and transportation.
Ninety-five million of the 135 million tonnes of carbon reductions in the utility sector comes
from retirement of coal power plants and carbon-ordered dispatching of the utility system
(including optimization of capacity expansion and unit commitment) and from repowering coal
plants with natural gas. These are cost-effective with a $50/tonne carbon charge. The
remaining 41 million tonnes are from renewables (wind, co-firing coal-based power plants with
biofuels, expansion of hydropower capacity), nuclear power plant life extensions, and power
plant efficiency improvements.
The 50 million tonnes of carbon reductions in industry and transportation from low-carbon
technologies are about equally divided among: (1) advanced combustion turbine cogenerators in
industry, (2) biomass and black liquor gasification and low-carbon industrial processes, and (3)
cellulosic ethanol/gasoline blends for automobiles.
Approximately 140 MtC of the increase in carbon emissions between 1990 and 2010 will have
occurred by the end of 1997; thus, it is useful to look at the 13-year forecast starting with 1997.
The carbon reductions incorporated in the efficiency case cut the overall growth in carbon
emissions between 1997 and 2010 from 250 million tonnes (as forecast in the BAU case) to 130.
The HE/LC scenario with $50/tonne carbon charge reduces carbon emissions in 2010 by an
additional 270 million tonnes.
1.6
September 22, 1997
Analysis Results
Chapter 1
Figure 1.2 Reductions in Carbon Emissions from Each Scenario
400
390
Other Low-Carbon Technologies
Electricity Supply Technologies
Energy-Efficient Technologies
Million Tonnes of Carbon Emissions Reduction
300
230
200
120
100
0
Efficiency
HE/LC Case
HE/LC Case
Case
25/tonne C
$50/tonne C
Figure 1.3 Reductions in Carbon Emissions from Each Type of Technology
400
HE/LC Case @ 50/tonne C
HE/LC Case @ 25/tonne C
Million Tonnes of Carbon Emissions Reduction
300
Efficiency Case
205
200
135
100
50
0
Energy-
Electricity
Other
Efficient
Supply
Low-Carbon
Technologies
Technologies
Technologies
Table 1.3 provides a comparison of the growth rate in energy and in carbon emissions for the four
cases, from 1997 to 2010. For the BAU and efficiency cases, the growth in carbon emissions is slightly
more rapid than the increase in energy demand. For the HE/LC case with a $50/tonne carbon
charge, carbon emissions decline while energy consumption rises. The carbon reduction reflects the
increased deployment of low-carbon fuels and technologies as a consequence of the relative increase
in price of carbon-based fuels precipitated by the $50/tonne incentive.
It is useful to compare the scenarios in this study to those of other studies. The 1991 report by the
Office of Technology Assessment (OTA) titled Changing by Degrees (U.S. Congress, 1991) analyzed
the potential for energy efficiency to reduce carbon emissions by the year 2015, starting with the
base year of 1987. Its "moderate" scenario results in a 15% rise in carbon emissions, from 1300
MtC/year of carbon in 1987 to 1500 MtC/year of carbon in 2015 (compared to a BAU forecast of 1900
MtC/year). Its "tough" scenario results in a 20% to 35% emissions reduction relative to 1987 levels,
September 22, 1997
1.7
Chapter 1
Analysis Results
or emissions levels of 850 to 1000 MtC/year of carbon in 2015. Our efficiency and HE/LC cases ranging
from 1.3 to 1.6 billion tonnes of carbon emissions in 2010 are comparable to OTA's "moderate" case and
show considerably higher emissions than OTA's "tough" case.
Table 1.3 Average Annual Energy and Carbon Growth Rates, 1997 to 2010, for Four Cases
High Efficiency/
High Efficiency/
Business-As-
Efficiency
Low Carbon Case
Low Carbon Case
Usual (BAU)
Case
($25/tonne)
($50/tonne)
Gross Domestic Product
(GDP)ᵃ
1.88%
1.88%
1.88%
1.88%
Energy Demand
1.09%
0.56%
0.34%
0.13%
Carbon Emissions
1.24%
0.65%
0.11%
-0.75%
Energy Consumption Per
-0.77%
-1.30%
-1.51%
-1.71%
GDP (E/GDP)
Carbon Emissions Per GDP
-0.63%
-1.20%
-1.73%
-2.58%
(C/GDP)b
a The Gross Domestic Product (GDP) in 1995 was $7251 billion in 1995 dollars. The 1.88% annual growth was
assumed to apply to the entire period, 1995-2010 to derive the results above.
b The carbon decrease per unit GDP growth for 1990 to 2010 is 0.7%, 1.1%, 1.4% and 1.9% per year for the
reference, efficiency, $25/tonne HE/LC, and $50/tonne HE/LC cases, respectively.
Another benchmark is provided by the 1992 National Academy of Sciences (NAS) report on Policy
Implications of Greenhouse Warming (National Academy of Sciences, 1992). This study identified a
set of energy conservation technologies that had either a positive economic return or that had a cost
of less than $2.50 per tonne of carbon. Altogether, NAS concluded that these technologies offer the
potential to reduce carbon emissions by 463 million tonnes, with more than half of these reductions
arising from cost-effective investments in building energy efficiency. Our efficiency and HE/LC
cases suggest the potential for reducing carbon emissions by between 120 and 380 million tonnes by the
year 2010. One reason that the NAS estimate is higher is because it is not limited to the 2010 time
frame, but rather characterizes the full potential for carbon reductions. Thus, it did not take into
account the replacement rates for equipment and processes, and other factors that prevent the
instantaneous, full market penetration of cost-effective energy-efficient and low-carbon
technologies.
1.4.2 R&D's Potential for Further Benefits by 2020
If carbon reductions in 2010 and beyond are to be sustained at reasonable cost, vigorous R&D efforts
are needed to fill the pipeline of next-generation energy technologies. It is difficult to estimate the
carbon savings that will accrue from these technologies; however, our effort to characterize their
features suggests that an aggressive pace of carbon reductions over the next quarter century can be
sustained, with a sufficient investment in R&D. Our analysis of R&D potential for the year 2020
focuses on opportunities for improved energy-efficiency and renewable energy technologies. The
potential long-term contributions of carbon sequestration, advanced coal technologies, and nuclear
Pt er may also be significant. However, the treatment of vigorous R&D initiatives to improve
these supply options after 2010 is beyond the scope of this report.
1.8
September 22, 1997
Analysis Results
Chapter 1
For an assessment of the broad range of R&D opportunities to reduce U.S. greenhouse gas emissions,
based on a 30-year planning horizon, the reader is referred to a report by 11 DOE national laboratory
directors (DOE National Laboratory Directors, 1997). That effort examines the potential of science
and technology-based developments in energy efficiency, clean energy, and carbon sequestration to
produce carbon reductions in each of the next three decades.
Renewable energy technologies will likely play a crucial role in limiting carbon emissions over the
long term. Low-carbon energy supply options are needed to fuel domestic and international economic
development without stimulating further global warming. Although renewable resources account for
only 7% of the nation's total energy consumption at present, many believe that they are at the
beginning of a long-term growth trajectory. With continuing technological development and cost
reductions, renewables could become preferred energy resources some time within the next several
decades. Early evidence of this transition is seen in the continuing adoption of renewable power
systems, including especially wind farms and biomass power systems, even in the face of low gas-
fired power generation costs and considerable uncertainty in today's electric energy sector.
With a vigorous and sustained program of research, development and deployment, biomass, wind,
photovoltaics, geothermal, and solar thermal technologies could deliver significant quantities of
electricity in 2020, thereby. substantially displacing carbon emissions. For example, the use of
forestry and agricultural residues in biomass power systems continues to be an attractive power
option where those residues exist. The successful development of higher-efficiency biomass
gasification systems would make this technology competitive in a wider range of applications,
including for power systems using dedicated feed stock supply systems. At the same time, biological
and agricultural research on biomass production will lead both to higher biomass yields and better
species for energy conversion purposes in the future.
A second area in which a vigorous and sustained R&D effort could spawn a range of key
improvements is in wind power systems. Potential improvements include:
Advanced blade shapes that increase wind power capture while reducing stress loads,
Elimination of gearboxes through development of direct-drive generators,
Variable speed turbines, and
Better resource prediction that will increase the value of wind power to power systems
operators.
A third area of renewables development that is at the beginning of a long-term growth path is the
use of renewables in buildings. Solar daylighting, passive solar designs, solar water heating, and
geothermal heat pumps already are cost-competitive in many applications, but are not yet widely
used. R&D advances could substantially accelerate their market penetration. In addition, building-
integrated photovoltaic products will benefit directly from advances in materials research. The
ultimate vision is that many buildings will become "net energy generators" through a combination of
renewable energy and energy-efficiency technologies.
In the next quarter century, improved energy-efficiency technologies will result from a combination
of incremental advances and fundamental breakthroughs. Incremental improvements in all sectors
can be achieved by the greater reliance on more precise and reliable sensors and controls or on lower-
cost sensors and controls, often integrated into industrial processes, transportation systems, and
buildings. Advanced manufacturing technologies, including rapid prototyping and ultraprecision
fabrication, also offer broad opportunities for continuous incremental improvements in energy
September 22, 1997
1.9
Chapter 1
Analysis Results
efficiency and renewable energy. Breakthroughs in bioprocessing, separations, superconductivity,
catalysts, and materials can have wide-ranging impacts on energy efficiency and carbon emissions by
the year 2020. Examples of specific technology opportunities are described in this report, by sector.
Six R&D areas offer great promise to reduce significantly the energy requirements of our nation's
buildings in 2020:
Advanced construction methods and materials,
Adaptive building envelopes,
Multi-functional equipment,
Integrated, advanced lighting systems,
Improved controls, communications and measurements, and
Self-powered buildings.
In addition to the broad application of better process modeling, sensors, and controls in industry,
many process/industry-specific opportunities for efficiency gains exist. These are described for each
of DOE's targeted industries of the future: pulp and paper, chemicals, petroleum refining, glass,
aluminum, iron and steel, and metal casting.
Many of the advanced technologies that have the potential to significantly improve the energy
efficiency of transportation need considerable R&D investment before they can become commercially
available in the year 2020. For example, to achieve fuel economies in the 60-80 miles per gallon
(MPG) range and remain affordable and safe, light-duty vehicles will need:
Breakthroughs in manufacturing processes for composite materials,
Large reduction in fuel cell costs and/or cost reductions and performance gains in batteries,
Utra-low rolling resistance tires,
High-efficiency accessories, and
Highly aerodynamic designs.
Opportunities for R&D to lead to improvements in the energy efficiency of other transportation
modes are also described in this report.
In all, the continued adoption of energy efficient and renewable energy technologies and a steady
flow of technology improvements from collaborative R&D programs with industry could make such
environmentally friendly technology an attractive option for domestic and global energy economies
in the future. With strong public-private partnerships to support the necessary R&D and market
transformation activities, ample cost-effective energy products and practices will be available in
2020.
1.10
September 22, 1997
Analysis Results
Chapter 1
1.5 ASSESSMENT OF COSTS, ENERGY SAVINGS, AND SOURCES OF CARBON
REDUCTIONS
The business-as-usual scenario projects an increase of 390 MtC/year between 1990 and 2010. In our
efficiency scenario, in which the nation actively pursues policies and programs to promote market
acceptance of energy efficiency while expanding commitments to research and development, energy-
efficient technologies reduce this growth in carbon emissions by 120 MtC/year. Under a carbon cap
and trading system, in which permits for carbon sell for either $25 or $50/tonne C, very substantial
carbon reductions appear possible. Detailed results for these cases, showing the sources of the carbon
reductions, are contained in Table 1.4. (Summaries of these results were presented in Figures 1.2 and
1.3.) Results indicate that, for the $50/tonne HE/LC case, there is a potential to roughly return to
1990 levels of carbon emissions in 2010. Almost two-thirds of the increase in carbon emissions is
eliminated in the case with a $25/tonne carbon charge (Table 1.4).
The estimates in Table 1.4 include ranges for most of the electricity supply options and the other
low-carbon technologies. There are no ranges for the efficiency technologies because the models used
to estimate their penetration are nonstochastic. When selecting a single estimate for the $50/tonne
case, numbers from the low end of the ranges were generally selected in order to be cautious. Because
we did not conduct an integrating analysis in which supply options compete against one another, we
felt it important to minimize potential overlap by entering the supply options in conservative
quantities. Also note that several renewable resources that could play a greater role by 2010 are
omitted from Table 1.4; these resources include include photovoltaics, geothermal, solar thermal,
and landfill gas.
One should not ascribe too much significance to specific entries in Table 1.4 There are many different
technologies, both on the supply and demand side of the energy system, that will compete to
achieve carbon reductions in an environment in which policies and economic signals favor such
reductions. Thus, for example, Table 4.1 shows advanced turbine systems in industry cutting carbon
emissions by 17 MtC/year in 2010, co-firing coal with biomass reducing emissions by the same
amount, and other low-carbon supply technologies (wind, nuclear plant extensions, hydropower
expansion, and power plant efficiency) contributing 24 MtC/year. The actual choice of technology
depends on how the economics of the different systems evolve over time, how the industry to supply
technology develops, the nature and speed of deregulation within the utility industry, and numerous
other factors that cannot be known today. As such, we do not intend the results in Table 1.4 to be
taken as a prediction of one technology over another to achieve carbon reductions. In this instance,
we have posited one of many possible mixes of supply technologies. These same comments apply to
the demand-side sectors and technologies.
In Table 1.5 we summarize the expected technology costs in 2010, as well as the cost of implementing
a carbon permit system. While these costs are necessarily uncertain, they are our best estimates and,
in our view, as likely to be high as to be low. We note, however, that we have focused our analysis
on technology costs, and have not assessed the viability of specific policies or programs to achieve
market acceptance. As described below, we do account for program and policy costs in an
approximate manner.
Appendix A-2 describes the calculations used to derive the direct costs and energy cost savings of the
cases. The costs considered include the incremental technology investment by consumers and
businesses, fuel price increases, and the estimated cost of federal, state, and local programs required
to achieve the carbon emission reductions. These constitute the direct costs of the scenarios. The
highest of these by far is the incremental investment costs. However, the generally higher first cost
September 22, 1997
1.11
Chapter 1
Analysis Results
of these technologies is counterbalanced by substantially lower operating costs. The benefits
considered are limited to the savings in operating (energy) costs from the technology investments.
Table 1.4 Potential Annual Reductions in Carbon Emissions in 2010, Compared to the Business-As-
Usual Forecast for 2010 (MtC)
High-Efficiency/Low-Carbon
Case
Efficiency
Case
$25/tonne
$50/tonne*
Buildings
Energy efficiency
25
42
59
Fuel cells
2
3
25
44
62
Industry
Energy efficiency
28
44
62
Advanced turbine systems
5
17 (14-24)
Biomass and black liquor gasification,
5
14 (13-16)
cement clinker replacement, and
aluminum technologies
28
54
93
Transportation
Energy efficiency
61
74
87
Ethanol
12
14
16
73
88
103
Utility Supply Options
Coal plant retirements and carbon-ordered
25
55
dispatching
Converting coal-based power plants to
9
40 (25-66)
natural gas
Co-firing coal with biomass
5
17 (16-24)
Wind
2
7 (6-20)
Extending the life of existing nuclear
3
5 (4-7)
plants
Hydropower expansions
2
4 (3-5)
Power plant efficiency
2
8 (7-13)
48
136
Total
126
234
394
"Numbers in parenthesis are ranges, as documented in the text of the report. See Appendix A-1 for a description of
the derivation of the results in this table.
We have presented the direct and most easily quantified of the costs and benefits, but have not
attempted a full benefit-cost calculation. We do not account for indirect effects of policies (e.g., the
reallocation of investment dollars to efficiency investments). We do not account for the increased
cost of some R&D programs that are needed to achieve the scenario results nor do we count the
benefit of reduced carbon and other pollutant emissions. Also, we have not analyzed any possible
redistribution of wealth that could arise from a carbon trading system or other policy to increase the
price of carbon-based fuel.
1.12
September 22, 1997
Analysis Results
Chapter 1
Considering only these direct costs and energy-saving benefits of the scenarios, we have analyzed
the economics of carbon emission reductions from two different perspectives in order to establish a
credible range of costs. In the first, which we label "optimistic," we evaluate direct costs and
energy-saving benefits with a real discount rate that approximates the cost of capital for efficiency
investments for the different end-use sectors: 7% for buildings, 10% for transportation, and 12.5% for
industry.
The lowest discount rate, for buildings, is based on the fact that the money for residential buildings
is derived from home mortgages or home improvement loans. The higher rate for industry reflects
the fact that energy-efficiency investments have to compete with investments for other projects.
These discount rates are not those that describe current market behavior, but rather are reflective of
costs of capital if the market did invest in the energy-efficiency measures. For the "optimistic"
case, we assume costs for efficiency measures brought about by utility, federal programs, and state
programs (e.g., demand-side management programs by utilities, federal market transformation
programs) to be 15% of technology costs. We also assume that at least half of the efficiency occurs as
a result of federal policies (e.g., standards or carbon permit charges) which add very low direct
program costs. Thus, the overall costs of implementation are taken to be about 7% in the "optimistic"
case. The electric supply-side technologies are assumed to add an incremental cost of $30/tonne
carbon in 2010, based on an average estimate of the incremental costs of the technologies from the
appropriate sections of this report.
These programs and policies are not specified in this study, but the broad nature of the actions could
include technology R&D partnerships such as the current Partnership for a Next Generation of
Vehicles and Industries of the Future; energy efficiency codes and standards; expanded partnerships,
technical assistance, and information programs to accelerate the adoption of energy-efficient
technologies; incentives through the tax system directed at investments in energy-efficient
technology in industry; and a variety of non-federal programs to accelerate market diffusion of
energy-efficient and low-carbon technologies.
The second perspective, which we label "pessimistic," assumes that there are hidden costs
associated with achieving widespread market acceptance of many of the efficiency and low-carbon
technologies, even after the imposition of a carbon charge and the implementation of major policies
and programs to promote a low-carbon future. In this perspective, we evaluate costs and benefits at a
real discount rate of 15% for buildings and 20% for transportation and industry. Program costs are
increased to 30% of the cost of efficiency measures, an estimate that is a high bound compared with
federal, state, and utility experience. Overall implementation costs (programs and directed
policies) are taken to be 15% of technology investments in this case. Other data and assumptions in
this case are the same as for the "optimistic" case.
The results of the economic analysis are presented in Table 1.5. Estimated direct costs are $26-$49
billion per year for the efficiency scenario and $51 to $88 billion per year for the high-
efficiency/low-carbon scenario. Estimated energy savings per year in 2010 are $42 to $51 billion per
vear in the efficiency case and $70-$88 billion per year for the high-efficiency/low-carbon case.
The costs, which are a small portion of annual gross private domestic investment of about $1.4
trillion in 2020, are likely to be more than balanced by savings in energy bills. Thus, net costs to the
U.S. economy are estimated to be near or below zero in this time frame.
The range of estimates in Table 1.5 reflects our attempt to "bound" optimistic and pessimistic
assessments. There are clearly other ways in which these bounds could be described, just as there are
many scenarios that could have been analyzed. We reflect a lower or pessimistic bound in three
wavs. First, we assume the investments in energy efficiency yield only 80% of the estimated energy
savings. Second, we value costs and benefits at discount rates noticeably higher than the likely cost
September 22, 1997
1.13
Chapter 1
Analysis Results
of capital. Third, we increase the estimated cost of programs and policies to twice that of typical
experience today. It is worth noting that if the implementation costs were taken to be much higher
than we believe to be reasonable - 50% of investments costs for programs and 25% overall - this
would add about $10 billion per year to the costs of the high-efficiency/low-carbon in the
pessimistic case.
Table 1.5 Estimated Costs and Energy Savings of the Efficiency and High-Efficiency/Low-Carbon
Scenarios Optimistic and Pessimistic View Estimates (billions of 1995$, annualized)
Efficiency
High-Efficiency/Low-Carbon
Caseᵃ
Caseᵇ
Direct
Energy
Direct
Energy
Costsd
Savingsᶜ
Carbonᶜ
Costs
Savings
Carbon
(billion
(billion
Savings
(billion
(billion
Savings
1995$)
1995$)
MtC
1995$)
1995$)
MtC
Energy Efficiency
Buildings
7-14
14-17
20-25
14-26
26-33
49-62
Industry
3-5
6-7
22-27
8-13
12-15
74-93
Transportation
16-30
22-27
58-73
23-43
32-40
82-103
Electricity Dispatch
0
0
0
2
0
44-55
Electricity Repowering
0
0
0
2
0
32-40
Other Low-Carbon Techologies
0
0
0
2
0
33-41
Total
26-49
42-51
100-125
51-88
70-88
314-394
a Energy efficiency category includes ethanol in transportation.
b Energy savings and carbon savings in the HE/LC case are relative to BAU case.
C In the "pessimistic" case, we have assumed that only 80% of the carbon savings are achieved, even though the
technology and implementation costs are unchanged. The range on carbon savings represents this assumption.
d Direct costs include the incremental technology investment cost and the cost of programs and policies required to
achieve the carbon emission reductions. Costs are calculated from differing viewpoints: the "optimistic" case uses
discount rates that vary between 7% and 12.5% for the different sectors, as described in the text. For the
"pessimistic" case, the discount rates used to annualize costs vary between 15% and 20%. Also in this case, the cost
of implementing programs (30%) and an overall package of programs and policies (15%) is taken to be twice that of
the "optimistic" case.
In addition to these costs, one needs to calculate the impact of the cases on natural gas demand. In
all of these cases, natural gas replaces very large quantities of coal. Higher natural gas demand
would result in higher natural gas prices, which in turn would increase the cost of substituting
natural gas for coal in power production, etc. As it turns out, our scenarios have somewhat reduced
gas demand compared with the BAU case (or with AEO97 baseline for 2010, on which the price of
natural gas in our work is based). Specifically, demand for natural gas in the HE/LC ($50/tonne)
case declines in 2010 by 2 quads compared with the business-as-usual case. This is the result of
declines of 0.5 quads for buildings, 1.0 quads for industry, and 0.5 quads for electricity. The latter
occurs because of the balance among three factors: increase in gas demand because of the large-scale
substitution of natural gas for coal, decrease of gas demand because of the use of many low-carbon
technologies that do not use natural gas (wind, nuclear power plant extensions, power plant
efficiency upgrades, hydropower expansion, co-firing with biofuels), and the large increase in
cogeneration, which reduces demand for natural gas for heating applications.
The sum of the second and third effects are somewhat greater than the first, and thus total natural
gas demand associated with electricity generation declines. This could reduce the cost of natural
gas, a benefit that we have not included in the analysis.
1.14
September 22, 1997
Analvsis Results
Chapter 1
The $50/tonne carbon charge, while not constituting a direct cost, does represent a potentially large
transfer payment. The magnitude of the transfer payment, as well as the losers and winners from
the transfers, depends on the nature of policy and its implementation as a cap and trade system or
some alternative. The amount of money that could be in play is very large: $50/tonne times 1.3
billion tonnes per year equals $65 billion per year.
In short, while there will surely be winners and losers for these energy-efficiency and low-carbon
scenarios, our analysis shows that their net economic costs - under a range of assumptions and
alternative methods of cost analysis - will be near or below zero.
The achievability of the cases depends on many factors. In all cases, carbon reductions require the
nation to embark on an aggressive set of policies and programs. Such efforts could occur in response to
an international agreement on climate change or to other events that result in a national
determination to reduce the growth of carbon emissions. In the high-efficiency/low-carbon. cases, we
assume a vigorous national program of research, development, demonstration, and diffusion, and a
trading regime for carbon with a domestic permit price of either $25/tonne or $50/tonne carbon.
Without some scheme that provides strong incentives for switching from coal to natural gas, and for
deploying other low-carbon technologies, much of the potential for carbon reductions will not be
realized.
Government policies and programs that encourage and/or require the adoption of energy-efficiency
and low-carbon technologies will be needed, along with incentives for industry to invest more in
these technologies. Additional private and public investments are necessary, not only to accelerate
the introduction of new technologies into the market before 2010 but also to ensure the availability
of technologies for the period after 2010. The transportation and utility sectors are especially
dependent on early technological advances to achieve the scenario results in 2010.
There is no assurance that these and other driving forces will cause the scenarios we have described
to take place. Our major conclusion is that technology can be deployed to achieve major reductions in
carbon emissions by 2010 at low or no net direct costs to the economy. Cost-effective energy efficiency
alone can take the nation 30 to 50% of the way to 1990 levels. Two additional utility sector measures
can reduce carbon emissions by another 30% at an estimated cost of $50/tonne carbon: carbon-based
dispatch and conversion of existing power plants from coal to natural gas.⁵ Finally, we identify
several additional technologies that can contribute up to 20% of the estimated carbon reductions,
also for less than $50/tonne. A next generation of advanced energy-efficiency and renewable energy
technologies promises to enable the continuation of an aggressive pace of energy and carbon
reductions over the next quarter century.
1.6 REFERENCES
DOE National Laboratory Directors. 1997. Technology Opportunities to Reduce U.S. Greenhouse
Gas Emissions, draft, September.
Energy Information Administration (ELA). 1996. Annual Energy Outlook 1997: With Projections to
2105, DOE/ELA-0383(97) (Washington, DC: U.S. Department of Energy), December.
National Academy of Sciences (NAS). 1992. Policy Implications of Greenhouse Warming:
Mitigation, Adaptation, and the Science Base (Washington, DC: National Academy Press).
Office of Technology Assessment (OTA). 1991. Changing by Degrees: Steps to Reduce Greenhouse
Gases, OTA-0-482 (Washington, DC: U.S. Government Printing Office) February.
September 22, 1997
1.15
Chapter 1
Analysis Results
ENDNOTES
1
The five national laboratories participating in the study were: Argonne National Laboratory
(ANL), Lawrence Berkeley National Laboratory (LBNL), National Renewable Energy Laboratory
(NREL), Oak Ridge National Laboratory (ORNL), and Pacific Northwest National Laboratory
(PNNL). LBNL and ORNL were the co-leaders of the effort.
2 The differences between the AEO97 BAU case and ours for 2010 are (1) 1.2 quads higher use of oil in
transportation (32.3 instead of 31.1 quads) because auto fuel economy does not increase and (2) lower
use of oil for electricity generation (declines from 1.5% of generation to 0.1%) and slightly higher use
of natural gas and coal. In all other regards, including price of all fuels and delivered energy, our
reference case and the AEO BAU case are essentially identical.
3 See Section 2.2.3 for a definition of cost-effective energy efficiency technology.
4
$50 per tonne of carbon corresponds to 12.5 cents per gallon of gasoline or 0.5 cents per kilowatt-hour
for electricity produced from natural gas at 53% efficiency (or 1.3 cents per kilowatt-hour for coal at
34% efficiency). $25 per tonne would cut these gasoline and electricity price increments in half.
5 The cost curve for repowering is relatively flat; as such, considerable additional reductions are
possible at a cost not too different from $50/tonne. The results are highly sensitive to the price
differential between coal and natural gas; at a lower (higher) price differential, a higher (lower)
permit price of carbon is needed.
1.16
September 22, 1997
Introduction & Background
Chapter 2
Chapter 2
INTRODUCTION AND BACKGROUND
This report presents the results of a multi-laboratory study aimed at quantifying the potential for
energy-efficient and low-carbon technologies to reduce carbon emissions in the United States. The
stimulus for this study derives from a growing recognition of the link between energy R&D and the
nation's ability to respond to international calls to reduce the growth of greenhouse gas emissions.
According to a recent report of the Intergovernmental Panel on Climate Change (IPCC), the earth's
surface temperature has increased about 0.2 degrees Celsius per decade since 1975. Further, the IPCC
report concluded that "the balance of evidence suggests that there is a discernible human influence
on global climate" as the result of activities that contribute to the production of greenhouse gases
(IPCC, 1996, P. 5). By preventing heat radiated from the sun-warmed earth from escaping into
space, the increased concentration of greenhouse gases in the atmosphere contributes to global
warming.
The major greenhouse gases are carbon dioxide (CO₂), methane (CH,), ozone (O₃), nitrous oxide
(N₂O), water vapor (H₂O), and a host of engineered chemicals such as chlorofluorocarbons (CFCs).
CO₂ accounts for a majority of recent increases in the heat-trapping capacity of the atmosphere,
with worldwide atmospheric concentrations of CO₂ increasing at about 0.5% annually.
Anthropogenic CO₂ has resulted in atmospheric CO₂ concentrations that exceed pre-industrial levels
by 30%. Of all the human activities that contribute to these increases, fossil fuel combustion is by
far the largest, accounting for almost 60% of the greenhouse warming resulting from anthropogenic
sources in recent years (NAS, 1992, Table 2.2, p. 8). Energy-efficient, renewable energy, and other
low-carbon technologies reduce CO₂ emissions by displacing the need for fossil fuel combustion;
hence, this report focuses primarily on this single greenhouse gas. Throughout the report, the
potential climate benefits of energy-efficient and low-carbon technologies are quantified in terms of
reductions in millions of metric tons of carbon (MtC) emitted.¹
Analysis by a number of key climate and energy modelers indicates that significant research and
development on greenhouse-friendly technologies is essential to achieving meaningful emission-
reduction targets at affordable costs. As a result, climate change is becoming a major impetus for
energy R&D programs and is likely to grow in importance in the future. By documenting the
emissions reductions that past energy-efficiency and renewable energy R&D can deliver by the year
2010, and by describing the potential for future research to reduce carbon emissions even farther, this
report is intended to inform a broad public about technology-based approaches to reduce greenhouse
gas emissions.
2.1 OBJECTIVES OF THE STUDY
The purposes of this study are threefold:
1. To provide a quantitative assessment of the reduction in energy consumption and carbon
emissions that could result by the year 2010 from a vigorous national commitment to accelerate
the development and deployment of cost-effective energy-efficient and low-carbon
technologies;
2. To document the costs and performance of the technologies that underpin a year 2010 scenario
in which substantial energy savings and carbon emissions reductions are achieved;
September 22, 1997
2.1
Chapter 2
Introduction & Background
3. To illustrate the potential for energy-efficiency and renewable energy R&D to lead to further
reductions in energy use and carbon emissions by the year 2020.
The report focuses on energy-efficiency and renewable energy R&D. The coverage of additional
selected low-carbon end-use and electricity supply options was based in large measure on their
perceived potential to contribute significantly to reducing carbon emissions by 2010.
2.2 METHODOLOGY
2.2.1 Overview
To achieve these objectives, we started with the Annual Energy Outlook 1997 (AEO97) reference case
forecasts for the year 2010 (Energy Information Administration, 1996). After thoroughly reviewing
these forecasts on a sector-by-sector basis, and working with ELA staff, we chose to accept the EIA
"business-as-usual" (BAU) scenario as is for buildings and industry and to modify some of the
assumptions and data and produce a new BAU case - not greatly different from the ELA case - for the
transportation and the electric utility sectors.
We then assembled existing information on the performance and costs of technologies to increase
energy efficiency or, for selected end-uses, to switch from one fuel to another (e.g., from electricity to
natural gas for residential end-uses or from gasoline to biofuels for transportation). For the buildings
sector, the technology performance and cost data base are extensive. For transportation, the data
base - although less fully developed than for buildings - is sufficient for our purposes. For industry,
only partial information on technologies and costs is presently available. As a result, the analysis
for industry relies primarily on historical relations between energy use and economic activity and
much less on explicit technological opportunities. The industrial analysis also includes some
examples of industrial low-carbon technologies. The analysis of low-carbon supply technologies in
the electricity sector is based on a review of the literature including detailed technology
characterizations prepared by DOE in conjunction with its national laboratories and industry.
Next we created scenarios of increased energy efficiency and lower-carbon emissions using the
technology data (or, in the industrial sector, historical relations) as a key input. We chose to run
three scenarios other than the BAU case. We have termed the first the "efficiency" case. It
assumes that the United States increases its emphasis on energy efficiency through enhanced public-
and private-sector efforts. The general philosophy of the efficiency case is that it reduces, but does
not eliminate, various market barriers and lags to the adoption of cost-effective energy-efficient
technology.
The other two cases, dubbed the $25 permit and the $50 permit "high-efficiency/low-carbon"
(HE/LC) cases, describe a world in which, as a result of commitments made on a climate treaty or
other factors, the nation has embarked on a path to reduce carbon emissions. Both of these cases
assume a major effort to reduce carbon emissions through federal policies and programs (including
environmental regulatory reform), strengthened state programs, and very active private sector
involvement. Both also include a focused national R&D effort to develop and transform markets for
low-carbon energy options (e.g., fuel cells for microcogeneration in buildings and advanced turbine
systems for combined heat and power in industry). The difference between the two HE/LC cases is in
the assumption of a carbon permit price resulting from a domestic trading scheme for carbon emissions
with a cap on U.S. emissions (or from equivalent policy measures that increase the price of carbon-
based fuels relative to those with less carbon). We assume a domestic permit price of $25 and $50
per tonne of carbon for the two cases. Both of these HE/LC cases include a program of research,
development, demonstration and diffusion that is more vigorous than in the efficiency case. In the
2.2
September 22, 1997
Introduction & Background
Chapter 2
buildings and industry sectors, the carbon price signal, combined with policies promoting energy
efficiency, is believed to trigger most of the additional carbon reductions. In the transportation
sector, it is the R&D-driven technology breakthroughs that generate the bulk of the carbon
reductions beyond the efficiency case. For the electricity sector, higher prices for carbon-based fuels
cause larger shifts from coal to natural gas; for this sector, these same higher relative prices
combined with federal and private research, development, and demonstration can bring advanced
low-carbon technologies to market.
Although the work focuses on 2010, we also look beyond this date. Here we describe new
technologies, materials, processes, manufacturing methods, and other R&D advances that promise
to offer significant energy benefits by the year 2020; for this time period, we make no effort to
forecast specific levels of market penetration, energy savings, or carbon reductions. Thus, instead of
creating scenarios we describe the technological innovations that could enable the continuation of an
aggressive pace of decarbonization well into the next quarter century, if appropriate investments in
R&D were made.
2.2.2 Time Frame
Analysis for all sectors focuses on two base years (1990 and 1997) against which future progress is
benchmarked, and a target year of 2010 for assessing emissions reduction potential. Energy use and
emissions for 1990 and 1997 are used to compare future energy consumption and carbon emissions. The
report examines a "snapshot" of energy use and carbon emissions, by sector, in 2010. The increased use
of energy-efficient technologies combined with the development of new technologies based on past
R&D plus an invigorated R&D effort initiated in 2000 are needed to achieve our 2010 scenarios.
Intermediate years between 1997 and 2010 are not examined.
We also highlight the likely post-2010 benefits of an intensified investment in energy R&D. This
captures the effects of technologies that may not be widely commercial for some years but that could
deliver cost-effective energy savings and emissions reductions, if public and privately supported
R&D were to accelerate their proof of concept and reduce their developmental risks.
2.2.3 End-Use Efficiency Scenarios
Each of the three end-use sector chapters is consistent in terms of overall approach, scope, and time
frame. They each analyze three scenarios for the year 2010: a business-as-usual case, an efficiency
case, and a high-efficiency/low-carbon case. (In the integration of this work, we later assess two
different HE/LC cases one with a $25/tonne carbon charge and the other with a $50/tonne carbon
charge.) The buildings sector also presents a "frozen efficiency" baseline, for additional comparison
purposes. While there is variation in the methodologies used to estimate the energy savings and
emission-reduction potential of each sector, the three sector chapters are similar in using a
combination of technology analysis and model-based forecasting. Specifically, the buildings and
transportation sectors use stock models with technology characteristics and other parameters taken
from assessments of individual technologies. The industrial sector forecasts conservation investment
behavior based on econometric modeling with industry-specific conservation supply curves as inputs.
All of the scenarios described in this report use the AEO97 forecasts of national economic output as
measured by gross domestic product (GDP), which is projected to increase by 1.9% per year through
2015. Similarly, the buildings sector uses the AEO97 forecast of annual growth in residential (1.1%)
and commercial (0.9%) floorspace; the industrial sector uses the AEO97 assumption of a 2.1% annual
growth rate for manufacturing production; and the transportation sector uses the AEO97 forecast of a
1.5% annual increase in vehicle miles traveled and a 3.7% annual increase in air travel.
September 22, 1997
2.3
Chapter 2
Introduction & Background
The scenarios for each sector also use the AEO97 energy price forecasts. World oil prices are assumed
to rise from $17 per barrel in 1995 to $20.4 per barrel (in 1995$) in 2010. In AEO97, natural gas prices
increase at annual rates of 1.4%, with larger increases in prices to the industrial, electricity, and
transportation sectors offsetting reductions in prices to residential and commercial consumers.
Between 1995 and 2010, the average price of electricity is projected to decline by 0.6% a year as a
result of competition among electricity suppliers. Electricity prices are forecast to decrease the most
for industrial customers and the least for residential customers.
Such macroeconomic and fuel price assumptions strongly influence the rate of penetration of energy-
efficient technologies in each sector. Further details regarding these assumptions can be found in
ELA (1996c).
Frozen Efficiency Baseline. This case, which is analyzed only for the buildings sector, assumes that
energy-consuming equipment and systems existing in the year 1997 remain at the same efficiency
until they are retired. This equipment and these systems retire over the 1997-2010 period at a rate
based on standard equipment lifetimes. It assumes that all new equipment employed after 1997
remains at the efficiency of new devices in the year 1997. The frozen efficiency baseline provides an
upper bound to likely energy demand (under the economic assumptions applied to all the cases),
because it ignores all forces leading to higher efficiency of new equipment in the business-as-usual
case. It also ignores any retrofits that might take place if there were economic reasons for early
retirement of equipment.
This case is presented primarily for heuristic reasons: it describes an easily-understood case in
which technology does not change. This is useful for exploring the impacts of technology change.
Also, the case is not necessarily divorced from reality: in the era of low energy prices preceding the
oil embargo of 1973-74, the energy efficiency of many household, transportation, and industrial
technologies changed very little.
Business-as-Usual Case. The business-as-usual (BAU) case represents the best estimate of future
energy use given current trends in service demand, stock turnover, and natural progress in the
efficiency of new equipment. It assumes that R&D and implementation programs at DOE and EPA
continue at more or less current:levels, without a significant influx of new funding. It captures likely
changes in efficiencies of new equipment over the analysis period. It also allows for some early
retirement of equipment where cost savings from new energy-efficient products are high relative to
purchase and installation costs, as in some industrial motor and drive systems and commercial
lighting retrofits.
To create this scenario, the buildings and industry sectors adopted the AEO97 reference case as their
BAU cases. For the transportation sector, we modified AEO97 somewhat. Specifically, the AEO97
reference case forecasts that the efficiency of passenger cars will increase from 27.5 MPG in 1997 to
31.5 MPG in 2010. We believe such improvements are unlikely in the absence of increases in real
gasoline prices and hence our BAU case for transportation leaves the MPG performance of light-duty
vehicles in 2010 unchanged from 1997 performance.
Efficiency Case. The efficiency case describes the potential for cost-effective, energy-efficient
technologies to penetrate the market by the year 2010, given an invigorated public- and private-
sector effort to promote energy efficiency through enhanced R&D and market transformation
activities. This case assumes that national policy, possibly in combination with exogenous events,
leads to an increase in the cost-effectiveness and deployment of energy-efficient technologies. Cost-
effectiveness is improved because R&D, in combination with increased deployment efforts, result in
declining capital costs. We do not specify the policies or exogenous events that could precipitate
2.4
September 22, 1997
Introduction & Background
Chapter 2
such changes. Instead, we examine the potential for technology-based energy and carbon reductions,
assuming that significant efforts are undertaken to enhance the attractiveness of these technologies.
To be attractive to manufacturers and consumers, a technology must be cost-effective. Thus, this
scenario limits itself to describing the potential for cost-effective technologies to reduce energy use
and carbon emissions. A technology is defined as "cost-effective" if it delivers a good or service at
equal or lower life-cycle costs relative to current practice.² Externalities are not internalized in this
definition of cost-effective. An energy-efficient technology may be societally cost-effective, for
instance by taking into account its air quality or safety benefits, but not be judged cost-effective by our
narrower economic criteria. This scenario reflects the view that "policy options exist that would
slow climate change without harming American living standards, and these measures may in fact
improve U.S. productivity in the longer run" (Arrow et al., 1997).
Compared to the business-as-usual case, the efficiency case assumes (1) better technology and (2)
higher penetration rates for energy-efficient and low-carbon technologies.
1. "Better technology" results from an invigorated public- and private-sector investment in R&D
such that energy-efficient technologies become more cost-competitive based on current fuel
prices. Performance improvements between 1997 and 2010 are mostly incremental in this
scenario, but by 2020 they could be revolutionary.
2. "Higher penetration rates" result from an invigorated set of policies and market
transformation programs that reduce market failures and allow markets to operate more
efficiently. Through improved information and risk reduction, capital markets for energy-
efficiency investments could be strengthened and consumer investment hurdle rates for the
purchase of high-efficiency equipment could be lowered.
Despite its assumption of an aggressive public commitment to energy efficiency, this scenario also
takes into account real-world experience and program implementation constraints which suggest
that it is not reasonable to assume that every consumer will purchase the least-cost, high-efficiency
technology option. There are many reasons to expect a shortfall from such a maximum case: capital
rationing, imperfect information, misplaced incentives, and the unevenness of supply, installation,
and maintenance networks (DOE, 1996b).
High-Efficiency/Low-Carbor Case. The high-efficiency/low-carbon (HE/LC) case assumes a
greater commitment to reducing carbon emissions through federal policies and programs,
strengthened state programs, and very active private sector involvement. One way to view this case
is to see it as an attempt to model a world where an international global warming treaty is
negotiated over the next few years and where the outcome for the United States (and other Annex I
nations) is to stabilize carbon and other greenhouse gas emissions in 2010 at 1990 levels. The United
States pursues those reductions by (1) aggressively instituting federal policies to develop and deploy
energy-efficiency and low-carbon technologies, such as increased funding for market transformation
and R&D efforts and (2) by issuing tradable emission permits.
In this rendition of the HE/LC case, policies are put into place by 2000 and progressively phased in
until they are fully in place by 2010. The permit price for carbon would presumably rise steadily
through 2010. Thus, we have multiple factors affecting consumer and business behavior, including
the following:
The recognition that policies to reduce carbon emissions will necessarily follow the signing of an
international agreement, including an anticipation of higher relative prices for carbon-based
fuels;
September 22, 1997
2.5
Chapter 2
Introduction & Background
The actual increases over time in the permit price of carbon (which we model as averaging
either $25 or $50 per tonne for much of this period);
Increased federal effort to accelerate R&D and diffusion of low-carbon technologies;
The development and introduction by other countries of advanced low-carbon technologies; and
The change in consumer preferences and behavior that would result from an international treaty
and national commitment to stabilize greenhouse gases, much like changes in consumer behavior
in the aftermath of the oil embargo of 1973-74.
In summary, this scenario for 2010 describes a combination of better technology, "readier" markets,
and a price of carbon that results in a significantly increased willingness to manufacture, purchase,
and use low-carbon technologies. It represents a vigorous national commitment that goes far beyond
current efforts.
2.2.4 Methodological Differences Across Sectors
The operational definitions used to model these scenarios for the individual end-use sectors reflect
the above conceptual definitions, but are nevertheless distinct (Table 2.1). These differences are due
partly to the modeling approaches used for each sector. They also reflect the authors' sense of what
could "drive" significant increases in energy efficiency in each sector. For instance, to achieve a
high-efficiency/low-carbon scenario, the transportation analysis postulates a set of technology
breakthroughs. The industrial analysis, on the other hand, achieves its high-efficiency/low-
carbon scenario by doubling market penetration rates and assuming that energy-efficiency decisions
are treated as strategic investments with correspondingly lower hurdle rates.
The sectors also differ in the way that life-cycle costs and benefits are calculated to determine the
cost-effectiveness of technologies in their efficiency scenarios.
The buildings sector employs a 7% real discount rate to value the stream of benefits accruing
from an investment. These benefits accumulate throughout the specific operational lifetimes
assumed for individual technologies. The efficiency case assumes market penetration of about
one-third of the technologies that are cost-effective at a 7% real discount but not adopted in
the business-as-usual case. The HE/LC case doubles this penetration.
The industrial sector assumes a capital recovery factor (CRF) of 15%, rather than 33% (which
is the BAU assumption). Thus, to be considered cost-effective in this sector, an investment must
pay back in no more than approximately seven years.
The transportation sector uses a 7% discount rate, but it is applied only to the first five years of
operation, even though the expected lifetime of a vehicle may be much longer. This five-year
period is meant to reflect the realities of purchase behavior in this sector, and results in
decisions that are based on considerably less than the full life-cycle of benefits.
2.6
September 22, 1997
Introduction & Background
Chapter 2
Table 2.1 Conceptual and Operational Definitions of Scenarios for 2010
Scenario/
Business-as-Usual
Efficiency
High-Efficiency/
Definition
(BAU)
(EFF)
Low-Carbon (HE/LC)
Conceptual
Best estimate of future
Potential for cost-
Optimistic but feasible
Definition
energy use given current
effective, energy-
potential for energy
trends in service demand,
efficient technologies to
efficiency and low-carbon
stock turnover, and
penetrate the market
technology based on a
natural progress in the
given an invigorated
greater commitment to
efficiency of new
effort to promote energy
reduce carbon emissions
equipment, including
efficiency through
resulting from actions that
advances supported by
enhanced public and
might include the creation
current public-sector
private-sector R&D and
of a market value for
programs; assumes no
market transformation
carbon of $25 and $50 per
changes in federal energy
activities.
tonne.
or environmental
policies.
Operational Definitions:
Buildings
AEO97 reference case
35% of the difference in
65% of the difference in
developed using the
total energy savings
total energy savings
NEMS model.ᵃ
between the BAU and
between the BAU and
cost-effective energy
cost-effective energy
b
savings potential.
savings potential.
Industry
AEO97 reference case;
The capital recovery
The CRF is lowered to
LIEF is calibrated to this
factor (CRF) for energy-
15% and the penetration
case and then is modified
efficiency investment
rates for energy-efficient
to produce the two
used in LIEF is lowered
technology used in the
efficiency scenarios.
from 33% to 15%. C
BAU are doubled.
Transportation
AEO97 reference case
Assumes earlier
Postulates breakthroughs
modified to hold new
introduction of advanced
in hybrid vehicle
light-duty vehicle fuel
fuel economy technology
technology, major
economy constant at
and adds certain key
aerodynamic and engine
current levels.
technologies that are not
efficiency gains for
in the BAU.
commercial aircraft, and
other technological
achievements.
a
NEMS = National Energy Modeling System developed by DOE's Energy Information Administration.
b
The cost-effective energy savings potential is defined as the difference between the energy demand that results
from using the most energy-efficient of the cost-effective technology currently available or forecasted to be
available by 2010, and the energy demand in 2010 assuming business-as-usual rates of technology change and use
in the economy.
c LIEF = Long-Term Industrial Energy Forecasting model developed by Argonne National Laboratory and
Lawrence Berkeley National Laboratory.
September 22, 1997
2.7
Chapter 2
Introduction & Background
2.2.5 What the Study Does Not Do
This report does not describe the policies that might be implemented to achieve higher penetrations
of energy-efficient and low-carbon technologies. (Reviews of a wide range of possible policy options
can be found in several recent publications, including OTA (1991), NAS (1992), and DOE (1996b)).
Rather, this report highlights the potential performance and impacts of technological
developments and transformed markets. The existence of cost-effective technologies is a
prerequisite for public policies to work. Without the technologies, policies to reduce greenhouse gas
emissions will be very costly. Indeed, this analysis suggests that carbon stabilization could produce
net benefits if the nation invests significantly in cost-effective energy-efficiency and low-carbon
technologies.
Thus, we believe it is critical to understand the availability of technologies, their performance, and
their costs for as many end-uses of energy as possible. Armed with this knowledge, discussion of
policies becomes much more meaningful. Without it, such discussion is less likely to lead to good
decisions. Thus, we choose to focus this report on the more narrow topic of technologies in the belief
that doing a credible job in this area will ultimately further the policy dialogue.
A second reason for focusing on technologies is our belief that insufficient attention has been given to
the role of R&D on energy-efficient and low-carbon technologies as a means to deal with climate
change and other environmental impacts. If effective energy technologies are not developed, then
the cost of reducing greenhouse gas emissions (and other environmental impacts of energy) will be
very high.
As in the AEO97 reference case, each of the scenarios is completed at the national level. Thus,
regional variations in population and economic activity are not considered, nor are regional
differences in fuel price, weather, or air quality and environmental conditions that might create
regional niche markets for particular technologies. As a result, our analyses have undoubtedly
overlooked the possible development of regional markets for advanced energy technologies. A
valuable next step would be to conduct analyses at a finer geographic scale to produce national
estimates that reflect such regional variations.
2.3 OVERVIEW OF THE REPORT
The rest of Chapter 2 sets the stage for the remainder of this report. It describes historical energy
and carbon trends, both at the national level and by sector, as a backdrop for assessing energy
consumption and carbon emission forecasts. It also discusses the government's role in energy R&D,
including the rationale for government support and some evidence of past energy-efficiency
technology successes that benefited from government sponsorship.
Chapters 3 through 5 address each of the major energy end-use sectors: buildings (Chapter 3),
industry (Chapter 4), and transportation (Chapter 5). Four tasks are completed for each sector:
1. Energy scenarios with and without a strong efficiency push, focusing on the year 2010, and
including comparisons with the AEO97 projections from the National Energy Modeling
System;
2. Documentation of the cost and performance assumptions for individual energy-efficient and
low-carbon technologies;
2.8
September 22, 1997
Introduction & Background
Chapter 2
3. Development of three scenarios (business-as-usual, efficiency, and high-efficiency/low-carbor
cases) for the year 2010 and an explanation of how the scenarios were developed; and
4. Descriptions of new technologies that could become available in the 2010 to 2020 time period,
as the result of R&D over the next two decades.
Each of these chapters is accompanied by appendices that provide detailed documentation of the
technology assumptions and the forecasting methodologies used. These are labeled Appendices C
(buildings), D (industry), and E (transportation).
Chapter 6 analyzes the electricity sector to forecast the effect of electricity and demand savings in
the year 2010 on CO₂ emissions from power plants. It also assesses the impact of a $50/tonne permit
price for carbon on the generation mix used by the electricity sector in 2010. The results of these
analyses are used in the buildings and industry sector chapters to convert electricity savings into
carbon reductions. Results from this chapter reveal the importance of fuel choice for new power
plants and fuel switching for existing power plants as determinants of carbon emissions in 2010.
Specifically, the cost and magnitude of fuel switching from coal to natural gas for power generation,
the possible early retirement of some coal-fired plants, and the upgrading/repowering of existing
plants were identified as key issues for Chapter 7.
The possible conversion of coal plants to natural gas combined cycle technologies is analyzed in
Chapter 7, as one of many electricity supply-side options for reducing carbon emissions by 2010.
Other options are addressed in Chapter 7, albeit more briefly, including renewable electricity
technologies, efficiency improvements in generation and T&D, advanced coal technologies, and
nuclear plant life extension. The chapter also characterizes the carbon reduction benefits that could
accrue by the year 2020 from a sustained renewable energy R&D effort.
2.4 HISTORICAL ENERGY TRENDS
2.4.1 National Trends
In studying historical trends in energy use and carbon emissions, we have chosen to highlight the
years 1973, the beginning of rising energy prices to the nation; 1986, the year in which energy prices
began a ten-year decline in real terms; 1990, the year generally used as a reference for carbon
emissions; and 1997, the first year of our forecast period.
Between 1973 and 1986, the nation's consumption of primary energy froze at about 74 quads while
the GNP grew by 35%.³ People purchased more fuel-efficient cars and appliances, insulated and
caulked their homes, and adjusted thermostats. Businesses retrofitted their buildings with more
efficient heating and cooling equipment and installed energy management and control systems.
Factories adopted more efficient manufacturing processes and purchased more efficient motors for
conveyors, pumps, fans, and compressors. These investments in more efficient technologies were
facilitated by higher energy prices and by federal and state policies that were enacted and
implemented to promote energy efficiency. About one-third of the freeze in energy use during this
period was the result of structural changes such as declines in energy-intensive industry and
increases in the service sector; two-thirds was due to increases in energy efficiency (DOE, 1995).
The gains in energy productivity achieved by the U.S. in the two decades following the 1973-74
Arab oil embargo represent one of the great economic success stories of this century. The extent to
which the U.S. economy improved its energy productivity can be quantified by examining the
relationship between total energy consumption and gross domestic product (GDP), as depicted in
September 22, 1997
2.9
Chapter 2
Introduction & Background
Figure 2.1. In Figure 2.1, primary energy use is measured in quads and energy consumption per dollar
of GDP is measured in thousands of Btus per 1992$. In 1970, 19.6 thousand Btu of energy were
consumed for each dollar of GDP (1992$). By 1995, the ergy intensity of the economy had dropped
to 13.4 thousand Btu of energy per dollar of GDP (1992$ ELA, 1996a, p. 17). DOE estimates that the
country is saving $150 to $200 billion annually as a result of these improvements.
Figure 2.1 Energy Consumption Per Dollar of Gross Domestic Product: 1973-1995
100
20
Primary Energy Use
Energy Consumption Per Dollar
of GDP
90
U.S. Primary Energy Use
(In Quads)
80
15
of GDP
Energy Consumption Per Dollar
70
60
10
1970
1975
1980
1985
1990
1995
Year
Starting in 1986, energy prices began their descent in real terms that has continued to the present. As
a result, energy demand grew from 74 quads in 1986 to 91 quads in 1995, and it continues to increase.
One of the major lessons of the period since 1973 is that the economy will and can respond to energy
price changes. In addition to prices, other factors are also important and can slow the decline in
conservation activity that otherwise would be expected with declining energy prices. Federal
policies, as well as federal, state, and utility programs and consumer preferences for energy-efficient
appliances, houses, and cars can increase the purchase and use of energy-efficient products.
Technological developments can improve the energy efficiency, reduce the carbon émissions, and
often improve the performance of the product. Demand for energy-efficient products and low-carbon
energy technologies is also strengthened by factors such as environmental concerns.
2.4.2 Sectoral Trends
Each end-use sector functions differently in the U.S. energy marketplace. One of the reasons for
these differences is the differing market structure for delivering new technologies and products in
each sector. Residential and commercial building technology is shaped by thousands of building
contractors and architectural and engineering firms, whereas transportation technology is in the
hands of a few manufacturers.
The principal causes of energy inefficiencies in manufacturing and transportation are not the same as
the causes of inefficiencies in homes and office buildings, although there are some similarities
(Hirst and Brown, 1990). For example, in the manufacturing sector, energy-efficiency investments
2.10
September 22, 1997
Introduction & Background
Chapter 2
are hindered by a preference for investments that increase output compared with investments that
reduce operating costs. The cost and relative difficulty of obtaining reliable information often
prevents energy-efficient features of buildings from being capitalized into real estate prices. This is
partly due to the lack of widely accepted building energy rating systems. These same information
gaps do not characterize the transportation sector, which has a well understood labeling system for
vehicles, in the form of miles per gallon. Misplaced incentives inhibit energy-efficient investments
in each of the sectors. Consumers often must use the energy technologies selected by others.
Specialists write product specifications for military purchases that limit access to alternatives.
Fleet managers select the vehicles to be used by others. And architects, engineers, and builders have
great control over the energy integrity of buildings, even though they do not pay the energy bills.
The involvement of intermediaries in the purchase of energy technologies limits the ultimate
consumer's role in decision making and leads to an emphasis on first cost rather than life-cycle cost
(DOE, 1996b).
The end-use sectors also differ in terms of their ability to respond to changing energy prices. The
transportation and residential sectors can respond relatively rapidly to price spikes, through
reduced driving and by adjusting thermostat settings, respectively.
The vast differences in the R&D capability of the various sectors also influence their ability to
respond quickly to changing energy prices and market signals. The private sector as a whole spends
more than $110 billion per year on industrial R&D, dwarfing the federal expenditure on non-defense
and non-space technology R&D (National Science Foundation, 1997). Of the private-sector R&D
expenditure, the automobile manufacturers stand out - Ford alone spends more than $8 billion per
year on R&D. Next comes the rest of the industrial sector. Here, manufacturers account for a
majority of the R&D expenditures. Finally, in the buildings sector, the construction industry has
virtually no indigenous R&D. The Council on Competitiveness in 1992 estimated that the
construction industry spends less than 0.2% of its sales on R&D, far less than other industries, which
average 3.5%.
Finally, each of the sectors is distinct in terms of their dynamics and primary societal benefits from
improved energy efficiency. Improving the efficiency of transportation is needed to improve air
quality and reduce dependence on imported oil. Improving the efficiency of the industrial sector
improves economic competitiveness and is often effective in preventing pollution. Opportunities for
energy-efficiency improvements are most widespread in the buildings sector because of market
barriers in the form of information that is difficult to obtain, energy consumers who do not make
purchase decisions on energy-using equipment, etc. Such differences make analysis by end-use sector
essential for understanding the U.S. energy, carbon, and innovation picture as a whole.
Table 2.2 presents the primary energy consumed annually by the buildings, industry, and
transportation sectors between 1973 and 1997. It shows significant sectoral differences in energy
consumption trends. For instance, during the 1973-86 period when the country's primary energy use
was steady at 74 quads, energy use in buildings and transportation increased by 2.7 quads and 2.2
quads respectively; industry experienced a compensating decline of 4.9 quads.
Over the entire period from 1973 to 1997, energy use increased in buildings from 24.1 to 33.7 quads
(40%); in industry, from 31.5 to 32.6 quads (3.5%); and in transportation, from 18.6 to 25.5 quads
(37%). As shown in Table 2.3, the growth in buildings and transportation has been relatively
steady, at less than 1% per year from 1973 to 1986, and between 1.3 and 2.9% per year from 1986 to
1997. Growth in energy demand in industry has been much more volatile during the period, showing
substantial declines during the period of rising prices (a negative 1.3% annual growth for the 13
years of increasing energy prices), an increase of 2.7% per year from 1986 to 1995, and a 2.9% per year
decline from 1995 to 1997.
September 22, 1997
2.11
Chapter 2
Introduction & Background
Table 2.2 Primary Energy Use in Quads: 1973-1997
1973
1986
1990
1995
1997
Buildings
24.1
26.9
29.4
32.1
33.7
Industry
31.5
26.6
32.1
34.5
32.6
Transportation
18.6
20.8
22.6
24.1
25.5
Total
74.3
74.3
84.2
90.6
91.8
Source: Energy use estimates for 1973-95 come from ELA (1996a, Table 1.1, p. 39). Energy use estimates for 1997
come from EIA (1996c).
Table 2.3 Historical Energy Growth Rates: 1973-1997
AAGR
AAGR
AAGR
AAGR
AAGR
1973-97
1973-86
1986-90
1990-95
1995-1997
Buildings
1.41%
0.85%
2.25%
1.77%
2.46%
Industry
0.14%
-1.31%
4.81%
1.45%
-2.87%
Transportation
1.32%
0.86%
2.10%
1.29%
2.86%
Total
0.89%
0.0%
3.18%
1.48%
0.66%
AAGR = Average Annual Growth Rate
The growth of carbon emissions during the period roughly follows that of energy demand growth.
Table 2.4 shows estimated carbon emissions from 1973 to 1997. Like energy, carbon emissions were
flat between 1973 and 1986. The increase in the fraction of coal in the final mix from 17.5% in 1973 to
23.2% in 1986 was offset by the increasing fraction of primary energy from nuclear power, from 0.1%
in 1973 to 6.0% in 1986. From 1986 to 1997, carbon emissions grew more slowly than energy
consumption. This was a result of an increase in the share of natural gas from 22.5% in 1987 to 25.4%
in 1997 and in electricity from nuclear power from 4.5% to 7.2%, combined with a small decrease in
coal (23.3% to 22.5%) and a larger decrease in petroleum (43.3% to 39.7%).
Table 2.4 Carbon Emissions from Fossil Energy Consumption: 1973 to 1997
1973
1986
1990
1995
1997
Carbon emissions from
energy in MtC
1260
1240
1344
1424
1480
1973-97
1973-86
1986-90
1990-95
1995-97
Average annual
growth rates (AAGR)
0.67
-0.12%
2.03%
1.16%
1.95%
for carbon emissions
Sources: Carbon emissions estimates for 1990 are from ELA (1996b, Table 6, P. 16), and for 1995 are from EIA
(1996b, Table A19, p. 120). Carbon emission estimates for 1973 and 1986 were derived using factors for carbon
emissions from combustion of oil, natural gas, and coal for 1990. For 1997, they are from the end-use sector
analyses described in Chapters 3 through 5 of this report.
2.12
September 22, 1997
Introduction & Background
Chapter 2
Although non-CO₂ industrial emissions of greenhouse gases are small by weight, they have global
warming potentials (GWPs) that range from 21 for methane to 23,900 for sulfur hexafloride (SF₆).
Carbon dioxide has a GWP of one by definition. Figure 2.2 shows the relative contribution of these
other gases in MtC equivalent. The largest non-CO2 greenhouse gas contribution is from methane
(CH4), which is responsible for 177.5 MtC equivalent and has a GWP of 21. Next is nitrous oxide
(N₂O), which is responsible for 39.1 MtC equivalent and has a GWP of 310. Finally, in 1994, various
halocarbons and other engineered chemicals amounted to 29.5 MtC equivalent. These engineered
chemicals are a source of concern since their emissions are growing rapidly - and the United States is
the major source. SF₆ alone is increasing at a rate of 0.5 MtC equivalent per year (ELA 1996b). Note
also that many of these emissions are seen not only in energy-intensive industries but also in "high-
tech" and service industries, as shown in Figure 2.2.
Figure 22 Non-CO₂ Greenhouse Gas Emissions by End-Use Sector and Industry
70
SF₆
60
HFCs
N₂O
50
CH₄
MtC equivalent
40
PFC
30
20
10
0
Chemicals
Steel
Other Mfgr*
Agriculture
Transp.
Aluminum
Forest
Metal
Service
Mining/
Buildings
Products
Casting
Energy Extraction
* Mainly semiconductors
Source: ELA (1996b)
2.5 THE GOVERNMENT'S ROLE IN ENERGY R&D
2.5.1 Rationale for Government Support
Most people agree that the federal government has a clear and important role in the funding of basic
research, and that it should not fund research that the private sector would conduct on its own.
Between these two extremes is a wide range of applied technology development and deployment
activities where the rationale for federal sponsorship is often unclear.
Economists have identified at least three situations in which the government's role in the R&D
process is justified. First is the situation where the potential aggregate benefits of the research are
large, but the uncertainties are simply too great for the private sector to shoulder the full research
costs. Second is the case where R&D activities will result in benefits that cannot be captured by
private entities. Although benefits might accrue to society at large, no single firm can realize
enough economic gain to justify the research costs. A recent Council of Economic Advisors report
September 22, 1997
2.13
Chapter 2
Introduction & Background
(CEA, 1995) estimated that the private returns from R&D are 20 to 30%, while social returns
(including energy and environmental benefits) are 50% or higher. This economic barrier limits the
extent to which the private sector can supplant a government role in maintaining nationally
beneficial R&D. The third situation occurs when the public sector is the primary consumer of the
results of the R&D. This is characteristic, for instance, of much defense and crime prevention
research.
Based on these three justifications, the rationale for government support of energy-efficiency and
low-carbon technology R&D is strong. Much of this research is both long-term and high-risk and
therefore cannot be afforded by private companies despite the possibility of substantial gains in the
long run Examples include high temperature superconductivity, fuel cell vehicles, and building
materials with switchable thermal and optical properties. Advances in energy research also offer
substantial public benefits that cannot be fully captured by private entities. Specifically, energy-
efficiency and low-carbon resources improve energy security by reducing the nation's reliance on
foreign sources of oil; they lead to reductions in waste streams; and they reduce greenhouse gas
emissions, which contribute to global warming. Finally, it is possible that governments will in the
future become the principal purchaser of greenhouse gas reductions as the result of future
international agreements. In this case, the third rationale for federal sponsorship of energy R&D
will also apply.
Industry's R&D priorities are shifting away from basic and applied research and toward near-term
product development and process enhancements. Business spending on applied research has dropped
to 15% of overall company R&D spending, while basic research has dropped to just 2%. In addition,
corporate investments in energy R&D, in particular, are down significantly (DOE, 1996a, P. 2).
Great potential exists for public-private R&D partnerships to produce scientific breakthroughs and
incremental technology enhancements that will produce new and improved products for the
marketplace. U.S. industry spends more than $100 billion per year on all types of R&D. The top 20
R&D performing companies all have R&D budgets exceeding $1 billion per year. These
expenditures dwarf the U.S. government's energy-related R&D appropriations. If climate
mitigation policies reoriented even a tiny fraction of this private-sector expenditure and capability,
it could have an enormous impact. One way to reorient private-sector R&D is through industry-
government R&D partnerships that involve joint technology roadmapping, collaborative priorities
for the development of advanced energy-efficient and low-carbon technologies, and cost-shared
R&D.
2.5.2 Past R&D Successes
Some indication of the cost-effectiveness of energy-efficiency R&D can be gleaned from the
experiences to date of DOE's Office of Energy Efficiency and Renewable Energy. From fiscal year
1978 through fiscal year 1994, DOE spent a total of about $8 billion on energy-efficiency R&D and
related deployment programs. Estimates of the benefits of several dozen projects supported by this
funding were published in DOE/SEAB (1995). In response to a detailed review of these estimates by
the General Accounting Office in 1995/96, DOE has revised and updated the estimated benefits
accruing from five technologies that were developed with DOE support. Altogether, these five
technologies alone have resulted in net benefits (i.e., the value of energy saved minus annualized
cost premiums for better equipment) of approximately $28 billion (1996$) and annual emissions
reductions of 16 MtC equivalent (Table 2.5).4
Thus, the value of the energy saved by these five technologies, alone, far exceeds the cost to the
taxpayers of DOE's entire energy-efficiency R&D budget over the past two decades. Additional
case studies and benefits are documented in Geller and McGaraghan (1996) and DOE/SEAB (1995).
2.14
September 22, 1997
Introduction & Background
Chapter 2
Table 2.5 Cumulative Net Savings and Carbon Reductions from Five Energy-Efficient Technologies
Developed with DOE Funding
Net Present Value
Annualized
of Savings Thru
Consumer Cost
Annual Carbon
1996
Savings in 1996
Reductions in 1996
Energy-Efficient Technology
(billions of 1996$)
(billions of 1996$)
(MtC equivalent)
Building Design Software
11.0
0.5
8
Refrigerator Compressor
6.0
0.7
3
Electronic Ballast
3.7
1.4
1
Flame Retention Head Oil Burner
5.0
0.5
3
Low-Emissivity Windows
3.0
0.3
1
Totals
28
3.4
16
Note: Savings for the refrigerator compressor and flame retention head oil burner are through 1996 only; the
remainder are savings from products in place by the end of 1996 and include estimated energy savings from the
product's years in operation beyond 1996.
In addition to funding the development of numerous energy-efficient technologies, including those
listed in Table 2.5, DOE has also developed and implemented energy-efficiency standards for
equipment and building shells. For example, building efficiency standards became possible as a
result of DOE's investment in "building design software" (the first line of Table 2.5). Because of a
potential problem with "double-counting", Table 2.5 includes only energy savings achieved beyond
the savings that resulted from the implementation of minimum energy-efficiency standards for
buildings.
Moreover, results recently reported by Elliott et al. (1997) indicate that the total benefits -
including both energy and non-energy savings - that accrue from so-called "energy-saving" projects in
industry are typically much greater than those from the energy savings alone. In fact, based on
numerous case studies, the authors conclude that the average total benefits received from these
"energy-saving" projects are close to two to four times the value of the energy savings alone. They
also noted that costs and benefits resulting from non-energy ramifications of energy-efficiency
projects are often not included in cost/benefit analysis of energy-efficiency projects.
Similarly, Romm and Ervin (1996) describe some of the public health benefits that have resulted
from advances in energy-efficient and renewable energy technologies, such as clean air and water.
Other collateral benefits include the productivity gains that have accompanied investments in
industrial efficiency improvements (Romm, 1994) and the growth in export markets for energy
technologies.
2.6 REFERENCES
Arrow, Jorgenson, Krugman, Nordhaus, Solow, et al. 1997.
Council of Economic Advisors (CEA). 1995. Supporting Research and Development to Promote
Economic Growth: The Federal Government's Role (Washington, DC: Council of Economic Advisors)
October.
September 22, 1997
2.15
Chapter 2
Introduction & Background
Elliott, R. N., S. Laitner, and M. Pye. 1997. "Considerations in the Estimation of Costs and Benefits
of Industrial Energy Efficiency Projects", presented at the Thirty-Second Annual Intersociety Energy
Conversion Engineering Congress, Honolulu, HI, July 27-August 1, Paper # 97-551.
Energy Information Administration (ELA). 1996a. Annual Energy Review, DOE/ELA-0384(95)
(Washington, DC: U.S. Department of Energy), July.
Energy Information Administration (ELA). 1996b. Emissions of Greenhouse Gases in the United States
1995. DOE/EIA-0573(95). U.S. Department of Energy, Washington, D.C., October.
Energy Information Administration (ELA). 1996c. Annual Energy Outlook 1997: With Projections to
2105, DOE/ELA-0383(97) (Washington, DC: U.S. Department of Energy), December.
Geller, H., and S. McGaraghan. 1996. Successful Government-Industry Partnership: The U.S.
Department of Energy's Role in Advancing Energy-Efficient Technologies. Washington, D.C.:
American Council for an Energy Efficient Economy.
Hirst, E. and M. A. Brown. 1990. "Closing the Efficiency Gap: Barriers to the Efficient Use of
Energy," Resources, Conservation and Recycling, 3: 267-281.
Intergovernmental Panel on Climate Change (IPCC). 1996. Climate Change 1995: The Science of
Climate Change (Cambridge, UK: Cambridge University Press), P. 5.
James, W. M. (The Procter and Gamble Company). 1997. Presentation at the AAAS S&T Policy
Symposium, Washington, D.C., April 25.
National Academy of Sciences (NAS). 1992. Policy Implications of Greenhouse Warming:
Mitigation, Adaptation, and the Science Base (Washington, DC: National Academy Press).
Office of Technology Assessment (OTA). 1991. Changing by Degrees: Steps to Reduce Greenhouse
Gases, OTA-0-482 (Washington, DC: U.S. Government Printing Office) February.
Romm, J.J. 1994. Lean and Clean Management (New York: Kodansha America Inc.).
Romm, J. J., and C. A. Ervin. 1996. "How Energy Policies Affect public Health," Public Health
Reports, 5: 390-399.
U.S. Congress, Office of Technology Assessment. 1991. Changing by Degrees: Steps to Reduce
Greenhouse Gases, OTA-0-482 (Washington, DC: U.S. Government Printing Office) February.
U.S. Department of Energy (DOE), Office of Policy. 1996a. Corporate R&D in Transition.
(Washington, DC: U.S. Department of Energy), March.
U.S. Department of Energy (DOE), Office of Policy and International Affairs. 1996b. Policies and
Measures for Reducing Energy Related Greenhouse Gas Emissions. DOE/PO-0047. U.S. Department
of Energy. Washington, D.C., July.
U.S. Department of Energy (DOE). 1995. Energy Conservation Trends, DOE/PO-0034 (Washington,
DC: U.S. Department of Energy, Office of Policy), April.
U.S. Department of Energy, Secretary of Energy Advisory Board (DOE/SEAB). 1995. Task Force on
Strategic Energy Research and Development, Annex 3. (Washington, DC: U.S. Department of
Energy), June.
2.16
September 22, 1997
Introduction & Background
Chapter 2
ENDNOTES
1 In this report, carbon dioxide is measured in carbon units, defined as the weight of the carbon
content of carbon dioxide. Carbon dioxide units at full molecular weight (typically, million tonnes of
carbon (MtC)) can be converted into carbon units by dividing by 44/12, or 3.67. This approach has
been adopted for two reasons: (1) carbon dioxide is most commonly measured in carbon units in the
scientific community, in part because it is argued that not all carbon from combustion is, in fact,
emitted in the form of carbon dioxide, and (2) carbon units are more convenient for comparisons with
data on fuel consumption and carbon sequestration (ELA, 1996b). Note that, in the U.S., a "ton"
(sometimes referred to as a "short ton") equals 2000 pounds; a metric ton, or "tonne," equals 1000
kilograms (approximately 2204 pounds).
2 We evaluate cost-effectiveness from several viewpoints, with real discounts between 7% and 20%.
Even with the high discount rates, the efficiency case is cost-effective.
3 Primary energy use is the chemical energy embodied in fossil fuels (coal, oil and natural gas) or
biomass, the potential energy of a water reservoir, the electromagnetic energy of solar radiation,
and the energy released in nuclear reactors. For the most part, primary energy is transformed into
electricity or fuels such as gasoline, jet fuel, heating oil or charcoal - these, in turn, are referred to as
secondary energy. The end-use sectors of the energy system provide energy services such as cooking,
illumination, comfortable indoor climate, refrigerated storage, transportation and consumer goods
using both primary and secondary energy (NAS, 1992, p. 3)
4 The net present value (NPV) of cost savings, cumulative through 1996, is calculated as follows:
end of service
NPV = Σ, = entry year 0.07(1996 t)
where: E, is the value in 1996$ of energy saved in year t
P, is the annualized cost premium (1996$) of the better product
0.07 is the 7% real interest rate recommended by the Office of Management and Budget
Note that, for future years (1996 - t<0), (Et - Pₜ) is discounted by 7% per year; for past years, (Eₜ- Pₜ) is
raised 7% per year.
September 22, 1997
2.17
The Buildings Sector
Chapter 3
Chapter 3
THE BUILDINGS SECTOR
3.1 INTRODUCTION
Energy is used in buildings to provide a variety of services such as lighting, space heating and
cooling, refrigeration, and electricity for electronics and other equipment. In the U.S., building
energy consumption accounts for nearly one-third of total primary energy consumption and related
greenhouse gas emissions. The cost of delivering all energy services in buildings (such as cold food,
lighted offices, and warm homes) will be over $220 billion in 1997.
Our analysis shows that substantial reductions in future greenhouse gas emissions can be realized
through the use of more energy-efficient technologies that save society money. In addition, these
technologies often supply other benefits beyond energy, carbon, and dollar savings, including the
following: (1) improved indoor environment, comfort, health, and safety, (2) reduced noise, (3)
improved process control, and (4) increased amenity or convenience (Mills and Rosenfeld 1994).
These indirect benefits, while difficult to quantify in economic terms, can be even more important
than the energy cost savings, particularly when they improve the comfort of homeowners or the
productivity of workers.
This chapter describes our detailed assessment of the achievable cost-effective potential for
reducing carbon dioxide emissions in 2010.' We calculate carbon, energy, and dollar savings
associated with adoption of more energy-efficient technologies. In addition, this chapter
qualitatively describes the role of research and development (R&D) in providing a stream of
advanced building technologies and practices after 2010 that will enable continued reduction in
energy use and greenhouse gas emissions.
All costs in this chapter are reported in 1995 U.S. dollars (1995$). Carbon dioxide emissions are
reported in terms of their carbon equivalent. To convert carbon dioxide units at full molecular weight
into carbon units, divide by 44/12 or 3.67. For further information on emissions data, see ELA (1995).
3.2 PROVEN AND NEAR-TERM TECHNOLOGIES
In developing scenarios of carbon dioxide emissions for the residential and commercial buildings
sectors, we drew from a wide range of information and models available on end-use energy demand,
consumption, efficiencies, and technologies (see Section 3.7 References). Using this information, we
developed a spreadsheet model that incorporates the work of existing models and analyses as
parameters while providing a transparent framework to display assumptions, calculations, and
results. This model, developed specifically for the project, is described in Appendix C-1.
3.2.1 Generic Assumptions
Our approach is based on a stock accounting framework of building and equipment types. For all
scenarios, base case growth in households and commercial floorspace tracks historical trends. This
results in a net total 2010 stock that is greater than 1997 levels by 15% and 12% in residential and
commercial buildings, respectively, taking account of new building construction and retirement of
existing stock. Retrofit or replacement of existing "shells" (walls, roofs, windows, doors) and
September 15, 1997
3.1
Chapter 3
The Buildings Sector
equipment is a function of their average lifetimes. We assume that, on average, residential and
commercial building shells last 100 and 50 years, respectively, and thus only a small portion of
buildings are replaced during the study period with a much larger fraction undergoing some shell
retrofit. In contrast, average equipment lifetimes range from one year (for lights) to 20 years (for
furnaces). All equipment with lifetimes significantly less than the forecast period (13 years), such
as residential lighting, will be replaced but only a portion of the equipment with lifetimes
comparable to or longer than the forecast period will be replaced. The combination of shell and
equipment turnovers results in four categories of buildings in our model: (1) old buildings with old
equipment; (2) old buildings with new equipment; (3) retrofit building shells with new equipment;
and (4) new buildings with new equipment.
After characterizing the building stock in 2010, we calculate energy intensities (end-use energy per
household or per unit floor area) for all end-uses for 1997 and, in our initial assessment, use the
factors from the Energy Information Administration's (ELA's) Annual Energy Outlook (AEO97) to
establish baseline values in 2010 (ELA, 1996). In general, average 2010 energy intensities are lower
than those in 1997, reflecting technology improvements that provide the same level of energy
service with less energy.
We multiply each equipment end-use energy in 1997 (e.g. water heating, cooling, lighting) in the
four building categories by applicable energy intensities to derive future energy use. If more services
per household or unit of commercial floorspace are required by consumers, or if the size of the overall
building stock (relative to 1997) increases, this will increase the energy required to provide energy
services. Thus, energy demand in 2010 is a product of the rates of change in energy service
requirements within the buildings and changes in the overall growth in the building stock.
To derive energy-efficiency scenarios, we use the cost of energy intensity improvements and
electricity and fuel prices in 2010 to assess cost-effective reductions in energy use. For the residential
buildings, the efficiency scenarios also account for fuel switching (the impact of switching from
electric to gas water heaters, clothes dryers, and ranges) and for the use of high-albedo roof
materials ("cool roofs") to reduce cooling requirements (see Appendix C-4). For the commercial
sector, we include the analysis of cool roofs but do not include fuel switching.
3.2.2 Scenario Definitions
The model was used to generate results for three scenarios: "business-as-usual" (BAU), "efficiency"
(EFF), and "high-efficiency/low-carbon" (HE/LC). The business-as-usual scenario was calibrated to
the National Energy Modeling System (NEMS) model outputs, so that it corresponds to the same
2010 baseline currently used in AEO97.
For both the efficiency and high-efficiency/low-carbon scenarios, we first calculate the 2010 energy
use assuming 100% implementation of maximum cost-effective efficiency improvements in new
building shells and equipment. This maximum efficiency potential was calculated as the difference
between the energy intensity of the most cost-effective energy-efficiency technologies currently
available, and the energy intensity of new equipment in 1997. The maximum cost-effective efficiency
improvements are based on detailed studies; measures were not included if they had a cost of
conserved energy greater than the average cost of purchased fuel or electricity.² For comparative
purposes, we have also analyzed a "frozen efficiency case" in which the efficiencies of all new
equipment and building shell measures are kept at 1997 levels of new products.
3.2
September 15, 1997
The Buildings Sector
Chapter 3
We then derive the efficiency scenario by assuming that 35% of the difference in total energy
savings between the business-as-usual case and the maximum cost-effective efficiency case is
achieved. For the high-efficiency/low-carbon scenario, we assume a 65% achievement rate.
Assessments of future policy impacts are inherently speculative. We chose these implementation
factors based on a review of program experience (Brown 1993, Brown 1994) and use of our judgment
regarding how energy service markets would respond to policies and programs associated with
aggressive commitments to reduce carbon emissions. We began with Brown's (1993) conclusion that
about half of the techno-economic potential could be captured given coordinated efforts on minimum
efficiency standards, utility programs, and information programs. Our choice of 35% and 65%
brackets this result. The lower number (efficiency case) matches Brown's most pessimistic
sensitivity case, while the higher number (high-efficiency/low-carbon case) corresponds to
aggressive implementation of non-price policies combined with the assumption of policies such as a
cap and trade system for carbon and other economic signals that would support these aggressive
efforts. Brown did not address price signals in his report, so the most optimistic scenario he considers
reaches about 60% of the maximum economic potential. We believe that the addition of these price
signals under an aggressive policy regime is consistent with our assumption of an achievable
efficiency level to 65%. Details of the scenario calculations are provided in Appendix C-2.
Emissions factors for fuel-fired end-uses are taken from EIA (1995), while electricity sector emissions
factors are calculated in the utility section of this report. Electricity carbon emissions factors in the
business-as-usual case are 163 gC/kWh of electricity at the meter. In the efficiency case, the
marginal generating plants are high-efficiency gas-fired combined cycle plants, which reduces the
carbon saved from each kWh to 95 gC/kWh. In the high-efficiency/low-carbor case, the carbon
saved per kWh (relative to the business-as-usual case) increases to 127 gC/kWh because of changes
in the electricity supply system brought about by the carbon permit price. (See Chapter 6, Tables 6.6
and 6.7, and accompanying discussion for an explanation of this factor.)
3.3 SCENARIOS FOR THE YEAR 2010
Three scenarios are presented for residential and commercial buildings carbon emissions in 2010:
business-as-usual, efficiency, and high-efficiency/low-carbon Tables 3.1 through 3.3 and Figure 3.1
provide the main results for the three scenarios.
On Figure 3.1, the x-axis shows the percent change in carbon emissions from 1990 levels. The y-axis
shows total cost of energy services in 2010, expressed on an annual basis. This cost includes the
annualized incremental cost of efficiency improvements beyond the business-as-usual case plus the
cost of electricity and fuel purchases.
September 15, 1997
3.3
Chapter 3
The Buildings Sector
Table 3.1 Primary Energy Use in the Buildings Sector (quads): 1990-2010
1990
1997
2010
Business-as-
High-
Usual
Efficiency
Efficiency/Low-
End-Use/Fuel
Case
Case*
Carbon Case*
Residential:
Electricity
10.2
11.9
13.0
12.0 (7.1%)
10.8 (16.9%)
Fossil
6.5
7.2
7.4
7.3 (1.4%)
7.2 (2.6%)
Subtotal
16.7
19.1
20.4
19.4 (5.0%)
18.0 (11.8%)
Commercial:
Electricity
9.4
10.6
11.4
10.7 (6.0%)
9.7 (14.9%)
Fossil Fuels
3.8
4.0
4.2
4.0 (4.7%)
3.9 (8.7%)
Subtotal
13.2
14.6
15.6
14.7 (5.6%)
13.5 (13.5%)
Sector Total:
Electricity
19.7
22.5
24.3
22.7 (6.6%)
20.6 (15.2%)
Fossil
10.2
11.2
11.7
11.4 (2.6%)
11.1 (4.8%)
Total
29.9
33.7
36.0
34.1 (5.3%)
31.7 (11.9%)
* Numbers in parentheses represent percent reductions from the business-as-usual (BAU) case.
Note: Table does not include effects of building-sector fuel cells. Numbers may not add to the totals due to rounding.
Table 3.2 Carbon Emissions in the Buildings Sector (MtC): 1990-2010
1990
1997
2010
Business-as-
High-
Usual
Efficiency
Efficiency/Low-
End-Use/Fuel
Case
Case*
Carbon Case*
Residential:
Electricity
162
183
213
202 (5.4%)
185 (13.5%)**
Fossil Fuels
91
102
106
104 (1.5%)
102 (2.9%)
Subtotal
253
285
319
306 (4.1%)
287 (10.0%)
Commercial:
Electricity
150
163
187
178 (4.7%)
165 (11.8%)*
Fossil Fuels
59
62
65
62 (4.5%)
59 (8.4%)
Subtotal
209
225
252
240 (4.7%)
225 (10.9%)
Sector Total:
Electricity
312
346
401
380 (5.1%)
350 (12.7%)**
Fossil Fuels
150
164
170
166 (2.7%)
162 (5.0%)
Total
462
511
571
546 (4.4%)
511 (10.5%)
* Numbers in parentheses represent percent reductions from the business-as-usual (BAU) case.
** A portion of the reduction in carbon emissions associated with the high-efficiency/low-carbon case is due to
changes in the electricity generation mix prompted by the charge of $50/tonne of carbon.
Note: Table does not include effects of building-sector fuel cells. Numbers may not add to the totals due to rounding.
3.4
September 15, 1997
The Buildings Sector
Chapter 3
Table 3.3 Annual Total Cost of Energy Services in the Buildings Sector (billions of 1995$): 1990-2010
1990
1997
2010
Business-as-
High-
Usual
Efficiency
Efficiency/Low-
Case
Case
Carbon Case
Annual Fuel
Cost
$226
$228
$251
$233
$218
Annual
Incremental Cost
of Efficiency
--
--
$0
$7
$13
Improvement
Annual Total
Cost of Energy
$226
$228
$251
$240
$231
Services
Note: All costs are expressed in 1995 dollars (1995$). The annual total cost of energy services equals the sum of
annual fuel cost and annualized incremental cost of efficiency improvement (i.e., the cost of purchasing and
operating higher-efficiency equipment in the efficiency and high-efficiency/low-carbon scenarios). Table does not
include effects of building-sector fuel cells.
September 15, 1997
3.5
Chapter 3
The Buildings Sector
Figure 3.1 Relationship Between Costs of Energy Services and Carbon Emissions in the U.S.
Buildings Sector in 2010
1997 1990
275
Frozen
Efficiency case
Total cost of energy services in 2010 (billions of 1995$/year)
250
BAU case
Efficiency case
High-Efficiency/
Low-Carbon case
225
200
50%
40%
30%
20%
10%
0%
-10%
-20%
-30%
-40%
Percent change in carbon emissions from 1990 level
1990 U.S. buildings sector C emissions = 462 MtC
1997 U.S. buildings sector C emissions = 511 MtC
3.6
September 15, 1997
fuldings Sector
Chapter 3
Business-as-Usual Scenario
ness-as-usual scenario provides an estimate of energy demand and carbon emissions in 2010 in
*** of any new efforts to promote the more rapid development, purchase, and use of high-
technologies in the residential and commercial buildings sectors. In this scenario, energy
grows by 20% from 1990 and 7% from 1997 levels (from 29.9 and 33.7 quads in 1990 and 1997,
to 36.0 quads in 2010). Carbon emissions in 2010 are 24% and 12% higher than in 1990
respectively (increasing from 462 MtC in 1990 and 511 MtC in 1997 to 571 MtC in 2010).
missions grow faster than primary energy use in the business-as-usual case, mainly
changes in the fuel mix used to produce electricity. Because there is no accelerated
improvement in the business-as-usual scenario, the total annual cost of energy services
billion) is only the annual energy cost paid by consumers during that year.³
sidential sector, energy use in the business-as-usual scenario grows from 16.7 quads in 1990
quads in 1997 to 20.4 quads in 2010, (a 22% and 7% increase over 1990 and 1997 levels,
ely). Carbon emissions are projected to grow from 253 MtC in 1990 and 285 MtC in 1997 to 319
the same time period (a 26% and 12% increase from 1990 and 1997, respectively). The
in emissions in this sector is due to moderate growth in the residential building and
"ment/appliance stock coupled with substantial growth in miscellaneous energy use. For
purposes, we divide these miscellaneous uses into three electricity categories
motors, and heating) and two non-electricity categories (natural gas and oil/other
products).
from the rise in miscellaneous electricity use grow nearly four times as fast as the
sector as a whole, resulting in the share of miscellaneous electricity use jumping from
"tal demand in 1997 to 29% in 2010. There exist important problems in the way that EIA
and calculates the size of the miscellaneous end-use which leads to uncertainties in the
alues. It would be possible with more research to allocate some of the miscellaneous energy
existing end-uses and to new ones; for example, electricity consumed by furnace fans should be
abies
as space heating. New end-uses for televisions and dishwashers might be appropriate. Even
energy is not correctly allocated among the end-uses, the estimates of the savings potential
significantly change. More research is needed to evaluate the amount of energy used for
asks as well as the technologies available to reduce energy use within the miscellaneous
rategory (for the most detailed recent assessments, see Sanchez (1997) and Koomey and
(1997)).
these increases in service demand, total residential energy demand will be tempered
improvements in key residential equipment efficiencies, mainly due to implementation of
efficiency standards between 1997 and 2010. In particular, energy intensities for gas and
water heaters, freezers, and refrigerators decrease by 34%, 29%, 18% and 15%, respectively,
the
period. Had these declines in intensities not occurred, energy use for these end-uses would
been 14% greater in 2010 than the current business-as-usual scenario results. Residential sector
lisp and carbon emissions in 1997 and 2010 are shown in Figure 3.2 below.
"mercial sector, there are even greater problems in the way that ELA defines and calculates
of the miscellaneous end-use than in the residential sector. Even given these accounting
"ttabities, our assessment of the opportunities for efficiency improvements is almost certainly
"mercial sector, energy use in the business-as-usual scenario is projected to grow by 18% from
"% from 1997 to 2010 (13.2 quads in 1990 and 14.6 quads in 1997 to 15.6 quads in 2010). Carbon
are projected to grow by 21% from 1990 and 12% from 1997 to 2010 (209 MtC in 1990 and 225
15,
1997
3.7
Chapter 3
The Buildings Sector
MtC in 1997 to 252 MtC in 2010). Miscellaneous electricity end-uses such as motors, electronics, and
small appliances are expected to increase from 9% of total commercial sector energy use in 1990 to
20% in 2010. This growth, which accounts for over 70% of the growth in carbon emissions in
commercial buildings, offsets nearly all carbon emission reductions from ene cy-efficiency
improvements in other end-uses. Miscellaneous energy use in the commercial sector is even less well
understood than in the residential sector. As mentioned above, more analysis and data collection are
needed to improve our understanding of this end-use category.
Although energy use from office equipment is expected to grow by 22% over the period, its share of
energy use in commercial buildings remains relatively small, growing to 6% in 2010. The greatest
increases in energy efficiency in the commercial sector come from continuing improvements in space
conditioning (due to improved equipment and controls) and water heating systems. Commercial
sector energy use and carbon emissions in 1997 and 2010 are shown in Figure 3.3 below.
Figure 3.2 Residential Sector Primary Energy Use and Carbon Emissions in 1997 and 2010 by End-Use
for the Business-As-Usual Scenario⁵
100
6.0
90
1997
2010
80
5.0
70
Energy Use (Quads)
4.0
60
50
3.0
40
2.0
30
Emissions (MIC) (1997 generation mlx)
20
1.0
10
0.0
0
Elec. Misc. Uses
Gas Spc. heating
Elec. Spc. cooling
Elec. Spc. heating
Gas Wat. heating
Refrigeration
Wat. heating
Lighting
Oil Spc. heating
Wood
Elec. Clths. Dryers
Freezers
Elec. Cooking
LPG Spc. heating
Gas Cooking
Coal + kerosene
Oil Wat. heating
Gas Misc. Uses
LPG Wat. heating
Gas Clths. Dryers
LPG Cooking
LPG Other Uses
Gas Spc. cooling
End Use
3.8
September 15, 1997
The Buildings Sector
Chapter 3
Figure 3.3 Commercial Sector Primary Energy Use and Carbon Emissions in 1997 and 2010 by End-Use
for the Business-As-Usual Scenario
4.5
70
40
1997
2010
60
3.5
50
3.0
Energy Use (Quads)
40
2.5
2.0
30
1.5
20
Emissions (MIC) (1997 generation mix)
1.0
10
0.5
0.0
0
Lighting
Elec. Misc. Uses
Elec. Space cooling
Gas Misc. Uses
Gas Space heating
Off. equip.-non-PCs
Ventilation
Gas Wat. heating
Refrigeration
Elec Wat. heating
Elec. Space heating
Coal + kerusene
Off. equip PCs
Gas Cooking
Oil Space heating
Oil Misc Uses
Elec Cooking
Oil Wat. heating
Gas Space cooling
Biomass
End Use
3.3.2 Maximum Cost-Effective Energy-Efficiency Potential
In determining the maximum cost-effective technical potential to be used as a baseline for
development of the efficiency and high-efficiency/low-carbon scenarios, we reviewed and updated,
as needed, the major recent sources of data on energy use and costs associated with upgrading to more
efficient energy-using technologies. The results of this work, as well as the references on which it is
based, are found in Appendix C-3. Once we determined the cost-effective energy-efficiency
measures, we then used the energy use and incremental cost of new 1997 equipment for that end-use to
calculate the potential efficiency improvement for that end-use. Table 3.4 lists the 1997 end-uses
and their potential for energy intensity reductions when replaced by these highly energy-efficient
technologies. As the table indicates, compared to 1997 new equipment, significant savings potential
exists for many end-uses in the residential and commercial sectors.
The difference in energy demand between the maximum cost-effective case (100% of the potential)
and the business-as-usual scenario for all buildings is 6.5 quads/year of primary energy in 2010. The
efficiency and high-efficiency/low-carbon scenarios discussed below are based on the assumption
that various shares of these savings are achieved.
Figures 3.4 and 3.5 show the percentage breakdown of savings for electricity and natural gas (these
results are independent of the efficiency scenario because these scenarios vary only in the percentage
of the maximum cost-effective resource assumed to be implemented, not in the character of that
resource). More than 50% of the electricity savings is in "miscellaneous", and about a quarter is in
lighting, with the remaining quarter split between space conditioning, water heating, and
refrigeration. About half of the natural gas savings is in residential space heating, with
commercial space conditioning and water heating splitting the remainder about equally.
September 15, 1997
3.9
Chapter 3
The Buildings Sector
Figure 3.6 shows a conservation supply curve for electricity savings in the high-efficiency/low-
carbon case. This graph shows the electricity savings by end-use associated with the cost of
achieving those savings. On the x-axis are the projected savings in 2010 in TWh, and on the y-axis is
the cost of conserved electricity (CCE) in cents/kWh (1995$). Total savings in this scenario are
about 16% of baseline electricity use. The most cost-effective savings come from commercial
lighting, which has a negative net CCE because of the labor savings associated with replacing
incandescent A-lamps with longer-lived halogen IR and compact fluorescent lamps. The costs of
savings in other end-uses range from 1.4 to 4.5 cents/kWh.
Table 3.4 Cost-Effective Energy Savings Potentials for Selected End-Uses in the Residential and
Commercial Buildings Sector
Energy Savings
Potential:
End-Use
Retrofitted Shell/
New Equipment
Residential
Fuel Switching - clothes drying
59%
Lighting
53%
Miscellaneous electric end-uses
33%
Fuel Switching - Cooking
33%
Refrigeration
33%
Fuel Switching - water heating
29%
Electric water heating
28%
Freezers
28%
Electric space heating
25%
Gas and oil water heating
23%
Electric space cooling
16%
Gas space heating
11%
Gas and oil cooking
15%
Miscellaneous gas and oil uses
10%
Commercial
Space heating (electric and gas & oil)
48%
Space cooling (electric and gas)
48%
Ventilation
48%
Miscellaneous electric end-uses
33%
Refrigeration
31%
Lighting
25%
Electric water heating
20%
Gas and oil water heating
10%
Miscellaneous gas and oil end-uses
10%
*
Energy savings potentials are calculated as the percent difference in energy intensity of maximum cost-effective
technology and new 1997 technology. Savings are achieved using technologies listed in Appendix C-3. It is
important to note that the impact these potentials have on reducing energy demand in the efficiency and high-
efficiency/low-carbon scenarios depends not only on savings potential but also on the magnitude of energy demand
by the particular end-use (see Tables in Appendix C-2) and the rate of turnover of equipment for that end-use.
**
Fuel switching energy savings potentials reflect the unit energy savings in switching from electric clothes dryers,
ranges, and water heaters to gas. Electricity energy is calculated as source energy using conversion factors from the
utility chapter.
Energy savings potential for residential space conditioning is greater with new shells than with retrofitted
shells. Our estimates for electric space heating, electric space cooling, and gas space heating with new shells show
additional incremental savings of 14%, 7%, and 8%, respectively, beyond savings achieved with retrofitted shells.
3.10
September 15, 1997
The Buildings Sector
Chapter 3
Figure 3.4 End-Use Electricity Savings, 2010
Figure 3.5 End-Use Natural Gas Savings, 2010
Residential
Space Cond.
Misc.
4%
Commercial
Commercial
33%
Space Cond.
Water Heating
11%
25%
Water Heating
6%
Residential
Res. Lighting
Space Cond.
11%
50%
Misc.
Residential
Commercial
33%
Com. Lighting
12%
Space Cond.
25%
Refrig./Freezers
4%
Note: The proportions of electricity and natural gas savings do not vary across scenarios. Total electricity
savings in 2010 in the high-efficiency/low-carbon case are about 400 TWh, while total natural gas savings in this
scenario are about 0.5 quads.
3.3.3 Efficiency Scenario Results
The efficiency scenario assumes that 35% of the maximum cost-effective efficiency savings are
achieved by 2010. This assumption is based on expected savings resulting from a moderately
vigorous effort to reduce energy use and carbon emissions using a combination of policy mechanisms
that may include higher prices resulting from a cap and trade system, energy-efficiency standards,
and information programs.
In the efficiency scenario, 2010 energy use drops to 34.1 quads while carbon emissions decline to 546
MtC. In this scenario, the total annual cost of energy services is $11 billion per year less than the
annual energy services cost in the business-as-usual scenario, reflecting the fact that the decrease in
energy expenditures resulting from more efficient technologies is greater than the increase in costs to
purchase and install the technologies in residential and commercial buildings. The largest energy
savings by end-use occur in miscellaneous electricity, lighting, water heating (residential) and space
cooling (commercial).
3.3.4 High-Efficiency/Low-Carbon Scenario Results
The high-efficiency/low-carbon scenario assumes that 65% of the maximum cost-effective efficiency
improvements are realized by 2010 as a result of a vigorous effort to reduce energy use and carbon
emissions. In this scenario, 2010 energy use and carbon emissions drop further, to 31.6 quads and 512
MtC, respectively, at a total cost savings of $20 billion per year below the business-as-usual
scenario. Annualized capital costs increase by $6 billion over the costs in the efficiency case, but
annual additional bill savings are about $15 billion. Some of the carbon savings in the high
efficiency/low-carbon case are associated with changes on the electricity supply side (see Chapter 6
for details).
September 15, 1997
3.11
Chapter 3
The Buildings Sector
Figure 3.6 Electricity Supply Curve By End-Use for Buildings in 2010, High-Efficiency/Low-Carbor
Case
8
2010 Buildings Sector Electricity Price - 7.4 cents/kWh
Discount Rate: 7%
6
Forecast Year: 2010
Start Year: 1997
Baseline Electricity Consumption
for year 2010 = 2453 TWh
10
Cost of Conserved Electricity
4
7
8 9
56
2
3
2
16% of
Baseline
Use
O
-2
1
-4
0
50
100
150
200
250
300
350
400
450
500
Savings in 2010 (TWh)
1 Commercial lighting
6 Commercial water heating
2 Commercial space conditioning
7 Commercial other uses
3 Commercial refrigeration
8 Residential refrigerators and freezers
4 Residential lighting
9 Residential water heating
5 Residential space conditioning
10 Residential other uses
Efficiency potential is calculated assuming 65% of technoeconomic potential is captured in the
high-efficiency/low-carbon case. Savings from reflective roofing are contained in the residential
and commercial space conditioning end-use categories.
3.12
September 15, 1997
The Buildings Sector
Chapter 3
Improving Efficiency and Saving Capital
Adding proven efficiency technologies to new homes can reduce monthly energy bills substantially.
What is less well known is that clever design of new homes can also result in capital cost credits
that can offset, in whole or in part, the additional capital costs of the more efficient technologies.
For example, adding improved insulation and windows can allow a builder to reduce the size of the
heating and cooling equipment and in some cases eliminate ductwork altogether. These credits can
only be captured by builders who take a whole systems approach to design, but the benefits of such
an approach are large, as shown by two real-world examples below.
Perry Bigelow, a builder in the Chicago area, has for years built highly energy-efficient homes
that cost only $300 to $500 more to construct, in spite of his guarantee that these homes will have
heating bills no higher than $200 annually (Andrews 1994). He accomplishes this goal by creating a
well-insulated building envelope with little air leakage (taking care to provide appropriate levels
of ventilation) and by replacing the furnace with a high-efficiency water heater that also doubles
as the space heater. By using hydronic heating, he can save $1000 on ductwork. He also can
downsize the air conditioner because the home's cooling load is so much lower than typical practice,
saving another $80 to $100. These savings totally offset the cost of the added insulation and the air
sealing, leaving a small additional cost to pay for low-emissivity gas filled windows and
fluorescent lighting.
Builder Barbara Harwood, whose company is based in Carrollton, Texas, built a block of homes in
Dallas called Esperanza Del Sol (Schwolsky 1997). The homes are small (1273 square feet) and
inexpensive ($80,000), but are so efficient that Harwood can guarantee that heating and cooling
costs will be no more than $1/day ($365/year). She upgraded insulation levels, reduced air
infiltration, and added an active ventilation system. To offset these costs, she used a smaller-
capacity geothermal heat pump and redesigned the ductwork. With these offsetting cost credits,
the more efficient homes cost only $150 more than their inefficient counterparts, but save about
$40/month in energy bills. The consumer who purchases these homes would have to pay another
$1.10/month on an 8%, 30 year mortgage to finance the added capital cost; the monthly energy
savings are almost 40 times larger, providing immediate positive cash flow to the homeowner.
These builders have discovered the benefits of an integrated design approach. They both use the
"hook" of guaranteed maximum energy bills to market efficiency to customers who might otherwise
be reluctant to spend more for it. They have shown that, with correct sizing of equipment and clever
redesign of building systems, highly efficient homes need only cost a little more up-front.
Commercial buildings can also benefit from HVAC equipment downsizing. Pacific Gas and Electric
Company's Advanced Customer Technology Test for maximum energy efficiency (ACT²) had one
pilot project in San Ramon, California (Houghton et al. 1992). This 20,000 square foot office building
was retrofit using improved glazing, more efficient lighting, and better controlled HVAC systems.
Chiller capacity was reduced by more than 40% because of better solar control from the windows and
the reduced internal loads from lighting. The savings from the smaller chiller offset some of the
cost of the window and lighting retrofits.
September 15, 1997
3.13
Chapter 3
The Buildings Sector
3.4 POTENTIAL FOR ADVANCED TECHNOLOGIES IN 2020⁶
To the casual observer, buildings in the year 2020 may look much like the buildings of today (Smith
and Rivera, 1989). This is because Americans prefer familiar forms for their buildings and because
new buildings amount to only 2-3% of the existing building stock in any given year. Nearly 90% of
the residential buildings, and 80% of the commercial buildings, that existed in 1997 will still be
occupied in 2010. By 2020, significantly more than half of the 1990 stock will still be in service.
However, beneath the surface, many significant changes are expected to occur that will affect how
buildings are constructed, the materials and systems used to build them, and the way in which
buildings are maintained and used (Smith and Rivera, 1989; Wendt, 1994). Without a sustained and
vigorous public-private research, development, and demonstration (RD&D) partnership, these
changes could lead to only modest improvements in energy efficiency. In contrast, an invigorated
buildings RD&D scenario over the next 25 years offers the potential to produce breakthrough
technologies that could dramatically reduce the energy requirements and environmental impacts of
buildings, while enhancing affordability, long-term durability, resistance to disasters, and indoor
environmental quality.
For advanced energy-efficiency technologies to penetrate the buildings industry by the year 2020,
they will have to be cost-effective, and passing the cost-effectiveness hurdle will be challenged by
energy prices that could decrease well into the 21st century. Thus, incorporation of additional
features to make energy-efficient technologies more attractive to consumers will be needed to ensure
success in the marketplace and should be part of the R&D planning process. RD&D will also be
instrumental in capturing the potential of existing technologies by establishing better programming,
design, and commissioning practices for buildings (Todesco 1996). Further, investments in training
and education will be required to enable technicians and engineers to keep pace with a new
generation of technologies and practices. New construction techniques, novel heating systems,
electronic appliance tuning and control, more sophisticated building wiring practices, and the field
installation of factory-built housing all require new talents for those who build, maintain, and
service buildings. There must also be a concerted effort to facilitate the integration of new
technologies.
This section identifies the potential improvements to energy-efficiency technologies that could
result by 2020 given a sufficiently vigorous R&D effort. Savings discussed here would be in addition
to savings estimated in the quantitative analysis for 2010 above.
3.4.1 New Technologies and Practices
Many of the changes in building technologies occurring over the next 25 years will be evolutionary in
nature, resulting from ongoing research that is continuously providing solutions to such issues as
moisture damage in structures, anomalous heat losses from envelopes, and indoor air quality
problems. By 2020, these solutions will have evolved into cost-effective practices and products that
will be the norm in new and existing buildings. In addition, a sustained, vigorous program of public
and private-sector RD&D could produce many novel building technologies and practices by the year
2020. The following six areas offer great promise to significantly reduce the energy requirements of
our nation's buildings through a combination of incremental and aggressive technology
improvements:
Advanced construction methods and materials;
Environmental integration and adaptive envelopes;
3.14
September 15, 1997
The Buildings Sector
Chapter 3
Multi-functional equipment and integrated system design;
Advanced lighting systems;
Controls, communications, and measurement; and
Self-powered buildings.
In each area, thought must be given not only to energy-efficient technologies and energy costs, but
also to the incorporation of other beneficial non-energy features that will accelerate their
introduction into the marketplace, such as lower first costs, ease of integration, time savings,
durability, comfort, and improved indoor environments.
3.4.1.1 Advanced Construction Methods and Materials
With sufficient RD&D support over the next 25 years, a systems engineering approach to the
building's life-cycle (programming, design, construction, commissioning, financing, operation,
renovation, reuse, and disposal) could become the norm. Such a transformation offers the potential
to deliver buildings with lower total first costs and lower energy consumption, as well as higher
overall quality and faster construction (Lawson, 1996; Lovins, 1992). The lower total first costs will
permit the reinvestment of some capital savings into additional cost-effective, energy-efficient
technologies. The total reduction in energy use could thereby be considerable.
By the year 2020, on-site labor for single-family homes, low-rise multi-family construction, and
commercial buildings of standard design (e.g., franchise restaurants and retail stores) will consist
primarily of assembling manufactured components and installing complete modules. This shift will
require less skilled, and more semi-skilled, on-site labor. The expanded use of CAD/CAM
technologies could enable "mass customization" capabilities, permitting the manufacture of
virtually all residences and many commercial buildings. Quality and material improvements that
are not affordable on a one-of-a-kind basis, can be assimilated into the high-volume manufacturing
process. Continued research into the manufacture of building components is needed to enable these
changes, to reduce waste, and to facilitate the recycling of unused materials.
Advanced modular construction methods will result in attractive, affordable, and flexible buildings
that will permit longer occupancy in homes, offices, and other commercial buildings. Modular and
easily installed heating, ventilating, and air-conditioning (HVAC) units with improved, leak-free,
insulated ducting will reduce installation and operation costs. By extending the average length of
stay in buildings, life-cycle costs become more important to decision makers. Durability and the
need for reusable and recyclable materials will therefore increase in importance, generating the
need for better durability testing tools and advances in materials, systems, and assemblies (Darrow,
1994). Better "engineered" wood, stress skin panels, optimized light-weight steel components, and
adhesive assembly techniques will be needed. Greater use of recycled materials requires the
development of higher "value added" uses for current wastes and the invention of low-value
recycled products. Examples being developed today include the following: (1) mixed paper waste in
lieu of pure newsprint to cellulose insulation and drywall; (2) wood wastes to engineered structural
members as opposed to only particle board; (3) flyash to lightweight masonry products as opposed to
site fill material; (4) corrugated paper to structural insulating panels; and (5) plastics to carpeting
and wood/plastic composites. The recycled materials must also be low- or non-emitting materials in
order to meet consumer concerns about indoor air quality. A program of vigorous materials research
could make these new materials commonplace by 2020.
September 15, 1997
3.15
Chapter 3
The Buildings Sector
By 2020, building life-cycle information management systems will create efficiency in the
architectural/engineering/construction process and in building operations. Information systems will
facilitate communication of programming and design intent through construction, commissioning,
maintenance, and operation of buildings. Performance tracking will insure persistence of savings
from efficient design and equipment. And, most significantly, continuous improvement in buildings
will occur through feedback of performance information to design of new buildings and renovations.
Over the next quarter century, there will be greater use of computer software in every aspect of the
building life-cycle. Design tools and building simulators will be more powerful and easier to use,
with improved graphical interfaces and links to manufacturer databases of equipment
specifications. There will be construction management and commissioning software for use in all
stages of a building's life-cycle including early design and commissioning. This software will be used
to create calibrated computer models to verify that actual building performance meets pre-specified
design targets that could be part of a performance contract. The calibrated model could have many
uses in operations and maintenance, including assisting in evaluation of the least-cost energy
supplies, optimization of existing control strategies, and analysis of possible retrofit options.
Finally, such data on actual as-operated conditions close the feedback loop that is problematic
today. Building designers will finally have an opportunity to learn how buildings they design
actually perform, and their future designs will benefit from lessons they learn based on existing
buildings.
3.4.1.2 Environmental Integration and Adaptive Envelopes
Advanced designs and technologies that intelligently integrate the performance of buildings with
the outdoor environment offer the potential to more efficiently heat, cool, insulate, ventilate, and
illuminate interior spaces. A variety of building designs tailored to the wide range of climates in
the U.S. will reduce first costs and operating costs. Equipped with these climate-specific and smart
technologies, the word "shelter" will no longer imply the exclusion of outdoor elements; instead it
will refer to structures that capitalize on fluctuating outside conditions to create interior comfort and
light.
One of the most significant changes in envelope performance from 1970 to 1995 was the development
of a new generation of window technology that involved high-transmittance low-emissivity (low-
E) glazings; the introduction of this new window technology resulted in a major shift in the window
marketplace. By 2020, the market penetration of such technologies could double as high-rate, thin-
film coating techniques make it possible to coat glass and plastic for cost-effective use in virtually
every climate. New types of highly insulating glazings (such as aerogel and honeycomb) will
compete for new markets if materials research is able to produce a window that, by enabling the
diffuse solar gain to exceed the winter thermal losses, outperforms a highly insulated wall even on
northern exposures in winter.
In most larger commercial buildings and in sunbelt housing, control of solar gain is critical. Since
building needs vary widely and climatic variables are unpredictable, one ideal component would be
a dynamically controllable "smart glass". The fundamental materials science technology base for
"active" and "passive" smart glazing technologies such as electrochromic coatings was developed in
the 1990s. However, RD&D resources are needed to develop viable and cost-effective materials
with optical properties that can be switched passively. In addition, research on switching
mechanisms is needed to assess the potential applicability of the range of alternatives, including
short wavelength switching to a reflective mode and long wavelength switching for thermal
comfort (Kammerud, 1995).⁷
3.16
September 15, 1997
The Buildings Sector
Chapter 3
To date, research on insulation has focused on static insulation systems, where insulation is simply
put in place to increase the thermal resistance of the roof, wall, or floor by a fixed amount. An
alternative is to consider dynamic systems, in which the performance of the building envelope
changes with the environment to minimize the building energy load. One study (Fine and McElroy
1989) found that dynamic building envelope systems (insulation, roofs, walls, and windows) could
reduce heating and cooling loads by 20 to 35%. Adaptive envelopes should be developed which
integrate other useful features, such as ventilation air intakes with heat exchangers and sensors
that are engineered as an integral part of the envelope, or energy-efficient windows as part of a unit.
Better use of thermal storage concepts would increase the ability of passive solar heating and
cooling to offset the use of mechanical systems. One possibility is to distribute natural heating and
cooling more uniformly over the day with resultant decreases in both heating and cooling
requirements. Development of phase-change materials with storage capacity and release rates
adapted to building use is needed. Applied R&D is needed to make such materials economically
competitive with standard building products, and to demonstrate their durability and safety. In
addition, to achieve the technical potential of these thermal mass strategies, design and
construction guidance is needed to identify how mass and insulation should be rearranged to optimize
thermal storage effects in specific climate regions (Christian, 1991).
Self-drying roof concepts are under development, and their commercialization offers significant cost
and energy benefits. Behind this work is the notion that roofs should be designed to accommodate
occasional leaks; that is, there should be a means to dry out the roof and restore it to its original
thermal performance after a leak is patched. One promising technique is to design roofs that dry to
the interior through evaporation. By extending roof life, self-drying promotes the installation of
better insulation, since the originally installed insulation will remain in place longer. In addition to
reducing energy loss, self-drying roofs also significantly reduce the cost of repairing, replacing, and
disposing of roofs.
The success of environmentally adaptive envelopes depends upon improved design and
commissioning practice, the development of advanced manufacturing techniques, new materials, and
sensor and control technologies to produce customized wall, roof, and floor panels that meet the
needs of buildings in different climates. Other important properties and features should be
simultaneously sought in the development of new materials such as reduced maintenance, resistance
to water condensation, and low emissions. Research is also needed to integrate the dynamics of such
advanced envelopes into total building energy management systems.
September 15, 1997
3.17
Chapter 3
The Buildings Sector
Mitigating Urban Heat Islands With Cool Roofs And Trees
The benefits of reducing urban heat islands through reflective roofing, white pavements, and tree
planting have gained increasing attention in recent years (Rosenfeld et al. 1996 and 1997, Konopacki
et al. 1997). These savings are both from the direct effect of sunlight being reflected (by white roofs)
or blocked (by trees) and hence prevented from entering the building envelope, and from the indirect
effect of cooler ambient conditions brought about by evapotranspiration from trees and increased
albedo. The cooler ambient conditions have the additional benefit of reducing smog formation
(which is directly related to air temperature).
The calculations above include estimates of savings from the direct and indirect effects of cool
roofing on building energy use but do not include the potential effects of large-scale tree planting. In
the efficiency case in 2010, cool roofs save about 4 TWh of cooling electricity, while increasing
heating gas use by 0.01 quads. In the high-efficiency/low-carbon. case in 2010, cool roofs save about 8
TWh of cooling electricity (worth more than $500 million per year), while increasing heating gas
use by 0.02 quads (worth more than $100 million per year). The associated net carbon savings (after
subtracting out the penalty for the increased heating gas use) are 0.2 MtC in the efficiency case and
1.3 MtC in the high-efficiency/low-carbon case. The cost of these reductions are negligible, because
changing roofing materials to be more reflective at the manufacturing stage is generally a zero cost
option. The development of advanced roofing, paving, and coating technologies would improve the
longevity and economics of these cool community options.
The additional savings from tree planting have not been included in the calculations, but the direct
and indirect effects from trees are generally of the same order of magnitude as for cool roofs
(Rosenfeld et al. 1996). The total savings from cool roofs and trees together would therefore be on
the order of 2-3 MtC in the high-efficiency/low-carbon case by 2010.
The cost of tree planting is more difficult to estimate, because of the sizeable unquantifiable benefits
of trees, as well as the long-term maintenance costs. Most people regard trees as a net positive
contribution to their local environment, and it is likely that the overall benefits (including the
energy and carbon savings benefits) substantially exceed the costs, but because of the uncertainties in
estimating these costs, we did not include tree planting in our savings estimates.
3.4.1.3 Multi-Functional Equipment and Integrated System Design
During the period through 2010, the efficiencies of HVAC equipment, water heating and other
appliances will continue to increase through incremental improvements. Efficiency improvements
will probably continue to be driven both by minimum efficiency standards as well as by marketplace
competition for technologies that have low operating costs because they are efficient. In many cases,
however, appliance and equipment efficiencies are reaching either their thermodynamic limits, or
can be made higher only at significantly higher first cost.⁸ For example, electric resistance water
heaters have become more than 90% efficient with 100% as the maximum. Gas water heaters and
refrigerators provide other examples where efficiencies may be reaching either an economic or
thermal limit. Condensing gas water heaters that have efficiencies above 90% have been
developed, but are generally too expensive for a mass market. In the case of refrigerators, applied
research and development has recently produced a 20 cubic foot refrigerator which consumes no more
electricity than a 40-watt light bulb running continuously (350 kWh/year). We anticipate that the
technologies used to reach this performance level will be available to the U.S. refrigerator market
in the next decade. To move refrigerators, as single-function appliances, beyond this level of
performance does not appear to be cost-effective in the near-term or beyond if real energy prices
continue to decrease.
3.18
September 15, 1997
The Buildings Sector
Chapter 3
Opportunities continue to exist for reducing losses in poorly designed hot water storage and
distribution systems. Improved tank/flue designs, improved piping layout and design, and advanced
circulation systems are some of the possibilities.
Based on the limits to performance for single-function equipment such as refrigerators, water
heaters, and HVAC equipment, RD&D efforts need to focus on multi-functional equipment and
appliances to provide the next quantum jump in efficiency improvement. Multi-functional equipment
needs to be developed that combines and integrates the functions of several appliances into a single,
highly efficient device. Such equipment promises to be highly efficient because the heating and
cooling that is rejected by a single-function device can be put to use in the integrated appliance, and
the component with the highest efficiency can be used to provide a dual function.
An example of multi-functional equipment is an integrated water heating/space conditioning system
which uses heat pumping to meet space heating, air conditioning, and water heating loads. As a
combined, integrated appliance, this unit's efficiency (as measured by the Seasonal Energy
Efficiency Ratio, or SEER) could be a full 70% higher than the combined efficiency of today's central
air-conditioning system and water heating system. Energy-efficient air filtration, as well as
humidity and temperature control, could be incorporated into HVAC systems to reduce indoor
concentrations of airborne particles such as pollen, other allergens, and infectious agents that cause
adverse health effects. This type of integrated technology can be applied to residential as well as
commercial buildings. As the efficiency of a single-function device is improved through incremental
development, as part of an integrated approach, this device is able to provide still higher
efficiencies.
There is also a large opportunity for integrated products that can control space humidity and
temperature independent of each other. Research on combined systems that use desiccants to control
humidity and vapor compression air conditioning to control temperature is expected to result in an
efficient, integrated system that can provide better comfort at reduced operating costs.
Further opportunities exist for improving the efficiency of heating and cooling systems in buildings
through integrated systems design, right sizing, modular/multiple equipment configurations, and
better integration of the process for distributing space heating and cooling within buildings
(Shepard 1995). As air conditioning and chiller efficiencies continue to improve with cascade,
multi-stage, and turbine-assisted compressors, the energy consumption and electrical demand
associated with oversizing, poor part-load performance, and the distribution of air and water
becomes a greater fraction of the total HVAC energy use in both residential and commercial
buildings. Research on load diversity, system integration, and design paradigms can reduce both
peak demand and energy use. In addition, research on advanced thermal distribution technologies
could enable the development and commercialization of higher-efficiency, quieter thermal
distribution systems, with air filtration to improve indoor air quality. At a higher level,
integrating heating/cooling devices as part of the distribution system itself, along with improved
integration of task/local environmental control systems, would provide efficiency benefits and
enable use of control technologies to target heating and cooling within a building.
There are other options for appliance integration, including combining water heating with
dehumidification, mechanical ventilation, and/or refrigeration. In these cases, heating or cooling is
produced for multiple applications and at much higher efficiency than would otherwise be possible.
In the 2000 - 2010 time period, research in fields of heat transfer, controls, component technology
development, and systems analysis will need to be conducted so that industry can take these results
and apply them to developing integrated products for both residential and commercial buildings. By
2020, we anticipate that efficient integrated and multi-function products could capture a substantial
fraction of the U.S. market for space conditioning, ventilation, water heating, and refrigeration.
September 15, 1997
3.19
Chapter 3
The Buildings Sector
3.4.1.4 Advanced Lighting Systems
Lighting is a dominant energy end-use in the commercial sector, an important use in houses, and an
essential element of roadway and outdoor use. At the national level, lighting accounts for 23% of all
U.S. electrical energy use. Through the development and intelligent use of more efficient lighting
technologies and design, lighting energy use could be reduced by over 50% by 2020 with equal or
improved health, comfort and productivity.
Lighting use is characterized by a tremendous diversity of applications and needs, and an equivalent
diversity of sources, fixtures, controls, and designs. Thus, energy efficiency can best be achieved by
an array of new and existing technologies intelligently matched to the appropriate lighting needs.
Unlike other aspects of the building infrastructure, most lighting system components are replaced at
a relatively high turnover rate within ten years, and thus provide opportunities to introduce more
efficient technologies on a regular basis. At the national scale, we spend $10 billion/year for new
lighting equipment but $40 billion/year for lighting energy consumption. By 2020, we must make a
transition to investing more each year in improved technology with the benefit of dropping the
annual consumption figure by 50%.
Changing the overall efficiency of U.S. lighting use can be viewed as improving four efficiency
parameters: (1) lamp or ballast efficacy, (2) fixture efficiency, (3) spatial task efficiency, and (4)
temporal control efficiency. There are large opportunities for improvements in each of these areas:
Lamp efficacies for fluorescent and other gas discharge sources have improved modestly over the
last 20 years, but are still well below the theoretical limit. The industry is exploring new
electrodeless solutions in both small sizes (10-100 watts) and in the kilowatt range. Large lamps,
such as the sulfur lamp, have demonstrated higher performance in prototype form. Some
technologies have other advantages, such as reduced maintenance due to long operating life, or
better environmental properties (e.g., mercury-free lamps). Most of the new discharge sources will
benefit from continued development of less expensive, smaller, and more efficient electronic power
supplies. Dimmability will also be more readily achievable using these new power supplies. Light
sources that use phosphors may be further improved by advances in the chemistry of phosphors.
By 2020, there will be many new CFL options with smaller size, better color rendition, higher
luminous output, and dimmability. But there will still be a tremendous market need for a long life,
very low cost, incandescent lamp replacement, perhaps utilizing improved filament technology or
halogen lamps with IR reflecting coatings. Finally, there are other contenders for the small source
market such as mini-HID sources and solid state light sources (LEDs or laser diodes).
There will be continued improvement in fixture design for both direct and indirect lighting systems
so that a greater fraction of the light is usefully extracted from the source, using innovations in
highly reflecting surfaces, refractive and diffracting materials, and non-imaging optical designs.
Two seemingly contradictory trends will continue through 2020. One trend will be towards localized
lighting that provides just the lighting needed at each task location and is flexible enough to adapt
to the ever-changing needs of today's office and factory environments. The other trend is towards
the use of centralized lighting in situations that require uniform light levels on a fixed schedule
over long periods of time. Hollow light guides and light pipes must be developed to meet these
needs and fiber optic designs can be used for smaller-scale centralized solutions.
Lighting controls have only recently advanced beyond simple on-off, multi-level, or time clock
controls to occupancy-based controls and photosensor controls that respond to daylight and lumen
maintenance. By 2020, new generations of smart control systems will respond automatically to
changing task and environmental needs. Voice-activated controls and flexible linkages (wired and
3.20
September 15, 1997
The Buildings Sector
Chapter 3
wireless) between light sources and tasks will provide new flexibility in both office and retail
environments. Controls linked to dimmable lighting systems and to building energy management
control systems (EMCS) will provide an equivalent spinning reserve load that can be used by owners
when negotiating utility contracts with electricity suppliers in the deregulated environment of 2020.
Some of the most important issues in the lighting community today are related to the human
dimension of occupant response to the indoor luminous environment. Lighting design has a direct
impact on performance, health, and satisfaction in the built environment; however, the nature of
that impact remains elusive. By 2020, the challenge is to conduct the research studies that will
establish definitive causal linkages between design parameters and occupant impacts, and then
apply these conclusions to the development of new technology and designs.
With only a modest RD&D effort, incrementally more efficient lighting components, including
improved bulbs, fixtures, and controls, will be in use throughout all building types in 2020. Important
improvements in lighting performance will result from using advanced techniques to improve the
performance of fluorescent lamps and expanded use of diodes as light sources. Systems will be
available to permit the integration of very-high efficiency lighting such as the sulfur lamp into
common interior spaces.
A more vigorous program of lighting research could ensure that, by 2020, the nation will be
discovering the virtues of lighting systems that deploy a mixture of centralized, energy-efficient,
artificial light sources, tracking sunlight concentrators, and light distribution systems for buildings
with high lighting usage. Offices and retail stores that require high lighting levels would be ideal
candidates to field test such systems. A few, high-intensity, super-efficient light sources, centrally
located, could then replace the numerous distributed light bulbs currently used. Whenever local
climatic conditions permit, the sun could provide the light source in lieu of artificial sources. This
piped lighting system could enhance many daylighting strategies based solely on architectural
design elements. These piped systems, which use sunlight supplemented by super-efficient
artificial light sources, could cut lighting-related power consumption in office buildings
dramatically, since sunlight is usually available during normal office working hours. In addition to
significant reduction in energy consumption for lighting, this system offers the potential to
dramatically reduce lighting maintenance costs by using fewer artificial light sources and for much
shorter periods.
Development of such lighting systems will require scientific breakthroughs and technical expertise
in advanced artificial light sources, optical systems design, materials development, thin film
coatings technology, fiber optics, photonics, manufacturing technology, systems engineering and
modeling, instrumentation and controls, and human factors.
3.4.1.5 Controls, Communications, and Measurement
Computer technology has made possible a revolution in equipment and capabilities for electronic
control of devices in homes, offices, and industry over the past 20 years. Similarly, significant
advances in communications and information capability have introduced major changes in life styles
and work practices over this same period. Over the next twenty years, this trend is expected to
continue, offering additional opportunities to increase the efficient use of energy in buildings. The
increasingly deregulated and converging energy and communications industries will play a major role
in defining, commercializing, packaging, and delivering these new energy services and technologies
to building owners. The fact that deregulation has resulted in greatly reduced RD&D investments
by utilities underscores the need for a sustained, vigorous public-private partnership to ensure that
energy-efficiency innovations emerge.
September 15, 1997
3.21
Chapter 3
The Buildings Sector
The communications industry has adopted programs for universal hardware and software
connections between most functional components. The controls industry has initiated similar
measures (ASHRAE Journal, November 1996, p.36). When universality is achieved, systems
designers can begin to lay out and wire buildings with centrally located communications/control
centers for all buildings including homes. This affords the opportunity to significantly reduce power
requirements by eliminating full replication at each building station. That is, there needs to be only
one video/audio receiver with low-power monitors at other sites, one computer central processing
unit with low-power (e.g., liquid crystal) terminals where needed, one energy management control
system (EMCS) with zone controllers where needed, and so on.
Developing and incorporating increased intelligence directed at energy use and control diagnostics in
future generations of EMCS will allow these devices to maintain higher quality building
environments with less expenditure of energy. Expected advances include EMCS with performance
evaluation and equipment status tracking ability, as well as predictive capabilities. For example,
EMCS with more powerful computational capability and with more sophisticated mathematical
modeling can couple weather predictions with building response characteristics and occupancy,
light, and moisture sensors to predict building performance and more closely match supply and
demand of HVAC and lighting. Energy management and control systems may also be developed to
enable the selection of least-cost energy service providers and rates (see further discussion under
"Self-Powered Buildings" below).
Future EMCS will utilize networks like the Internet to transmit data, sound, and video for real-time
remote analysis. This will permit integrated buildings service providers to track the performance of
heating and cooling plants, diagnose failures, test machinery, and to communicate findings to
building owners and operators, all without setting foot in the building. Some "full service" providers
would also offer other services including energy management, security, and property and facilities
management.
For appliances such as clothes washers and dryers, control and communications capabilities will
allow for remote programming and cycle control as needed. Delayed start, checking on cycle progress
from a remote location, and modification of settings remotely are all examples of potential
capabilities. Additional research to develop more sophisticated sensors and control logic will
increase future ability to measure and control energy use in the ever-widening pool of appliances and
equipment used in buildings. Advanced sensors can check the status of food being cooked, room
lighting levels, and thermal comfort and instruct controllers to automatically adjust appliances for
optimum operation.
The development of advanced sensors, controls, and communications equipment needs to reflect the
nature of changing "plug load" devices in buildings. The forecasted rapid growth in miscellaneous
electricity consumption in buildings suggests an important future role for a broad range of novel
control strategies to promote energy efficiency. In addition, advances in office equipment
performance could mitigate potential increases in these miscellaneous electricity uses in many parts
of the commercial sector (Komor 1996).
3.4.1.6 Self-Powered Buildings
The move toward a competitive marketplace for energy services such as gas and electricity will be
essentially complete by 2010. By 2020, that market will have matured to accommodate complex buy-
sell utility service arrangements monitored and administered by automated systems. This, combined
with the advent of power production and improved energy storage technologies, will give building
owners new levels of flexibility in meeting their energy requirements, as well as the possibility of
revenue streams from the sale of energy or ancillary services. Buildings will cease to be simply
3.22
September 15, 1997
Buildings Sector
Chapter 3
mers of electric utility services but may supply all or a portion of their own energy requirements
the economics are right, sell to others. Removal of utility and environmental regulatory
ars would also accelerate the adoption of combined heat and power systems.
turbines running on natural gas are likely to be the first step in this process. These will allow
sings to generate their own electricity, with the reject heat from the turbines being used for
"estic hot water or building space conditioning. Six manufacturers have announced actual or
availability of gas turbine electric generators in the 50 kW range. Costs are uncertain, but
kely mature in the $750-$1000/kW range, including heat recovery equipment. Barriers to
mentation include mechanical maintenance requirements as well as cost. The advent of
ated control and diagnostic systems will make these distributed power plants as "forgettable"
other piece of space conditioning equipment.
ext step in the development of the self-powered building will be the advent of low-cost fuel
The fuel cell is a unique technology that can revolutionize the way building power, heating,
and water heating are generated and maintained.
Potential Additional Savings From Advanced Fuel Cell Technologies
high-efficiency/low-carbon scenario, fuel cell technology is also likely to make a contribution
carbon emissions by 2010. While we have not included fuel cells in our main building
scenarios, we examined recent technology projections from Arthur D. Little (ADL) and
the potential carbon savings from fuel cells in our high-efficiency/low-carbon case.
are several different fuel cell technologies under development, including the phosphoric acid
(PAFC), the proton exchange membrane fuel cell (PEMFC), the molton carbonate fuel celli
and the solid oxide fuel cell (SOFC). In addition, there are advanced gas turbines under
""Iment that could supply the same services as fuel cells, for comparable costs. We do not
the exact mix of technologies that might deliver carbon savings by 2010, but calculate the
impacts assuming that some combination of these technologies would contribute savings.
1). Little created what they termed an "optimistic" scenario that resulted in 8200 MW of
fuel cell capacity in commercial buildings by 2010. This estimate assumes a $50/tonne
harge and an aggressive commitment to building sector fuel cell development at or above
levels of funding. Their results imply that about 5% of all commercial building floor area in
have heat and power supplied by fuel cells.
penetration of a new and untried technology is ambitious by any measure. Because we are
in a "best estimate", not an optimistic scenario, we chose to reduce the expected
ion to 65% of ADL's forecasted levels for our high-efficiency/low-carbor case (4.9 GW). For
cautious case, we reduced the penetration again to 35% of ADL's forecasted levels (2.45
is
described in Table C-2.9 in Appendix C-2, implementation of this technology (or some
of fuel cells and small advanced gas turbines in buildings) at the efficiency case level
"sult in primary energy savings beyond the high-efficiency/low-carbon scenario of about 0.14
and additional carbon savings of about 2.5 MtC. The savings in the cautious case would be
half of the efficiency case savings. (See also Appendix D-3, in which the technical potential
mercial-sector advanced turbine systems in the 5-15 MW size range is estimated to be about 12
"10 at an estimated cost of $350/kW.)
no other system identified provides all the benefits of the fuel cell. The fuel cell can
electricity, provide heat and hot water, offer fuel flexibility, and operate quietly; in
15, 1997
3.23
Chapter 3
The Buildings Sector
addition, the fuel cell is modular, is a non-polluter, and has an overall conversion efficiency
potential of 80% or better (Fiskum, 1997). Unlike gas turbines, fuel cells have no moving parts and
are therefore inherently quiet. The ability to tailor the installation to the thermal needs of the
building by selection of fuel cell technology will also be attractive. For example, proton exchange
membrane (PEM) fuel cells, whose operating temperature does not exceed 100 degrees Centigrade,
will be used in installations with only low-level waste-heat applications such as domestic water
heating. Other types, such as molten carbonate and solid oxide fuel cells, operate at higher
temperatures for applications requiring a higher quality heat resource.
Fuel cell prices currently range from $3000/kW to $5000/kW for commercially available phosphoric
acid and near-term PEM cells, respectively. An aggressive RD&D program could cut these costs in
half in less than ten years. Research needs include work on high-risk components and processes,
including heat exchanger development to bring the high-temperature hydrogen stream in line with
PEM cell stack temperature, and catalyst development to increase CO tolerance and to mitigate
carbon monoxide contamination degradation of the catalyst (Fiskum, 1997).
Another key component of the self-powered building will be building-integrated photovoltaic (PV)
panels, an application which will become more widespread as the costs of PV cells decline. Full
implementation of this concept will require storage to achieve full flexibility, and such systems
could include compact, high-efficiency flywheels as a means of taking advantage of the diversity
between load and resource peaks. In some applications, notably commercial buildings located in
high solar resource areas, the coincidence between the mid-afternoon resource peak and the demand
for such services as air conditioning may minimize the need for storage. In any case, the
availability of an electric power spot market, accessed by the building's automated energy
management computer, will allow real time purchases of power when needed or sales of excess
power when available. PV system costs are still in the range of $7000/kW without storage, but
improvements in solar cell manufacturing processes and inverter technologies support program goals
calling for reductions of more than 50% in ten years or less.
3.4.2 Best Practice Buildings in the Year 2020
3.4.2.1 "Best Practice" Housing in 2020
By the year 2020, a vigorous RD&D program could produce many advanced technologies that
together will greatly reduce the average annual energy budgets of American families. The "Best
Practice" home of the year 2020 is defined as a home that employs those energy technologies that
are predicted to have the lowest life-cycle costs when purchased in the year 2020, under the
assumption that a "high-efficiency/low-carbon" scenario unfolds between now and then. A collage
of these best practice features is shown in Figure 3.7.
The best practice home in the year 2020 will be factory built and shipped to its site as modules or
subassemblies. The use of integrated systems design and CAD/CAM technologies for "mass
customization" will have produced these components and modules to reflect the particular
requirements of the home buyer. On-site construction work will consist primarily of assembling these
manufactured components and modules, rather than fabrication from raw materials.
3.24
September 15, 1997
The Buildings Sector
Chapter 3
Figure 3.7 "Best Practice" Home of the Year 2020
Manufactured wall system
Photovoltaic roof with
Home designed and optimized
with integrated superinsulation
with sophisticated but user-friendly
reflective roofing and
and electrochromic windows.
strategic positioning of trees
computer tools.
to reduce cooling costs.
External communications
link for:
-adaptive comfort controls
Fuel cell for power generation
-energy management
and space conditioning:
-utility purchasing
multi-functional and
-telecommuting
integrated appliances;
-telemedicine & shopping.
sensor-controlled ventilation
system with air filtration and
heat exchanger.
"Grey" water irrigation.
"Superwindows"
Heat pump clothes dryer and
(optimized for
horizontal axis clothes washer.
orientation. external
temperature, and
internal needs).
Integrated natural and
electric light system.
Healthful house construction:
Energy storage
-radon-resistant
(underneath garage).
-non-allergenic
-low-emission materials
-recycled materials.
The best practice home will use affordable, modular, and therefore flexible techniques to permit
longer occupancy. Durability and quality of the basic structure will significantly improve over the
year 1997, and adaptive envelopes will provide significant energy advantages. Material
consumption in residential structures will be reduced through the use of recycled materials and
engineering advances in materials, systems, and assemblies which provide stronger, more durable,
lighter, and less expensive structures. HVAC systems will be right-sized and refined to match
reduced cooling and heating loads and improved comfort features of the envelope. Thermal
distribution systems will effectively transport heating and cooling to the conditioned space.
Climate-appropriate advanced ventilation strategies will range from passive ventilation systems
to filtered systems to heat exchange systems.
Thermal mass will be strategically used to improve comfort and efficiency. "Smart" windows will
see widespread use in upscale houses and for specific rooms and orientations in general housing.
When properly linked via controls and sensors to HVAC systems, improved comfort can be provided
with downsized systems.
Widespread use of paneling and shingles with built-in PV arrays, fuel cells, and advanced energy
storage systems will significantly reduce overall building sector non-renewable energy needs and
will either deliver electricity back to the grid or will provide energy for family electric vehicles.
Building-integrated photovoltaics will be widely employed in new home construction, and a strong
retrofit market for PV shingles will have developed as well.
September 15, 1997
3.25
Chapter 3
The Buildings Sector
Advanced high-efficiency lighting systems actively operating with an array of daylighting and
site/task strategies will optimize building luminosity and reduce energy consumption. Appliances,
lighting, and building control systems will all incorporate smart technology to closely match energy
and water supply and ambient conditions with need. The best practice home in 2020 will be low in
volatile organic pollutants due to the use of low-emitting building materials, and will be equipped
with sensor-controlled energy-efficient ventilation and air cleaning to provide good air quality.
Automatic load modulation of heating and cooling systems in response to varying weather,
environment, and occupant demands will be installed in best practice residences. In addition to
improved sensors and controls, zoning and variable loading of the heating and cooling system will be
used.
The home may have a new generation of high-efficiency gas appliances operating much closer to
combustion temperatures, or it may. be equipped with an integrated water heating/space
conditioning electric heat pump system that minimizes waste heat. These multi-functional systems
will focus on occupant thermal comfort rather than conditioning the space.
Distributed water heating capability (i.e., instant heating at the faucet) may provide
supplemental "on-demand" water heating. Water use and energy efficiency will also be enhanced
by improved design and technology for distribution systems In addition, a greywater irrigation
system equipped for sterilization of effluent may reduce the water required for landscaping, gardens,
and lawns in arid or water-constrained regions of the country.
Home computers and sophisticated communication systems will begin to permit the use of the home
as the location of office, secondary school, routine medical treatment, and selected shopping
activities. This will begin to change the "mix" of building types as well as the need to commute to
these activities.
3.4.2.2 "Best Practice" Commercial Buildings in 2020
By the year 2020, "best practice" commercial buildings will have many advanced technologies that
will greatly reduce the cost of their utility requirements. More advanced programming, design,
construction and commissioning processes will enable both reduced first costs and reduced operating
costs. Varying designs will match building systems with the wide range of climate conditions found
in the U.S. Commercial buildings will be designed and constructed to provide indoor environments
that increase the productivity of workers. A collection of alternative technologies and options that
could be cost-effective in the year 2020 - under the high-efficiency/low-carbon scenario - are
illustrated in Figure 3.8. The drawing shows a composite commercial building containing retail,
office, laundry, and dining facilities.
Commercial buildings will continue to look similar to those existing today. The primary change
will be in the "mix" of these facilities as the advances in electronic information dissemination
reduce the need for physical interaction, and therefore the size, of some commercial buildings. Some
"traditional" commercial buildings, involved primarily in the transfer of information and
knowledge (e.g., offices and libraries) will be significantly down-sized as their physical interaction
(people-related) activities are replaced with electronic communication capabilities. Improved
communications (combined with just-in-time inventory control) will also permit the reduction or
elimination of many stock rooms as well as warehousing and distribution facilities. Many
commodities will flow directly from production to end-use.
3.26
September 15, 1997
The Buildings Sector
Chapter 3
Figure 3.8 "Best Practice" Composite Commercial Building of the Year 2020
Energy Storage Flywheel
Spectral Selective
Tracking
(below floor)
Weather Sensors
Reflective Roofing
Daylight
Collector
Hybrid Gas/Electric
Central Electric Light Source
External Communications
(multiple
Link
Space Cooling
Waste Water
locations)
Filtration/Recovery/Reuse
Self-Drying
Heat Pump Clothes Dryers
Roof System
Horizontal Axis Clothes Washers
Communications & Control
Energy Management
Piped Lighting
Utility Purchasing
Central Computing
Fuel Cell for
Fresh Air Distribution
Power Generation
with Local Heating
Space & Water Heating
& Cooling
Absorption Cooling
Dessicant Regeneration
Piped Lighting
Thermochromic
Photovoltaic
Wall Panels
Wall Panels
& Awning
Light Pipe
Advertising
Variable
Transmissivity
Shade Trees
Windows
Reflective
Pavement
Photovoltaic
"Grey" Water Irrigation
Lamp Posts
Porous Pavement
Turf Paving
The state-of-the-practice commercial buildings will rely heavily on manufactured components for
their construction. One-of-a-kind structures may continue to have many site-built components, but
construction of commercial buildings of standard design (e.g., franchise restaurants and retail stores),
will primarily involve assembly of manufactured components or installation of complete modules.
To make school buildings more affordable to build and operate, such modular construction of schools
may also become commonplace. Quality and material improvements, that cannot be afforded on a
one-of-a-kind basis, will be assimilated into the high-volume manufacturing process.
Low-emissions construction materials and furnishings will be used in the building to reduce the
energy used for ventilation as well as adverse health effects in occupants. Ventilation air will be
filtered to remove infectious agents and allergens that cause illness in workers and lost productivity,
and the use of recirculated air will be minimized. Individual controls will enable workers to adjust
lighting to the most comfortable intensity for their work and for reduced glare. Daylighting will be
more widely used to enhance worker satisfaction and comfort. "Best practice" commercial buildings
will deal effectively with issues of moisture, thermal bridges, thermal distribution, air
infiltration, and air quality.
By the year 2020, "best practice" buildings will also be delivering major performance improvements
through the use of an integrated systems-oriented and optimizing design process. The energy
performance improvements from an increased emphasis on design and commissioning will be
accompanied by improved building energy services and lower overall first costs.
September 15, 1997
3.27
Chapter 3
The Buildings Sector
Improved information about building performance will allow informed design. Right-sizing and
modular staged-operation designs with flexible uses and good part-load operating characteristics
will reduce peak electrical demands as well as overall energy use. Information management systems
for tracking equipment performance and status will ensure persistence of savings from energy-
efficiency measures throughout the building life-cycle.
Larger commercial buildings will have many space conditioning equipment choices, including hybrid
gas/electric space cooling systems and fuel cells for power generation, space and water heating,
absorption cooling, and desiccant regeneration. Chlorofluorocarbon refrigerants will be completely
removed from the buildings sector by 2020 and hydrogenated chlorofluorocarbons will be found only
in older equipment.
The "best practice" commercial building will have highly-efficient centralized electric light
sources combined with tracking daylight collectors connected to "piped" light distribution systems.
In addition, natural lighting through windows and skylights will illuminate interior spaces during
daytime hours.
Most new and existing buildings will use smart control technologies to optimize the building load
configuration in response to weather, occupant demands, and utility rate structures. Natural
conditions and building supply systems will be automatically balanced to adjust for predicted
weather and occupant use. In order to permit greater use of the external environment to improve
internal comfort conditions and reduce energy use, load control will also regulate the variable R-
value wall panels and variable transmittance fenestration. Photovoltaic roofing shingles, wall
panels, and awnings will contribute to the power requirements of state-of-the-practice commercial
buildings.
The widespread use of "cool community" principles will mitigate the impact of urban heat island
effects on major new developments and communities. In addition to reflective roofing and pavement,
this may include using porous pavement, interspersing grass with concrete in lightly used parking
areas, and installing grey water irrigation systems.
3.5 IMPROVEMENTS TO THIS ANALYSIS
There are a few areas where additional work could improve the accuracy of the calculations
described in sections 3.2 and 3.3 above.
Ducts in residential buildings typically leak 15-30% of the air passing through them. In
addition, many of these ducts are inadequately insulated. The end result is that significant
amounts of heating and cooling energy are wasted, particularly when ducts are in unconditioned
spaces. A few relatively inexpensive measures (particularly the aerosol duct sealing
technology) can reduce duct air and heat leakage significantly, even in existing buildings
(Modera et al. 1996). Such measures are not included in the savings estimates for space
conditioning equipment discussed above, and it is likely that an additional 0.5 to 1 quad of
primary energy savings could be achieved by 2010 by widespread implementation in the
residential sector.
The savings estimates for commercial water heating and cooking, as well as for miscellaneous
natural gas use, could be refined significantly. The data available on these end-uses are sparse.
3.28
September 15, 1997
The Buildings Sector
Chapter 3
No savings have been estimated for commercial office equipment, but opportunities may arise to
use voluntary programs (such as the highly successful ENERGY STAR office equipment
program) to promote efficiency as this end-use evolves over the next decade.
No savings have been included for commercial building shell measures. Windows strongly
influence heating, cooling, and lighting loads in all commercial buildings, and insulation can be
important for smaller commercial buildings.
No savings have been included for ground source heat pumps in residential and small
commercial buildings.
No savings have been included for the advanced heat exchanger technology currently being
commercialized by Modine, which reduces air conditioner and heat pump energy use by 15-20%
and reduces the cost of the heat exchanger.
No savings have been included for integrated systems that combine heating and water heating,
or heating, cooling, and water heating.
No savings have been included for district heating and cooling systems with combined heat and
power.
More data are needed on the effects of large-scale tree planting on energy use, and this policy
option needs to be incorporated into the estimates of potential 2010 impacts.
No credits have been calculated for downsizing of HVAC equipment associated with more
efficient building shells.
No attempt has been made to correct for changes in internal gains associated with energy
savings for appliances located within conditioned spaces. Recent work in U.S. commercial
buildings indicates that the heating penalties roughly offset the cooling benefits in both
primary energy and dollar terms when averaged across the entire commercial sector). There is
no comparable analysis for average residences in the U.S., but an analysis for Europe (Krause et
al. 1995) finds that this effect leads to small net energy penalties in residences.
Because energy savings from miscellaneous electricity use are so important to the results of the
buildings sector, it is crucial that more research be carried out, both to characterize how energy
is used in the miscellaneous category and to identify technologies for improving the efficiency
of sub-categories within the miscellaneous category of electricity use.
On balance, we believe that adding these items to the analysis would increase savings and decrease
costs.
3.6 SUMMARY AND CONCLUSIONS
Our analysis leads to the following key results for 2010:
The "efficiency" scenario results in 1.9 quads (5.3%) less energy use and 25 MtC (4.4%) fewer
carbon emissions than the "business-as-usual" scenario in 2010. This represents a savings of $18
billion in fuel costs in 2010, which is purchased with an annualized incremental cost of $7
billion in efficiency improvements.
September 15, 1997
3.29
Chapter 3
The Buildings Sector
The "high-efficiency/low-carbon" scenario results in 4.3 quads (12%) less energy use and 60 MtC
(11%) fewer carbon emissions than the "business-as-usual" scenario in 2010. This represents a
savings of $33 billion in fuel costs in 2010 resulting from an annualized incremental expenditure
of $13 billion on efficiency improvements.
In the residential sector, the greatest energy and carbon savings are achieved in miscellaneous
electricity, lighting, space conditioning, and water heating. In the commercial sector, the
greatest energy and carbon savings are achieved in miscellaneous electricity, space
conditioning, and lighting.
For both residential and commercial buildings, about 90% of the primary energy saved is
electricity in both the "efficiency" and the-"high-efficiency/low-carbon' scenarios.
The time frame of the study (13 years) limits the penetration of efficiency technologies, because
we only consider efficiency upgrades at the time of equipment retirement (no early retirements).
About one-fifth of buildings sector primary energy consumption is not affected in our efficiency
scenarios because the lifetimes of certain types of equipment are comparable to or longer than
the analysis period (see Table C-2.11 in Appendix C-2). Savings from this "untouched energy"
would eventually be achieved in our efficiency and high-efficiency cases, but only after 2010.
Six R&D areas offer great promise to reduce significantly the energy requirements in U.S. buildings
in 2020:
Advanced construction methods and materials will provide increased efficiency and improved
building energy services, often with lower overall first costs. Construction methods in this time
frame will consist primarily of factory-manufactured modules and components assembled on-
site, enabling systems engineering to deliver greater energy efficiency, more affordable
construction, and increased use of recycled materials. Building information management
systems will improve life-cycle performance including feedback for continuous improvement in
design.
Environmental integration will produce buildings matched to the wide range of climatic
conditions, and adaptive envelopes will capitalize on changing outdoor conditions to reduce
energy use and improve occupant comfort and productivity. In addition, environmental
integration strategies such as reflective roofing materials and turf paving will reduce urban
heat island effects.
Multi-functional equipment and integrated systems design offer the opportunity for a quantum
leap in efficiency improvements. For example, combining the functions of several appliances
into a single, highly effective device that puts to use waste heat and employs high-efficiency
components to perform dual functions. Also, the use of integrated systems-oriented design and
commissioning processes will provide efficiency improvements along with improved energy
services and reduced first costs.
Advanced lighting systems in 2020 will include a range of improved technologies such as
improved controls; more high-efficiency small sources matched to improved luminaires;
daylighting systems; and centralized sources with advanced distribution systems. Appropriate
combinations of such systems will have the potential to employ highly efficient artificial
light sources in combination with tracking sunlight concentrators, light pipes, and daylighting
to meet the occupants' precise functional needs for lighting with an order-of-magnitude
reduction in energy use.
3.30
September 15, 1997
The Buildings Sector
Chapter 3
Controls, communications, and measurement capabilities will enable greatly reduced energy
requirements by matching current and predicted weather conditions, utility rates, and internal
environmental measurements to meet fluctuating occupant requirements while expending less
energy.
Finally, self-powered buildings will have fuel cells or small turbines, PV building components,
and energy storage devices to provide building owners with new levels of flexibility in meeting
their energy needs and generating revenues from electricity sales.
Achieving this promise will require significant R&D expenditures over the next twenty years, but
will yield benefits that more than offset these expenditures.
3.7 REFERENCES
American Home Appliance Manufacturers. AHAM Fact Book, 1996, Chicago, IL.
Andrews, Steve. 1994. "Perry Bigelow: Energy Efficiency Maestro". Home Energy. March/April. P.
13.
Brown, R. E. 1994. Estimates of the Achievable Potential for Electricity Efficiency Improvements in
U.S. Residences. In Proceedings of the 1994 ACEEE Summer Study on Energy Efficiency in Buildings.
Washington, DC: American Council for an Energy-Efficient Economy. Vol. 7, page 55.
Brown, R. E. 1993. Estimates of the Achievable Potential for Electricity Efficiency in U.S.
Residences. M.S. Thesis, Energy and Resources Group, University of California, Berkeley.
Christian, J. E. 1991. "Thermal Mass Credits Relating to Building Envelope Energy Standards,"
ASHRAE Transactions, Vol. 91.
Darrow, C. 1994. "A House of the Future," The International Panel and Engineered-Wood
Technology Exposition Conference Proceedings, Atlanta, Georgia, October 5-7.
Energy Information Administration (EIA). 1996. Annual Energy Outlook 1997, with Projections to
2015. Washington, DC: U.S. Department of Energy, DOE/EIA-0383(97), December.
Energy Information Administration (EIA). 1995. Emissions of Greenhouse Gases in the United States,
1987-1994. Washington, DC: DOE Energy Information Administration, DOE/EIA-0573(87-94).
Fine, H. A. and D. L. McElroy. 1989. "Assessment of the Energy Conserving Potential of Active
(Variable Thermal Resistance and Switchable Emittance) Building Thermal Envelope Systems,"
Thermal Performance of the Exterior Envelopes of Buildings IV, Dec. 4-7, 1989. ASHRAE.
Hanford, J.W., J.G. Koomey, L.E. Stewart, M.E. Lecar, R.E. Brown, F.X. Johnson, R.J. Hwang, and L.K.
Price. 1994. Baseline Data for the Residential Sector and Development of a Residential Forecasting
Database. Berkeley, CA: Lawrence Berkeley National Laboratory, LBNL-33717.
Houghton, David J., Robert C. Bishop, Amory B. Lovins, Bristol L. Stickney, James J. Newcomb,
Michael Shepard, and Bradley J. Davids. 1992. The State of the Art: Space Cooling and Air
Handling. Competitek. August.
Kammerud, R. C. 1995. Energy Use Impacts of Advanced Envelope Materials: Assessment of
Mechanisms for Altering Radiation Absorptivity. Oak Ridge, Tennessee: Oak Ridge National
Laboratory, ORNL/Sub/86X-SP110V
September 15, 1997
3.31
Chapter 3
The Buildings Sector
Komor, P. 1996. Cooling Demands from Office Equipment and Other Plug Loads. E-Source Tech
Update TU-96-9. July 1996. Boulder CO. ESource.
Konopacki, Steven, Hashem Akbari, Mel Pomerantz, S. Gabersek, and Lisa Gartland. 1997. Cooling
Energy Savings Potential of Light-Colored Roofs for Residential and Commercial Buildings in 11
U.S. Metropolitan Areas. Ernest Orlando Lawrence Berkeley National Laboratory. LBNL-39433.
May.
Koomey, J.G., C. Atkinson, A. Meier, J.E. McMahon, S. Boghosian, B. Atkinson, I. Turiel, M.D.
Levine, B. Nordman, and P. Chan. 1991. The Potential for Electricity Efficiency Improvements in
the U.S. Residential Sector. Berkeley, CA: Lawrence Berkeley National Laboratory, LBNL-30477.
Koomey, J. G., C. Dunham, and J. Lutz. 1994. The Effect of Efficiency Standards on Water Use and
Water Heating Energy Use in the U.S.: A Detailed End-Use Treatment. Berkeley, CA: Lawrence
Berkeley National Laboratory, LBNL-35475.
Koomey, J.G., F.X. Johnson, J. Schuman, E. Franconi, S. Greenberg, J.D. Lutz, B.T. Griffith, D.
Arasteh, C. Atkinson, K. Heinemeier, Y.J. Huang, L. Price, G. Rosenquist, F.M. Rubinstein, S.
Selkowitz, H. Taha, and I. Turiel. 1994. Buildings Sector Demand-Side Efficiency Technology
Summaries. Berkeley, CA: Lawrence Berkeley National Laboratory, LBNL-33887.
Koomey, J. G., D. A. Vorsatz, R. E. Brown, and C.S. Atkinson. 1997. Updated Potential for Electricity
Efficiency Improvements in the U.S. Residential Sector. Berkeley, CA: Lawrence Berkeley
Laboratory, LBNL-38894 (DRAFT).
Koomey, J. G., and M. C. Sanchez. 1997. Trends in Carbon Emissions for Buildings. Submitted to
Energy Policy in May, 1997.
Koomey, J. G., M. C. Sanchez, D. A. Vorsatz, R. E. Brown, and C.S. Atkinson. 1997. The Potential for
Natural Gas Efficiency Improvements in the U.S. Residential Sector. Berkeley, CA: Lawrence
Berkeley Laboratory, LBNL-38893 (DRAFT).
Krause, Florentin, Eric Haites, Richard Howarth, and Jonathan Koomey. 1993. Energy Policy in the
Greenhouse. Volume II, Part 1. Cutting Carbon Emissions-Burden or Benefit?: The Economics of
Energy-Tax and Non-Price Policies. El Cerrito, CA: International Project for Sustainable Energy
Paths.
Krause, Florentin, David Olivier, and Jonathan Koomey. 1995. Energy Policy in the Greenhouse.
Volume II, Part 3B. Negawatt Power: The Cost and Potential of Low-Carbon Resource Options in
Western Europe. El Cerrito, CA: International Project for Sustainable Energy Paths.
Lawrence Berkeley National Laboratory. 1996. Draft Report on Potential Impact of Alternative
Efficiency Levels for Residential Cooking Products. Berkeley, CA: Lawrence Berkeley National
Laboratory.
Lawson, C. 1996. The Price of Commissioning Equals Cost Savings. ASHRAE Journal 38:1:30-34.
Atlanta GA. ASHRAE.
Lovins, A. 1992. Energy-Efficient Buildings: Institutional Barriers and Opportunities. E-Source
Strategic Issues Paper. SIP1. December 1992. Boulder CO. ESource.
Mills, Evan, and Arthur H. Rosenfeld. 1994. "Non-Energy Benefits as a Motivation for Making
Energy-Efficiency Improvements". Presented at the American Council for an Energy-Efficient
Economy (ACEEE) 1994 Summer Study on Energy Efficiency in Buildings in Asilomar, CA.
3.32
September 15, 1997
The Buildings Sector
Chapter 3
Modera, M.P., D.J. Dickerhoff, O. Nilssen, H. Duquette, and J. Geyselaers. 1996. "Residential Field
Testing of an Aerosol-Based Technology for Sealing Ductwork". Presented at the ACEEE 1996
Summer Study on Energy Efficiency in Buildings in Asilomar, CA. Published by American Council for
an Energy Efficiency Economy (also LBNL-38554).
Rosenfeld, Arthur H., Joseph J. Romm, Hashem Akbari, Melvin Pomerantz, and Haider Taha. 1996.
"Policies to Reduce Heat Islands: Magnitudes of Benefits and Incentives to Achieve Them".
Presented at the ACEEE 1996 Summer Study on Energy Efficiency in Buildings in Asilomar, CA.
Published by American Council for an Energy Efficiency Economy.
Rosenfeld, A. H., J.J. Romm, H. Akbari, and A. C. Lloyd. 1997. "Painting the Town," Technology
Review, February.
Sanchez, Marla. 1997. Miscellaneous Electricity Use in U.S. Residences. Master's Thesis (MS).
University of California, Berkeley, Energy and Resources Group.
Schwolsky, Rick. 1997. "Positive Tradeoffs". Builder Magazine. January. P. 332.
Sezgen, A. O., E. M. Franconi, J. G. Koomey, S. E. Greenberg, and A. Afzal. 1995. Technology Data
Characterizing Space Conditioning in Commercial Buildings: Application to End-Use Forecasting
with COMMEND 4.0. Berkeley, CA: Lawrence Berkeley Laboratory, LBL-37065.
Sezgen, A. O. and J. G. Koomey. 1995. Technology Data Characterizing Refrigeration in Commercial
Buildings: Application to End-Use Forecasting with COMMEND 4.0. Berkeley, CA: Lawrence
Berkeley Laboratory, LBL-37397.
Shepard, et al. 1995. Commercial Space Cooling and Air Handling Technology Atlas. June 1995.
Boulder CO. ESource.
Smith, J. A. and J. D. Rivera. 1989. "Beyond the Year 2000: A Broad Analysis of the Future Trends
that will Affect Buildings," Buildings of the 21st Century: Developing Innovative Research
Agendas, Gersau, Switzerland, May 16-18.
Todesco, G. 1996. Super-Efficient Buildings: How Low Can You Go? ASHRAE Journal. 38:12:35-40.
Atlanta GA. ASHRAE.
U.S. Department of Commerce. 1996. Statistical Abstract of the US. Washington, DC: Economics and
Statistics Administration, Bureau of the Census.
U.S. Department of Energy (DOE). 1990. Technical Support Document: Energy Conservation
Standards for Consumer Products: Dishwashers, Clothes Washers, and Clothes Dryers. Washington,
DC: DOE, DOE/CE-0299P.
U.S. Department of Energy (DOE). 1993. Technical Support Document: Energy Efficiency Standards
for Consumer Products: Room Air Conditioners, Water Heaters, Direct Heating Equipment, Mobil
Home Furnaces, Kitchen Ranges and Ovens, Pool Heaters, Fluorescent Lamp Ballasts, and Television
Sets. Washington, DC.: DOE, DOE/EE-0009.
U.S. Department of Energy (DOE). 1995. Technical Support Document: Energy Efficiency Standards
for Consumer Products: Refrigerators, Refrigerator-Freezers, & Freezers. Washington, DC: DOE.
Vorsatz, D. A. and J. G. Koomey. 1997. The Potential for Efficiency Improvements in the U.S.
Commercial Lighting Sector. Berkeley, CA: Lawrence Berkeley Laboratory, LBNL-38895 (DRAFT).
September 15, 1997
3.33
Chapter 3
The Buildings Sector
Webber, C. 1997. Personal Communication regarding LBNL Technical Analysis of Reduction of
Standby Electricity Use in Televisions and Video Cassette Recorders. Berkeley, CA: Lawrence
Berkeley National Laboratory.
Wendt, R. L.. 1994. "Housing 2020: The Future of Wood in U.S. Dwellings," The International Panel
and Engineered-Wood Technology Exposition Conference Proceedings, Atlanta, Georgia, October 5-7.
Westphalen, D., R. A. Zogg, A. F. Varone, and M. A. Foran. 1996. Energy Savings Potential for
Commercial Refrigeration Equipment. Prepared by Arthur D. Little, Inc. for the Building Equipment
Division, Office of Building Technologies, US Department of Energy. ADL Reference No. 46230.
ENDNOTES
1
A "cost-effective technology" in our analysis is generally defined as a technology that is the
minimum life-cycle cost option using a 7% real discount rate and the lifetime of the option. Life-
cycle cost is the discounted sum of incremental capital costs and operating costs over the life of the
option. This criterion is the equivalent of the cost of conserved energy equaling the value of
displaced or saved energy.
2
To determine which measures are less expensive than the average price of purchased fuel or
electricity and hence cost-effective, we calculate cost of conserved energy (CCE) using the following
equation:
d
Capital
Cost
%
CCE
(c/kWh)
=
Annual Energy Savings
where d is the discount rate and n is the lifetime of the conservation measure. The numerator in the
right hand side of the equation is the annualized cost of the conservation investment. Dividing
annualized cost by annual energy savings yields the CCE.
3 Carbon emissions are derived from the product of end-use energy (by fuel) and carbon emissions
factors of MtC/quad of primary energy taken from ELA (1996). The total cost of energy services is the
estimated amount spent on energy consumption plus the incremental efficiency cost for purchasing
and operating high-efficiency technologies. In the business-as-usual scenario, the incremental
efficiency cost is defined to be zero.
4 Miscellaneous energy use involves end-uses in buildings that are not currently allocated to other
end-uses, namely refrigeration and freezing, space conditioning, lighting, cooking, drying, and water
heating. In order to more accurately estimate energy savings potential, we divided the
miscellaneous end-use into three electricity categories and two fuel categories. The three electricity
categories were: electronics (e.g., color televisions and video cassette recorders), motors (e.g., fans
and pumps), and heating (e.g., waterbed heaters, coffee makers, etc.). About 20% of miscellaneous
electricity is associated with standby losses of equipment that are turned off but still draw a small
amount of power (the so-called "leaking" component of miscellaneous). See Sanchez (1997) for more
details.
5 The scale for 2010 carbon emissions for electricity end-uses in Figures 3.2 and 3.3 is slightly
different than shown for 1997, since a 2.5% decline in the carbon intensity of electricity generation is
projected for 2010, but this does not significantly change the results shown in the figures. For
example, residential miscellaneous electricity carbon emissions in 2010 are 92 MtC but appear
slightly greater (-94 MtC) in Figure 3.2.
3.34
September 15, 1997
The Buildings Sector
Chapter 3
Major contributions to this section were made by George Courville, Mike MacDonald, Jeff Muhs,
John Tomlinson, Jim VanCoevering, and Bob Wendt (Oak Ridge National Laboratory).
=
With thermal switching, the absorptivity and emissivity change between a high and a low value
at a set material temperature; with short wavelength switching, the solar absorptivity changes at
a specific wavelength radiation flux; and with long wavelength switching, the emissivity changes
when the temperature of the radiative environment satisfies certain conditions.
8
Heat pump water heaters are an exception to this general pattern. They have been demonstrated
in the field to deliver up to three times as much energy in hot water as is provided to the unit in
electricity; however, the technology's relatively high cost is a major market barrier. Technology
breakthroughs could result in significant reductions in first costs, enabling greater market
penetration of heat pump water heaters.
September 15, 1997
3.35
The Industrial Sector
Chapter 4
Chapter 4
THE INDUSTRIAL SECTOR
4.1
INTRODUCTION
This chapter presents an assessment of the possible contribution that an invigorated effort to move
energy-efficient technologies that are commercially available, or near commercialization, into the
market could make to reducing greenhouse gas emissions from the US. industrial sector by 2010. We
begin with some background information on our approach to the assessment and how that approach
is shaped by the complexities of the U.S. industrial sector and the available analytical tools for this
sector. We then describe the results of our model-based scenario analysis for the year 2010. In
subsequent sections we provide examples of the types of technologies that need to come into
widespread use to achieve the scenario results. Widespread adoption of these technologies requires
appropriate policies (e.g., accelerated research and development (R&D), fiscal incentives, and market
conditions). Finally, we describe qualitatively, and illustrate with examples, the role of R&D in
providing a steady stream of advanced technologies that can continue to reduce industrial energy
intensity and greenhouse gas emissions, into the foreseeable future. Details of the models used in the
analysis and the technologies described in this chapter are provided in appendices.
4.1.1 Approach
The industrial sector is extraordinarily complex and heterogeneous. By definition, it includes all
manufacturing, as well as agriculture, mining, and construction activities. The manufacturing
industries range from those that transform raw materials into more refined forms (e.g., the primary
metals and petroleum refining industries) to those that produce highly finished products (e.g., the
food processing, pharmaceuticals, and electronics industries). Hundreds of different processes are
used to produce thousands of different products. The U.S. chemical industry alone produces more
than 70,000 different products at over 12,000 plants. Even within a manufacturing industry, individual
firms vary greatly in the outputs they produce and how they produce them. Further, two plants
producing identical outputs can use different processes, and two plants using identical processes can
use different vintages and types of equipment. In some industries, plants employing the same basic
processes can produce a different mix of outputs.
This complexity makes it difficult to conduct this assessment in a "bottom-up" fashion.¹ The available
time and resources do not allow us to (1) catalog all of the advanced technologies whose use might be
increased under an invigorated effort to move them into the market, (2) identify all the processes in
which these technologies might be used, (3) estimate the fraction of the plants that are not already
using these technologies, and (4) determine which of these plants would be likely to choose to invest
in them under the invigorated effort noted above. Instead, we rely on publicly-available computer-
based models to develop rough estimates of the potential for increased investment in energy efficiency
more generally, and then supplement these estimates with examples of technologies, the adoption of
which would achieve the model results under an invigorated effort to move them into the market.
4.1.1.1 Scenario Analysis
For the scenario portion of the analysis, the ideal analytical tool would be an industrial model that is
publicly-available, complete and up-to-date, and has a stock-adjustment mechanism as well as
detailed, technology-specific conservation supply curves for all important industrial processes that are
affected by energy prices, capital recovery rates, and other economic parameters. We would also like
September 18, 1997
4.1
Chapter 4
The Industrial Sector
to be able to relate the modeling results to those reported in the US. Department of Energy's Annual
Energy Outlook 1997 (AEO97), which is prepared by the Energy Information Administration using the
National Energy Modeling System (NEMS) (ELA 1997b).
No existing modeling tool has all of these features. Instead, we employ two modeling tools that, when
used together, provide us with the features we need: (1) the Long-Term Industrial Energy Forecasting
(LIEF) model, which provides a mechanism for evaluating general investment in conservation
technology as a function of energy prices, capital recovery rates, and other parameters, and (2) the
NEMS Industrial Module (NEMS-IM), which captures the effects on energy intensity of groups of
specific technologies, but does not model investment in these technologies as functions of energy
prices or any other factors. (See Appendix D-1 for a description of these two models and the industry
disaggregation scheme used in each.)
We used these two models to develop three scenarios: a "business-as-usual" (BAU) case, an "efficiency"
(EFF) case, and a "high-efficiency/low-carbon" (HE/LC) case. These cases are defined, and their
results described, in Section 4.2. Our general approach was to use the AEO97 reference case
(developed using the NEMS model) as our BAU case. Using the macroeconomic and energy price
assumptions in the AEO97 reference case, we adjusted the LIEF model's base case slightly to more
closely approximate the overall energy forecast in the AEO97.2 We then ran the adjusted LIEF model
to obtain an efficiency and high-efficiency/low-carbon case. We computed the difference between the
LIEF BAU case and the LIEF efficiency case ("delta one"), and between the LIEF BAU case and the
LIEF HE/LC case ("delta two"). We applied the LIEF model "deltas" to the NEMS (AEO97 base)
results to compute our final estimates for potential greenhouse gas emissions reductions. We also used
the NEMS model to explore the extent to which capital stock turnover and technology performance
would have to increase to correspond to "delta one" and "delta two."
4.1.1.2 Technology Examples
The technology discussion focuses on energy-conserving technologies that, as a result of past R&D, are
currently available for purchase in the market or are highly likely to enter the market within the next
few years. While these technologies are available, they have not necessarily been widely adopted and,
under current circumstances, may not be - thus the need for an accelerated effort to encourage their
adoption and achieve the savings that the models suggest are possible. While there are many reasons
for an invigorated effort to adopt these technologies, some of which we discuss later, we temper our
expectations to be sensitive to the slow turnover of heavy equipment in industry.³ Another timing
issue is that some energy-intensive industries also have "windows of opportunity" during the next
few decades where aging capital equipment must be replaced for environmental or competitive
reasons.
We focus on seven energy-intensive industries that are either modeled in detail by the NEMS and
LIEF models or are the focus of the DOE Office of Industrial Technologies' (OIT) Industries of the
Future process, sometimes referred to as "Vision Industries": forest products,4 glass, iron and steel,
metal casting, aluminum, chemicals, and petroleum refining. These major energy-using sectors
account for about 80% of manufacturing energy use (see Figure 4.1). We also look at cross-cutting
technologies (such as energy-efficient motors) that affect all industries. These energy-intensive
industries are briefly described in the box below.
4.2
September 18, 1997
The Industrial Sector
Chapter 4
Energy-Intensive Industries
Industries are characterized using data collected by the Bureau of the Census from establishments (plants) that
are classified in a particular industry based on the value of the production of the plant and the industry that is
identified as the origin of that product. This classification system, known as the Standard Industrial Classification
(SIC), is being superceded this year by the North American Industry Classification System (NAICS). In addition
to economic information collected by the Census, energy consumption is collected for the Energy Information
Administration in the Manufacturing Energy Consumption Survey (MECS).
According to the 1994 MECS, the most energy-intensive industries were, in descending order, Petroleum and
Coal Products (NAICS 324); Paper and Allied Products (321); Chemicals and Allied Products (325); Primary
Metals (331); and Stone, Clay and Glass Products (327). The range of intensity of these industries is from 44.3 to
13.3 thousand Btu per dollar of output (TBtu/$). A brief description of these five most energy-intensive
industries follows.
Petroleum and Coal Products. The major activity in this industry is converting crude petroleum into the
petroleum products widely used in our economy - gasoline, diesel, fuel oil and lubricants. The process is a
complex one of first separating the crude into different products, then recombining these components into the
desired products. The separation is done through distillation and cracking that requires high temperatures and
pressures, and is affected by the density of the original crude. Environmental considerations have greatly
increased the complexity of this process, as reformulated and oxygenated fuels are increasingly needed to assure
clean air quality. Another factor that has made for increased energy use in this industry is the declining
availability of light crude and the greater processing requirements for heavy crude. Petroleum refining is the
most energy-intensive industry with an intensity of 44.3 TBtu/$.
Paper and Allied Products. This industry converts fiber, usually from wood, into paper, pulp or paperboard, and
then into a variety of products. The process begins with wood, which is first debarked and chipped, then either
mechanically or chemically reduced to a slurry that is bleached, then formed into pulp, paper, or board. Though
paper making is a very energy-intensive process, much of the energy used is derived from the biomass that is the
basic feed stock for the process. The Forest Products Vision process combines this industry with wood products
manufacturing, which includes saw mills, plywood mills and engineered wood products. In 1994, energy
intensity was 18.5 TBtu/$.
Chemical and Allied Products. The major segments of this industry are basic chemicals; resins, synthetic rubber
and manmade fibers; pesticides, fertilizer and other agricultural chemicals; pharmaceuticals and medicines;
paints, coatings, sealants and adhesives; soap, cleaning compounds and toilet preparations; and other chemical
products. Basic chemical production includes petrochemicals, industrial gases, and other inorganic chemicals,
and other organic chemical manufacture. Basic chemical production uses the bulk of the energy required by this
industry and creates the largest volume of products In all of chemical manufacturing, heat and pressure are used
to separate and combine chemical building blocks into saleable products, either for final consumers or to other
manufacturing. In 1994, energy intensity was 16.0.TBtu/$_When only basic chemicals are considered, the
intensity is about twice as high.
Primary Metals. This industry includes the production of iron and steel (a Vision industry), aluminum (another
Vision industry), and a variety of non-ferrous metals - lead, copper and zinc are the most important. The
production of iron and steel falls into three sub-industries. Integrated producers transform mon ore into pig iron,
then convert this to steel. The refined steel is cast or rolled into primary products such as sheet, bars, and billets.
Specialty steel producers convert pig iron or steel into special products such as stainless and other alloy steels. Mini-
mills produce primary steel products from scrap steel, usually in an electric arc furnace. Aluminum producers
convert alumina (aluminum oxide) into aluminum metal using an electrolytic process. The major producers also
convert ore, usually bauxite, into alumina, but that operation falls within the chemical industry classification. The
intensity of this industry in 1994 was 15.3 TBtu/$.
Stone, Clay and Glass Products. "Nonmetallic Mineral Products," under NAICS, includes cement, glass (a Vision
Industry), bricks, lime, and other stone and ceramic products. Pyroprocessing, or the application of heat to assure
a chemical reaction, is required in most of these subindustries, which is what makes them so energy-intensive.
Cement and lime are formed at high temperatures in a kiln; glass is produced by melting silica sand; bricks, china
and pottery are just clav until fired. The intensity of this industry is 13.3 TBtu/$.
September 18, 1997
4.3
Chapter 4
The Industrial Sector
4.1.1.3
A Continuing Stream of New Technologies
We assess qualitatively, again through the use of illustrative examples, the contributions that R&D
might also make to reducing greenhouse gas emissions over a longer time frame. We describe R&D
efforts that can lead to advanced technology offering energy-intensity and greenhouse gas emissions
reductions beyond those described in the efficiency and high-efficiency/low-carbon cases,
accompanied by rough quantitative estimates where possible. In this portion of the discussion, we
again focus on the energy-intensive industries and on cross-cutting technologies. Input for the R&D
assessment was sought from technology experts, particularly the OIT Industry of the Future teams
and their industry and laboratory partners.
It is worthwhile to think of these more advanced technologies as the source of future emissions
reductions if the pipeline of R&D is kept full and productive over the entire time horizon. Technology
that is currently available to contribute to reduced energy use and emissions exists because R&D in the
past is now paying benefits in the form of new technology. If there are to be future benefits, this
pipeline must remain full. R&D focusing on efficiency improvements and carbon emissions
reductions is needed to generate the new technologies of the future.
Figure 4.1 Share of Energy-Intensive Industries in Manufacturing End-Use Energy: 1994
Aluminum
Forest
Steel
1.6%
Products
Metal
9.1%
14.8%
Casting
0.9%
Other
21.4%
Petroleum
Refining
26.4%
Glass
1.1%
Chemicals
24.7%
Total=22.6 Quads (1994),
4.2 ENERGY EFFICIENCY EMISSIONS REDUCTIONS
The LIEF model contains conservation supply curves for various industries that correlate energy
conservation investment as a function of energy prices. These curves have been calibrated to historical
industry data using an implicit Capital Recovery Factor (CRF) of 33%. CRFs and associated discount
rates at this level or higher - representing a requirement that these investments pay back the capital
outlay within a few years - have been found to characterize much of the decision-making in industry
on investments in energy-efficiency technologies and on similar investments. At the same time, firms
have another class of investment decisions - termed "strategic" investments - that are
4.4
September 18, 1997
The Industrial Sector
Chapter 4
characterized by a lower CRF or discount rate
(i.e., the initial investments are allowed to pay
back over a longer period) (see Ross 1990).
One way, then, to simulate an increased
investment in energy-efficient technology is to
Increasing the Use of Advanced, Energy-
postulate a policy or set of policies that would
Efficient Technology in Industry
lead industry to apply something like this
more "strategic" discount rate to energy-
Many aspects of business decision-making
efficiency investments. This effect could be
may slow the adoption of energy-efficient
induced via policies that served to decrease
technology. They include
the first cost of such investments or that
High capital intensity of process industry
resulted in increased annual cost savings.
leading to slow capital stock turnover,
Perceived riskiness of new technology,
Lack of internal funding resulting in less
capital for energy projects,
Another way to simulate such an increase in
Lack of information.
technology investment is to directly increase
the factor that represents the penetration rate
Policies that might reduce these effects are:
of new technologies. The penetration rate
Accelerated depreciation,
parameter in LIEF provides a measure of the
Better demonstration and showcase efforts
rate at which industry adopts conservation
to prove technology reliability,
projects. Firms do not immediately adopt all
Reducing first costs and/or achieving
technologies that meet their criteria for cost-
better performance through aggressive
effectiveness and other factors - delays may
R&D,
represent a lack of capital, other priorities for
Rebates or tax credits,
the use of available capital funds, scheduling
Information programs and energy
concerns, or simply a lack of awareness of the
management services,
technologies. The box to the right discusses
Regulation and efficiency standards,
some of the factors that may affect the
Pricing and fiscal policies,
adoption of new, more energy-efficient
Other economic incentive programs,
technologies and policies that could be used to
The exemplary role of governments.
influence them. An increase in this
penetration rate reflects a higher priority
These policies can be interpreted as changing
placed on energy conservation by industry as
the effective or perceived hurdle rates for
well as better information dissemination (Ross
efficiency investments or increasing the old
et al. 1993).
capital turnover and adoption rates for new
technology.
We have used both of these factors to simulate
the efficiency case and the high-
efficiency/low-carbon case for the industrial
sector. We assume that either the discount rate or the penetration rate is affected in the efficiency case,
and that both may be affected in the high-efficiency/low-carbon case. Further details on how the
models were used to simulate these cases are provided in Appendix D-1.
4.2.1 Business-as-Usual Case
Our business-as-usual (BAU) case is the AEO97 reference case. Under this case, national economic
output, measured by gross domestic product (GDP) is projected to increase by 1.9% annually to the
year 2010. Within this overall growth, the manufacturing sector growth rate is projected at 2.1% per
year, with energy-intensive industries growing at half the rate of non-energy-intensive industries, 1.3
versus 2.6%. The leading growth sectors within manufacturing are projected to be industrial
machinery, electronic equipment, and transportation equipment. Of all the manufacturing subsectors,
September 18, 1997
4.5
Chapter 4
The Industrial Sector
electronic equipment is expected to have the highest growth rate, twice that of the manufacturing
sector as a whole.
Total energy intensity, to the year 2010, is projected to decline by 1.1% per year. Among industry
sectors, the largest declines in total energy intensity are projected for the pulp and paper and glass
industries, with the cement industry third. Electricity intensity is projected to decline by 0.5% overall
but with considerable inter-industry variation. The largest decline, 1.1% in the pulp and paper
industry, contrasts with an increase of the same magnitude in the iron and steel industry. The
distribution of primary energy consumption among end-uses is expected to remain stable, with more
than two-thirds of industrial sector use accounted for by manufacturing heat and power requirements
and the remaining third split about equally among non-manufacturing heat and power applications
and use as process feed-stocks. For manufacturing heat and power, the largest energy-consuming
industries are petroleum refining, chemicals, and pulp and paper production. The long-term trend of
declining energy intensity in manufacturing is expected to continue, representing an 18% decline in
energy intensity between 1995 and 2010. This trend is due to both adoption of energy-efficient
technologies and relatively lower growth rates in the more energy-intensive industries. The effects of
industry mix shifting toward less energy-intensive industries is stronger than the efficient-technology
effect on the overall rate of change in energy intensity.
The AEO97 reference case assumes that there are no changes in federal energy or environmental
policies over the forecast period. To the extent that the NEMS model reflects recent historical trends in
industrial technology R&D performance, availability, and introduction, current and future private and
government R&D funding for new and emerging technologies consistent with recent history
contributes to the reference case decline in energy intensity.
4.2.2 Efficiency and High-Efficiency/Low-Carbon Cases
The industrial sector forecasts for the efficiency and high-efficiency/low-carbon (HE/LC) cases use
the AEO97 energy prices and macroeconomic activity forecasts as a starting point. We assume no
changes in economic activity that might arise from changes in energy markets.⁵ Moreover, we assume
no changes in the energy prices that could occur under conditions of lower energy demand. Energy
markets adjust to changes in demand. This means that reduced demand in the EFF and HE/LC cases
would lead to lower energy prices, thereby reducing incentives for efficiency gains.
The efficiency case assumes that industrial firms apply a "strategic" discount rate (or hurdle rate) to
energy-savings investments. We simulate this effect in LIEF by changing the Capital Recovery Factor
(CRF) from 33% to 15% to reflect the lower hurdle rate. Not all cost-effective technologies are
assumed to instantaneously penetrate the market. The HE/LC case is based on the assumption that
the penetration rate of the technologies that are cost-effective under a CRF of 15% doubles on average.⁶
The LIEF model penetration factor was set initially at 3%, roughly calibrated to the NEMS BAU. The
NEMS model uses rates of capital stock turnover that are similar in magnitude. This implies that, in
the high-efficiency/low-carbon scenario, some acceleration of stock turnover is expected. This could
occur under policy incentives for early retirement or economic incentives attributable to the costs and
performance of new process technology that would make old equipment economically obsolete earlier
than has been the case historically.
Table 4.1 summarizes the results in the energy consumption levels forecast by the AEO97. The overall
change in energy use between 1997 and 2010 is shown for the BAU case in the first two columns for
fossil fuels and electricity use (including system conversion losses). Renewables, feedstocks and non-
energy uses of petroleum (e.g. asphalt, waxes, lubricants, etc.) are also shown, but are unaffected by
the LIEF analysis. The next two columns show the effects of the efficiency case and the HE/LC case,
as forecast by LIEF, on the AEO97 BAU case. Figure 4.2 shows that the HE/LC case approaches zero
4.6
September 18, 1997
The Industrial Sector
Chapter 4
growth with energy use increasing by only 1.4 quads (4%) between 1997 and 2010, in spite of an
output increase of 30% over the period.
Table 4.1 Industrial Energy Use: AEO97 Business-as-Usual Case, and Efficiency and High-
Efficiency/Low-Carbon Forecasts by LIEF (Quads)
AEO97
LIEF
BAU
Efficiency Case
HE/LC Case
1997
2010
2010
2010
Electricity (incl. related losses)
11.3
13.2
12.2 (7.6%)
11.2 (15.2%)
Fossil Fuels
16.0
18.2
17.2 (5.4%)
16.3 (10.4%)
Subtotal
27.3
31.4
29.4 (6.5%)
27.5 (12.5%)
Renewables*
1.8
2.3
2.3
2.3
Petrochemical Feedstocks and non-
energy uses of petroleum
5.3
6.0
6.0
6.0
Total
34.4
39.7
37.6
35.8
Note: Numbers in parentheses represent the percent reduction compared to 2010 BAU case.
Expanded renewable use is considered in Section 4.3.
Figure 4.2 BAU Energy Use and Projected Efficiency Cases in 2010 (quads)*
50.00
Electricity (incl. related losses)
Fossil Fuels
45.00
Renewables
40.00
Feedstocks and non-energy uses of
35.00
30.00
Quads
25.00
20.00
15.00
10.00
5.00
0.00
AEO 1997
AEO 2010
Efficiency 2010
HE/LC 2010
Scenario and Year
Table 4.2 shows the results of these analyses for ten major economic sectors of U.S. industry.⁷ The
results in Table 4.2 are expressed in terms of an additional annual percentage reduction in sectoral
September 18, 1997
4.7
Chapter 4
The Industrial Sector
energy intensity compared with the BAU case. The efficiency case reduces total energy intensity
growth by an additional 0.5% per year. The HE/LC case reduces the growth in energy intensity by
over 1% per year, relative to the BAU case, and reduces the growth in electricity use by more than 1%
annually (1.3%).
Table 4.2 LIEF Results: Change in Energy Intensity**, Annual Average Rate, 1997-2010, Compared
with the Business-as-Usual Case (% change)
Efficiency Case
HE/LC Case
CRF = 15%
CRF = 15%
Penetration = Normal
Penetration = Double
Electric
Fuels
Total
Electric
Fuels
Total
Heavy Manufacturing
-0.36%
-0.28%
-0.31%
-0.70%
-0.60%
-0.63%
Pulp & Paper
-0.35%
-0.28%
-0.31%
-0.72%
-0.60%
-0.64%
Bulk Chemicals
-0.40%
-0.28%
-0.33%
-0.81%
-0.60%
-0.69%
Petroleum
-0.47%
-0.28%
-0.31%
-0.78%
-0.60%
-0.63%
Glass
-0.39%
-0.29%
-0.34%
-0.71%
-0.56%
-0.63%
Cement
-0.28%
-0.27%
-0.27%
-0.65%
-0.65%
-0.65%
Iron & Steel
-0.43%
-0.29%
-0.34%
-0.78%
-0.56%
-0.64%
Aluminum
-0.15%
-0.29%
-0.16%
-0.30%
-0.56%
-0.31%
Other
-0.35%
-0.28%
-0.31%
-0.75%
-0.63%
-0.69%
Light Manufacturing
-0.86%
-0.61%
-0.78%
-1.76%
-1.16%
-1.56%
Non-Manufacturing*
-0.67%
-0.67%
-0.67%
-1.26%
-01.27%
-1.27%
All Industry
-0.64%
-0.43%
-0.52%
-1.28%
-0.84%
-1.04%
"Non-manufacturing includes agriculture, construction, and mining (including energy extractions).
- Excludes renewables, feedstocks and non-energy uses of petroleum.
Table 4.3 translates these changes in energy intensity into percentage changes (reduction) in energy
consumption in 2010, relative to the BAU case. In the HE/LC case, overall energy consumption
decreases by more that 12% in 2010 relative to the BAU case, while the decrease in the Efficiency case
is more than 6%. The results for individual industries vary; the declines in energy intensive industries
are close to the average for all of industry, but non-energy intensive sectors show percentage declines
of about twice that of heavy industry.
That the percentage reduction in energy use is higher in light industry stems from two reasons. The
first is that energy is a very small part of the costs in these sectors so that energy efficiency investment
is often overlooked. The LIEF model represents this by a large difference between the average light
manufacturing plants and the most efficient ones. The high growth sectors in light manufacturing
have relatively larger opportunities to make significant percentage reductions than do their energy
intensive counterparts, who have already done so in response to rising energy prices in the 1970s. In
addition, light industries' energy use is dominated by electricity. Electricity savings in light
manufacturing comes largely from computer controls, motor systems, as well as contributions from
lighting and HVAC that are similar to technologies discussed in the buildings chapter
The second is that the growth in output for light industry is much higher than for heavy industry.
Output grows more than 80% by the year 2010 for light industry, but only 30% for heavy industry. As
a result, in the BAU case, electricity demand nearly doubles for light industry and fossil fuel use
4.8
September 18, 1997
The Industrial Sector
Chapter 4
grows more than 60%. Fossil fuel demand for heavy industry only increased by 12% in the BAU case,
while electricity demand increases by 48%.
The difference between light and heavy manufacturing is a major source of the difference in the
energy savings (on a percentage basis) between fossil fuels and electric energy. One should note that,
while these percentage savings vary, a significant portion of the energy savings in absolute terms still
come from fossil fuel use reduction in heavy industry, e.g. fossil fuel reductions in heavy
manufacturing is about 8% while the industry total for fossil fuels is about 12%.
Table 4.3 LIEF Results: Energy** Savings in the Year 2010 Compared with the Business-as-Usual
Case (% reduction)
Efficiency Case
HE/LC Case
CRF = 15%
CRF = 15%
Penetration = Normal
Penetration = Double
Electric*
Fuels
Total*
Electric*
Fuels
Total*
Heavy Manufacturing
4.6%
3.6%
4.0%
8.7%
7.5%
7.9%
Pulp & Paper
4.5%
3.6%
3.9%
9.0%
7.5%
8.0%
Bulk Chemicals
5.0%
3.6%
4.2%
9.9%
7.5%
8.5%
Petroleum
5.9%
3.6%
3.9%
9.7%
7.5%
7.8%
Glass
5.0%
3.7%
4.3%
8.8%
7.0%
7.8%
Cement
3.5%
3.5%
3.5%
8.2%
8.1%
8.1%
Iron & Steel
5.5%
3.7%
4.4%
9.6%
7.0%
8.0%
Aluminum
2.0%
3.7%
2.0%
3.8%
7.0%
4.0%
Other
4.4%
3.5%
3.9%
9.3%
7.9%
8.5%
Light Manufacturing
10.6%
7.6%
9.6%
20.4%
14.0%
18.3%
Non-Manufacturing*
8.3%
8.3%
8.3%
15.1%
15.2%
15.2%
All Industry
8.0%
5.4%
6.6%
15.30%
10.4%
12.5%
These numbers are based on electricity system-average energy loss from the business-as-usual case.
"Non-manufacturing includes agriculture, construction, and mining (including energy extraction).
- Excludes renewables, feedstocks and non-energy uses of petroleum.
Table 4.4 illustrates how the energy use by fuel type is affected in each scenario. Natural gas use, the
dominant fuel use by industry, declines the most in absolute terms. Petrochemical feed stocks, other
non-energy uses of petroleum, and renewables are assumed to be unaffected in the efficiency and
high-efficiency/low-carbon. cases and do not contribute to the carbon emissions.
Table 4.4 Change in Industrial Energy Use by Fuel Type
AEO
Efficiency
HE/LC
1997
2010
2010
2010
Natural gas (billion cu ft)
9,914
11,103
10,303
9,564
Coal and coke (1000 short tons)
104,716
113,741
10,551
97,976
Liquid fuels - incl. LPG (1000 bbl)
695,160
697,300
647,090
600,648
September 18, 1997
4.9
Chapter 4
The Industrial Sector
Petrochemical feed stocks & other
925,536
1,180,979
1,180,979
1,180,979
petroleum (1000 bbl)
Table 4.5 provides carbon emissions estimates for 2010 in metric tons. Because LIEF does not model
fossil fuel choice, estimates of carbon reductions are based on the fossil fuel mix and emission factors
in NEMS. For fossil fuels, there are two ways to compute carbon emissions. The first is to assume that
efficiency affects fuel reductions through the average fuel mix. The second is to assume that most
energy-efficiency reductions operate on the margin (i.e., they affect those fuels that constitute the
growth in the BAU forecast).
Table 4.5 Carbon Emissions Estimates (MtC per year)
AEO97
Efficiency Case
HE/LC Case
1997
2010
2010
2010
Electricity
172
213
204 (4.5%)
186 (12.7%) *
Fossil Fuels
311
335
317 (5.4%)
300 (10.4%)
Industry Total
482
548
521 (5.1%)
486 (11.3%)
*A portion of the reduction in carbon emissions associated with the high-efficiency/low-carbon case is due to
changes in the electricity generation mix prompted by the charge of $50/tonne of carbon (see Chapter 6).
Numbers in parentheses represent the percent reduction compared to 2010 BAU case.
An examination of the change in fossil fuel mix in industry in the AEO97 found that no fuel's share
changed by more than 1%. Consequently, using the average industrial fossil fuel mix from the AEO97
is a reasonable approach to compute the change in greenhouse gas emissions. However, the electric
utility industry shows an increasing share of natural gas. Therefore the carbon reductions for
electricity use in Table 4.5 are based on the marginal carbon emission rates, rather than the average
(see Chapter 6 for more details).
These overall carbon reductions are translated into industry-specific carbon reductions in Table 4.6.
Heavy manufacturing contributes about one-third of the savings in both the efficiency and HE/LC
cases. The large contribution of carbon savings from light industry comes mostly from electricity
efficiency. Electricity use in this sector is growing rapidly - almost doubling - in the BAU case.
Table 4.6 Industry-Specific Reductions in Carbon Emissions (MtC per year in 2010)
Efficiency
HE/LC
Electric
Fuels
Total
Electric
Fuels
Total
Heavy Manufacturing
2.1
7.1
9.2
5.9
14.8
20.6
Pulp & Paper
0.3
1.1
1.5
1.0
2.3
3.3
Bulk Chemicals
0.7
1.5
2.2
1.9
3.2
5.1
Petroleum
0.3
2.5
2.7
0.6
5.2
5.8
Glass
0.1
0.2
0.2
0.2
0.3
0.5
Cement
0.0
0.2
0.3
0.1
0.5
0.6
Iron & Steel
0.4
1.2
1.6
1.1
2.3
3.4
Aluminum
0.1
0.0
0.2
0.4
0.0
0.4
Other
0.2
0.4
0.6
0.6
0.9
1.5
4.10
September 18, 1997
The Industrial Sector
Chapter 4
Light Manufacturing
6.3
5.2
11.5
17.8
9.7
27.4
Non-Manufacturing*
1.3
5.7
7.0
3.4
10.4
13.8
Total
9.6
18.1
27.7
27.0
34.9
61.9
* Non-manufacturing includes agriculture, construction, and mining (including energy extraction).
4.2.3
Comparison with the NEMS model
The NEMS model provides a different approach and perspective on the EFF and HE/LC cases. The
NEMS model uses a stock turnover approach to project the change in energy use. New technology is
projected to be more efficient; thus, as capital is replaced, the overall energy requirements in the
industry decline. To compare the scenarios, the NEMS industrial model was run under alternative
assumptions and compared to those corresponding industry sectors in LIEF (see Table 4.7).
When the retirement rate of capital is doubled in the NEMS industrial model, the decline in total
energy use ranges from 1-8%, depending on the sector. On the other hand, when the performance of
new technology is assumed to double (i.e., the relative energy intensities of new technologies in NEMS
decline twice as fast as in the BAU case), even larger reductions in energy use are.achieved for all
sectors except cement and steel. These parametric variations in the NEMS model illustrate, in rough
magnitude, what rate of technology improvement or stock turnover would be consistent with the EFF
and HE/LC case. For example, only in the iron and steel industry does the doubling of the retirement
rate result in energy savings comparable to those in the HE/LC case; for all other industries, it would
require more effort than simply doubling the capital stock turnover to achieve comparable savings.
For aluminum and glass, the energy savings resulting from the NEMS run that doubles technology
performance are higher than the energy savings in the HE/LC case, suggesting that for these sectors
more rapid technology development is an important part of future savings. This is particularly true of
the aluminum sector.
Table 4.7 Comparison of Year 2010 Total Energy Savings Relative to BAU in the NEMS and LIEF
Models
LIEF
NEMS
Efficiency Case
HE/LC Case
Doubled
Doubled Technology
Retirement
Performance
Paper
3.9%
8.0%
4.9%
7.5%
Chemicals
4.2%
8.5%
1.3%
5.0%
Glass
4.3%
7.8%
3.6%
9.9%
Cement
3.5%
8.1%
5.7%
3.6%
Iron and Steel
4.4%
8.0%
8.2%
2.9%
Aluminum
2.0%
4.0%
1.2%
7.8%
4.2.4
The Historical Context of Energy Efficiency in Industry
Over time, both the "what" and the "how" of industry output changes. Buggies and whips have
disappeared, but automobile production has taken their place. And while the Model T was mass-
produced, today's methods of production are only vaguely reminiscent of Henry Ford's assembly line.
Energy use in manufacturing and other industry sectors has changed due to both product and process
transformation. Energy use changes occur because of energy-efficiency improvements over time as
well as changes in the mix of industries. Rough approximation of the importance of these two factors
September 18, 1997
4.11
Chapter 4
The Industrial Sector
indicates that efficiency accounts for about two-thirds of the change, while the shift in the mix of
industries accounts for about one-third. Put into historical perspective, forecasts of energy use and
energy intensity changes used for this analysis are modest changes and, we believe, more than just
possibilities. With appropriate and effective policy measures to accelerate the adoption of
technologies that are currently, or will soon be, available, the efficiency gains and energy and carbon
savings projected could easily be achieved.
A study published by DOE (1995) illustrates how rapidly energy intensity in the industrial sector can
decline. Between 1972, the last full year prior to the effect of the first oil price shock, and 1985, when
energy prices fell, the rate of decline in energy intensity in industry was 2.74% per year. During the
period of the most rapid decline, from 1975 to 1983, industrial sector energy intensity fell by 3.12% per
year. These numbers show that, when industry has a major incentive to reduce energy use, it will do
so. By the same token, when the incentives are reduced, so are the improvements. Between 1984 and
1991, energy intensity in the industrial sector declined by less than 1% per year, and in four of these
years, the intensity actually increased.⁸ Of the energy savings that occurred in the industrial sector
between the mid-1970s and the early 1990s, this report suggests that about one-third of the total was
attributable to compositional shifts (i.e., shifts from high energy-intensive industries to industries with
lower energy intensity). The remainder was attributable to reductions in energy intensity within
industries.
In the BAU forecast, total energy intensity declines at about 1.1% per year, with more than half of this
decline (0.6%) attributable to projected composition effects. If one takes the efficiency component of
the total energy intensity decline forecast for the BAU case (0.5% per year) and adds the additional
0.85% per year from the high-efficiency/low-carbon case, the HE/LC case has a rate of energy
intensity decline (1.35%) that is slightly below the historical rate over the period 1972-1991 (1.89%).
4.2.5 The Costs of Achieving the Efficiency and HE/LC Cases
The LIEF model conservation supply curves can be used to compute the investment implied by the
forecast energy reductions. These estimates, shown in Table 4.8, are the additional investment
required to achieve the energy savings presented above. Due to the long-lived nature of industrial
capital goods, this cumulative investment in more efficient and productive industrial plant and
equipment continues to generate energy and costs savings, relative to the base case, after the 2010 time
horizon.
LIEF projects that this level of investment is profitable with the BAU forecast energy prices and a CRF
of 15%. The energy savings provides about a seven-year payback on the initial investment. The
magnitude of the up-front costs, which are paid back only over time, may be an issue in designing
policies to spur this enhanced technology penetration.
To put this level of investment in energy efficiency into context, we compare it to total investment in
manufacturing. If the cumulative investment in energy efficiency is spread out evenly over the 13-
year time period, the HE/LC case would require a $3.6 billion increase in annual investment in
efficiency technology. In 1992, total investment in manufacturing (not including agriculture,
construction, and mining) was $110.1 billion (1995$). Thus, the incremental annual investment needed
to achieve the HE/LC case represents a 3.3% increase over the level of manufacturing investment for
1992.
Table 4.8 Cumulative Incremental Investment (1998-2010) for Energy Efficiency Implied by the
LIEF Model to Achieve the Forecast Energy Reductions (billions of 1995$)
Efficiency Case
HE/LC Case
4.12
September 18, 1997
The Industrial Sector
Chapter 4
Fossil Fuels
7.4
15.2
Electricity
15.8
32.0
Total
23.2
47.2
September 18, 1997
4.13
Chapter 4
The Industrial Sector
Historical behavior with respect to energy efficiency investments has been characterized by implicit
marginal discount rates equivalent to 33% capital recovery. The efficiency case is based on the notion
that the marginal return on energy efficiency will be closer (or equal) to a strategic discount rate,
represented here as a 15% CRF. For example, this translates to a marginal real return of 12.5% per
year on a 15-year investment, which we will use for illustrative purposes. It is from this perspective
that the efficiency case reflects 'cost-effective' investments. The 'last' investment will produce cost
savings that will provide a return of 12.5%; other investments will generate higher returns. On
average, the return will be higher than the marginal, or 'last', energy-efficiency project.
Table 4.9 shows the private investment cost of an investment in efficiency in a single year compared to
the value of the energy savings that would continue to accrue thereafter. The first line in the table is
the incremental investment in the last year of our forecast, 2010. The second and third lines are the
change in consumption and expenditure of energy for that year, which are negative since energy
consumption is reduced. One can see that investments generate annual savings of about a third of the
initial outlay. This is an average return that is quite a bit higher than the assumed marginal return of
12.5%. Recall that the marginal return is the 'last' cost-effective investment, which just pays for itself
at the 12.5% rate.
Table 4.9 also shows the total energy savings and direct private costs of the scenario. These costs are
generated using the cost of conserved energy (CCE) method detailed in Appendix A-1.3. For the
efficiency scenario the energy savings exceed the direct private investment costs by $4 billion. The
HE/LC scenario has energy savings in excess of direct investment costs of $7 billion.
Table 4.9 Net Costs of Private Investment for Energy Savings in the Efficiency and High-
Efficiency/Low-Carbon Cases (millions of 1995$)
Efficiency
HE/LC
Electric
Electric
Units
Fossil Fuels
(End-use)
Fossil Fuels
(End-use)
Investment in 2010
M$
$800
$1,700
$1,500
$3,200
Annual Energy Reduction
TBtu
94
47
178
82
Annual Reduction in Energy Costs
M$
$300
$600
$600
$1,100
Total Energy Redirection
TBtu
900
336
1800
685
Total Investment Cost
M$
$1,100
$1,800
$2,400
$4,100
Note: Costs are based on the annualized costs over the time period, not the cummulative investments.
We believe that most, if not all, of the difference between the observed behavioral CRF of 33% and the
15% CRF is due largely to factors that preclude firms from using these lower marginal rates for
energy-efficiency investments, such as transaction costs, agency costs, the lack of information or the
cost of acquiring it, perceived risk, etc. However, policies will be required to remove these factors and
shift investment behavior to prioritize energy efficiency the same as other corporate investment.
These policies will have a public cost.
The HE/LC case also focuses on 'cost-effective' investment under the same notion of these lowered,
strategic, marginal rates of return. However, one important difference in the HE/LC scenario is that a
higher adoption rate is assumed. While some additional penetration, relative to the BAU, may be
accounted for by further transformation of the market of energy-efficient practices we feel that some
accelerated retirement may also take place. When the economic losses of accelerated retirement are
accounted for, this implies that, at the margin, all investments are not likely to be cost-effective at our
4.14
September 18, 1997
The Industrial Sector
Chapter 4
assumed 15% CRF. Since we do not have a model to account for this potential early retirement and
the economic losses, we must caveat our estimates of investment. The energy savings from the
HE/LC scenario in Table 4.9 does not change, but the investment cost may be understated by the
amount of loss due to any early retirement that may occur. Because the net benefit is still greater than
the annualized investment we calculate, then unaccounted costs may be about twice our estimated
energy-efficiency investment, with the HE/LC scenario remaining 'cost-effective' on average.
A carbon-based fuel price increase was considered and simulated using LIEF for a number of carbon
shadow prices. Energy price increases alone do not have a very dramatic effect on energy use in the
LIEF model. While they do have some affect on the options to reduce energy use, they have no
endogenous affect on the rate of penetration of new technology in the model. For example, a $50
shadow price for carbon increases shifts the "ideal" energy-output ratio by only 8.5% for electricity
and 5% for fossil fuel. The gap between the ideal and actual energy-output ratios is a measure of the
conservation potential for the sector. Under the BAU case, this gap is 3.8% for electricity and 4% for
fossil fuel. Under the EFF case, this gap is 27.6% for electricity and 15.3% for fossil fuels. Under the
$50 shadow price case, the gap is 9.5% for electricity and 7% for fossil fuels. To achieve the same ideal
energy-output ratio as the HE/LC case would require a shadow price of $250 for fossil fuels and $300
for electricity. Table 4.10 shows the carbon reduction and the percentage reduction in electricity and
fossil fuels that result from simulation of different carbon shadow prices.
Table 4.10 Effect of Different Carbon Shadow Price Simulations on Electricity and Fossil Fuel
Reductions
Shadow
Electricity
Fossil Fuels
Price of Carbon
% of BAU
Carbon Saved
% of BAU
Carbon Saved
25
98.4
3
99.0
3
50
97.1
6
98.2
6
100
95.1
10
96.9
10
200
92.2
16
94.8
17
300
90.1
20
93.3
22
400
88.6
23
92.3
26
The HE/LC case reduces electricity to 85.2% of the BAU case and fossil fuel to 92.5%. The
energy/carbon savings in the table would be larger if these higher prices systematically affect the
penetration rates of new technology, which one would expect. However, penetration rates are
currently parametric in LIEF, and since we have very little information about how price changes affect
penetration rates, we have not altered that parameter for this exercise. Given the belief that the rise in
prices would increase penetration, the estimates of energy and carbon savings from LIEF would
represent an upper bound on the required carbon tax or a lower bound on the savings.
The implications of the 'standalone' analysis of carbon shadow prices is that a variety of polices well
beyond a carbon permit charge would be required to achieve the savings projected in these scenarios.
4.3 ADDITIONAL EMISSIONS REDUCTIONS FROM INDUSTRIAL LOW-CARBON
TECHNOLOGIES
4.3.1 Introduction and Summary
Industrial low-carbon technologies reduce greenhouse gas emissions through means other than
traditional energy efficiency. We separate low-carbon technologies into three types:
September 18, 1997
4.15
Chapter 4
The Industrial Sector
Power-system efficiency maximization (PSEM) technologies: such technology systems comprise
mainly existing technologies assembled in an innovative way so as to maximize energy efficiency
at certain types of locations for particular industries' heat and power needs.
Fuel-switching technologies: these reduce carbon emissions by using low- or no-carbon fuels
instead of high-carbon fuels. Many energy forecasting models, including LIEF and NEMS,
incorporate switching from oil, coal or electricity to less carbon-intensive gas. They do not,
however, generally incorporate switching to new advanced biomass or other new renewable
technologies. Both of these low-carbon technology types are often grouped with energy-
efficiency technologies. We separate them from efficiency technologies in this chapter because
their additional contributions to carbon reductions are not generally included in traditional
energy models.
Low process carbon technologies: this type reduces or avoids the emission of CO₂ and other
greenhouse gases from industrial processes, not from combustion. They are clearly not included
in energy models. We have found that most of these emissions are non-CO2 greenhouse gases.
Because these emissions do not involve energy, they have not been included in energy-focused
carbon analyses. However, as shown in Section 4.3.4, these non-energy emissions account for a
third of total greenhouse gas equivalent emissions in the industrial sector. (Industrial CO₂
emissions from energy are projected to be 482 MtC equivalent in 1997 (ELA 1996) and non-
energy-related carbon equivalent emissions were 244 MtC equivalent in 1994).
This section provides examples, rather than a comprehensive survey, of low-carbon technologies.
Such a survey would have been difficult because, unlike traditional energy-efficiency technologies,
these technologies do not have a long history of being analyzed from the perspective of reducing
carbon equivalent emissions. However, as shown in Table 4.11, just these examples showed great
potential reductions. Thus, a comprehensive survey of these technologies is an important area for
future analysis in the industrial sector. Note that the carbon reductions presented are in addition to
the carbon savings of Section 4.2.2. Some of these technologies also feature carbon reductions due to
traditional energy efficiency. We used the energy-efficiency projections for the various traditional
markets presented in Section 4.2.2 to subtract these carbon savings from the technologies' estimated
overall carbon reduction. Greenhouse reductions from "low process carbon" technologies are not
included in this report's summary tally of carbon reduction potential because of the report's focus on
combustion-related emissions.
In the following sections, we provide examples of the three types of low-carbon industrial
technologies. The Advanced Turbine System (ATS) described in Section 4.3.2 is an example of a PSEM
technology. It is a combined heat and power (CHP) system that replaces grid electricity and steam
from industrial boilers with a highly efficient on-site natural gas-fired turbine that generates both
electricity and steam. The carbon reductions from on-site CHP were not included in Section 4.2.2. The
ATS may also further maximize system efficiency by replacing electricity used to drive motors that
drive equipment with direct power for the equipment. Even when used as a power-only technology,
ATS reduces carbon emissions because it is located on-site - avoiding transmission and distribution
(T&D) losses. The ATS is also a fuel-switching technology if it replaces high-carbon fuels such as coal
used in the boilers with natural gas or no-carbon biomass gas.
Section 4.3.3 gives an example of a fuel-switching technology. Black liquor and biomass gasifiers
integrated with combustion turbines replace biomass boilers and grid electricity. In the near and
medium time frame, biomass and black liquor gasification technologies provide the option of
switching from a high-carbon to a "no-carbon" fuel. Note that the advanced technologies described in
Section 4.3.3 are also PSEM technologies because they replace inefficient biomass boilers and grid
electricity with biomass gasification combined heat and power systems.
4.16
September 18, 1997
The Industrial Sector
Chapter 4
Section 4.3.4 describes two low process carbon technologies. The first, the advanced aluminum
production cell, shows that for some industrial processes there are multiple opportunities for reducing
carbon equivalent emissions. The second involves the substitution of waste products - fly ash and
blast furnace slag - for a portion of the calcined cement clinker intermediate product in cement
production. Both of the examples reduce carbon through improved energy efficiency in addition to
reducing or eliminating carbon equivalent process emissions.
A summary of the carbon reductions from these technologies is given in Table 4.11 for both the
efficiency and the high-efficiency/low-carbon (HE/LC) cases.
Table 4.11 Examples of Additional Carbon Equivalent Reductions by 2010 Resulting From Low-
Carbon Technologies* (MtC equivalent)
Efficiency
High-Efficiency/Low-Carbon
Case
Case
Power System Efficiency
Maximization Technology (PSEM)
Advanced Turbine Systems
5-7
14-24
Fuel-Switching Technology
Forest Products - IGCC
5
10
Low Process Carbon Technologies
New Aluminum Production Cell
0-1
2-4
Cement Clinker Replacement
-
1-2
Total
10-13
27-40
"These reductions are not accounted for in Section 4.2.2.
4.3.2 Power System Efficiency Maximization Technologies
Power-system efficiency maximization technologies are grounded in the second law of
thermodynamics. PSEM technologies take advantage of the fact that waste heat is always produced.
Such systems also reduce or avoid extra energy conversion and process steps that waste energy. The
key to PSEM is the system. Instead of using a separate technology for electricity for the company's
PCs, building heating and cooling, process steam and electricity for motors, a company could use a
PSEM technology. For example, the Advanced Turbine System (ATS) described in Section 4.3.2.1,
could provide all these system needs. The ATS could provide reliable high-quality electricity to the
PCs; ATS steam coupled with a heat exchanger could provide building heating and cooling and steam
for process uses; and the turbine could be hooked directly to the drive shaft of the machine formerly
driven by a motor that used grid electricity. District energy sites, where businesses group together
and share electricity and steam from the same turbine, are also examples of PSEM technologies in the
industrial sector. A recent study (IDEA 1997) indicates that, of the nearly 6000 current U.S. district
heating installations generating more than 1.1 quads, 8% are classified as industrial. We expect that
well-crafted policies to increase energy efficiency and reduce carbon will spur creative uses of both
heat and power in such systems. In addition to multiple incremental improvements, we expect that
some PSEM will be breakthrough technologies.
September 18, 1997
4.17
Chapter 4
The Industrial Sector
4.3.2.1 Advanced Turbine Systems (ATS) for Industrial Applications
Advanced Turbine Systems (ATS) are high-efficiency, next-generation gas turbines that produce less
carbon per kWh than technologies used in conventional power markets. When commercialized in the
year 2001, the emissions of CO2 from ATS are projected to be 600 lb/MWh, 29-73% lower than
conventional technologies (see Figure 4.3).9 ATS is one of the major low-carbon technologies for the
industrial sector between now and 2010 because it is a natural gas-fired turbine that cogenerates
electricity and steam. The ATS's high energy efficiency stems from multiple incremental
improvements applied in a novel manner.¹⁰ Cogenerated steam displaces industrial steam boilers and
their associated emissions. The steam can also be put back into the system for additional electricity
generation. Further emissions reductions are due to the ATS being.gas-fired and located on-site.
Although not included here because of possible double counting with Section 4.3.3.1, the ATS
technology is also well suited for biomass and landfill gas fuels. The ability of ATS to burn biomass
without turbine fouling and maintenance problems is being explored via new turbine materials,
including ceramics and single crystal and directionally solidified turbine blades. Substantial
reductions in greenhouse gas emissions will result if ATS is fired with biomass fuel - especially in
combined heat and power mode. It will require the evolution of a biomass fuel supply infrastructure,
or its penetration will be limited to those industries that already have access to biomass fuels, such as
forest products and some food processing sectors. We provide an example of biomass-based
cogeneration in the paper industry in Section 4.3.3.
We divided the ATS "markets" into three types. The first type includes high electricity-to-thermal
(E/T) ratio "power only" opportunities. These are sites where there is little or no steam demand and
most of the steam from the ATS is fed back into electricity generation. The second type, "combined
heat and power" (CHP), includes sites where ATS provides both steam and electricity needed on-site.
The third type is a "new steam" market, where the steam and electricity needs vary.11
This "new steam" market is a new market not included in most energy forecasting models. It is new
CHP capacity in which power and heat are not balanced and where the desire to generate electricity
may be more important than getting the perfect steam match. Unlike traditional cogeneration
equipment that is only efficient at a particular E/T ratio, ATS CHP systems run at high efficiency in a
variety of steam and electricity configurations. As detailed in Appendix D-3, this market will spur
creative uses of both heat and power. For analytic purposes, we have analyzed the "new steam"
market as if it were two separate CHP and power-only markets. We decomposed new steam into
traditional CHP (cogeneration assuming heat/power balance) and Power-Only (PO):
New steam = a*CHP + b*PO
While some sector-specific studies (Appendix D-3) show a and b values around 0.5 for the entire
market, the values of a and b are not well known except that they are both significant. As detailed in
Appendix D-3, this decomposition also simplifies the calculation of the carbon offset. Figure 4.4
depicts simplified diagrams that allow comparison of the following: (1) a traditional steam boiler
system, (2) a steam boiler that produces power using an ATS, (3) an ATS used for combined heat and
power, and (4) an ATS used for power only. There are many other combinations, such as a turbine
with a recuperator not shown here.¹²
4.18
September 18, 1997
The Industrial Sector
Chapter 4
Figure 4.3 Carbon Equivalent Emissions for Several Electric Generation Technologies
(pounds per MWh)
598
506
370
344
337
295
229
164
COAL
OIL
NATURAL GAS
Utility
Indus.
Recip.
Utility
Advanced
Combined
Cogen
Boiler-Steam
Gas
Engine
Boiler-
Indus.
Cycle-Gas
Indus.
Turbine
Turbine
< 1000 hp
Steam
Gas
Turbine
Gas
< 20 MW
Turbine
Turbine
> 200 MW
Turbine
Source: Gas Research Institute (1994) and Onsite Energy (1997)
Considering the large markets not yet served by this type of CHP, industry experts predict that the
availability of advanced turbines will double the growth rate of new CHP capacity (Carroll 1997).
This growth will greatly exceed the historic industrial market penetration of cogeneration,¹³
particularly for smaller power technologies used to meet internal energy requirements. Under the
efficiency or high-efficiency/low-carbon scenarios, the change in the market will occur even faster.
Relatively higher prices for carbon-based fuels will encourage dispatching of electricity from low-
carbon fuels, reform of environmental permitting, and utility regulations and will thus accelerate the
replacement of boilers by on-site ATS cogeneration. The turbine's low installed costs, low NOₓ
emissions, and ability to generate electricity when steam is not needed will also contribute to the rapid
growth of this new steam market.
Table 4.12 shows the contributions of these two "markets" to the total carbon reductions. As described
in Appendix D-3, the power-only carbon reductions are much smaller because we assume that the
power being displaced is also quite efficient.¹⁶ Thus, the ATS only takes credit for carbon reductions
due to avoidance of transmission and distribution losses (7%). In addition, we assume the grid
electricity (see utility chapter for details) and the steam boilers displaced have higher carbon emissions
than those displaced in the efficiency case. For both cases, we subtracted the same traditional
cogeneration that is contained in the NEMS BAU.
September 18, 1997
4.19
Chapter 4
The Industrial Sector
Figure 4.4 Simplified Diagrams of Advanced Turbine Systems in Power-Only and Cogeneration
Mode Compared to Steam Boiler
Steam Boiler
Natural Gas, Coal, Oil, Other
Water
Boiler
Steam
Typical Boiler Cogen System
Natural Gas, Coal, Oil, Other
Water
Boiler
Steam
Generator
Combined Heat and Power
Steam
Steam Turbine
Electricity
Advance Turbine System
Natural Gas
Air
Steam
Heat Recovery
Steam Generator
Combustor
Generator
Combined Heat and Power
Compressor
Turbine
Electricity
Advanced Turbine System
Natural Gas
Water
Heat Recovery
Air
Steam Generator
Combustor
Steam
Generators
Power-Only
Two-Stage
Two-Stage
Steam Turbine
Compressor
Turbine
Electricity
4.20
September 18, 1997
The Industrial Sector
Chapter 4
Table 4.12 Calculation of 2010 ATS Carbon Savings (MtC) and Corresponding ATS Electricity
Generation (TWh)**
Combined Heat
Power Only
Total
and Power*
Efficiency
4-6 (29-59)
1 (120)
5-7 (150-180)
High-Efficiency/Low-carbon
12-21 (60-120)
2 (220)
14-24 (280-340)
Numbers may not add up exactly due to rounding. TWh shown above in parentheses.
-
Excludes carbon reductions and electricity generation from traditional cogeneration that is contained in the
NEMs BAU case as well as forest products biomass cogeneration which is considered in Section 4.3.3. Other ATS
markets where ATS electricity generation did not result in substantial carbon savings were also excluded.
** See Table D.3-4 for details.
We estimate ATS carbon reductions of 5-7 MtC equivalent for the efficiency case (see Appendix D-3).
This corresponds to an electric capacity of 23-27 GW and requires 0.5 TCF of additional natural gas
(5% of 2010 BAU industrial demand) due to fuel switching from oil and coal biolers. For the high-
efficiency/low-carbon (HE/LC) case we assume, similar to Section 4.2.2, that the penetration of ATS in
these markets will double over that of the efficiency scenario. In addition, we assume the grid
electricity (see utility chapter for details) and the steam boilers displaced have higher carbon emissions
than those displaced in the efficiency case. This results in an ATS HE/LC carbon reduction of 14-24
MtC equivalent per year by 2010. This corresponds to an electric capacity of 42-51 GW and 1.0 TCF of
additional natural gas (11% of projected BAU 2010 industrial demand).
Most of the carbon reduction comes from the fact that the ATS has a combined efficiency that is 5-10%
greater than boilers. This greater efficiency also results in electricity costs that are 10% lower than
current generation systems. Equipment costs are projected to be approximately $350/kW ($1.8M for a
5 MW unit) for a recuperated simple cycle unit and somewhat higher for a combined cycle unit. The
major turbine manufacturers in the U.S. project that ATS will have captured 15% of U.S. power
generating capacity by 2010 (Major 1997). In power-only mode, the system will be competitive against
electricity prices of $0.03-0.04/kWh (Brent and Davidson 1996, Hoffman 1997). More specifically,
Figure 4.5 shows that the ATS is the least-cost option for a wide range of gas and electricity prices, but
it does not compete favorably with very low gas prices (where the large combined cycle turbine is less
expensive) or with high gas prices (where coal gasification systems are less expensive). Note that the
breakeven point between ATS and combined cycle systems is very close to the projected price of
natural gas to industrial consumers ($2.60 per million Btu) in the AEO97 BAU case.
Even though the ATS is 2-3 years from being commercialized, some of the ATS manufacturers already
have significant orders for ATS (Parks 1997). Since the average order/delivery time is 18 months, this
means that the ATS customers are willing to wait at least 18 additional months for a superior
technology. This suggests that the ATS may penetrate far more rapidly than traditional energy
technologies.
In addition to carbon reduction, these turbines have other environmental benefits. ATS's low-emission
combustion systems generate less than 9 ppm NOx through lean premix combustion and less than 5
ppm NOx with catalytic combustion, with no other major pollutants. When deployed in 2001, ATS
systems, per MW, will produce 77-95% less NOx per megawatt than competing power generation
technologies (Major and Davidson 1997b).
September 18, 1997
4.21
Chapter 4
The Industrial Sector
Figure 4.5 Electric Generation Cost Comparison
0.07
225 MW
Combined Cycle
0.06
500 MW Coal
0.05
Gasification
Cost of Electricity ($/kWh)
0.04
5 MW
CHP ATS
0.03
0.02
0.01
0
2
3
4
5
6
7
8
Cost of Gas ($/MMBtu)
Source: Onsite Energy (1994)
4.3.3 Fuel-Switching Technologies
In the very near-term, fuel switching from high-carbon fuels such as coal to lower-carbon fuels such as
natural gas is feasible and is already included in most energy forecasting models. In the near and
medium time frame, biomass and black liquor gasification technologies described in Section 4.3.3.1
provide the option of switching from a high-carbon to a "no-carbon" fuel. Biomass is considered "no-
carbon" because we assume the CO₂ produced will be rapidly resequestered by growing biomass feed
stock (see Chapter 7 for more detail on biomass). These technologies can also be considered PSEM
technologies because they replace inefficient biomass boilers and grid electricity with biomass
gasification cogeneration. Black liquor technology utilizes black liquor gasification instead of
improved efficiency recovery boilers (which are the replacements implicit in the modeling calculations
of Section 4.2.2). Biomass gasifiers replace inefficient boilers for steam and electricity. These
technologies allow the industry to generate more of its own electricity which leads to the offset of
purchased electricity. The extra generation of biomass-based electricity is not included in the
modeling calculations of Section 4.2.2 and is responsible for the carbon offsets calculated here.
Although no examples are provided, other renewable energy-powered industrial technologies (e.g.,
solar detoxification) could also be considered low-carbon fuel-switching technologies.
4.22
September 18, 1997
The Industrial Sector
Chapter 4
4.3.3.1
Integrated Gasification Combined Cycle Technology for the Forest Products Industry
Integrated gasification combined cycle (IGCC) technologies can significantly impact the carbon
reductions expected in the forest products industry in two ways: (1) by increasing energy self-
generation and (2) by better utilizing residues from the forest management and manufacturing
processes. Potential offsets of carbon emissions by 2010 are approximately ten MtC equivalent per
year in the high-efficiency/low-carbon scenario. The efficiency scenario could achieve offsets of about
5 MtC equivalent per year. To achieve the carbon reductions in the high-efficiency/low-carbor
scenario, it will be necessary to facilitate early commercialization to reduce investment risk and
provide an incentive for industry to commit the resources necessary to implement these advanced
technologies.
The pulp and paper industry purchases 43% of its energy and uses a diverse mix of resources
including electricity, steam, coal, residual and distillate fuel oil, liquid propane gas, and natural gas. In
1972, the industry used oil for nearly a quarter of its purchased energy but this proportion decreased
to 6.9% in 1994 by doubling purchased electricity and increasing coal purchases by 50%. This complex
purchased fossil fuel and energy pattern is shown in Figure 4.6.
Figure 4.6 Purchased Energy in the U.S. Pulp and Paper Industry by Fuel Type, 1972-1994
Elecnticity
Steam
Trillion Btu
Sold
1400
Other
Coal
1200
Nat Gas
1000
Resid
LPG
800
Distill
Distill
600
Resid
LPG
400
Coal
200
Nat Gas
Steam
0
1972 1983 1986 1987 1988 1989 1990 1991 1992 1993 1994
Elecriticity
Other
Year
Sold
Source: Miller Freeman, Inc. (1972-1994)
The industry self-generates the remaining 57% of its required energy through the recovery of energy
and chemicals in spent black liquor, use of residues such as hog fuel and bark in boilers, and
cogeneration of heat and power (see Figure 4.7). The American Forest and Paper Association (AF&PA)
estimates that use of these energy sources displaced more than 227 million of barrels of oil in 1994
(Miller Freeman, Inc. 1996).
These fuel switches, increased cogeneration, and energy conservation measures resulted in a decrease
in energy intensity. Even though total energy consumption increased over the period 1972-1994,
energy consumption per ton of product output decreased by 21% (Miller Freeman, Inc. 1997).
Two opportunities for further improvements were analyzed in detail: increased self-generation from
black liquor and increased recovery of usable energy from hog fuels and bark coupled with increased
recovery of forest residues and pre-commercial thinnings. Increased self-generation offsets purchases
September 18, 1997
4.23
Chapter 4
The Industrial Sector
of electricity and coal, and thus offsets CO₂ emissions.¹⁷ These higher-efficiency processes could also
increase the industry's electricity production for return to the grid.
Figure 4.7 Self-Generated Energy in the U.S. Pulp and Paper Industry by Fuel Type, 1972-1994
Trillion Btu
1600
1400
Other
1200
1000
Hydro
800
Bark
600
Hogged
400
BLiquor
200
0
1972 1983 1986 1987 1988 1989 1990 1991 1992 1993 1994
Year
Source: Miller Freeman, Inc. (1972-1994)
Kraft Recovery Boiler Replacements. Traditionally, about 40% of the energy used in a mill is
generated from burning the lignin solids. Lignin is the portion of wood that holds the fibers together
and makes them stiff. The pulping process separates the lignin from the pulp fiber. The lignin is a
dilute solution which is evaporated and burned in a boiler designed to recover the pulping chemicals;
heat from combustion is used to make steam. Some of the steam is used to supply the mill's needs and
some is used to generate electricity for the mill.
In the black liquor gasification combined cycle process, a little less steam is generated but two to three
times more electricity is produced. Process changes designed to make mills more environmentally
friendly tend to change the balance of energy forms that a mill uses. Mills are using less steam energy
and more electrical energy; the combined cycle process fits right into the future process needs.
The technology is coming on the scene at an opportune time because most of the existing recovery
boilers in the industry are reaching the end of their useful safe operating life. The majority of recovery
boilers were put into service between 1955 and 1980, with a peak period around 1967 (see Figure 4.8).
For environmental and safety reasons the industry is developing altérnative technologies in
anticipation of replacing these boilers after a 40-year service life. The need for capital replacement
creates an opportunity for penetration of new, high-performance, environmentally acceptable
technologies. The gasification component of the replacement technology is already at the early stages
of commercial deployment, mainly as a means of expanding mill electric generation capacity in
situations where the current recovery boiler limits throughput. There is a need for chemicals recovery
cycles to be tested and for the integrated cycle to be demonstrated. Expediting RD&D could allow
significant carbon emissions offsets by matching the timing of technology development and
commercialization to the need for boiler replacement.
4.24
September 18, 1997
The Industrial Sector
Chapter 4
Figure 4.8 Kraft Boilers in Service in the United States
20
18
16
14
Number of boilers entering service
12
10
8
6
4
2
0
1946
1951
1955
1959
1963
1967
1971
1975
1979
1983
1987
1991
Source: American Forest and Paper Association
A major barrier to the adoption of black liquor IGCC systems is the central role that the current
recovery boiler plays in the chemical and energy recovery of the mills. Typically, this part of the
pulping process has to reliably operate at full throughput with annual capacity factors of greater than
95%. A further barrier is the need for process heat. Increasing the electricity output will require a
concomitant improvement in process heat utilization since the steam output of the black liquor IGCC
system will be 21% less than that of the recovery boiler, even though the electricity output is
effectively doubled.
Replacement of the current recovery boilers by new technology based on gasification to recover both
process chemicals and the energy content of the dissolved lignin has the potential to produce 104 TWh
of electricity per year, offsetting about 100 Mt of CO₂ emissions. Full replacement of the current
recovery boiler capacity at the 1996 production volume would offset 26 MtC equivalent per year.
Based on a rate of recovery boiler replacement that assumes a 40-year life for the existing recovery
boilers, the 2010 displacement is 5.2 MtC equivalent per year, and the 2020 displacement is 8.7 MtC
equivalent per year. The methodology used to determine the replacement rate, on which the projected
carbon reductions are based, is discussed in Appendix D-4. The black liquor IGCC system is designed
to meet New Source Performance Standards (NSPS), and would also have low NOₓ and SOx
emissions. Investment costs for integrated gasification combined cycle are forecast to be less than
those for replacement with a conventional recovery boiler system, on a dollar per kilowatt-hour basis.
It is anticipated that IGCC systems would be competitive against electricity purchases at $35/MWh.
Residual Biomass Boiler Replacements. Food processing, wood products, and pulp and paper are
industries that generate large amounts of residual biomass (e.g., waste wood and bark). While much
of this biomass is currently being used, if it were gasified and used to cogenerate steam and electricity,
it would substitute for (largely) fossil fuel-produced electricity. Advances in turbine efficiency (see
Section 4.3.2.1) make this an economically attractive option. By using residues from pulping processes
as well as biomass from forestry operations in conjunction with gasification and combined cycle
September 18, 1997
4.25
Chapter 4
The Industrial Sector
technologies, 2.3 GW of capacity can be put in place by 2010, offsetting 4.8 MtC equivalent per year.
This would represent about one-third of the potential mill conversions projected to need replacement
by that time. Because of the stage of development of the technology and its markets, a conservative
estimate would reduce replacements from one-third to one-quarter of the potential mill conversions.
Using the more conservative penetration, the carbon replacement potential from gasification of
residual biomass is 3.6 MtC equivalent per year.
Approximately 200 mills are already producing heat and some power from the use of residual
biomass in their processes. 18 The majority of in-place boiler units entered service between 1965 and
1975 and need replacement; they are either reaching the end of their service lives or they may have
difficulty meeting environmental regulations (or both). Residual biomass gasification can penetrate
this replacement market with the potential to double the net rate of electricity generation - from a
generation efficiency of about 15% to 35%. The technology is already in the early stages of
commercialization with the first 18 MW IGCC operating in Sweden. Prototype units are being
demonstrated elsewhere in Scandinavia and the United States.
The current cost of this technology is approximately 50% over the plant cost when the technology is
mature. Incentives will be necessary to facilitate entry of the technology into the replacement market.
One proposal is a capital cost buydown to bring technology costs down.
The gasification system is designed to meet New Source Performance Standards (NSPS) and would
have low NOₓ and SOx emissions. Biomass growth and harvesting would be according to best
practices, and to some extent the biomass fuel source could include materials that are currently
landfilled and thus contribute to landfill methane emissions.
4.3.4 Low Process Carbon Technologies
Low-process carbon technologies reduce or avoid the emission of non-combustion CO₂ and other
greenhouse gases in industrial and other processes. As shown in Table 4.13, 92% of the carbon
equivalent emissions of process carbon are due to non-CO2 greenhouse gases that have far higher
global warming potentials (GWP) than CO2.
4.3.4.1 Industrial Sources of Non-CO₂ Greenhouse Gasses
Although non-CO2 industrial emissions of greenhouse gasses are small by weight, they have GWPs
that range from 21 for methane to 23,900 for sulfur hexafloride (SF₆). Figure 4.9 shows the relative
contribution of these other gases in MtC equivalent. The largest non-CO₂ greenhouse gas contribution
is from methane (CH4), which is responsible for 177.5 MtC equivalent and has a GWP of 21. Next is
nitrous oxide (N₂O) which is responsible for 39.1 MtC equivalent and has a GWP of 310. Finally, in
1994, various halocarbons and other engineered chemicals amounted to 29.5 MtC equivalent. These
engineered chemicals are a source of concern since their emissions are growing rapidly - and the
United States is the major source. As shown in Table 4.13, emissions of these other greenhouse gases
from agriculture (27%), mining/energy extraction (25%), service (24%), and transportation (8%)
sectors are important.
The manufacturing sector accounts for 14% of carbon equivalent emissions due to other greenhouse
gases. The manufacturing processes that generate GHG emissions include:
Waste emissions of CF₄, C₂F₆, C₃F₈, NF₃, and CHF₃ from plasma etching, chemical vapor
deposition (CVD), and CVD chamber cleaning in semiconductor manufacturing;
4.26
September 18, 1997
The Industrial Sector
Chapter 4
Waste emissions of SF6 from the manufacture of transformers, circuit breakers/load-shedding
devices, and electrical distribution components where SF₆ is used as an insulator;
By-product emissions of N₂O from adipic acid manufacture,19
Waste methane emissions from production of ethylene and styrene;
PFC emissions from aluminum production (see Section 4.3.4.3); and
Waste emissions of SF₆ from magnesium casting in which SF₆ is used as a cover gas to protect
against catastrophic oxidation.
Table 4.13 Process Carbon Emissions and Energy Use by Sector
Carbon Emissions (MtC equivalent)
Other GHG
Process CO2*
Carbon
Total Carbon
Energy Use
Equivalent
(quads)
Manufacturing
18.4
33.0
51.4
22.4
Service
0
58.2
58.2
Agriculture
0
66.7
66.7
Mining/Energy
0.9**
61.5
62.4
Extraction
Construction
2.0
0
2.0
Subtotal Industry
21.3
219.4
240.7
32.6
Buildings
0.0
5.3
5.3
33.7
Transportation
0.0
19.0
19.0
25.5
Total
21.3
243.7
265.0
91.8
*Source: ELA 1996
**Gas flaring.
While none of the manufacturing emissions are particularly large, we note that global emissions of SF6
are increasing at a rate of 7-8% per year. This is of particular concern because SF6 has a very high
global warming potential of 23,900 and an expected lifetime of 3,200 years, making it a very potent
greenhouse gas. Thus SF₆ emissions alone are increasing at a rate of 0.5 MtC equivalent per year (ELA
1996). Several emerging technologies may be immediately helpful in avoiding these emissions. For
example, applications of high temperature superconductor technologies include transformers and
current limiters that act as circuit breakers (Platt 1997). Many of these emissions are seen not only in
energy-intensive industries but also in "high-tech" manufacturing industries. These non-energy-
intensive industries include semiconductor manufacturing and equipment manufacturing for the
electric utility industry. Due to scope and time constraints, technology options to reduce these
emissions are not addressed in this report but are an important area for future analysis.
September 18, 1997
4.27
Chapter 4
The Industrial Sector
Figure 4.9 Non-CO₂ Greenhouse Gas Emissions in the United States (MtC equivalent)
180.0
Buildings
Transportation
160.0
Mining/ Energy Extraction
140.0
Agriculture
Service
120.0
Other Manufacturing (mainly
100.0
semiconductors)
Metal Casting
80.0
Steel
Forest Products
60.0
Chemicals
40.0
Aluminum
20.0
0.0
PFC
CH4
N20
HFCs
SF6
Source: ELA (1996)
4.3.4.2
Process CO₂ Emissions
Compared to other greenhouse gases, process CO₂ emissions are relatively small, accounting for only
9% of process carbon emissions and less than 5% of industrial combustion-related CO₂ emissions.
Overall, the industrial sector directly emitted about 23 MtC from CO₂ industrial processes.
The primary industrial processes that generate process carbon emissions include:
The calcination of limestone in cement manufacture (largest single source);
The manufacture and consumption of limestone (e.g., in lime kilns, iron smelting, steel making,
glass manufacture and flue gas desulfurization);
Dolomite consumption;
Soda ash manufacture and consumption (e.g., in glass manufacture, flue gas desulfurization, and
chemicals production);
CO₂ manufacture;
4.28
September 18, 1997
The Industrial Sector
Chapter 4
Gas flaring; and
Aluminum production.
There has also been a disproportionate increase in process CO₂ emissions relative to combustion-
related CO₂ emissions. Over the past eight years, process CO₂ emissions have increased nearly 16%
while combustion-related CO₂ emissions have increased only 4%. However, these process carbon data
are highly uncertain due to their variability across sites due to non-uniform measurement technique.
For example, the carbon emissions could be flat in reality but appearing to rise because measurements
are more comprehensive today.
The following sections describe carbon savings in the aluminum and cement industries that are
possible given aggressive RD&D and commercialization strategies.
4.3.4.3 Low-Carbon Technologies in Primary Aluminum Production
Because of the very high chemical stability of aluminum oxide and other aluminum compounds, the
production of aluminum metal was not-feasible until the nineteenth century when electrical power
generation facilities became available to permit commercial electrolytic reduction operations. Creation
of today's world-wide aluminum industry occurred after the simultaneous inventions by Hall and
Heroult of a process for high-temperature reduction of aluminum oxide dissolved in a molten fluoride
salt using a carbon anode which is consumed during the process reacting to form carbon dioxide.
The global warming potential associated with aluminum production results from several factors
First, carbon dioxide is generated at fossil fuel plants that produce the electricity required for the
electrolysis process. 20 State-of-the-art Hall-Heroult cells achieve power consumption levels as
low as 13,200 kWh/tonne of aluminum produced; however, most aluminum plants require more
electricity per tonne of product.
Second, the production of one metric ton (or tonne) of aluminum leads to the generation of at
least 1.22 tonnes of process carbon dioxide (or 0.33 tonnes of carbon) from the reduction cell
operation.
Third, global warming effects also result from the generation of perfluorocarbons (CF₄ and C₂F₆)
during instabilities in the cell operation (called "anode effects") that occur when oxide
concentration in the cell bath becomes undesirably low. In 1994, aluminum smelting is estimated
to have emitted the equivalent of 4.2 million metric tons of carbon equivalent, from
perfluorocarbon (PFC) byproducts (ELA 1996).
Further reductions in the levels of greenhouse gas emissions associated with primary aluminum
production will require:
1. Development of reduction technologies that require less energy for primary metal production;
2. The development of inert, non-carbonaceous anodes that are not consumed through the
reduction process; and
3. The development of improved cell designs and operating control strategies to reduce PFC
emissions.
September 18, 1997
4.29
Chapter 4
The Industrial Sector
On the basis of ongoing research on aluminum reduction technology, the desired improvements will
require the development and commercialization of retrofit advanced cell technology with wettable
cathode and inert anode components. Two scenarios have been developed: an efficiency scenario,
based on the development and use of wettable cathodes with conventional carbon anodes, and a high-
efficiency/low-carbon scenario, based on the addition of inert anodes to the advanced cell.
Under both the efficiency and high-efficiency/low-carbon scenarios, R&D on advanced aluminum
production cells would be funded by both the federal government and the private sector. However,
under the high-efficiency/low-carbon scenario, the development of inert anodes and the associated
control systems would be pursued more aggressively. In either case, alternative cathode and anode
materials, advanced cell designs, and advanced operating control methods would be developed with
the overall goal of reducing the cell voltage (electrical energy requirements and associated power
plant CO2 emissions), eliminating CO2 cell emissions, and significantly reducing emissions of PFCs
arising from cell operating instabilities. In the discussion below, we analyze the incremental energy
efficiency improvements and reduced carbon gas emission savings from these advanced low-carbon
technologies for primary aluminum production.
Under the efficiency scenario, the wettable cathode part of the advanced cell is forecast to be ready for
commercial operations by approximately 2005. Conventional, non-wettable cathode cells operate with
thick metal layers above the cathode surface. In contrast, use of wettable cathodes permits cell designs
in which product metal can be drained from the cathode to collection sites within the cell leaving only
a thin film of metal at the cathode surface. Normal undulations at the metal surface resulting from
electromagnetic stirring and gas bubble driven circulation are virtually eliminated with wettable
cathodes permitting cell operations with reduced anode-cathode spacings. In combination with
advanced process sensors and control systems to optimize cell operation, the potential energy savings
are estimated to be as high as 15-20% over conventional cells (DOE 1990). These same sensors and
control systems would yield reduced levels of PFC gas emissions. These technologies will be designed
for simultaneous or independent retrofit use on existing cells.
The high-efficiency/low-carbon scenario forecasts the additional development of inert anodes. The
most promising materials presently being evaluated are ceramic/metal composites consisting
primarily of nickel oxide and nickel ferrite with a copper/nickel metal phase (Windisch and Strachan
1991). These permanent anodes would also eliminate CO₂ emissions associated with the manufacture
and consumption of carbon anodes. If successful, the advanced cell would result in an approximate
27% reduction in the electricity requirements for primary aluminum production.
Research would be scheduled so that commercial-scale demonstration tests (individual commercial
sized reduction cells up to the actual conversion of an operating potline) would be in operation by
approximately 2005. To re-engineer an existing smelter site with radically different production may
require a capital investment ranging from $500,000 to $2 billion. For investments of this scale,
conclusive demonstrations defining operating performance, operating costs, and equipment lives
must be completed to achieve industry acceptance and widespread adoption.
The economic feasibility of the advanced technology would be enhanced if federal policies promoting
further reductions in carbon emissions were established. Even without such policies, the U.S.
aluminum industry has expressed a goal of eliminating process CO₂ emissions in primary aluminum
(Energetics 1997). Furthermore, trends toward increased use of aluminum in the transportation sector
to improve vehicle fuel efficiency through weight reduction could significantly increase demand for
primary aluminum, further increasing the economic feasibility of the advanced cell technology.
4.30
September 18, 1997
The Industrial Sector
Chapter 4
Under the efficiency scenario, we assume that five of the existing 22 aluminum plants operating in the
U.S. (American Metal Market 1997) are retrofitted to use the advanced wettable cathode cell. With an
average plant capacity of 190,000 tonnes of aluminum per year and an average annual electricity
consumption of 13,200 kWh per tonne of aluminum, electricity efficiency improvements of 17% in
these five plants would result in 0.19 million tonnes of reduced carbon-equivalent emissions in 2010.21
This is 0.09 MtC more than the efficiency scenario described in Section 4.2.2.²² In addition, the PFC
emissions from anode effects are projected to be halved in those five plants where wettable cathodes
are installed. This would represent an 11.4% (or 0.48 MtC) reduction in the aluminum industry's
carbon equivalent emissions of 4.2 MtC.²³
Under the high-efficiency/low-carbon scenario, use of the advanced, inert anode by 10 of 22 plants
could lead to reduced carbon emissions by 1.6 MtC equivalent, of which 1.00 metric tonnes of carbon
savings are due to the reduced consumption of electricity.²⁴ This is equivalent to 0.67 MtC over the
high-efficiency/low-carbon scenario described in Section 4.2.2.²⁵ An additional 0.6 Mt of carbon
savings result from the elimination of carbon emissions from the production cell.26 The use of inert
anodes to eliminate the process CO₂ emissions from smelting was not considered in Section 4.2.2;
thus, all of these carbon reductions are accounted for here. In addition, the PFC emissions from anode
effects are projected to be eliminated in those 10 plants where inert anodes are installed. This would
represent a 45.5% (or 1.91 MtC) reduction in the aluminum industry's carbon equivalent emissions of
4.2 MtC."
These carbon reduction estimates are summarized in Table 4.14. The advanced aluminum production
cell in the efficiency scenario accounts for 0.6 MtC (or 0.5 to 1.0 MtC) of reductions above the carbon
reductions already incorporated in Section 4.2.2. The high-efficiency/low-carbon scenario accounts
for 3.2 MtC ( or 3 to 3.5 MtC) more than the carbon reductions already incorporated in Section 4.2.2.
Technical details of the advanced aluminum production cell are discussed in Appendix D-6.
Table 4.14 Carbon Reductions from Advanced Aluminum Production Cells, in 2010 (MtC)
Sources of Carbon Reductions
Efficiency Scenario
High-Efficiency/Low-
Carbon Scenario
Electricity Savings
Included in Section 4.2.2
0.1
0.3
Increment above Section 4.2.2
0.1
0.7
Cell Production
0
0.6
Reduced Perfluorocarbons
0.5
1.9
Total
0.7
3.5
4.3.4.4 Replacing Cement Clinker with Solid Wastes
The cement industry is the single largest source of U.S. process CO₂ emissions and a major energy
user. The annual process CO₂ emissions from the U.S. cement industry are 9-10 MtC equivalent (ELA
1996). Energy-related CO₂ emissions are of similar magnitude depending upon the cement kiln
technology. Some estimates indicate that each ton of cement clinker produced results in the direct
emission of one ton of CO2. Other estimates with different kiln technologies have a much higher
energy/process CO₂ ratio. Of the process emissions, about 60% of the direct emissions are from
calcination of limestone and the other 40% are from combustion products from fossil fuels that directly
or indirectly supply the energy for calcination.
September 18, 1997
4.31
Chapter 4
The Industrial Sector
Nearly all cement in the United States is made from ground clinker intermixed with gypsum. One
technically straightforward and cost-saving way to reduce energy input and carbon emissions per ton
of cement is to replace some of the clinker with abundant utility and steel plant wastes such as fly ash
or granulated blast-furnace slag. Such a replacement makes cement with somewhat different
properties, but still a satisfactory building material. Most European countries allow such cements and
have found that these cements last longer and are more tolerant to salt water than pure clinker cement.
However, U.S. product specifications (Standard Specification for Portland Cement, ASTM C150) do
not allow any extra ingredients in cements. These specifications are difficult to change because the
small minority of those who might lose markets (e.g., non-integrated cement producers) can easily
stop changes under the current system. A recent study (Sauer 1997) estimates that changing the U.S.
specifications to permit inter grinding could reduce both energy and process CO₂ emissions by 5-20%
per year by reducing demand. If the specifications were changed, it is likely the new technology could
be rapidly adopted by U.S. cement manufacturers, especially the multi-national firms that are already
using this type of cement in Europe. Under the high-efficiency/low-carbon scenario, the barriers to
this technology could be overcome. In addition, there would be motivation to conduct further
research, development and demonstration activities exploring a wide range of cement inter grinding
materials and percentages and to ensure that they provide the same or improved performance. Based
on these studies and assuming a low-carbon, aggressive R&D scenario, our estimate is that by 2010, 1-
2 MtC equivalent of industrial carbon emissions could be avoided due to cement inter grinding and
replacement.
Though the manufacturing process has remained the same, the U.S. cement industry has changed over
the past 20 years. The number of kilns in operation has dropped by 50% since 1975. There has been a
28.3% improvement in fuel efficiency since 1975, dropping the energy required per metric ton of
cement from an average of 7.26 MMBtu in 1975 to 5.20 MMBtu in 1994. Over 60% of U.S. clinker
capacity is foreign owned or affiliated with foreign firms, and most of these are integrated European
cement companies. The primary customer, accounting for 60% of shipments, is the ready-mix concrete
industry which supplies concrete, mixed to customer specifications, to construction sites (Bureau of
Mines 1994). Concrete typically contains 10-15% cement as a binder. Cement demand is projected to
grow at 1% per year, half the rate of GDP.
On average, energy accounts for between 30 and 40% of cement manufacturing cost. Electricity
represents about 10% of energy input, but frequently accounts for close to 50% of total energy cost.
Integrated cement producers and ready-mix concrete suppliers would benefit from replacing high cost
clinker with low- or negative-cost materials. The cement industry is already a leader in waste
utilization. More than'half of plants responding to a 1994 survey reported the use of one or more types
of waste as fuel. This technology could, however, speed the decline of non-integrated cement
producers.
In addition to reducing CO₂ emissions, this technique also reduces NOx, SO₂, and particulate
emissions associated with electricity use. It also reduces solid waste by replacing quarried raw
materials with wastes and by-products such as fly ash, foundry sands, and mine tailings.
4.4 PROVEN INDUSTRIAL TECHNOLOGIES
Although our forecasting methodology does not draw directly from detailed representation of
individual technologies, the forecast savings that are expected in each sector will be drawn from a
variety of sources of new technologies and business practices. This section illustrates the range of
commercially available and near commercial innovations that firms in these industries can draw upon
to achieve the additional reductions in energy use that are considered feasible in the HE/LC case and
4.32
September 18, 1997
The Industrial Sector
Chapter 4
could contribute to this projected decline. In addition, we provide examples of technologies that
directly displace carbon in Section 4.3.
We provide illustrative examples of currently-available technologies that we believe could be
integrated into industry to provide the savings suggested by the model simulations for each energy-
intensive industry; we also provide examples of cross-cutting technologies. We describe how each
technology is used and from what type of efficiency it draws its energy and cost savings.
Some technologies recover or reduce the production of waste heat in high-temperature applications
while others optimize the process load to the energy-using equipment. Many of the most successful
technologies have multiple benefits, including pollution prevention or productivity-enhancing
features. A technology that reduces product loss or increases process throughput will often reduce
labor or material costs as well as energy costs. For example, continuous casting, widely adopted by
the steel industry, is cost-effective based on its energy savings alone but industry has adopted
continuous casters in large measure because of the improvement in steel quality and because it
reduces losses. Similarly, impulse drying, an emerging technology, saves energy, but also allows
additional throughput on the paper-making machines and will improve the quality of the product.
While it is felt that these technologies are representative and have the potential to be readily accepted
by industry, the estimates of energy savings provided below do not represent any industry consensus of
the relative difference between the new technology and average practice. Instead we rely on available,
published literature that assesses the performance of these technologies and business practices.
The diversity of industries, businesses, plants, and processes implies that not all of these examples will
be universally cost-effective, or even applicable. Site- or plant-specific constraints may prevent the use
or economic acceptability of a technology for retrofit applications that would be readily accepted in a
new plant design. In many of the most energy-intensive process industries, few green-field plants are
being built in this country, further limiting some applications. While we do not consider explicitly the
economics of when to replace old equipment, we understand that a variety of considerations enter into
this business decision, including:
How learning curves tend to continually lower the costs (including energy costs) as cumulative
production experience with new technology is gained;
Countervailing factors like "wear and tear" that tend to increase costs over time;
How the introduction of new equipment can alter the economics of existing equipment; and
Available design trade-offs between capital and other costs, especially energy costs.
New and replacement capacity will be put into place at many existing plants based on these and other
decision variables. The opportunity for new technology to be adopted occurs at the point in time
when these decisions are made. It is at this point that energy prices and capital discount rates can
influence the decision to purchase new technology and thus the adoption of technologies for which
examples are given below.
Many of these technology examples exhibit energy savings of more than 5-10% relative to current
average practice, but the turnover rates of the capital stock in the energy- and capital-intensive
industries require our projections to take this into account. In 13 years, many of these technologies
(and many others not listed here) are capable of reaching higher levels of penetration, but most will
not achieve 100% penetration. In addition, the technology examples often account for some fraction of
September 18, 1997
4.33
Chapter 4
The Industrial Sector
the energy use in that sector. However, the examples show that there are many ways in which
efficiency in industry can be increased, given the right incentives.
Brief descriptions of energy-efficient technology opportunities for the industrial sector are provided in
the following sections; more details are available in the associated appendices and references.
4.4.1 Cross-Cutting Technologies
There are a variety of cross-cutting technologies that are not process- or product-specific in operation
in industry. Some include lighting and heating, ventilation, and cooling technologies that are also
commercial applications and are not discussed here (see Chapter 3). Others include sensors and
computer control systems which have a common underlying technology, but have a variety of
configurations and benefits depending on the industry. There are two major ways that all of industry
can benefit from improved efficiency: cogeneration and improved motor systems.
4.4.1.1 Combined Heat and Power
Combined heat and power (CHP) is the joint production of useful steam and electricity, either for on-
site use or sale back to the electric grid. There are substantial thermodynamic advantages to the joint
production of heat and power that could greatly reduce generation losses from traditional power
production and would reduce carbon emissions system-wide. The advantage of such an approach is
that little additional fuel is required for the electricity generation over that required for simple steam
production. Thus, the efficiency for use of the thermal energy available from the fuel is higher than
with separate electricity generation and steam production, and the net greenhouse gas emissions can
be reduced by the application of cogeneration. Based on a typical boiler configuration, the gas turbine
with heat recovery steam generation is typically the most cost-effective (Boyd et al. 1996). CHP can
also help reduce carbon through fuel switching to low- or no-carbon fuel. Under the BAU case, CHP
power production will grow to 333 TWh by 2010. See Section 4.3.2 for an example of a CHP system
that can reduce carbon emissions far more than predicted in the BAU.
4.4.1.2 Motor Systems
Energy-efficiency opportunities associated with electric motor drives derive not so much from the
replacement of motors with high-efficiency motors as from energy-conscious design throughout the
system employing the motor drive. Such a systems approach (see Section 4.3.2) has also resulted in
significant non-energy savings when motor systems are improved.²⁹ The system includes power
supply lines, controls, motor feed cables, the electric motor, the drive and transmission system, and
the driven load. Each of these system elements may present a significant opportunity to conserve
energy.
The power supply and control systems affect efficiency in three ways. First, power is consumed by
resistance losses in the supply wires. Second, losses in the supply wires may contribute to voltage
imbalance in the power supplied to a polyphase motor, leading to reduced efficiency and possible
motor damage. Third, other system loads and certain control devices, particularly adjustable speed
drives, can distort the sinusoidal AC voltage provided to the motor, resulting in efficiency and torque
losses, vibration, and possible bearing damage, which is accompanied by increased friction.
Losses associated directly with the electric motor include electrical resistance losses, magnetic losses,
friction and air flow losses, and stray losses associated with manufacturing quality limitations. High-
efficiency motors address these losses, though efficiency improvement over standard motors may only
average 5% to 7%. While an electric motor consumes less than full power when the load it serves is
less than the motor rating, the efficiency of the motor declines dramatically as the load declines below
4.34
September 18, 1997
The Industrial Sector
Chapter 4
40% of rated load. Since motor over sizing is common practice, this provides a significant efficiency
improvement opportunity.
Losses associated with drive systems are frictional losses in belt and gear systems. Higher losses are
associated with greater speed reductions, which may improve the relative economics of adjustable
speed drives (motor speed control). While drive transmission efficiency may be well over 90%, it may
be below 50% as well. Thus, drive system design may offer more savings opportunity than motor
replacement.
The most important savings opportunities will often lie in specification and design of the driven load.
In the extreme, process changes may eliminate the need for the load entirely or equipment substitution
can reduce power requirements. For instance, mechanical conveyors may be used rather than
pneumatic conveyors at a substantial energy savings. More commonly, proper selection of loads such
as fans, pumps, and compressors to match the intended application requirements will result in the
equipment operating at higher efficiency and presenting less load to the electric motor. Then, proper
matching of the remaining load to a motor, perhaps with variable speed control, will result in optimal
overall system efficiency.
4.4.2 Pulp and Paper
Paper manufacturing was one of the most energy-intensive industries in the United States in 1994,
using more than 18,500 Btu per dollar value of shipments. The manufacturing of paper requires that a
fiber source, normally wood, be chipped, digested, bleached, and then formed as a slurry from which
paper or board is made. Once formed as paper, the product must be dried. Large amounts of steam
and power are used to debark and chip the wood, digest the wood, bleach the pulp, and dry the paper
products. Much of this energy source (over 50%) comes from the reprocessing of lignins from the
wood, bark, and unusable portions of the tree. In lumber and wood products, the fraction of biomass
energy sources is nearly 70%.
In paper manufacturing, any technology that will economize the use of steam, reduce the need for
heat, better utilize the biomass fuel sources available, or help to balance both steam and power needs
will improve the performance of the industry. The technologies that hold promise to reduce energy
and carbon emissions in the near-term continue to economize on the use of heat. Longer-term options
alter the balance between steam and power. The most promising near-term options are discussed
below.
Impulse Drying: Impulse drying reduces the huge energy requirements of evaporative drying by
removing more water in the pressing section and reducing the amount of water which must be
evaporated. The total energy savings for full implementation of this technology are estimated to be
approximately 0.25 quad/yr. Without an invigorated effort, the net energy savings are estimated to be
about 12 trillion Btu annually from a market penetration of only 65 drying units by 2020. Impulse
drying methods allow papermaking machines to run at higher speeds, thereby increasing production
rates. This drying method reduces energy use by one-third, reduces production costs by $5 per ton of
paper, improves paper strength by 25%, increases productivity by as much as 80%, and reduces
carbon dioxide emissions as well.
Multiport Cylinder Drying: The evaporative drying in a paper mill is accomplished by winding the
continuous sheet of paper serpentine over a series of rollers. The rollers are pressurized with steam
which condenses on the inside of the roller. The multiport cylinder drying concept uses an alternative
method to remove the condensate from the drier, which reduces the condensate film thickness inside
the drier to 25-30% of conventional technology. This improves heat transfer and increases drying.
September 18, 1997
4.35
Chapter 4
The Industrial Sector
On-Machine Sensors for Paper Properties: The development of new sensors to provide real-time
feedback on whether the process and product are within specification can save the energy of
reprocessing off-grade material and allow the use of greater amounts of recycled fiber. With an on-line
sensor for strength properties the process variability can be reduced and greater proportions of
recycled fiber utilized. A 10% reduction in refiner energy at a single mill saves more than 70 billion
Btu/year. Reducing the normal off-grade production rate by 50% (from a typical 5% to 2.5%) can save
an additional 118 billion Btu/year. If 300 plants adopted these sensors, the annual savings would be
about 60 trillion Btu.
Biomass Gasification Cogeneration: The pulp and paper industry is about 57% energy self-sufficient,
due to the use of wood residues (i.e., hog fuel and bark, pulping wastes, and cogenerated electricity).
The gasification of biomass and electricity generation through a combined cycle would increase the
electricity output of the paper industry, further reducing purchased electricity needs. To meet the in-
plant process steam requirements, this biomass-based integrated gasification and combined cycle (BM-
IGCC), would require an increased utilization of wood residues (about double) possibly from wastes
in plantation forestry or other sources. If one-third of the current population of hog and bark boilers
were to be replaced with BM-IGCC, many of which will be retired by 2010, then cogeneration output
from the paper industry would increase by 17 billion kWh, about 27% compared to 1994 levels. This
would reduce total U.S. industrial electricity purchases by 1.3% in 2010 and carbon emissions by about
1.3 million metric tons.
4.4.3 Chemicals
The chemical industry is almost too complex to characterize as a single industry. Some products -
chlorine and other industrial gases - are made electrolytically or using electricity to compress and
liquefy gases. Other processes, such as petrochemical processing, require high temperatures and
pressures to effect the chemical combination or separation that is required. Within chemical
manufacturing there are over 30 industries and more than 10,000 products. A recent study by
Steinmeyer (1997) found that, in the chemicals industry, simple capital-energy tradeoffs (e.g., using
larger pipes and heat exchangers) result in a 37% reduction in process energy consumption for a cost
of less than 1.5% of total production costs; this study examined only energy-related costs. Another
recent study by Elliot (1997) showed that productivity savings are often far larger than energy savings.
For example, at the Louisiana Division of Dow Chemicals from 1982 to 1993, the average total annual
savings from efficiency projects was 3.2 times the energy savings (Nelson 1993).
Reaction and separation are at the heart of most chemical engineering processes, and they typically
require heat, high pressure, or both. Because of these requirements, the industry in 1994 used 5.3
quads of energy (second only to Petroleum Refining) and required nearly 16,000 Btu per dollar of
product shipped. Promising technologies for the near-term are those that economize on the use of
heat or cooling or bring the two in better balance. Examples are:
Pinch Analytical Techniques: The "pinch" technique was originally a method for optimizing heat
recovery in thermal processes and has more recently been applied as a general optimization tool.
Energy savings occur because of the heat recovery process (waste heat from one process is used to
provide needed heat to another). In the classic case of heat exchanger networks, the pinch point helps
to define the best match between available and needed heat, allowing the heat exchange system to be
optimally sized for greatest cost-effectiveness. In early applications, energy savings averaged 30%,
with capital cost savings in new plant designs, and one year paybacks in retrofits are common.
Refinements to the technique have resulted in typical savings of 50% in new plants and retrofit
paybacks of six months. By the mid-1980s the use of pinch analysis was widespread in the chemical
industry, and its use has broadened further since then (WEC 1995).
4.36
September 18, 1997
The Industrial Sector
Chapter 4
Advanced Distillation Control Techniques: Distillation in refining and chemical industries consumes
3% of total U.S. energy use, which amounts to approximately 2.4 quads of energy annually. In
addition, distillation columns usually determine the quality of final products and many times
determine the maximum production rates. Distillation columns commonly use 30% to 50% more
energy than is necessary to meet the product specifications. It has been estimated that an overall
average 15% reduction of distillation energy consumption can be attained if better column controls are
applied.
4.4.4 Petroleum Refining
The most energy-intensive processes are: distillation; catalytic hydrocracking, reforming and
hydrotreating; alkylation; and hydrogen production. Efficiency improvements can be achieved in the
following ways: (1) introduction of more efficient equipment; (2) reducing process activation energies
(through improved catalysts); (3) improving equipment integration to recover more heat; and/or (4)
adopting improved process control.
4.4.4.1 Monitoring Overall Energy Performance
Refineries could promote energy efficiency by rigorously pursuing a program to monitor
equipment/process/overal refinery energy performance to identify when a system or piece of
equipment begins to become inefficient so that corrective actions can be initiated.
4.4.4.2 Utility System Improvements
The principal utility systems in a refinery are the cooling, steam power, and fuel-gas systems; they are
integrated with virtually every process subsystem. While their impact on the overall refinery
operating profit margin is relatively small, the potential for energy savings is substantial (see
appendix for details).
4.4.4.3 Process/Equipment Modifications
Major opportunities to reduce energy usage also exist through retrofitting and/or replacement of
existing equipment nearing the end of its useful life. Examples of such opportunities are as follows:
Fired (Process) Heaters. Over 60% of the energy used in refineries is obtained from burning gaseous
fuels in refinery heaters. For higher temperature processes such as steam reforming, the application of
advanced oxy-fuel combustion systems such as Dilute Oxygen Combustion can result in net fuel
savings of 25%. These gains can be enhanced further by converting natural gas to hydrogen and
carbon monoxide, making use of waste heat generated by the Dilute Oxygen Combustion System.
Boilers. About 20% of all energy used by petroleum refiners is used for generation of steam. One
route for improving boiler efficiency is through improved sensors and controls. For example,
balancing the burners in a multi-burner boiler and reducing excess air can cut fuel use by 10 to 25%.
In single-burner boilers, controlling excess air can lead to similar gains. The technology to automate
excess air firing is available, but a practical system remains several years away.
4.4.4.4 Fluid Catalytic Cracking
Fluid catalytic cracking (FCC) is currently the most energy-efficient and widely used of the cracking
processes. Improved computer simulations of cracking kinetics should result in an improved
commercial technology by the year 2008. Introduction of improved catalysts and other process
September 18, 1997
4.37
Chapter 4
The Industrial Sector
modifications would occur somewhat later. FCC improvements could eventually lead to CO₂
reductions of up to 8 MtC.
4.4.4.5 Fouling Mitigation in Heat Exchangers
Seven percent of the total energy consumed in petroleum refining is due to extra energy needed to run
heat exchangers that have a fouling build-up. Research indicates that improved operations and
retrofits can reduce fouling. An accelerated program of heat exchanger retrofits and better
understanding of fouling conditions could reduce CO₂ emissions by 0.5 MtC by 2010.
4.4.5 Glass
The glass industry is comprised of several major product segments, each with their own processes for
producing final products. The segments include container, flat glass, wool and textile fiber, specialty,
lighting, and hand glass. The major common energy-intensive stage of the glass industry is the glass
furnace. There are nearly 500 furnaces in over 200 plants in the glass industry (ignoring the smaller
hand glass segment). While there are other stages of product finishing which also require significant
amounts of energy, the examples below focus on the glass furnace as the primary area of concern for
energy efficiency. Other process and product specific areas of energy efficiency are also possible.
4.4.5.1 Oxy-Fuel Process
Since 1991, the fiber, container, and specialty glass industries have accepted the oxy-fuel process as an
alternative to regenerative and recuperative air-fuel furnaces. According to one source, more than 50
major furnaces (20 ton/day) have been converted to oxy-fuel combustion technology (Geiger 1996). In
the oxy-fuel process, oxygen or oxygen-enriched air is used in combustion in the melting furnace. It is
reported that fuel savings from oxy-fuel conversions are typically 10-15% for well designed soda-lime
regenerative furnaces, and at least 30-40% for direct fired or regenerative boro-silicate or lead glasses
(Ross 1996). Currently, approximately 15% of the large commercial furnaces in the U.S. have been
converted to the oxy-fuel process (Ross 1996).
Oxy-fuel technology also increases furnace productivity by 25%, reduces defects, and eliminates the
need for heat recovery (DOE/OIT Impacts, December 1996). There is also a waste-heat-driven thermal
swing absorption (TSA) process for producing low-cost oxygen for this process. The TSA system can
be used in both the glass and steel industries. This low-cost absorption system selectively absorbs
oxygen from air at a cost 30% lower than the best conventional system. This new technology increases
productivity dramatically, reduces fuel use by 60%, nitrogen oxides (NOx) emissions by 50%, and
particulate emissions by 30%. The system also eliminates the need for other more costly add-on NOₓ
and particulate control equipment to meet increasingly stringent environmental regulations for glass
and metal melting. The expected energy savings are 28 trillion Btus ($70 million) annually.
4.4.5.2 Advanced Burner Technology
Adoption of newly developed burners in the oxy-fuel process further improves the energy efficiency
of the process. Some recent burner designs have shown as much as a 30% decrease in fuel use, as well
as improvement of product quality.
4.4.5.3 Glass Batch/Cullet Preheater Technology
The dual batch/cullet preheater uses the oxy-gas furnace's waste heat to preheat cullet and batch
before feeding it to the furnace. Preheating cullet and batch reduces the amount of energy and oxygen
required in the overall melting process (GRID 1996).
4.38
September 18, 1997
The Industrial Sector
Chapter 4
4.4.6 Aluminum
Aluminum smelting is highly capital-intensive, with capacity cost estimates ranging from $3,000 per
metric ton for expansion of existing facilities to $5,000 per metric ton for new facilities (DOI 1993).
Low energy costs in countries such as Brazil, Canada, and Australia have made the international
aluminum industry extremely competitive, and near-term construction of smelting capacity is not
expected in the United States. Investment in state-of-the-art technology has also been limited by
capital constraints. A variety of technologies exist, however, that have the potential to incrementally
reduce energy intensity in the aluminum industry in the time frame to 2010.
4.4.6.1 Improving Hall-Heroult Cell Efficiency
The current U.S. composite baseline energy intensity for aluminum smelting is estimated at 15.2
kWh/kg of aluminum, with the potential near-term reduction using retrofit technology estimated at
13 kWh/kg (Energetics 1997). Performance in the range of 13 to 15 kWh/kg has been achieved in
domestic smelters through a variety of techniques including enhanced potline controls, better anode
rod connections, improved cathode block materials, and increases in anode size resulting in lower
current density (Newsted et al. 1992, Jeltsch and Franklin 1992). Additional research to design
dimensionally stable cells and to optimize materials use for internal control of cells, and to use signal
analysis to analyze cell voltages in potlines, are seen as areas which can improve smelting
performance in the next ten years (Energetics 1997). The primary barriers to adoption of high-
efficiency technologies may be economic.
4.4.6.2 Materials Recycling
Remelting aluminum scrap requires only a small fraction of the energy required to smelt aluminum
from alumina. Remelting is also far less capital-intensive than smelting, which reduces barriers to
modernizing. In 1995, aluminum recovered from old scrap was equivalent to about 35% of apparent
consumption in the U.S. (DOI 1994). While some of the barriers to higher recycling rates are
institutional (e.g., perceived value of recycling beverage containers), technological barriers also exist
for some products like aluminum in cars. These include problems with scrap sorting, separation,
cleaning, and pre-treatment, which inhibit the increased use of different types of scrap and also
contribute to problems with metal quality. Byproduct recycling (e.g., salt cake and spent potlining) is
also inhibited by a lack of knowledge of byproduct characteristics. A critical review of the U.S.
recycling industry infrastructure could identify ways to enhance aluminum recycling rates (Energetics
1997). Given the magnitude of energy savings associated with recycled aluminum versus virgin
aluminum, enhanced recycling may offer the greatest energy savings and greenhouse gas emissions
reduction opportunities in the short term.
4.4.6.3 Improve Furnace Efficiency
Improving energy efficiency of melting and holding furnaces offers potential for energy savings in the
secondary aluminum industry. Several commercially available technologies exist for reducing energy
use in furnaces, including heat recuperators and regenerators and the use of oxygen-assisted
combustion. Heat recuperators operate by passing the combustion products through heat exchanger
tubes, thus allowing the preheating of inlet combustion air and recovery of heat that would otherwise
be exhausted to the atmosphere. Heat regenerators accomplish heat recovery through a paired
burner/exhaust system in which the burners alternate in the firing mode in cycles lasting about 20
seconds. Oxygen-assisted combustion uses oxygen in a dual-firing burner to increase furnace melt
rates, reduce energy use, and reduce emissions. Energy savings from oxygen-assisted combustion can
be substantial (Heffron et al. 1993).
September 18, 1997
4.39
Chapter 4
The Industrial Sector
4.4.7 Iron and Steel
Iron and steel industry comprises the ore-based integrated steel plants, the dominantly scrap-based
"mini-mills," and specialty steel mills. Steel production via integrated plants has been decreasing,
while that of the electric arc furnace (EAF) based mini-mills has been increasing. At present, the
production capacity of the mini-mills is comparable to some of the smaller integrated plants. Mini-
mills are more energy-efficient, since they use scrap or directly-reduced iron or hot-briquetted iron. If
the mini-mill relies mainly on scrap, the range of products that can be produced is somewhat limited
by scrap quality issues.
4.4.7.1 Direct Smelting / Direct Reduction
The ongoing process development activities in iron making in the U.S. and abroad clearly indicate a
need to minimize coke consumption and increase the use of natural gas and/or coal as a reductant for
making solid and/or liquid iron. Energy savings from such technologies arise from by-passing the
coke-making stage and frequently from very high throughput. For example, Kobe Steel and Midrex
Direct Reduction Corp. have developed a production approach for molten iron that reduces the
process from hours to minutes (Metals Industry 1996). Because the product is in molten form, there
are savings in downstream steel making operations and the material can be cooled to iron shot or
ingots without reoxidation.
This technology eliminates the production of coke and reduces the need for ore preparation by
integrating three steel processes into one. Coke-making and ore preparation are responsible for the
largest portion of emissions in primary steelmaking. This technology reduces energy consumption by
20-30% and capital costs by 25-50% compared to conventional blast furnace technology. The first
commercial applications of this technology are operating in Europe.
4.4.7.2 Scrap Preheating
Energy consumption in EAF operations can be reduced by preheating scrap to approximately 400°C
with EAF offgases. Heated metal charges comprising 20-30% of inputs can result in power
consumption rates of less than 300 kWh/tonne of liquid steel (Scheidig 1995). The potential energy
savings is roughly 90 kWh/ton of liquid steel. For a DC Fuchs shaft furnace, compared to a
conventional DC furnace, energy savings of 13.5% and reduced electrode consumption of 29% are
estimated. Baghouse dust reduction is estimated at 30% (Haissig 1994). In the dual shaft furnace
design, iron particles in the offgas tend to adhere to the scrap, resulting in iron recovery in the melt
and leaving the offgas zinc-enriched (Burgmann and Pelts 1995). If zinc levels are enriched to above
25%, the dust may be an acceptable input to zinc refining, rather than requiring disposal as a RCRA-
listed hazardous waste (Center for Metals Production 1987). Preheating also reduces furnace tap-to-
tap time (normally about an hour) by 12 to 15 minutes (Scheidig 1995), resulting in increased raw steel
production capacity, measured in terms of sustainable annual production.
4.4.7.3 Hot Connection
Depending on plant layout, moving forms from the continuous casting operation to the rolling
operation with minimal cooling may provide energy savings. Reheat furnaces are generally employed
to bring the cast forms back to rolling temperature. Adjusting plant layout to move the cast semi to
the rolling operation at a temperature of 600° to 800°C can result in an energy savings of 0.4 to 0.6
GJ/tonne of semi based on the IISI reference plant defined in 1982 (Etienne and Irving 1985). A Dutch
study based on a transport or connection temperature of 700°C estimated an 18% reduction in energy
for reheating, for a savings of 0.3 GJ/tonne of crude steel (De Beer et al. 1994).
4.40
September 18, 1997
The Industrial Sector
Chapter 4
4.4.7.4 Near Net Shape Casting
Near net shape casting provides an example of an innovative and energy-efficient technology that has
experienced rapid penetration in a capital- and energy-intensive industry. It is the direct casting of the
metal into (or near to) the final shape (e.g., strips or sections), replacing the present energy- and
capital-intensive processes of continuous slab casting, slab reheating, and hot rolling. Near net shape
casting uses 25% less energy than the current best practice conventional technology. The first
commercial application, thin slab casting, was introduced in 1989 and now accounts for one-quarter of
all U.S. thin slab production capacity. Using this technique, sheet steel can be produced at a cost of
$250/ton compared to conventional technology costs of $350/ton.
4.4.8 Metal Casting
Metal casting is not a single industry segment according to the SIC system, but covers a diverse group
of products and metals. Products range from cast pipes, motor vehicle components, and tools. Iron,
steel, aluminum, copper and zinc are all metals used by the industry. The industry is labor intensive,
with many small plants; four out of five have fewer than 100 workers. Over half of the energy use is
in melting metal. Technologies which improve the melting stage or reduce waste/recasting have
important energy implications.
4.4.8.1 Computer-Aided Casting Design
Rapid advances in computer modeling of the casting process and in computer-aided drafting of
castings have led to an increased use of computers in foundries, and hence, an increased need for
integration in casting design systems. Increased integration in the casting design functions is needed
to realize the full potential for improving both casting designs and production lead time. Two kinds
of information are produced by the casting analysis and simulation function: (a) predicted outcome of
casting the current design; and (b) the processing parameters for the casting process, if the casting
design appears sound. The predictive results allow the foundry engineer to evaluate the filling of the
mold cavity, the potential for defects such as porosity in the casting to occur, the sequence of
solidification, and the time for complete solidification. With computer modeling, an average of 25%
improvement was found in casting yield (Lensen 1996, Lensen et al. 1995), which would comparably
reduce energy use for metal remelting.
4.4.8.2 Optimized Coreless Induction Melting
Most foundries can dramatically reduce a major portion of their energy through optimization of their
induction melting equipment. It has been estimated that foundries are only operating their induction
furnaces at 50-80% of their optimal efficiency (Horwath et al. 1996). A foundry melting 1000
tons/month could reduce its monthly melting costs by $5/ton by installing sensors and computer
optimization of its melting practice. Four major variables are important in determining the power
required for melting: (1) charge makeup, (2) furnace cover, (3) power application, and (4) furnace
condition. In some cases, optimal material use resulted in higher energy use (22% more). Use of a
furnace cover reduced energy consumption by 12%. Furnace condition (i.e., hot, medium, or cold)
interacts with the charge to significantly affect energy consumption. Maintaining the furnace in hot
condition resulted in 15.4% less energy consumption for melting the charge (Horwath et al. 1996).
4.5 THE LONGER TERM: FUTURE TECHNOLOGIES AND R&D POTENTIAL
The technologies cited above are currently available, or soon will be, because of past R&D. For future
technologies to contribute to increased energy and emissions reductions presumes a continued stream
September 18, 1997
4.41
Chapter 4
The Industrial Sector
of R&D activities into the future. Recent efforts by the Department of Energy are directed at ensuring
that steady stream of R&D by partnering with industry.
The Office of Industrial Technologies, in an effort to garner support and make their research and
development activities more in line with the needs of industry, has initiated a joint government-
industry planning process called the "Industries of the Future." The vision of the way that future
industry will function and the technologies that the industry will use shapes, in part, the organization
and implementation of government R&D efforts. It is this process that may lead to an invigorated
effort to develop future technologies that will improve energy efficiency and reduce carbon emissions.
In this section we discuss the potential for additional decreases in energy intensity in the future as a
result of the continuation of future R&D efforts. Here we draw heavily on the vision documents that
have been published or are being prepared by the energy-intensive industries under the OIT's
Industries of the Future process. We discuss general areas of potential advancement or provide
specific examples of some of the technologies or technology areas that show particular promise for
reducing energy consumption and concomitant greenhouse gas emissions.
4.5.1 Pulp and Paper
The Vision process for the Forest Products Industry of the Future was developed by the industry in
collaboration with the Department of Energy's Office of Industrial Technologies, and is called
"Agenda 2020 - A Technology Vision and Research Agenda for America's Forest, Wood, and Paper
Industry." Two of the major concerns of this document are Environmental Performance and Energy
Performance. One way these objectives might be met is through the use of polyoxometalate bleaching.
4.5.1.1 Polyoxometalate Bleaching
Traditionally, the last remnants of lignin from the pulp have been removed with a chlorine bleaching
process. However, the environmental impacts of chlorine have lead to significant efforts to find
alternative methods to produce a desirable soft white fiber. Among these have been ozone bleaching
and peroxide bleaching. Unfortunately, nothing has come to market which is as effective and selective
as chlorine or chlorine dioxide. Polyoxometalates may be just such a new process. They are highly
selective and can be regenerated within the process. In addition to desirable performance
characteristics, the polyoxometalate system is consistent with the goals of increasing recycling of
process water and reducing the effluent load from pulp mills. Compared to chlorine based systems,
the new process promises to reduce electrical energy consumption of pulp bleaching by 50%.
4.5.2 Chemicals
4.5.2.1 Biological/Chemical Caprolactam Process
Nylon-6 is currently produced from caprolactam. The chemical synthesis of caprolactam from
cumene is a complex, multi-step process that is energy-intensive and generates considerable waste.
Nylon-6 could also be produced from caprolactone. However, the current market price for
caprolactone makes this route uneconomical.
A laboratory-demonstrated biological process has been developed that would provide a one-step,
cost-effective production process for caprolactam manufacture that requires 50% less energy than the
current process, costs half as much (considering both capital and energy costs), and produces almost
no waste byproducts. Research on this process has established the technical feasibility of the
biomanufacturing process for converting inexpensive cyclohexane into caprolactone. Under this
project, the feasibility of the laboratory-demonstrated biomanufacturing process was established, and
4.42
September 18, 1997
The Industrial Sector
Chapter 4
the process is now available to be optimized for possible scale-up to pilot plant scale. It is estimated
that, by the year 2020, this technology can provide annual energy savings of 12 trillion Btu (DOE
1997). While this is a modest total savings (the chemical industry used over five quads in 1991), this is
just one of tens of thousands of chemical processes.
4.5.2.2 Flexible Chemical Processing of Polymeric Materials
Waste textiles and recycled waste materials from automobiles, appliances, and furniture contain
polymers (such as nylon-6, nylon-66, PET, and polyurethanes) that can be converted into valuable
chemical feed stocks. However, processes that can only convert a single type of recycled material can
face high costs for material collection and for transportation of the resulting feed stocks. Because these
costs are the major contributors to process costs, processes are needed that can convert a variety of
recycled materials.
Research in this area is working toward developing a thermochemical process that can convert a wide
variety of recycled materials into valuable chemicals. A two-stage process is envisioned: the first will
use selective catalytic pyrolysis to recover chemicals such as caprolactam, hexamethylendiamine, and
dimethyl-terephathalate; the second will convert the unreacted organic material into synthesis gas,
which can be converted to a variety of chemicals of use to the chemical industry.
Because the process can address a wide variety of recycled materials, large regional recycling plants
can be developed, lowering material collection and transportation costs, and thereby increasing the
viability of recycling many materials. It is estimated that, by the year 2020, the use of this technology
will save 265 trillion Btu annually (DOE 1997).
4.5.2.3 Genetic Engineering
Many chemicals firms are investing heavily in genetic engineering and, over the next decade, many
expect to commercialize products. Low-carbon biotechnologies include engineered plant systems to
allow crops to fix their own nitrogen from the air (thus avoiding N₂0 emissions associated with
fertilizer manufacture); agricultural "petroleum plants" that grow feed stocks for the chemicals
industry; and intermediate products such as polymers.
4.5.3 Petroleum Refining
The National Petroleum Council issued a report in 1995, "Research, Development, and Demonstration
Needs on the Oil and Gas Industry", which identifies the future of the industry in 2020. It stresses,
among other things, the need for flexibility in processes as well as new chemistries and materials.
Changing input feed stocks and environmental requirements will tend to push the industry toward
higher energy use in 2020, without developments such as new catalysts or other process changes that
are on the horizon.
4.5.3.1 Development of Improved Catalysts
The purpose of a catalyst is not to lower the energy needs of a reaction (which are governed by
thermodynamics) but to lower the energy required to activate a process and thereby increase the
kinetics and/or product selectivity. If it accomplishes either or both of these tasks, the energy
demands on a given process should decrease either due to lower heat demand (lower energy of
activation) or from greater throughput. Most of the energy use in a refinery that could benefit from
improvements in catalyst technology is consumed in one of three major process areas: (1)
hydroprocessing, (2) catalytic cracking, and (3) alkylation.
September 18, 1997
4.43
Chapter 4
The Industrial Sector
In hydroprocessing, much energy is utilized in heating up heavy oils and resids to temperatures at
which the catalyst activity is high enough. Additional energy is expended in the compression of
hydrogen to pressures up to 2000 psi. Improved catalysts (capable of functioning at lower
temperatures and pressures) could reduce the energy used by decreasing the reaction temperature of
this process.
Energy usage could be improved for catalytic cracking in terms of product selectivity. Cracking
catalysts are extremely efficient at converting "good" gas oils to gasoline and distillate. However,
when significant fractions of resid and the metals that accompany these resids are used as fluid
catalytic cracking (FCC) feeds, the selectivity (in terms of gasoline yield) drops precipitously. This
gasoline loss comes at the expense of increased coke and dry gas production, which in turn requires
catalyst coolers in order to keep the temperature of the catalyst bed down (required by increased coke
burn) and higher compressor capacity to handle the increased dry gas yield. If catalysts were
designed to handle higher amounts of heavy oils without the detrimental effects outlined above, then
more resid could be handled in the highly efficient FCC with resulting decreased utilization of the less
efficient hydrotreaters.
The largest energy demand in the alkylation units are in the refrigeration units used to keep the
hydrofluoric acid temperature down. Here the need is for a catalyst which will operate at temperature
above ambient. Many solid alkylation catalysts which are in pre-commercial testing and evaluation
function at temperatures around 150°C. Many of the streams requiring alkylation are at or near this
temperature when they exit their respective processing units. Such heat is normally considered waste
heat and thus could easily be utilized for the alkylation process. Therefore, even though the reaction
temperature would go up, the energy demand would decrease.
4.5.4 Glass
The glass industry vision of itself in 2020 is defined in "Glass: A Clear Vision for a Bright Future".
This vision document includes, as one of many goals, reducing process energy use from present levels
to 50% toward the theoretical limit of 2.2 million Btu required to melt a ton of glass. On April 29, 1996
a compact between the DOE and the major glass producing companies was signed to enable
collaboration in such areas as waste reduction, energy efficiency, and quality control. The technology
road map is currently under preparation. The technologies below are just a few examples of areas of
glass industry technology development.
4.5.4.1 Optimizing Electric Boost to Reduce Total Energy Consumption
High energy efficiency, through conversion of electric energy into useful heat, and low volatilization
are the primary advantages of electric melting. Current operating practice has shown that effective
use of electricity near the back end of the furnace, where the batch is added, can reduce fossil fuel
needs. Research needs for optimizing electric boost include, but are not limited to, investigating new
electrode and electric arc melting processes, modeling of the current technology to fine-tune operation
conditions, such as energy inputs and locations of the electrodes, and improving the electrode control
system (Glass Industry Working Group).
4.5.4.2 Recovering and Reusing Waste Heat from Oxy-Fired Furnaces
Recovery and reuse of waste heat from the oxy-fuel process will further increase energy efficiency of
the process. Preheating the batch and cullet, described above, is one method to recover heat from the
flue gas. Other options, such as regenerative oxygen heat recovery (Browning and Nabors 1996) and a
"synthetic air" concept (Argent 1997), have been proposed and need to be tested and evaluated. A
Thermal Swing Adsorption (TSA) oxygen production process has been demonstrated in the laboratory
4.44
September 18, 1997
The Industrial Sector
Chapter 4
with enrichments of up to 89%. The process is based on synthetic chemicals that can reversibly bind
oxygen at low temperatures and release it at elevated temperatures. The operation is in a temperature
range of 70° to 220°F, so low grade waste heat can be used to drive the process, and the external
energy required for produce oxygen can be reduced.
4.5.5 Iron and Steel
"Steel - A National Resource for the Future" broadly defines four areas of R&D to shape the industry
in 2020. These include production efficiency (which encompasses energy efficiency), recycling,
environmental engineering, and product development. The goal of increasing steel production to over
70% of recovered scrap would have major implications for energy use. DOE and the two major steel
industry trade groups have signed a R&D collaborative compact to work together on the first three of
the four research areas. Below, we discuss some of the process areas within which energy and other
savings are likely to be achieved from technical breakthroughs.
Activity will be largely dictated by the viability of different iron making processes that are under
development. R&D effort should focus on developing a process scheme that incorporates both iron
making and steel making into one system with thin strip casting as a final product. The effort should
incorporate a coal-based reductant process which can be coupled with steel making operations and
simultaneously produce power in a combined cycle that includes both gas and steam turbines.
Steel making processes currently utilize computer technology, primarily to implement prespecified
procedures in a timely manner. There is very little feedback in these systems to either enhance process
efficiency or improve the product quality. Key process parameters should be identified so that
interactive logic and high-speed computer systems can be used to control/modify/maintain these
process parameters to obtain a quality product. Such an intelligent-processing approach is essential
for the production of so called "cleaner steel" with low residual elements.
The development of sensors for all aspects of process control and for enabling process changes with a
feedback system is essential for improving process efficiency and optimizing different stages of the
melting, casting, thermomechanical processing, and final heat treatment. Applications of novel ideas
and approaches need to be explored and transfer of technologies available from defense and chemical
processing industries may be a fruitful approach.
4.5.6 Metal Casting
A diverse group of CEOs and presidents from the foundry, die casting, and foundry supply
companies co-authored "Beyond 2000: A Vision for the American Metal Casting Industry." This
vision of the industry identifies six critical areas: production efficiency; recycling; pollution
prevention; application development; process controls; and new technology development. The
specific goals include increasing productivity by 15% and reducing energy consumption by 3-5% by
2010. The Cast Metals Coalition is preparing a R&D strategy to achieve these and other goals
identified in the industry vision. Some examples of technology areas are given below.
Electromagnetic Casting: An electromagnetic field in a casting is used to induce eddy currents in the
liquid metal that, together with the field, stir and contain the liquid metal in the casting. Two
examples are discussed below:
EM Stirring: In continuous casting, the solidification process can be improved by EM stirring,
producing better metallurgical results, improved internal quality of the casting, and even reduced
meniscus instability and surface defects (Beitelman and Mulcahy 1994, Chang et al. 1995). The benefit
September 18, 1997
4.45
Chapter 4
The Industrial Sector
from EM stirring takes the form of reduced wastage per cast. As a minimum, we expect that the
present average yield of 55% for the industry can be increased to 65%, a savings of 130,000 tons per
year, with an associated energy savings of 25 trillion Btu per year (American Foundrymen's Society
1995).
EM Confinement: In the presently dominant sheet-forming process, thick steel slabs are cast and then
hot-rolled. Twin-roll casting with EM confinement has the potential to cast thin sheets by eliminating
the hot-rolling stage, giving the sheet product an enormous economic advantage over products made
by competing methods (Saucedo and Blazek 1994, Blazek et al. 1994) and completely by-passing an
energy-intensive stage of production.
4.6 CONCLUSIONS
This chapter presents an approach to assessing the potential for efficiency to reduce energy use in the
most diverse sector of the economy, the industrial sector; this approach represents a compromise
between the desire for technology detail and the need to evaluate sector-wide energy use. The
approach uses two publicly available models, Argonne's Long-term Industrial Energy Forecasting
(LIEF) model and the Energy Information Administration's industrial module from the National
Energy Modeling System (NEMS), to simulate a plausibly optimistic set of scenarios for additional
energy savings, relative to an established base case (AEO97). The models are used to project what
energy savings could arise from an 'invigorated effort' to put currently available or near commercial
technologies into practice in industry. This invigorated effort is loosely characterized by either a
combination of new policy initiatives or a more serious consideration of efficiency as a strategic
concern of industrial decision makers.
Two efficiency cases are presented in order to project overall reductions in energy use by 2010. A
reduction of 5-10% is projected to be technically feasible, given adequate policies or other incentives to
expand the adoption of cost-effective measures. This is about 2.5 quads in the high case. The LIEF
model projects that these reductions could arise from cost-effective investments defined by a capital
recovery factor of 15% (about a seven year pay-back). The LIEF model does not assume that in every
case all energy-efficiency investments are made, but an increased penetration rate of efficiency
investment is assumed relative to the base case as a result of this 'invigorated effort'. For many of the
energy-intensive industrial sectors, these projected energy savings are consistent with roughly
doubling the current rates of capital stock replacement or doubling the rate of energy technology
efficiency improvement that is currently represented in the NEMS model.
Since the models used to conduct the scenario analysis do not have a detailed, technology-specific
representation of each major industrial sector, the chapter also provides illustrative examples of
technologies for most of the energy-intensive industries. These are examples of technologies that have
the potential to reduce energy use relative to current practices if widely adopted. These technology
examples exhibit substantial energy savings relative to current industry practice, so they reinforce the
fact that the model results are feasible. But one cannot expect these technologies to be adopted widely
unless there is some invigorated effort to encourage their adoption. The slow turnover of the capital
stock in the energy and capital intensive industries is one reason that this invigorated effort would be
needed. Under conservative projections, in the near-term, many of these technologies (and the many
others not listed here) are capable of reaching high levels of penetration but most will not achieve
100% penetration. However, the examples show that there are many ways in which efficiency in
industry can be increased, given the right incentives; the examples help establish the technical
plausibility of the projections.
4.46
September 18, 1997
The Industrial Sector
Chapter 4
The efficiency case projections also show that, on a percentage basis, there are more savings in 'light'
non-energy-intensive industry vs. the 'heavy', energy-intensive sectors. This result arises from the
LIEF model scenarios but, due to the structure of the model, does not have an analog in NEMS.
Because the share of total production costs attributable to energy use in the non-energy-intensive
sectors is very low (the manufacturing average is about 3% and most light industry is less), it is not
surprising that the range of energy performance is quite broad. Energy-efficient technologies, in the
form of motor systems as well as lighting and HVAC options (similar to those discussed in the
commercial section of Chapter 3), represent cost-effective investment opportunities in light
manufacturing. However, there may not have been a managerial or technical focus on energy
efficiency in those industries. An 'invigorated effort' could provide this focus. On the other hand, to
reduce energy use in 'heavy' industry, where considerable attention to efficiency has already been
paid, low capital turnover rates and difficulty in financing medium to large investments may be the
major impediments to accelerated improvements in energy utilization. This 'invigorated effort' in
these sectors might require tax incentives, alternative financing arrangements, new developments that
lower first cost, or demonstration projects that lower perceived risk. The diversity among these broad
categories of industry implies that the mix of policies required to achieve the high-efficiency case may
differ for the various types of industries, based on their current business and technical practices as
well as current domestic and international market conditions.
For all of the industries discussed above, further progress in energy efficiency beyond 2010 requires
further developments in technology. These developments may be incremental improvements (e.g.,
sensors, controls, and system/process modeling) or may be fundamental breakthroughs (e.g., catalysts,
direct smelting, or bioprocessing). Incremental improvements need not be associated with 'small'
efficiency changes. The ability to sense and adjust a process to achieve optimal operating conditions
can have large effects on productivity and energy consumption. However, the search for totally new
methods to produce a product with fundamental breakthroughs in chemistry, metallurgy, or biology
offers another route to enhance productivity and lower energy use. These two avenues of R&D to
create the manufacturing sector of 2020 are both being sought by private and private/public
partnerships.
Table 4.15 summarizes the technology examples presented above. A rough categorization of
incremental (I) and fundamental (F) has been made. Many of the underlying concepts in the examples
apply to other sectors, while others are very process specific. This identification is made as well. In
should be noted that the year 2010 designates current (on very near commercial) technologies, while
the year 2020 designates technologies that will require further R&D, with no prediction of a
commercialization date.
The range and types of technological solutions in industrial applications is quite large. Since energy
represents a cost, and energy efficiency a potential source of profit, these technical solutions can fit
within the economic goals of business. With the right incentives, higher energy efficiency of the
magnitude projected here in the industrial sector is an achievable goal.
September 18, 1997
4.47
Chapter 4
The Industrial Sector
Table 4.15 Summary of Technology Examples
Type of
change:
Concept
Saves fossil
Example Taken
incremental
applicable
or electric
From
Technology Example
Year
or
to other
energy
fundamenta
sectors
I
Aluminum
Improve Furnace Efficiency
2010
I
Y
EF
Aluminum
Materials Recycling
2010
I
Y
E
Aluminum
Improving Hall-Heroult Cell Efficiency
2010
I
N
E
Aluminum
Wettable Cathodes
2010*
I
N
E
Aluminum
Inert Anodes
2010*
I
N
E
Chemicals
Pinch Analytical Techniques
2010
I
Y
F
Chemicals
Advanced Distillation Control Techniques
2010
I
N
F
Chemicals
Flexible Chemical Processing Of Polymers
2020
F
N
F
Chemicals
Biological/Chemical Caprolactam Process
2020
F
N
F
Cross-cutting
Combined Heat and Power
2010
I
Y
EF
Cross-cutting
Motor Systems
2010
I
Y
E
Glass
Glass Batch/Cullet Preheater Technology
2010
I
Y
F
Glass
Advanced Burner Technology
2010
I
Y
F
Glass
Oxy-Fuel Process
2010
I
N
F
Glass
Producing Oxygen More Efficiently
2020
I
Y
E
Glass
Recovering Waste Heat
2020
I
Y
F
Glass
Maximize Combustion Efficiency
2020
I
Y
F
Glass
Optimizing Electric Boost
2020
I
N
F
Iron and Steel
Process Controls
2010
I
Y
EF
Iron and Steel
Hot Connection
2010
I
Y
F
Iron and Steel
Scrap Preheating
2010
I
Y
EF
Iron and Steel
Use Of DC, Rather Than AC, EAFs
2010
I
N
E
Iron and Steel
Coal Or Natural Gas Injection
2010
F
N
F
Iron and Steel
Direct Smelting Reduction
2010
F
N
F
Iron and Steel
Process Controls And Sensors
2020
I
Y
EF
Iron and Steel
Direct Smelting & Thin Strip Casting
2020
F
N
EF
Metal Casting
Computer-Aided Casting Design
2010
I
Y
EF
Metal Casting
Optimized Coreless Induction Melting
2010
I
N
E
Metal Casting
Electromagnetic Stirring
2020
I
N
EF
Metal Casting
Electromagnetic Casting
2020
F
N
EF
Petroleum Refining
Utility System Improvements
2010
I
Y
F
Petroleum Refining
Process/Equipment Modifications
2010
I
N
F
Petroleum Refining
Development Of Improved Catalysts
2020
F
Y
F
Pulp and Paper
On-Machine Sensors For Paper Properties
2010
I
Y
F
Pulp and Paper
Multiport Cylinder Drying
2010
I
N
F
Pulp and Paper
Impuise Drying
2010
I
N
F
Pulp and Paper
Biomass Gasification
2010*
I
Y
EF
Pulp and Paper
Black Liquor Gasification
2010*
I
N
EF
Pulp and Paper
Sulfur Free Pulping
2020
F
N
EF
Pulp and Paper
Polyoxometalate Bleaching
2020
F
N
EF
* Based on the accelerated deployment described in Section 4.3.
4.48
September 18, 1997
The Industrial Sector
Chapter 4
4.7 REFERENCES
Aluminum Association, "Patterns of Energy Usage in the U.S. Aluminum Industry, Full Year - 1989,"
August, 1991.
Aluminum Association. 1993. Aluminum Statistical Review for 1992, Aluminum Association,
Washington, D.C.
American Foundrymen's Society. 1995. Foundry Industry Research Plan.
American Metal Market. 1997. Metal Statistics, 1997: The Statistical Guide to North American Metals,
89th Edition (New York, NY: American Metal Market, Chilton Publications): P. 113.
Argent , R.D. 1997. "Synthetic Air" for Oxy-Fuel Glass Melting Furnaces with Filtration and
Regeneration," Presented at the Annual Meeting of the Society of Glass Technology, January 17, 1997,
Clearwater, FL.
Beitelman, L. and J. A. Mulcahy. 1994. "Flow Control in the Meniscus of Continuous Casting Mold
with an Auxiliary A.C. Magnetic Fields," International Symposium on Electromagnetic Processing of
Materials, EPM'94, Iron and Steel Institute of Japan, pp. 235-241.
Blazek, K. E., H. G. Gerber, and I. G. Saucedo. 1994. "Application of Alternating Magnetic Fields for
Edge Containment in Strip Casting," International Symposium on Electromagnetic Processing of Materials,
EPM'94, Iron and Steel Institute of Japan, pp. 197-202.
Boyd, G., J. Molburg, P. Thimmapuram. 1996. Investigating Cogeneration Potential Using Industrial Fuel
Use Data, draft report, Argonne National Laboratory, Argonne, IL.
Brent, R. And K. Davidson, Market Outlook and Application for Industrial Gas Turbines in Distributed
Generation, paper presented at the Fortieth Annual Engineering and Operations Workshop, American
Public Power Association, Salt Lake City, UT, March 25-28, 1996.
Browning, R. and J. Nabors. 1996. "Regenerative Oxygen Heat Recovery for Improved Oxy-Fuel
Glass Melter Efficiency," Presented at the 57th Conference on Glass Problems, October 8th and 9th,
1996, Columbus, OH.
Burgmann, W. and B.B. Pelts. 1995. "Scrap preheating shaft furnaces - development and results,"
Steel Times International, 19(1):16-17, Jan.
Carmichael, I.F. 1992. "An Introduction to Blast Furnace Coal Injection," Iron & Steelmaker, 19(3):67-
73.
Carroll, Peter, Solar Turbines, Personal Communication, May 1997.
Carroll. P., and K. Davidson, The Role for Advanced Turbine Systems in Atmospheric Pollution Prevention,
presentation to the U.S. Department of Energy, Washington, D.C., May 15, 1997.
Center for Metals Production. 1987. Technoeconomic Assessment of Electric Steelmaking Through the Year
2000, EPRI EM-5445, Electric Power Research Institute, Palo Alto, CA.
Chang, F. C., J. R. Hull and L. Beitelman. 1995. "Simulation of Fluid Flow Induced by Opposing AC
Magnetic Fields in a Continuous Casting Mold," Process Technology Conference Proceedings, Vol. 13,
Iron and Steel Society, pp. 79-88.
September 18, 1997
4.49
Chapter 4
The Industrial Sector
De Beer, J.G., M.T. van Wees, E. Worrell, and K. Blok. 1994. The Potential of Energy Efficiency
Improvement in the Netherlands up to 2000 and 2015, Utrecht University, Utrecht, Netherlands.
Decision Analysis Corporation, NEMS Industrial Module: Methodology to Incorporate the ABMA Boiler
Database, December 14, 1995.
Dillich, S., Personal Communication, May 27, 1997.
Eisenhauer, J.et al, Report of the Aluminum Technology Roadmap Workshop, November 19-20, 1996,
Alexandria, Virginia. Energetics, Inc. Columbia MD, February 1997.
Emad, F. et al. 1996. "In-Line Inspection of Carbon Anodes for Use in Aluminum Production," Light
Metals 1996.
Energetics, Inc. 1990. Industry Profiles Final Report: Energy Profiles for U.S. Industry, prepared for
the Office of Industrial Technologies, U.S. Department of Energy, Columbia, MD.
Energetics, Inc. 1997. Aluminum Technology Roadmap Workshop, Final Draft, Report of the
Aluminum Technology Roadmap Workshop, November 19-20, 1996, Alexandria, Virginia, Columbia,
MD.
Energy and Environmental Analysis, Inc., Industrial Market Evaluation, Topical Report, Analysis of the
Industrial Boiler Population, prepared for the Gas Research Institute, Report # GRI-96/0200, June 1996,
Chicago, IL.
Energy Information Administration, 1997, Manufacturer's Energy Consumption Survey, 1994. MECS
1997 U.S. DOE, Washington D.C. preliminary tables available on the Internet.
http://www.eia.doe.gov
Energy Information Administration (ELA). 1994. Manufacturing Consumption of Energy: 1991,
DOE/ELA-0512(91), Washington, D.C.
Energy Information Administration (ELA). 1996. Emissions of Greenhouse Gases in the United States,
DOE/ELA-0573(95), Washington, D.C.
Energy Information Administration (ELA). 1997. Annual Energy Outlook 1997, with Projections to 2010.
DOE/ELA-0383(97). Washington, D.C.
Etienne, A. and W.R. Irving. 1985. "The Status of Continuous Casting," in Institute of Metals, 1985,
Continuous Casting '85, conference proceedings May 22-24, 1985, London, UK.
Evans, J.W., "Electricity in the Production of Metals: From Aluminum to Zinc", Metallurgical and
Materials Transactions B. Volume 26B, April 1995 PP 189-208.
Gas Research Institute, Final Draft of the 1998 Edition of the GRI Baseline Projections, Arlington, VA,
September, 1997.
Gas Research Institute, Light Duty Vehicle Full Fuel Cucle Emissions Analysis, 1994.
Gee, J.T. et al, "Long-Term Testing and Evaluation of Cathode Components in a Commercial
Aluminum Cell," August 1989, Great Lakes Research, DOE/IE/12689-1; Church, K.D. et al,
Addendum DOE/ID/12689-1 (vol-Add.).
Geiger, G., ed. 1996. "Glass Problems Conference Focuses on Oxy-Fuel," The American Ceramic Society
Bulletin, Vol. 75, No. 3, March.
4.50
September 18, 1997
The Industrial Sector
Chapter 4
General Electric, 1997, secondary source from Solar Turbines and Onsite Energy Presentation to the
U.S. Department of Energy, May 1997.
Gibbs, M. and C. Jacobs. 1996. "Reducing PFC Emissions from Primary Aluminum Production in the
United States," Light Metal Age.
Glass Industry Working Group. Adapted from discussion in the Energy Efficiency and Conservation
Subcommittee.
GRID. 1996. "License Granted to Corning, Inc.," Summer.
Haissig, M. 1994. "The d-c shaft furnace," Iron and Steel Engineer, May, PP. 25-27.
Haupin, W.E., "Principles of Aluminum Electrolysis," Light Metals, 1995.
Heffron, J., R. Hewertson, and E. Riley. 1993. "Benefits to Aluminum Furnaces Through Oxygen-
Assisted Melting," in Proceedings of the 8th International Sheet & Plate Conference, Louisville, KY,
October 5-8, 1993, Aluminum Association, Washington, D.C.
Hogan, W.T., and F.T. Koelble. 1996. "Fewer blast furnaces, but higher productivity," New Steel,
November.
Horwath, J. A., T. Klemp III, and J. M. Svoboda. 1996. "Variables Identified for Optimal Coreless
Induction Melting", Modern Casting, May, PP. 33-35.
International District Energy Association, Encouraging Supply-Side Energy Efficiency in Federal
Restructuring Legislation, White Paper, February 20, 1997 Washington D.C..
Jeltsch, R., and T. Franklin. 1992. "Retrofit of Kaiser's Mead Smelter," in Light Metals, Proceedings
of the 121st TMS Annual Meeting, San Diego, CA, March 1-5, 1992, Minerals, Metals, & Materials
Society, Warrendale, PA.
Kenchington, H, et al, Reduced Greenhouse Gas Emissions through Advanced Aluminum Cell
Technology. DOE proposal, April 1997.
Larson, E.D. 1991. Biomass-Gasifier/Gas-Turbine Cogeneration in Pulp and Paper Industry. ASME
paper International Gas Turbine and Aeroengine Congress and Exposition, Orlando, FL. June 3-6,
1991.
Lensen, D. H. 1996. "Survey Provides Profile Casting Design Software Use", Modern Casting,
September, PP. 29-31.
Lensen, D. H., C. Beckermann, and G. W. Fischer. 1995. "Implementation Issues for Computer-Aided
Casting Design", Summary Report to American Metalcasting Consortium.
Livesay, P., "Strength Characteristics of Portland-Limestone Cements," Construction and Building
Materials, 5(3), PP. 147-150, 1991.
Lympany, S.D. and J.W. Evans: Metallurgical and Materials Transactions B., 1983, vol. 14B, pp. 63-70.
Major, Graham, Learning from experiences with small-scale cogeneration, Centre for the Analysis and
Dissemination of Demonstrated Energy Technologies, CADDET, 1995, The Netherlands.
Major, W and K. Davidson, Gas Turbine Power Generation: Environmental Analysis and Policy
Considerations, DRAFT Report by Onsite Energy Corporation prepared for Oak Ridge National
Laboratory, February 19, 1997a.
September 18, 1997
4.51
Chapter 4
The Industrial Sector
Major, W. and K. Davidson. 1997b. "Gas Fired Power Generation: Environmental Analysis and
Policy Consideration," draft report by Onsite Energy, Carlsbad, CA, prepared for the U.S. Department
of Energy, Office of Industrial Technology.
Major, W., Basis for 60 GW of Remaining Cogeneration Potential in the Industrial Sector May 22, 1997.
Margolis, Nancy., Personal Communication. 1997.
Mathur, V.K. date? Thermal Swing Absorption Process for Oxygen Separation from Air, DOE/CE40927-3,
Prepared for U.S. Department of Energy, Office of Industrial Technologies.
Miller Freeman, Inc., 1972-1997 North American Pulp and Paper Fack Books, San Francisco.
Nelson, Kenneth E. 1993. "Creating an Empowered Conservation Culture," Proceeding of the Workshop
on Partnerships for Industrial Productivity through Energy Efficiency, PP. 209-224, September 19-22,
Portland, OR.
Newsted, G., H. Meyer, R. Hawkins, and J. Johnson. 1992. "Twenty-Five Years of Progress at
Intalco," in Light Metals, Proceedings of the 121st TMS Annual Meeting, San Diego, CA, March 1-5,
1992, Minerals, Metals, & Materials Society, Warrendale, PA.
Onsite Energy, ATS Market Assessment, 1994.
Onsite Energy. 1997. Gas Turbine Environmental Analysis and Policy Considerations.
Parks, W., Personal Communication, May 16,1997.
Parks, William, Personal Communication, July 17, 1997
Portland Cement Association, Economic Research Department, Skokie, Illinois, 1996d.
Portland Cement Association, Potential Reduction of CO2 Emissions in the Manufacture of Portland
Cement (R&D No. 2010), Research and Development Department, PCA, Skokie, Illinois, 1996a.
Portland Cement Association, Properties of Concretes Made with Fly Ash and Cements Containing
Limestone (R&D Serial No. 2052a), Research and Development Department, PCA, Skokie, Illinois,
1996b.
Portland Cement Association, The Reduction of Resource Input and Emissions Achieved by Addition
of Limestone to Portland Cement, Research and Development Department, PCA, Skokie, Illinois, 1996.
Portland Cement Association, The Use of Limestone in Portland Cement: A State-of-the-Art Review
(PCA Serial No. 2025a), Research and Development Department, PCA, Skokie, Illinois, 1996c.
Reed, J., Personal Communication, May 21,1997.
Richards, N., 1994, "Strategies for Decreasing the Unit Energy and Environmental Impact of Hall
Heroult Cells," Light Metals 1994.
Romm, Joseph, Personal Communication, July 17, 1997
Ross, C.P. 1996. "Oxy-Fuel Conversion Challenges for Glass Manufactures," Presented at American
Flame Research Committee Meeting, May 6-7, 1996, Orlando, FL.
Ross, M. 1990. "Capital Budgeting Practices of Twelve Large Manufacturers", in Advances in Business
Financial Management, Ed. Philip Cooley, Dryden Press, Chicago.
4.52
September 18, 1997
The Industrial Sector
Chapter 4
Ross, M., P. Thimmapuram, R. Fisher, W. Maciorowski. 1993. Long-Term Industrial Energy Forecasting
(LIEF) Model (18-Sector Version), ANL/EAIS/TM-95, Argonne National Laboratory, Argonne, IL.
Saucedo, I. G. and K. E. Blazek. 1994. "Development of an Electromagnetic Edge Dam for Twin-Roll
Casting," METEC-94, pp. 457-462.
Sauer, G. "Cement, Concrete and Greenhouse Gas" paper presented at the CGLI Second Roundtable
on North American Energy Policy, April, 1997.
Scheidig, K. 1995. "Hot metal from oxygen cupola furnaces as an alternative charge material for
electric arc furnaces," Stahl und Eisen, 15 May, 115(5): 59-64.
Steinmeyer, Dan. 1997. Proceedings of the 1997 ACEEE Summer Study, July 8-11, Saratoga Springs
NY.
Tresouthich, S.W. and Mishulovich, A. "Energy and Environmental Considerations for the Cement
Industry, Proceedings of the Energy and the Environment in the 21st Century Conference,
Cambridge, MA, 1990.
U.S. Department of Agriculture (USDA). 1990. An Analysis of the Timber Situation in the United States:
1989-2040. Technical Report RM-199.
U.S. Department of Energy (DOE). 1987. A Guide to Investment Appraisal of Energy Efficiency Projects in
the Steel Industry, Energy Efficiency Office, Washington, D.C.
U.S. Department of Energy (DOE). 1995. Energy Conservation Trends: Understanding the Factors
Affecting Energy Conservation Gains and Their Implication for Policy Development, DOE/PO-0034,
Washington, D.C.
U.S. Department of Energy (DOE). 1997. 1996 Chemicals Team Annual Report, Office of Industrial
Technologies, Washington, D.C.
U.S. Department of the Interior, Bureau of Mines (DOI). 1993. Aluminum Availability and Supply: A
Minerals Availability Appraisal, Information Circular 9371, Washington, D.C.
U.S. Department of the Interior, Bureau of Mines (DOI). 1994. Mineral Commodity Summaries: 1994,
Washington, D.C.
U.S. Environmental Protection Agency (EPA), 1996. Primary Aluminum Industry: Technical Support
Document for Proposed Maximum Available Control Technology (MACT) Standards.
U.S. Environmental Protection Agency, 1995, Profile of the Nonferrous Metals Industry, EPA 310-R-95-
010.
Weyerhaeuser, Stone and Webster, Amoco, and Carolina Power and Light. 1995. New Bern Biomass to
Energy Project: Phase 1. Feasibility Study. NREL/TP-42-7942
Wilken, L and R. Brent, Advanced Turbine Systems (ATS): A Partnership of the U.S. Department of
Energy and Solar Turbines Inc. To Develop and Commercialize the Premier Power Source for the 21st
Century., Solar Turbines, 960294/296, San Diego CA, 1996.
Windisch, C. and D. Strachan. 1991. Inert Electrodes Program, Fiscal Year 1990 Report, PNL-777,
Battelle Pacific Northwest Laboratory, Richland, WA.
World Energy Council (WEC), 1995. Energy Efficient Improvement Utilizing High Technology: An
Assessment of Energy Use in Industry and Buildings, London
September 18, 1997
4.53
Chapter 4
The Industrial Sector
ENDNOTES
1 Because they become very important in a low-carbon scenario, we have made an exception to this
approach in the case of low-carbon technologies which are examined from the bottom-up. Unlike
energy-efficiency technologies, there is no established modeling procedure or analysis method for
assessing the penetration of low-carbon (especially low process carbon) technologies. Thus, we simply
provide case studies and take their results as a lower bound on the potential.
2 Differences in the industry subsector detail prevented us from doing a precise calibration of the two
models. It is not clear that such a calibration would substantially improve our analysis for these
purposes.
3 Technologies that supplement the existing process (e.g., process controls) might penetrate rapidly,
but most that are replacements for existing process will more likely follow the 'normal' turnover
patterns. Technologies achieving rapid penetration include sensors and process control software and
technologies that can save significant amounts of energy.
4 The forest products industry includes pulp and paper, as well as lumber and wood products. This
report focuses on the relatively more energy-intensive pulp and paper segment of the forest product
industry.
5 In particular, we do not change the underlying forecast for activity in the refining sector.
6 The HE/LC case also includes low-carbon technologies described in Section 4.4. These low-carbon
technologies are not explicitly captured in the computer model runs.
7 Not all energy-intensive industries are "heavy." The metal casting industry, for example, consists of
many small shops owned by small businesses. It is important to distinguish OIT's Industries of the
Future and the "heavy industry" of Table 4.1. These "heavy" industries include non-vision industries
such as food (SIC 20), non-refining petroleum (SIC 295, 299), stone, clay and cement (SIC 324-329),
non-aluminum non-ferrous (SIC 3331, 3339, 3351, 3356, 3357, 3364, 3366, 3369). Conversely, the vision
industries include wood and lumber (SIC 24), miscellaneous paper (SIC 265, 267), and miscellaneous
chemicals (SIC 283, 285, 2879, 289), which are not included in the "heavy" industry of Table 4.1.
8 These numbers are calculated from the numbers in Appendix A for Figure 16 in DOE (1995).
9 Note that, in Figure 4.3, ATS in simple cycle power-only mode has a higher heat rate than the most
efficient combined cycle turbine. However, even the power-only ATS emits far less CO₂ than the
existing sources of power to the grid. In our calculation of the carbon reductions for ATS in power-
only applications, we make the conservative assumption that the ATS has about the same emissions as
new power plants. The carbon savings in this market come only from avoided T&D.
10 As an average of 3 to 18 MW units (9 to 80 MW in multiple units), ATS are 43% efficient in simple
cycle and can be 80 - 85% efficient (combined thermal and electric efficiency) when used for
cogeneration (Hoffman, 1997).
11 Constraints to growth of cogeneration have derived in part from the traditional requirement that
steam and electricity loads be matched to maintain efficient and cost-effective operation of the
cogenerator. The new ATS overcomes this problem by running efficiently at a wide range of
electricity to thermal ratios. Cogeneration has also been constrained by environmental permitting,
utility regulation, and utility competition. Together these factors explain why very efficient CHP
technology still comprises a relatively small fraction of electricity and steam generation. When
policies to promote CHP are instituted, however, this fraction can grow dramatically (Major 1995).
Finland, Denmark, and the Netherlands have each achieved a contribution of about 30% of electricity
production based on CHP.
4.54
September 18, 1997
The Industrial Sector
Chapter 4
12 In the recuperator configuration hot exhaust gas from the turbine is used to preheat the air leaving
the compressor prior to combusion, thereby reducing the amount of fuel required to reach the design
turbine inlet temperature.
13 See Appendix D-3 for details.
14 Such reform would include standardized permits to reduce costs for small sites and a life cycle
approach that takes into account power plant emissions and T&D for on site power permits.
15 Current regulations, and some proposed utility deregulation legislation, include barriers to small on-
site CHP. Both scenarios assume policies that elicit the cooperation of utilities in the increase in on-site
generation. One scenario that could be imagined is that the utilities themselves finance and service
these industrial ATS's for CHP and power generation.
16 Carbon reductions from fuel switching were not included - a conservative assumption discussed in
Apendix D-3.
14The calculations performed for both scenarios were reviewed by the American Forest and Paper
Association and industry representatives (David Cooper and Delmar R. Raymond).
18 The makeup of residual biomass and residue generation rates in various forest product and paper
industries are described in Appendix D-5.
19 Three large plants in the U.S. manufacture adipic acid and they are working with the EPA to reduce
emissions (Boyd 1997). However, emissions from this process have increased by 9.9% since 1990.
20 Many of the U.S. aluminum smelters are located in regions of the country with large hydropower
resources, notably the Pacific Northwest (served by the Bonneville Power Administration), the
Southeast (served by the Tennessee Valley Authority), and Northern New York. Under both scenarios,
the aluminum smelters that would likely be converted first would be those in regions such as the Ohio
River Valley that are dominated by coal-powered plants.
21 0.2 MtC = [(5 plants) * (190,000 tonnes of AL/plant)* (13,200 kWh/tonne of AL) * (89 gr of carbon
per kWh) * 0.17]/1,000,000 grams/tonne.
= The carbon savings from the aluminum industry efficiency improvements that are already included
in Section 4.2.2 must be subtracted in order to identify the increment that can be added by this analysis
of alternative aluminum production cells. We calculate this as follows. Table 4.2 estimates that the
aluminum industry as a whole will be 2% more efficient in its use of electricity in 2010 under the
efficiency scenario, compared to the business-as-usual case. Under the assumption here that 5 of 22
smelters (23%) will be retrofitted with wettable cathodes that offer a 17% improvement in electricity
efficiency, the nation's aluminum smelters as a whole would be 3.9% more efficient (3.9% = 0.23 *
17%). This represents a 1.9% efficiency improvement (or 0.09 MtC of emissions reductions) over the
2% that is assumed in the efficiency case in Section 4.2.2.
23 0.48 = (5 of 22 plants) * (50%) * 4.2 MtC.
24 1.0 MtC = [(10 plants) * (190,000 tonnes of AL/plant)* (13,200 kWh/tonne of AL) * (160 gr of carbon
per kWh) * 0.25]/1,000,000 grams/tonne.
25 The carbon savings from the aluminum industry efficiency improvements that are already included
in Section 4.2.2 must be subtracted in order to identify the increment that can be added by this analysis
of alternative production cells. We calculate this as follows. Table 4.2 estimates that the aluminum
industry as a whole will be 3.8% more efficient in its use of electricity in 2010 under the high-
efficiency/low-carbon scenario, compared to the business-as-usual case. Under the assumption here
that 10 of 22 smelters (45%) will be retrofitted with inert anodes that offer a 25% improvement in
electricity efficiency, the nation's aluminum smelters as a whole would be 11.4% more efficient (11.4%
= 0.45 * 25%). This represents a 7.6% efficiency improvement (or 0.67 MtC of emissions reductions)
over the 3.8% efficiency improvement that is assumed in Section 4.2.2.
September 18, 1997
4.55
Chapter 4
The Industrial Sector
26 0.6 MtC = (10 plants) * (190,000 tonnes of Al/plant) * (0.33 tonnes of C/tonne of AL).
27 1.91 MtC = (10 of 22 plants) * (100%) * 4.2 MtC.
28 Note that almost none of the carbon savings due to Section 4.2.2's high-efficiency/low-carbon
(HE/LC) case should be subtracted from this amount. The clinker replacement is not an efficiency
increase, but a demand reduction. Thus, the drop in carbon emissions is nearly additive. We
subtracted a very small amount because of the 8.1% reduction in cement industry energy use under
the HE/LC scenario. The carbon savings due to a 5-20% reduction in demand is 2-4 MtC depending
on the kiln technology. The amount we subtracted is about 5% of the total (0.06-0.08 MtC). If we
assume that the penetration by 2010 is limited to the European owned firms (roughly half), then the
carbon reduction is 1-2 MtC. Note that the HE/LC energy efficiency savings of Section 4.2.2 are
equivalent to a carbon emissions reduction of just over 1 MtC depending on cement kiln technology.
29 The Greenville Tube Company (GT) realized non-energy benefits ten times greater than the energy
benefits when the company upgraded its motors. GT is a manufacturer of high-precision, small-
diameter, stainless steel tubing. GT replaced an old motor and inefficient eddy current clutch drive
with an energy-efficient motor with vector control. This new motor required fewer runs and
produced far less scrap than the old system. The motor reduced annual energy consumption by 37%
and resulted in savings of more than $77,000 annually from increased productivity, reduced scrap
generation, and reduced energy costs.
4.56
September 18, 1997
Transportation Sector
Chapter 5
Chapter 5
TRANSPORTATION SECTOR
5.1 INTRODUCTION
The trend of more than a decade of continuous energy-efficiency improvements in transportation,
marked by a sharp decoupling of energy consumption and economic growth, appears to have come to
an end. The transportation sector's energy use now appears to be growing at nearly the same rate as
the gross domestic product (GDP).
From 1949 until 1973, energy use in the U.S. transportation sector grew at an average annual rate of
3.6% per year (ELA, 1996a, Table 2.1). In the years following the oil crisis of 1973-74 until the oil
price collapse of 1986, that rate fell to only 0.6% per year.¹ This sharp decrease in growth was
caused by a combination of market and non-market factors - sharply rising oil prices and, perhaps
more important, strong expectations that prices would continue to escalate for the foreseeable future;
threats of gasoline rationing and actual (though largely government-caused) local gasoline
shortages; successes in government-sponsored R&D, especially in aeronautics; and new regulations,
particularly the Corporate Average Fuel Economy (CAFE) standards for automobiles and light
trucks. Some manifestations of the decrease in the growth of energy use during this period were:
Between 1973 and 1988, new passenger cars increased their fuel economy from about 14 MPG to
28.6 MPG (EPA rated) (Heavenrich and Hellman, 1996, Table 1), a rate of 5% per year.
During 1970-1987, commercial aviation decreased in energy intensity from 10,351 Btu per
passenger-mile to 4,753 Btu/pm (Davis and McFarlin, 1996, Table 2.16) again at an average rate
of 5% per year.
During 1970-1994, the energy intensity of rail freight decreased from 691 Btu/ton-mile to 388
Btu/ton-mile, or 44% (Davis and McFarlin, 1996, Table 2.17), a rate of 2.4% per year.
Although changes in travel behavior, choice of vehicle size, changes in vehicle occupancy rates and
other non-technological factors have a role in the rate of growth in transportation energy use,
improved technological efficiency has been the most critical factor in energy trends. For example,
had energy intensities not changed since 1972, commercial airlines would be using over twice the
energy they use today (assuming today's number of passenger-miles of travel), and three quarters of
the savings are due to technological improvements in aircraft (Davis and McFarlin, 1996, Table
2.21). Similarly, examination of the causes of substantial fuel economy gains by automobiles during
the 1970s and 1980s show that the majority of the gains were achieved by improving technical
efficiency, not by consumers moving to small cars. Between 1978 and 1984, only 7.8% of the period's
MPG gain was achieved by shifts to smaller cars (Westbrook and Patterson, 1985). Between 1976 and
1989, the combination of weight reduction, improved transmissions, tires, and aerodynamics,
widespread use of fuel injection, various engine improvements, improved lubricants, and wider use of
front wheel drive accounted for about 70% of the total 8.4 MPG improvement during the period
(Westbrook, 1989). In fact, the technology of automobiles has improved so much over the past few
decades that if the 4,000 pound plus, 15.8 MPG automobile of 1975 were to be built with today's
technology but without any change in weight or horsepower, it would get 26.4 MPG (Greene and Fan,
1994)! And although 85% of the improvement in rail freight energy efficiency came from increased
loadings per car, much of the 85% resulted from improved communications and computing capability
September 17, 1997
5.1
Chapter 5
Transportation Sector
(other factors included changing composition of freight during this period and other operational
improvements), and improved vehicle technology accounted for the remaining 15% (Greene, 1996).
Over the past ten years (1986-1996), the rate of growth of transportation energy use has averaged
1.6% per year, but in the past three years it accelerated to 2.2% per year, just below the rate of
growth of GDP. Transportation energy efficiency, which improved significantly during the decade
of the 1980s, appears to be stagnant (U.S. DOT/BTS, 1996, p. 87). The average fuel economy of new
passenger cars has not improved significantly over the past decade. The average fuel economy of
light-duty vehicles, new cars, and light trucks combined has not changed significantly since 1982
(Heavenrich and Hellman, 1996, Table 1) and, as a consequence, the average on-road fuel economy of
the entire on-road light-duty vehicle fleet was only 1% higher in 1995 (the most recent year for
which data are available) than in 1991 (U.S. DOT/FHWA, 1996, Table VM-1). Gasoline prices are
now at pre-1973 levels and fuel economy standards have not been raised over 1985 levels. There are
exceptions, however: commercial air travel and rail freight continue to make meaningful efficiency
gains (U.S. DOT/BTS, 1996, P. 101). Overall, the transportation sector appears to have entered a
period of growth in activity only slightly slower than that of GDP with only modest gains or no
improvement in energy efficiency.
Despite these recent trends, the 1997 Annual Energy Outlook (AEO97) reference case forecast to 2015,
which serves as the backdrop for this analysis, foresees very slow growth in transportation energy
use (1.4%/yr.) accompanied by virtually no change in the prices of transportation fuels (0.2%/yr.).
A modest rate of growth in vehicle travel (1.4%/yr.) together with MPG gains of 5.1 MPG for new
passenger cars and 3.7 MPG for new light trucks over 1995 levels, combine to hold the growth of
light-duty vehicle energy use to 1% per year through 2015. Every year since at least 1989, the AEO
(among others) has forecasted continued light-duty vehicle fuel economy gains yet the actual fuel
economy of light-duty vehicles as a whole has not improved. In some cases, energy prices have
turned out to be lower and in other cases higher than expected. Apparently, technology that could
have been used to improve fuel economy is either not being implemented, or is being used to provide
some other feature that consumers value, such as performance. We expand on this point below in
explaining why, in our "business-as-usual" (BAU) case, we forecast no improvement in light-duty
vehicle fuel economy. We believe that, given low energy prices, plentiful oil supplies, no market
disruptions, and no new energy policy initiatives, it is optimistic to expect continued energy-
efficiency improvement and slow growth of energy use.
Current policy initiatives and activities to increase future transportation energy efficiency are
relatively modest. Except for light-duty highway vehicles, the federal government does not
regulate transportation fuel efficiency. The National Highway Traffic Safety Administration has
the power to raise CAFE standards for autos and light trucks, but there seems little chance that it
will do so at the present time. The Energy Policy Act contains provisions to move alternative fuel
vehicles into the fleet (fleet vehicle requirements and altfuel tax credits), but these provisions are
limited, and congressional support for coercive action is nonexistent. On the other hand, there are
important R&D initiatives that could play a role in improving transportation fuel efficiency,
particularly the long-standing NASA and Defense Department programs in aeronautic design and
the Partnership for a New Generation of Vehicles (PNGV), a joint government/industry research
effort aimed primarily at developing vehicles with up to three times current fuel economy levels.
The newest of federal initiatives aimed at improving transportation fuel efficiency, PNGV has
reorganized and redirected the federal government's R&D effort in advanced automotive
technologies towards the ambitious goal of tripling automotive fuel economy and reducing pollutant
emissions while at the same time preserving consumer amenities and holding down costs. Current
PNGV spending is on the order of $250 million dollars (the exact amount is subject to debate because
of definitional problems of which efforts are actually dedicated to PNGV goals) (U.S. Congress,
5.2
September 17, 1997
Transportation Sector
Chapter 5
OTA, 1995), with the largest government share coming from DOE's Electric and Hybrid Vehicle
Program. Current PNGV thinking seems aimed at an advanced hybrid-electric vehicle, with
research efforts aimed particularly at advanced materials, high-power energy storage devices, fuel
cells and improved engines, lean NOx catalysts (to allow necessary emission control for lean-burn
engines including diesels and direct injection stratified charge engines), and improved electric
drives, including power electronics.
In this chapter, the potential for these and other energy-efficient and low-CO2 technologies to
cost-effectively reduce transport sector greenhouse gas emissions is examined. Three transportation
sector scenarios were developed using the Energy Information Administration's (ELA) National
Energy Modeling System (NEMS) model, AEO97 version (see Overview of Methodology box), with
reference case assumptions about macroeconomics and energy prices (Decision Analysis Corp., 1996;
ELA, 1994). These are labeled the -(1) "business-as-usual" (BAU), (2) "efficiency" (EFF), and (3)
"high-efficiency/low-carbon" (HE/LC) cases. Our business-as-usual case differs from the AEO97
reference case only in that new light-duty vehicle fuel economy is held constant at current levels
throughout the period of the forecast. In the reference case, it improves at an average annual rate of
0.4%.
The efficiency and high-efficiency/low-carbon scenarios differ from each other less in effort than in
outcome. In our view, the improvements postulated in the efficiency scenario are likely to be
forthcoming if appropriate policy measures are undertaken and research efforts intensified. In
contrast, because the outcomes postulated in the high-efficiency/low-carbon scenario require
technological breakthroughs, they require a certain degree of luck to be achieved by 2010. There are
no credible methods to accurately gauge the probability of such breakthroughs; we believe they
stand a decent chance of occurring with an intensification of research efforts, but we stop short of
claiming that they are a likely outcome of such an intensification. In other words, the efficiency
scenario represents what is often called a "most likely" or "probable" scenario, in the authors'
judgment The high-efficiency/low-carbon scenario is better described as an "optimistic" or
"possible" scenario. However, both are predicted on a major intensification of R&D effort plus
significant policy measures aimed at pushing the market towards giving fuel efficiency a much
higher priority.
The efficiency scenario is created by assuming earlier introduction of advanced fuel economy
technology and by adding certain key technologies that are absent from the AEO97 reference case. It
assumes the introduction of advanced ethanol-from-biomass technology in 2005, technology which
the U.S. DOE is currently intensively involved in developing. In the efficiency case, technology
development is incremental rather than revolutionary. Nonetheless, the efficiency case does
presume a major energy technology R&D effort, perhaps two to ten times the level of current
government programs. It also assumes that policies necessary to draw energy-efficiency technology
into the market are implemented, as needed. In other words, effective policy actions, whether they
be increased fuel economy standards, revenue neutral feebates, fuel taxes, public information or some
other initiative, are assumed to have been put in place. This point is critical, because AEO97
forecasts inexpensive, plentiful fossil fuels, and because the goal of preventing global climate
change is a classic public good that markets on their own will generally ignore.
September 17, 1997
5.3
Chapter 5
Transportation Sector
Overview of Methodology
Producing scenarios for this analysis comprised three principle steps: (1) developing assumptions about future
advances in energy technology for transportation, (2) entering these assumptions into an integrating model to
predict their market acceptance and impact on transportation energy use and, (3) adjusting the model's predictions
for analyses and forecasts done "on the side.". Because of time and budget constraints; no attempt was made by the
transportation sector team to integrate our scenarios with those of other energy using sectors to produce an
economy wide scenario. The methodology is therefore a partial analysis of the effects of technology on the
transportation sector, assuming no interaction with other sectors of the economy.
Obviously, there is no sure.way to predict the evolution of technology. Thus, the key to developing a useful
technology scenario ris clearly documenting assumptions, and also demonstrating that the assumptions are
consistent with recent advances in technology by referencing published scientific and technical reports. Wherever
possible, we base ur-assumptions on objective technology assessments, such as the Office of Technology
Assessment's (1995) examination of the potential for advanced automotive technology. The result of this step is a
list of specific technologies with the following data for each, (1) date of initial market introduction, (2) quantitative
impact OIT energy efficiency (e.g., % fuel economy improvement over a baseline vehicle) and, (3) incremental cost to
the buyer.
We, used the Energy Information Administration's (EIA's) National Energy Modeling System (NEMS),
Transportation Sector Model as a. tool for integrating the technology assumptions and predicting their impact on
energy use. NEMS is undoubtedly the most fully documented (U.S. DOE/EIA, 1994; 1995a, 1995b, 1996a, 1996b),
most rigorously peer reviewed (e.g., NRC.) and most thoroughly tested comprehensive, national energy model. The
NEMS Transportation Sector Model comprises a set of submodels for each transport mode that range in complexity
from the highly detailed light-duty vehicle model to much simpler models for waterborne transport and rail freight.
The NEMS light-duty vehicle model requires an itemization of each technology, as well as its applicability to each
of six:passenger car and six light truck rclasses. In addition to introduction date, cost: and fuel. economy
improvement potential, interactions (incompatibilities, complementarities, etc.) among technologies must be carefully
specified. NEMS predicts market penetrationrover time based on cost-effectiveness and time since introduction, but
also by applicability and interactions with other technologies These predictions reflect normal requirements for
testing of new technologies, as well as turnover of the stock of manufacturing capital The Freight Truck and Air
Travel Models also require a list of technologies,"introduction dates and efficiency improvement estimates, but
market penetration is handled somewhat more mechanistically. In both cases, new technologies are introduced
when the price of fuel crosses a threshold price For Rail Freight and Waterborne Freight, one must directly specify
a rate of efficiency improvement.
Given technology assumptions, othersmacroeconomic inputs; and the energy and economic, predictions of the
1997 Annual Energy Outlook Reference Case Projection, the NEMS ransportation Model predicts new vehicle
sales, used vehicle scrappage,1 vehicle utilization and fuel iconsumption? by: model and vehicle type These
endogenous predictions are sensitive to economic variables. For example, improving energy efficiency will result in
some degree of increased vehicle travel due to the lower cost of fuel permile. Based on the composition of demand
by fuel type, NEMS also forecasts. carbon, emissions, as well *For the largest transport modes, the evolution of
vehicle stocks over time are explicitly calculated "in:great detailstire generalz theitechnological characteristics of a
vehicle are determined in the year in which it is manufactured NEMS' exhaustively accounts for the numbers
of vehicles by type, class, and vintage, as well.as which technologies.have been applied to these vehicles. As a
result, NEMS's representation of the dynamics of technological change are quite meticulous
IF
int
Finally two parts :of the scenario ana weresdone "off line thus&necessitating some straightforw
adjustments to the NEMS forecasts. The estimation of market: supply: and demand for cellulosic ethanol as a
blending component of conventional gasoline was:calculated by-means taispreadsheet model. The supply and
demand studies upon whichthis analysis were:based were simplystoo/recently.prodiced (they are based on draft
reports). to have already been incorporated by the ETA' into the NEMS model structure at the time. Also we chose
to introduce advanced direct injection diesel passenger cars andilight trucks,using the NEMS model algorithm for
conventional gasoline vehicles rather than as alternative fuelvehicles This reflects belief that the new
advanced diesels will bealmost tinguishable from gasolinervehicles fromithe consumers:/perspective (with the
exception of their cost and fuel economy variables the NEMS model takes intolaccount): drawback of this choice
is that we sacrificed the NEMS model's ability to automatically account.for the additional diesel use and,
therefore, had to adjust for: after the fact. Several additional calculations were:made based on NEMS outputs.
The NEMS model's output.includes the market penetration of each technology by vehicle type and class. Using this
information together with the input assumptions about technology costs and fuel reconomy improvement we were
able to compute measures of the overall cost-effectiveness of the sum total of all technologies applied to passenger
cars and light trucks.
5.4
September 17, 1997
Transportation Sector
Chapter 5
The high-efficiency/low-carbon case begins with the efficiency case assumptions and then goes
beyond incremental technological advances and postulates breakthroughs in fuel cell technology for
light-duty vehicles, as well as major aerodynamic and engine efficiency gains for commercial
aircraft, among other selected technological achievements. It also includes more optimistic
assumptions about biomass ethanol production costs. It is not the intent of this scenario to include all
possible technological advances, but rather to focus on a few that could have major long-run
implications for greenhouse gas emissions from the transportation sector. We could, as well, have
assumed technological breakthroughs for battery-electric or compressed or liquefied natural gas
vehicles, both of which have some potential to reduce carbon emissions compared to petroleum-
based fuels. The more breakthroughs one assumes, however, the lower the probability that the
scenario will actually occur. Furthermore, in the long-run, no single technology appears to have a
greater potential to reduce carbon emissions from transportation than the fuel cell. We do not assume
a target and tradeable permit system equivalent to $50/T of carbon in the high-efficiency scenario.
We do assume in both scenarios that significant policies similar to this are in place to encourage
producers to produce and consumers to choose fuel-efficient, low-carbon technologies.
Although the focus of this study is on the year 2010, forecasts to 2015 are also presented because
changing the technology of transportation energy use takes more than one decade. Once a technology
is market ready, two to three years of testing and certification are still required prior to
introduction. Even then, most technologies will not appear on all makes and models simultaneously
due to the need to replace plant and equipment in an efficient manner. Finally, expected lifetimes
for transportation vehicles are counted in decades. The median expected lifetime of a passenger car
is now 14 years, truck lifetimes average 16 years, marine vessel and aircraft life expectancies are at
least twice that (Davis and McFarlin, 1996, Tables 3.6 and 3.7). Thus, the full impact of
technologies introduced between now and 2010 will not be apparent in 2010. We include the year
2015 to illustrate this fact. In all cases, a normal rate of replacement of capital stock is assumed,
both in the production of transportation vehicles and in their purchase and scrappage. That is to
say, no changes are made to the NEMS model to accelerate the turnover of capital stocks.
Results of the three scenario projections are compared with ELA's AEO97 projections in Table 5.1. In
the business-as-usual case, transportation energy use grows from 25.5 quads in 1997 to 32.3 quads in
2010 and to 34.0 quads in 2015. Emissions of carbon increase as well, up 26% in 2010 and 33% higher
by 2015. The efficiency scenario achieves roughly a 10% reduction in energy use and a 12% reduction
in transportation sector emissions versus the business-as-usual case by 2010. Reductions in 2010 versus
the AEO97 reference case are slightly less, 7% for energy and 9% for carbon emissions. Use of
cellulosic ethanol as a blending component in gasoline reduces greenhouse gas emissions by 2-3% over
and above the reduction in energy use. The greatest reductions in fossil fuel use are achieved by rail
freight (-16%), light-duty highway vehicles (-12%), and commercial air travel (-11%). Energy use
in 2015 is actually below that of 2010 in the efficiency scenario because of the greater penetration of
new, efficient equipment into the stocks of transportation vehicles. Transportation uses 28.2 quads of
energy, 17% below the business-as-usual case but still 10% over 1997 levels. Emissions of carbon are
down by 20% over the business-as-usual case, still 6% higher than in 1997. The high-
efficiency/low-carbon scenario reduces energy use and carbon emissions by another 4% in 2010 and by
an additional 5% in 2015. By 2015, transportation sector carbon emissions are projected to be below
the 1997 level in the high-efficiency/low-carbon scenario.
September 17, 1997
5.5
Chapter 5
Transportation Sector
Table 5.1 Comparison of Three Transportation Energy Scenarios to the AEO97 Reference Case
Energy Use (quads)
1997
2010
2015
Business-as-Usual
25.5
32.3
34.0
Reference Case
25.4
31.4
32.3
Efficiency
25.4
29.2
28.6
High-Eff/Low-Carbon
25.3
27.8
26.4
Carbon Emissions (MtC)
1997
2010
2015
Business-as-Usual
487
616
646
Reference Case
485
598
614
Efficiency
485
543
532
High-Eff/Low-Carbon
484
513
485
Note: Carbon emissions include emissions from the generation of electricity for electric vehicles. Reference case
assumptions about electric vehicle market penetration have not been changed in any of the three scenarios.
Similarly, transportation energy use includes electricity generation losses.
We wish to emphasize that, in our judgment, the reductions in carbon emissions described in these
scenarios are unlikely to be achieved by advances in technology alone, in the absence of meaningful
additional policy measures to insure that cost-effective and near cost-effective technologies to
improve energy efficiency and to expand the production of biomass fuels are in fact implemented.
This is not only our conclusion. The 1995 Asilomar Conference on Energy and Sustainable
Transportation, organized by the National Research Council (NRC), Transportation Research
Board's Committees on Energy and Alternative Fuels, addressed the question, "Is technology enough
to achieve sustainable transportation?" The conference's consensus, to be published in a forthcoming
volume of proceedings, was that technologies capable of creating a sustainable transport system
could be developed over a reasonable time period but that the marketplace on its own would be
unlikely to adopt such technologies in the absence of specific policy measures to make it happen
(McNutt et al., 1997). Because of the inertia inherent in the nation's transportation system, and
because reducing greenhouse gas emissions is a public good, meaningful policy action is likely to be
essential to achieving the carbon emissions reductions described in these scenarios.
We also believe that research and development of low-carbon emission technologies will have to be
expanded to achieve the results of the efficiency and high-efficiency scenarios. Support for this
view can be found in the NRC's just-published review (NRC, 1997) of the research program of the
PNGV, the most significant national effort to advance technology to improve transportation energy
efficiency. The views of the standing committee charged with reviewing the progress of the
program are unambiguous:
"The PNGV is experiencing severe funding and resource allocation problems that will
preclude the program from achieving its objectives on its present schedule if they are
not resolved expeditiously."
The panel comments on the serious underfunding of PNGV in at least nine different places in its
report. In Table H-1, summarizing its assessment of the status and prospects for the key PNGV
5.6
September 17, 1997
Transportation Sector
Chapter 5
technologies, all technologies save fuel cells were categorized as having a basic need for additional
resources. Noting that PNGV has been unresponsive in providing the committee with estimates of
the funding that would be required, the committee notes that the industry consortium of the PNGV
stated that it would like to see government funds available to PNGV doubled (NRC, 1997, p. 107).
Elsewhere, the committee notes that funding for ultracapacitor research would have to be increased
by at least ten times for a period of 10 to 15 years in order to catch up with the status of battery
research with respect to PNGV goals. While the technological progress assumed in our efficiency
case does not require that PNGV goals are attained, continued advances by industry and government
R&D programs will be essential. PNGV, of course, addresses only light-duty vehicles. R&D
support for low-greenhouse gas technologies for other modes is even more modest. In the view of the
transportation sector analytical team, substantial additional funding for R&D will be required,
perhaps two to ten times what is presently being spent, depending on the area of investigation.
5.2 PROVEN AND ADVANCED TECHNOLOGIES
Despite the fact that the fuel economies of successive model years of U.S. new cars and light trucks
have been essentially constant for the past decade (Heavenrich and Hellman, 1996), technologies
positively affecting vehicle efficiency have continually entered the fleet. These include fuel
injection, 4-valve per cylinder engines, 4-speed electronically controlled automatic transmissions
with lockup, growing use of lightweight materials and structural redesign for weight reduction, tires
with lower rolling resistance, and improved aerodynamics. Efficiency improvements offered by
these technologies have been counteracted, however, by increased acceleration performance and top
speed; weight increases due to increased body stiffness and more power and safety equipment (e.g.,
air bags); and other factors. In other words, auto makers and purchasers have been willing to trade
off fuel economy for competing vehicle amenities such as weight and power.
There is wide agreement that new efficiency technologies will continue to enter the fleet, and that
technologies recently entered will gain market share. Table E.1 in Appendix E lists those
technologies that appear in the NEMS data base and are expected to either gain market share or
enter the market during the next decade or so. With a few exceptions, these are proven technologies
whose costs and impact on efficiency can be reliably specified. The most important of these
technologies, from the standpoint of their potential impact on fleet fuel efficiency during the next
few decades, are described briefly below. Documentation for costs and projected fuel efficiency
improvements for these and the other technologies in the NEMS data base is contained in Energy
and Environmental Analysis, Inc. (1994).
5.2.1 Material Substitution
Weight reduction has been a key factor in the U.S. automobile fleet's fuel economy improvement
since the early 1970s, and will likely play an important role in future improvements. Past weight
reductions involved a combination of a widespread conversion to front-wheel drive, which
eliminated the drive shaft and rear axle and allowed important packaging gains; a significant
downsizing of the fleet, made possible by changing consumer demands; the shift to unit body
construction from a chassis on frame structure; and material substitution, largely from plain carbon
steel to high strength low alloy (HSLA) steels, but also including shifts to plastic parts and some
aluminum as well. Recently, structural redesign using supercomputers has allowed significant
weight savings. However, much of these savings have been taken back by increases in body rigidity,
which enhances ride quality and safety, as well as the addition of safety and power equipment.
Accordingly, the average weight of the fleet has begun to increase.
September 17, 1997
5.7
Chapter 5
Transportation Sector
Despite past improvements, there remain substantive possibilities for large weight reductions
without sacrificing vehicle interior space or safety. The Office of Technology Assessment (U.S.
Congress, OTA, 1995)2 identified an array of weight reduction scenarios including the following: a
"clean sheet" design using advanced steel alloys that might achieve greater than a 10% weight
reduction in a mid-sized auto; all-aluminum vehicles using successively more optimized designs
achieving up to a 30% reduction; and a technically-optimistic design using polymer composites
achieving a 35-40% reduction (though OTA considered this last scenario to be quite uncertain from a
commercial standpoint because it requires breakthroughs in manufacturing technology).
Material substitution is treated in a series of steps in the NEMS model, with each step representing
a 5% weight reduction relative to the baseline. The first step (now complete in the current new car
fleet) represents increased use of HSLA, while the next four steps represent increasing use of plastics
and aluminum over time, to achieve a total reduction of 20% relative to a modern 1990 vehicle (more
with older non-unit body designs).
5.2.2 Aerodynamic Drag Reduction
Improvements in vehicle aerodynamics have been an important part of the overall fuel economy
improvement of the U.S. light-duty vehicle fleet, with average drag coefficients (Cds) being
reduced from 0.45-0.50 in 1979/1980 to between 0.30 and 0.35 today, with some models in the 0.27-0.29
range. These reductions are important to vehicle fuel economy because a 10% reduction in Cd
typically will yield a 2.0-2.5% increase in fuel economy at constant performance.
Prototypes with extraordinarily low Cds (e.g., 0.18 for the Chevrolet Citation IV and 0.15 for the
Ford Probe IV ("Going with the Wind," 1984)) have been shown, and the General Motors EV1
electric car attains a Cd of 0.19. There is a strong consensus among auto makers, however, that mass
market vehicles will likely be limited to Cds of about 0.25 because of limits on the practical slope of
windshields, need for cargo space (low Cds require tapered rear ends), and other factors, including
customer design preferences. Further, reductions in Cds for light trucks are limited by factors such as
need for high ground clearance and large tires, open beds in pickup trucks, and so forth. Also, the
short length of subcompact autos limits the degree to which their Cd can be reduced.
In NEMS, aerodynamic drag reduction is also implemented in a series of steps starting from a 1990 Cd
baseline of 0.37, with each step representing a 10% reduction over the previous level (i.e., to 0.33,
0.30, 0.27, and 0.245, respectively).
5.2.3 Improved Automatic Transmissions
A range of potential improvements to automatic transmissions can offer fuel economy benefits of up to
about 6% in automobiles. Key areas of improvement are design changes that reduce hydraulic losses
in the torque converter and transmissions with added numbers of gears, with continually variable
transmissions possible.
Five-speed automatic transmissions were introduced in Japan and Europe a few years ago and have
recently been introduced to the United States in a few luxury models. Nissan and Mercedes have
experienced fuel economy gains over a 4-speed automatic in the 2-3 MPG range (Hattori et al., 1990).
A number of continuously variable transmissions (CVTs) have been tested with widely varying
results, and Suburu sells a small car with a CVT in the U.S. market. OTA estimates that a CVT
should be capable of achieving approximately a 6% fuel economy increase over a 4-speed automatic.
5.8
September 17, 1997
Transportation Sector
Chapter 5
Electronic transmission control of both conventional automatic transmissions and CVTs will add some
benefits over the older mechanical controls. First generation controls selected only the shift points
and provided about 0.5% benefit in fuel economy, and such controls were in most transmissions by
1995. More advanced second generation controls have appeared, and they interact with the engine
control to optimally select torque converter lock-up and shift points while also determining engine
calibration. Such controls provide 1.5% benefit over mechanical controls.
5.2.4 Engine Friction Reduction
Reducing mechanical friction is an ongoing process in engine development, and steady reductions in
friction have occurred as engine designers continually modify existing engines and introduce new
engine families. There is substantial potential for fuel economy gains as existing friction reduction
improvements are rolled into the fleet. Primary areas for further improvement are:
Piston and connecting rod weight reduction using lightweight materials,
Lightweight valves and valve springs,
Use of two rings instead of three,
Improved oil pumps,
Improved lubricants,
Low friction crankcase seals, and
Roller cam followers.
Only roller cam followers and two-ring pistons are discrete technologies, with specific benefits of 2%
in fuel economy, while other benefits are based on design evolution.
Fuel economy improvements of as much as 4.5% (compared to current engines) should be available
using the full range of evolutionary technologies. The NEMS model has separate representation of
roller cams, while all other technologies are modeled as engine friction reduction in discrete steps of
1.5% benefit in fuel economy, with steps in the order of increasing cost and complexity.
5.2.5 Variable Valve Timing
In conventional engines, the timing and extent of opening of the intake and exhaust valves are fixed,
and are compromises between the very different needs of high and low power settings. Variable
valve control allows substantial efficiency improvement; for example, closing the intake valves
early can substitute for throttling to reduce air intake, thus reducing pumping losses at low load.
Also, variable valve control boosts engine power, allowing engine downsizing while maintaining
power levels.
Honda uses a system called VTEC that controls both lift and timing of intake and exhaust valves.
VTEC is not a fully variable system, offering only two settings for valve timing and lift, but it still
obtains an 8% fuel economy improvement at constant performance. It has been used in the U.S.
market both for boosting power (Acura NSX, Prelude VTEC) and improving fuel economy (Civic VX).
Although VTEC was introduced to the U.S. market in 1991 (in the NSX), neither VTEC nor
competing systems (Mitsubishi uses a system, MIIVEC, that combines valve control with cylinder
September 17, 1997
5.9
Chapter 5
Transportation Sector
shutdown at low loads) have gained significant market share since then. The major concerns are cost
and complexity. Second generation VVT systems that offer wider control of lift and timing are
expected to increase fuel economy benefits at constant performance to 10%.
5.2.6 Lean-Burn Engines
Lean-burn engines reduce engine power by reducing fuel flow without throttling back airflow, thus
increasing the air/fuel ratio; in contrast, conventional engines maintain air/fuel ratios at or below
"stoichiometric" (i.e., the ratio - about 14.6:1 - where there is just enough air to fully combust the
fuel). Aside from the reduced pumping loss obtained by foregoing throttling, engine thermal
efficiency is increased and hydrocarbon and carbon monoxide emissions are reduced. The primary
challenges facing lean-burn engines are difficulties in maintaining stable combustion at high
air/fuel ratios and the need to develop new NOₓ catalysts that will work in an oxygen-rich exhaust
environment. The former challenge generally is handled by designing the
cylinder/piston/valves/fuel injector system and operation in such a way as to stratify the fuel
charge so the region around the spark plug has a richer fuel mixture than in the rest of the
combustion chamber and ignites readily. An alternative method is to use high swirl combustion
chambers that promote combustion. For the emissions challenge, most automobile manufacturers are
working to develop "lean NOₓ catalysts," and, as discussed below, both Toyota and Mitsubishi have
sold vehicles that combine lean operation and new NOₓ catalyst technology since the early 1990s in
Japan.
Low cost lean-burn systems that do not need "direct injection" of fuel into the cylinder head can
provide up to a 10% benefit in fuel economy by utilizing advanced cylinder head designs and lean
air-fuel sensors.
5.2.7 Advanced Tires
Rolling resistance accounts for approximately a third of the loads on an automobile during the EPA
test procedure. The magnitude of this resistance is approximately linearly related to the rolling
resistance coefficient of the vehicle's tires, so reducing this coefficient through changes in tread
design, tire materials, and tire structure will have a significant positive impact on fuel economy.
Tire design and materials have improved steadily throughout the years, with the switch to radials
from bias-ply tires beginning in the late 1970s, then the shift to second generation radials beginning
in the mid-1980s each achieving about a 20-25% reduction in rolling resistance and a 3-4%
improvement in fuel economy.
Additional improvements have recently been introduced by Michelin and other companies and are
beginning to penetrate the fleet. Use of these and other, further-improved designs can yield about a
25% reduction in rolling resistance by 2005, with 5% improvement in fuel economy resulting; an
additional 3% fuel economy improvement may be possible by 2015 (Hattori et al., 1990). Some of
these gains are likely to be offset by manufacturer design decisions that increase tire traction and
durability, so that only about half the potential fuel economy gains are likely to be realized. The
NEMS model has the improvements occurring in four discrete steps over time to achieve a total 4%
benefit in fuel economy.
Aside from these proven technologies, there are a few additional technologies that are not expected
to enter the fleet in commercially significant amounts before 2010 under the business-as-usual case
assumptions, but that have the potential to impact fleet fuel economy in this time frame if there are
appropriate incentives. These are:
5.10
September 17, 1997
Transportation Sector
Chapter 5
Advanced drag reduction (to a Cd of 0.22 for mid-sized vehicles),
Hybrid-electric power trains,
Direct injection stratified charge (DISC) gasoline engines,
Direct injection (DI) diesel engines, and
Proton exchange membrane (PEM) fuel cell power trains.
All but the PEM fuel cell power trains are considered likely to be introduced into the U.S. in small
numbers before 2010 (e.g., in limited edition or luxury models). In fact, Volkswagen has already
introduced DI diesel engines into the U.S. market as options in its Passat, Jetta, and Golf models. DI
diesels cannot meet current NOx standards for gasoline-fueled automobiles. At this time, diesels
have an exemption to U.S. rules on NOx emissions; however, this exemption is unlikely to stand if
large numbers of diesels are sold in the U.S. market. Similarly, DISC engines have been introduced
into the Japanese fleet by Toyota and Mitsubishi, but their high cost and U.S. emissions
requirements should keep them out of the U.S. fleet for the immediate future - except perhaps in
very limited numbers. As discussed below, however, these technologies could make an impact on
U.S. fleet fuel economy before 2010 either in the efficiency scenario, which postulates both increased
R&D spending and increased market or regulatory incentives for fuel economy, or in our high-
efficiency/low-carbon scenario that postulates better-than-expected luck in technology
development.
5.2.8 Advanced Drag Reduction
In our view, significant market pressure on fuel economy could reduce Cd values a bit further than
projected by the auto makers. Some existing vehicle designs that have attained lower Cds without
some of the design compromises of the prototypes noted above indicate that a Cd of 0.22 should be
practical for a mid-size car without requiring wheel skirts or a sharply tapered rear end.³ This
value has been adopted as successfully entering the mass market automobile fleet in both the
efficiency and high-efficiency/low-carbon scenarios, and is modeled as an additional 10% reduction
in drag over the lowest Cd value in NEMS of 0.245.
5.2.9 Hybrid-Electric Power Trains
Hybrid-electric power trains combine two energy sources with an electric drivetrain, with one or
both sources providing electricity to the electric motor. Although many configurations are possible,
all have some form of energy storage (battery, flywheel, ultracapacitor, etc.). Hybrids offer a
theoretical efficiency advantage over conventional internal combustion engine (ICE) drivetrains for
the following reasons:
They offer the potential to recapture some of the vehicle's potential energy that is normally
lost (as heat) when the vehicle is braked. In a hybrid, the electric drive motor can be operated
in generator mode to brake the vehicle; the electric energy produced is stored in the battery or
other storage device.
The hybrid drivetrain allows the vehicle powerplant to be smaller and to operate more
efficiently than the powerplant in a conventional drivetrain. In a conventional drivetrain, the
engine is sized for the maximum load (usually short-term rapid acceleration) and can produce
many times the power it uses during the great majority of its operation. For example, during
idle, low speed cruise, or deceleration, the powerplant may be operating below 10% of its
September 17, 1997
5.11
Chapter 5
Transportation Sector
maximum power capability, and most engines (especially gasoline engines) are very inefficient
at such lower power levels. Because the storage device can absorb any excess power (over that
needed to operate the vehicle) produced by the engine, the engine can continue to operate at an
efficient power level even when the vehicle loads are low. Also, in a hybrid, the storage
device can provide part of the power for maximum acceleration, allowing the hybrid
powerplant to be sized for average power requirements or for power requirements in operations
where the battery can't help (e.g., during sustained hill-climbing), which are generally lower
than acceleration loads - so the hybrid's engine can be smaller.
The net energy gains from the regenerative braking, smaller and lighter powerplant, and improved
powerplant cycle efficiency are counteracted by losses in the electrical components (storage device,
generator, motor/controller) and their added weight (in particular, weight of the storage device
and electric motor). The wide variety of hybrid configurations and component designs, the
relatively early stage of development of hybrid powertrain systems, and the ongoing redesign of
hybrid powertrain components to satisfy the unique requirements of hybrid operation has yielded a
wide range of estimates of the potential efficiency benefits of shifting to hybrid drivetrains.
Further, ongoing changes in engine design for conventional drivetrains shift the relative value of
hybridization, with reduction in pumping losses achieved by variable valve control, for example,
reducing the benefit of hybridization because these are the same losses hybridization is designed to
counter. The OTA has estimated that a battery/ICE hybrid can achieve about a 25-35% gain over a
conventional drive vehicle with the same type of powerplant, assuming what it considered
optimistic values for the efficiencies of the battery and electric motor (U.S. Congress, OTA, 1995).
Current examples of operating hybrids that satisfy normal vehicle safety and performances
requirements have not achieved efficiency improvements this high (U.S. Congress, OTA, 1995). On
the other hand, the Department of Energy's (DOE's) goal for its hybrid drivetrain R&D program is
a doubling of fuel economy, and theoretical analyses of hybrid configurations using simulation
models have projected gains ranging as high as the DOE goal (Burke, 1995; Ross, 1996). In our view,
gains this high are unlikely without sacrificing some aspects of performance or operational
flexibility. On the other hand, there are active R&D efforts on hybrid components such as
ultracapacitors and high-efficiency electric motors that, if successful, could raise the efficiency
advantage of hybridization to somewhat higher levels than OTA projects. The efficiency case
conforms approximately to the OTA projections; the high-efficiency/low-carbon case assumes
exceptional success at improving drivetrain components and reducing costs. This translates to a 28%
fuel economy benefit over a 1995 conventional gasoline-fueled car, and a 10% benefit over a DI diesel
vehicle for the efficiency case; in the high-efficiency/low-carbon case, the assumed gains are 43%
and 23%, respectively.
The primary barriers to successful commercialization of hybrid-electric vehicles are the current
high costs of electric motors, controllers, and batteries, and the need for additional progress in
reducing the specific power and increasing the efficiency of these electrical components. In
particular, there is an urgent need for reliable high-efficiency, high specific power batteries. There
recently has been progress on such batteries, but considerable work remains. In addition, there are
relatively few suitable engines in the right size category (one liter or so) for hybrids, since
automotive engines typically are sized to meet the higher power requirements of conventional
drivetrains.
5.2.10 Direct Injection Stratified Charge (DISC) Gasoline Engines
Conventional spark ignition (gasoline) engines are inefficient at part load in large part because they
reduce power by throttling back on their air supply, creating large drag losses (so-called "pumping
losses") in the stream of intake air. Direct injection stratified charge engines do not throttle intake
air; instead, they reduce only fuel flow at part load, operating at fuel/air ratios as low as 1:50.
They manage this by injecting fuel directly into each cylinder at high pressures (700 psi or higher
5.12
September 17, 1997
Transportation Sector
Chapter 5
compared to 50 psi in a conventional fuel injection system (Markus, 1997)) in such a way that the
fuel/air mixture is stratified (thus, "stratified charge"), with high fuel concentrations near the
spark plug so as to maintain stable combustion. The combination of zero throttling losses, low fuel
use at light loads because of the very lean fuel mixture, and some added benefits of direct injection -
particularly, more precise control of combustion and fewer problems such as fuel condensation on
intake-port walls - yields substantial fuel efficiency improvements rivaling those of DI diesels.
Concerns with DISC engines include problems with increased NOₓ emissions because normal
reduction catalysts will not operate in the oxygen-rich exhaust environment of a lean-burn engine;
the expense and durability of the fuel injectors, which have to operate at very high pressures
ranging up to 2000 psi; and the need for extremely precise control of combustion to maintain smooth
performance from the engine as it shifts back and forth between lean to stoichiometric operation.
Both Toyota and Mitsubishi have introducing DISC engines into their fleets in Japan, Mitsubishi
with a 1.8 liter, 148 hp engine in its Galant sedan and Legnum wagon, Toyota with a 2.0 liter, 143 hp
engine in its Carina sedan (Markus, 1997). Both companies use catalysts to reduce NOₓ emissions:
Mitsubishi's is a true lean-NOx catalyst that reacts hydrocarbons with NOₓ to form nitrogen,
oxygen, water, and carbon dioxide; Toyota's system stores NOₓ and reduces it to nitrogen during high
power operation when the engine uses a stoichiometric (no excess air) air/fuel mixture (Markus,
1997). Neither system is believed ready to meet U.S. emissions requirements, especially for
catalyst longevity. The Toyota system would likely experience difficulties with high levels of
sulfur in U.S. fuels, which can poison the catalyst material.
Available data suggest that Toyota's DISC engine provides a 25% fuel economy benefit in the
Japanese 10-mode cycle, which could translate to an 18% benefit in the U.S. FTP if emissions
problems are solved. This benefit has been used in the efficiency case; in the high-efficiency/low-
carbon case, a benefit of 23% is assumed.
5.2.11 Turbocharged Direct Injection (TDI) Diesel Engines⁵
Until recently, all diesel powertrains used in light-duty vehicles in the United States were indirect
injection diesels (IDI). In an IDI diesel, fuel is sprayed into a prechamber, mixed with air, and
partially burned before the charge is passed into a main combustion chamber where the combustion
continues. This design was desirable for automobiles because it yields smoother combustion with less
noise and lower NOₓ emissions than direct injection designs. These advantages are purchased at the
expense of some efficiency losses from heat transfer from the prechamber and pressure losses as the
partially burned gases flow through the passages between the prechamber and main combustion
chamber.
Advances in fuel injection technology and combustion chamber design, coupled with turbocharging
and intercooling, have allowed direct injection diesels to attain smoothness and noise levels
comparable to IDI diesels with low NOₓ emissions and high specific power (power/weight) levels,
approaching that of naturally aspirated 4-valve per cylinder gasoline engines. The best 4-valve
turbocharged DI diesels can attain fuel economy improvements of 40% or more over current 2-valve
per cylinder engines, though conversion to gasoline equivalent fuel economy yields closer to a 30%
gain (diesel fuel is a more energy-dense fuel than gasoline). The 40% value has been used in our
analysis, but it assumes that lean-NOx catalysts will be successfully adapted to diesels to meet
NOₓ standards. Catalyst researchers generally are considerably less optimistic about success for
diesels than they are for gasoline-fueled vehicles.
September 17, 1997
5.13
Chapter 5
Transportation Sector
As noted above, Volkswagen has introduced DI diesels into the U.S. fleet in its Golf, Jetta, and
Passat models. These engines are 1.9 liter and produce 105 horsepower. Audi produces a larger 2.5
liter engine for its European models.
5.2.12 Proton Exchange Membrane (PEM) Fuel Cell Powertrains
Fuel cells are electrochemical devices that convert the chemical energy in fuels to electrical energy
directly, without combustion. This process avoids the thermodynamic limitations imposed by the
Carnot cycle, and fuel cells theoretically can have efficiencies of 90% or greater. With hydrogen as
a fuel, fuel cells have emissions only of water; with fuels such as methanol or hydrocarbons,
reforming to obtain hydrogen will produce small quantities of carbon monoxide and other pollutants
as byproducts and larger quantities of carbon dioxide.
For the immediate future, PEM fuel cells appear to be the clear choice among alternative fuel cell
technologies for light-duty vehicle applications because they operate at moderate temperatures
(20-120 degrees C) and developers have been able to rapidly improve their power density (from .085
kW/liter in 1989 to about 1 kW/L today) and decrease their costs (platinum loadings, a major cost
factor, have been reduced from about 4 mg/cm² in 1990 to current levels of about 0.15 mg/cm2) (Oei,
1997).
Despite rapid progress, fuel cells must overcome major hurdles before they can succeed commercially
in the light-duty market. Costs must be sharply reduced. Even with mass production, PEM fuel cells
would cost at least $200/kW to manufacture with today's production technology and cell designs -
nearly ten times the cost of ICE engines (Oei, 1997), disregarding the additional cost of needed
hydrogen storage or reformers.6
Key needs are development of low-cost membranes, size and cost reduction of hydrogen reformers or
onboard storage, and improvement of "balance of plant." Also, there are several "engineering" issues
that will have to be dealt with once stack design has gotten to the point where serious vehicle
design is contemplated - for example, cooling (the low temperature operation of fuel cells means
that the heat being rejected is very low grade heat, requiring lots of air movement or large radiator
surface areas, neither very appealing to vehicle designers (Borroni-Bird, 1997)) and prevention of
freezing in cold weather.
On-board fuel storage represents a significant barrier because hydrogen's energy density is very low,⁷
and the easiest fuel to reform into hydrogen onboard the vehicle, methanol, has no significant
supply infrastructure. Chrysler in partnership with DOE recently announced significant progress
towards onboard production of hydrogen from gasoline, which would solve the supply infrastructure
problem and allow much easier fuel storage than hydrogen. Not surprisingly, however, the
selection of gasoline as the preferred "hydrogen carrier" for fuel cells is by no means an easy call.
For example, gasoline's availability and easier fuel storage must be traded off against the cost and
space occupied by the reformer (Jost, 1997).⁸ Toyota has claimed a substantial improvement in
hydrogen storage technology using an advanced metal hydride adsorbent that matches the energy
density of liquefied hydrogen storage with only 10 atmospheres of pressure required (Yamaguchi,
1997). Presumably, however, this type of storage would be extremely heavy. Other options being
pursued by various researchers include direct methanol fuel cells, which preclude the need for a
reformer, and the use of ethanol in place of methanol or gasoline as a hydrogen source. The latter
option is especially attractive if the ethanol can be produced from cellulosic materials, because the
effect on reducing greenhouse gas emissions is particularly large for this technology.
5.14
September 17, 1997
Transportation Sector
Chapter 5
We expect the rate of progress and probability of commercialization of fuel cell powertrains to be
sensitive to the level of R&D funding and market pressures to improve overall vehicle fuel economy.
Progress has in fact been rapid, as shown by the improvements in power density discussed above.
Ford, GM, and Chrysler are all pursuing fuel cell vehicle R&D, as are Japanese and European
companies, with Toyota's and Mercedes Benz's programs being the most visible. A Canadian
company, Ballard, appears to be in a leading position in PEM fuel cell R&D, and has supplied
systems to most of the vehicle R&D programs. Given current funding levels and the market's lack of
pressure on fuel economy levels as well as the large amount of development work that remains to be
done, however, introduction of fuel cells into mass market vehicles appears likely to be beyond the
2010 time frame, and the base scenario adheres to this projection. This, in fact, was the conclusion of
the NRC's advisory panel overseeing the PNGV program (NRC, 1997, Table H-1). On the other
hand, increased funding and market pressure and/or particularly fortuitous progress in the ongoing
R&D program might move the date of introduction forward. Further, the newness of the technology
and the dependence of the basic fuel cell stack costs on manufacturing design leaves open the
potential that the eventual cost of the fuel cell system might be somewhat lower than competing
ICE drivetrains; this depends on substantial cost reduction over a range of technologies, because the
costs of hydrogen storage or reforming, the electric motor, and even the battery that is likely to be
necessary for startup power, all play a significant role in total system costs.
The efficiency case assumes that fuel cells will not be introduced in mass market vehicles before
2010; we note that the major auto makers are not projecting a pre-2010 commercial introduction of fuel
cell vehicles even assuming a high level of success in their development programs. The PNGV
program envisions that the earliest fuel cells will use a reformer to produce hydrogen from gasoline.
We assume that fuel cells in conjunction with a gasoline reformer will be about 70% more efficient
than current gasoline engines, but only slightly more efficient than a diesel hybrid drivetrain. The
high-efficiency/low-carbon case assumes introduction of commercial gasoline fuel cell vehicles by
2007. Although ethanol from cellulosic material would make an excellent fuel for the fuel cell
hybrid and would result in further reductions in greenhouse emissions, we assume the first fuel cell
vehicles will use widely-available gasoline.
5.2.13 Fuel Cells in Heavy Trucks and Locomotives
In many ways, fuel cell propulsion may be attractive for large transportation vehicles, such as
locomotives or ships, before it is ready for use in light-duty vehicles. Use of fuel cells in heavy
trucks will require a breakthrough in hydrogen production, distribution, or on-board storage, or else a
breakthrough in reforming technology before it will be competitive with the diesel engine. The
drive-cycle thermal efficiency of current heavy-duty diesel truck drivetrains is in the range of 35%
to 40%. The drive-cycle thermal efficiency of current methanol steam-reforming fuel cell
drivetrains (including electric motor/controller and battery) is also 35% to 40%. Thus, there is
likely to be little incentive for heavy trucks to switch to fuel cells until hydrogen fuel cells, with
drivetrain efficiencies in the range of 45% to 50%, become available.
Fuel cells may succeed in the locomotive market first because, (1) fuel costs are more important to
rail carriers than to truckers, (2) locomotives already use electric traction drive and, (3) fuel cells of
the size necessary for locomotive powerplant output (4000 HP) are already commercially viable in
stationary powerplant applications. Therefore, we consider the use of fuel cells in locomotives in
the high-efficiency/low-carbon scenario. Fuel cells may also have applicability to marine vessels,
again because of their size. We do not introduce marine fuel cell applications in the high-
efficiency/low-carbon technology scenario, simply because we are not aware of suitable
applicability studies. We believe that analysis of the potential for fuel cell technology in heavier
transportation vehicles would likely reveal additional promising applications. Thus, the
locomotive fuel cell analysis is intended more to be indicative of potential large-scale fuel cell
applications in transportation than a reflection of our judgment of the true potential market.
September 17, 1997
5.15
Chapter 5
Transportation Sector
Locomotives can be broadly classified into two types: local service and line haul. Local service
locomotives are primarily older line-haul locomotives that are low powered (2000-3000 HP), and
are typically utilized in light load applications. Local service locomotives consume about 120,000
gallons/year of fuel per locomotive. Line-haul locomotives are more powerful (4000 HP) and
consume 375,000 gallons/yr of fuel per locomotive. Both types spend considerable amounts of time at
idle, over 70% of the time for local service locomotives and over 50% of the time for line-haul. Idle
fuel consumption accounts for 38% of fuel consumed in local service locomotives, and only 6.3% of fuel
consumed for line-haul locomotives (CARB, 1991; CARB, 1992).
Locomotives have very long useful lives and the engines are rebuilt several times. The diesel engine
alone costs about $400,000, but a complete rebuild costs only about $100,000 to $150,000. Engines are
typically rebuilt every eight years, and the entire locomotive is rebuilt every 24 years and/or
moved to local service at that point. Hence, a useful life calculation of 24 years may be reasonable in
terms of a replacement cycle.
A fuel cell based locomotive could utilize methanol, ethanol, or liquefied natural gas (LNG) and
could be of the PEM type being considered for cars or the Phosphoric Acid type used for power
generation. It is believed that for high power applications, the Phosphoric Acid Fuel Cell is in a
relatively greater state of maturity. Several large units are currently operating as prototypes for
power generation. Estimates of future fuel cell costs are highly uncertain, but megawatt size
Phosphoric Acid units could be manufactured at low volumes for approximately $1000/kW in the
near future, and perhaps at $400-500 per kilowatt in ten years. At this cost, a typical locomotive
unit of 3000 kW would cost $1.2 to $1.5 million in 2007 reflecting a cost premium of $800,000 to $1.1
million over a diesel engine powered locomotive. It is also possible that "rebuilds" would not be
required every eight years so that net cost differences may be smaller over the lifetime of the
locomotive. Also, developers of PEM cells for highway vehicles are aiming at sharply lower costs,
in the range of $50 per kilowatt (for the fuel cell only) or lower; although the size, duty cycle, and
manufacturing volume of locomotive and automotive power plants are clearly very different,
presumably a portion of any cost reductions achieved in automotive fuel cells would be applicable to
fuel cells designed for locomotives.
The average efficiency of the fuel cell over the duty cycle would be 1.6 to 1.7 times as high as the
diesel engine (whose cycle average efficiency excluding idle is about 35%). Fuel savings of 40% are
possible, which is approximately 150,000 gallons/year for a line-haul locomotive. Hence fuel
savings alone could pay for the capital cost increases over about eight years, making the technology
reasonably cost-effective in the context of a 24-year useful life. Of course, major uncertainties exist
in the actual cost of the fuel cell for a 3000 kW unit, the life of the fuel cell, and the maintenance
requirements relative to a diesel engine.
If successful, fuel cell locomotives could have a 5 to 6% market share by 2010, and 16 to 18% by 2015,
for the total fleet. In the high-efficiency/low-carbon scenario, we assume a 5% share in 2010 and
20% by 2015. We further assume use of cellulosic ethanol, although methanol and LNG would also
be likely candidates.
5.2.14 Costs and Timing of Technology
As part of the OTA study, the cost of all the above-described automotive technologies was derived
on the basis of near-term estimates, though at high production volume.. One possible area for
improvement in costs is the effect of research and additional learning to provide an "experience"
based cost reduction. Lipman and Sperling (1997) have analyzed cost reductions based on cumulative
total production, and concluded that many new technologies experience a 20 to 35% cost reduction for
every order of magnitude increase in cumulative production (i.e., the cost decline function is linear
5.16
September 17, 1997
Transportation Sector
Chapter 5
with respect to the logarithm of cumulative units produced). If these new technologies are
manufactured at typical automotive volumes from their introduction, then an order of magnitude
increase in cumulative production will occur over a span of five to seven years with sales growth over
the period. The next order of magnitude increase in cumulative production will take much longer
unless the technology essentially increases market share to 100% over the next decade. We have
utilized the data from the Lipman and Sperling paper to conclude that a 30% cost reduction over the
1997-2005 period is possible relative to the costs derived for OTA.
A second factor is the timing of technology introduction. The contrast here is between new
technology introduction in a business-as-usual scenario relative to one where both business and
government invest in research and development at rates consistent with an accelerated PNGV,
coupled with changes in market preferences for fuel economy driven also by changes in government
policy (e.g., new fuel economy standards, high motor fuel taxes, etc.). The resulting reduction in lead
time is assumed to be 30% relative to the earliest introduction dates forecast by OTA, starting from
1997. In other words, a technology forecast by OTA to be commercialized in 2010 (13 years from 1997)
would be expected to arrive in 2006 (1996 + [13*0.7]) under the regime of increased R&D spending and
market changes. This factor has been incorporated for all post-2005 technologies defined in NEMS
or added to the NEMS technology list for the efficiency and high-efficiency/low-carbor cases.
5.2.15 Alternative Transportation Fuels
Alternative Fuels derived from fossil energy sources have limited potential to reduce greenhouse gas
emissions. The full fuel cycle greenhouse gas emissions of fossil fuels have been compared in detail
by Delucchi (1991, Tables 9a-e), Wang (1996) and others. Several fossil fuel alternatives have
somewhat lower CO₂ emissions than conventional or reformulated gasoline (RFG), most notably
liquefied petroleum gases and natural gas, whether compressed (CNG) or liquefied (LNG). On the
basis of emissions of CO₂ equivalent greenhouse gases per vehicle mile, CNG and LPG offer moderate
reductions both for light (Figure 5.1) and heavy-duty (Figure 5.2) vehicles. Methanol from natural
gas, while it is a relatively attractive alternative fuel for spark-ignited internal combustion
engines, seems to offer no CO2 reduction potential.
Battery-powered electric vehicles (EVs) can also lower greenhouse gas emissions, depending on the
energy source used to produce the electricity stored in the vehicle's batteries. Electricity obtained
from nuclear or solar power would very nearly eliminate greenhouse gas emissions. Use of nuclear
power is unlikely, however, since nuclear power plants tend to operate at capacity at present and are
not likely to supply a marginal increase in demand due to electric vehicle use. Electricity from
current natural gas-fired plants would achieve roughly a one-third reduction, and electricity from
advanced combined cycle natural gas generations could do even better. Estimating CO2 emissions
reductions from electric vehicles is highly dependent on assumptions about when vehicles will be
recharged and how utilities will choose to operate different kinds of generating units. One such set
of estimates, developed based on technologies and generation mixes projected for 2015, is shown in
Figure 5.3. There are no CO₂ emissions from vehicle operation and emissions from vehicle
manufacture are the same for all regions. Largely due to greater use of natural gas in advanced
generating units, the south central and west regions are expected to produce the lowest greenhouse
gas emissions for EVs operated there.
September 17, 1997
5.17
Chapter 5
Transportation Sector
Figure 5.1 Fuel Cycle Greenhouse Gas Emissions for Light-Duty Vehicles
800
CO₂ Equivalent Emissions (g/mi)
600
400
200
0
-50
Gasoline
Methanol/NG
LPG
Ethanol/Wood
CNG/Wood
Ethanol/Com
CNG
EVs
Methanol/Wood
Vehicle Operation
Fuel Production & Distribution
Vehicle Manufacture
Source: Leiby et al., 1996, Table D-4
Figure 5.2 Fuel Cycle Greenhouse Gas Emissions for Heavy-Duty Vehicles
3000
CO₂ Equivalent Emissions (g/mi)
2000
1000
0
-500
Diesel
Methanol/NG
LPG
Methanol/Wood
Ethanol/Corn
CNG
Ethanol/Wood
CNG/Wood
Vehicle Operation
Fuel Production & Distribution
Vehicle Manufacture
Source: Leiby et al., 1996, Table D-4
5.18
September 17, 1997
Transportation Sector
Chapter 5
Figure 5.3 Projected Fuel Cycle Greenhouse Gas Emissions of Battery-Powered Electric Vehicles by
Region in 2015
800
CO₂ Equivalent Emissions (g/mi)
600
400
200
0
Gasoline
LPG
EV: NE
EV: SE
EV: SC
CNG
EV: U.S.
EV: EC
EV: WC
EV: W
Vehicle Operation
Fuel Production & Distribution
Vehicle Manufacture
Source: Singh (1997)
The analyses in Chapters 6 and 7 indicate that there is considerable opportunity to reduce carbon
emissions in the electric utility sector. A substantial shift towards lower-carbon electric generating
facilities will increase the carbon-reducing benefits of electric vehicles. For example, large shifts
away from coal and towards natural gas, especially with combined cycle technology, will tend to
push the relatively high EV emissions in regions whose dominant fuel is now coal (Figure 5.3) down
towards the lower emissions prevalent in areas with primarily gas-fired electricity (e.g.
California).
The AEO97 reference case already projects large increases in the numbers of electric and natural gas
vehicles on the road. Primarily as a result of zero emission vehicle (ZEV) regulations in California,
AEO97 foresees annual sales of 75,000 battery electric cars and 150,000 battery electric light trucks in
2010. To this is added more than a quarter million hybrid electric vehicles. By 2010, the AEO97
reference case projects nearly 2 million battery-electric and over 2 million hybrid electric light-duty
vehicles in operation. Given the recent relaxation of ZEV mandates in California, this projection
now seems optimistic. The AEO97 reference case also projects compressed natural gas vehicle sales
at 325,000 units in 2010 with a total on-road stock of 2.6 million light-duty vehicles. This is more
than thirty times the 82,000 CNG vehicles estimated to be on the road today (ELA, 1996c, Table 1).
We retain these alternative fuel vehicles in all three scenarios, but do not expand them.
Among the alternative transportation fuels under consideration, biomass fuels derived from wood
appear to have the greatest potential to reduce greenhouse gas emissions. Whereas ethanol derived
from com may actually produce higher levels of CO₂ equivalent emissions than conventional
gasoline (depending on the fuel used to power the distillation plant, and other factors), ethanol
derived from cellulosic sources (wood, switchgrass, wood wastes, agricultural residues, municipal
solid waste), can reduce carbon emissions by about 90% for both light-duty and heavy-duty vehicles
(Figures 5.1 and 5.2). Cellulosic ethanol has the potential to be more effective than compressed
synthetic natural gas derived from wood, partly because of the energy that must be used to compress
methane for storage on board the vehicle, and partly because cellulosic ethanol production yields
by-products that can be used to generate more electricity than is required to produce the ethanol
(Delucchi, 1991, Table 9b; Wang, 1996).
September 17, 1997
5.19
Chapter 5
Transportation Sector
Both battery electric vehicles and compressed natural gas vehicles, but especially battery-powered
vehicles, are likely to cost more than conventional gasoline vehicles, will require more frequent
refueling, and will have reduced range (Greene, 1994). It does not appear likely that most consumers
will consider these drawbacks to be outweighed by the likely lower fuel costs for these vehicles.
Thus, we expect these potential low CO₂ fuel technologies will not easily achieve the business-as-
usual forecast market shares (of course, technological breakthroughs in batteries or gaseous fuel
storage could make these vehicle technologies much more attractive). It is for these reasons that we
focus below on the use of cellulosic ethanol as a transport fuel.
5.3 SCENARIOS FOR 2010
5.3.1 The Business-as-Usual Scenario for Transportation
The AEO97 reference case serves as the business-as-usual case, except for its forecast of increasing
light-duty vehicle MPG through 2015. The ELA AEO97 reference case projects an increase in
passenger car MPG from 27.5 in 1997 to 31.5 in 2010 and 32.6 in 2015. Light truck MPG is projected to
increase from 20.5 to 22.9 MPG in 2010 and 24.2 MPG in 2015. We view this as inconsistent with the
historical record, which appears to us to indicate that, without increasing fuel prices or a policy
intervention such as fuel economy standards, MPG is not likely to increase. Thus, we incorporate zero
MPG improvement after 1997 for light-duty vehicles into our business-as-usual case, reflecting the
view that the current level of CAFE standards are and probably will remain a binding constraint on
light-duty vehicle fuel economy throughout the business-as-usual forecast.
From 1982 to 1997, light-duty vehicle fuel economy remained essentially constant, as shown in Figure
5.4. Of course, motor fuel prices declined sharply at the beginning of the 1983-1997 period, but are at
about the same levels today as they were in 1986, and as they were in the early 1970s prior to the
first oil price shock. Given that the AEO97 oil price forecast projects no significant increase in oil or
gasoline prices through 2015, it is reasonable to ask why fuel economy should increase. The ELA's
view is that advances in motor vehicle technology will permit not only fuel economy but other
vehicle attributes such as performance and weight to be increased at lower costs, resulting in greater
consumer satisfaction. There is a very small increase in the price of gasoline through 2010, and this
together with a slowing of income growth may allow the rate of technological advance to catch up
with and pass the effect of consumer demand for larger, more powerful vehicles. Because a
significant slate of cost-effective current and future fuel economy technologies are represented in the
reference case input data, the model takes advantage of them even though fuel prices do not
increase. NEMS would make greater use of the technologies if prices increased significantly, but the
model is driven partly by technology availability and partly by changes in economic parameters.
To some extent, the fuel economy benefits of these technologies are offset by a predicted increase in
demand for performance. Nonetheless, a 5 MPG gain remains.
5.20
September 17, 1997
Transportation Sector
Chapter 5
Figure 5.4 New Light-Duty Vehicle Fuel Economy and Gasoline Prices, 1967-1996
30
250
MPG
Price (Lag 1):
200
20
Average Miles per Gallon
150
1992 Cents per Gallon
100
10
50
0
0
1968
1972
1976
1980
1984
1988
1992
1996
It is difficult to separate out analytically the impacts of CAFE standards and the effect of the
marketplace in pushing fleet fuel economy one way or the other. However, we believe that the most
likely explanation for the stagnation of fuel economy levels over the past decade is that the CAFE
standards have tended to act as a floor on fuel economy, that without the standards the market
level of fuel economy would have been lower than it was. We note that important fuel economy
technologies, such as fuel injection, front-wheel drive, lock-up torque conversion, 4-valves per
cylinder, overhead cam design, improved aerodynamics, and others, all increased their market
penetration over the 1983-1997 period (Figure 5.5). Fuel economy technologies were adopted, yet
average fuel economy did not increase. There are two major reasons. First, much of the potential to
improve fuel economy was used instead to increase average light-duty vehicle horsepower by 55%
and weight by 13% from 1983 to 1996 (Heavenrich and Hellman, 1996, Table 1). The second reason is
that the impact of a technology on fuel economy depends on how that technology is implemented. To
some extent, the fuel economy benefit of a technology is inherent in it. But to a degree, the benefit
also depends on the details of vehicle design, specifically whether the technology is implemented
with the purpose of increasing MPG or with some other purpose in mind.
If CAFE was in fact a binding constraint during the past decade and remains so today, fleet MPG will
not begin to increase significantly until a market equilibrium is reached wherein actual fleet fuel
economy becomes equal to the fuel economy level that would be achieved in the absence of CAFE
standards. In our view, estimating "free market" fuel economy levels is basically a judgment call.
We have assumed that market equilibrium will not be reached in the base case, so that fleet fuel
economy will remain unchanged. In other words, we assume that, although fuel economy technology
will continue to be adopted, it will be used to provide other benefits than fuel economy, particularly
increased size and performance. Note that the AEO97 reference case also projects increased
September 17, 1997
5.21
Chapter 5
Transportation Sector
performance and size over the forecast period; the difference here is a matter of degree, not one of
radically different visions of the most likely future.
Figure 5.5 Use of Fuel Economy Technology In New Light-Duty Vehicles
I
Front Wheel Drive
Torque Converter Lock-Up
Port Fuel Injection
0.8
Multivalve/Cylinder
0.6
New Vehicle Market Share
0.4
0.2
0
1975
1978
1981
1984
1987
1990
1993
1996
5.3.2 The Efficiency Scenario For Transportation
The efficiency case was created by making reasonable, incremental assumptions about how a
concerted effort to accelerate the development and promote the adoption of low greenhouse gas
technologies could reduce emissions by the U.S. transportation sector. In this section, the specific
changes made to the business-as-usual case are described in detail.
5.3.2.1 Changes to the Modal Models
The efficiency scenario assumes that the time required for market introduction of advanced
technologies can be reduced by 25% through increased emphasis on technology R&D, and that
several new technologies will be developed that would otherwise not be available in significant
numbers before 2010. For light-duty vehicles, these technologies include the following:
A direct-injection stratified charge (DISC) gasoline engine,
A turbocharged direct-injection clean diesel engine (TDI Diesel) that meets current and future
emissions standards,
Advanced drag reduction, materials substitution, and engine friction reduction (Drag VI),
5.22
September 17, 1997
Transportation Sector
Chapter 5
A gasoline/electric hybrid vehicle (Gasoline Hybrid), and
A diesel/electric hybrid drive vehicle (Diesel Hybrid).
In fact, the diesel hybrid and the 2-stroke engine were not included in the efficiency scenario in order
to reduce the number of new engine technologies introduced. We chose the gasoline over the diesel
hybrid because its emissions of conventional pollutants can very likely be reduced to extremely low
levels, making it attractive for air quality reasons. In the high-efficiency/low-carbon scenario,
both the 2-stroke and the diesel hybrid are included (the 2-stroke is assumed to be applicable only
in compact or smaller-sized vehicles). The result is that new powerplant technologies all but
entirely replace today's conventional gasoline engine by 2015 in the high efficiency scenario. This
seems a very ambitious undertaking and one that would require greater expense and a higher degree
of technical success than is consistent with the efficiency scenario.
The efficiency scenario assumes a cost reduction of about one-third over estimates developed by OTA
(1995) for the advanced technologies shown in Table 5.2, based on the potential for learning-based
cost reductions discussed earlier. Among conventional technologies, the cost of CVTs was reduced
from $250 to $150 and the cost of VVT was cut in half for passenger cars and left unchanged for light
trucks. The cost reductions are intended to reflect the success of an enhanced R&D effort.
In the truck freight sector, several new technologies were brought into the forecast by reducing the
fuel price threshold at which they would become attractive to buyers. These include:
The LE-55 diesel engine with a 21% efficiency improvement for heavy trucks,
Reduced empty weight,
The turbo compound diesel engine, and
Advanced drag reduction.
The low-emission, 55% thermal efficiency (LE-55) diesel engine is a research target of the U.S.
Department of Energy's Office of Transportation Technologies. Compression ignition (diesel) engines
are the most efficient heat engines currently available. Very large units (in stationary or marine
applications) achieve thermal efficiencies (work output as a ratio to energy content of fuel) of 50%.
The best turbocharged diesel engines for heavy trucks achieve 45% thermal efficiency, versus 24%
for gasoline engines. The DOE's Offfice of Heavy Vehicle Technology has established a goal of 55%
thermal efficiency for heavy truck engines as an intermediate target on the way to a long-term goal
of 63%. These improvements are to be achieved through a combination of increased peak pressure,
insulation of pistons, cylinder walls and heads to reduce heat loss, effective recovery of exhause
heat, friction reduction, and improved turbocharger efficiency (U.S. DOE/OHVT, 1996).
For commercial aircraft, an efficiency improvement of 40% was projected for 2015 for new aircraft,
comprised of 25% engine efficiency gains and 15% aerodynamics and materials substitution. Also,
load factors were assumed to increase to 70% in accord with industry projections as a result of
advanced informational and operational technologies. Finally, railroad freight efficiency per ton-
mile was assumed to improve at 2% per year, actually somewhat lower than the 2.8%/yr. rate
experienced over the past 20 years.
September 17, 1997
5.23
Chapter 5
Transportation Sector
5.3.2.2 New Technologies
Table 5.2 shows the fuel economy benefits, price impacts, years of introduction, effects on vehicle
weight, and effects on vehicle performance of the five new technologies that were added to the
AEO97 reference case set. Detailed assumptions underlying the cost of fuel economy improvement
estimates shown in Table 5.2 are provided as an appendix to this chapter. In order to meet current
and future emissions standards, the DISC and TDI Diesel engines, as well as the two-stroke engine
included in the business-as-usual case, will require the development of practical, lean-combustion
nitrogen oxide catalysts. Catalyst technology for treatment of exhaust emissions has been advanced
significantly over the past few years and, with further research, the prospects for its early
commercialization appear to be very good (e.g., Buchholz, 1997; Strehlau et al., 1997). Achieving
equivalent results for diesel exhaust NOₓ appears to be more difficult, and commercialization of
diesel catalysts is likely to occur several years after introduction of gasoline-engine catalysts (U.S.
Congress, OTA, 1995). In addition, the DI Diesel will require advances in fuel and emissions control
technology in order to meet likely future particulate standards.
Fuel economy benefits, incremental costs and other changes are calculated with reference to a 1995
technology gasoline vehicle. In the NEMS model, light-duty vehicles are classified into passenger
cars vs. light trucks, domestic vs. imported, with six size classes for each category. In each class, the
1995 base vehicle has the average characteristics of cars in its class. For example, half of the
passenger cars in 1995 had 4-valve per cylinder engines, but less than 10% of the light trucks did
(Heavenrich and Hellman, 1996). Thus, the 1995 base vehicle is credited with half of the fuel
economy improvement potential and half of the increased cost of 4-value technology. One hundred
percent of passenger cars and 99% of light trucks had port fuel injection, and so the base year
vehicles are given 100% and 99% of the fuel economy benefit and cost of fuel injection technology.
Future fuel economy improvements are calculated based on the additional penetration of fuel
economy technologies beyond the business-as-usual case. Thus, the ability of further use of port fuel
injection to improve fuel economy is negligible, while considerable potential remains for 4-valve
technology.
Table 5.2 New Light-Duty Vehicle Technologies Added to the Efficiency and High-
Efficiency/Low-Carbon Scenarios*
Technology
MPG Benefit (%)*
OTA Price
Scenario
Introduction Date*
Increase
Price
(EFF, HE/LC)
(EFF, HE/LC)
DISC
18, 23
$450
$300
2000, 2000
Turbo DI Diesel
40,40
$1100
$750
2004, 2004
Hybrid/Gasoline
33,42
$3000
$2000
2005, 2005
Hybrid/Diesel
54,72
$3500
$2300
2005, 2005
Drag VI
12, 12
$256
$256
2012, 2012
Gasoline Fuel Cell
- 84
-
$800
2007
+
For an explanation of the assumptions underlying these estimates please see the appendix to this chapter.
5.3.2.3 Valuing Energy Savings
The NEMS model values the fuel economy savings of advanced technology by computing the
expected discounted value of annual fuel savings over a payback period. We used a 7% real discount
5.24
September 17, 1997
Transportation Sector
Chapter 5
rate over five years whereas the reference case assumes an 8% real discount rate over a four-year
payback period. The issue of discounting fuel savings is discussed in greater detail in Section 5.3.5.
5.3.2.4 Trends in Vehicle Performance
The NEMS model predicts consumer demand for increased performance and then adjusts new car MPG
downward to reflect the effect of higher horsepower on fuel economy. The model's predictions are
consistent with recent trends in light-duty vehicle performance since the early 1980s. Over this
period, new vehicle fuel economy was constrained by the federal Automotive Fuel Economy
Standards (CAFE) to levels higher than the market would otherwise have demanded. Gasoline
prices fell precipitously, starting in 1983 and reaching pre-1974 levels by 1987 (Figure 5.4). As a
result, new technology adopted since the mid-1980s, that could have increased fuel economy, was
instead used to hold fuel economy constant while increasing vehicle horsepower and weight. The
ratio of horsepower to weight for passenger cars increased by 50% from 1982 to 1996). The NEMS
horsepower equations essentially continue this trend of ever-increasing performance.
Continued use of new technology to increase performance without increasing fuel economy is
consistent with the continued low motor fuel prices projected in the AEO97 reference case. The
reference case foresees gasoline prices rising from $1.15 in 1995 to $1.23 in 2010 and falling to $1.18
per gallon in 2015 (1995$). Such variations are within the noise of year-to-year fluctuations. For
example, the actual average price of gasoline in 1995 was $1.20 and the average price for 1996 will
likely exceed $1.30 per gallon (ELA, 1997, Table 9.4). With no increase in price and binding fuel
economy standards, it is likely that performance and weight will continue to increase and fuel
economy will not.
In the efficiency case, the trend toward ever greater horsepower is questionable. In the presence of
higher fuel economy standards, voluntary commitments by manufacturers to meet GHG targets,
"greener" consumers, externality-based fuel taxes, or some other change in policies or preferences
focusing consumers' and manufacturers' attention on efficiency, it is likely that performance trends
would change. Nonetheless, we retain the NEMS performance projection in the efficiency case, but
relax it in the high-efficiency/low-carbor case by permitting only half of the projected increase in
horsepower. This results in new vehicle fuel economy levels 1-2 MPG higher in the high-
efficiency/low-carbon case than would otherwise be the case.
5.3.2.5 NEMS New Light-Duty Vehicle Fuel Economy Estimates
Transforming the technology of transportation energy use takes time. First, manufacturers must
implement a new technology. New designs must be engineered, tested, and certified to meet
government standards. Generally, capital equipment will also have to be replaced or retooled. The
orderly replacement of long-lived production facilities (engine production lines may last 15 years, or
more) is important to holding down the cost of technological change. Second, consumers must become
accustomed to the new technology, and the supporting infrastructure of maintenance and repair must
be developed. Finally, new technologies must compete with existing technologies and with other
new technologies. In general, a single technology will not dominate all possible applications
(vehicle types and consumer preferences). For all these reasons, new technologies rarely achieve
100% (or even 10%) market penetration of the new vehicle fleet in the first year of introduction. The
NEMS model simulates the gradual evolution of technology market shares toward their eventual
equilibrium levels by means of technology adoption curves calibrated to historical rates of adoption.
As a result, the NEMS forecast of average fuel economy for new vehicles will lag behind the full
technological potential. This is illustrated in Table 5.3, which lists all of the best technology
predicted to be available in 2010 and 2015 in the efficiency scenario, except that the diesel rather
September 17, 1997
5.25
Chapter 5
Transportation Sector
than the gasoline hybrid is included. The effects of regulations that are likely to reduce fuel
economy are also included, but further increases in performance (horsepower/weight) predicted by
the NEMS model are not, i.e., horsepower-to-weight ratios are assumed to remain constant at 1997
levels. (This applies only to Table 5.3 - all scenarios incorporate substantial increases in hp/wt
ratios.) The combined effect of all technologies could improve the fuel economy of the average
passenger car by 100% to 55 MPG in 2010, and by another 20% to over 60 MPG in 2015. Yet, even in the
high-efficiency/low-carbon scenario, these levels are not achieved by the new car fleet in the
NEMS forecasts.
Table 5.3 Maximum Technological Fuel Economy Potential Versus NEMS New Car Average
Estimates
Technology
2010
2015
Fuel Economy
Fuel Economy
Improvement (%)
Improvement (%)
Material Substitution IV
9.9
13.2
Drag Reduction V
9.2
12.0
Engine Friction III
5.0
6.5
Tires III
5.0
7.0
ACC II
1.0
1.0
Electric Transmission II
1.5
1.5
Electric Power Steering
1.5
1.5
Air Bags
-1.0
-1.0
Emissions Tier II
-1.0
-1.0
ABS
-0.5
-0.5
Side-Impact
-0.5
-0.5
Roof Crush
-0.3
-0.3
Diesel Hybrid
54.0
60.0
Total % Improvement*
100.0
123.0
1997 MPG
2010 MPG
2015 MPG
27.5
Maximum Use of All Fuel Economy Technology
Miles per Gallon
55.0
61.3
Percent Improvement
100
123
New Car Salesweighted Average Fuel Economy: Low CO2 Scenario
Miles per Gallon
37.5
41.4
Percent Improvement
36
51
New Car Salesweighted Average Fuel Economy: Breakthrough Scenario
Miles per Gallon
43.1
50.2
Percent Improvement
57
83
* Total percent improvement is computed as [(1 * 100 54 ) (1 + 100 30 - 1)] * 100. Summing rather than multiplying the
smaller percentage improvements yields a more conservative estimate.
Clearly, faster rates of fuel economy improvement than predicted in either scenario are achievable,
but at added cost. The constraint that fuel economy improvements be approximately cost-effective
5.26
September 17, 1997
Transportation Sector
Chapter 5
requires that the changeover of technologies and manufacturing capital occur at approximately
normal rates. This causes realized new car MPG levels to lag considerably behind the full
technological potential. On the one hand, this implies that considerable additional energy-
efficiency improvement can be made beyond 2015. On the other, it implies that markets must be
encouraged, through public policy measures, to make continuous improvements if cost-effective
reductions in CO₂ emissions are to be realized.
5.3.2.6 Changes to the Heavy Truck Model
In contrast to the business-as-usual case, the efficiency scenario for heavy trucks:
Advances introduction dates for two fuel economy technologies,
Introduces one additional technology,
Expands the applicability of several truck technologies,
Reduces the "trigger price" at which the technologies are assumed to become cost-effective, and
Accelerates the rate at which new technologies are assumed to penetrate the new truck market.
The AEO97 reference case assumes that the Turbocompound Diesel Engine and the Advanced LE-55
Heat Engine will not be available through 2015. The efficiency scenario assumes these technologies
will be introduced in 2003. Advanced drag reduction, which is also excluded from the reference case
for heavy trucks is assumed to have become available in 1997.
The additional technology introduced is reduction in vehicle empty weight through material
substitution. Reducing vehicle empty weight by 10% should be possible, with a consequent 3%
increase in fuel economy (Roberts and Greene, 1983; Greene, 1996a, Table 5.5). Reduced empty weight
is assumed to be applicable to all types of heavy trucks.
The AEO97 reference case assumes that advanced drag reduction, the turbocompound diesel, and the
LE-55 heat engine will be applicable only to the heaviest diesel trucks. The efficiency case extends
the applicability of these technologies to medium-heavy diesel trucks, as well. However, the fuel
economy benefits of advanced drag reduction are cut from 18% to 10% for medium trucks to reflect the
fact that they are generally operated at lower speeds.
A key factor governing the use of fuel economy technology in the NEMS Heavy Truck Model is the
"trigger price." Until market fuel prices reach the "trigger price" level specified for a technology,
the technology will not be introduced. Diesel fuel prices never exceed $8.70 (1995$) per million Btu
($1.21/gal.) in the AEO97 reference case. Trigger prices for all but existing technologies, however,
are $9/MMBtu, or greater in the reference case. The efficiency case assumes that all of the new
technology can be made cost-effective at $8/MMBtu.9
Other parameters controlling the rate and extent of market penetration for technologies were also
changed. One of these is the number of years until 99% of the maximum potential market
penetration is achieved. For improved tires and lubricants, electronic engine controls, and-electronic
transmission controls, a value of 20 years is assumed in the AEO97 reference case. But for advanced
drag reduction, turbocompound diesel, and the LE-55 engine, 99 years is the assumed value. For the
efficiency case, all were set at 20 years. The AEO97 reference case assumed that the LE-55 engine
would have a maximum market potential of 50% for heavy-duty diesels. The efficiency scenario
assumes a 100% maximum for heavy diesels, but only 50% for heavy gasoline, LPG, and CNG trucks.
September 17, 1997
5.27
Chapter 5
Transportation Sector
Likewise, the maximum market potential for other advanced technologies was increased to 100% for
the heavy diesel market, but left at the reference case values for other fuel types. For medium
diesel trucks the maximum penetration for new technologies was raised to 90%, but left at the
reference case levels for other fuel types. These changes do not imply that any of these technologies
will actually reach maximum market penetration over the forecast time period. Table 5.4
summarizes the primary fuel economy technologies for heavy trucks in the efficiency scenario for
2010.
Table 5.4 Key Heavy Truck Fuel Economy Technologies for the Efficiency Scenario in 2010
Maximum
Year of
Trigger Price
Market
Fuel Economy
Technology
Introduction
(1995$/MMBtu)
Potential
Improvement %
(other / diesel)
(medium/heavy)
Improved Tires & Lubes
1994
$7.75
80% / 100%
10% / 6%
Electronic Engine Controls
1994
$7.75
70% / 100%
2%
Elec. Transmission Controls
1994
$7.75
75% / 100%
5% / 2%
Advanced Drag Reduction
2000
$7.75
25% / 100%
7% / 18%
Turbocompound Diesel
2000
$7.75
25% / 100%
15% / 17%
LE-55 Heat Engine
2003
$7.75
50% / 100%
19% / 21%
Reduced Empty Weight
1997
$7.75
90% / 100%
3%
5.3.2.7
Changes to the Rail Model
The AEO97 reference case scenario assumes an annual rate of reduction in rail freight energy use per
ton-mile of 1%. Since 1972, the average annual rate of reduction in energy use per ton-mile has been
2.8% per year. The vast majority of this improvement has been due to operational efficiency
improvements reflected in increased load factors per car (Greene and Fan, 1995, p. 15). Higher load
factors are partly due to the restructuring of the rail industry following deregulation in 1980, and
partly due to the use of advanced technology for managing operations. Technologies such as lighter
weight and higher capacity cars, lower resistance axle bearings, rail-wheel lubrication and
improved efficiency locomotives also played an important role (Cataldi, 1995). These technologies
are, as yet, still only partially implemented. Based on Cataldi (1995), advanced technologies that
can play a role in substantially reducing rail energy use in the future include the following:
Flywheels: Trains presently give up large amounts of kinetic energy on downgrades that could
be transferred to flywheels and later used to power the train. The volume and mass necessary to
store huge quantities of power can be readily accommodated on trains.
Oxygen-enrichment to increase engine thermal efficiency: Membranes that exclude part of the
free nitrogen in the air, thereby enriching the oxygen concentration, can be incorporated into
locomotives' air filtration systems. This technology should benefit new, higher power density
engine designs, while helping to hold down their nitrogen oxide emissions.
Alternative fuels: Railroads and locomotive manufacturers have been studying and testing the
use of natural gas fired locomotives. Once again, the ability of trains to accommodate the
volume and mass of storage systems for liquefied natural gas gives them a distinct advantage
over smaller vehicles in the application of this technology. Although natural gas locomotives
are not expected to provide energy-efficiency gains over diesels, natural gas will produce fewer
CO₂ and NOₓ emissions and reduce U.S. dependence on oil.
5.28
September 17, 1997
Transportation Sector
Cha: ter 5
Fuel cells: Beyond 2010, fuel cells for locomotives hold promise. Locomotives alread; use
electric drive systems. And carrying fuel, even compressed hydrogen in large volumes, is less of
a problem for trains than for highway vehicles.
Because existing energy-efficiency technologies have yet to achieve full utilization, because other
promising options exist, and because further operational efficiency gains are likely with the
advance of information technology and some additional railroad consolidation, rail energy-
efficiency improvements could continue at a substantial rate. A concerted effort to develop and
implement cost-effective technologies is represented here by a 2% annual improvement in to: mile
efficiency in the efficiency case compared with the 1% rate assumed in the AEO97 reference case.
5.3.2.8 Changes to the Air Model
No new technologies were introduced in the NEMS Air Travel Model, but several important cl ages
were made to promote and accelerate the introduction of fuel efficient technology in accordance with
goals set by the Committee on Aeronautical Technologies, Aeronautics and Space Engineering "hard
of the NRC. Broadly, these goals call for a reduction in fuel burn per seat of about 40% by the 2.10 to
2015 time period, to be achieved through a combination of improved propulsion system perfor nance
(25%) and aerodynamic and weight improvements (15%) (NRC, 1992, P. 49).
Once again, in the AEO97 reference case, new technologies do not enter the commercial ai craft
market because trigger prices are set well in excess of $1.00 per gallon and jet fuel prices never ceed
$0.80/gal. over the forecast period. Trigger prices for ultra-high bypass turbo-fans, already use
on the new Boeing 777s, were lowered to $0.58/gal., just slightly above current jet fuel rices.
Advanced aerodynamics, weight reduction through advanced materials use, and improved gine
thermodynamics, were all given the same, lower trigger price. The prices of turboprop engin and
laminar flow control were left at levels high enough to prohibit their introduction on new air raft
through 2015.
Ultra-high bypass turbofans were introduced in 1995. The other three technologies were assu .ed to
be introduced in 2000. Consistent with estimates presented in NRC (1992), Greene (1992, Tal le 4),
and Greene (1996b), the efficiency improvement potentials for all four new technologies were et at
15%.
Finally, the AEO97 reference case predicts no changes in aircraft load factors. Aircraft in stry
analyses foresee commercial load factors increasing to 70% by 2015 (Boeing, 1995, p. 25; McD mell
Douglas, 1996, P. 18). The industry view is adopted in the efficiency scenario, on the grounds that it
will very likely be advances in information technology that permit the increase in load facto On
the other hand, although the industry predicts an increase in aircraft size (seats/aircraft) of bout
15% by 2015 while the AEO97 reference case does not, no such increase is included in the effi ney
scenario on the grounds that more seats per aircraft will be less a reflection of technological Mange
than of airframe choice.
5.3.2.9 Introduction of Cellulosic Ethanol
Alternative fuels derived from fossil fuels have limited potential to reduce greenhou gas
emissions. The full fuel cycle greenhouse gas emissions of fossil fuels have been compared in detail
by Delucchi (1991, Table 9a), Wang (1996), U.S. DOE (Leiby et al., 1996, Tables D-4 and D-5), and
others; see Wang (1996) for a review. Several fuel alternatives have lower CO2 emissions than
conventional or reformulated gasoline (RFG), most notably liquefied petroleum gases l'G),
methane and battery-powered electric vehicles in certain regions, whether compressed (Cr G) or
liquefied (LNG). Estimates of greenhouse gas emissions are strongly dependent on conte and
September 17, 1997
5.29
Chapter 5
Transportation Sector
assumptions. Absolute levels and sometimes the relative rankings of fuels vary across studies.
Several general patterns seem to hold up, however. For example, fossil-fuel based alternatives to
gasoline or diesel fuel, including battery-electric vehicles where substantial amounts of coal are
used for electricity generation, offer about a 20% net reduction in greenhouse gas emissions per mile
(Figure 5.3).
In the context of this analysis, a 20% reduction in greenhouse gas emissions will not create a strong
incentive to adopt an alternative fuel. For light-duty vehicles, if society's willingness to pay for
greenhouse gas emissions reductions were on the order of $25-$50 per tonne of carbon, this could
justify up to a $0.06 to $0.12 per gallon subsidy¹⁰ for a fuel that produced no greenhouse gas emissions.
A 20% reduction would therefore be worth $0.01 to $0.02 per gallon, hardly enough to get motorists'
attention. Also, the principal near-term alternative fuels entail some increase in vehicle cost or loss
of amenity (Leiby et al., 1996). Thus, unless much higher incentives are introduced, it is unlikely
that enough substitution of alternative fossil fuels for conventional gasoline will occur to produce
significant greenhouse gas reductions in transportation (Leiby et al., 1996).
Alternative fuels produced from renewable biomass feed stocks can yield significant reductions in
greenhouse gas emissions. The most recent estimates indicate that ethanol derived from cellulosic
feed stocks (as opposed to grain) produces less than 1% as much greenhouse gas emissions on a fuel
cycle basis as conventional gasoline or diesel fuels (Singh, 1997). 11 Table 5.5 shows the greenhouse
gas emission coefficients used to estimate the effects of cellulosic ethanol use and increased demand
for diesel fuel on transportation sector greenhouse gas emissions. Ethanol from cellulose generates
negligible amounts of greenhouse gases in comparison to fossil fuels or ethanol from grain. Whether
ethanol is derived from grain or woody biomass, the carbon in the fuel itself does not count because
equivalent carbon will be recaptured from the atmosphere by the next rotation of crops. The
differences lie in feed stock cultivation, fertilizer manufacture, and fuel production. Com requires
more cultivation and more fertilizer than woody crops, and fertilizer production, in particular,
generates significant greenhouse gas emissions. Whereas distillation of alcohol after the
fermentation of grain is energy intensive, by-products from the wood-to-alcohol process will produce
excess power, on net, resulting in a greenhouse gas credit for replacing fossil fuels with biomass in the
generation of electricity. Indeed, given current practice, ethanol from corn may produce more
greenhouse gas emissions than gasoline, on a per Btu basis. Thus, ethanol from cellulosic feed stocks
will not only reduce greenhouse gas emissions by replacing gasoline, but might achieve even greater
benefits by replacing ethanol from com. However, the net greenhouse gas balance of ethanol
production from com is strongly dependent on future com yields, the market for distillation
byproducts, and the efficiency of and fuel used in distillation (currently, coal is often the preferred
fuel because of corn-based ethanol's disadvantage in greenhouse gas emissions, but future widespread
use of com stillage as fuel would swing ethanol's greenhouse gas emissions strongly towards a
positive balance).
A new process for producing ethanol from cellulosic biomass that appears to have the potential to
dramatically reduce costs is under development by the U.S. DOE's National Renewable Energy
Laboratory (Chem Systems, Inc., 1993). After initial preparation of the biomass, pretreatment with
sulfuric acid and then steam is used to expose the cellulose and convert xylan to xylose. Two percent
of the resulting mixture is separated for conversion to cellulase, an enzyme that hydrolyzes
cellulose. The cellulase is then combined with the rest of the mixture fermented in a key step known
as simultaneous saccharification and fermentation (SSF) because the hydrolyzation of cellulose and
the fermentation of xylose occur simultaneously. The inclusion of xylose fermentation in this step
increases the output of ethanol by about 25% over previous processes. Effluent from the SSF process
goes to an ethanol purification and solids separation phase, which produces ethanol and solids.
After removal of water, the solids are burned as fuel to cogenerate steam and electricity required for
the plant, with surplus electricity that can be sold as a byproduct.
5.30
September 17, 1997
Transportation Sector
Chapter 5
Table 5.5 Greenhouse Gas Emissions Factors for Transportation Fuels
Fuel
g/Btu
Btu/gallon
g/gallon
Conventional Gasoline
Summer
0.10554
114,500
12,084
Winter
0.10304
112,700
11,613
Average
0.10421
113,537
11,832
Diesel
0.09617
128,700
12,377
Ethanol from corn
0.13390
76,100
10,190
Ethanol from cellulose
0.00076
76,100
58
Source: Singh (1997)
Initial estimates of the cost of ethanol produced by the NREL process ranged from $0.78 to $1.27
(1990$) per gallon, plant gate price (Chem Systems, 1993, Tables II-9 to II-13). However, recent cost
projections (Bowman et al., 1997) based on a comprehensive assessment of feed stock supply in the
United States (Walsh et al., 1997) and anticipated improvements in the ethanol conversion process
predict that much lower production costs can be achieved by 2010 or 2015. Ethanol can be produced
from a variety of cellulosic feed stocks: short rotation woody crops, switch grass, softwood and
hardwood wastes, agricultural residues, and even municipal solid wastes. Selecting the lowest cost
feedstock at each level of output, aggregate ethanol supply curves were constructed for 2000, 2005,
2010, and 2015 under "moderate" and "optimistic" assumptions. The optimistic curves assume that
the yield improvements of the moderate case are accelerated by five years, with the net result that
the real cost of feed stocks does not rise over time. The moderate scenario curves are used in the
efficiency scenario. The optimistic case, being similar in intent to our high-efficiency/low-carbon
scenario, is used in that scenario.
In the moderate scenario, ethanol production costs drop dramatically after 2005, the year in which
advanced ethanol conversion technology is assumed to be introduced. In 2000, the first billion
gallons cost $1.10 (1995$) per gallon at the plant gate, which rises to almost $1.25 per gallon at a 10
billion gallon output level. These prices exclude motor fuel taxes and transportation costs. By
comparison, the average refinery price of all grades of gasoline in 1995 was $0.63 per gallon (EIA,
1996a, Tables 5.20 and 5.21). Because ethanol has only about two-thirds of the energy content of
gasoline, the comparable price of ethanol per gasoline energy equivalent gallon would be $1.63 for
the first billion gallons and $1.85 at the 10 billion gallon level of output. By 2010, the cost of
ethanol drops to about $0.75 per gallon ($1.11 per gasoline equivalent gallon) at the 1 billion gallon
output level, $0.79/gallon ($1.17 equivalent) at 10 billion gallons of production. Even in the
optimistic case, the first billion gallons cost $0.67/gallon ($.99 equivalent), increasing to
$0.73/gallon ($1.08 equivalent) at the 10 billion gallon output level. Despite dramatic reductions in
the cost of producing ethanol from biomass, because of the lower energy content of ethanol, ethanol
still cannot compete with gasoline as a pure fuel.
We conclude that the market for cellulosic ethanol in 2010 will be largely as a blending component
for gasoline. Demand curves for ethanol for blending with gasoline have been estimated by Hadder
(1997) for the year 2010. These show the value to refiners of being able to produce a gasoline refined
to be blended with alcohol downstream. Ethanol increases the gasoline's octane rating and adds
oxygenates that are required in certain areas under the Clean Air Act. The demand for ethanol as a
blending component turns out to be sensitive to the market share of RFG. The more RFG that is
required, the lower the demand for ethanol. We assume that RFG's market share remains at its
September 17, 1997
5.31
Chapter 5
Transportation Sector
current level of about 30%. Estimated ethanol demand increases as its price declines, from 2 billion
gallons per year at an ethanol price of just over $1 per gallon to 5 billion gallons at $0.80/gallon and
9 billion at $0.65/gallon. From this point, increases in demand associated with further price
decreases drop off sharply as the limits of economical blending are reached. The moderate 2010
supply curve for cellulosic ethanol intersects the demand curve at about 5 billion gallons per year
(Figure 5.6).
Figure 5.6 Biomass Ethanol Supply and Demand for Ethanol in Gasoline Blending
1.2
1.0
0.8
1995 $/gal. (delivered)
Moderate Supply
0.6
Optimistic Supply
Demand
0.4
0.2
0.0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Billions of Gallons Per Year
These calculations include no tax subsidy for cellulosic ethanol from biomass. If the projected supply
curves are correct, cellulosic ethanol would require no subsidy to be economically attractive as a
blending component for gasoline. The current tax subsidy for ethanol - now produced from grain is
due to expire, and the future of the industry is uncertain. Assuming discontinuation of the subsidy,
cellulosic ethanol would displace corn-based ethanol from the gasohol, yielding significant
greenhouse gas emissions reductions.
5.3.2.10 Adjustment of NEMS Gasoline Forecast
The 5.5 billion gallons of cellulosic ethanol demanded in the efficiency case reduce greenhouse gas
emissions by 13 million tonnes of carbon equivalent in 2010 compared to the business-as-usual case.
Cellulosic ethanol is assumed to replace first corn-based ethanol, and then conventional gasoline.
The Federal Highway Administration (FHWA) estimates that 1.214 billion gallons of ethanol
were used in gasohol in 1995 (U.S. DOT/FHWA, 1996, Table MF-33E). If gasohol made from corn-
based ethanol were to maintain a constant share of the gasoline market, then corn-based ethanol use
would grow to 1,263 million gallons in 2010, then shrink to 1,119 million gallons in 2015. Table 5.6
shows the projected demand for cellulosic ethanol, the corn-based ethanol assumed to be replaced
and the impact on fuel cycle greenhouse gas emissions. Because the upstream effects cannot be
5.32
September 17, 1997
Transportation Sector
Chapter 5
assumed to be accounted for in the other sectoral models, they are included here. Note that the
reduction is shown in tonnes of carbon, while the emissions before and after are in tonnes of CO2.
Table 5.6 Impact of Cellulosic Ethanol on Greenhouse Gas Emissions from Light-Duty Vehicles in
2010
Efficiency
Optimistic
Cellulosic Ethanol (million gallons)
5,514
7,480
Corn Ethanol Displaced (million gallons)
1,263
1,119
Gasoline Equivalent Energy Displaced (million gallons)
3,696
5,014
GHG Emissions Before (million tonnes CO₂ per year)
46.6
62.2
GHG Emissions After (million tonnes CO₂ per year)
0.3
0.4
GHG Emissions Reduction (million tonnes C per year)
12.6
16.8
5.3.2.11 Adjustments for Increased Light-Duty Vehicle Diesel Use
Because the TDI Diesel engine and the Diesel-hybrid technologies were introduced in the NEMS
Transportation Sector Model as fuel economy technologies, the fuel-type accounting algorithms of
NEMS were bypassed. We introduced the advanced diesel in this way because we believe that its
characteristics will be more similar to gasoline engines than the diesels available in the past. The
TDI will fully meet all gasoline vehicle standards and will be quite similar in terms of performance,
noise, and odor.¹² Thus, an adjustment must be made ex post, to transfer an appropriate amount of
energy from the gasoline to the distillate category. The adjustment affects the energy use projections
in three (relatively minor) ways. First, the TDI Diesel's impact is specified in terms of a change in
miles per gallon. Since diesel fuel contains more Btu per gallon than gasoline and since the NEMS
model assumes that gasoline is being consumed, the energy use transferred from gasoline to diesel
must be increased by the ratio of diesel to gasoline Btus per gallon. Second, distillate fuel produces
slightly less carbon emissions per Btu than gasoline. Therefore the estimated carbon emissions must
be adjusted both for the slight increase in energy use and the slightly lower emissions per Btu for
that greater energy use (the net result is a very small increase in carbon emissions). Third, and
finally, the reduction in gasoline use reduces the potential pool for ethanol blending in gasoline. As
a result, the demand for ethanol must be adjusted downward to reflect the lower level of gasoline
use. The net result of all of these changes on energy use and carbon emissions is less than 1%.
5.3.3 The High-Efficiency/Low-Carbon Scenario for Transportation
The high-efficiency/low-carbon scenario postulates the introduction before 2010 of several new
technologies and combines them with other changes to reflect greater success in developing and
implementing low greenhouse gas technologies and greater public concern over greenhouse gas
emissions. Note that successfully achieving these outcomes requires some technological
breakthroughs, implying that the outcomes are significantly less certain than those in the
efficiency case. As we pointed out in the introduction to this chapter, the high-efficiency/low-
carbon scenario is best characterized as an "optimistic" version of the efficiency scenario's "most
likely" assumptions. Both must be considered responses to intensified R&D efforst and new policy
measures to push the market toward low-carbon technologies. A $50/ton carbon tradable permit
price could be one of the necessary policies, but it does not define the difference between our efficieny
and high-efficiency scenarios.
September 17, 1997
5.33
Chapter 5
Transportation Sector
5.3.3.1 Light-Duty Vehicles
Changes for light-duty vehicles include introducing a fuel-cell hybrid in the year 2007 and
reintroducing the diesel hybrid and the 2-stroke engine for smaller vehicles. In the projections
shown here, we assume that the fuel cell hybrid vehicle uses gasoline which is reformed to provide
hydrogen for the fuel cell's operation (e.g., Jost, 1997). The vehicle could just as easily have been
designed to operate on alcohol fuels. The gasoline fuel cell hybrid achieves an 84% efficiency gain
over a conventional gasoline vehicle, assuming major progress not only in fuel cell and gasoline
processor technology, but also in electric motors and other electric drivetrain components. Because a
major breakthrough would be required to make this vehicle marketable, we do not attempt to
estimate its cost. Instead, we assume that it will be cost-effective on a life-cycle cost basis - that is,
that its incremental cost will be equal to its lifetime fuel savings. This implies a price increment of
$800. Note that this value is not meant to be interpreted as a forecast of likely future fuel cell costs;
instead it allows us to evaluate the consequences of such an optimistic outcome.
Some of the technologies necessary to produce an 84% efficiency gain for the fuel cell hybrid would
also make the internal combustion engine hybrids, both gasoline and diesel, somewhat more
efficient (e.g., ultra high-efficiency electric motors, improved energy storage devices with high
specific power and high in/out efficiency). Fuel economy gains for the gasoline and diesel hybrids
are boosted to 42% and 72%, respectively. A more optimistic assumption is made for the DISC
engine, as well. Its fuel economy benefit is increased to 23% from 18%.
If Intelligent Transportation Systems technologies are highly successful, they should be able to
improve traffic flow, resulting in higher on-road fuel economy. To reflect this, the on-road fuel
economy factor, which otherwise deteriorates by 3% from 1997 to 2015, is held constant. The high-
efficiency/low-carbon case further assumes that the current emphasis on horsepower (HP) will
abate substantially, although increased HP will still be valued. This case is consistent with a
change in attitudes favoring "greener" automobiles or policies to encourage higher MPG. To reflect
greater public concern over greenhouse gas emissions, the demand for increased horsepower is reduced
by decreasing its sensitivity to income, from an elasticity of 0.9 to 0.5.
As noted earlier, there are other potential technology breakthroughs capable of significantly
reducing greenhouse emissions (e.g. breakthroughs in batteries for electric vehicles, or in gas storage
for natural gas vehicles (see box)). These were left out of the high-efficiency/low-carbon scenario
not because they are necessarily less plausible than fuel cells, but because the inclusion of large
numbers of technology breakthroughs in a single scenario would be implausible.
5.34
September 17, 1997
Transportation Sector
Chapter 5
Other Potential Breakthrough Technologies
Aside from the new technologies postulated in the high-efficiency/low-carbon scenario, other
potential technologies could yield substantial reductions in greenhouse gas emissions with
technology breakthroughs or, in some cases, with a substantial market push. In the light-duty
vehicle market, for example, battery electric vehicles have potential to reduce greenhouse gases if
they can greatly increase their market share and improve their energy efficiency. For example,
several recent studies have concluded that, under plausible assumptions about EV efficiency and the
mix of fuels and technology used to generate recharge electricity, use of EVs will yield net reductions
of greenhouse gases. Delucchi (1997) estimates a national average reduction of 26% in 2015, with
power generation heavily weighted to coal; whereas Wang (1997) estimates a 19% reduction in 2005.
Areas with predominately natural gas-generated electricity could have much larger savings. Note,
however, that these results are dominated by assumptions about EV and baseline gasoline vehicle
efficiency, type of fuel and technology used for power generation, inclusion or exclusion of non-CO₂
greenhouse gases, and the types of trips replaced by EV use; it is relatively easy to construct
plausible scenarios with much higher or lower reductions in greenhouse gases, or even increases
(with coal-fired electric power and extremely efficient competing gasoline vehicles).
Crucial technological roadblocks for EV market penetration are:
Battery improvements especially higher specific energy and power, lower cost, improved
longevity, higher in/out efficiency,
Power electronics - especially lower cost, and
Electric motors - especially higher efficiency over a range of driving cycles and higher specific
power.
There are claims that transportation use of alternative fuels other than electricity (particularly
compressed natural gas) will yield strong greenhouse benefits. In natural gas's case, recent analyses
have shown contrasting greenhouse effects. For example, Delucchi (1997) estimates a 20% reduction
in greenhouse gases compared to gasoline use in 2015, whereas Wang (1997) estimates a 5% increase
in 2005. The primary difference in the two analyses is that Wang computes a 10% energy-efficiency
penalty associated with switching to CNG, based on recent test data; Delucchi estimates an 11%
improvement in energy efficiency based on potential efficiency gains from higher compression CNG
engines. Delucchi's optimism may well be the more appropriate approach for the longer term, but at
best CNG offers only a modest greenhouse emissions improvement.
Although we selected fuel cell vehicles fueled by gasoline (with onboard fuel processors) as the
"breakthrough" technology in the high-efficiency/low-carbor scenario, some analysts believe that
the direct use of hydrogen as a fuel is sufficiently more attractive to outweigh the disadvantages of
hydrogen's low energy density (complicating onboard storage) and lack of a supply infrastructure
(Ogden, 1977). The advantages of direct hydrogen include avoidance of the added weight and cost of
the fuel processor and larger fuel cell required (fuel cell performance is reduced because the processor
does not produce pure hydrogen), and reduced vehicle efficiency because of the energy losses in the
processor and added vehicle weight (assuming the higher fuel storage weight for hydrogen is less
than the weight savings from removing the processor and reducing fuel cell size). Although lack of
infrastructure still represents a barrier, there have been advances in small scale-steam reforming of
natural gas that could greatly ease the introduction of a viable hydrogen supply infrastructure
(Ogden, 1977).
September 17, 1997
5.35
Chapter 5
Transportation Sector
5.3.3.2 Changes to the Medium and Heavy Truck Model
Medium heavy trucks are typically operated locally in pick-up and delivery mode. For such
vehicles, hybrid technology, with regenerative braking and energy storage capabilities, should
offer significant advantages. It is assumed that a diesel hybrid becomes available to heavy trucks
starting in the year 2005. This technology is assumed to offer the same 72% fuel economy benefit as
the light-duty vehicle version.
Greater success in materials, aerodynamics, tires, and engines, should make these technologies more
economically attractive to truckers. Since the NEMS Heavy Truck Model does not explicitly include
an economic trade-off between fuel savings and technology penetration, this effect was simulated by
shortening the time to 99% penetration for each technology by 30%. For most technologies, this
implies a 15 year period from time of introduction to nearly full market penetration.
5.3.3.3 Changes to Other Modes
Several changes were made to the commercial air model inputs. Starting in 2005, propfans were
assumed to be available for smaller commercial aircraft. Propfans offer a 20-30% efficiency
improvement over high bypass turbofan engines, and 10-15% over even ultra-high bypass engines.
Development of propfans has been hindered by concerns about initial cost, maintenance, and
vibration. Propfans are made available to only one-third of new aircraft in the high-
efficiency/low-carbon scenario. Additionally, partial success in hybrid laminar flow (HLF)
technology to reduce drag is assumed by 2010. Although HLF has the potential to reduce fuel use by
15% or more, only a 9% efficiency gain is assumed due to the continuing difficulties in developing a
practical system. In the efficiency case, ultra-high bypass engines are assumed to give a 10%
efficiency gain, thermodynamic improvements provide a 15% gain, and advanced aerodynamics
yield an 18% improvement. In this case, those are increased to 17%, 18% and 27%, respectively,
certainly optimistic but not implausible estimates (for example, see Greene, 1992, Table 4).
The annual efficiency improvement rate for railroads is increased to 2.5%, still slightly lower than
the 2.8% rate achieved over the past two decades. Waterborne freight's efficiency improvement
rate is bumped up to 1% per year from 0.05% to reflect a 10% total efficiency gain achievable
through improved hull designs and coatings. In fact, these modes have substantial potential to use
alternative power plants and fuels, as reflected in the 2020 technology discussion below.
5.3.4 Comparison of Forecasts
The efficiency and high-efficiency/low-carbon scenarios indicate that advanced energy
technologies could reduce emissions of greenhouse gases from transportation by 12-17% by 2010 and by
18-25% by 2015 (Table 5.7). Although these are large changes, they may appear modest compared to
the changes in new vehicles, the "leading edge" of changes in the entire transportation fleet.
Changing the technology of transportation requires turning over a vast stock of vehicles, and this
requires decades. As a result, the impact of advanced technologies introduced between now and 2010
will only just begin to be felt in 2010 and will still not have achieved its full effect by 2015. This
phenomenon can be most easily seen by comparing the fuel economy of new cars and light trucks to
that of the entire fleet of light-duty vehicles. In the efficiency case in 2015, for example, new cars
average 41.4 MPG and light trucks 31.9 MPG (EPA-rated fuel economy), but the fleet as a whole lags
behind at 28.2 MPG (24.0 MPG onroad). Given enough time to turn over the stock of vehicles, the
eventual light-duty fleet MPG will climb about one-third higher to nearly 38 MPG (32 MPG onroad).
The time lag required for new technology to penetrate the light-duty vehicle fleet is a common
feature of all modes. Thus, the energy savings and greenhouse gas reductions shown in Tables 5.7 and
5.8 for 2010 and 2015 reflect less than half of the ultimate savings that the technology introduced
over this period will eventually achieve.
5.36
September 17, 1997
Transportation Sector
Chapter 5
Passenger car and light truck fuel economy improvements are, in general, attributable to the
combined effect of many fuel economy technologies rather than a single, dominant technology. A
number of improvements to conventional engines combine to increase average new vehicle MPG in
2010 by almost 20% for passenger cars and by about 10% for light trucks. These include engine friction
reduction, greater use of multi-valve engines, and variable valve timing and lift control.
Substitution of lighter weight materials, aerodynamic drag reductions, various transmission
improvements, and the combined effects of advanced lubricants, tires, and accessories, each
contribute 2-5% gains. Of all the technologies added to the efficiency and high-efficiency scenarios,
the lean-burn gasoline engines (DISC and 2-stroke) deliver the greatest fuel economy benefits, about
15% for passenger cars and 12% for light trucks. These numbers represent sales weighted average
effects, taking into consideration the fact that even in 2010 new vehicles are not equipped with
these technologies. Diesel and hybrid technologies each boost average new car and light truck fuel
economy by about 5% in 2010, their smaller impact being due to their smaller market shares.
The sales weighted average impacts of nine classes of fuel economy technologies in the high-
efficiency/low-carbon case are illustrated in Figures 5.7 and 5.8. The measured percent fuel economy
gain applies to the impact of the technology on the average fuel economy of all new passenger cars or
light trucks and, thus, takes into account the estimated market penetration for each category of
technologies. In 2010, passenger cars get a considerably greater benefit from engine efficiency
improvements than light trucks, but the gap narrows considerably by 2015. Although the DISC and
2-stroke technologies are the most significant new technologies in 2010, the gasoline fuel cell comes
on strong by 2015. The impact of the fuel cell in 2010 is obviously limited by the assumption that it
would be first introduced in 2007. The impacts shown in Figures 5.7 and 5.8 depend entirely on the
cost, fuel economy benefit, and introduction date assumptions shown in Table 5.2, and the way the
NEMS model translates those assumptions into market success. Thus, the graphs do not represent a
prediction of what specific technologies will achieve, but rather an illustration of what could
happen given the outstanding successes in fuel economy technology R&D, as reflected in our high-
efficiency/low-carbon scenario assumptions.
The 23% gain for light-duty vehicles in 2015 is just slightly higher than the 21% and 22%
improvements by freight trucks and rail in the efficiency scenario. Aircraft efficiency gains seem to
lag behind at a mere 9% in 2015, but this is due to the fact that air passenger efficiencies increase
the most (17%) in the business-as-usual case. In 2010, rail and air have made the greatest efficiency
gains over 1997. This is consistent with the record of the past quarter century, during which time
these two modes have led all others in energy-efficiency improvement.
September 17, 1997
5.37
Chapter 5
Transportation Sector
Table 5.7 Transportation Sector Projections to 2010 and 2015 Efficiency Scenario (cont. next page)
1997
2010
Change V. BAU
%
BAU
BAU
Efficiency
Change
Change
Energy Use (quads)
25.5
32.3
29.3
-3.1
-9%
Carbon Emissions (MMtC/Yr.)¹
487
616
543
-73
-12%
Passenger Cars**
171
184
160
-24
-13%
Light Trucks
113
166
143
-23
-14%
Other Modes
203
266
240
-26
10%
Fuel Use by Fuel Type (quads)
Motor Gasoline
15.1
18.0
15.2
-2.8
-15%
Cellulosic Ethanol (in motor gasoline)
0.0
0.0
0.5
0.5
Distillate
4.6
5.8
5.7
-0.1
-2%
Jet Fuel
3.6
4.7
4.2
-0.5
-11%
Residual
1.2
1.6
1.6
0.0
0%
Other
1.1
2.2
2.1
-0.1
-4%
Energy Use by Mode (quads)
Light-Duty Vehicles
14.6
18.2
16.3
-2.0
-11%
Passenger Cars**
8.8
9.6
8.6
-1.0
-11%
Light Trucks
5.8
8.6
7.7
-2.0
-11%
Freight Trucks
5.6
6.8
6.3
-0.5
-8%
Air
3.6
4.7
4.2
-0.5
-11%
Rail
0.5
0.5
0.4
-0.1
-16%
Marine
1.7
2.3
2.3
0
0%
Pipeline
0.8
0.9
0.9
0
0%
Other
0.2
0.3
0.3
0
0%
Energy-efficiency Indicators
New Car MPG
27.5
27.8
37.5
9.7
35%
New Light Truck MPG
20.5
20.6
27.1
6.5
32%
Light-Duty Fleet MPG
19.6
19.4
21.5
2.1
11%
Aircraft Efficiency (Seat-Miles/Gal.)
51.8
58.2
61.6
3.4
6%
Freight Truck Fleet MPG
5.6
6.0
6.8
0.8
13%
Rail Efficiency (ton-miles/1,000 Btu)
2.7
3.0
3.6
0.6
20%
Transportation Activity Levels (billions)
Light-duty Vehicle Miles of Travel
2262
2762
2774
12
0%
Freight Truck VMT
173
237
238
1
0%
Commercial Air Seat-Miles
1116
1729
1608
-121
-7%
Rail Ton-Miles
1208
1459
1464
5
0%
Marine Ton-Miles
892
1047
1050
3
0%
Note: Because some light truck energy use is included in the freight truck sector, the totals by mode will not add to
the totals by fuel type.
+
After all scenarios had been completed, a minor error was discovered in the NEMS passenger car fuel economy
technology input data. This error allowed four wheel drive improvements to be applied to certain categories of cars
to which they are, in fact, not applicable. The overall effect on new car fuel economy is less than 0.3 MPG in 2010
and less than 0.5 MPG in 2015.
** Motorcycles, which are always less than 1%, are included with passenger cars.
5.38
September 17, 1997
Transportation Sector
Chapter 5
Table 5.7 Transportation Sector Projections to 2010 and 2015 Efficiency Scenario (Continued)
1997
2015
Change V. BAU
%
BAU
BAU
Efficiency
Change
Change
Energy Use (quads)
25.5
34.0
28.7
-5.2
-15%
Carbon Emissions (MMtC/Yr.)
487
646
532
-114
-18%
Passenger Cars
171
192
154
-38
-20%
Light Trucks
113
174
133
-41
-24%
Other Modes
203
280
245
-34
12%
Fuel Use by Fuel Type (quads)
Motor Gasoline
15.1
18.7
13.5
-5.3
-28%
Cellulosic Ethanol (in motor gasoline)
0.0
0.0
0.4
0.4
Distillate
4.6
6.0
6.5
0.5
8%
Jet Fuel
3.6
5.0
4.2
-0.7
-15%
Residual
1.2
1.8
1.8
0.0
0%
Other
1.1
2.5
2.4
-0.2
-6%
Energy Use by Mode (quads)
Light-Duty Vehicles
14.6
19.1
15.5
-3.6
-19%
Passenger Cars
8.8
10.0
8.3
-1.7
-17%
Light Trucks
5.8
9.1
7.2
-1.9
-20%
Freight Trucks
5.6
7.1
6.3
-0.8
-12%
Air
3.6
5.0
4.3
-0.7
-15%
Rail
0.5
0.5
0.4
-0.1
-20%
Marine
1.7
2.5
2.5
0.0
0%
Pipeline
0.8
0.9
0.9
0.0
0%
Other
0.2
0.3
0.3
0.0
0%
Energy-efficiency Indicators
New Car MPG
27.5
27.9
41.4
13.5
48%
New Light Truck MPG
20.5
20.6
31.9
11.3
55%
Light-Duty Fleet MPG
19.6
19.5
24.0
4.5
23%
Aircraft Efficiency (Seat-Miles/Gal.)
51.8
60.6
66.1
5.5
9%
Freight Truck Fleet MPG
5.6
6.1
7.4
1.3
21%
Rail Efficiency (ton-miles/1,000 Btu)
2.7
3.2
3.9
0.7
22%
Transportation Activity Levels (billions)
Light-duty Vehicle Miles of Travel
2262
2914
2937
23
1%
Freight Truck VMT
173
250
251
1
0%
Commercial Air Seat-Miles
1116
1923
1759
-164
-9%
Rail Ton-Miles
1208
1535
1540
5
0%
Marine Ton-Miles
892
1099
1102
3
0%
September 17, 1997
5.39
Chapter 5
Transportation Sector
Table 5.8 Transportation Sector Projections to 2010 and 2015 High-Efficiency/Low-Carbon Scenario
(cont. next page)
1997
2010
Changes V. BAU
%
BAU
BAU
HE/LC
Change
Change
Energy Use (quads)
25.5
32.3
27.9
-4.5
-14%
Carbon Emissions (MMtC/Yr.)
487
616
512
-104
-17%
Passenger Cars**
171
184
147
-37
-20%
Light Trucks
113
166
132
-34
-21%
Other Modes
203
266
233
-33
-12%
Fuel Use by Fuel Type (quads)
Motor Gasoline
15.1
18.0
13.9
-4.2
-23%
Cellulosic Ethanol (in motor gasoline)
0.0
0.0
0.7
0.7
Distillate
4.6
5.8
5.7
-0.1
-2%
Jet Fuel
3.6
4.7
4.0
-0.7
-14%
Residual
1.2
1.6
1.6
0.0
0%
Other
1.1
2.2
2.1
-0.2
-8%
Energy Use by Mode (quads)
Light-Duty Vehicles
14.6
18.2
15.2
-3.0
-17%
Passenger Cars**
8.8
9.6
8.0
-1.6
-17%
Light Trucks
5.8
8.6
7.2
-1.4
-17%
Freight Trucks
5.6
6.8
6.2
-0.6
-9%
Air
3.6
4.7
4.1
-0.7
-14%
Rail
0.5
0.5
0.4
-0.1
-25%
Marine
1.7
2.3
2.3
-0.0
-1%
Pipeline
0.8
0.9
0.9
0.0
0%
Other
0.2
0.3
0.3
0.0
0%
Energy-efficiency Indicators
New Car MPG⁺
27.5
27.8
43.1
15.3
55%
New Light Truck MPG
20.5
20.6
30.8
10.2
50%
Light-Duty Fleet MPG
19.6
19.4
23.2
3.8
20%
Aircraft Efficiency (Seat-Miles/Gal.)
51.8
58.2
64.6
6.4
11%
Freight Truck Fleet MPG
5.6
6.0
7.0
1.0
17%
Rail Efficiency (ton-miles/1,000 Btu)
2.7
3.0
4.0
1.0
34%
Transportation Activity Levels (billions)
Light-duty Vehicle Miles of Travel
2262
2762
2806
44
2%
Freight Truck VMT
173
237
238
1
0%
Commercial Air Seat-Miles
1116
1729
1619
-110
-6%
Rail Ton-Miles
1208
1459
1467
8
1%
Marine Ton-Miles
892
1047
1051
4
0%
Note: Because some light truck energy use is included in the freight sector, the totais by mode will not add to the
totals by fuel type.
+
After all scenarios had been completed, a minor error was discovered in the NEMS passenger car fuel economy
technology input data. This error allowed four wheel drive improvements to be applied to certain categories of cars
to which they are, in fact, not applicable. The overall effect on new car fuel economy is less than 0.3 MPG in 2010
and less than 0.5 MPG in 2015.
** Motorcycles, which are always less than 1%, are included with passenger cars.
5.40
September 17, 1997
Transportation Sector
Chapter 5
Table 5.8 Transportation Sector Projections to 2010 and 2015 High-Efficiency/Low-Carbon Scenaro
(Continued)
1997
2015
Change V. BAU
%
BAU
BAU
HE/LC
Change
Change
Energy Use (quads)
25.5
34.0
26.7
-7.3
-21%
Carbon Emissions (MMtC/Yr.)
487
646
484
-162
-25%
Passenger Cars
171
192
134
-58
-30%
Light Trucks
113
174
114
-59
-34%
Other Modes
203
280
236
-44
-16%
Fuel Use by Fuel Type (quads)
Motor Gasoline
15.1
18.7
11.2
-7.5
-40%
Cellulosic Ethanol (in motor gasoline)
0.0
0.0
0.7
0.7
Distillate
4.6
6.0
6.7
0.7
-12%
Jet Fuel
3.6
5.0
4.0
-1.0
-19%
Residual
1.2
1.8
1.8
-0.0
-1%
Other
1.1
2.5
2.2
-0.3
-12%
Energy Use by Mode (quads)
Light-Duty Vehicles
14.6
19.1
13.8
-5.3
-28%
Passenger Cars
8.8
10.0
7.4
-2.6
-26%
Light Trucks
5.8
9.1
6.4
-2.7
-29%
Freight Trucks
5.6
7.1
6.2
-0.9
-13%
Air
3.6
5.0
4.1
-0.9
-19%
Rail
0.5
0.5
0.4
-0.2
-38%
Marine
1.7
2.5
2.4
-0.0
-1%
Pipeline
0.8
0.9
0.9
0.0
0%
Other
0.2
0.3
0.3
0.0
0%
Energy-efficiency Indicators
New Car MPG
27.5
27.9
50.2
22.3
80%
New Light Truck MPG
20.5
20.6
37.8
17.2
83%
Light-Duty Fleet MPG
19.6
19.5
27.1
7.6
39%
Aircraft Efficiency (Seat-Miles/Gal.)
51.8
60.6
70.7
10.1
17%
Freight Truck Fleet MPG
5.6
6.1
7.5
1.4
23%
Rail Efficiency (ton-miles/1,000 Btu)
2.7
3.2
4.8
1.6
51%
Transportation Activity Levels (billions)
Light-duty Vehicle Miles of Travel
2262
2914
2974
60
2%
Freight Truck VMT
173
250
252
2
1%
Commercial Air Seat-Miles
1116
1923
1923
-152
-8%
Rail Ton-Miles
1208
1535
1542
7
0%
Marine Ton-Miles
892
1099
1103
4
0%
September 17, 1997
5.41
Chapter 5
Transportation Sector
Figure 5.7 Sources of Fuel Economy Improvements in High-Efficiency Scenario, 2010
20
15
Percent Fuel Economy Gain
Passenger Cars
10
Light Trucks
5
0
Materials
Transmissions
Tire/Lube/Acc
TDI Diesel
Fuel Cell
Drag
Engine
DISC/2-stroke
Hybrids
Improvements
Figure 5.8 Sources of Fuel Economy Improvements in High-Efficiency Scenario, 2015
20
Passenger Cars
15
Percent Fuel Economy Gain
Light Trucks
10
5
0
Materials
Transmissions
Tire/Lube/Acc
TDI Diesel
Fuel Cell
Drag
Engine
DISC/2-stroke
Hybrids
Improvements
5.42
September 17, 1997
Transportation Sector
Chapter 5
Transportation activity increases at moderate rates in the business-as-usual case and, indeed, in all
the other scenarios as well. Transportation activity in the NEMS model is relatively insensitive to
energy prices. In the business-as-usual scenario, light-duty vehicle travel increases by 22% from
1997 to 2010, an average annual rate of just 1.5%. In the high-efficiency/low-carbon scenario, the
increase is 24%, reflecting a very small increase due to the lower fuel cost per mile of vehicle travel
(1.7%/year). Growth from 2010 to 2015 is slower still, 1.1% per annum. Air travel is the fastest
growing mode, with seat-miles growing at 3.4% annually through 2010 and slowing to 2.1% annually
from 2010 to 2015. Efficiency improvements in the efficiency and high-efficiency/low-carbon
scenarios include increased load factors (passenger-miles per seat-mile) so that seat-miles are
actually 7% lower in the efficiency case than in the business-as-usual case in 2010 (Tables 5.7 and
5.8). Freight truck vehicle miles increase at a faster rate than light-duty vehicle miles, 2.5% per
year through 2010, slowing to 1.1% from 2010 to 2015. These levels are almost unchanged by further
increases in truck freight energy-efficiency. NEMS measures rail and marine activity in ton-miles,
and these are up 21% and 17%, respectively, by 2010. Once again, the growth rate from 2010 to 2015
is at the much slower rate of about 1% per year.
The combined effects of moderately increasing transportation activity and significant efficiency
gains are still not enough to reduce energy use or carbon emissions below present levels by 2010.
Overall, transportation energy use in the business-as-usual case grows from 25.5 quads in 1997 to 32.3
in 2010 and 34.0 in 2015. The efficiency scenario lowers energy use by 9% in 2010 and carbon emissions
by an additional 3%, due to the success of cellulosic ethanol as a gasoline blending component Still,
energy use is up 15% over the 1997 level, and carbon emissions are 12% higher. In 2015, however,
energy use and carbon emissions are reduced compared to 2010 but still higher than in 1997.
Although the 1997 version of the NEMS model does not forecast beyond 2015, it is reasonable to
assume that energy use and emissions will continue to fall for a decade or so beyond 2015 as
technological improvements penetrate the stock of transportation vehicles.
Motor gasoline use, on the other hand, is only 0.15 quads higher in 2010 than in 1997, and is a full 1.6
quads lower than the current level in 2015. The use of 0.4 quads of cellulosic ethanol and an
equivalent shift to diesel are partially responsible for the reduction in gasoline consumption.
Because cellulosic ethanol produces almost no net greenhouse gas emissions, it is far more effective
than any fossil-based alternative fuel at reducing transportation's carbon emissions. Demand for
distillate and jet fuel combined is up 1.7 quads in 2010 and is 2.6 quads higher than the 1997 levels in
2015. The slower growth of gasoline demand suggests a change in refinery operations would be
required, but no analysis of the impacts of this change has been made.
The high-efficiency/low-carbon scenario achieves the milestone of reducing CO2 emissions below
1997 levels, but by 2015 rather than 2010 (Table 5.7). In 2010, CO₂ emissions are 17% (a full 100 MtC
per year) below the business-as-usual case, but still 4% above 1997 levels. With new cars at 43 MPG
(EPA test value), new light trucks at 31 MPG, and the fleet average at only 27 MPG (23 MPG onroad),
efficiency is improving rapidly and still has a long way to go. New passenger car MPG hits a fleet
average of 50 in 2015 in this scenario, buoyed by market shares of 25-30% for hybrid vehicles, and
15-20% for turbo-charged direct-injection diesel vehicles. Two-stroke engines are also popular in
this scenario, capturing about one-third of the small-car market. By 2015, all remaining new light-
duty vehicles are equipped with DISC engines, the gasoline engine of today having been all but
entirely squeezed out by newer technologies.
Yet even the high-efficiency/low-carbon case, with its breakthrough technology assumptions,
illustrates how much time it takes to fundamentally change the technology of transportation energy
use. Though fleet average light-duty vehicle MPG is up from 19.6 to 27.1 (onroad) by 2015, there is
another 10.3 MPG to go before the fleet achieves equilibrium with the efficiency of new vehicles.
Similarly, in the rail mode, use of fuel cells has penetrated only 5% of the stock of locomotives by
September 17, 1997
5,43
Chapter 5
Transportation Sector
2010 and 15% by 2015. In most cases, the majority of CO2 emission reductions have yet to be realized,
even by 2015. The point is not that little can be done to reduce transportation's CO₂ emissions. The
point is that if CO₂ emissions must be reduced, the sooner one gets started, the better.
5.3.5 Cost-Effectiveness of Light-Duty Vehicle Fuel Economy Improvement
The cost-effectiveness of technological changes that improve fuel economy is a very complex issue,
depending not merely on the value of fuel savings and the increase in retail price, but on how each
technology affects the performance, reliability, appearance and feel of a vehicle. Even such a
seemingly simple matter as computing the value of fuel savings is not straightforward, since it
depends on car buyers' expectations about future fuel prices, vehicle lifetime (or, alternatively,
market valuation of remaining fuel savings when the vehicle is traded in or resold), consumer
discounting of future savings, expectations about future depreciation of the vehicle's value, and
expected utilization rates.
Technological advances are likely to create new opportunities to provide other benefits of
importance to car buyers and to society. For example, multi-point fuel injection is generally held to
be not cost-effective solely on the basis of fuel savings - yet every new car sold and nearly every new
light truck is equipped with it. The reason for including fuel injection is that it improves
drivability and also is a critical technology for meeting emissions standards. Technologies included
in the efficiency scenarios also have the potential to create social benefits. By reducing oil
consumption, they will decrease the volume of U.S. oil imports. By making it easier and cheaper to
improve efficiency and substitute alternative energy sources for oil, these technologies will improve
U.S. energy security. Technologies such as hybrid vehicles and fuel cells will help vehicles meet
increasingly stringent emissions standards. Most importantly, technological advances will be
essential to creating a sustainable world transport system.
The cost of supplying technologies is also not a simple matter, since it depends on the rate at which
capital equipment must be replaced. If the rate of adoption exceeds the normal rate of turnover of
manufacturing equipment, the costs of technological change increase. Also, new technologies must
often be certified to meet safety and environmental standards, which takes time and involves some
degree of risk. Consumers expect a high degree of reliability of vehicles, and this might be
threatened by too rapid introduction of novel technologies.
For all these reasons, the NEMS model does not base technology adoption on a simple cost-
effectiveness calculation, but rather attempts to simulate the complex process described above. The
market penetration of fuel economy technologies is a function of cost-effectiveness, but is not solely
determined by it. Market penetration follow an s-shaped curve that predicts 50% market
penetration for precisely cost-effective technologies, with increasing or decreasing market share as
cost-effectiveness increases or decreases, respectively. This simulates the fact that consumers are
not identical in their valuation of technology (e.g., high mileage drivers such as sales
representatives might tend to value fuel economy more than would average drivers), and that
technologies have other characteristics that consumers may, or may not, value. Also, introduction is
not immediate when cost-effectiveness is reached, but is rather phased in over time, simulating a
normal process of retirement and replacement of manufacturing capital.
The phasing in of new technologies can be seen in Figures 5.9 to 5.11, which show the predicted
market penetrations of engine technologies. Engine technology penetrations in the efficiency case
are shown for passenger cars and light trucks in Figures 5.9 and 5.10. Although the DISC, TDI
Diesel, and Gasoline Hybrid technologies eventually come to dominate the market, it takes about a
decade for this to occur, allowing time for orderly introduction of the technologies. For comparison,
5.44
September 17, 1997
Transportation Sector
Chapter 5
the historical market penetration rates of fuel injection technologies are shown in Figure 5.11.
Although it took less time for multi-point fuel injection to replace carburetted fuel systems, this
technological change was urged on by emissions regulations. Nonetheless, as a point of comparison,
it suggests that the rates predicted by the NEMS model are comparable to similar historical
transitions.
Figure 5.9 Market Penetration of Advanced Engines for Domestic Passenger Cars - Efficiency
Scenario
1
Gasoline
0.9
Hybrid
0.8
TDI Diesel
0.7
DISC
0.6
Market Share
Conventional
0.5
0.4
0.3
0.2
0.1
0
1997
2000
2003
2006
2009
2012
2015
For all the reasons noted above, simple cost-effectiveness calculations based solely on incremental
first cost and the value of future fuel savings can be misleading. Indeed, the NEMS model outputs do
not include direct measures of the costs of technological changes or their value to vehicle purchasers.
However, for light-duty vehicles, approximate technology cost estimates can be derived from the
market shares of each technology and from the initial cost estimates. By comparing the weighted
average cost of fuel economy technology in the efficiency and high-efficiency cases in 2010 with the
weighted average cost in 1997 for the BAU case, we can obtain an estimate of the increase in retail
price per vehicle due to the adoption of fuel economy technology. The incremental costs must be
adjusted, however, to reflect the fact that a significant fraction of the potential MPG increase is
used in the NEMS model to produce higher horsepower or increased vehicle weight, or to offset
small MPG losses due to safety and emissions improvements. The cost adjustment is made by
multiplying the full cost increase by the ratio of the actual MPG gain to the potential MPG. For
example, for automobiles in the efficiency case, this ratio is 0.7. Using the same assumptions
employed in the model to calculate cost-effectiveness, we can also estimate the value to the average
consumer of the change in fuel economy. These estimates are summarized in Table 5.9.
September 17, 1997
5.45
Chapter 5
Transportation Sector
Figure 5.10 Market Penetration of Advanced Engines for Domestic Light Trucks - Efficiency Scenario
1
Gasoline
0.9
Hybrid
0.8
TDI Diesel
0.7
DISC
0.6
Market Share
0.5
Conventional
0.4
0.3
0.2
0.1
0
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
Figure 5.11 Penetration of Fuel Injection Technology
100
Diesel
80
Throttle-body
Fuel Injection
Market Share (%)
Port Fuel
60
Injection
Carburetor
40
20
0
1975
1980
1985
1990
1995
5.46
September 17, 1997
Transportation Sector
Chapter 5
Table 5.9 Simple, Total Cost-Effectiveness Estimates for Light-Duty Vehicle Fuel Economy
Technology
Value of Fuel Savings to Consumer
Scenario
MPG
Full Incremental
Adjusted*
10% Implicit
20% Implicit
Cost
Incremental Cost
Discount Rate
Discount Rate
Passenger Cars
Business as Usual
27.5
$0
$0
-
-
Efficiency
37.5
$850
$600
$1,600
$1,000
HE/LC
43.1
$900
$900
$2,150
$1,350
Light Trucks
Business as Usual
20.5
$0
$0
-
-
Efficiency
27.1
$800
$650
$1,950
$1,200
HE/LC
30.8
$950
$900
$2,700
$1,700
Gasoline prices assumed to remain constant at $1.20 per gallon. Vehicle usage rate of 15,500 miles per year,
declining with vehicle age at 4% per year, and lifetime of 14 years. For calculating value to consumers, MPG
estimates are reduced by 15% to reflect actual operating conditions.
*Adjusted to account for the use of fuel economy technology to increase horsepower instead of increasing miles per
gallon. The cost effectiveness estimates in Table 5.9 show that even at the higher 20% implicit discount rate, the
light-duty vehicle fuel economy improvements are, as a whole, cost effective. This is not surprising since the NEMS
model bases its technology market penetration predictions on a similar measure of cost effectiveness. Discounting
future fuel savings at a lower rate of 10% only improves cost effectiveness.
Based on a simple comparison of incremental vehicle costs to the value of fuel savings to the
consumer, fuel economy improvements in the efficiency scenarios appear to be cost-effective as a
whole. Savings exceed costs for both discounting formulas shown. Choosing the correct discount rate
is somewhat controversial since it depends on whether one believes that there are imperfections in
the market for fuel economy. In the buildings chapter, for example, a 7% real rate is used to discount
future fuel savings. We believe that a 20% implicit discount rate should be used for valuing light-
duty vehicle fuel economy savings for the following reasons. When a consumer invests in vehicle
technology to improve fuel economy, his or her decision-making calculus is analogous to a firm's
capital investment decision. Indeed, consumers can be though of as producing their own vehicle
travel with inputs of vehicles, materials, and labor. In making this decision, the consumer must not
only consider his or her discount rate (time preference or opportunity cost for money) but also the
depreciation of capital. In other words, there are two costs of capital that must be accounted for, the
time cost of money tied up in the capital and the depreciation of the capital. In general, the
depreciation in a car's value is much greater during the first few years of its life. Indeed, a very
significant depreciation occurs instantaneously when the first owner takes over possession from the
dealer. After that time, the car is no longer "new". The initial owners of vehicles tend to hold them
for about four years, on average, so that they bear a disproportionate share of the cost of
depreciation.
The tendency of used car markets to "bundle" vehicle attributes, rather than price each separately
may create a market imperfection that, when combined with the greater depreciation in value
during the first few years of ownership, implies that new car buyers may reasonably be expected to
demand a high rate of return in fuel savings for an investment in fuel economy technology. According
to this hypothesis, with the exception of a few highly visible items, used car prices are determined
by initial prices and the average rate of depreciation. That is, the value of fuel economy in the used
car market is determined not so much by the present value of future fuel savings, as by the
depreciated value of the initial investment in fuel economy. Assuming this market imperfection
exists, the cost to the new car buyer of an investment in fuel economy technology is determined by the
September 17, 1997
5.47
Chapter 5
Transportation Sector
depreciation in its value over the first four years, rather than by the consumption of its fuel savings
potential.
The combination of these two factors may lead new vehicle buyers to demand a rate of return much
higher than the simple discount rate. If one assumes a 20% depreciation during the first year of
vehicle ownership and 10%/yr. thereafter, then a consumer with a 7% real discount rate would, in
effect, discount the full 14 years of fuel savings at about 15% to16% to compensate for the cost of
depreciation during the first four years of ownership. If future fuel savings are computed using the
average usage rate for new vehicles, then future savings must be further discounted by 4% to 5% per
year to reflect the typical rate of decline in vehicle use with age. Taken together, these factors
imply that a new car buyer may appear to behave as if his discount rate for valuing future fuel
savings were 20%, when in fact his simple real discount rate is only 7%
These rough estimates should be treated with considerable caution. First, they represent a
comparison of total costs of fuel economy changes with total benefits, taxes included, rather than the
more correct comparison of marginal costs and benefits, excluding taxes. Markets will, in theory, stop
improving fuel economy when the marginal costs equal the marginal benefits. In general, this will
be at a lower level of fuel economy than the point at which total costs equal total benefits. Second,
the NEMS model represents technology adoption as a more complex process than a simple
computation of monetary costs and benefits, and attempts to simulate actual market behavior. Thus,
the calculations reported above do not correspond to the NEMS technology adoption methodology.
5.3.6 Oil Imports and Oil Market Benefits
The reductions in energy use achieved in the efficiency and high-efficiency/low carbon scenarios
represent significant reductions in U.S. petroleum demand which should result in reduced U.S. oil
import dependence and lower oil prices to consumers. Because of transportation's continuing
dependence on petroleum in the business-as-usual scenario, 95% of transportation energy is still
derived from petroleum in 2010. In the BAU scenario, transportation uses 30.6 quads (14.5 million
metric barrels per day (MMBD)) of petroleum products. Technological advances contained in the
efficiency scenario reduce petroleum consumption by 3.4 quads (1.6 MMBD) in 2010, and those in the
high-efficiency/low-carbor scenario produce total oil savings of 4.9 quads (2.3 MMBD).
Lower U.S. oil consumption due to more energy-efficient technology and substitution of cellulosic
ethanol should reduce U.S. oil imports, or reduce world oil prices, or both. The exact world oil
market response is indeterminate because it depends on the actions of the OPEC cartel. In a
competitive world oil market, the response to reduced U.S. demand could be predicted based on
knowledge of the U.S. and rest-of-world supply and demand curves for oil. But because the cartel's
supply does not necessarily follow the rules of competitive market behavior there is, in effect, no
OPEC oil supply curve. Faced with reduced demand, competitive producers would lower prices,
encouraging demand and driving out the higher cost producers until a new equilibrium were reached.
But a cartel can choose to cut production, raise production, or do nothing, making the ultimate
outcome uncertain. Cutting production would raise world oil prices but the cost to OPEC would be loss
of market share, a key determinant of market power.
No matter what the OPEC cartel chose to do, however, either U.S. imports would fall, or oil prices
would fall, or both, as a result of the technological advances reflected in the efficiency and high-
efficiency/low-carbon scenarios. This is illustrated in Figure 5.12, which shows U.S. long-run supply
and demand curves for petroleum derived from the 1997 Annual Energy Outlook's Low, High, and
Reference Oil Price Cases for 2010 (ELA, 1996b, Table C11). The curves clearly show that at the
reference case oil price of $20.41 per barrel, domestic supply and demand curves do not intersect, with
the result that the 12.9 MMBD shortfall must be imported.
5.48
September 17, 1997
Transportation Sector
Chapter 5
Figure 5.12 U.S. Oil Supply and Demand in 2010
25
20
Quantity (MMBD)
15
-
Supply
-
Demand
10
---
High Efficiency
5
0
0
10
20
30
Price (1995 $/BBL)
The advanced technologies of the high-efficiency/low-carbor scenario shift the demand curve
towards lower demand, and may also change its slope (perhaps making demand more responsive to
price). If we assume that the world price of oil does not change (to achieve this result, OPEC would
have to cut production by an amount roughly equivalent to the reduction in U.S. consumption), then
U.S. imports would be lower by about 2 MMBD. If OPEC maintains previous production levels or
increases its output, world oil prices would fall. As prices fall, U.S. domestic supply will decline
and demand will increase, pushing imports back up. However, to achieve the original level of
imports (12.9 MMBD), prices would have to fall by about $5 per barrel (given the supply and
demand curves shown in Figure 5.12). The $5/bbl. price cut would reduce the total cost of oil to the
economy by about $35 billion, and reduce the cost of oil imports by about $20 billion, in comparison to
the AEO97 reference case. The possible outcomes are: 1) U.S. imports are reduced by 2.3 MMBD or
more, 2) world oil prices fall by about $5/bbl or more, or 3) a combination of reduced imports of up to
2.3 MMBD and a price decrease of up to $5/bbl. occurs.
5.4 R&D POTENTIAL FOR ADVANCED TECHNOLOGIES IN 2020
5.4.1 Light-Duty Vehicles
Many of the advanced technologies that have the potential to impact U.S. automotive fuel use after
2010 or 2020 need considerable research and development work before they can attain
commercialization. The federal government has supported work on many of these technologies for
more than 20 years, beginning with the Energy Policy and Conservation Act of 1975. The current U.S.
R&D effort on the more exotic of the new technologies has been characterized as "the most
comprehensive, best organized, and best funded in the world" (U.S. Congress, OTA, 1995).
Nevertheless, over the years, federal funding for vehicle technology R&D has been erratic, and
there are continuing budget battles over funding for DOE's Electric and Hybrid Vehicle Program and
the PNGV as a whole. As noted above, the National Research Council Committee that is reviewing
the PNGV program has stated in no uncertain terms that they believe the program is seriously
under-funded relative to its ambitious goals.
The OTA identified several R&D areas that will require considerable new resources including:
safety; analysis and development of infrastructure for manufacturing, refueling, servicing, recycling,
and so forth; and development of new standards for new materials and fuels (U.S. Congress, OTA,
September 17, 1997
5.49
Chapter 5
Transportation Sector
1995). Also, OTA concluded that the current federal program may not take appropriate advantage
of the innovative capabilities of small business, especially with budget pressure on the National
Institute of Science and Technology's Advanced Technology Program and other R&D efforts that
focus on smaller companies.
Although there are many hurdles to overcome, a strong R&D effort coupled with a market or
regulatory incentive to improve fuel economy should be capable of producing, by 2020 or earlier, mid-
sized vehicles with fuel economies in the 60-80 MPG range and performance similar to current
vehicles - that is, "PNGV territory." Note that continuing fleet increases in power and performance
will tend to reduce future fuel economy potential, since generally there is a direct tradeoff between
performance and fuel economy. An optimistic vision of a potential high-efficiency/low-carbon
vehicle in 2020, assuming the necessary breakthroughs in a number of areas (e.g., manufacturing
processes for composite materials, two orders of magnitude reduction in fuel cell costs) would combine
the following characteristics:
Highly aerodynamic design with Cd of 0.22 or below;
Lightweight body with composite body structure (safer alternative: optimized aluminum);
Ultra-low rolling resistance tires, CR of 0.005 (about half that of today's tires);
Hybrid drivetrain with lightweight, highly efficient storage device (ultracapacitor or
flywheel) and electric motor/controller;
Fuel cell powerplant with advanced hydrogen storage or efficient fuel reformer (safer
alternatives: DISC engine or DI diesel with lean NOx catalysts); and
Use of high-efficiency/low-carbon accessories and low-energy-use design (e.g., advanced
window coatings and insulation).
The current PNGV program is addressing many of the remaining R&D roadblocks though some need
considerably more attention and the solution to others might be accelerated with greater resources.
For example, development of manufacturing processes for composites has been hit hard by budget
cutbacks; as noted, without major breakthroughs, composites will likely be too expensive to play a
major role in vehicle light-weighting. In addition, there is some concern that Japanese and European
firms are devoting more resources to DISC and DI diesel engines than are U.S. companies, and these
engines may play a critical role in future high-efficiency/low-carbon vehicles, especially if fuel
cell development is delayed or is unsuccessful at reducing costs sufficiently for commercialization.
Fuel cells are widely believed to be the most attractive powerplant option for future vehicles, and
recent progress in increasing their power density and lowering costs through reducing their platinum
requirements has been extremely promising. Nevertheless, as discussed previously, many hurdles
remain, and their costs must decline remarkably for them to compete successfully with internal
combustion engines. In fact, they would revolutionize the power generation industry long before they
reached the $30/kW level of ICEs.¹³
Although some may view a fuel cell hybrid vehicle as an "ideal" vehicle, there are sufficient
uncertainties in the potential of the technologies needed for such a vehicle, and sufficient
heterogeneity in regional requirements and markets, to imply that an ideal R&D program in light-
duty vehicle technologies should be a broad program incorporating a range of alternative technology
pathways to high vehicle efficiency and low emissions. A breakthrough in high-specific-energy
battery technology coupled with significant progress in electric motors and power electronics, for
5.50
September 17, 1997
Transportation Sector
Chapter 5
example, could put large numbers of efficient electric vehicles into many urban markets; in some of
those markets (e.g., California) both the overall emissions effects and the greenhouse gas emissions
effects could be extremely positive. Similarly, breakthroughs in on-board storage technology for
natural gas might allow substantial penetration of natural gas vehicles into many markets,
although the positive greenhouse gas emissions impact of such vehicles would likely be
substantially less than for EVs or fuel cell hybrids.
5.4.2 Freight Trucks and Locomotives
The diesel cycle engine will dominate the freight truck sector at least until 2020 because of its high
thermal efficiency, potential fuel flexibility, and durability. DOE's Office of Heavy Vehicle
Technologies (OHVT) within the Office of Transportation Technologies is attempting to develop
the enabling technologies needed to achieve fuel flexibility, ultra-low emissions, and high fuel
efficiency in all classes of trucks, buses, and other heavy vehicles such as off-highway vehicles.
The typical new Class 8 tractor trailer in 2020 is expected to achieve an on-road fuel economy of over
10 MPG, compared to about 7 MPG today, assuming a high-efficiency/low-carbon, low emission
diesel cycle engine (thermal efficiency of at least 55% at rated speed and load at the flywheel) and
other technologies such as reduced aerodynamic drag, low rolling resistance tires, and lightweight
materials (such as magnesium) become an economic reality.14 While many of these technologies
have already been demonstrated to a limited extent, a key enabler is a durable highly efficient
NOX catalyst capable of operating at high-efficiency/low-carbon in an oxidizing atmosphere.¹⁵
Fuel reformulation is envisioned, as well as nonpetroleum fuels, during this period (2000-2020).
However, as the efficiency of the diesel cycle becomes fully exploited (thermal efficiencies of over
63% will be highly unlikely), the hydrogen fuel cell, unconstrained by the Carnot cycle, may be the
next powerplant of choice for freight trucks, locomotives, and passenger cars. Significant R&D
efforts at DOE have enabled the demonstration of methanol-fueled fuel cell buses and other
vehicles. However, significant development of the fuel cell itself, power management strategies,
and hydrogen fuel production, distribution and storage are required, and economical solutions are
hard to envision before 2020. Particularly problematic are the low cost, efficient production,
delivery, and storage of hydrogen fuel (carbon-containing fuels significantly degrade fuel cell
thermal efficiency - in many cases to efficiencies below that of current-production diesel engines).
The fuel cell powerplant, combined with low aerodynamic drag, low rolling resistance tires, and
lightweight materials may raise Class 8 tractor trailer fuel efficiency to 15 MPG or more.
Locomotive engines may be an ideal test bed and an early entry for fuel cell powerplant technologies,
because sizes needed (4000 hp-equivalent) are on the scale of smaller stationary electrical power
generation plants which are already commercial. In addition, locomotives are already driven by
computer-driven electric motors for traction control.
5.5 SUMMARY
Cost-effective or near cost-effective technologies and alternative energy sources have the potential
to significantly restrain the growth of the U.S. transportation sector's greenhouse gas emissions
through 2010. There remains a substantial reservoir of proven technology for improving motor
vehicle fuel economy, and technologies that are very nearly market-ready (such as the DISC engine
with lean-NOx catalytic converter) will almost certainly further expand the potential to increase
MPG by 2010. New technologies and operational efficiency gains hold out similar potential for air
passenger travel and for truck and rail freight. Ethanol derived from cellulosic feed stocks instead
of grain could also make a significant contribution by 2010 as a blending component for conventional
gasoline if cost reductions foreseen by energy researchers are achieved. Overall, the combined
impact of such technologies could be to reduce greenhouse gas emissions by 10% in 2010 and by almost
September 17, 1997
5.51
Chapter 5
Transportation Sector
20% in 2015, relative to the business-as-usual case. If important breakthroughs can be achieved in
fuel cells and other key technologies, transportation's greenhouse gas emissions in 2015 could be held
below current levels.
In the business-as-usual case, transportation energy use grows from 25.5 quads in 1997 to 32.3 in 2010
and 34.0 in 2015. Carbon dioxide emissions, in million tonnes of carbon, increase from 487 in 1997 to
616 in 2010, and to 646 in 2015. As mentioned earlier, the business-as-usual case anticipates rates of
growth in transportation activity that are slow by historical standards. The actual outcome could
easily be 10% higher. The efficiency case holds transportation energy use to 29.3 quads in 2010 and
28.7 in 2015. Accordingly, carbon emissions grow to only 543 MtC in 2010 and 532 in 2015. The fact
that emissions are lower in 2015 than in 2010 reflects the fact that changing the technology of
transportation energy use requires the orderly turnover of durable capital stock. The high-
efficiency/low-carbon scenario holds 2010 carbon emissions from transport to 512 MtC, and reduces
2015 emissions to 484 MtC, just slightly below the 1997 level.
Changes in the mix of transportation fuels in the three 2010 scenarios are summarized in Table 5.10.
Although petroleum fuels are still the predominant source of energy for transportation, use of
alternative fuels expands in all three scenarios. Natural gas consumption for transport grows from
0.75 TCF in 1997 (about 98% of which is used in natural gas pipelines) to roughly 1.2 TCF in 2010. In
2010 pipelines still account for 70-75% of natural gas use, but CNG vehicles consume about 0.25 TCF,
and natural gas used to produce methanol for motor fuel accounts for nearly all of the rest. Biofuels
in the form of cellulosic ethanol come on strong in the efficiency and high-efficiency/low-carbon
scenarios, providing from one-fourth to one-third of an MMBD oil equivalent. In accord with the
AEO97 reference case projections, all scenarios foresee substantial increase in electricity use,
essentially all going to electric vehicles. The lower levels of electricity use in the efficiency and
high-efficiency/low-carbon scenarios, like those of natural gas use, are due to the general
improvements in vehicle technology in those scenarios.
Table 5.10 Transportation Energy Use by Fuel Type
2010
Fuel
1997
BAU
Efficiency
High-Efficiency
Petroleum Fuels (MMBD)
11.77
14.59
12.91
12.18
Natural Gas (TCF)
0.75
1.22
1.19
1.16
Biofuels (MMBD OE)
0.001
0.04
0.25
0.34
Electricity (MMBD OE)
0.04
0.23
0.22
0.20
Note: Petroleum fuels converted to million barrels per day oil equivalent using a heat content of 5.738
MMBtu/barrei. Natural gas includes pipeline fuel and natural gas used to produce methanol for use as a neat fuel,
but does not include natural gas used to produce methanol for use in Methyl Tertiary Buty! Ether. It is assumed
that, to produce one quad of methanol, 1.44 quads of natural gas are required. For electricity generation, 3.38
quads of primary energy are assumed to be required for each quad of electrical energy consumed.
Carbon emissions by mode are summarized in Table 5.11. Light-duty vehicles account for the vast
majority of carbon emissions reductions versus the business-as-usual case, with significant
contributions also being made by trucks and commercial aircraft. Rail freight shows the greatest
relative reduction, while emissions from shipping, military and "other" are essentially constant
across the three scenarios.
5.52
September 17, 1997
Transportation Sector
Chapter 5
Table 5.11 Carbon Emissions in 2010 (MtC)
2010
1997
BAU
Efficiency
High-Efficiency
Light-Duty Vehicles
278.7
346.3
297.3
273.0
Freight Trucks
73.3
95.0
83.4
80.9
Freight Rail
8.9
9.6
8.1
7.1
Shipping
30.8
42.7
42.3
41.6
Air Transport
60.0
83.5
72.9
69.6
Military, Transit, Other
35.3
39.0
39.3
39.3
TOTAL
486.9
615.9
543.3
511.5
Note: Breakdown into modal carbon emissions based on emissions factors taken from EIA (1996d) and DOE
(1996)
Most of the reduction in energy use and carbon emissions comes from light-duty highway vehicles.
There are four reasons for this. First, light-duty vehicle technology has been far more intensively
studied, so that a great deal more is known about the technological potential for this mode. Second,
the level of expenditure on technology R&D is greatest for this mode, with the possible exception of
aerospace R&D, including defense aerospace. Third, the commercial modes are believed to be more
sensitive to fuel costs and more aggressive in the adoption of energy-efficient technology. Therefore,
the rates of energy-efficiency improvement in the business-as-usual case are higher for these modes.
Finally, light-duty vehicles simply use more energy than any other mode: 60% in the business-as-
usual case. The other modes cannot be ignored, however, and should probably be given much greater
attention with respect to R&D investment.
Although technological improvements have the potential to cost-effectively restrain greenhouse
gas emissions from transportation, it is not likely that the changes will come about without a major
public policy initiative. There are two reasons for this. First, the problems posed by greenhouse gas
emissions are what economists term a classical public good externality. This means that the market
economy will not provide the right price signals either for the development or the adoption of low-
carbon technologies. Second, the AEO97 projections foresee a world where fossil fuels are abundant,
available and inexpensive. In particular, none of the oil market upheavals of the past quarter
century are present in the forecast. As a result, there are no other economic incentives to encourage
either energy-efficiency or alternative fuels. In such an environment, it is not reasonable to expect
either that appropriate technology will be developed or that success in the marketplace will result.
As a result, the efficiency case is based on the assumption that policies are implemented to promote
the development of cost-effective low-carbon technologies and to spur the adoption of these
technologies. In our view, this would include at a minimum a greatly increased public sector
investment in R&D addressing energy-efficient and low greenhouse gas technologies, perhaps two to
ten times the current level of effort. There are other public interests in developing such technologies
(e.g., energy security and environmental sustainability) that, we believe, could easily justify such a
level of investment. But policies to insure the adoption of low-carbon technologies in the market
would also be necessary. It is not the purpose of this study to recommend what those policies should
be; nonetheless, we are obliged to point out that meaningful policies will be necessary.
Indeed, technology has enormous potential to reduce transportation's greenhouse gas emissions.
Cost-effective technological change will take time however, and its full effects will not be felt for
two decades or more. Because the problems that may result from increased carbon emissions affect
September 17, 1997
5.53
Chapter 5
Transportation Sector
the global environment, significant reductions will demand meaningful public policy initiatives.
These must include a greater effort to develop low-carbon technologies and a commitment to
implement policies that will insure their adoption in the market.
5.6 REFERENCES
Boedecker, E., J. Cymbalsky, C. Honeycutt, J. Jones, A.S. Kyders, and D. Le. 1996. "The Potential
Impact of Technological Progress on U.S. Energy Markets", Issues in Midterm Analysis and
Forecasting 1996, U.S. Department of Energy, Energy Information Administration, Office of
Integrated Analysis and Forecasting, Washington, DC.
Boeing Commercial Airplane Group. 1995. Current Market Outlook 1995, Seattle, Washington,
May.
Borroni-Bird, C. 1997. Chrysler Corp., personal communication, March 14.
Bowman, D., P. Leiby, and J. Rubin, 1997. "Methodology for Constructing Aggregate Ethanol Supply
Curves", Draft report, Energy Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee,
March 24, 1997.
Buchholz, K. 1997. "Chrysler Updates Two-Stroke Engine Progress," Automotive Engineering, vol.
105, no. 1, p. 84, January.
Burke, A.F., 1995. "Electric/Hybrid Super Car Design Using Ultracapacitors," 30thh IGCEC
Meeting, Orlando, Florida, August.
California Air Resources Board (CARB). 1991. Locomotive Emission Study, Booz-Allen Hamilton,
report to ARB, Sacramento, California.
California Air Resources Board (CARB). 1992. Evaluation of Emission Controls for Locomotives,
ARB Report, Sacramento, California.
Cataldi, R. 1995. "Integrating Steel Wheels Into Sustainable Transportation," presented at the
Asilomar Conference on Transportation and Energy, Pacific Grove, California, Association of
American Railroads, Washington, DC.
Catalysts for Gasoline Fueled European Cars," Automotive Engineering, vol. 105, no. 2, pp. 133-135,
February.
Chem Systems, Inc. 1993. Technical Report Eleven: Evaluation of a Potential Wood-to-Ethanol
Process, DOE/EP-0004, Office of Domestic and International Energy Policy, U.S. Department of
Energy, Washington, DC, January.
Davis, S.C. and D.N. McFarlin. 1996. Transportation Energy Data Book: Edition 16. ORNL-6898,
Oak Ridge National Laboratory, Oak Ridge, TN.
Decision Analysis Corporation of Virginia. 1996. "NEMS Transportation Sector Model:
Documentation Update," Final, Subtask 18-2, prepared for the Energy Information Administration,
U.S. Department of Energy, Washington, DC, December.
Delucchi, M.A. 1991. Emissions of Greenhouse Gases from the Use of Transportation Fuels and
Electricity, vol. 1, ANL/ESD/TM-22, Center for Transportation Research, Argonne National
Laboratory, Argonne, Illinois, November.
5.54
September 17, 1997
Transportation Sector
Chapter 5
Delucchi, M.A. 1997. A Revised Model of Emissions of Greenhouse Gases from the Use of
Transportation Fuels and Electricity, University of California Institute of Transportation Studies,
February, Draft).
Energy and Environmental Analysis, Inc. 1994. Documentation of the Fuel Economy, Performance,
and Price Impact of Automotive Technology, Prepared for Martin Marietta Energy Systems, July.
Energy and Environmental Analysis, Inc., and Decision Analysis Corporation of Virginia. 1996.
NEMS Fuel Economy Model LDV High Technology Update, Draft Documentation, Subtask 9-1,
prepared for the Energy Information Administration, U.S. Department of Energy, Washington, DC,
June.
Energy Information Administration (ELA). 1994. Model Documentation Report: Transportation
Sector Model of the National Energy Modeling System, DOE/EIA-M070, Office of Integrated
Analysis and Forecasting, Washington, DC.
Energy Information Administration (ELA). 1996a. Annual Energy Review 1995, DOE/EIA-0384(95),
Washington, DC, July.
Energy Information Administration (ELA). 1996b. Annual Energy Outlook 1997, DOE/EIA-0383(97),
Washington,DC, December.
Energy Information Administration (ELA). 1996c. Alternatives to Traditional Transportation Fuels
1995, DOE/EIA-0585(95), Washington, DC.
Energy Information Administration (ELA). 1996d, Emissions of Greenhouse Gases in the United
States 1995, ELA-0573(95), Table B1.
Energy Information Administration (ELA). 1997. Monthly Energy Review February 1997, DOE/EIA-
0035(97/02), Washington, DC.
"Going with the Wind" 1984. Car and Driver, August.
Greene, D.L. 1992. "Energy-Efficiency Improvement Potential of Commercial Aircraft," Annual
Review of Energy and Environment, vol. 17, PP. 537-573.
Greene, D.L. 1996a. Transportation and Energy, Eno Transportation Foundation, Inc., Lansdowne,
Virginia.
Greene, D.L. 1996b. "Commercial Air Transport Energy Use and Emissions: Is Technology Enough?"
forthcoming, Sustainable Transportation: Is Technology Enough?, American Council for an Energy
Efficient Economy, Washington, DC.
Greene, D.L. and Y. Fan. 1995. "Transportation Energy Intensity Trends, 1972-1992," Transportation
Research Record, No. 1475, Energy and Environment, Transportation Research Board, Washington,
DC, 1995, PP. 10-19.
Greene, D.L., 1994. Alternative Fuels and Vehicles Choice Model, ORNL/TM-12738, Center for
Transportation Analysis, Oak Ridge National Laboratory, Oak Ridge, Tennessee, October.
Greene, D.L., and Y. Fan. 1994. Transportation Energy Efficiency Trends, 1972-1992, ORNL-6828,
Oak Ridge National Laboratory, Oak Ridge, TN.
Hadder, G. 1997. CITATION TO COME. Oak Ridge National Laboratory, Oak Ridge, Tennessee.
Hattori, et.al. 1990. A New 5-Speed Automatic Transmission for Passenger Cars, SAE Paper 900551.
September 17, 1997
5.55
Chapter 5
Transportation Sector
Heavenrich, R.M., and K.H. Hellman. 1996. Light-Duty Automotive Technology and Fuel Economy
Trends Through 1996, EPA/AA/TDSG/96-01, Office of Mobile Sources, U.S. Environmental
Protection Agency, Ann Arbor, Michigan.
Jost, K. 1997. "Gasoline-Reforming Fuel Cell," Automotive Engineering, vol. 105, no. 2, pp. 51-52,
February.
Leiby, P.N., D.L. Greene, and .Vidas. 1996. Market Potential and Impacts of Alternative Fuel Use
in Light-Duty Vehicles: A 2000/2010 Analysis, DOE/PO-0042, Technical Report Fourteen in the
Assessment of Costs and Benefits of Flexible and Alternative Fuel Use in the U.S. Transportation
Sector, Office of Policy, Office of Energy Efficiency and Renewable Energy, U.S. Department of
Energy, Washington, DC, January.
Lipman, T.E. and D. Sperling, 1997. "Forecasting the Cost Path of an Electric Vehicle Drive System:
A Monte Carlo Expreience Curve Simulation", Institute for Transportation Studies, University of
California at Davis, January.
Markus, M. 1997. "A Tale of Two-Strokes," Car and Driver, March, pp. 115-118.
McDonnell Douglas, Douglas Aircraft Company. 1996. "Outlook for Commercial Aircraft 1995-
2014," Market Planning, Long Beach, California.
McNutt, B., L. Fulton, and D. Greene. 1997. "Epilogue: Is Technology Enough? A Synthesis of Views
Expressed at the Conference" in Transportation, Energy, and Environment: Can Technology Sustain
Us? American Council for an Energy Efficient Economy, Washington, DC.
Murrell, J.D., J.A. Foster and D.M. Brister. 1980. "Passenger Car and Light Truck Fuel Economy
Trends Through 1980", SAE Technical Paper Series no. 800853, Society of Automotive Engineers,
Warrendale, Pennsylvania.
National Research Council, Aeronautics and Space Engineering Board. 1992. Aeronautical
Technologies for the Twenty-First Century, report of the Committee on Aeronautical Technologies,
National Academies Press, Washington, DC.
National Research Council, Standing Committee to Review the Research Program of the
Partnership for a New Generation of Vehicles. 1997. Review of the Research Program of the
Partnership for a New Generation of Vehicles: Third Report, National. Academy Press,
Washington, DC.
Oei, D-G. 1997. "Fuel Cell Engines for Vehicles," Automotive Engineering, Ford Motor Company,
February.
Ogden, J., et al, "Hydrogen as a Fuel for Fuel Cell Vehicles: A Technical and Economic Comparison,"
National Hydrogen Association 8th Annual Conference, March 1977.
Roberts, G.F. and D.L. Greene. 1983. Trends in Heavy Truck Energy Use and Efficiency, ORNL/TM-
8843, Oak Ridge National Laboratory, Oak Ridge, Tennessee, October.
Ross, M. Et al., 1996. "A Parallel Hybrid Automobile with Less Than 0.1 kWh of Storage," Society
of Automotive Engineers, Technical Paper Series, #961282, Warrendale, Pennsylvania.
Singh, Margaret. Argonne National Laboratory. Personal communications. April/May 1997.
Strehlau, W., J. Leyrer, E.S. Lox, T. Kreuzer, M. Hori and M Hoffman. 1997. "Lean NOₓ
5.56
September 17, 1997
Transportation Sector
Chapter 5
U.S. Congress, Office of Technology Assessment (OTA). 1995. Advanced Automotive Technology:
Visions of a Super-Efficient Family Car, OTA-ETI-638 (Washington, DC: U.S. Government Printing
Office, September.
U.S. Department of Energy. 1995. Overview of the U.S. Fuel Cell Program, DOE ATDCCM, October.
U.S. Department of Energy, Office of Heavy Vehicle Technologies. 1996. Office of Heavy Vehicle
Technologies Strategic Plan. Washington, DC.
U.S. DOE/PO-0042, 1996, Market Potential and Impacts of Alternative Fuel Use in Light-Duty
Vehicles: A 2000/2010 Analysis, Table D-4.
U.S. Department of Transportation, Bureau of Transportation Statistics. 1996. Transportation
Statistics Annual Report 1996, Washington, DC.
U.S. Department of Transportation, Federal Highway Administration. 1996. Highway Statistics
1995, FHWA-PL-96-017, Washington, DC, November.
Walsh, M., R. Perlack, D. Becker, A. Turhollow and R. Graham. 1997. "Evolution of the Fuel
Ethanol Industry: Feedstock Availability and Price", Draft quoted with permission of M. Walsh,
Oak Ridge National Laboratory, Oak Ridge, Tennessee, February 12.
Wang, M.Q. 1996 and 1997. GREET 1.0 - Transportation Fuel Cycles Model: Methodology and Use,
Argonne National Laboratory, ANL/ESD-33, June 1996, updated in 1997 (GREET 1.2).
Westbrook, F.W. 1989. "Allocation of New Car Fuel Economy Improvements, 1976-1989: Synopsis,"
submission to Oak Ridge National Laboratory, November.
Westbrook, F.W., and Patterson, P.D. 1985). Dynamics of Light-duty Vehicle Fuel Economy - 1978-
1984, SAE Technical Paper Series 850527.
Yamaguchi, J. 1997. "Toyota RAV4 EV - Today and Tomorrow," Automotive Engineering, February.
ENDNOTES
1 This rate applies to the period 1973 (18.605 trillion Btus) to 1985 (20.067 trillion Btus). Source is
Table 2.2, Monthly Energy Review, April 1997, DOE/EIA-0035(97/04), U.S. Department of Energy,
Energy Information Administration, Washington, DC.
2 This is a summary report. The full report, which presents this material, was not published due to
OTA's closure, but it is now available as part of a three-CD set that contains all of OTA's reports
since its inception. U.S. Congress, Office of Technology Assessment, OTA Legacy: 1972 through 1995,
U.S. Government Printing Office, Washington, DC, Stock no. 052-003-01457-2, $23 U.S.
3 The Toyota AXV5, with a Cd of 0.20, appears to avoid sacrifices in interior and cargo space.
Removing its wheel skirts, which might inhibit maintenance and restrict the vehicle's turning
circle, would likely raise its Cd to about 0.22. Because the vehicle's underbody cover adds weight,
the net positive effect on fuel economy will be reduced somewhat (U.S. Congress, OTA, 1995).
4 In particular, requirements for 0-60 mph acceleration and sustained gradeability.
September 17, 1997
5.57
Chapter 5
Transportation Sector
5 Also referred to as compression ignition direct injection (CIDI) engines and turbocharged direct
injecuon (TDI) diesel engines.
6 An accurate cost comparison would have to account for the transmission needed by the engine versus
the electric motor needed to convert the fuel cell's output electricity into shaft power. Also, the fuel
cell drivetrain may need a powerful battery to drive the vehicle until the cell can warm up.
7
At about 3700 psi storage pressure, storage volume for hydrogen is about 5 times that needed for
gasoline (Oei, 1997).
8 An additional cost may be the loss in system efficiency associated with onboard reforming as well
as the original refining of the gasoline. However, onboard hydrogen storage has energy costs in the
form of hydrogen production (probably at a large scale, and more efficient than the onboard
reformer) and pressurization if stored in-high pressure tanks.
9 Despite what is implied in the NEMS Transportation Model documentation, we were informed by
Mr. David Chien, principal in charge of the Transportation Model, that the model's calculations
were in 1987$. Thus, $8 in 1995 dollars equates to approximately $6 in 1987 dollars.
10 For carbon only; or $.08-$0.16 per gallon if the tax applies to all greenhouse gases, on a carbon
equivalent basis.
11 Communication from Margaret Singh of Argonne National Laboratory, April 3, 1997. Her
calculations were made using the August, 1993 version of M. A. Delucchi's greenhouse gas emissions
model and exclude any vehicle efficiency gains which might occur with the use of an ethanol
vehicle.
12 An alternative approach would have been to introduce these technologies using the Transportation
Model's alternative fuel vehicle capabilities. This approach was not taken on the grounds that
diesel is more conventional than an alternative fuel. Consumers are familiar with it, it is widely
available and, especially for the advanced, clean, TDI technology considered here, its performance
would be essentially identical to that of a gasoline vehicle.
13 Actually, for an accurate comparison, an ICE plus a transmission and inexpensive fuel tank should
be compared with a fuel cell, hydrogen storage or liquid fuel storage/reformer system, battery for
warmup power and power buffer, and electric traction motor, making the task of commercializing
fuel cells all the more onerous.
14 The Class 8 truck is very efficient already. Considering an energy per ton-mile measure of
performance, an equivalent passenger car needs to travel about 140 miles on a gallon of gas to be as
efficient as a 7 MPG Class 8 truck.
15 This technology is required for direct-injection gasoline engines as well.
5.58
September 17, 1997
The Electricity Sector
Chapter 6
Chapter 6
THE ELECTRICITY SECTOR'S RESPONSE TO END-USE
EFFICIENCY CHANGES
6.1 INTRODUCTION
Electricity consumption accounts for about 36% of both total primary energy consumption and carbon
emissions in the United States (EIA 1996a). As a consequence, converting efficiency-induced
electricity savings in the residential, commercial, and industrial sectors into carbon reductions is a
critical part of this study.
This task is complicated by several factors. First, the U.S. electricity industry is in the midst of a
major restructuring, from a highly regulated, vertically integrated industry to a largely
competitive, deintegrated industry. Because this transformation is far from complete, it is difficult
to predict the structure and operation characteristics of electricity markets for the year 2010.
Second, electricity production in the year 2010 will depend on the generating units that are retired,
repowered, and constructed between now and then, as well as on how those units are operated in 2010.
The decisions made by the profit-maximizing owners of individual generating units are likely to be
different than the cost-minimizing decisions made in the past by utility owners of large generation
and transmission systems. As a result of these changing dynamics of capacity expansion and system
operation, one cannot assume that the average and marginal carbon intensities of electricity use
(tonnes of carbon/GWh) will be the same in 2010 as they are today. Indeed, they are likely to be
quite different. Third, electricity prices in the year 2010 are likely to vary from hour to hour based
on current spot-market prices; consumer response to such time-varying prices is likely to be
substantial but is largely unknown.
The next section describes some of the key changes in the structure of the U.S. bulk-power system
that are likely to occur over the next decade. Section 6.3 describes the Oak Ridge Competitive
Electricity Dispatch (ORCED) model that is used here to project the characteristics of the
electricity sector in the year 2010. Section 6.4 compares ORCED's projections for the electricity
sector with those developed by ELA in its Annual Energy Outlook 1997 (AEO97). We then develop a
competitive-electricity market case for 2010, which is used as the base case against which the
efficiency and high-efficiency/low-carbon cases are compared.
6.2 BACKGROUND
In response to the 1992 Energy Policy Act, the Federal Energy Regulatory Commission (FERC) issued
a major order (Order 888) in April 1996, which it slightly revised in March 1997. This order requires
utilities to unbundle their generation and transmission services. A utility cannot offer preferential
transmission pricing for electricity generated by its own power plants. A key purpose of this order is
to eliminate problems associated with vertical market power in bulk-power markets and thereby
assure open access to the nation's transmission facilities.
Other factors are also forcing the U.S. electricity industry to change. These factors include low
natural gas prices (both today and over the next 10 to 15 years), substantial improvements in the
efficiency of gas-fired combustion turbines, and broad public sentiment to deregulate economic sectors
wherever possible.
September 15, 1997
6.1
Chapter 6
The Electricity Sector
We see a future in which the generation sector will be driven primarily by competitive forces rather
than by regulatory mandates. Decisions on whether, when, and where to build, repower, or retire
generating units will be made by investors, not by regulators.¹ Historically, vertically integrated
utilities have planned, built, and operated power plants to minimize the life-cycle costs of the
entire electric-power system over a long time (e.g., 20- to 30-year horizon). In tomorrow's
competitive environment, this decision rule will be replaced by one that emphasizes the
profitability of individual generating units over a much shorter time horizon, using a higher
discount rate to reflect the increased riskiness of power-plant ownership.
Our view of the future calls for most of today's utility-operated control centers to be replaced by
independent system operators (ISOs) that cover larger areas. As a consequence, the number of control
areas will decline from about 140 to perhaps only 20 to 50. Because these ISOs perform a monopoly
function, they will be regulated by FERC.
Similarly, transmission will remain a monopoly function, also regulated by FERC. Increasingly,
transmission will be separated from generation. Today, FERC requires utilities to "functionally"
unbundle generation from transmission. In the future, utilities will increasingly divest themselves of
their generation assets and will become "pure" transmission or transmission-plus-distribution
utilities. In this environment, transmission will become a common carrier.
6.3 MODEL DESCRIPTION
ORCED is an expanded version of part of a previously developed model called ORFIN (Oak Ridge
Financial Model) (Hadley 1996). Whereas ORFIN is a comprehensive electric-utility planning
model, ORCED deals only with generation. We developed ORCED to aid in the analysis of the
operation of competitive (as opposed to the traditional regulated) bulk-power markets. The model
allows the following issues to be examined:
Horizontal market power: concentration of generation assets among a few owners;
Generator profitability: which units will be retired because their expected revenues will not
cover the sum of their fuel costs, variable operations and maintenance (O&M) costs, and
(avoidable) fixed O&M costs, as well as repowering and new construction decisions;
Carbon emissions and other environmental effects of changes in the U.S. bulk-power sector; and
Optimal mix of new and existing generators, including new generating technologies.
The model is structured to allow simulation of different bulk-power market structures. In particular,
the user can specify various generation pricing schemes:
An energy-only spot price as proposed by the three California investor-owned electric utilities.
When unconstrained demand exceeds available supply, what would otherwise be unserved
energy is "curtailed" because spot prices rise sufficiently to suppress demand to match the level
of available generating capacity. The user simulates this situation by specifying a value for
the price elasticity of demand during these time periods. ORCED uses the amount of demand to
be curtailed and the price elasticity to calculate the value of unserved energy in e/kWh.²
An energy-only spot price plus the loss-of-load probability (capacity) component used in the
United Kingdom. Here, the user specifies a value for unserved energy (e.g., 200c/kWh), which
6.2
September 15, 1997
The Electricity Sector
Chapter 6
the model multiplies by the hourly value of the loss-of-load probability to produce a time-
varying increment to the energy-only spot price.
An energy-only spot price plus a capacity reservation price (in $/kW-year), as proposed by the
PJM Interconnection and the New England Power Pool. In this case, the user specifies an amount
of generating capacity needed for planning reserve, which determines the annual capacity
payments (in $/kW-year) required.
We are using ORCED to examine the issues listed above as functions of the following factors (in
addition to the pricing schemes noted above):
Characteristics of individual generators: type of unit, differences in capital and other fixed
costs ($/kW-year) vs. fuel and variable O&M costs (c/kWh), dispatchability (e.g., fully
dispatchable coal plant vs. must-run nuclear unit vs. stochastic wind plant), forced and planned
outage rates (%).
Customer and load characteristics: shape of load curve, price elasticities of demand, value of
unserved energy.
Generating-resource portfolio: mix of generating units and relationship between available
generating capacity and unconstrained peak demand.
ORCED includes a production-costing model that uses load-duration curves rather than
chronological loads as inputs. The model is run twice for each year of simulation, once for an on-peak
season and a second time for an off-peak season (Figure 6.1). We define the on-peak season as June
through August, and the off-peak season as the remaining nine months (September through May),
although the model can accept alternative definitions of the two seasons. The model can
incorporate disaggregate inputs on loads and load shapes for the residential, commercial, and
industrial customer classes. Data on these class loads are aggregated for use within ORCED, which
builds and dispatches generating units to meet aggregate load.
A load-duration curve is created by ordering demand (in MW) in terms of magnitude from highest to
lowest. The resultant curve shows the percentage of time that demand exceeds a particular value,
ranging from the one-hour peak demand down to the minimum demand.
Use of a load-duration curve to calculate production costs is much simpler and computationally much
less burdensome than use of chronological loads (i.e., hour by hour loads). This simplification,
however, has a price: because it obscures the timing of loads, one cannot accurately calculate
production costs on the basis of generating-unit details, such as minimum and maximum loading
points, startup times and costs, and minimum shutdown times. To partially remedy these problems,
ORCED analyzes production costs using the two load-duration curves, one for the three-month
summer peak period and the other for the nine-month off-peak period. ORCED also simulates the
effects of startup costs for those units with capacity factors of less than 10%.
For each season, the model has available to it 26 generating units. The first 25 units are
characterized in terms of capacity, forced and planned outage rates, fuel type, heat rate, variable
O&M costs, fixed O&M costs, and annual capital costs (based on initial construction cost, year of
completion, and capitalization structure). The 26th unit is an energy-limited hydro unit, for which
the inputs include, in addition to those noted above, the plant's capacity factor (equivalent to its
maximum energy output for the year). This treatment of hydro as energy-limited ensures that hydro
displaces the most expensive energy (i.e., at the top of the load-duration curves).
September 15, 1997
6.3
Chapter 6
The Electricity Sector
Figure 6.1 Example Load-Duration Curves for Peak and Off-Peak Seasons
100
90
80
70
60
50
Percent Peak
40
Peak Season
30
Off-peak Season
20
Annual
10
0
0
20
40
60
80
100
Percent Season
The model dispatches these 26 generating units separately for the two seasons. Although the
calculation process is the same for the two seasons, the results differ because of differences in the
load-duration curves and because all the planned maintenance is assumed to occur in the off-peak
season.
The plants are first dispatched against the load-duration curve on the basis of bid price, the default
for which is variable (fuel plus variable O&M) costs. (If the user bids a zero price for a unit, the
generator is treated as a must-run unit and is dispatched first by the model.) Because plants are not
available 100% of the time, we also model forced outages on a probabilistic basis.³ Thus, the
higher-cost plants will see demands not only from customers, but "equivalent demands" based on the
probability that plants lower in the dispatch order (i.e., less expensive) will be undergoing a forced
outage. The model creates an equivalent load-duration curve for each plant, which extends the
amount of time the plant runs based on the forced-outage rates of the plants lower in the dispatch
order.
Model results include spot prices for each point on the two load-duration curves. These prices are
based on the bid prices for each generator. The prices also reflect any externally imposed uplift
charge, capacity charge, or emissions taxes. Finally, the prices during high-demand hours reflect
generating-unit startup costs and the costs of unserved energy for those hours that unconstrained
demand exceeds supply. See Appendix F for more details on the inputs and results from ORCED.
ORCED can be run iteratively to estimate the response of customers to changes in overall and time-
of-use electricity prices. User inputs include an overall price elasticity of demand and a time-of-use
elasticity. The overall price elasticity adjusts the entire load-duration curve up or down in response
to decreases or increases in the average price of electricity. The time-of-use elasticity adjusts each
point on the load-duration curve up or down based on price decreases or increases during that time
period.
6.4
September 15, 1997
The Electricity Sector
Chapter 6
In addition, the model can use the time-of-use elasticity to compute the value of unserved energy (in
e/kWh) that equilibrates supply and demand when unconstrained demand would otherwise exceed
online supply. Alternatively, the user can input an estimate of the value of unserved energy, which
is then used to calculate the costs associated with those times when unconstrained demand would
exceed supply. A third approach involves user specification of a minimum reserve margin and
associated annual capacity payment (in $/kW-year) to pay for this "extra" capacity.
In addition to dispatching power plants and computing production costs, the model can also
"optimize" the mix of generating units available for the year of analysis. (That is, the model
includes a capacity-expansion module as well as a production-costing module.) We put optimize in
quotes because the factor on which to optimize is almost certain to be different for a competitive
electricity industry than it was for the regulated electric utility industry. For example, the model
could choose from among the following optimization functions:
Minimize total costs;
Minimize avoidable costs (fuel, variable and fixed O&M);
Minimize electricity prices; or
Maximize generator earnings.
In Figure 6.2 we show the impact of different objective functions for optimization. Minimizing price
does not necessarily minimize cost, because prices are based on the variable costs only and ignore
fixed costs. Given choices of technology, ORCED would select low variable cost but high total cost
technologies. Conversely, maximizing earnings does not raise total cost. Instead, the model selects
high variable/low fixed cost technologies. For our analysis we chose to optimize on minimizing
avoidable costs; we made this choice because it was conceptually the most appealing and the results
were the most reasonable for a system-wide optimization. For plants not yet built, their capital
costs as well as all operating costs are avoidable. Since no costs have been expended to build them,
their construction costs are not "sunk" and can be avoided by not building them.
In addition to specifying the optimization function, the user can also specify constraints on
individual generating units or the mix as a whole. For example, the user could set minimum and/or
maximum capacity levels for each generator. Maximum levels could be specified for those units that
were built earlier. Minimum levels could be specified for those units that must be available for
policy reasons (e.g., renewable resources that might not be fully cost-effective but are deemed
desirable from a broad societal perspective). Also, minimum levels for new plants may be specified
to represent plants built between the current year and the study year. Otherwise, the model may
choose not to build these intermediate plants, selecting only those with the most advanced
technologies. Constraints could be specified for a minimum capacity reserve margin, for a maximum
carbon emission allowance, or to ensure that each generating unit recovers at least its variable plus
avoidable fixed costs, and so on.
Since ORCED is written in Microsoft Excel, there are several methods that can be used for
optimization. The easiest is to use the built-in Solver tool. A single cell can be identified as the
objective function to be minimized, and other cells can be identified as variables, with constraints
placed on their values and/or other parameters within the spreadsheet. Since the problem is non-
linear, Solver uses a Generalized Reduced Gradient method, running the model thousands of times
searching for a solution. Another method, which is generally slower but avoids problems of local
optima, is the use of genetic algorithms.
September 15, 1997
6.5
Chapter 6
The Electricity Sector
Figure 6.2 Generation Price and Costs Using Different Optimizations (10% Fixed Reserve Margin)
3.0
2.5
2.0
1.5
Min Avoid. Cost
Min Total Cost
Min Price
1.0
e/kWh
Max Earnings
0.5
0.0
Price
Variable
Avoidable
Total Cost
-0.5
Cost
Cost
Earnings
-1.0
6.4 SCENARIOS FOR 2010
6.4.1 Calibration to EIA AEO97
Before analyzing the two end-use efficiency scenarios, we first calibrated ORCED results to those
produced by ELA's NEMS model for 1995 and 2010. Unfortunately, reconciling the two sets of results
to each other is difficult because of differences in the ways that the two modeling systems classify
various costs (e.g., fuel, variable O&M, fixed O&M, and capital costs associated with generation) as
well as ELA's inclusion of administrative and general (A&G) and customer service costs in the basic
categories of generation, transmission, and distribution costs.4 Because of these difficulties, our
numbers do not always match the ELA numbers exactly.
This calibration ensures that the assumptions concerning the mix of generating units, fuel prices,
customer demand, environmental regulations, and so on are consistent between ORCED and those
developed by ELA. For example, both sets of results assume the continuation of current economic and
environmental policies affecting the U.S. electricity industry. However, EPA's proposed regulations
to tighten standards for emissions of nitrogen oxides and small particulate matter are reflected in
neither the ELA nor the ORCED results.
We first developed a base case for the year 2010 that includes the same mix of generating units (in
both capacity and energy) as that produced by ELA, with the same reserve margin (11%), as shown
in Table 6.1. In addition to data from the AEO97, we used other data from ELA (ELA 1996b, EIA
1996c, ELA 1995), as well as data from the North American Electric Reliability Council (NERC
1996), the Electric Power Research Institute (EPRI 1993), and a compilation of various official
databases by Resource Data International, Inc. (RDI 1996).
6.6
September 15, 1997
The Electricity Sector
Chapter 6
Table 6.1 Comparison of Year 2010 AEO97 and ORCED Estimates of U.S. Generating Capacity and
Generation
Percent of generating capacity
Percent of generation
EIA
ORCED
ELA
ORCED
Hydro+other renewables
13.4
11.3
9.6
9.2
Coal
35.0
36.9
50.1
50.8
Nuclear
10.2
11.1
15.8
15.5
Oil
3.2
3.1
1.5
0.1
Gas
38.3
37.6
23.0
24.4
ORCED analyzes the generation sector only; the model is silent with respect to the costs of
transmission, distribution, and customer service. ORCED produces the following cost estimates for
2010, all expressed in 1995 dollars and adjusted upward by 1/0.93 to reflect the 7% T&D losses
between the generator busbar and the customer meter.⁵
Fuel
1.35
Variable O&M
0.18
Fixed O&M
0.51
Capital
1.00
Total
3.04c/kWh
We developed an estimate of the ELA capital cost of generation by subtracting estimates of the
capital costs associated with transmission, distribution, and administrative and general (A&G)
services from ELA's total capital cost:
[2.32 - (0.52 + 1.46) * 0.63 - 0.08] = 1.00c/kWh. 6
Total (Trans + Dist)
A&G
Our estimate of the ELA fuel cost is based on the sum of ELA's fuel cost plus 88% of its wholesale
purchase cost:
(0.98 + 0.67*0.88) = 1.57c/kWh. 7
Fuel Wholesale
We used the ORCED estimates of fixed and variable O&M costs to impute a comparable (i.e., equal)
value for ELA.
The net result is very close agreement between the ORCED and EIA scenarios for 2010 (Table 6.2).
ELA's estimate of the total cost of generation (3.26e/kWh) is 7% higher than the ORCED result
(3.04c/kWh). The ORCED cost is lower because ORCED dispatches fewer expensive oil-fired
resources than does the ELA model (Table 6.1). These differences in dispatch and variable costs occur
because ORCED dispatches generation nationwide and ignores transmission constraints. The close
agreement between ELA and ORCED results, in spite of all the adjustments required to produce a set
of internally consistent and comparably defined terms, is reassuring. It lends confidence to our
September 15, 1997
6.7
Chapter 6
The Electricity Sector
development of alternative cases in which we intended to reflect more fully than ELA did the effects
of competition in bulk-power markets.
Table 6.2 Comparison of EIA and ORCED Estimates of Generation Costs (1995e/kWh)
Generation costs
ELA AEO97
ORCED
Capital
1.00
0.99
Fuel
1.57
1.36
O&M
0.69
0.69
Total
3.26
3.04
6.4.2 The Base Case for a Competitive Market
Beginning with the AEO97 case, we developed a case intended to reflect the workings of a fully
competitive bulk-power market in the year 2010. (The AEO assumes a continuation of current
economic regulation, as indicated above, and therefore does not account for the possible effects of a
restructured and largely competitive U.S. electricity industry.) To reflect these changes, we let the
model select the amounts of each of the generating units that minimize the sum of variable plus
avoidable costs. Instead of specifying the amount of generating capacity that must be online in 2010
(to yield a 10.7% reserve margin in the AEO97 case), we allowed the model to select the amount of
capacity that minimized the cost of the power-supply system plus the cost of unserved energy. We
used a demand elasticity of -0.05 for those time periods when capacity is insufficient to meet
unconstrained demand. The resultant optimization yielded a reserve margin of 6.8%.
In general, we set prices equal to their real-time (hourly) values based on the variable (fuel plus
variable O&M) cost of the unit on the margin each hour of the year, adjusted overall electricity
demand to reflect lower prices using an assumed overall elasticity of -0.5, and adjusted the load
shape to reflect the response to real-time pricing using a value of -0.1 for the price elasticity within
each time period.
Beginning with the ORCED run that matches the AEO97 values for 2010, we first reran the model
allowing it to select the "optimal" amounts of generating capacity from among all the plants that,
according to ELA, are scheduled to come online after the year 1998. (The optimization was based on a
minimization of avoidable costs.) We also allowed the model to select plants for retirement. For
each of the new plants, we use the levelized fixed charges rate to calculate the annual capital cost
of the plant and treat all fixed costs (both capital and O&M) as avoidable.
Next, we adjusted the load-duration curves for the two seasons simulated by the model (peak and
off-peak). The new system load has a peak demand that is 3.4% below the AEO97 case and a total
demand that is 1.2% higher. We then reran ORCED using the new load shapes. Table 6.3 compares
the two sets of results. Although demand is higher in the restructuring case than in the AEO97 base
case, carbon emissions are lower. The lower carbon emissions are the result of reduced coal use and
increased natural gas use in the restructuring case.
Notice that, under restructuring, the total generating cost goes down 0.3c/kWh and yet the system
average price increases slightly. In a deregulated market, prices will be based on the variable cost
(or bid price) of the most expensive plant at that time. This is known as the "market clearing price."
Consequently, the link between total costs and prices is broken. Whereas current electric utility
6.8
September 15, 1997
The Electricity Sector
Chapter 6
regulation sets prices to recover total costs, future prices may, or may not, be sufficient to recover all
costs.
Overall, ORCED prices under restructuring did not differ greatly from the AEO97 price forecast.
Part of the reason for this similarity of results is that ELA did assume some cost decreases in their
model; we did not assume further O&M cost decreases on a plant-by-plant basis due to the pressure of
competition. Also, the objective function used in our cases was to minimize avoidable costs on a
system-wide basis. This did not necessarily create the lowest-priced scenario, as discussed in section
6.3. Because prices were similar to those in the AEO97 (less than 5% difference when factoring in
transmission and distribution prices), we simply used the AEO97 price forecasts for analysis of
energy-efficiency savings in the other chapters of this report.
Table 6.3 Comparison of Year 2010-Forecasts: AEO97 and the ORCED Base Case for a Competitive
Electricity Industry
AEO97
Competitive Industry
Restructuring
Peak demand, GW
734
709
Total end-use electricity sales, TWh
3,784
3,828
System load factor, %
62.8
65.7
Reserve margin, %
10.7
6.8
Generation shares, %
Coal
50.8
47.4
Gas
24.4
29.2
Other
24.8
23.4
Generation prices and costs, c/kWh
Retail price
3.04
3.02
Variable cost
1.43
1.45
Total cost
2.81
2.51
Carbon emissions, MtC
631
625
6.4.3 Efficiency and High-Efficiency/Low-Carbon Cases
We next applied the electricity-savings estimates, described in Chapter 3 for the residential and
commercial sectors and in Chapter 4 for the industrial sector, to adjust the aggregate load shape for
the United States as a whole. We then reran ORCED using the new, lower load shapes. As in the
reference case, capacity was optimized to minimize avoided costs, and dispatched on the basis of
lowest variable costs. Avoided costs for existing plants only included their variable, start-up, and
fixed O&M costs, while plants to be built between 1997 and 2010 also included the annualized
capital cost of their construction. As part of the high-efficiency/low-carbon case, we included an
additional cost of $50 per tonne of carbon, to be consistent with the rationale used in the demand-
side efficiency scenarios.
September 15, 1997
6.9
Chapter 6
The Electricity Sector
Once power plant production levels were determined, carbon production and primary energy use could
be calculated. We calculated marginal carbon savings, the carbon saved by the reduction in energy,
bv taking the difference in carbon production and dividing by the reduction in energy-demand. This
is in contrast to using the average carbon intensity as an approximation of the carbon saved per unit
of energy saved. The marginal carbon and energy savings take into account the change in production
mix that occurs with energy- efficiency and carbon-reduction measures. A plot of the carbon savings
and primary energy used in each scenario shows the changes as a function of the end-use energy
saved (Figures 6.3 and 6.4). The slopes of the lines represent the marginal carbon and primary
energy saved.
Table 6.4 summarizes the results for the three scenarios. Table 6.5 describes these results in terms of
percent change relative to the base case for a competitive utility industry. The efficiency case
yielded a 8.5% reduction in electricity. end-use energy and the high-efficiency/low-carbon case
reduced end-use energy by 16.3% relative to the restructuring case summarized in Table 6.3.
Figure 6.3 Carbon Production Versus End-Use Demand: Reductions Due to Energy Efficiency and
Carbon
Management
640
Restructure
620
Efficiency
600
Slope = 90 tC/GWh
580
Carbon, MI
560
540
Slope = 350 tC/GWh
520
500
Hi Efficiency/
Low Carbon
480
3000
3200
3400
3600
3800
4000
2010 Electricity End-Use Demand, TWh
Note: Slope is equal to marginal carbon savings in tonnes of carbon per gigawatt-hour.
6.10
September 15, 1997
The Electricity Sector
Chapter 6
Figure 6.4 Primary Energy Production Versus End-Use Demand: Reductions Due to Energy Efficiency
and Carbon Management
40
39
Restructure
38
37
Efficiency
36
Slope = 6200 BTU/kWh
35
34
33
32
Primary Energy, Quads
Slope = 13600 BTU/kWh
Hi Efficiency/
31
Low Carbon
30
3000
3200
3400
3600
3800
4000
2010 Electricity End-Use Demand, TWh
Note: Slope is equal to marginal energy savings in Btu per kilowatt-hour.
Table 6.4 Comparison of Year 2010 Forecasts: Results of Efficiency and High-Efficiency/Low-
Carbon Scenarios
Competitive
Efficiency
High-Efficiency/
Industry
Low-Carbon
Optimization
Peak demand, GW
709
651
596
Total Primary Energy used, quads
37.9
35.9
31.8
Total electricity generated, TWh
4,090
3,740
3,420
Total end-use electricity demand, TWh
3,830
3,500
3,200
System load factor, %
65.7
65.5
65.5
Reserve margin, %
6.8
7.9
12.9
Generation shares, %
Coal
47.4
52.1
46.2
Gas
29.2
22.2
26.0
Other
23.4
25.7
27.8
Generation prices and costs, e/kWh
Retail price
3.02
3.03
3.66
Variable cost
1.45
1.43
2.07
Total cost
2.51
2.46
3.21
Carbon emissions, MtC
625
596
492
Average carbon emissions, kg/MWh
163
170
154
Marginal carbon saved, kg/MWh
I
89
350
September 15, 1997
6.11
Chapter 6
The Electricity Sector
Table 6.5 Comparison of Year 2010 Forecasts: Effects of Efficiency and High-Efficiency/Low-
Carbon Scenarios
Change relative to the competitive utility industry base case
Efficiency
High Efficiency/Low Carbon
Peak demand
-8.2%
-16.0%
Total energy
-8.5%
-16.3%
System load factor
-0.2% points
-0.2% points
Reserve margin
+1.1% points
+6.1% points
Generation shares
Coal
+4.7% points
-1.2% points
Gas
-7.0% points
-3.2% points
Other
+2.3% points
+4.4% points
Generation prices and costs
Retail price
+1%
+21%
Variable cost
-1%
+43%
Total cost
-2%
+28%
Carbon emissions
-4.6%
-21.2%
The efficiency scenario forecasts a lower percentage reduction in carbon (4.6%) than in end-use energy
(8.5%) (Figure 6.5). The difference occurs because lower end-use demands translate into less
construction and operation of high-efficiency gas-fired combustion turbines and combined-cycle units
(Figure 6.6). (Because of their high capital costs, ORCED selects only the minimum amounts of
advanced coal and renewable technologies in the base restructuring case.) In the efficiency case,
relative to the restructuring case, capacity and generation for combined-cycle units declines by about
27%; capacity for combustion turbines drops 28% (generation drops by 68%); and coal and gas-
powered steam plants actually increase their production slightly.
Identification of high-efficiency gas-fired combustion turbines and combined-cycle units as the
marginal plants under consideration has the effect of substantially lowering the avoided carbon
from electric efficiency improvements. The forecasts contained in AEO97 are consistent with this
assumption through the year 2010. For example, page 49 of the AEO97 states that, "of the new
capacity [required through 2015], 81 percent is projected to be combined-cycle or combustion turbine
technology fueled by natural gas or both oil and gas."
In contrast, the HE/LC case forecasts a higher percentage reduction in carbon (21.2%) than in end-use
energy (16.3%). This is because the inclusion of the charge of $50 per tonne of carbon changes the mix
of technologies used to produce electricity so that low-carbon supply options are favored. Over 16%
of the coal capacity is retired in this scenario, and the remaining, more-efficient coal plants operate
at a lower capacity factor. Overall, generation from coal declines 18% compared to the reference
case. Capacity from combined cycle plants actually increases over the amount in the efficiency case,
but still has a 14% decline from the reference case. Combustion turbine generation declines 84% from
the reference case while combined cycle generation declines only 13%.
6.12
September 15, 1997
The Electricity Sector
Chapter 6
Figure 6.5 Energy and Carbon Changes from Restructure Case
- Hi Efficiency/ Low
Efficiency
Carbon
0
-5
Percent Change
-10
-15
-20
Electricity Demand
Carbon Emissions
-25
Figure 6.6 Generation by Technology Under Each Scenario
4500
Other
Gas-Combustion
4000
Turbine
3500
Gas-Combined
Cycle
Electricity Production, TWh
3000
Gas-Steam
2500
2000
Coal
1500
1000
Nuclear
500
0
Hydro
Restructure
Efficiency
Hi Efficiency/ Low
Carbon
Compared to the reference case, generating prices and costs remain about the same under the
efficiency scenario. In the HE/LC scenario, prices and costs increase about two-thirds of a cent
because of the additional charge of $50/tonne of carbon. This carbon charge represents a 1.4c/kWh
increase to the more expensive coal plants, but only a 0.5c/kWh increase to the better gas-fired
combined cycle plants. ORCED redispatched and changed capacities so that the cost increase would
be minimized. Note that, although the generation price increases by 21%, the price of generation
September 15, 1997
6.13
Chapter 6
The Electricity Sector
represents only about half of the total price of electricity. Keeping transmission, distribution, and
customer service prices the same, the total price increase would be only 10%.
Because electricity prices are essentially unchanged under the efficiency scenario (changing from
only 6.20c/kWh to 6.22e/kWh including other components), total electricity costs to consumers
decrease by an amount that is proportionate to the reduced electricity demand (i.e., 330 TWh).
Thus, the cost reduction is approximately $20 billion. This value takes into account the savings in
transmission, distribution, and customer service costs included in the full retail price of electricity.
The high-efficiency/low-carbon case yields slightly lower savings of around $18 billion. Total
energy savings almost double to 630 TWh, but the price increase of 0.65c/kWh due to the carbon
charge cuts into the overall savings.
One way to measure the cost impact of the $50/tonne of carbon cost is to evaluate the extra cost due to
plant operation changes (redispatch, retirements, and new construction). In the high-
efficiency/low-carbon case, we construct and operate the plants as optimized with the carbon
charge. If we remove the charge, but construct and dispatch plants as in the high-efficiency/low-
carbon case, we are no longer operating at minimum cost. The excess above the minimum cost is about
$2.2 billion. Dividing this amount by the tons of carbon saved from the minimum-cost case yields an
average cost of $30 per tonne of carbon saved.
Allocating the carbon savings in the efficiency case between the buildings and industrial sectors, we
find that the buildings sector (residential and commercial) saved 19.4 Mt of carbon while the
industrial sector saved 9.6 Mt (Table 6.6). The carbon savings for each were proportional to their
savings in electricity, since electric generation is determined by the system load.
Table 6.6 Carbon Reductions from Electricity Savings by Sector under the Efficiency and High-
Efficiency/Low-Carbon Cases (MtC)
Sector
Efficiency
High-Efficiency/Low-
Case
Carbon Case
Buildings (Residential)
10.9
49.2
Buildings (Commercial)
8.5
38.3
Industrv
9.6
45.0
Total
29.0
132.5
Note: Transport is not included since electricity use in that sector is negligible.
Some of the 133 Mt of carbon that is forecast to be displaced by the high-efficiency/low-carbon case
can be attributed to the end-use efficiency improvements in the buildings and industrial sectors. The
remaining savings are attributed to the change in electricity generation mix that resulted from the
charge of $50/tonne of carbon. Two methods of allocating the carbon savings between the end-use
and supply sectors were examined.
In the first method, ORCED modeled the high-efficiency/low-carbon case without the
$50/tonne charge in the supply sector. The result was a savings of 56 MtC, attributed to energy
efficiency alone. The rest of the carbon savings (77 MtC) is then attributed to the electricity
sector.
In the second method, ORCED modeled the restructure case with the $50/tonne charge in the
supply sector, but without the demand reduction due to efficiency improvements. The result was
a savings of 33 MtC, which is attributed to the electricity sector. The rest of the carbon savings
(100 MtC) is attributed to the end-use sectors based on their energy savings.
6.14
September 15, 1997
The Electricity Sector
Chapter 6
The results of these alternative allocations (including the distribution of carbon savings across the
buildings and industrial sectors) are shown in Table 6.7 and Figure 6.7. The total carbon that is
displaced by the lower electricity demand is the same (133 MtC), but the allocation between the
end-use and supply sectors varies widely depending on the method used to allocate the savings.
Because the savings involve a synergy between increased energy efficiency and changes to supply
dispatching, it is difficult to identify the appropriate allocation of savings for the end-use vs.
electricity supply sectors. To simplify matters, we used the average between the two methods
described above. Specifically, the following averaging was used:
The minimum carbon reduction attributed to electricity end-use efficiencies is 56 MtC (Method 1)
and the maximum is 100 MtC (Method 2). The average of 78 MtC is therefore assumed for the
end-use sector.
The minimum carbon reduction attributed to the electricity sector is 33 MtC (Method 2) and the
maximum is 77 MtC (Method 1). The average of 55 MtC is therefore assumed for the electricity
supply sector.
Table 6.7 summarizes this allocation process.
Table 6.7 Allocation of Carbon Reductions from the Electricity Saved by the High-Efficiency/Low-
Carbon Case (MtC)
Method 1: Carbon
Method 2: Carbon
Final Allocation by
Sector
Reduced by Energy
Reduced by $50/tC
Averaging
Efficiency First
Charge First
Buildings (Residential)
21
37
29
Buildings (Commercial)
16
29
22
Industry
19
34
27
Subtotal
56
100
78
Electricity Sector
77
33
55
Total
133
133
133
Based on the results shown in Figure 6.7, 160 tonnes of carbon are displaced for each GWh of
electricity saved in the high-efficiency/low-carbon case with carbon permit price of $50/tonne,
reflecting the introduction of new low-carbon technologies. This is the marginal carbon-to-energy
ratio that is therefore used for analyzing the impacts of other carbon management strategies in the
electricity sector in Chapter 7. This value is significantly higher than the marginal carbon-to-
energy ratio used in the efficiency case (90 tonnes of carbon per GWh of electricity, as also shown in
Figure 6.3), for the reasons noted earlier.
September 15, 1997
6.15
Chapter 6
The Electricity Sector
Figure 6.7 Carbon Reductions Due to Energy Efficiency and Carbon Management: (A) Without a
Carbon Charge and (B) With a Carbon Charge of $50/Tonne
640
Restructure
Slope = 90 tC/GWh
620
Restructure/
Efficiency
600
Low Carbon
Slope = 95 tC/GWh
580
(A)
High Efficiency
Carbon, Mt
560
540
Slope = 160 tC/GWh
520
500
(B)
Hi Efficiency/
Low Carbon
480
3000
3200
3400
3600
3800
4000
2010 Electricity End-Use Demand, TWh
Note: Slope is equal to marginal carbon savings in tonnes of carbon per gigawatt-hour. In (A), the high-efficiency
option represents all the assumptions of the high-efficiency/low-carbon scenario except that there is no charge for
carbon.
6.5 SUMMARY
The U.S. electricity industry is undergoing massive change. Because the process is far from complete,
it is even more difficult to make estimates about electricity production and use for the year 2010 than
it would otherwise be. However, we developed a reasonable and internally consistent picture of
electricity demand and supply for the year 2010 on the basis of ELA's AEO97 projection and
additional simulations with the Oak Ridge Competitive Electricity Dispatch (ORCED) model.
ORCED was used to simulate the operation of the U.S. electric power supply system in 2010. We
first calibrated our input data so that our results closely matched those produced by ELA for its
Annual Energy Outlook 1997. We then developed a case for 2010 that is intended to reflect the ways
in which a fully competitive industry might operate. Compared to the AEO97, these results suggest
greater electricity use, lower peak demand, and a generation mix that includes more natural gas and
less coal. Thus, although consumption is higher, carbon emissions are lower.
We then simulated the operation of this competitive electricity industry given the efficiency-
induced reductions in electricity use in the residential, commercial, and industrial sectors as
described in Chapters 3 and 4. The efficiency case reduced electricity demand by 9%, which led to a
5% reduction in carbon emissions. The high-efficiency/low-carbon case reduced electricity use by
16%, which led to a 21% reduction in carbon emissions.
6.16
September 15, 1997
The Electricity Sector
Chapter 6
6.6 REFERENCES
Electric Power Research Institute (EPRI). 1993. TAG™ Technical Assessment Guide Volume 1:
Electricity Supply-1993, Palo Alto, California: TR-102276-V1R7.
Energy Information Administration (ELA). 1996a. Annual Energy Outlook 1997 with Projections to
2015, Washington, DC, DOE/EIA-0383(97).
Energy Information Administration (ELA). 1996b. Emissions of Greenhouse Gases in the United
States 1995, Washington, DC, DOE/EIA-0573(95).
Energy Information Administration (ELA). 1996c. International Energy Outlook. U.S. Department of
Energy.
Energy Information Administration (ELA). 1995. Inventory of Power Plants in the United States
1994, Washington, DC, DOE/EIA-0095(94).
Hadley, S. W. 1996. ORFIN: An Electric Utility Financial and Production Simulator, Oak Ridge,
Tennessee: Oak Ridge National Laboratory, ORNL/CON-430.
North American Electric Reliability Council (NERC). 1996. Generating Availability Report, 1991 -
1995, Princeton, NJ, July.
RDI (Resource Data International). 1996. Powerdat Database, Boulder, Colorado: Resource Data
International, Inc.
ENDNOTES
1 States will continue to oversee power-plant siting and environmental emissions.
2
The value of unserved energy is the price that customers would be willing to pay for electricity that
is unavailable as a result of demand exceeding supply.
3
The amount of computer time required for a full simulation depends strongly on the number of
generators treated probabilistically. We found a reasonable tradeoff between computing time and
accuracy when about 10 plants are modeled probabilistically and the other 16 are derated.
"Inclusion or exclusion of the data for cogenerators (which account for about 4% of electricity in the
year 2010) is another source of confusion.
5
According to AEO97, total generation in 1995 was 3246 billion kWh and sales totaled 3008 billion
kWh, implying a loss of 7.3%.
6 The 63% multiplier for T&D represents the percentage of T&D costs attributable to capital as
opposed to O&M.
7 This 88% multiplier is derived from ORCED's estimates of fuel and O&M costs of 1.35c/kWh for
fuel and 0.18 e/kWh for O&M. The 88% then is [=1.35/(1.35+0.18)].
September 15, 1997
6.17
Electricity Supply Technologies
Chapter 7
Chapter 7
ELECTRICITY SUPPLY TECHNOLOGIES
7.1 INTRODUCTION
The electricity industry has many supply-side options at its disposal to reduce or offset carbon
dioxide emissions from electricity production by the year 2010. One of these options, reconfiguring
the generation mix to reflect a $50/tonne charge for carbon, was discussed in Chapter 6. We labeled
this option "carbon-ordered dispatching" because it involves the same technologies that were
considered in the AEO97 reference case. Electricity was redispatched from the existing generation
mix, and the construction and retirement of power plants also changed, but no new technologies were
introduced. Chapter 7 considers other electricity supply technology options, including:
Repowering coal-based power plants with natural gas;
Implementing renewable electricity technologies;
Improving efficiency in generation and transmission and distribution (T&D) systems;
Extending the life of existing nuclear plants; and
Constructing new power plants using advanced coal technologies.
Each of these options is assessed independently. Because interactions among the options are not
taken into account, there is a likely possibility of double-counting with respect to the actual
emissions reduction potential.
The viability and costs of these supply options in 2010 are based on the assumption that the
electricity grid is transformed by the carbon-ordered dispatching that occurs under the "high-
efficiency/low-carbon" (HE/LC) scenario, as described in Chapter 6. Thus, since we assume that
considerable decarbonization has already taken place, this chapter addresses the question: What
additional supply technology options now make sense in a scenario in which carbon has acquired a
value of $50/tonne? We conclude by discussing the significant contribution that renewable energy
technologies can make by the year 2020.
7.2 REPOWERING COAL-BASED POWER PLANTS WITH NATURAL GAS
The conversion of existing coal-fired power plants to operate on natural gas (via repowering) is one
option to significantly increase the efficiency of power generation and reduce carbon emissions in the
U.S. electric power sector.¹ Natural gas is a less carbon-intensive fuel and its use also reduces
emissions of the following criteria air pollutants: sulfur dioxide (SO₂), nitrogen oxide (NOx), total
suspended particulates (TSP), and hazardous air pollutants (HAPs). Our analysis shows that
natural gas combined cycle (NGCC) is a cost-effective power generation technology and carbon
emission reduction option. Depending upon assumptions regarding the differential in the delivered
price between natural gas and coal, the price of carbon permits, and environmental externality
values for criteria air pollutants, we found that carbon emissions of up to 238 MtC could be reduced
annually through repowering.
September 15, 1997
7.1
Chapter 7
Electricity Supply Technologies
7.2.1 Repowering Approachs
The simplest repowering approach is site repowering, where the existing power plant site is reused
with an entirely new NGCC system. Cost and performance data for the General Electric "H" frame
turbine was used; this class of turbine will be the most efficient in the post-2000 period, with the
lowest cost per kilowatt of capacity. While site repowering provides the highest cycle efficiency
(since none of the existing boiler island equipment is reused), it also requires a greater capital
investment (see Appendix G-1).
The more conventional approach is referred to as steam turbine repowering. In this case, a new gas
turbine and heat recovery steam generator (HRSG) are integrated with the existing steam turbine
and auxiliary equipment from the coal plant. Due to age of equipment and the fact that the steam
turbine was designed for linkage with a coal-fired boiler, the efficiency of a repowered steam
turbine plant would be lower than at a site repowered plant. The steam turbine repowering option
has a higher operating cost (due to the lower efficiency) but a lower capital cost (see Appendix G-1).
The cost-effectiveness ($/tC) of both repowering options was examined for all coal-fired power
plants greater than 50 megawatts (MW).² Included in the cost calculation were the cost of
repowering, hook-up, and transmission. We analyzed the site repowering results for the two
alternative gas/coal price differentials: $0.72 and $1.18 per million Btu (MBtu)³, three price ranges
for carbon permits (<$50/tonne, $50-100/tonne, and $101-150/tonne), and three environmental
externality values for SO₂ and NOx (none, low, and high). In addition, a sensitivity analysis was
performed to examine the impact on cost-effectiveness if additional natural gas pipeline
infrastructure (hook-up and transmission) were not needed to ensure gas deliverability to repowered
plants. This sensitivity analysis (referred to as the "no additional transmission cost" case) was
conducted only for those power plants that are currently connected to the natural gas pipeline
network (i.e., dual-fuel). Appendix G-3 contains a complete description of the methodological steps
and key data parameters.
The analytical approach was static in that the cost of repowering was computed for each candidate
power plant but the analysis did not optimize unit/plant production cost, dispatch, or system load.
Moreover, for the steam repowering case, the largest steam turbine (not each individual steam
turbine) at the plant was repowered to generate the equivalent of 1995 plant output (kilowatt-hours,
kWh), since this is both more economic and consistent with industry practice than repowering each
turbine. Lastly, the gas delivery infrastructure costs (hook-up and transmission) were derived
assuming (1) no excess capacity in the current delivery system, and (2) that if such a fuel-switching
strategy were implemented, the natural gas pipeline industry would build capacity (even if done
incrementally) to meet the total estimated gas requirements of repowering all candidate plants and
allocate appropriate delivery costs to each repowered plant. Assumptions regarding gas
deliverability are described below in Section 7.2.2.2.
7.2.2 Repowering Issues
In 1995, there was 335 GW of coal-fired capacity at 404 power plants in the United States. Figure 7.1
indicates that this capacity was comprised of:
319 dual-fuel units (units that can burn both coal and natural gas),
122 multi-fuel units (coal-fired units at sites with natural gas or petroleum units), and
711 coal-fired units (units at coal-only plant sites).
7.2
September 15, 1997
Electricity Supply Technologies
Chapter 7
Figure 7.1 Candidate Coal-Fired Power Plants for NGCC
Repowering
Based on unit number
Based on capacity
Dual-Fuel
Dual-Fuel
319 Units
75,593 MW
28%
23%
Coal Only
Coal Only
711 Units
229,777 MW
Multi-Fuel
62%
69%
25,326 MW
Multi-Fuel
8%
122 Units
10%
These categories reflect differences in the investment cost of conversion and deliverability of natural
gas (i.e., those plant sites consuming gas in 1995 would have a natural gas pipeline connection,
thereby resulting in a lower hookup cost.)
7.2.2.1 Increase in Natural Gas Demand
Utility gas consumption in 1995 was 3.5 trillion cubic feet (TCF). Figures 7.2 and 7.3 show the
increases in natural gas demand from this base that would result from either site or steam turbine
repowering for each of three cost-effectiveness values: less than $50/tC, $50-100/tC, and greater
than $150/tC. The increase in gas demand ranges from 1.0 TCF (<$50/tC) to 4.9 TCF ($50-100/tC) in
the low gas/coal price differential case without externalities. This quantity of gas for repowered
plants represents 29% and 140% increases in 1995 utility gas consumption, respectively.
If all the candidate coal-fired power plants were repowered with NGCC, natural gas demand in the
utility sector would increase by 9.0 TCF/yr (site repowering) or 9.4 TCF/yr (steam turbine
repowering) to either 12.5 TCF/yr or 12.9 TCF/yr, respectively, an increase of over 250% compared to
current consumption levels.
The potential gas price increase resulting from NGCC repowered plants was not analyzed in this
study. Only the current and projected gas/coal price differentials expected under AEO97 were
included in the cost analysis. However, ELA has prepared a preliminary estimate; they found that
an 11 TCF increase in demand would increase natural gas prices by $3.09/MBtu over 20 years (1995-
2015), if coal-fired power plants were converted to natural gas when scheduled for life
extension/refurbishment and there was considerable demand-side energy-efficiency investment.
September 15, 1997
7.3
Chapter 7
Electricity Supply Technologies
Figure 7.2 Incremental Increase in Gas Consumption Resulting from Coal to Gas Conversion with
Constant 1995 Gas/Coal Price Differential ($0.72/MBtu)
60
No Environmental Externality Credits
Low Environmental Externality Credits
5.3
High Environmental Externality Credits
50
49
42
4.2
42
4.0
Increase in Gee Consumption TCF
3.0
2.3
2.0
1.3
10
1.0
05
00
< 50
50-100
101-150
Carbon Cost Ranges, S/tC
Figure 7.3 Incremental Increase in Gas Consumption Resulting from Coal to Gas Conversion with
Constant 2010 Gas/Coal Price Differential ($1.18/MBtu)
60
No Environmental Externality Credits
Low Environmental Externality Credits
High Environmental Externality Credits
50
4.9
47
4.5
40
3.8
Increase In Gas Consumption, TCF
30
2.9
3
2.0
12
10
0.8
0
00
< 50
50-100
101-150
Carbon Cost Ranges, S/tC
7.4
September 15, 1997
Electricity Supply Technologies
Chapter 7
7.2.2.2 Gas Deliverability
The spatial distribution of the initial 404 candidate plants is depicted in Figure 7.4. Some of the
candidate plants were not considered for repowering since they were (1) not considered economic by
ELA, or (2) determined to be unnecessary due to reductions in demand arising from end-use efficiency
improvements.² Most of the plants are located in the Mid-Atlantic, South Atlantic, Midwest, and
Plains regions. While these are also primary gas-consuming regions served by major trunk lines,
many industry experts believe there is limited unused or underutilized capacity in the current 1.2
million mile pipeline system (transmission - 264,900 miles; distribution - 935,000 miles; field -
62,200 miles). Since this capacity is necessary to accommodate peak winter demand and non-utility
growth, it is of little value to power plants considering conversion, since these power plants require
firm pipeline commitments.
Figure 7.4 Location of Candidate Plants for Coal/Gas Repowering in the U.S.
Due to the potentially significant increase in utility gas demand that could result from repowering
(either site or steam turbine) coal-fired power plants, and the uncertainties regarding when
repowering would take place, new pipeline capacity sufficient to serve all candidate plants was
developed to ensure deliverability. A detailed assessment was performed (using a geographical
information system, GIS) to compute the distance of each candidate power plant to its nearest trunk
line. Cost estimates were derived for the cost of upgrading the lines to meet the increased gas
demands (see Appendix G-4). Table 7.1 summarizes the distance of the candidate plants to their
closest production zone.
September 15, 1997
7.5
Chapter 7
Electricity Supply Technologies
The requirement to add new pipeline capacity could affect the attractiveness of repowering as a
carbon mitigation strategy. During 1994 and 1995, 1,200 to 1,500 miles of new pipeline were added to
the system. According to Federal Energy Regulatory Commission (FERC) filings of pipeline projects,
there are a considerable number of new pipelines and pipeline expansions that have been proposed,
some of which are still pending approval. While mileage is not included with each filing, in the
regions of concern (Central, Midwest, Northeast, and Southeast), more than 8,200 miles of pipe is
projected to be added; this level of expansion is greater than the 1994-95 rate of addition. However,
it is not known how long it will take to complete these proposed pipelines. Consequently, an accurate
assessment of the ability to increase the rate of pipeline expansion/construction could not be
estimated as a part of this study.
Table 7.1 Plant Distance from Production Zone
Range
Dual-Fuel
Multi-Fuel
Coal Only
Total
(Miles)
# Units
%
# Units
%
# Units
%
# Units
%
60 440
48
37
5
12
55
22
108
26
440 - 620
33
25
8
19
64
26
105
25
620 890
30
23
15
35
59
24
104
25
890-1,480
19
15
15
35
67
27
101
24
Total
130
100
43
100
245
100
418
100
7.2.3
Emissions Reductions
Based on our analysis, repowering of coal-fired power plants with NGCC is a cost-effective carbon
reduction strategy. Tables 7.2-7.4 summarize the site repowering results for the two alternative
gas/coal price differentials: $0.72 and $1.18 per million Btu (MBtu), three price ranges for carbon
permits (<$50/tonne, $50-100/tonne, and $101-150/tonne), and three environmental externality
values for SO₂ and NOx (none, low, and high). The price differential of $0.72/MBtu represents the
1995 gas/coal price differential held constant, while $1.18/MBtu is EIA's forecasted price
differential for the year 2010.⁵ In addition to the "no externalities" case, two alternative market
values were used for SO₂ and NO.: low externalities represent $0 per ton of SO₂ and $700 per ton of
NOx; high externalities represent $100 per ton of SO₂ and $1400 per ton of NOₓ.
As can be seen in Table 7.2, given a carbon permit price of less than $50/tC and a gas/coal price
differential of $0.72/MBtu, 30 to 119 MtC could be removed via NGCC site repowering, depending
upon externality assumptions. When the price differential increases to $1.18/MBtu, 0 to 83 MtC
could be removed from utility emissions. Consequently, we see that a increase of $0.46/MBtu in the
price differential decreases carbon reductions from NGCC repowering by approximately 30 MtC.
Although the disaggregated data are not presented, most of the carbon reduction in the <$50/tC
range actually occurs in the $25-50/tC range.
An ancillary benefit of switching from coal to gas and improving conversion efficiency is reduction in
SO₂ and NOx, two criteria pollutants. At the <$50/tC level, approximately 50% of the SO₂ and
NO, would be removed (depending on the externality value); at $50-100/tC and higher almost all
the remaining coal-fired SO₂ and NO, emissions would be eliminated. If all the candidate plants
were repowered, almost all of the SO₂ and most of the NOx would be removed.
The economic value of the SO₂ and NO, emissions reductions that would result from repowering of
the plants was also assessed in this study. Using the methodology described in Appendix G-2, SO₂
7.6
September 15, 1997
Electricity Supply Technologies
Chapter 7
was valued from $0-100/ton; NO, was valued at from $700-1400/ton. These values were used as the
basis for the environmental externality credits to offset the investment cost of repowering.
Table 7.2 Summary Statistics: Coal to Gas Repowering with a Carbon Permit Price of <$50/tonne
Constant 1995 Gas/Coal Price Differential ($0.72/MBtu)
Incremental
Incremental
Incremental
Externality
Carbon
SO₂ Removed
NOₓ Removed
Gas Consumed
Cases*
Removed (MtC)
(Mt)
(Mt)
Affected GW
(TCF)**
None
30.3
0.5
0.7
26.8
1.0
Low
66.0
1.2
1.4
63.3
2.3
High
118.6
4.0
2.6
122.6
4.2
Gas/Coal Price Differential in 2010 ($1.18/MBtu)
Incremental
Incremental
Incremental
Externality
Carbon
SO₂ Removed
NOx Removed
Gas Consumed
Cases*
Removed (MtC)
(Mt)
(Mt)
Affected GW
(TCF)**
None
0
0
0
0
0
Low
23.6
0.3
0.6
20.2
0.8
High
83.4
2.5
1.9
83.3
2.9
Two alternative market values were used for SO₂ and Nox: low externalities represent $0 per ton of SO₂
and $700 per ton of NOx; high externalities represent $100 per ton of SO₂ and $1400 per ton of NOx.
**TCF = trillion cubic feet
Table 7.3 Summary Statistics: Coal to Gas Repowering with a Carbon Permit Price of $50-100/tonne
Constant 1995 Gas/Coal Price Differential ($0.72/MBtu)
Incremental
Incremental
Incremental
Externality
Carbon
SO₂ Removed
NOₓ Removed
Gas Consumed
Cases*
Removed (MtC)
(Mt)
(Mt)
Affected GW
(TCF)**
None
134.6
4.9
2.7
147.3
4.9
Low
140.4
6.7
2.8
165.6
5.3
High
106.7
5.0
1.8
130.8
4.2
Gas/Coal Price Differential in 2010 ($1.18/MBtu)
Incremental
Incremental
Incremental
Externality
Carbon
SO₂ Removed
NOx Removed
Gas Consumed
Cases*
Removed (MtC)
(Mt)
(Mt)
Affected GW
(TCF)**
None
109.6
2.5
2.2
108.9
3.8
Low
134.1
5.0
-
2.7
146.4
4.9
High
123.9
5.5
2.3
147.6
4.7
. Two alternative market values were used for SO₂ and Nox: low externalities represent $0 per ton of SO₂
and $700 per ton of NO,; high externalities represent $100 per ton of SO₂ and $1400 per ton of NOx.
**TCF = trillion cubic feet
September 15, 1997
7.7
Chapter 7
Electricity Supply Technologies
Table 7.4 Summary Statistics: Coal to Gas Repowering with a Carbon Permit Price of
$101-150/tonne
Constant 1995 Gas/Coal Price Differential ($0.72/MBtu)
Incremental
Incremental
Incremental
Externality
Carbon
SO₂ Removed
NOₓ Removed
Gas Consumed
Cases*
Removed (MtC)
(Mt)
(Mt)
Affected GW
(TCF)*
None
69.4
4.0
1.2
93.9
2.8
Low
31.0
1.6
0.5
43.8
1.3
High
13.1
0.6
0.2
20.9
0.5
Gas/Coal Price Differential in 2010 ($1.18/MBtu)
Incremental
Incremental
Incremental
Externality
Carbon
SO₂ Removed
NOx Removed
Gas Consumed
Cases*
Removed (MtC)
(Mt)
(Mt)
Affected GW
(TCF)**
None
117.2
6.6
2.3
148.4
4.5
Low
75.1
4.0
1.2
98.6
3.0
High
29.8
1.5
0.4
41.4
1.2
. Two alternative market values were used for SO₂ and Nox: low externalities represent $0 per ton of SO₂ and
$700 per ton of NOx; high externalities represent $100 per ton of SO₂ and $1400 per ton of NOₓ.
**TCF = trillion cubic feet
7.2.4
Cost-Effectiveness
Figure 7.5 portrays the cost-effectiveness of site repowering with NGCC and the corresponding
cumulative carbon removed for the two alternative gas/coal price differentials when no
environmental externalities are considered. With a price differential of $0.72/MBtu,
approximately 30 MtC can be removed for <$50/tC, an additional 135 MtC can be removed for $51-
100/tC, and an additional 77 MtC can be removed for >$100/tC. When the price differential
increases to $1.18/MBtu, 0 MtC of carbon are removed at <$50/tC, 110 MtC are removed at $51-
100/tC, and an additional 132 MtC are removed at >$100/tC.
Figures 7.6 and 7.7 depict the effect of environmental externality credits for SO₂ and NOₓ on carbon
cost-effectiveness. As mentioned above, in addition to the "no externalities" case, two alternative
market values were used for SO₂ and NOx: low externalities represent $0 per ton of SO₂ and $700 per
ton of NOx; high externalities represent $100 per ton of SO₂ and $1400 per ton of NOx. The rationale
for these values is explained in Appendix G-3. Both Figures 7.6 and 7.7 (together with Tables 7.2 -
7.4) illustrate that the effect of the environmental externality credit is to shift the carbon cost curve
downward and to the right causing more capacity (GW) and carbon removal (MtC) to occur at lower
carbon permit price levels.
Because dual-fuel plants are already receiving natural gas (although at lower volumes than a
repowered plant), a sensitivity analysis was conducted wherein no hook-up or transmission costs
were incurred to deliver an increased quantity of natural gas to these repowered plant sites. This "no
additional transmission cost case" is illustrated in Figures 7.8 and 7.9, which depict the two
alternative gas/coal price differentials and include externality credits for site and steam turbine
repowering. Since transportation costs comprise approximately 30% of the total investment cost, the
carbon cost curves shift downward considerably when these costs are removed. In Figure 7.8,
approximately 45 GW of coal-fired capacity can be repowered at <$50/tC, removing 42 MtC of
carbon, 1.2 Mt of SO₂ and 0.9 Mt of NOx. The amount of natural gas required by these repowered
plants is 1.5 thousand cubic feet; approximately 50% of 1995 utility consumption.
7.8
September 15, 1997
Electricity Supply Technologies
Chapter 7
The cost-effectiveness numbers derived in this study are optimistic. These numbers should be used
with caution because they do not (or do not adequately) consider the following factors that will
determine the ultimate cost-effectiveness of the coal-to-gas repowering:
Potential increase in gas prices from NGCC repowering,
Actual cost of repowering the candidate coal-fired power plants,
Excess transmission capacity, and/or economies of scale in delivering the required gas,
Capacity utilization of the converted plants,
Costs associated with breaking long-term coal contracts, and
Other socioeconomic factors (e.g., differential state/federal tax effects, displaced coal miners).
In addition, the effectiveness of repowering as a carbon control strategy will depend on whether and
to what extent the converted plants are dispatched. If, because of the costs associated with
conversion, the repowered plants are not dispatched or their utilization is minimized, the
associated carbon reductions will depend on the fuels and technologies used at the plants dispatched
ahead of the repowered plants.
Figure 7.5 Carbon Curve for Coal/Gas Site Repowering: No Environmental Externality Credits
250
200
150
Incremental Cost ($/tC)
100
50
0
0
50
100
150
200
250
300
Total Carbon Removed (MtC/yr)
$1.18 per MMBtu
$0.72 per MMBtu
September 15, 1997
7.9
Chapter 7
Electricity Supply Technologies
Figure 7.6 Carbon Curve for Coal to Gas Site Repowering: Effect of Environmental Externality
Credits on Cost of Carbon Removal with Constant 1995 Gas/Coal Price Differential ($0.72/MBtu)
250
200
150
Incremental Cost ($/IC)
100
50
0
0
R
100
150
200
250
300
Total Carbon Removed (MtC/yr)
High Externalities
Low Externalities I
Figure 7.7 Carbon Curve for Coal to Gas Site Repowering: Effect of Environmental Externality
Credits on Cost of Carbon Removal with Gas/Coal Price Differential in 2010 ($1.18 MBtu)
250
200
150
Incremental Cost ($/IC)
100
50
0
0
50
100
150
200
20
ID
Total Carbon Removed (MtC/yr)
High Externalities
ow Externalities
7.10
September 15, 1997
Electricity Supply Technologies
Chapter 7
Figure 7.8 Carbon Curve for Partial Repowering: Constant 1995 Gas/Coal Price Differential
($0.72 MBtu) Low Environmental Externality Credits
150
140
Gas Differential: $0.72/MBtu
SO2 Credit: $0/ton SO2
130
NOx Credit: $700/ton NOx
120
110
100
90
Carbon Cost ($/tC)
80
70
60
50
Site Repowering
40
30
20
10
Steam Turbine
0
0
10
20
30
40
50
60
70
Total Carbon Removed (MtC)
Figure 7.9 Carbon Curve for Partial Repowering?: Constant 2010 Gas/CoalPrice Differential
($1.18 MBtu) High Environmental Externality Credits
150
Gas Differential: $1.18/MBtu
140
SO2 Credit: $100/ton SO2
NOx Credit: $1,400/ton NOx
130
120
110
100
90
Carbon Cost ($/tC)
80
70
60
50
40
Site Repowering
30
20
Steam Turbine
10
0
0
10
20
30
40
50
60
70
Total Carbon Removed (MtC)
September 15, 1997
7.11
Chapter 7
Electricity Supply Technologies
7.3 RENEWABLE ELECTRICITY TECHNOLOGIES
Over the long term, renewable energy technologies are likely to play a crucial role in limiting carbon
emissions and global warming. While aggressive energy efficiency and fuel switching can reduce
domestic carbon emissions to approximately 1990 levels by 2010, controlling or reducing carbon
emissions beyond that date will require greater energy contributions from low-carbon technologies
such as renewables. In other words, renewables will play an essential role in helping the United
States to cut carbon emissions in the years beyond 2010.
Renewables will also make important contributions to both domestic and international carbon
emission controls by 2010. Renewable technology contributions to domestic electricity and carbon
savings in 2010 under the HE/LC scenario are summarized in Table 7.5.
Table 7.5 Additions to Generating Capacity Electricity and Carbon Emission Reductions from
Renewables for the HE/LC Case in 2010
Renewable Technology
Capacity Additions
Electricity
Carbon Emission
(GW)
(TWh)
Reduction (MtC)ᵃ
Included in Scenario:
Biomass Cofiring
8-12
58-88
16-24
Wind
8-23
28-81
6-20
Hydropower
10-16
23-35
3-5
Subtotal
25-49
Excluded from Scenario:
b
Landfill Gas
3-7
20-50
25-53
PV
3-5
6-10
1-2
Geothermal
6-14
47-110
6-16
Solar Thermal
0-2
0-6
0-1
Subtotal
32-72
Total
38-79
182-380
57-121
a
business-as-usual forecast for 2010.
These carbon emissions reductions represent the difference between the high-efficiency/low-carbon case and the
b
The carbon emission reduction in this case represents the equivalent derived from the prevention of the methane
release coupled with its radiation-trapping properties.
This section examines the potential for renewable electricity technologies to reduce U.S. carbon
emissions. The contributions of renewables in various end-use sectors, such as transportation, are
discussed in other chapters in this report.
Renewables are in the midst of a major, long-term transition, from being "advanced technologies"
with only a peripheral market role to becoming mainstream "technologies of choice" in the energy
marketplace early in the next century. One clear marker of this transition is the changing cost of
electricity from renewable power technologies. Figure 7.10 displays these costs for the period from
7.12
September 15, 1997
Electricity Supply Technologies
Chapter 7
1980 to 2005, based on both historical data and recent projections (Office of Utility Technologies,
1997).
Figure 7.10 Historical and Projected Costs of Electricity from Four Renewable Power Technologies
Photovoltaics
Wind
100
40
80
Cost of electricity
30
(e/kWh)
60
40
Cost of electricity
(e/kWh)
20
20
10
0
0
1980
1985
1990
1995
2000
2005
1980
1985
1990
1995
2000
2005
Solar Thermal
Geothermal
40
10
8
30
Cost of electricity
(e/kWh)
(e/kWh)
6
20
Cost of electricity
4
10
2
0
0
1980
1985
1990
1995
2000
2005
1980
1985
1990
1995
2000
2005
The pace and timing of this transition is difficult to project, however, because it is strongly
dependent on such variables as the progress made through research and development, the evolution
of energy economy policies, and the magnitude and impact of consumer interest in "green" energy. For
example, under the $50/MtC cost-of-carbon scenario assumed in this study, the adoption of wind
power in the United States is likely to increase rapidly on an economic basis. In addition, increasing
attention is being focused on consumer interest in green energy. As the electric utility sector moves
toward competitive markets, consumers probably will have the option of purchasing power that is
environmentally cleaner.
The rate of change will impact the role of renewables in 2010 at least as much as the specific energy
contribution of renewables in that year. Therefore, this section discusses the trends as well as the
predicted contributions of renewables to the energy supply and to carbon emission reductions in 2010.
A thorough analysis of the role of renewables in 2010, which captures the complexity of their
transition, has not been conducted as a part of this study. Instead, this section presents analyses of a
few renewable technologies whose role is likely to be quite significant by 2010, and includes a
general discussion of the other renewable technologies. A more thorough analysis of the
relationship between renewables and reductions in carbon emissions over a longer time frame is the
subject of a future study.
September 15, 1997
7.13
Chapter 7
Electricity Supply Technologies
Thus, this section discusses the role of renewables in two time frames: (1) developments and
contributions by 2010, and (2) the long-term outlook.
7.3.1 Renewable Electricity in 2010
As stated earlier, renewable electric technologies will make important contributions to carbon
emission reductions in 2010 in the context of a policy that imposes a $50/tonne cost on carbon
emissions. Estimates of those contributions are presented here. While the scope of this study did not
include a thorough and systematic analysis of this issue, the estimates given are based on a number
of directly relevant studies. These include, in particular, carefully developed performance and cost
projections for renewable electric technologies (Office of Utility Technologies, 1997) and projections
of future market penetrations of these technologies (Office of Energy Efficiency and Renewable
Energy, 1997).
The potential of biomass cofiring was assessed because that technology provides an opportunity for
reasonably straightforward displacement of a significant amount of coal. This assessment, which
draws upon another recent analysis, is presented first.
An analysis of the impact of a $50/tonne cost of carbon on wind power was also conducted, and those
results are discussed second. Wind was selected because cost projections for wind power indicate that
this technology will be competitive with other electricity generation sources according to the
electricity costs modeled in the HE/LC scenario of this study. In addition, wind power is already
successfully penetrating electricity markets in the United States and abroad.
This analysis is followed by estimates of carbon emission reductions that would be likely to result
from hydropower upgrades and landfill gas capture and use. These estimates are derived from DOE
and EPA studies relevant to market projections for those two technologies, respectively (Rinehart et
al., 1997; EPA, 1993).
Finally, other key renewable power technologies are discussed briefly. We present estimates of the
likely contribution of these technologies in 2010, developed through comparisons and extrapolations
from earlier projections (Office of Energy Efficiency and Renewable Energy, 1997).
7.3.1.1 Cofiring Coal with Biomass
Cofiring biomass with coal has the technical and economic potential to replace at least 8 GW of the
nation's coal-based generating capacity by 2010, and as much as 26 GW by 2020. Though the current
substitution rate is negligible, a rapid expansion is possible with the use of wood residues (urban
wood, pallets, and secondary manufacturing products) and dedicated feedstock supply systems
(DFSS) such as willow, poplar, and switchgrass.
The current coal-fired power-generating system represents a direct opportunity for carbon mitigation
by substituting biomass-based renewable carbon for fossil carbon. Extensive demonstrations and trials
have shown that biomass can replace up to about 15% of the total energy input with little more than
burner and feed-intake system modifications to existing stations (CONEG, 1996). Since large-scale
power boilers in today's 310-GW-capacity fleet range from 100 MW to 1.3 GW, the biomass potential
in a single boiler ranges from 15-150 MW.
Preparation of biomass to an appropriate size of less than one-quarter inch, with a moisture content
of less than 25%, can be achieved using existing commercial technologies. "Tuning" the combustion
output of the boilers causes little loss in total efficiency, implying that the biomass-to-electricity
7.14
September 15, 1997
Electricity Supply Technologies
Chapter 7
combustion efficiency is close to the 33-37% range of an unmodified coal plant, an efficiency that
stand-alone biomass generating capacity has yet to demonstrate.
Economics
The cost of implementing biomass cofiring varies from site to site. It is influenced by the space
available for yarding and storing the biomass, the installation of size-reduction and drying
facilities, and the nature of the boiler burner modifications required. The cost is expected to be in the
range of $100-$700/kW of biomass capacity. Early trials indicate that a median value of about
$180/kW is likely. A 100-MW coal plant with 10% biomass substitution would then require an
investment of $1.8 million. There is an O&M cost increase of $70,000/year over coal, as a result of
the need for an additional yard worker to handle the biomass. Assuming a GENCO recovers its
investment cost in three years, the annual fuel offset then has to be $670,000 to cover capital
recovery ($1.8 million) and increased O&M costs ($210,000 for three years). If the average price of
coal is about $1.40/MBtu (million Btu), the annual fuel cost of coal is $1.081 million (10 MW of
biomass capacity at 85% capacity factor and 32.9% thermal efficiency, 10,337 Btu/kWh). The
allowable cost of biomass then is $411,000, or about $9/tonne. Above this cost, the biomass would
have to be subsidized to encourage a GENCO to use biomass cofiring.
Fuel Costs
Near-term potential biomass feedstocks are those residues available within a radius of about 50
miles around a plant. Data from existing biomass power plants in the Northeast and California
indicate that there are extensive sources of biomass residues available for about $0.50/MBtu (less
than $9/tonne). Transportation costs limit the range over which such biomass feedstocks can be
acquired and, in the long term, there are likely to be dedicated feedstock systems much closer to the
power plants. By definition, residues (e.g., urban wood residues, rights-of-way clearance,
construction and demolition wood, pallets, and sawdust shavings from secondary wood processing)
are finite and will respond to the prices offered for them.
Dedicated feedstocks would not be bound by this constraint. However, such feedstocks are much more
expensive than residues. With current technology the price is about $2/MBtu, although the current
development goal is in the range of $1-$1.50/MBtu. It is assumed that an estimated 10.4 million
acres will be needed to reach a nominal production of 86 Mt by 2020. Because DFSS is in an early
stage of development, the model assumes that the initial planting will yield only about 6
tonnes/acre by 2002 (today's state-of-the-art), and that by 2010 the yield will be closer to 8
tonnes/acre. Today's costs are high; $45/tonne is feasible, but a combination of learning-curve
improvements and economies of scale should bring the cost down to about $32/tonne by 2010. The
competing coal price is assumed to be $1.40/MBtu ($1.33/GJ) throughout.
Biomass Substitution Potential
The cofiring estimates in this section were derived from a 30 GW scenario for all biomass
technologies, developed by NREL for the current Biomass Power Program Strategic Plan. This
scenario is for a mix of steam, cofiring, and integrated gasification/combined cycle (IGCC) biomass
generation. However, the resource plan that was developed, which included residues and DFSS, is
independent of the end use and involves the development of 11-12 million acres of land for DFSS by
2020, or just under 3 million acres by 2010. The resource development shown in Figure 7.11 is used as
the basis for this carbon assessment. This indicates that DFSS would come-on rapidly after the year
2001 and assumes that residues would be capable of only a small increase in quantity, since much is
already being utilized. The average cost of residues is expected to increase gradually, while the cost
of DFSS crops is expected to demonstrate a strong learning curve and large economies of scale.
September 15, 1997
7.15
Chapter 7
Electricity Supply Technologies
Figure 7.11 30 GW Strategic Plan Scenario
Note: Biomass consumption estimates by NREL
160
120
80
Biomass (M tonnes)
40
0
0
4
Land impact (M acres)
8
12
1990
1995
2000
2005
2010
2015
2020
Residue
DFSS
M acres
Timing
While a coal-fired station could be modified for cofiring in less than one year (including
environmental permitting), a biomass resource assessment, contractual arrangements, and logistics
for biomass residues could take the better part of 18 months, based on actual project experience.
Although the availability of residues is assumed to be significant and would ultimately supply
about 50 Mt, price and availability are likely to be variable. The price will no doubt increase with
the level of demand; therefore, the biomass feedstock supply is expected to be a blend of DFSS and
residues.
The DFSS component is predicated on making a start on land accumulation (whether purchases,
leases, or cooperatives), with land preparation and planting in 1999. A significant effort will be
required to initiate development of the 11-12 million acres proposed for 2020; today, discussions are
about DFSS demonstrations at the 1000-acre level. Adequate clonal material and management
systems for planting, tending, and harvesting will also need to be developed. The crops of choice in
much of the Northeast and Southeast are probably woody species, which would require extensive
nursery activity to put the needed clonal material in place for planting out. With willow, the first
harvest cycle would be four years after planting and a rotation of three years thereafter. For poplar,
the cycle is likely to be in the range of six to eight years.
Environmental Issues
Because biomass generally contains significantly less sulfur than coal, cofiring with biomass could
reduce SOx emissions. Early results suggest that there is also a NOx reduction potential using woody
biomass. However, most coal-fired power stations have efficient precipitators and some have
sulfur-capture technologies, so the net environmental effect of 10% biomass substitution (on an energy
basis) appears to be negligible. The solid wastes (ash) would be little changed in either composition
or mass (most biomass has considerably less ash than coal). But some stations sell fly ash to Portland
7.16
September 15, 1997
Electricity Supply Technologies
Chapter 7
cement manufacturers, so there may be a need to negotiate the acceptance of mixed biomass and coal
ash in such applications with respect to ASTM standards.
The DFSS environmental impact is dependent on the choice of lands for plantations. Replacing
annual cropland with perennial DFSS appears to result in a net environmental gain. Results for
pasture land are probably negligible and replacement of forest may result in some increased impacts.
The use of residue has the potential to offset landfilling as well as potential methane emissions
from landfilling clean biomass materials. Experiences in California indicate that the issue will be
one of rationalizing the cost distribution between the waste generator, the hauling contractor, and
the generating station receiving the residue rather than it going to a landfill. If such negotiations
were successful, and the generating station could guarantee reception of the residues at all times
(many urban wood residue generators do not have storage facilities), both residue costs and their
availability could improve significantly.
Impact on Carbon Emissions
Given the technical and economic potential described above, it is probably reasonable to assume
additional biomass-cofired capacity of 8-12 GW by 2010, which should reduce carbon emissions by
16-24 MtC.
7.3.1.2 Wind Power
The development of wind power systems has progressed quite rapidly since 1980. There are
approximately 1800 MW of wind capacity operating in the United States today, and another 4300
MW of capacity overseas (Figure 7.12). This capacity growth is almost certain to continue because of
continuing decreases in the cost of wind-generated electricity as well as growing interest in emission-
free power derived from local, renewable resources. Figure 7.13 shows the projected cost of wind-
generated electricity for wind farms located in Class 4 and Class 6 resource sites (as presented in
DOE's 1997 Technology Characterizations). Class 4 sites have average wind speeds of 5.6-6.0 m/s,
Class 6 sites have average wind speeds of 6.4-7.0 m/s, both measured at a height of 10 meters.
Figure 7.13 also displays the median, 10th percentile, and 90th percentile of electricity generation
prices in 2010 based on the HE/LC case described in Chapter 6. As these projections indicate, wind
power prices are projected to drop below the median 2010 price for that scenario before 2005. Thus,
strictly on a price-of-energy basis, in this scenario wind power will be able to compete favorably
with other power sources for several years prior to 2010.
In addition to the price of energy, a number of other factors will affect the extent to which wind
power systems will be adopted between now and 2010. These include, for example, the overall
market for new power systems, the price penalty that wind power will encounter for providing
intermittent power, and the price advantages that wind power will realize because it is a "green"
power source and because it is not subject to the risk of future fuel price increases. Because the level of
influence of each of these factors has not been analyzed, it is difficult to project their combined
impact.
September 15, 1997
7.17
Chapter 7
Electricity Supply Technologies
Figure 7.12 Domestic and International Wind Power Capacity, Grid-Connected
4,500
4,000
3,500
Domestic
3,000
International
Megawatts
2,500
2,000
1,500
1,000
500
0
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
Source: Department of Energy. Office of Utiltry Technologies Wind Division
Figure 7.13 Projections of Wind Power Costs
6
5
4
e/kWh
3
2
COE, Class 4
COE, Class 6
5-lab, median
1
5-lab, 90-%ile
5-lab, 10-%ile
0
1996
2000
Year
2005
2010
7.18
September 15, 1997
Electricity Supply Technologies
Chapter 7
In the AEO97 reference case forecast, ELA projects electricity generation to increase by about 800
TWh by 2010, from 3083 TWh in 1995 to 3874 TWh in 2010 (ELA, 1996). In this context, ELA projects a
total of only 3800 MW of wind generating capacity. The Quality Metrics analysis by the Office of
Energy Efficiency and Renewable Energy (EERE) of the impacts of its R&D program estimates that
additional installed wind power generating capacity will reach 8 GW and contribute
approximately 29 TWh to the electricity market by 2010 if wind program goals are met. This will
result in a carbon emission reduction of 6 MtC. Neither of these projections, however, assumes a major
new market policy to promote wind power.
In the HE/LC case of this study, electricity generation will increase by significantly less than in the
business-as-usual case. Thus, the HE/LC scenario presents a much smaller target market for new
power-generation sources. However, under the transition of the utility sector to a competitive
market, it is very likely that newer technologies with lower generating costs will displace some
existing generation capacity with higher generating costs. Moreover, in the HE/LC scenario,
generation costs are projected to be more than 25% higher than in the base case. Thus, as illustrated
in Figure 7.13, wind power generation costs will be highly competitive in this marketplace, so
displacement of higher-cost existing generation by wind is likely. In an attempt to model the
penetration of wind under these conditions, the ORCED model was run using the wind technology
characteristics developed by the Office of Utility Technologies (Office of Utility Technologies,
1997). As might be expected, the results indicated that the level of wind penetration in this case is
quite sensitive to the actual input parameters. According to the model, for example, using the cost
and performance characteristics projected by DOE for the year 2005 could lead to the adoption of as
much as 50,000 MW of wind power by 2010. Using 160 kg/MWh, the average carbon intensity of the
generation displaced (see Chapter 6), this would result in a carbon emission reduction of 28 MtC.
Because the ORCED model indicated that this new wind capacity would displace coal-fired
generation, a higher conversion ratio (275 kg/MMh) is used to estimate carbon emission reductions,
resulting in an estimated reduction of 48 MtC.
This level of penetration would require wind-turbine manufacturing capacity to expand at a rate of
approximately 25% per year. As Figure 7.12 indicates, this level of wind capacity expansion has
been reached in the past. Europe's experience with wind power also indicates that this technology
can expand quite rapidly. It is possible for the manufacturing industry to respond quickly to market
demands, since most of the components of wind systems (generators and gearboxes) are readily
available, and not specific to wind technology. In 1991, the European Wind Energy Association set a
goal of 4 GW of wind by 2000. This goal has been realized already in 1997, and the new targets are 8
GW by 2000 and 40 GW by 2010. Given that Europe is a much more land-constrained continent with
generally lower wind resources than the United States, this comparison suggests that 50 GW of wind
power capacity can be realized in the United States by 2010 in the context of a strong policy
environment.
The HE/LC context of this analysis assumes a policy environment that acknowledges the need to
address global warming. In such an environment, renewable energy, including wind power, will be
able to demand somewhat higher prices because of consumers' preferences for green power. The value
of that premium is not yet known.
It is well-documented that wind resources in the United States are quite extensive. For example, an
assessment of wind resources and access to transmission indicates that more than 115 GW of Class 5
and Class 6 sites are within 5 miles of existing lines, and more than 1,000 GW Class 4 sites are
within that same range (Parsons et al., 1995). This assessment excludes sites that are not suitable for
wind farm development, such as cities and wilderness areas. Thus, 50 GW could probably be
developed primarily in Class 5 and Class 6 areas, which means that they will operate with
September 15, 1997
7.19
Chapter 7
Electricity Supply Technologies
relatively high capacity factors and low costs of energy. (The Draft Technology Characterizations
indicate that capacity factors will be 45% in Class 6 regions and 35% in Class 4 regions by 2005.)
This analysis does not take into account the fact that wind-generated electricity will probably face
at least a partial price discounting because wind power is not fully predictable. At this time, the
level of this discounting is simply not known. To date, with low levels of penetration into grid-
connected generation, intermittency has not been an issue. There are some indications that the range
of electricity prices in a competitive market will be fairly narrow. For example, prices for
electricity transactions on the Continental Power Exchange during peak hours generally vary only by
about 2 cents/kWh (Continental Power Exchange, 1997). This implies that price variations between
different generation sources cannot vary by more than that, and it is likely that the difference will
be much smaller under full competition.
In summary, analyses indicate that total wind power capacity in 2010 could range from as low as 5
GW, based on a simple extrapolation of today's energy economy, to as high as 50 GW in an
environment that includes competitive pricing and policies emphasizing control of carbon emissions.
Given these results, it is probably reasonable to estimate that additional wind capacity will be 8-23
GW in 2010. This translates into electricity contributions of 28-81 TWh, resulting in reductions of
carbon emissions of 6-20 MtC relative to the BAU forecast for 2010.
7.3.1.3 Increasing Generation and Capacity at Existing Hydropower Plants
Hydroelectric power currently supplies about 10% (78 GW) of the nation's electricity and constitutes
84% of the nation's generation from renewables (ELA, 1996). Hydroelectric power plants produce no
greenhouse gas emissions during operation (DOE, 1994). In the 1940s, 40% of the country's electricity
came from hydropower plants (Williams and Bateman, 1995). The adverse environmental affects of
some hydropower projects are now relatively well known (e.g., Mattice, 1991), but significant
progress is also being made in mitigating these problems (Sale et al., 1991).
Hydroelectric power uses the energy of falling water to generate electricity. Hydroelectric
generation technologies for utility-scale applications are generally considered to be mature, with
turbine efficiencies typically in the 75%-85% range (OTA, 1995). There are three types of
hydropower facility:
1. Most hydropower plants use dams to raise the water level, which increases the water's
potential energy, and allows for regulation of the water availability. Conventional
hydropower (with reservoir storage) can provide baseload, intermediate, or peaking
power, depending on the availability of water and project design (OTA, 1995).
2. Run-of-river systems do not use large dams or storage reservoirs. Instead, smaller
diversion structures are used to channel some of the water through a canal or penstock to
a powerhouse, after which the water is returned to the river. Run-of-river systems
avoid some of the costs and environmental impacts associated with large hydro
facilities.
3. Pumped storage projects use off-peak electricity (usually from a baseload power plant)
to pump water to an upper reservoir; this water is later released to flow through a
generator during periods of peak demand. Such plants are net consumers of energy.
Although pumped storage is not a renewable energy technology, it can result in a net
reduction in greenhouse gas emissions when the fuel providing electricity for pumping
has a lower carbon content than the fuel being displaced by the pumped storage
generation (DOE, 1994).
7.20
September 15, 1997
Electricity Supply Technologies
Chapter 7
The main challenge for hydropower in recent years has been the growing concern over its local
environmental impact. By damming rivers to create storage reservoirs, hydro facilities can have an
adverse effect on terrestrial and aquatic ecosystems. Wildlife habitats can become inundated. Fish
migration routes can be cut off, and fish can die in the generating turbines or because the downstream
water quality and habitat are changed. Plants that grow along the riverbanks can be disrupted by
changes in the natural water level, both above and below the dam, and large or rapid variations in
the amount of water being discharged can disrupt aquatic habitats and accelerate erosion
downstream.
Regulatory measures - such as the licensing of non-federal hydropower projects and the Endangered
Species Act - are reducing the environmental impact of hydropower projects, but they are also
reducing total electricity production from this energy source. Between 1995 and 2010, 19 GW of
hydropower at non-federal projects will be subject to relicensing. Recent trends indicate that
relicensing results in an average 8% loss in generation due to the imposition of new environmental
constraints on operation.
Under the HE/LC scenario, and assuming a sustained regulatory reinvention effort between now and
the year 2010, incentives could be in place to increase hydroelectric power generation in either of two
ways. Neither of these opportunities involves the construction of hydropower plants at new sites.
However, both will require continued R&D to improve the design of turbine systems and to minimize
adverse environmental effects:
Increasing generation at existing hydropower plants. This option consists of modernizing and
upgrading existing turbines and generators to increase their efficiency and/or electrical output.
With enabling incentives, upgrading hydropower plants can result in energy production gains of
5%-10%. Hydropower upgrades would also have significant environmental benefits, because
new generating technologies offer improved fish passage, better water quality, and new
opportunities for improving downstream aquatic habitats.
Adding generating capacity at existing dams. A recent resource assessment identified 20 GW of
undeveloped hydropower capacity at existing dams (Rinehart et al., 1997). About 36,000 GWh
of new hydropower generation could be added by developing these sites between 1995 and 2010
(Office of Conservation and Renewable Energy, 1990).
Further expansion of hydropower capacity is possible, but unlikely until after 2010. The national
hydropower resource assessment (Rinehart et al., 1997) has identified an additional 11 GW of
environmentally acceptable hydropower at undeveloped sites (those requiring the construction of
new dams or diversions). These resources may eventually be developed, given more advantageous
economics, regulatory reinvention, and/or technology improvements. Further development of
efficient low-head generating technologies would encourage deployment at the many low-head sites
that are currently unsuitable for hydropower additions.
Considering only the near-term options, and the fact that there may be some loss of hydropower
capacity due to relicensing issues and environmental mitigation regulations, net capacity additions
by 2010 could be 10-16 GW, reducing emissions by 3-5 MtC. Additional carbon savings can be achieved
after 2010 with continuing advancements in generating technologies and environmental mitigation
techniques.
7.3.1.4 Landfill Gas
When food scraps and other organic wastes in landfills decompose, they produce methane. Methane
is a potent greenhouse gas that is also the main ingredient of natural gas. According to the
September 15, 1997
7.21
Chapter 7
Electricity Supply Technologies
Intergovernmental Panel on Climate Change, each kilogram of methane is about 21 times more
effective at trapping radiation in the atmosphere than a kilogram of carbon dioxide. Landfills are
the largest source of anthropogenic methane emissions in the United States; they are responsible for
almost 40% of these emissions each year (EPA, 1997).
New EPA regulations require operators to seal larger, closed landfills with a special cap, collect the
gas, and burn it to prevent atmospheric releases of methane. But wells sunk into landfills can capture
the gas before it escapes the surface. It can then be used for a variety of applications, including
generating electricity.
Today, about 165 landfills recover and utilize methane as a fuel. Various estimates (Governmental
Advisory Associates, 1994; EPA, 1997) indicate that between 300 and 750 of the country's 3500
landfills could economically recover methane using currently available technologies. The
development of more efficient, less expensive technologies for gas recovery, clean-up, and utilization
could accelerate the adoption of landfill gas-to-energy systems. For example, highly efficient,
experimental fuel cells have operated on landfill gas processed using new clean-up technology.
By 2010, 0.2-0.5 quads of energy per year could be recovered from the methane in landfills and
converted to electricity. Taking into account the difference in the radiative effects of methane and
CO₂, this represents the equivalent of 25-53 MtC in reduced emissions (DOE, 1994).
7.3.1.5 Other Renewable Power Technologies
This section examines three more renewable electric technologies: photovoltaics (PV), geothermal
power, and solar thermal power. Figure 7.10 illustrates that the costs for these three technologies
have also decreased sharply over the past 15 years. It is very likely that this trend will continue.
While none of these technologies are expected to contribute as much electricity as biomass cofiring or
wind power by 2010, their role in 2010 electricity markets may be significant and growing.
Photovoltaics
Photovoltaics (PV) uses solar cells to generate electricity from sunlight without any emissions or
moving parts. This technology has made substantial progress since its first successful application in
space. While PV power costs are still significantly higher than the costs of other renewable
technologies, sales of PV power systems have been growing steadily, probably because of the many
unique advantages of PV. These include modularity (applications can range from solar calculators to
power stations), widespread applicability (since adequate solar resources are widely available),
and ease of integration into the built environment (through incorporation into building facades,
roofing materials, highway sound barriers, parking-lot structures, etc.). The most important
application of PV today is in stand-alone systems that provide power to remote water pumps or off-
grid residences, for example. Because approximately two billion people live in villages without
grid electricity, remote power represents a very large and important market for PV in developing
countries.
For grid-connected applications, one of the most promising trends in the past few years is "building-
integrated" PV. Numerous buildings have been constructed - primarily in Europe, Japan, and the
United States - that incorporate PV panels in windows, awnings, or roofing materials. Thus, the PV
panels serve a dual function, which effectively lowers the cost of their role as power generators. In
these applications, the PV power directly displaces grid electricity at the end point of the delivery
system, where it has the greatest value. Another advantage of PV is that its peak power output
generally coincides with peak electricity demand, which further enhances its market value.
7.22
September 15, 1997
Electricity Supply Technologies
Chapter 7
Worldwide sales of PV power systems have grown to nearly 100 MW per year, up from 10 MW in
1982, an average annual growth rate of about 20%. This rate of growth is likely to increase as a
result of numerous programs promoting PV for village power in developing countries as well as
programs promoting greater use of PV in several developed nations. One example is the U.S. Million
Solar Roofs program, announced by President Clinton at the United Nations on June 26, 1997. Others
include Japan's Sunshine Project, Germany's subsidy of up to 70% of PV system costs, and
Switzerland's PV Schools Program.
ELA estimates that total installed PV capacity in the United States will be only 0-2 GW in 2010
(ELA, 1996). However, an independent assessment of the impact of DOE's R&D programs indicates
that, by 2010, installed U.S. PV capacity will be approximately 1.3 GW under a BAU scenario
(Office of Energy Efficiency and Renewable Energy, 1997). Using this as a starting point, and
considering the many advantages of PV, an estimate of installed capacity of 4-7 GW in 2010 is
probably reasonable for the HE/LC scenario. This would provide 6-13 TWh of electricity and reduce
carbon emissions by 1-2 MtC. One important addition to PV capacity will come from the recently
announced Million Roofs Initiative, which will result in 1-2 GW of new capacity.
The market trends for PV in 2010 are probably more significant than its energy and carbon
contributions. By 2010, PV energy prices will be substantially lower than they are today, and we
will have had considerably more experience with the development and use of building-integrated
PV products. In this context, the United States will be moving into a situation in which a significant
and increasing fraction of construction includes PV generation capabilities.
Geothermal Electricity
Geothermal power technologies use the thermal energy from underground reservoirs of hot water or
steam to produce electricity. With higher temperature resources, the steam is used to drive a turbine
directly; with lower temperature resources, a binary technology is used in which the hot water
vaporizes another working fluid, which then drives a turbine. These geothermal power-generation
technologies are considered fairly mature. The major challenge lies in locating and characterizing
the size and longevity of specific geothermal reservoirs.
Approximately 3 GW of geothermal capacity is installed in the United States today. While EIA
estimates that geothermal capacity will increase by only 0.2 GW by 2010 (ELA, 1996), DOE's recent
Quality Metrics Study indicates that geothermal power capacity will increase by 5.8 GW, and
electricity production will increase about 45 TWh, in a BAU scenario (Office of Energy Efficiency
and Renewable Energy, 1997). While a $50/tonne cost of carbon would improve the economics for
geothermal, it is not expected to provide as much of a boost as it does for wind or biomass. It is
probably reasonable to use the DOE estimate as the lower boundary, and project that total installed
geothermal power capacity in 2010 under the HE/LC case will be 8-16 GW. The 6-14 GW increase in
geothermal capacity over today's level would reduce carbon emissions by 6-16 MtC.
Solar Thermal Electric Technology
Solar thermal power technologies use mirrors to concentrate direct sunlight onto a thermal receiver,
thus creating a high-temperature energy source that can be used with a heat engine to generate
electricity. There are three types of solar thermal power systems: parabolic troughs, power towers,
and dish/engine systems. Parabolic trough systems use large fields of linear parabolic reflectors,
each of which heats a fluid flowing through a receiver pipe located along the focal line of the
reflector. About 350 MW of these systems are operating in California. A 10-MW demonstration of a
solar thermal power-tower system, which uses large mirrors to direct solar rays to a thermal
receiver atop a tower, is also operating in California. The third technology uses individual
September 15, 1997
7.23
Chapter 7
Electricity Supply Technologies
parabolic dish reflectors to provide thermal energy to a Stirling engine mounted at the focal point of
the dish. A few individual prototype units, which have power outputs of about 10-25 kW each, are
being tested in the United States.
While ELA projects negligible gains for solar thermal generating capacity by 2010 (ELA. 1996),
DOE's recent Quality Metrics Study suggests that solar thermal systems will provide
approximately 2 TWh of electricity in 2010 in a BAU scenario (Office of Energy Efficiency and
Renewable Energy, 1997). Under the HE/LC scenario of this study, an estimate of 0-2 GW capacity
and 0-6 TWh electricity generation in 2010 is probably reasonable. This would reduce carbon
emissions about 0-1 MtC.
7.3.2 The Long-Term Role of Renewables
As indicated at the beginning of this section, it is quite likely that renewable energy technologies
will play a crucial role in limiting carbon emissions and global warming in the long term. Continued
domestic and international economic development that does not foster further global warming will
require greater energy consumption coupled with lower carbon emissions. The only options are thus
low-carbon energy supplies, such as nuclear power or renewables, or the sequestration of carbon
emissions from the use of fossil fuels. With the continuing technological development and cost
reductions of renewables, renewables may become preferred energy resources some time within the
next one to three decades. Moreover, they will probably expand to become the world's primary
energy resource in the latter half of the next century. In fact, just such a transition was suggested
recently by Shell International (Figure 7.14) (Royal Dutch/Shell Group of Companies, 1996).
Figure 7.14 Sustained Growth Scenario from Shell International (Reproduced courtesy of Shell
International Petroleum Company)
1500
Surprise
Geothermal/Ocean
Solar
1000
New biomass
exajoules
Wind
Nuclear
500
Hydroelectric
Gas
Oil & natural gas liquids
Coal
0
Traditional biomass
1860
1880
1900
1920
1940
1960
1980
2000
2020
2040
2060
M68-B220902
This subsection describes the future direction and likely accomplishments of continuing R&D in
renewables. This discussion should lend credence to the prediction that non-hydro renewables will
make the transition from a minor to a major contributor to the world's electricity supplies.
7.24
September 15, 1997
Electricity Supply Technologies
Chapter 7
Biomass Power
The most important R&D areas for biomass power are in gasification/conversion systems and in
feedstock production. Gasification involves converting the solid biomass feedstock material to a gas
that is cleaned and then burned in a combustion turbine or used in a combined-cycle plant. This
technology is currently in the initial demonstration stage of development.
The importance of this technology is that it can take advantage of advanced turbine designs and
heat-recovery steam generators to achieve almost twice the efficiency of currently installed biomass
technologies (NREL estimates, 1997). High-pressure gasification technologies yield the highest
efficiencies, but they also require the development of efficient, cost-effective methods for cleaning
the hot gases before they enter the turbine.
On the biomass production side, genetic research is likely to produce energy crop species that
provide consistently higher biomass yields on an energy-content basis, thus providing a proportional
reduction in biomass feedstock costs. Related research into new species designed for better fuel
production also looks promising in terms of significantly decreasing biofuels costs over time.
Research into advanced agricultural methods will also lower feedstock production costs over time.
Finally, the development of simpler feedstock handling and processing methods will also lead to
lower costs. Whole-tree processing methods, for example, which avoid the cost of chipping the
wood before processing or use, could reduce the cost of harvesting and delivering the biomass to the
power plant by about one-third (OTA, 1995).
Taken together, improvements in biomass power conversion as well as feedstock production and
processing could reduce the cost of electricity from biomass to about 3-4 cents/kWh. This would make
biomass power very economical in comparison to other mainstream electricity sources. As biomass
power expands, most of it will employ dedicated feedstocks. In this context, biomass use will entail
low net carbon emissions. These net emissions primarily result from the combustion of fossil fuels in
production and delivery, because the carbon emitted during conversion will be reabsorbed as new
feedstock grows. Thus, biomass power can become a major contributor to reducing overall carbon
emissions from electricity generation in the coming decades.
Wind Power
The technological and economic feasibility of wind power - both in the United States and abroad
- has already been well established, as the wind generation capacity curves in Figure 7.12 indicate.
Nonetheless, major advances for wind power technology in both the short-term and long-term are
likely. These are predicted for the short-term by DOE's cost and performance projections (Office of
Utility Technologies, 1997), as illustrated in Figure 7.13.
Wind turbine design is the most critical R&D area. In general terms, the research goals are to
produce turbine designs that have half of the material content of today's turbines, at perhaps three-
quarters of the material cost (to account for more expensive materials), but with higher efficiencies
and longer lifetimes. Such design improvements will not only lower the cost of wind-generated
electricity, they will also make it economically practical to utilize the widespread, somewhat
lower quality wind resources found in Class 4 wind regimes. Some of the critical research needed to
achieve these goals includes continued empirical research into the air-turbine blade interface,
computational fluid dynamics modeling of that interaction, and fatigue testing and structural
modeling coupled with materials research. This is all aimed at producing more efficient turbine
blades that minimize material utilization while extending blade operating lifetimes.
September 15, 1997
7.25
Chapter 7
Electricity Supply Technologies
Another important area of R&D concerns the development of direct-drive generators and improved
power electronics. This will yield higher conversion efficiencies and more durable power-conversion
CO: onents, eliminating the need for a gearbox in the drive train. A major challenge will be the
gration of advanced components and controls into large-scale, utility-class hardware.
There is also considerable room for improvement in turbine manufacturing processes through process
development and automation, since today's turbine blades are still largely built by hand.
A fourth critical research area is that of wind prospecting and prediction. Wind regimes are
extremely site-specific, so even though wind resources have been broadly categorized for the nation
and the world as a whole, the siting of individual wind farms requires detailed information in order
to select the best site. Wind speeds can vary dramatically over the course of seconds (due to
turbulence), hours (diurnal variations), days (weather fronts), and months (seasonal variations).
The best locations are those with strong, sustained winds having little turbulence. Finding such
locations requires extensive prospecting and monitoring (OTA, 1995). The development of better tools
for resource characterization and prediction will both improve the economics of wind power and
enhance its value by enabling utilities to more reliably predict the power output from specific wind
power plants.
Another important thrust for research is to address siting issues. For example, the tops of ridges are
often good wind sites, but such a visible location for a wind farm can be a cause for concern when the
site is either close to a population center or in an area of particularly great scenic value. To date,
there have been virtually no studies to understand the local values associated with the visual
impact of wind systems relative to other energy technologies in the United States. Yet such analysis
could play a key role in decisions about the adoption of wind power in specific regions. Another
environmental consideration affecting site selection is the potential risk to birds, particularly
raptors, which sometimes fly into the rapidly turning rotor blades. This, too, seems like an issue
that may well be resolved through research to understand the scope of the problem relative to other
threats to bird species as well as the development of ways to keep birds a safe distance from moving
turbine blades.
In summary, the research front for wind power technology is very broad. Achievements are likely to
lead to widespread adoption and application of this electric power technology throughout the
world, wherever resources are adequate, over the next few decades.
Geothermal Electricity
Both current geothermal power systems and advanced geothermal power technology concepts will
benefit from continuing R&D.
Today's geothermal power plants use the thermal energy from hot water and steam in
hydrothermal reservoirs to generate electricity. While the power conversion and drilling
technologies related to these power plants are considered relatively mature, they will also benefit
from R&D in heat exchangers, hot fluid management systems, and new thermal conversion cycles.
These activities alone could result in energy cost reductions of at least 20% in the next few years
(NREL estimates, 1997).
The most important R&D area for conventional geothermal technology is resource exploration and
caracterization. The cost of geothermal electricity is highly dependent on resource characteristics
uch as temperature, depth, sustainable extraction rate, fluid chemistry, and ease of drilling. By
1.020, improvements in drilling technology, advanced seismic data gathering, and better computer
7.26
September 15, 1997
Electricity Supply Technologies
Chapter 7
modeling and interpretation of the data could lower the average cost of locating and assessing
geothermal resources by 50% (NREL estimates, 1997).
In the long term, geothermal power plants could make use of hot dry rock resources - areas of
exceptionally hot rock (above 150°C) that have little or no water in them. Energy can be extracted
from these zones by injecting water from the surface underground, where it is heated. Although the
engineering feasibility of extracting energy from hot dry rock has already been demonstrated
(Secretary of Energy Advisory Board, 1995), further R&D is necessary to make the technology
commercially viable. With success in that endeavor, the potential for geothermal power would be
vastly expanded because hot dry rock resources are widely available.
Photovoltaic Power Systems
Although PV power technology has already experienced major gains in both performance and
economics as a result of R&D conducted over the past 30-40 years, there is still considerable
potential for further improvements. This is true for essentially all aspects of PV power systems,
including research on basic photovoltaic materials, development of high-efficiency PV cells and
modules, development of better PV power products, lower cost manufacturing processes, and
improvements in the various components of PV systems.
A good example of the potential of PV R&D is found by comparing the module efficiencies of current
commercial PV modules with the efficiencies of individual solar cells. For crystalline silicon PV
technology, the technology representing about 90% of current sales, commercial module efficiencies
are generally between 10% and 15%, while the best laboratory cell efficiencies are well above 20%.
For thin-film PV technology, which includes amorphous silicon, copper indium diselenide, and
cadmium telluride modules, current module efficiencies are generally well under 10%, but cell
efficiencies are above 15%. Thus, in all cases, progress in commercial products would be virtually
assured through the replication of established laboratory results. There is also clearly the
potential for greater increases in cell efficiencies over today's laboratory results. Some of the
research tools that are being applied include computer modeling of various semiconductor materials
and atomic-level engineering of new devices to better understand their photovoltaic and electronic
properties.
Looking ahead, we find that significantly greater efficiencies are possible. For example, multi-
junction cells have been tested with efficiencies above 30%. At this time, these are small,
laboratory-scale devices whose initial application is expected to be with concentrators, in which
the cost of the cell is significantly offset by the increased solar energy captured by the optical
concentrator. However, in a decade or two, it is certainly conceivable that low-cost processes for
making similar high-efficiency multi-junction devices will have been developed, which will make
it possible to use them in conventional, flat-plate PV modules.
In the area of manufacturing processes, considerable effort is being made to perfect processes that
provide uniform, high-quality materials for the thin-film technologies. The fruits of these efforts
are likely to be realized in the next few years as a number of firms construct fairly large (5-20 MW
per year) manufacturing plants based on the results of process research and development.
There is still considerable progress to be made in the development of PV power products. For
example, many PV power systems today are still being individually designed for specific
applications. Off-the-shelf PV power systems and consumer products (such as PV walk-lights,
lanterns, and battery chargers) are becoming more available, but the commercial PV industry is still
a long way from making it as easy to purchase a residential PV system as it is to buy a refrigerator.
The development of products that are readily applied to such individual needs will have an
September 15, 1997
7.27
Chapter 7
Electricity Supply Technologies
important effect on PV electricity costs because it will increase the volume of sales. A particularly
important set of PV products is likely to be PV components for building shells. These include
wind ows, wall materials, awnings, and roofing materials that incorporate PV and are as readily
installed as the components they replace in today's building industry. A reasonable long-term target
is to have a large fraction of new construction incorporate such building-integrated PV products.
Finally, we will continue to see improvements in the balance-of-system components of PV systems.
Examples include power conditioners and controllers, which serve as the electrical operating and
interface system for integrating PV power modules with the load and/or the power grid. These
components will continue to improve as well as benefit from developments in power electronics.
Greater system integration is also likely, simplifying overall system design. A good example is the
development of PV modules that incorporate dc-to-ac inverters, an activity that is currently under
way.
In summary, PV technology will benefit from major R&D advances for many years to come, and these
advances will significantly improve the economics of PV power. Among the implications of these
advances, it is likely that PV power systems will reach prices of $3000/kW by 2010, which is less
than half the current average price. Further price reductions will no doubt occur beyond that.
Solar Thermal Electricity
Solar thermal technologies will benefit from R&D in a broad range of areas. For example, successful
development of durable silver/polymer reflectors will reduce reflector costs by 25% to 50% for all
three technologies, reducing system costs by 10% to 20%. Improved reflectors and receivers will also
allow higher operating temperatures and thus higher solar-to-electric conversion efficiencies.
Technology advances for Stirling engines will directly benefit dish/engine systems; one of the most
important areas is the extension of operating lifetimes between overhauls. The development and
application of hybrid solar/natural gas systems will be particularly important for power tower and
parabolic trough technologies. These will make it possible to provide dispatchable power and to use
combined-cycle technology, as well as smaller solar fields without being penalized by smaller steam
turbines, which tend to be less efficient.
By 2020, we are likely to see power-tower conversion efficiencies around 30%, compared with about
15% today, and dish/engine conversion efficiencies of about 35%, up from about 25% in current
prototypes. At the same time, these technologies will cost less and be more durable. At this stage,
they are likely to be fully competitive with other mainstream power technologies in areas with
good solar resources throughout the United States and the rest of the world.
7.4 EFFICIENCY IMPROVEMENTS IN GENERATION AND TRANSMISSION &
DISTRIBUTION
Lowering the heat rates of fossil-fueled generation results in greater efficiency (i.e., less fuel burned
per electricity generated) and lower carbon emissions. OTA (1991, p. 320), for instance, suggests that
improved maintenance could reduce heat rates by 5%, resulting in a reduction of 22 million tonnes of
carbon emissions by the year 2000. OTA includes this measure in their "moderate" scenario, viewing
it as either a low-cost or a no-cost measure. The rate of improvement assumed by OTA is consistent
with a power plant performance monitoring and improvement project conducted by the Electric Power
Research Institute (1986; 1989). Hirst and Baxter (1997) also note the value of cutting heat rates for
fossil-fuel power plants, as a carbon reduction option. No efficiency improvements to existing fossil
plants are assumed in the 1997 Annual Energy Outlook's reference case (Schouberlein, 1997).
7.28
September 15, 1997
Electricity Supply Technologies
Chapter 7
The Southern Company has had extensive experience with improving the efficiency of their electric
utility system. Over a thirteen-year period, the Southern Company was able to reduce its heat rate
by 5.8%, lowering it from approximately 10,300 Btu/kWh (in 1982) to less than 9,700 Btu/kWh (in
1994) (Southern Company, 1993; Siegel, 1997). This represents an improvement in fossil system
efficiency from approximately 33.2% (in 1982) to 34.8% (in 1994). The current level of efficiency in
U.S. fossil-fired power plants is approximately 33%. In addition to improving the company's
system-wide heat rate, the Southern Company was able to increase its reliability from 88% (in
1982) to 96% (in 1994) (Southern Electric International, 1996), and was able to increase its
availability to approximately 86%. The current availability of U.S. fossil-fired power plants is
approximately 81%.
These heat rate and availability improvements to the Southern Company's electric system have
provided benefits valued at $1.1 billion/year. One of the largest benefits to the Southern Company
has been from the deferral of 6,000 MW of new capacity. The cost of these heat rate and reliability
improvements to the Southern Company is estimated at approximately $325 million/year. The
operation and maintenance activities that comprise these costs include: establishing a heat rate
improvement training program, creation of a plant heat rate review board and a system heat rate
technical network, assignment of an efficiency engineer at each plant, instituting a program of heat
rate monitoring, and investing in design upgrades (Siegel, 1997).
The Southern Company's experience is consistent with the OTA and EPRI estimate that a 5% heat
rate improvement is technically feasible at a low cost or at no cost. Such an improvement would
result in a concomitant reduction of 5% in the carbon emissions of the utility sector. Based on
Chapter 6's HE/LC case (Table 6.4), the electricity sector's carbon emissions in 2010 would be 492
MtC. Although coal generation accounts for only 46.2% of the electricity generation forecasted for
2010, coal plants account for 81% (or 400 MtC) of the carbon produced by the electricity sector. A 5%
reduction would represent 20 Mt of carbon emissions. Assuming that 35-65% of this total is feasible,
a realistic estimate of the potential reduction is 7-13 MtC.
Improving the efficiency of transmission and distribution (T&D) systems is another supply-side
option available to utilities. As with generation, T&D improvements can include both capital
investments (for example, new transformers and conductors) and improved operations. Because T&D
losses account for only about 7% of total generation, the opportunities to reduce CO₂ emissions
through such mechanisms are limited. However, they could nonetheless be cost-effective.
Improving T&D efficiency by 10% would cut emissions by almost 1% (Hirst and Baxter, 1997).
7.5 NUCLEAR PLANT LIFE EXTENSION
In both the AEO97 reference case and the restructured case described in Chapter 6, nuclear plants are
projected to lose market share in the national mix of electricity generation. Similar trends are
forecast worldwide, with the forecasted decline in nuclear power in Europe being particularly large
(South, et al., 1997).
In the U.S., the nuclear power capacity of 99.2 gigawatts that existed in 1995 is projected to drop to
88.9 gigawatts in both the AEO97 reference case and the restructured case in 2010. This drop is
primarily the result of the retirement of 17 plants whose licenses expire between 1999 and 2010. The
combined capacity of these 17 plants is 11.5 gigawatts. The average capacity factor of the
remaining plants ranges from 76-79% throughout the forecast, deviating little from the current
capacity factor of 77%.
September 15, 1997
7.29
Chapter 7
Electricity Supply Technologies
No additional nuclear units are actively under construction in the U.S. Therefore, no new planned
units are assumed to come into service during the 2010 forecast. One nuclear unit, Watts Bar 1 owned
by closed. the Tennessee Valley Authority, received its license in 1996, but a few plants have also recently
Nuclear power is a carbon-free source of electricity. Retaining as much as possible of its current
power generation would therefore be an important carbon mitigation strategy in an economy where
carbon emissions bear a charge of $50 per tonne, as in the HE/LC scenario.
AEO97 defines a "high nuclear case" which assumes that every nuclear plant operating in 1996 has
an additional 10 years of operation, as long as their operating costs do not exceed 4 cents/kWh. This
2010 forecast results in the closure of only three nuclear plants (totaling 1.3 gigawatts of capacity)
due to license expirations and the addition of 10.2 gigawatts of new capacity from 14 plant lifetime
extensions (ELA, 1996; Nuclear Regulatory Commission, 1996). Thus, nuclear capacity in ELA's
forecast for 2010 grows from 88.9 gigawatts (in the reference case) to 99.1 gigawatts (in the "high
nuclear case"). Based on a capacity factor of 77%, this 10.2 gigawatts of capacity expansion from
nuclear plant life extensions results in 69,000 GWh of additional nuclear energy in 2010, compared to
the reference case.
According to ELA's "high nuclear case," 12 Mt of carbon would be offset by this additional carbon-
free source of electricity. Using the capacity on the margin in the HE/LC case (with carbon
emissions averaging 160 tonnes/GWh), we estimate that the carbon reductions from this additional
nuclear resource drop to 11 MtC. A range of 4-7 MtC (from 35-65% of this potential) would appear to
be a more realistic forecast for the HE/LC scenario. This range recognizes that it will not be
economical or politically feasible to extend the operation of nuclear power plants with licenses that
expire by the year 2010.
The AEO97 reference case forecasts that nuclear capacity in the U.S. will decline at an increasing
pace after 2010, decreasing from 88.9 gigawatts in 2010 to 62.7 gigawatts in 2015. Thus, with the
demand for energy continuing to grow, the impact of nuclear power as a carbon offset declines
precipitously over this slightly longer planning horizon. Under the "high nuclear case," the
assumed 10-year nuclear plant licensing extensions (subject to the 4 cents/kWh maximum cost)
increases nuclear capacity in 2015 from 62.7 gigawatts (in the reference case) to 94.7 gigawatts (in
the "high nuclear case"), Thus, the magnitude of carbon offsets offered by this strategy becomes
quite significant after 2010.
Figure 7.15 illustrates the accelerated role that nuclear power life extension could have in offsetting
carbon emissions after 2010. Only 45 of the nation's 105 nuclear plants have licenses that extend
beyond 2020 (Nuclear Regulatory Commission, 1996). An effort to maintain the viability of this
capacity could result in a very large contribution to carbon reductions over the next quarter century.
AEO97 does not estimate the cost of its "high nuclear case," although it acknowledges that the
physical degradation of some units would have to be reversed. OTA (1991) also notes the potential
carbon savings of extending the useful life of all nuclear plants to 45 years, but assumes that this
option involves either low costs or saves money. Understanding the effects of aging in order to better
manage the aging nuclear infrastructure is an important R&D topic. Pressure vessel embrittlement
and the degradation of cables, pumps, and valves can be better managed by advances in materials
science and by developing and implementing advanced monitoring technologies. Such technologies
are the result of R&D and help maintain the current licensing basis of the nation's nuclear power
plants, thereby enabling their operation to extend beyond the initial 40-year licensing period.
7.30
September 15, 1997
Electricity Supply Technologies
Chapter 7
Figure 7.15 U.S. Commercial Nuclear Power Reactor Generating Capacity
120
Licensed Capacity
Licensed Capacity-Extended 20 Years
100
LICENSED CAPACITY (GWe)
75
50
25
0
1960
1970
1980
1990
2000
2010
2020
2030
2040
2050
2060
YEAR
EFG 96-7400
7.6 ADVANCED COAL TECHNOLOGIES
To test the possible effects on carbon emissions of other advanced fossil-fired electricity generation
technologies, we replaced the advanced technologies used by ELA with estimates from DOE's Office
of Fossil Energy (see Table 7.6). These estimates changed the construction costs and heat rates for
advanced combustion turbines, combined-cycle units, and coal units. ORCED did not select the
advanced coal unit with either the ELA or the Fossil Energy estimates of this unit's costs and
operating characteristics; in both cases, its initial cost was too high to warrant inclusion in the
generation mix. The only significant change to occur was the replacement of the most advanced
combustion turbine as specified by ELA with an older combined cycle unit. The net effect of this
change on carbon emissions was negligible.
This limited analysis suggests that between now and the year 2010, highly efficient (i.e., a heat
rate of about 7000 Btu/kWh) but expensive (i.e., a cost of over $1000/kW) advanced coal units cannot
compete economically with either the generation mix that remains from the 1990s or with gas-fired
combined-cycle units.
September 15, 1997
7.31
Chapter 7
Electricity Supply Technologies
Table 7.6 Base Case Technologies Compared to Advanced Technologies (costs in 1995$)
Original
Alternative
Original
Alternative
Advanced Gas Combined Cycle
Year of construction
2005
2005
2009
2010
Capital Cost, $/kW
410
525
410
500
Heat Rate
6284
5688
5817
5538
Fixed O&M, $/kW-yr
27
16
27
16
Variable O&M, e/kWh
0.05
0.015
0.05
0.015
Advanced Gas Combustion Turbine
Year of construction
2002
2005
2008
2010
Capital Cost, $/kW
339
400
374
364
Heat Rate
10873
8699
7793
8533
Fixed O&M, $/kW-yr
11.9
17.6
16.9
17.6
Variable O&M, c/kWh
0.010
0.012
0.05
0.012
Advanced Coal
Year of construction
2006
2005
Capital Cost, $/kW
1340
1050
Heat Rate
9600
7064
Fixed O&M, $/kW-yr
34
26
Variable O&M, c/kWh
0.25
0.2
7.7
SUMMARY
ole 7.7 summarizes the potential reductions in carbon emissions that could occur as the result of the
chnology options discussed in this chapter. Each option is intended to reflect roughly the amount
that could be achieved under aggressive policies combined with a carbon incentive of approximately
$50/tonne. The total carbon reductions from the options shown in Table 7.7 range from 80 to 117 MtC
by the year 2010. Additional carbon reductions may result from landfill gas recovery, photovoltaics,
geothermal, and solar thermal resources.
The analysis of renewable energy potential over the next quarter century indicates that with a
vigorous and sustained program of research, development, and deployment, renewable energy
technologies could be providing a greater and rapidly growing contribution to electricity generation
by the year 2020. The potential contributions of carbon sequestration, advanced coal technology, and
nuclear power were not explored in this report.
7.32
September 15, 1997
Electricity Supply Technologies
Chapter 7
Table 7.7 Carbon Reduction Potential of Selected Electricity Supply Technology Options in the
HE/LC Scenario with Carbon Permit Price of $50/tonne
High-Efficiency/Low-Carbor Case
(MtC)
Converting coal-based power plants to natural gas
44
Cofiring coal with biomass
16-24
Wind
6-20
Hydropower
3-9
Efficiency Improvements
7-13
Extending the life of existing nuclear plants
4-7
Total
80-117
7.8 REFERENCES
CONEG. 1996. Utility Coal Biomass Cofiring Plant Opportunities and Conceptual Assessments.
Report available from the Northeast Regional Biomass Program, CONEG Policy Research Center,
Inc., Washington, DC. (Work performed by ANTARES Group, Inc., and Parsons Power.)
Continental Power Exchange (CPEX). 1997. See, for example, the electricity transaction prices
quoted in The Energy Daily.
Electric Power Research Institute (EPRI). 1986. Power Plant Performance Monitoring and
Improvement, Report CS/EL-4415, Vol. 3 (Palo Alto, CA: Electric Power Research Institute),
February.
Electric Power Research Institute (EPRI). 1989. Power Plant Performance Monitoring and
Improvement, Report CS/EL-4415 (Palo Alto, CA: Electric Power Research Institute), May.
Electric Power Research Institute and U.S. Department of Energy. 1997. Renewable Energy
Technology Characterizations, draft, February.
Energy Information Administration (EIA). 1995. Inventory of Power Plants in the United States
1994, Washington, DC, DOE/EIA-0095(94).
Energy Information Administration (EIA). 1996. Annual Energy Outlook 1997: With Projections to
2015. DOE/ELA-0383(97). U.S. Department of Energy, Washington, DC.
Farhar, B. 1996. Energy and the Environment: The Public View. Renewable Energy Policy Project
(REPP) Issue Brief No. 3. University of Maryland at College Park, October.
Governmental Advisory Associates, Inc. 1994. Methane Recovery From Landfill Yearbook, 1994-95.
Utility Data Institute.
Hadley, S. W. 1996. ORFIN: An Electric Utility Financial and Production Simulator, Oak Ridge,
Tennessee: Oak Ridge National Laboratory, ORNL/CON-430.
Mattice, J. S. 1991. "Ecological effects of hydropower facilities," in Hydropower Engineering
Handbook. Gulliver, J. S., and R. E. A. Amdt (eds.). McGraw-Hill, Inc., New York, New York.
September 15, 1997
7.33
Chapter 7
Electricity Supply Technologies
North American Electric Reliability Council (NERC). 1996. Generating Availability Report, 1991 -
1995, Princeton, NJ, July.
Nuclear Regulatory Commission. 1996. The Information Digest, July.
Office of Conservation and Renewable Energy. 1990. Renewable Energy Technology Evolution
Rationales, Internal Working Draft. U.S. Department of Energy, Washington, DC, October. pp. 4-5.
Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy. 1997. Quality Metrics
Database for FY 1998. U.S. Department of Energy, Washington, DC.
Office of Technology Assessment (OTA), U.S. Congress. 1995. Renewing Our Energy Future. OTA-ETI-
614. U.S. Government Printing Office, Washington, DC.
Office of Utility Technologies, U.S. Department of Energy. 1997. Draft Technology
Characterizations. To be published, August 1997. U.S. Department of Energy, Washington, DC.
Parsons, B., Y. Wan, and D. Elliott. 1995. Estimates of Wind Resources Located in Close Proximity to
Existing Transmission Lines. NREL/TP-463-7096. Internal publication. National Renewable Energy
Laboratory, Golden, CO.
RDI (Resource Data International). 1996. Powerdat Database, Boulder, Colorado: Resource Data
International, Inc.
Rinehart, B. N., J. E. Francfort, G. L. Sommers, G. F. Cada, and M. J. Sale. 1997. DOE Hydropower
Program, Biennial Report, 1996-1997. DOE/ID-10510, U.S. Department of Energy, Idaho Operations
Office, Idaho Falls, ID.
Royal Dutch/Shell Group of Companies. 1996. The Evolution of the World's Energy Systems.
London, United Kingdom.
Rueckert, Thomas. 1997. Personal communication from Tom Rueckert, U.S. Department of Energy
Solar Thermal Program, May.
Sale, M. J., G. F. Cada, L. H. Chang, S. W. Christensen, J. E. Francfort, B. N. Rinehart, S. F.
Railsback, and G. L. Sommers. 1991. Environmental Mitigation at Hydroelectric Projects, Vol. I:
Current Practices for Instream Flow Needs, Dissolved Oxygen, and Fish Passage. DOE/ID-10360.
U.S. Department of Energy, Idaho Falls, ID.
Secretary of Energy Advisory Board. 1995. Task Force on Strategic Energy Research and
Development, Annex 1: Technology Profiles. U.S. Department of Energy, Washington, DC, June.
Siegel, J. 1997. "The Economic and Environmental Benefits of Improving the Performance of Coal-
Fired Power Stations," Paper presented at the Workshop on Coal and the Global Environment,
Amsterdam, The Netherlands, June 16-17.
Schouberlein, D. (Energy Information Administration). 1997. Personal communication, July 1.
South, D. W., T.B. Meade, R. L. Camp, and M. H. Schwartz. 1997. Role of Nuclear Energy in
Reducing U.S. and Global Greenhouse Gas Emissions (Washington, DC: Energy Resources
International, Inc.), April, draft.
Southern Company. 1993. The Southern Company and the Southern Environment, September.
U.S. Department of Energy (DOE). 1994. Climate Challenge Options Workbook. U.S. Department of
Energy, Washington, DC, October.
7.34
September 15, 1997
Supply Technologies
Chapter 7
ironmental Protection Agency (EPA). 1993. Opportunities to Reduce Anthropogenic Methane
in the United States: Report to Congress. EPA 430-R-93-012. U.S. Environmental Protection
October.
nvironmental Protection Agency (EPA). 1997. EnviroSense Web site, at
>s.inel.gov/partners/xgw01154.html#meth. Landfill Methane Outreach Program.
S., and B. G. Bateman. 1995. Power Plays. Investor Responsibility Research Center.
EN"
approaches include (1) repowering with an advanced coal technology (integrated coal
pen
"on combined cycle (IGCC), or pressurized fluidized bed combustion (PFBC)) or (2) plant
Will
*nce (efficiency) improvements through various management and technical adjustments.
th of these "other" repowering options, the carbon emissions reduction potential is not as
with,
with NGCC due to (1) the magnitude of efficiency improvement and (2) the carbon (together
and nitrogen) content of coal versus natural gas.
2
All
considered
I-fired power plants greater than 50 MW, and projected to remain in operation, were
uneconnimic
for NGCC repowering: 22.5 gigawatts (GW) of capacity identified by ELA in AEO97 to be
were deleted, as were 47.5 GW determined to be unneeded due to end-use energy
the
/
improvements (see Section 6). Appendix G-2 discusses the deletion of this capacity from
?,as repowering analysis.
3
One
trilli
Btu (MBtu) is the equivalent of one thousand cubic feet (MCF) of natural gas. One
bic feet of natural gas is abbreviated as TCF.
4
inter
stimate of gas transmission cost may be high, since it may overestimate the amount of
since
and intrastate pipeline that is needed to serve the repowered capacity. Alternatively,
"10
repo,
costs are averaged over all candidate plants based on gas volume delivered to the
site, it may approximate the diseconomies of scale that might arise in expanding
or building new pipeline to serve only a limited amount of repowered capacity.
5
A
1995 gas/coal price differential assumes that (1) end-use energy efficiency has an
effect on increased utility gas consumption and/or (2) extraction/production costs for
HAS decline at the same rate as the increase in demand.
The
price differential of $0.72/MBtu represents the 1995 value as reported by ELA in its
nergy Outlook (AEO97). It represents a lower bound value, since the differential remains
over time (and demand), reflecting no price response by the natural gas industry with
AEO.4
utility fuel demand. The $1.18/MBtu reflects the 2010 gas/coal price differential within
I
1995
his differential reflects a real natural gas price increase of $0.40/MBtu ($2.04/MBtu in
44/MBtu in 2010) and a 1.9 TCF increase in utility gas demand.
7
"Partent
Repowering" is equivalent to the "no additional transmission cost" case.
Septe
15, 1997
7.35
Clinton Presidential Records
Digital Records Marker
This is not a presidential record. This is used as an administrative
marker by the William J. Clinton Presidential Library Staff.
This marker identifies the place of a tabbed divider. Given our
digitization capabilities, we are sometimes unable to adequately
scan such dividers. The title from the original document is
indicated below.
J
Divider Title:
Economic Policy Institute
1000 STREET. NW
SUITE 1200
WASHINGTON DC 20036
202/775-8810
FAX 202-775-0819
Accelerating Globalization?
The Economics Effects of Climate Change Policies on U.S. Workers
Robert E. Scott
September 17, 1997
Executive Summary
The Framework Convention on Climate Change, which the U.S. and other countries
signed in 1992, commits the signatory countries to limit, and ultimately reduce, their emissions of
greenhouse gases. These measures were designed to limit the damages which could result from
climate changes associated with increasing concentrations of these gases in the world's
atmosphere. Most manmade GHGs are generated through the burning of fossil fuels, so reducing
their emissions will require limits on energy use. The signatories to the Convention are now
engaged in negotiations that could lead to a treaty with binding commitments to reduce GHG
emissions at a Conference of the Parties in Kyoto, Japan in December 1997.
This report examines the costs to the U.S. economy of some of the policies that have been
proposed to limit GHG emissions. While there have been a number of studies of the costs of these
policies, inadequate attention has been given to the effects of greenhouse gas policies on the
distribution of income. We analyze a set of possible GHG policies using the DRI/McGraw-Hill
model of the economy, and we find that these policies could have significant costs for the
economy, especially in the next ten to fifteen years, and that their distributional effects would be
even larger than their effects on the overall level of output. Policies designed to stabilize U.S.
emissions of greenhouse gases at 1990 levels in 2010 - the least restrictive policies now being
discussed for the Kyoto negotiations - could reduce the level of national output by 1.8 to 2% in
the short run, relative to the base case of no GHG policies. Because of the increase in real
energy prices, consumption would fall even more, from 2.2 to 2.5%, relative to the base-case
forecast. If GHG emissions are reduced by 10% in 2010, then the short-run costs increase
sharply, to 2.9% of GDP and 3.6% of consumption. The key impacts of GHG policies in the
three cases analyzed in this report are summarized in Table A.
The effects of GHG policies on the distribution of income and employment should be
considered in the broader context of overall labor market trends. U.S. production workers have
experienced a steady erosion in their real wages since at least 1973 (Mishel, et al 1997). Wages
-i-
are eroding for a number of reasons, including declining rates of unionization, falling real values
of the minimum wage, deregulation and increased rates of immigration. Rapid growth in foreign
trade and investment are also important causes of declining real wages. Unless these issues are
considered in the development of GHG policies, those measures will exacerbate the effects of
globalization on American workers. Without offsetting measures, these policies can lead to
further downward pressure on the wages of the three-quarters of the workforce that does not
have a college degree.
The DRI/McGraw-Hill model shows that GHG policies could cut wage growth in half
over the next two decades. GHG policies will have a strikingly consistent, negative impact on
real wages. These reductions in real wage growth, relative to the base case, will continue for
over a decade. Real wages decline in all three scenarios between 2000 and 2006, and by 2020
they are only 3% to 4% higher than they were in 1995 (versus a 9% increase in the base case).
This translates into an average growth rate of less than 0.2% per year. With such low rates of
wage growth the continued increase in inequality will lead to falling wages for most workers.
In the short run, 1.5 to 2.6 million fewer jobs would be created, and in the long run there
will be substantial shifts in employment and population between industries and regions of the
country. Even in the long run, after the initial shock of adjusting to higher energy prices is
absorbed, total employment would still be lower with GHG policies than without them. In
addition, 400,000 to 500,000 high-paying job opportunities in mining, transportation and public
utilities would be eliminated (relative to base-case forecasts), while 240,00 to 290,000 jobs would
be created in low-wage retail trade sectors. In addition, 180,000 to 340,000 high-wage
manufacturing jobs would also be eliminated if GHG emission quotas were allocated to industry
(rather than auctioned to raise government revenues).
Coal, in particular, is the most carbon-intensive of all the fossil fuels, so all GHG policies
will have a disproportionately large impact on this industry. Coal output is projected to fall by 45
to 50% in 2020, relative to a base case with no restrictions on carbon emissions. This will
eliminate 35,000 to 38,000 jobs in this industry by 2020, which is already forecast to lose 36,000
jobs in the base case. GHG policies will double projected job losses in this hard-hit sector. Other
individual industries that would be especially hard hit by GHG policies include rubber and plastics
footwear, leather footwear and related goods, kitchen products, apparel, toys, and various other
mining industries (fertilizers, nonferrous metals and iron and ferroalloy ores).
GHG policies will sharply increase energy prices in the U.S., especially for coal and
natural gas, as shown in Table A. As a result of higher energy prices, firms and individuals will
increase their investments in new and improved homes, cars and other equipment in order to
reduce energy use. As a result, the overall capital stock of the U.S. will increase by 8% to 10%
by 2020. The increase in investment will create jobs in some industries, while destroying jobs in
others.
The DRI/McGraw-Hill model, like most other macroeconomic models. is not designed to
-ii-
fully capture the effects of greenhouse gas policies on patterns of trade or foreign investment
flows. Most models assume that the U.S. will continue to purchase most of its imports from
other developed countries, which will also need to limit greenhouse gas emissions. However,
higher energy prices will increase the competitiveness of exporters based in unaffected low-wage
developing countries, who will capture a growing share of the U.S. market. Consequently, the
trade deficit will expand as rising energy prices make imports more attractive and our exports less
competitive. Increases in energy prices, and in the trade deficit, will cause the shift of resources
out of mining and manufacturing into services and other low-wage sectors of the economy that
was noted above. This could amplify labor-market problems that will result from GHG policies.
The model forecasts that the trade deficit, as a share of GDP, will increase by 1.3 to 2.1
percentage points if GHG emissions are reduced to 1990 levels in 2010. Imports would increase
by $175 to $240 billion per year. The trade deficit would increase by $149 to $240 billion per
year. Deficits of this size would have huge impacts on traded goods, especially manufacturing-
sector products. Using input output multipliers from the U.S. Department of Commerce, we have
estimated that a gross total of 2 to 4 million fewer jobs will exist or be created in industries
producing traded goods, and related inputs because of these increases in the trade deficit, as
compared with the base case forecast. Jobs will also be created in some sectors, and as a result,
gross job displacement is likely to exceed the net job losses described above.
These trade estimates may even be too low. Multinational firms in developed countries
will have additional incentives to shift investment to the developing countries, if those countries
do not face limits on their emissions of greenhouse gases. The DRI/McGraw-Hill model does not
include the effects of plant closures or investment diversion to countries excluded from the
agreement, so it may underestimate the effects that GHG policies will have on the U.S. trade
deficit, and on manufacturing output, employment, and wages.
Greenhouse gas policies are likely to apply market-based mechanisms to discourage
energy use. These could include a tax on the carbon content of fuels, or a system of permits that
are auctioned to the highest bidder. This report considers two systems of permits, rather than a
direct carbon tax, in which markets ration energy use. Either system will generate substantial
amounts of revenue for either the government, or the holders of private permits. If revenues are
collected by the government and used to reduce public borrowing, then sustained losses in
consumption will result. If the permits are simply granted to businesses, free of charge, then
corporate profits will increase dramatically and labor's share of GDP will decline sharply. The
DRI model projects that total before tax corporate profits would increase by over $1 trillion per
year in 2020, if the permits are simply granted to business, and the share of profits in GDP will
increase by 9.4 to 11.3 percentage points. However, if revenues are collected by the government,
then profits will not rise as a share of output, and labor's share will not fall.
This report makes no assessment of the potential benefits of GHG policies, in terms of
potential risks or damages that could be avoided, nor does it take a position on the advisability of
adopting any further policies to restrict GHG emissions. However, we recommend that if such
-iii-
policies are implemented. two specific, additional measures regarding adjustment assistance and
border taxes are needed to limit their disruptive effects on the economy, and on the lives of
American workers.
The first concerns the impact of GHG policies on globalization of trade and investment.
If two different sets of standards or timetables for emission reductions are established for
developed and developing counties, they will create incentives for firms to increase imports and
move production and investment out of the U.S., in order to take advantage of lower energy
prices and less restrictive emission limits in the developing countries. There are policies available
which can reduce the trade and investment distorting effects of GHG policies.
In order to reduce or eliminate the negative effects of GHG policies on U.S. trade and
investment, it could even be necessary to establish a border equalization tax policy. This tax
would rebate the cost of emission limits on the energy content of exports, and assess an equivalent
fee on products imported into the U.S. This kind of policy could eliminate the incentive to import
goods and move plants abroad to escape U.S. emission limits. Other policies which could reduce
the effects of GHG policies on trade and foreign investment include quotas on energy intensive
imports, and safeguard measures to limit import surges. However, each of these measures could
conflict with our obligations under the GATT, increasing the difficulty of developing a
comprehensive GHG policy package.
The second type of policy that would be needed is a set of measures that would provide
substantial assistance to the workers and communities damaged by the creation of greenhouse gas
policies. For workers, such a program should involve meaningful retraining opportunities, and
several years of income support while involved in those activities. For particularly hard hit sectors
and regions, special programs could be required to provide for early retirements, buyouts and to
ensure continuation of health and pension benefits. For communities, there are many options
including, but not limited to, industrial and R&D policy support.
Worker retraining policies, especially in the trade adjustment area, have established a poor
track record, which may raise questions about the wisdom of incorporating such measures into a
GHG policy or treaty package. These policies have failed for several reasons. They have
historically been poorly funded, not providing sufficient resources for meaningful retraining nor
the time and family income needed to sustain it. In addition, commitments have often been
dishonored in the past, when funding for adjustment programs was terminated or drastically
reduced after the policy had been implemented. For these reasons, a much larger and more
dependable commitment of resources would be needed to compensate workers and communities
for the costs of GHG policies.
Finally, we conclude by noting three areas of uncertainty. First, there is substantial
uncertainty about the pace of global warming and the precise climactic impact of a given volume
of CO₂ emissions. Second, this report analyzes the costs, but not the potential benefits of climate
change policy, including any possible economic benefits. Third, there are many unknowns about
-iv-
the costs of reducing emissions. The DRI model provides one way to estimate those costs. This
model could overstate or understate the costs of reducing GHG emissions. If the energy
efficiency of the economy grows more slowly than is expected in the base case, then greater
reductions in energy use will be required to meet any given standard. which would dramatically
increase the costs of compliance. On the other hand, in the past such modeling exercises have
produced estimates that have proved, in retrospect, to significantly overestimate the costs of new
pollution controls. However, even if the models discussed in this report overestimate the costs of
GHG policies by 50% or more, the actual costs would still be large and significant.
-v-
Accelerating Globalization?
The Economics Effects of Climate Change Policies on U.S. Workers
Robert E. Scott
Introduction
The Framework Convention on Climate Change, which the U.S. and other countries
signed in 1992, committed the U.S. and other countries to a voluntary plan to reduce emissions of
greenhouse gases. In 1995, at the first Conference of the Parties (COP-I), a document was
produced which established a process designed "to elaborate policies and measures" which would
"set quantified limitation and reduction objectives within specified time frames, such as 2005,
2010, and 2020" for greenhouse gas (GHG) "emissions by sources and removals by sinks."¹ At
COP-II in 1996, the U.S. first endorsed the concept of "legally binding" emission targets. The
signatories to the Convention are now engaged in negotiations that could lead to a treaty with
binding commitments at the third Conference of the Parties in Kyoto, Japan in December 1997.
There is a high level of uncertainty about the benefits of policies to avoid climate change.
In fact, it is precisely because of the existence of this uncertainty about the effects, and the costs,
of climate change, that some economists have taken an interest in the theoretical aspects of
climate change modeling (see, for example, Nordhaus 1994). Some have found that there are
greater incentives to implement policies to reduce GHG emissions than would exist if the costs of
climate change were known with certainty.
Because of the uncertainly surrounding the benefits of climate change policy, most
analysts have focused on an assessment of the costs of GHG policies.² However, little attention
has been paid to the distributional implications of climate change policies. This paper examines
-1-
these issues using a model developed by DRI/McGraw-Hill (1997). The results of this modeling
exercise are analyzed and contrasted with a recent administration study of the costs of climate
change policies (Interagency Analytic Team 1997). DRI/McGraw-Hill also prepared the most
extensively cited model used in the IAT report. We refer here to that earlier effort as the IAT
model.
There have been a number of surveys of the cost of climate change policies.³ We rely on
some of these studies in our analysis. This report contributes an analysis of the distribution of the
costs of climate change policies across society to the debate. We make no assessment of the
potential benefits of GHG policies, in terms of potential risks or damages that could be avoided,
nor do we take a position on the advisability of adopting any further policies to restrict GHG
emissions. However, we recommend that if such policies are implemented, two specific,
additional measures regarding adjustment assistance and border taxes are needed to limit their
disruptive effects on the economy, and on the lives of American workers.
We begin by describing the baseline forecasts for the economy (business as usual case) and
then review the overall effects of potential GHG policies on output (GDP) and consumption,
including a comparison of the DRI/McGraw-Hill (1997) results with those of the IAT. The next
section considers the effects of these policies on key sectors of the economy, including
employment, wages, profits, and foreign trade. We then examine GHG policy-effects on
industries and regions of the economy. The paper concludes with a discussion of policy
implications and areas for future research.
The Macroeconomic Impacts of GHG Policies
-2-
The model used in this study is the DRI/McGraw-Hill macro-economic model. This is a
short-run forecasting model with an input-output-based structure. Some analysts have criticized
the use of the DRI model to analyze issues such as GHG policies because of its short-run
orientation, arguing that the policies proposed should be designed to reduce energy use and
carbon emissions over the next century.⁴ However, the policies discussed in IAT (1997) would
take effect in the year 2000 and be phased in over 10 years. They would have substantial short-
run impacts on economic output, which would exceed the costs identified in the other long-run
models used in the IAT report. Therefore, it is appropriate to examine these adjustment costs
when considering the distributional effects of GHG policies.
Many models have estimated the costs of stabilizing the emissions of carbon, on a national
basis, at 1990 levels by the year 2010. This scenario has become the standard of comparison for
modeling of the costs of GHG policies. The IAT chose this scenario as a baseline, and also ran a
number of "sensitivity-test" scenarios which varied key assumptions, including a range of targets
between 90% and 110% of 1990 level emissions in the year 2010. Most scenarios assume that
emissions are held constant after the target is achieved in 2010. The European Union has
proposed emission reductions of 15% of 1990 levels by 2010 as a starting point for the the
Kyoto-round negotiations. It is also important to point out that ultimate stabilization of
atmospheric CO₂ may require further reductions in global emission levels beyond these targets,
with attendant increases in the costs of GHG policies.⁶ Comparison of cases 1 and 2 suggests that
the costs of emissions control increase exponentially. Therefore adjustment costs could continue
and expand if future COP meetings results in additional, increasingly restrictive agreements to
reduce GHG emissions in the Annex-I countries, as seems likely.
-3-
Baseline forecasts
DRI/McGraw-Hill (1997) assumes that real GDP growth will average two percent per
year between 2000 and 2010, and 1.5% per year thereafter. GDP growth declines after 2010
because the rate of growth of the labor force declines from about one percent per year between
2000 and 2010 to 0.5% per year thereafter, as the baby boom generation begins to exit the labor
force. The energy intensity of the economy is forecast to decline by one percent per year
throughout the forecast period. Carbon emissions therefore increase 1.3 percent per year between
2000 and 2010 and 0.8% per year thereafter.
Total carbon emissions in the baseline case are 1,746 million metric tons (MMT) in 2010
and 1,894 MMT in 2020. Actual emissions in 1990 in the U.S. were 1323 MMT, according to
DRI/McGraw-Hill (1997).⁷ Therefore total emissions in the baseline case grow 32.0% and
43.2%, respectively, in 2010 and 2020 in this model. These are the amounts by which carbon
emissions would have to be reduced to stabilize emissions at 1990 levels.⁸ Some fuels, such as
natural gas and petroleum, emit much less carbon per unit of energy released, so energy use
would not have to fall by this amount if fuel switching were to occur (as is assumed in the DRI
model).
Policy Scenarios
It is assumed that a system of tradeable carbon emission permits is announced and
implemented beginning in 2000. The number of permits in circulation is reduced each year from
2000 to 2010, until the emissions target is achieved. Permits are not traded internationally, but
are tradeable within the domestic market, to facilitate adjustment at least possible cost. Three
cases are analyzed⁹:
-4-
Case 1:
(Grandfathered Permits). Allowance are issued to industry at no cost.
Emissions are stabilized at 1990 levels in 2010.
Case 2:
(Grandfathered Permits, 90% of '90 Emissions). Allowances are issued
to industry at no cost. Emissions stabilized at 90% of 1990 levels in 2010.
Case 3:
(Auction, Deficit Reduction) Allowances auctioned by U.S. Government,
no revenue recycling. Emissions stabilized at 1990 levels in 2010.
Carbon prices range from $180 to $200 per ton in 2010 and $270 to $320 per ton 2020 in
DRI/McGraw-Hill (1997), as shown in Table 1. These prices have very different effects on
different types of fuel, as noted above. Gasoline prices increase only 31.5% to 49.2% in 2010, for
example ($.41 to $.63/gallon). However, coal prices rise much more rapidly, 450% to 680%,
because coal has a much higher carbon content per unit of energy, as noted above. Natural Gas,
which is one of the most carbon-efficient fuels, rises substantially more in price than gasoline
because fuel switching sharply increases demand, and capacity is slow to respond.
Energy prices in the IAT draft report are compared in the bottom two sections of Table 1.
The base case prices are nearly identical in the two studies. However. carbon prices are much
lower in the IAT scenarios. There are two reasons for this. First, the IAT model assumes slightly
higher baseline improvements in energy efficiency (1.25% VS. 1.0% per year). Therefore, carbon
emissions in 2020, for example, are projected to be 1805 MMT in the IAT base case, versus
1894 MMT in DRI/McGraw-Hill (1997), a difference of about 5%. Second, and more
importantly, the IAT report apparently assumes larger price and substitution elasticities than
DRI/McGraw-Hill (1997). The IAT report has been criticized because "The price elasticities for
energy demands need further explanation. Some very rough calculations suggest that they may
-5-
be on the high side. 10" These comments suggest that the estimates in DRI/McGraw-Hill (1997)
may be more reliable than those developed for the IAT."
Effects on Output and Consumption
Figure 1 shows that in the cases considered here, GHG policies could have significant
costs for the economy, especially in the next ten to fifteen years. Policies designed to stabilize
U.S. emissions of greenhouse gases at 1990 levels in 2010 - the least restrictive policies now
being discussed for the Kyoto negotiations - could reduce the level of national output by 1.8 to
2% in the short run. If GHG emissions are reduced by 10% in 2010, then the short-run costs
increase sharply, to 2.9% of GDP and 3.6% of consumption. The maximum impact occurs in
2007 in all three scenarios, seven years after the permit system begins to be phased in, but three
years before emissions are fully reduced to the target level. Output falls because of the increase in
energy prices, shown in Table 1 above, which depresses consumption and investment spending.
Overall inflation rates, shown in Figure 2, rise by 0.25 to 1 percentage points during the
initial implementation period. By 2020, in the grandfathered permits cases, the cumulative effect
of these price effects causes a 7% to 9% increase in the level of the GDP price deflator, relative
to the base case, because businesses will pass along the cost of permits to their customers. This
increase in the level of the GDP deflator is permanent, even though inflation returns to base case
levels by 2015.
If permits are auctioned by the government, the policy will have a more depressing effect
on output, because the permit fees will cause a sharp, permanent increase in government savings.
The effect of the permits is similar to that of a tax increase. Inflation increases in the short run,
and then falls below the base case after 2008, reflecting the permanent suppression of demand
-6-
caused by this policy.
In all three scenarios, investment recovers and is increased as a share of GDP, relative to
the base case, as shown in Figure 10, below. Firms and consumers increase their spending on
capital goods to increase energy efficiency, effectively substituting capital for energy and
accelerating the replacement of the existing capital stock. For example, total purchases of
vehicles in 2020 in the three policy scenarios are 200,000 to 500,000 units higher than in the base
case. As a result of the increased investment, the available stock of effective non-residential
capital increases steadily throughout the forecast period, as shown in Table A, above. By 2020, it
is 85 to 10% higher than in the base case.
The increase in investment is effectively financed through a forced increase in national
savings in the three policy scenarios. If the permits are simply granted to businesses, free of
charge, then business revenues and corporate profits will increase dramatically. Part of these
retained earnings will then be available to finance increased investment spending. Savings through
the federal budget also increase in all three models, after an initial adjustment period. If permits
are grandfathered, then corporate profit tax revenues will increase. If the permits are auctioned
by the government, then the revenues will flow directly into the treasury. The model assumes in
all cases that the government uses the increased revenues to reduce the deficit or, if there is a
surplus, to reduce the stock of outstanding federal debt, increasing the effective rate of
government savings.
The increase in national savings and investment that would result from a system of energy
permit sales is a key element in all models of the effects of GHG policies. In the long run, the
increase in savings and investment causes GDP to be higher in 2020 than in the base case, in all
-7-
three scenarios, as shown in Figure 1. Furthermore, GDP is growing 0.1 to 0.2 percentage points
faster than in the base case in all three scenarios in 2020. Hence, the stimulus to investment
provided by GHG policies could slightly increase the underlying rate of growth of the domestic
economy, in the long run. However, the assumption that permit sales would increase national
savings is highly questionable, for reasons discussed below. In the absence of the stimulus to
savings and investment, GDP levels and growth rates would be lower throughout the forecast
period.
Figure 3 demonstrates that the increase in real energy prices has a bigger effect on
consumption than it does on output. GHG policies have a larger impact on consumption than on
GDP for at least two reasons. First, a larger share of output is devoted to investment, for reasons
explained above, and thus consumption must decline as a share of national income. Second, the
increase in energy prices will force consumers to reduce real purchases of other goods, holding
everything else constant (unless overall demand for energy is price elastic, which is not the case in
most models).
Consumption losses reach a peak of 2.2 to 3.6 percent in 2007, relative to the base case,
as shown in Figure 2. Consumption also recovers, following the GDP path in figure 1. However,
overall consumption levels remain about one-half percent below the base case in 2020 in all three
scenarios, despite the increased levels of output in that year, reflecting the increase in savings
discussed above.
Comparisons with IAT Estimates.
The IAT estimates generated much smaller projections of the losses than would result
under similar policy scenarios. For example, Figure 4 compares predicted changes in total
-8-
consumption from the base case. assuming grandfathered permits in both the IAT and the
DRI/McGraw-Hill (1997) models. The maximum loss in the IAT model, of 1.4%, occurs two
years earlier (2006) and is only 55% as large as the maximum loss in the DRI/McGraw-Hill
(1997) model of 2.5% of base consumption. The difference between the two forecasts reflects
the two key differences in energy forecasts: 1) lower base case energy use in the IAT model; and
2) higher price elasticities in that model.
The Reliability of Critical Assumptions
There are substantial economic and practical reasons to question some of the assumptions
underlying both the IAT and the DRI/McGraw-Hill (1997) models. One key assumption concerns
they way that revenues from the sale of permits are used in the economy. Figure 5, which
contrasts two scenarios from the IAT report, illustrates the effects of different ways of handling
permit revenues on total consumption in the U.S. If permit revenues are recycled to consumers,
rather than being used for deficit reduction, losses to consumers will not fade away over time. 12
In this case, consumption will fall and remain at a lower level throughout the forecast period.
Given the political realities of the federal budget processes. it is highly unlikely that permit
revenues of the type envisioned here would be used to reduce the budget deficit or the federal
debt. For example, forecasts in mid-1997 that the federal budget could achieve a small surplus
within the next few years if the economy continues to grow at its current pace have generated a
wide range of proposals for using projected revenues for new tax cuts, and/or public spending
increases. Therefore, national savings are unlikely to increase as assumed in these models.
Higher levels of federal spending for consumption purposes, for example for defense
spending, will not stimulate growth in the way envisioned in the DRI model. Thus the permits are
-9-
not likely to stimulate as much growth in GDP. nor as rapid a recovery, as is forecast in either the
DRI/McGraw-Hill (1997) or IAT (1997). On the other hand, some type of public investment, in
areas such a R&D, educational programs and other types of public infrastructure, could have an
even larger impact on productivity and economic growth rates than the private investment
spending that drives the DRI/McGraw-Hill results. Therefore, the ultimate effects of GHG
policies on the economy will be heavily effected by the ways in which the resulting revenues are
utilized.
For these reasons it is unlikely that all of the GHG revenue collected by the government
would go to deficit reduction. Some federal revenues will probably be returned to consumers
(through tax cuts) or otherwise result in new public consumption expenditures. While tax cuts or
other expenditures could stimulate the economy, they could not fully offset the depressing effects
of GHG policies. Therefore, output is likely to fall with GHG policies, in the long run, in contrast
with the case reviewed here. 13 Figure 4 shows that consumption will fall when the permit system
is put in place in 2000, recover slightly, and then begin to decline again. The sustained decline is
caused by the increase in the cost of carbon permits and energy prices shown in Table 1, above. 14
The increased cost of energy puts an increasing drag on the rate of growth of economic output.
This is the most important long-run cost of GHG policies.
The initial decline in consumption shown in Figure 4 is smaller if all the revenues are
returned to consumers, perhaps through reductions in income or other labor taxes (such as the
social security tax). This "revenue neutral" policy has a less depressing effect on total output and
consumption than the grandfathered permits policy. Losses begin to increase in 2015 because of
the long-run effects of higher energy prices. The losses and distortion effects shown in Figure 4
-10-
would increase with the size of the effective carbon tax. Permit fees as large as those estimated in
DRI/McGraw-Hill (1997, $180 to $320 per ton, Table 1 above) would have proportionately
larger effects on consumption than the IAT cases (in which permit fees are only $95 per ton in
2010 and $125 in 2020).
One final economic reason for questioning the results of both the IAT and DRI/McGraw-
Hill (1997) is that investment may not respond in the way assumed in the models. In each case
(except for the auction, 100% to consumer case shown in Figure 4), it is assumed that increases in
savings by the business and government sector stimulate a proportionate increase in domestic
investment. 15 However, even though corporate retained earnings would rise sharply if permits
were grandfathered to existing energy users, there is no guarantee that the resulting excess cash
would be invested in the U.S. Investment opportunities elsewhere in the world could be more
attractive, particularly if there are parts of the world not subject to GHG policies (as discussed
below). In fact, recent empirical research has shown that in the long run, about one-quarter of
any given change in savings has flown out of the U.S. in the past. This ratio could change in the
future, as well. For these reasons, investment (and therefore GDP and consumption) might not
receive the stimulus expected in the IAT and DRI/McGraw-Hill (1997) models even if savings are
increased. In fact, increased corporate earnings could accelerate capital flight problems that are
discussed below.
On balance, we expect that GHG policies designed to stabilize emissions at 1990 levels
will reduce consumption by at least $50 to $100 billion per year, on average, between 2000 and
2020. Losses are likely to continue and grow in the future because the policies will reduce the
economy's potential growth rate, because permit revenues are likely to be spent on tax cuts or
-11-
new public expenditures, and because savings and investment will be diverted to other countries
because of the effects of GHG policies on economic incentives. However, the Kyoto negotiations
also need to consider the distribution of the benefits and the costs of these policies.
Distributional Effects of GHG Policies
Since 1973, U.S. production workers have experienced a steady decline in their real wages
and earnings, as shown in Figure 6. After rising steadily at an average rate of $0.14 per hour (in
$1982) through 1973, real compensation began to decline steadily thereafter at a rate of $.035 per
hour. This decline in production worker incomes contrasts sharply with the continued growth in
real hourly earnings of the top twenty percent of the labor force, which has been maintained
throughout the 1980s and 1990s. Earnings inequality, the gap between the top and bottom classes
of wages earners, has increased sharply as a result. For example, the ratio of the top to the
bottom deciles of workers has increased from 3.7 to 4.5 since 1973, an increase of more than
20%.
Wages have been eroding and income inequality has been increasing for a number of
reasons, including declining rates of unionization, falling real values of the minimum wage,
deregulation and increased rates of immigration. Rapid growth in foreign trade and investment
are also important causes of declining real wages (Mishel, Bernstein and Schmitt, 1997). Unless
these issues are considered in the development of GHG policies, those measures will exacerbate
the effects of globalization on American workers. Recent reports suggest that trade may be
responsible for at least 20 to 25% of the increase in U.S. income inequality in this period (Tyson,
1997). 16 With these facts in mind, we turn next to an analysis of the effects of GHG policies on
labor and other factors of production, and then examine the effects of GHG policies on trade in
-12-
some detail.
The effects of GHG policies on employment in DRI/McGraw-Hill (1997) are illustrated in
Figure 7. As GDP and consumption fall after the permit system is introduced, unemployment
(not shown) rises initially and peaks (relative to the base case) in 2006. Overall, unemployment
(not shown) is increased by 0.6 to 1.1 percentage points in. To put this in perspective, in the
1990-1991 recession in the U.S., unemployment increased by 2.2 percentage points from its 1989
low of 5.3 percent. Thus, the effects of GHG policies on unemployment are projected be one-
quarter to one-half as large as the last U.S. recession.
Job displacement peaks in 2007 or 2008 and persists, as shown in Figure 7, even as excess
unemployment is eliminated. Displacement persists, though at lower levels, throughout the
forecast period because real wages are depressed by the rise in energy prices (and the overall price
level), causing workers to drop out of the labor force. Job losses are especially large in the
scenario with the lowest emission targets (grandfathered permits. 90% of '90 emissions), in which
energy prices must rise most rapidly.
GHG policies have a strikingly consistent, negative impact on real wages, as shown in
Figure 8.¹⁷ The rate of growth in real wages is reduced by 0.4 to 0.55 percentage points
immediately after the introduction of the permit system in 2000. These reductions in real wage
growth, relative to the base case, persist for over a decade. The losses are largest with the most
restrictive emission limits (90% of '90 emissions), and least for the case with grandfathered
permits and 1990 emission targets. In the latter case there is a small, offsetting, positive effect
from the increased ability of corporations to grant wage increases when permits are
grandfathered, which increases corporate revenues.
-13-
The decade-plus suppression of wage growth rates could cumulatively reduce the growth
in the level of real wages by more than 50% between 1995 and 2020, relative to the base case, as
shown in Figure 9. Real wages decline in all three scenarios between 2000 and 2006, and by
2020 they are only 3% to 4% higher than they were in 1995 (versus a 9% increase in the base
case). This translates into an average growth rate of less than 0.2% per year. With such low
rates of wage growth the continued increase in inequality will lead to falling wages for most
workers. Only the top groups of wage earners are likely to experience any real wage gains over
this period.
Impacts on the Business Sector
If GHG policy simply grants permits to business at no cost, then businesses currently using
energy will effectively capture the stream of income that is equal to the value of those permits.
They can choose to either use the permits themselves, or sell them to other users on the open
market. The advantages of this system are that: 1) as the number of permits in circulation is
reduced between 2000 and 2010, the lowest cost means of eliminating energy use will be
identified by market forces; and 2) the grandfathering of permits will provide some compensation
to businesses that will face higher costs of energy because of the implementation of GHG policies.
Under this policy scenario, businesses will capture revenue stream, which will grow in
value with the prices of permits, if the permits are given away, as shown in Figure 10. Corporate
profits (before taxes) are forecast to rise steadily in the base and auction, deficit reduction cases
from $500 billion in 1995 to $1.6 trillion in 2020. This reflects both the growth of the capital
intensity of the economy (the capital stock is growing faster than in the base case) and an increase
in the expected returns to capital, as well as the general growth in real output. If permits are
-14-
grandfathered, annual corporate profits in 2020 will increase by an additional $1.1 to $1.2 trillion
dollars, an increase of about 66%, as a result of the infusion of GHG permit revenues and
associate increases in the prices of domestic products.¹⁸
Investment as a share of GDP will increase in all three GHG scenarios, as shown in Figure
11. Higher energy prices will increase the levels of investment in both residential and non-
residential goods, including houses, cars and producer durable equipment (production machinery).
This will have important side effects, by stimulating the production of capital goods and
construction sectors, as shown below. The increase in investment, if it materializes, is the key to
generating increased levels of GDP in the DRI/McGraw-Hill forecasts.
A policy induced increase in the level of profits will cause profits to rise sharply as a share
of GDP in the two cases where permits are simply given to businesses (grandfathered), as shown
in Figure 12. This sizeable increase in profits reflects returns on the incremental investment
shown in Figure 11, and also returns on the asset value of carbon permits in circulation. In the
DRI/McGraw-Hill model, increasing profits have a positive macroeconomic impact, because they
provide savings to fund the investment growth. However, rising profits, as a share of national
income, also imply a fall in the labor share of income, providing further evidence that production
workers will be squeezed by GHG policies. The profit share of output increases by 9.4 and 11.2
percentage points, respectively, in the two cases with grandfathered permits in 2020, with the
larger effect occurring with the restrictive target of 90% of '90 emissions. In the base case,
profits rise gradually from 7.4% to 15% of GDP. In Scenario's 1 and 2, profits rise to 24.6% to
26.5% of GDP. Such large increases in capital's share of national income would virtually ensure
a sizeable increase in the inequality of income distributions in the U.S.
-15-
Effects of GHG Policies on Trade and Investment flows
By 2020, U.S. Imports would rise sharply under GHG policies because higher energy
prices would make foreign goods (made with cheaper energy inputs) more attractive, as shown in
Figure 13. In the short run, as output falls due to the introduction of the permit system, imports
will decline slightly. After 2007 they surpass the base case (no GHG policies), and by 2020 they
are $175 to $275 billion higher than in the base case. Imports rise most rapidly under the scenario
with the most restrictive limits on GHG gases (90% of '90 emissions), because the high costs of
carbon permits. The auction, deficit reduction case has the smallest effect on imports, because
output is slightly lower than in the case with grandfathered permits (at '90 emission levels).
Real exports, shown in Figure 14 would fall below base case exports until at least 2013,
for all three policy scenarios. Exports recover somewhat thereafter, because of the effects of
increased domestic productivity. In particular, in case 2 (permits auctioned, revenues used for
deficit reduction) exports are $25 billion higher than in the base case by 2020.
On balance, the increase in imports is considerably larger than the increase of exports, in
the long run, in all three cases. Therefore the trade deficit (imports less exports) increases sharply
as a share of GDP, as shown in Figure 15. The balance improves slightly during the initial period
of adjustment to GHG policies, between 2000 and 2010, because of the decline in consumption,
but the deficit expands rapidly thereafter. By 2020, the deficit has increased to between 6.0%
and 7.5% of GDP, as compared with a 1997 deficit of approximately 2% of GDP. The trade
deficit is forecast to grow substantially in the base case, as well. However, the GHG policies
would add between $149, $287 and $240 billion to the trade deficit in 2020 in cases 1 through 3,
respectively. The trade deficit would increase by 1.3, 2.7 and 2.1 percentage points of GDP in the
-16-
respective cases.
Deficit increases of this size would have huge impacts on traded goods, especially
manufacturing-sector products. We have estimated that $1 billion worth of exports would
generate approximately 14,000 domestic jobs, including direct and indirect employment effects, in
1995 (Scott and Lee, 1997). Using this multiplier, the forecast increases in the trade deficit could
eliminate a gross total of 2 to 4 million job opportunities in goods producing and related
industries. As shown below, the DRI/McGraw-Hill industry projections for 2020 do not reflect
employment losses of these magnitudes, because the model assumes that the economy returns to
full employment and displaced workers are reabsorbed into other industries by that time.¹⁹
However, if the forecast changes in trade flows are correct, then 2 to 4 million fewer jobs will
exist or be created in industries producing traded goods, and related inputs, could occur with
GHG policies, as compared with base case employment forecasts. In the long run, increased
imports will cause shifts in employment within the manufacturing sector, and between
manufacturing and services. As a result, gross job displacement is likely to exceed the net job
losses that are discussed below.
If GHG policies do eliminate such a large number of manufacturing jobs, they will have a
much larger impact on wages and incomes, especially for non-college educated workers, than is
suggested by the DRI/McGraw-Hill macroeconomic model, because wages and incomes are much
higher in manufacturing than in other sectors of the economy. Further research is needed to
separately estimate the effects of GHG policies on the wages and incomes of blue- and white-
collar workers.
There is also a more fundamental set of international trade and investment issues that are
-17-
not addressed in the DRI model, as suggested by outside reviewers of the IAT, noted above. The
DRI model assumes that the share of imports coming from the Annex-I countries (that will
impose GHG emission limits) and non-Annex-I countries (which will presumably not be limited,
for some time to come) will remain fixed in the base case. However, the structure of trade flows
has shifted rapidly in the past 10 years, even without the incentives provided by GHG policies.
The share of U.S. imports originating in the developing countries increased from 29.1% in 1978
to 36.4% in 1990, even while the ratio of imports to value added in the manufacturing sector was
increasing sharply, from 18.3% to 30.7% (Sachs and Shatz 1994, 10 and 12).
GHG policies will also create incentives to accelerate the pace of globalization and
integration of the world economy. Globalization will take at least two forms. First, the share of
U.S. imports originating in developing countries will grow, even more rapidly than it has in the
past (for example, Mexico and China have experienced some of the most rapid rates of growth in
U.S. imports in the 1990s). Second, investment will be diverted from the U.S. to non-Annex I
countries, especially Mexico and other countries with large supplies of low-wage, semi-skilled
labor. The pool of investment capital available for investing abroad could also be increased if
grandfathered permit policies are implemented, as shown above. For the past two decades trade
has grown twice as fast as world income, and foreign direct investment has grown four times as
fast as income. Globalization is likely to accelerate if GHG policies are implemented that do not
limit or otherwise control production shifts designed to escape GHG limits.
Sectoral and Regional Impacts of GHG Policies
Tables 2 and 3 forecast the effects of GHG policies on total and sectoral employment, at
a fairly aggregated level in the economy (1- and 2-digit SIC categories of output). Employment
-18-
changes in 2007 (Table 2) will be dominated by the initial reduction in output, relative to the base
case, as noted above. Employment changes will be almost universally negative, and will be spread
more or less evenly across the economy. Total losses are greatest when GHG limits are most
restrictive (90% of '90 emissions). DRI/McGraw-Hill predicts that a net total of 1.5 to 2.6
million jobs, or about 1.0% to 1.8% of total employment in the base case, will be lost in this first
recession. At the high levels of aggregation shown in Table 2, only the energy-related sectors are
predicted to experience very large job losses in the short run, including mining (down 11.5% to
15.4%) and petroleum refining (down 2.5% to 7.4%)
In 2020 employment changes are more mixed (Table 3). Total employment declines by
400,000 to 1 million jobs (0.1% to 0.4% of base case employment in 2020), because of the
reduction in real wages discussed above. However, employment grows in some sectors (such as
other services, retail trade and durable goods) while declining in others (including mining,
transportation and non-durables manufacturing). Contract construction employment increases in
the two cases with less restrictive limits on carbon emissions because of increased levels of
investment. Significant employment losses spread to a much wider range of industries, as the
effects of higher energy costs and related changes in other costs are spread throughout the
economy. Mining losses rise to 22% to 24% of the base case and petroleum refining employment
declines by 17% to 24%. Several industries are forecast to have significant job losses under
grandfathered permits, but gains when permits are auctioned and the revenues used to reduce the
federal deficit or increase savings, including apparel products, leather products, electrical
machinery and miscellaneous manufacturing. These are all sectors that are open to international
competition, where production is highly sensitive to wage levels. Nominal wages would be about
-19-
six percent lower than in the base case by 2020 in the auction, deficit reduction scenario.
However, wages are higher than in the base case when permits are grandfathered, which explains
why employment impacts differ at the sectoral level in these scenarios.²⁰
Employment shifts between industries could contribute to the downward pressure on
wages discussed above. Table 4 reproduces forecasts for employment changes in 2020 from
Table 3 for a selected group of the most highly aggregated industries, and also includes
information on average production worker wages in these sectors in 1995. This table strikingly
illustrates one reason why wages grow so slowly in these scenarios. 400,000 to 500,000 job
opportunities in mining, transportation and public utilities, where wages average between $14.22
and $15.32 per hour, would be eliminated (relative to base-case forecasts), while 240,000 to
290,000 jobs would be created in retail trade, which had wages averaging $7.70 per hour in 1995.
In addition, 180,000 to 340,000 manufacturing jobs, paying between $11.60 and $12.90 per hour,
would also be eliminated if GHG emission quotas were allocated to industry (rather than
auctioned to raise government revenues). This process of structural change resulting from GHG
policies will contribute to increasing inequality within labor markets in the U.S.
DRI/McGraw Hill also provides detailed forecasts at the 3 digit level for industries, and at
the 2 digit level for the 10 census regions of the U.S. Table 5 reports the industries that are
predicted to gain or lose more than 5% of base-case employment in 2020. Losses outweigh gains
in Table 4, in several ways. The number of industries experiencing losses of more than 5% of
output is more than twice as large as the number with gains of more that 5%. In addition, there
are 13 industries that lose more than 10% of output in 2020, including Rubber and Plastics
footwear (-56%) coal mining (-45%), other leather goods (-34%) and electric utilities (30%).
-20-
However, there are no industries that gain more than 7.1% in total output. These trends reflect
the broader tendency for employment to fall in manufacturing, especially durable goods, and rise
in services sectors, including education services, retail and wholesale trade and state and local
governments, as shown in Tables 3 and 4, above. Other individual industries that would be
especially hard hit by GHG policies include kitchen products, apparel, toys, and various other
mining industries (fertilizers, nonferrous metals and iron and ferroalloy ores).
Coal mining is one of largest and most heavily affected industries under GHG policies.²¹
Coal output is projected to fall by 45 to 50% in 2020, relative to a base case with no restrictions
on carbon emissions. The coal industry employed 116,000 workers in 1995. DRI/McGraw-Hill's
(1997) base case forecasts that coal industry employment will decline to 80,000 workers in 2020.
GHG policies will eliminate another 35,000 to 38,000 jobs in this industry by 2020, reducing total
employment to between 42,000 and 45,000 workers. GHG policies will double projected job
losses in this hard-hit sector between 1995 and 2020.
The tendency for GHG policies to shift employment from manufacturing to services would
be much more pronounced if trade and foreign investment effects were larger, for the reasons
noted above. In addition, given the large increase in imports and the trade deficit, shown in
Figures 12 and 14 above, DRI may be underestimating the effect of the GHG policies on output at
the sectoral level. The shift of employment from manufacturing to services, the growth of imports
and the decline in the price competitiveness of U.S. products would all have depressing effects on
U.S. wages, for the reasons discussed above.
The shift from manufacturing to services, especially retail trade, is illustrated in Table 6,
which summarizes forecast changes in population and employment in the 10 census regions of the
-21-
U.S. in 2020, under the case 1 assumptions (grandfathered permits, 1990 emission limits).
Manufacturing employment (shown in the last column) declines in 9 of 10 regions, rising only in
the Pacific Northwest, where it is attracted by the availability of relatively cheap hydro-powered
electricity. Retail trade experiences employment gains in 7 of the 9 regions that experience
declining manufacturing employment. Mining, especially coal mining, declines sharply in all
regions of the country, as noted above. Construction absorbs some of excess labor in regions that
are forecast to enjoy stable or rising population levels because of GHG policies. Overall,
employment declines in 6 of the regions (more than 1% in the East South Central and West South
Central areas), and population declines in 3 of the 10 regions. Changes in energy costs, and the
location of employment opportunities drive population shifts. In general, there is a substantial
shift of jobs and population to the Pacific Northwest, with the other 7 regions not noted here
experiencing small net gains or losses of jobs and employment in Case 1.
Regional and industrial shifts in production and employment will reinforce the tendency of
GHG policies to increase income disparities in the U.S. The movement of labor out of high-wage
sectors, such as manufacturing and mining, and into low wage sectors, such as retail trade, will
tend to increase income inequality across the U.S., as the supply of good jobs for the bottom
three-fourths of the labor force is reduced. Regionally, the Pacific Northwest, which already
enjoys relatively high standards of living, will experience employment and population gains, while
relatively low-income regions such as the East South Central will experience job and populations
losses. These changes will tend to increase the disparity which exists in wages and income
distributions among these regions, by increasing labor demands in the regions that gain and
reducing labor demand in the regions that lose jobs and population.
-22-
Policy Implications and Areas for Future Research
To date, the discussion of GHG policies has paid inadequate attention to sectoral, regional
and other adjustment policies needed to minimize the negative effects of GHG policies on the
distribution of income. Previous research on GHG policies has focused almost exclusively on the
identification of costs and benefits at the macroeconomic level of the economy. This analysis adds
consideration of equity issues to the debate.
Optimal GHG Control Strategies
As noted above, policy makers are discussing a range of goals for the Kyoto-COP-III
meetings and potential agreements. These range from stabilizing emissions in the Annex-I
countries at 110% of 1990 levels by 2010 (IAT, 1997), or 2020, to reducing them by up to 20%
as soon as 2005. 22 DRI/McGraw-Hill (1997) estimates that stabilizing or reducing emissions at
1990 levels will require an effective carbon permit price of $180 to $192/ton in 2010, and $270 to
$281 per ton in 2020 (Table 1, above). Further emission reductions would require substantially
higher carbon prices.
Nordhaus (1994, 93-96) has developed a simple simulation model of the costs and benefits
of global climate change which identifies optimal emission control rates and the carbon taxes
needed to achieve those levels of control (the carbon tax would have essentially the same effect as
a carbon permit-fee on energy use and carbon emissions). His model suggests that very small and
gradual controls are needed, with carbon taxes of only $5.29 per ton (in 1989 $), rising to $17.75
per ton in 2075. 23 He reviews a number of other, earlier studies which also reached similar
conclusions. Fankhauser and Tol (1996) also review a number of recent estimates of the marginal
damage of a ton of carbon emissions of $5 to $125 per ton of carbon, "with most estimates in the
-23-
lower end of this range. This literature suggests that the emission limits considered in IAT
(1997) and DRI/McGraw-Hill (1997) may be too restrictive, in the sense that marginal costs
would exceed marginal benefits at the given levels of carbon emission limits.
Measures to Offset the Effects of GHG Policies on Income Distribution
This report makes no assessment of the potential benefits of GHG policies, in terms of
potential risks or damages that could be avoided, nor does it take a position on the advisability of
adopting any further policies to restrict GHG emissions. However, we recommend that if such
policies are implemented, two specific, additional measures are needed to limit their disruptive
effects on the economy, and on the lives of American workers.
The first concerns the impact of GHG policies on globalization of trade and investment.
If two different sets of standards or timetables for emission reductions are established for
developed and developing counties, they will create incentives for firms to increase imports and
move production and investment out of the U.S., in order to take advantage of lower energy
prices and less restrictive emission limits in the developing countries. There are policies available
which can reduce the trade and investment distorting effects of GHG policies.
In order to reduce or eliminate the negative effects of GHG policies on U.S. trade and
investment, it could even be necessary to establish a border equalization tax policy. This tax
would rebate the cost of emission limits on the energy content of exports, and assess an equivalent
fee on products imported into the U.S. This kind of policy could eliminate the incentive to import
goods and move plants abroad to escape U.S. emission limits. Other policies which could reduce
the effects of GHG policies on trade and foreign investment include quotas on energy intensive
imports, and safeguard measures to limit import surges. However, each of these measures could
-24-
conflict with our obligations under the GATT, increasing the difficulty of developing a
comprehensive GHG policy package.
The second type of policy that would be needed is a set of measures that would provide
substantial assistance to the workers and communities damaged by the creation of greenhouse gas
policies. For workers, such a program should involve meaningful retraining opportunities, and
several years of income support while involved in those activities. For particularly hard hit sectors
and regions, special programs could be required to provide for early retirements, buyouts and to
ensure continuation of health and pension benefits. For communities, there are many options
including, but not limited to, industrial and R&D policy support.
Worker retraining policies, especially in the trade adjustment area, have established a poor
track record, which may raise questions about the wisdom of incorporating such measures into a
GHG policy or treaty package. These policies have failed for several reasons. They have
historically been poorly funded, not providing sufficient resources for meaningful retraining nor
the time and family income needed to sustain it. In addition, commitments have often been
dishonored in the past, when funding for adjustment programs was terminated or drastically
reduced after the policy had been implemented. For these reasons, a much larger and more
dependable commitment of resources would be needed to compensate workers and communities
for the costs of GHG policies.
Limitations and Areas for Future Research
Finally, we conclude by noting three areas of uncertainty. First, there is substantial
uncertainty about the pace of global warming and the precise climactic impact of a given volume
of CO₂ emissions. Second, this report analyzes the costs, but not the potential benefits of climate
-25-
change policy, including any possible economic benefits. Third. there are many unknowns about
the costs of reducing emissions. The DRI model provides one way to estimate those costs. This
model could overstate or understate the costs of reducing GHG emissions. If the energy
efficiency of the economy grows more slowly than is expected in the base case, then greater
reductions in energy use will be required to meet any given standard, which would dramatically
increase the costs of compliance. On the other hand, in the past such modeling exercises have
produced estimates that have proved, in retrospect, to significantly overestimate the costs of new
pollution controls. However, even if the models discussed in this report overestimate the costs of
GHG policies by 50% or more, the actual costs would still be large and significant.
In addition to these general areas of uncertainty, this report has raised a number of specific
concerns that require future research. These include the effect of GHG policies on foreign trade
and investment, including sectoral details and relationships between the major accounts
(merchandise, services, current and capital accounts); and if permits are traded internationally, or
joint implementation is allowed (neither of these alternatives was considered in this report), the
implications for trade and exchange rates in a general equilibrium context. Another major area of
concern is the system for implementing permits, how they will be created and traded, the nature of
the property rights involved, and the wealth effects of grandfathering or other forms of
distribution to the private sector. There is a related set of issues concerning the distribution of
income between labor and capital, and the future rate of return on capital, that GHG policies will
also intersect with. Finally, further disclosure, review and debate is needed about the technical
assumptions of the models used to analyze GHG policies. This review should include the relevant
price and/or substitution elasticities and the potential (and cost) of increasing the energy-efficiency
-26-
of the domestic economy without GHG policies. Differences in base case level of energy use,
during the forecast period, can have a very large impact on the cost of limiting carbon emissions
and this issue has not been adequately addressed in the literature.
-27-
References
Bernstein, Paul M., W. David Montgomery and Thomas F. Rutherford. 1997. "World Economic
Impacts of US Commitments to Medium Term Carbon Emissions Limits," Washington,
D.C.: Charles River Associates Incorporated, CRA No 837-06.
Bovenberg, A. Lans and Ruud A. de Mooij, 1994. "Environmental Levies and Distortionary
Taxation," The American Economic Review, 94(4), 1085-1089, (September).
Boyd, Roy, Kerry Krutilla and W. Kip Viscusi. 1995. "Energy Taxation as a Policy Instrument
to Reduce CO₂ Emissions: A Net Benefit Analysis," Journal of Environmental Economics
and Management, 29, 1-24.
DRI/McGraw-Hill. 1997. "The Impact of Carbon Mitigation Strategies on Energy Markets, the
National Economy, Industry and Regional Economies," July.
Fankhauser, Samuel and Richard S. Tol. 1996. "Climate Change Costs: Recent Advancements in
Economic Assessments," Energy Policy, 24(7), 665-673.
Goulder, Lawrence H. 1995. "Effects of Carbon Taxes in an Economy with Prior Tax
Distortions: An Intertemporal General Equilibrium Analysis," Journal of Environmental
Economics and Management, 29, 271-297.
Interagency Analytic Team. 1997. "Economic Effects of Global Climate Change Policies." Draft
of May 30, as released to the House Commerce Subcommittee on Energy and Power.
Washington, D.C.: Climate Change Task Force. Manuscript.
Hoeller, Peter, Andrew Dean and Jon Nicolaisen. 1991. "Macroeconomic Implications of
Reducing Greenhouse Gas Emissions: A Survey of Empirical Studies." OECD Economic
Studies, 16, Spring, 45-78.
Hoerner, J. Andrew and Frank Mueller. 1993. "The Impact of a Broad-Based Energy Tax on the
Competitiveness of U.S. Industry," The Natural Resources Tax Review, July-August
1993, 428-458.
Hoerner, J. Andrew and Frank Mueller. 1996. "Carbon Taxes for Climate Protection in a
Competitive World," Environmental Tax Program, Center For Global Change, College
Park, MD. Manuscript, June.
Maddison, David. 1995. "A Cost-Benefit Analysis of Slowing Climate Change," Energy Policy,
23(4/5), 337-346.
-28-
Mendeloff, John. 1988. The Dilemma of Toxic Substance Regulation. Cambridge, MA: The
MIT Press.
Mishel, Lawrence, Jared Bernstein and John Schmitt. 1997. The State of Working America,
1996-97. Armonk, NY: M.E. Sharpe.
Nordhaus, William D. 1994. Managing the Global Commons: The Economics of Climate
Change. Cambridge, MA: The MIT Press.
Nordhaus, William D. and Zili Yang. 1996. "A Regional Dynamic General-Equilibrium Model of
Alternative Climate-Change Strategies," The American Economic Review, 86(4), 741-
765. September.
Repetto, Robert and Duncan Austin. 1997. The Costs of Climate Protection: A Guide for the
Perplexed. Washington, D.C.: World Resources Institute.
Sachs, Jeffrey, and Howard J. Shatz. 1994. "Trade and Jobs in U.S. Manufacturing." Brookings
Papers on Economic Activity: 1. 1-84.
Scott, Robert and Thea Lee. 1997. "Trade Deficit, Job Losses Soar Since NAFTA."
Washington, D.C.: Economic Policy Institute Trade Fax, February 19.
Tyson, Laura. 1997. "Inequality Amid Prosperity," The Washington Post. July 8.
Wigley, T.M.L., R. Richels and J.A. Edmonds. 1996. "Economic and Environmental Choices in
the Stabilization of Atmospheric CO₂ Concentrations," Nature, 379, 18 January, 240-243.
Yellen, Janet. 1997. "Statement before the House Commerce Subcommittee on Energy and
Power," July 15.1 Typescript.
-29-
Endnotes
1. This section is based on, and the quotes are drawn from Bernstein, Montgomery and Rutherford
(1997).
2. Nordhaus (1994), which estimates optimal GHG policies from an explicitly defined welfare
model is a notable exception. Other examples include Boyd, Krutilla and Viscusi (1995) and
Madisson (1995).
3. See, for example, Hoeller, Dean and Nicolaisen (1991) and Repetto and Austin (1997).
4. See, for example, comments on the IAT draft report in letters released to the House
Subcommittee on Energy and Power, as an appendix to Yellen (1997):
Dale Jorgenson and William D. Nordhaus (April 24, 1997) "This model is highly
inappropriate extremely sensitive to assumptions about energy prices. "The letter also criticized
the "use of fixed-coefficient, input-output model." and Raymond J. Kopp (May 15, 1997) stated
that the DRI model is " useful only for the analysis of short-term adjustments, and even here,
one should proceed with caution because the model is "driven from fixed coefficient input-
output tables that can give very distorted views of the adjustment behavior of the economy."
5. The IAT report utilized three models. In addition to the DRI model discussed here, the IAT
also developed forecasts for a similar set of policy scenarios using the Markal-Macro and Second
Generation Models (SGM). These are simulation-type models that calculate equilibrium output
levels, given constraints (such as available energy resources) and a set of initial conditions and
assumptions about the structure of the economy. The Markal-Macro and SGM models are not
designed to estimate adjustment costs or economic disequilibria that will result from policy
changes. Hence the DRI model is better suited to the analysis of displacement issues, while the
general class of simulation models are more appropriate for estimating long-run consequences.
Note, however, that the IAT was criticized by reviewers for: 1) failing to utilize a full computable
general equilibrium model, rather than the Markal-Macro or SGM; and 2) because neither of
those models included a well developed foreign trade sector. See comments by Nordhaus and
Jorgenson (previously cited), Ray Kopp, John Weyant and Richard Richels, in letters released to
the House Subcommittee on Energy and Power, as an appendix to Yellen (1997).
6. See Wiley, T.M.L., R. Richels and J.A. Edmonds 1996.
7. Note that IAT (1997, Table 1) reports that total carbon emissions in 1990 were 1338 MMT,
1.1% more than is reported for 1990 consumption levels in DRI/McGraw-Hill (1997).
8. These target reduction levels are 5% to 8% larger than those required in IAT (1997). The
difference is explained by the combination in the IAT report of more rapid declines in the energy
intensity of the economy (1.25 percent per year vs. 1.0 percent per year), and the slightly higher
baseline level of carbon emissions. The efficiency assumption is the most important difference, by
far, between the two baseline cases.
-30-
9. A fourth case and a new base case are also analyzed in DRI/McGraw-Hill (1997). These cases
assume that the energy intensity of the domestic economy in the "heroic base case" improves at
1.75% per year. This corresponds to a similar scenario in the IAT report. Results parallel those
of the cases discussed below. The scenario in case 4 is otherwise identical to case 1 (grand
fathered permits). Case 4 and its comparable base case are not analyzed here because the "heroic
base case" scenario is highly unlikely to occur.
10. Comments on the IAT draft report released to the House Subcommittee on Energy and
Power, as an appendix to Yellen (1997) in letter from Richard Richels, May 30, 1997. See also
letter from Larry Goulder, June 2, 1997, "The report seems consistently to employ assumptions
and policy scenarios that contribute to low costs of emissions reductions."
11. This view is also supported by Ray Kopp, in a letter of May 30, 1997, who noted that "All of
the departures from AEO97 used in the draft report serve to decreased carbon emissions [in the
base case] and therefore make attainment of a 1990 stabilization less costly." (Letter released to
the House Subcommittee on Energy and Power, as an appendix to Yellen, 1997).
12. Figure 4 is derived from two of the scenarios shown in Figure 16 (p. 28) of the IAT draft
report.
13. If public investment were to be increased using GHG revenues, then output, consumption and
other segments of the economy could achieve higher levels in the long run than in the cases
considered here. However, increases in public investment would appear less likely to obtain than
general or targeted tax cuts, in the present political environment in the U.S.
14. Bovenberg and de Mooij (1994) have shown that, in theory, revenue raising environmental
taxes can lower economic output because they "exacerbate, rather than alleviate, preexisting tax
distortions." Goulder (1995) develops a general equilibrium model which shows that the problem
with the carbon tax is its "focus on intermediate inputs and its relatively narrow base in
comparison with income taxes." In principle, raising the cost or limiting the supply of energy
inputs will change the shape and position of the production possibility frontier, generally reducing
the amount of output that can be generated with any given set of other inputs.
15. For example, comparing Case 1 and the Base Case, the increase in investment is about 90% as
large as the increase in total national savings.
16. Mishel, Bernstein, and Schmitt (1997, 195) estimate that trade could be responsible for more
than 30% of the increase in income inequality in the U.S. between 1979 and 1989.
17. The effects on real wages are estimated by dividing the employment cost index by the
consumer price index series in DRI-McGraw-Hill (1997, 20 and exhibit 15). The base case value
for this series is then contrasted with each of the forecasts to determine the change in the real
value of compensation. Note that DRI-McGraw Hill draws a sharp contrast between changes in
the nominal wage series for case 3 (permit auctions) which decline sharply, and the other policy
cases, where nominal wages increase because of increases in corporate revenues and hence the
-31-
ability to pay. However, the variation in these results is eliminated when real wage series (and the
corresponding differences of consumer prices in each case) are considered.
18. This increase in before-tax corporate profits appears to primarily reflects an increase in the
implicit rate of return on capital. The cumulative increase in fixed nonresidential capital between
2000 and 2020 in the three cases analyzed here is between $1.0 and $1.4 trillion, roughly
equivalent to the increase in annual corporate profits by 2020. Therefore most of the increase in
profits would appear to reflect higher returns on existing capital stock. This implicit forecast of
increased returns is unrealistic, unless the world-wide corporate rate return on invested capital
rises sharply because of GHG policies. However, an alternative view might treat the incremental
profits as the returns on the imputed capital value of carbon emission permits. This view suggests
that grandfathering of permits will not only transfer revenues to permit holders, but also new
assets with a significant capital value. The implicit rate of return of capital, and wealth transfers
associated with the permits, are important issues for future research.
Note that the DRI/McGraw-Hill (1997) forecasts for the three scenarios contain puzzling and
potentially contradictory results. The Merchandise trade balance worsens by less than $100
billion in each case, although the calculated trade balances (including services) decline much more
sharply, as indicated in the text. It is unlikely that services trade can or will account for these
differences. In addition, the current account balance improves by $87 to $175 billion in the three
GHG scenarios, presumably because of the increase in national savings. However, improvements
of this magnitude in the current account are unlikely unless they are mirrored in changes in the
trade balances, which dominate current account transactions.
20. Exchange rates are essentially unchanged in the DRI/McGraw-Hill (1997) forecasts, so
changes in nominal wages and prices directly effect competitiveness. The rigidity of exchange
rates in these forecasts is a potential source of error.
21. There is a greater percentage drop in employment in rubber and plastic footwear under GHG
policies, as shown in Table 4. However, output in this industry was less than $1 billion in 1995,
as compared with coal industry output of over $28 billion in that year.
22. Association of Small Island States, 1995, "Draft Protocol Submitted to the United Nations
Framework Convention on Climate Change," as referenced in Repetto and Austin (1997, 1).
23. Nordhaus and Yang (1996) develop a regional bargaining model which generates slightly
higher optimal permit prices of roughly $30 per ton of carbon in 2090, versus $21 in the DICE
model developed in Nordhaus (1994).
-32-
Table A
Costs of Greenhouse Gas Policies
Summary of Key Impacts
Case 1
Case 2
Case 3
Grandfathered
Grandfathered Permits,
Permits
factor/year:
Permits
90% of emissions
Auctioned
Change in GDP
2007
-2.0%
-2.9%
-1.8%
2020
0.2%
0.0%
0.5%
Change in Consumption
2007
-2.5%
-3.6%
-2.2%
2020
-0.3%
-0.7%
-0.2%
Change in Total Employment
2007
(1,800,000)
(2,600,000)
(1,500,000)
2020
(600,000)
(500,000)
(100,000)
Change in Real Wages,
Relative to Base Case
2007
-3.0%
-4.3%
-3.7%
2020
-4.4%
-5.5%
-5.2%
Change in the Profit Share
of GDP
2007
3.3%
4.6%
-0.1%
2020
9.4%
11.3%
-0.4%
Change in the Effective
Non-Residential Capital Stock
2007
0.3%
0.3%
-0.1%
2020
8.2%
10.4%
8.2%
Change in Gasoline Prices
2010
33.1%
49.2%
31.5%
2020
43.1%
50.0%
43.1%
Change in Natural Gas Prices
2010
129.6%
183.0%
124.8%
2020
157.8%
179.1%
159.2%
Change Coal Prices
2010
482.2%
679.4%
451.4%
2020
760.6%
760.6%
729.3%
Sources: EPI analysis of DRI (1997)
Table 1
Energy Price Comparisons
2010
2020
DRI/McGraw-Hill Base Case
Implicit Price of Carbon ($1995)
$0.00
$0.00
Gasoline-Retail ($/gallon)
$1.30
$1.44
Coal-Electric Utility ($/mmBtu)
$1.07
$0.99
Natural Gas-Utility ($/mmBtu)
$2.30
$2.77
Electricity-Weighted Average (cents/
$4.90
$4.95
DRI Case 1- Grandfathered Permits
Implicit Price of Carbon ($1995)
$192.00
$281.00
Gasoline-Retail ($/gallon)
$1.73
$2.06
Coal-Electric Utility ($/mmBtu)
$6.23
$8.52
Natural Gas-Utility ($/mmBtu)
$5.28
$7.14
Electricity-Weighted Average (cents/
$8.13
$9.58
Percentage Changes from Base:
Gasoline
33.1%
43.1%
Coal
482.2%
760.6%
Natural Gas
129.6%
157.8%
Electricity
65.9%
93.5%
DRI Case 2-- Grandfathered Permits, 90%
of '90 Emissions
Implicit Price of Carbon ($1995)
$270.00
$320.00
Gasoline-Retail ($/gallon)
$1.94
$2.16
Coal-Electric Utility ($/mmBtu)
$8.34
$8.52
Natural Gas-Utility ($/mmBtu)
$6.51
$7.73
Electricity-Weighted Average (cents/
$9.29
$10.51
Percentage Changes from Base:
Gasoline
49.2%
50.0%
Coal
679.4%
760.6%
Natural Gas
183.0%
179.1%
Electricity
89.6%
112.3%
DRI Case 3- Permits Auctioned
Implicit Price of Carbon ($1995)
180
270
Gasoline-Retail ($/galion)
$1.71
$2.06
Coal-Electric Utility ($/mmBtu)
$5.90
$8.21
Natural Gas-Utility ($/mmBtu)
$5.17
$7.18
Electricity-Weighted Average (cents/
$8.00
$9.47
Percentage Changes from Base:
Gasoline
31.5%
43.1%
Coal
451.4%
729.3%
Natural Gas
124.8%
159.2%
Electricity
63.3%
91.3%
Administration-IAT Draft of May 30, 1997
(1990 Level Case, No International Trading, Permits Auction
Implicit Price of Carbon ($1995)
95
125
Gasoline-Retail ($/gallon)*
$1.56
$1.73
Percentage Change in Gasoline Pri
19.8%
19.8%
Comparison, DRI (1997) vs. IAT Estimates:
Implicit Price of Carbon ($1995)
189%
216%
Percentage Change in Gasoline-Ret
160%
218%
"Estimated, assuming "26 cents gailon of refined petroleum product (IA
Note that IAT Draft assumes slightly lower gasoline prices in its base
Source: EPI analysis of DRI (1997)
Table 2
Employment Change by Industry in 2007
Relative to Base Case
Grandfathered Permits,
Grandfathered Permits
90% of '90 Emissions
Auction, Deficit Reduction
Percent
Jobs Lost
Percent
Jobs Lost
Percent
Jobs Lost
Change
(thousands)
Change
(thousands)
Change
(thousands)
Civilian Employment - HH Survey
-1.3%
(1,800)
-1.8%
(2,600)
-1.0%
(1,500)
Nonagricultural Establishments
-1.3%
(1,800)
-2.0%
(2,700)
-1.2%
(1,600)
Contract Construction
-1.6%
(100)
-2.8%
(170)
-1.3%
(80)
Finance, Insurance & Real Estate
-1.2%
(90)
-1.4%
(110)
-0.3%
(20)
Mining
-13.5%
(70)
-15.4%
(80)
-11.5%
(60)
Transportation and Public Utilities
-2.5%
(170)
-4.2%
(280)
-2.8%
(190)
Total Services
-1.3%
(570)
-2.0%
(880)
-1.0%
(450)
Health Services
-1.3%
(160)
-2.0%
(240)
-1.3%
(150)
Education Services
-0.9%
(20)
-1.3%
(30)
-0.4%
(10)
Other Services
-1.3%
(390)
-2.0%
(610)
-1.0%
(290)
Retail Trade
-1.2%
(290)
-1.9%
(480)
-0.9%
(220)
Wholesale Trade
-0.6%
(50)
-1.3%
(100)
-0.5%
(40)
Federal Government
0.0%
0
0.0%
0
0.0%
0
State and Local Governments
-1.2%
(240)
-1.7%
(340)
-2.3%
(450)
Manufacturing
-1.0%
(180)
-1.2%
(210)
-0.2%
(30)
Nondurables Manufacturing
-1.5%
(110)
-1.7%
(130)
-0.4%
(30)
Food and Products
-0.9%
(15)
-1.2%
(19)
-0.4%
(7)
Tobacco Products
0.0%
0
0.0%
0
0.0%
0
Textiles and Products
-1.8%
(11)
-1.8%
(11)
-0.2%
(1)
Apparel and Products
-3.2%
(23)
-2.5%
(18)
-0.3%
(2)
Paper and Products
-1.2%
(8)
-1.6%
(11)
-0.4%
('
Printing and Publishing
-1.1%
(18)
-1.3%
(22)
-0.4%
"
Chemicals and Products
-1.3%
(13)
-2.0%
(20)
-0.5%
(5)
Petroleum Products
-4.1%
(5)
-7.4%
(9)
-2.5%
(3)
Rubber and Plastics Products
-1.2%
(12)
-1.4%
(15)
-0.3%
(3)
Leather and Products
-6.0%
(4)
-4.5%
(3)
0.0%
0
Durables Manufacturing
-0.7%
(70)
-0.9%
(90)
0.0%
0
Lumber and Wood Products
-1.3%
(10)
-1.6%
(12)
0.0%
0
Furniture and Fixtures
-0.8%
(4)
-0.8%
(4)
0.0%
0
Stone, Clay, and Glass
-1.4%
(7)
-2.0%
(10)
-0.4%
(2)
Primary Metal Industries
-1.4%
(9)
-2.0%
(13)
-0.5%
(3)
Fabricated Metal Products
-0.9%
(13)
-1.1%
(16)
-0.1%
(2)
Nonelectrical Machinery
-0.2%
(3)
-0.2%
(4)
-0.1%
(2)
Electrical Machinery
-0.8%
(13)
-0.9%
(14)
0.2%
3
Transportation Equipment
-0.1%
(2)
-0.1%
(2)
0.1%
2
Instruments and Parts
-0.1%
(1)
-0.3%
(2)
0.1%
1
Miscellaneous Manufacturing
-2.3%
(9)
-2.3%
(9)
-0.5%
(2)
Source: EPI analysis of DRI (1997)
Table 3
Employment Change by Industry in 2020
Relative to Base Case
Grandfathered Permits,
Grandfathered Permits
90% of '90 Emissions
Auction, Deficit Reduction
Percent
Jobs Lost
Percent
Jobs Lost
Percent
Jobs Lost
Change
(thousands)
Change
(thousands)
Change
(thousands)
Civilian Employment - HH Survey
-0.4%
(600)
-0.3%
(500)
-0.1%
(100)
Nonagricultural Establishments
-0.3%
(500)
-0.3%
(500)
-0.1%
(100)
Contract Construction
0.3%
20
-0.2%
(10)
1.5%
90
Finance, Insurance & Real Estate
-0.5%
(40)
-0.4%
(30)
0.7%
60
Mining
-22.2%
(100)
-24.4%
(110)
-22.2%
(100)
Transportation and Public Utilities
-5.6%
(350)
-6.4%
(400)
-5.0%
(310)
Total Services
-0.9%
(490)
-0.9%
(480)
-0.5%
(280)
Health Services
-2.2%
(320)
-2.3%
(330)
-2.0%
(290)
Education Services
0.9%
20
1.4%
30
1.4%
30
Other Services
-0.5%
(180)
-0.5%
(190)
-0.1%
(20)
Retail Trade
0.9%
240
1.1%
290
0.9%
250
Wholesale Trade
0.1%
10
-0.2%
(20)
0.8%
70
Federal Government
0.0%
0
0.0%
0
0.0%
0
State and Local Governments
1.7%
390
2.3%
520
-0.8%
(180)
Manufacturing
-1.1%
(170)
-2.1%
(330)
1.5%
240
Nondurables Manufacturing
-2.5%
(180)
-3.1%
(220)
0.4%
30
Food and Products
-1.3%
(20)
-1.1%
(17)
-1.1%
(17)
Tobacco Products
0.0%
0
0.0%
0
9.1%
1
Textiles and Products
-2.5%
(12)
-2.5%
(12)
3.3%
16
Apparel and Products
-8.8%
(49)
-9.1%
(51)
3.9%
22
Paper and Products
-1.3%
(8)
-2.4%
(15)
0.0%
0
Printing and Publishing
-1.0%
(17)
-0.9%
(15)
0.4%
7
Chemicals and Products
-2.4%
(24)
-4.9%
(48)
0.1%
1
Petroleum Products
-20.0%
(20)
-24.0%
(24)
-17.0%
(17)
Rubber and Plastics Products
-1.7%
(20)
-2.6%
(30)
0.9%
10
Leather and Products
-21.6%
(11)
-27.5%
(14)
5.9%
3
Durables Manufacturing
0.0%
0
-1.2%
(110)
2.4%
210
Lumber and Wood Products
0.1%
1
-0.3%
(2)
3.8%
27
Furniture and Fixtures
1.5%
6
1.0%
4
5.9%
24
Stone, Clay, and Glass
-0.9%
(4)
-2.4%
(11)
1.5%
7
Primary Metal Industries
-2.8%
(15)
-5.9%
(32)
-0.2%
(1)
Fabricated Metal Products
-0.2%
(3)
-1.0%
(15)
2.3%
33
Nonelectrical Machinery
2.5%
38
2.2%
33
0.8%
12
Electrical Machinery
-4.0%
(51)
-7.0%
(90)
1.9%
25
Transportation Equipment
3.1%
45
2.3%
34
3.7%
55
Instruments and Parts
1.3%
8
-0.5%
(3)
3.6%
22
Miscellaneous Manufacturing
-5.7%
(22)
-7.3%
(28)
2.1%
8
Source: EPI analysis of DRI (1997)
Table 4
Employment by Industry in 2020
Change from the Base Case
Case 1
Case 2
Case 3
Average
Grandfathered Permits,
Auction, Deficit Reduction
Hourly
Grandfathered Permits
90% of '90 Emissions
Grandfathered Permits
Earnings
percent
thousands
percent
thousands
percent
thousands
1995*
Total Civilian Employment
-0.4%
(606.3)
-0.3%
(491.0)
-0.1%
(127.3)
Nonagricultural Establishments
-0.3%
(498.8)
-0.4%
(579.0)
-0.1%
(153.5)
$11.46
Contract Construction
0.4%
23.3
-0.2%
(10.1)
1.4%
88.5
$15.04
Finance, Insurance and Real Estate
-0.5%
(39.6)
-0.3%
(26.9)
0.8%
64.9
$12.33
Mining
-23.4%
(106.2)
-17.5%
(116.6)
-22.0%
(99.8)
$15.32
Transportation and Public Utilities
-5.6%
(346.6)
-6.4%
(395.9)
-4.9%
(305.9)
$14.22
Services
-0.9%
(490.5)
-0.9%
(488.8)
-0.5%
(282.5)
$11.41
Retail Trade
0.9%
236.5
1.1%
287.5
0.9%
248.6
$7.70
Wholesale Trade
0.2%
17.8
-0.2%
(18.6)
0.9%
75.2
$12.40
Federal Government
0.0%
0.0
0.0%
0.0
0.0%
0.0
n.a
State and Local Government
1.7%
394.7
2.3%
527.6
-0.8%
(176.8)
n.a
Nondurables Manufacturing
-2.5%
(181.4)
-3.2%
(226.2)
0.3%
24.3
$11.60
Durables Manufacturing
0.0%
2.2
-1.3%
(110.9)
2.4%
210.1
$12.90
"Employment and Earnings", March 1996. Average hourly earnings for production and non-supervisory workers in the
private sector, from the establishment survey.
Source: EPI analysis of DRI (1997)
Table 5
Industrial Winners and Losers
Sectors Experiencing Gains or Losses of
More than 5% in Output in 2020
with Grandfathered Permits
Change in Output
from Base Case
Losses Exceeding 5%
2010
2020
086 Rubber & Plastics Footwear
-54.3%
-55.9%
012 Coal Mining
-45.9%
-45.3%
091 Other Leather Goods
-24.0%
-33.7%
207 Electric Utilities
-19.7%
-29.6%
014 Natural Gas
-13.7%
-28.1%
208 Gas Utilities
-13.4%
-27.3%
160 Household Audio & Video Equip.
-12.5%
-22.4%
081 Petroleum Refining Ex. Fuel Oil
-11.7%
-19.0%
090 Leather Footwear
-15.9%
-16.0%
203 Pipelines, Ex N. Gas
-10.1%
-13.7%
082 Fuel Oil
-11.3%
-13.0%
096 Kitchen Pottery Products
-11.9%
-12.5%
190 Watches & Clocks
-8.9%
-12.4%
013 Crude Petroleum
-5.2%
-8.7%
042 Apparel From Purchased Mat'l
-8.8%
-8.3%
195 Toys & Sporting Goods
-7.2%
-7.6%
016 Chemical & Fertilizer Mining
-5.7%
-6.6%
089 Leather Tanning & Finishing
-7.7%
-6.5%
011 Misc. Nonferrous Ores
-4.4%
-6.4%
010 Copper Ore Mining
-5.0%
-6.0%
108 Nonferrous Metals NEC
-4.3%
-5.8%
009 Iron & Ferroalloy Ores
-6.0%
-5.6%
083 Lubricating Oils and Greases
-5.4%
-5.0%
Gains Exceeding 5%
017 New Construction
0.0%
5.0%
130 Elevators & Mat'ls Handling Eq.
1.8%
5.2%
174 Motor Vehicles
1.5%
5.4%
146 Other Service Industry Machinery
2.1%
5.4%
141 Packaging Machinery
2.7%
5.8%
052 Partitions & Fixtures
1.9%
6.0%
127 Lawn & Garden Equipment
2.4%
6.1%
148 Computer Peripheral Equipment
2.4%
6.6%
172 Truck & Bus Bodies
3.2%
7.0%
173 Truck Trailers
3.6%
7.1%
Source: EPI analysis of DRI (1997)
Table 6
Regional Gains and Losses
Changes in Population and Employment in 2020
(Grandfathered Permits, Percent Change from Base)
Civilian
Resident
Employment
Contract
Mining
Total
Retail
Region
Population
(HH Survey)
Construction
total
Coal
Services
Trade
Manufacturing
New England
0.3%
0.1%
0.4%
-3.7%
-51.2%
-3.0%
1.1%
-1.7%
Middle Atlantic
0.0%
-0.3%
6.7%
-23.3%
-41.5%
-0.8%
0.7%
-1.9%
South Atlantic
-0.3%
-0.8%
-0.5%
-28.3%
-40.3%
-1.3%
0.3%
-1.6%
East North Central
0.3%
-0.1%
0.5%
-19.8%
-39.5%
-0.7%
1.2%
-0.5%
East South Central
-1.1%
-2.0%
-1.4%
-33.1%
-39.8%
-3.5%
-0.5%
-2.1%
West North Central
0.5%
0.2%
1.5%
-10.1%
-43.5%
-0.9%
2.1%
-0.4%
West South Central
-0.6%
-1.1%
-0.9%
-20.3%
-65.5%
-1.8%
-0.3%
-1.1%
Pacific Northwest
1.7%
1.6%
1.4%
-20.9%
-47.5%
1.2%
4.5%
1.5%
Pacific Southwest
0.1%
-0.3%
-0.2%
-14.3%
-36.1%
-0.5%
1.0%
-1.5%
Source: EPI analysis of DRI (1997)
Figure 1
Green House Gas Limits Will Reduce
GDP 1.8 to 2.9 Percent During Phase-in
1.0%
0.5%
percentage change from base case
0.0%
-0.5%
-1.0%
-1.5%
-2.0%
-2.5%
-3.0%
1995
2000
2005
2010
2015
2020
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 2
GHG Policies Increase GDP Inflation
Rates 0.25 to 1.0 Percentage Points
4.5%
4.0%
3.5%
3.0%
2.5%
2.0%
1995
2000
2005
2010
2015
2020
Base Case
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 3
GHG Policies Will Reduce Consumption
by 2.2 to 3.6 Percent During Phase-in
0.0%
Percent change from base case
-1.0%
-2.0%
-3.0%
-4.0%
1995
2000
2005
2010
2015
2020
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 4
Consumption Losses in DRI/McGraw-Hill
are Much Bigger than in IAT Report
0.0%
Change in Consumption from Base Case
-0.5%
-1.0%
-1.5%
-2.0%
-2.5%
1995
2000
2005
2010
2015
2020
Grandfathered Permits, DRI/McGraw-Hill
Grandfathered Permits, IAT (1997)
Source: EPI analysis of DRI (1997) and Interagency Analytic Team (1997)
Figure 5
Losses Persist and Grow if Costs of
GHG Permits are rebated to Consumers
0.0%
-0.2%
Change in Consumption from Base
-0.4%
-0.6%
-0.8%
-1.0%
-1.2%
-1.4%
1995
2000
2005
2010
2015
2020
Results of Administration Draft Model
Grandfathered Permits
Auction, 100% to consumers
Source: EPI analysis of Interagency Analytic Team (1997)
Figure 6
Real Hourly Earnings of Production and
Nonsupervisory Workers, 1969-1996
$9.00
$8.00
(1982-84=100)
$7.00
$6.00
$5.00
$4.00
1940 1950 1960 1970 1980 1990 2000
Source: EPI analysis of Bureau of Labor Statistics, "Employment and Earnings." April 1997.
Figure 7
GHG Policies will Reduce Employment
Growth by as much as 2.7 Millions Jobs
0
base case, millions
-0.5
-1
-1.5
-2
-2.5
Chang
-3
1995
2000
2005
2010
2015
2020
Total Nonagricultural Employment
Grandfathered Permits
Grandfathered Permits, 90% of 90 Emis.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 8
GHG Policies will Depress the Rate of
Growth of Real Wages
0.6%
0.4%
Real Wage Growth Rate
0.2%
B
0.0%
-0.2%
-0.4%
1995
2000
2005
2010
2015
2020
Employment Cost Index/Consumer Prices
Base Case
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 9
Cumulative Real Wage Growth Will be
Cut 50% by GHG Policies through 2020
110
Real Wages, 1995 = 100
108
106
104
102
100
98
1995
2000
2005
2010
2015
2020
Base Case
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 10
Corporate Profits Surge if Permits are
Given to Business (Cases 1 and 2)
3000
2500
2000
Billions
1500
1000
500
1995
2000
2005
2010
2015
2020
Base Case
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 11
GHG Policies Will Increase Investment
to Improve Energy Efficiency
2,200
2,000
1,800
hs of dollars
1,600
1,400
1,200
1,000
800
600
1995
2000
2005
2010
2015
2020
Fixed Non-Residential Investment
Base Case
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 12
Profit Shares of GDP will Rise Sharply
if Permits are Given to Businesses
30.0%
Profits before Tax as Share of GDP
25.0%
20.0%
15.0%
10.0%
5.0%
1995
2000
2005
2010
2015
2020
Base Case
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 13
Real Imports Would Rise Sharply Under
GHG Policies
300
m base case, $ billions
250
200
150
100
50
Difference
0
-50
1995
2000
2005
2010
2015
2020
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
Source: EPI analysis of DRI (1997)
Figure 14
Real Exports Would Decline Because of
GHG Policies in Most Cases
30
20
Difference from base case, $ billions
10
0
-10
-20
-30
1995
2000
2005
2010
2015
2020
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
ource: EPI analysis of DRI (1997)
Figure 15
GHG Policies will Increase the Trade
Deficit's Share of GDP by 28% to 56%
0.0%
+
-2.0%
hare of GDP
-4.0%
-6.0%
-8.0%
1995
2000
2005
2010
2015
2020
Real Net Exports of Goods and Service
Base Case
Grandfathered Permits
Grandfathered Permits, 90% of 90 emms.
Auction, Deficit Reduction
ource: EPI analysis of DRI (1997)