United States
Environmental Protection
Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S9-91/019 Sept. 1991
EPA Project Summary
Analysis of Historical Radiatively
Important Trace Gases (RITG)
Emissions: Development of a
Trace Gas Accounting System
(T-GAS) for 14 Countries
S. Piccot, T. Lynch, R. Kaufmann, C. Cleveland, and B. Moore
In September 1989, a study was com-
pleted which focused on evaluating the
feasibility of developing a country-spe-
cific CO, emissions forecast model. One
objective of this 1989 study was to de-
velop a pilot scale emissions model
which could be used to estimate energy
consumption and CO, emissions for
specific energy end-use sectors In a
country. Consistent with this objective,
a pilot scale model was developed for
Poland, South Korea, France, and India.
The second objective of this study was
to test or validate the methodology used
in the model and, If the methodology
proved to be viable, to develop a full
scale model development plan. Analy-
sis of the results from the pilot model
showed that the methodology was a
potentially viable tool for developing a
country-specific global emissions
model. Based in part of this finding,
EPA decided to Initiate a more compre-
hensive, Phase 2 study. This report sum-
marizes the results of the Phase 2 study.
The objectives of the Phase 2 study
were to: (1) develop and test a CO,
emissions model for 14 countries; (2)
conduct a limited test of the model's
forecasting capability by estimating and
comparing emissions forecasts for Po-
land with forecasts developed by other
models; and (3) use the model and ac-
companying global energy use data-
bases to summarize and assess
historical energy use and emissions
patterns for the 14 countries. A key
outcome of the Phase 2 study was the
development of model algorithms and
databases for the 14 countries. Other
key outcomes were the development of
software systems which facilitate the
use of the algorithms and databases
developed under this program, and
Which assist In the manipulation and
analysis of the model resuts. These al-
gorithms, databases, and software sys-
tems are referred to as the Trace Gas
Accounting System (T-GAS).
This Project Summary was devel-
oped by EPA's Air and Energy En-
gineering Research Laboratory,
Research Triangle Park, NC, to an-
nounce key findings of the research
project that Is fully documented In a
separate report of the same title (see
Project Report ordering Information at
back).
Introduction
Identifying and assessing the most ap-
plicable and effective mitigation strategy
for a country requires that country-specific
emission patterns be examined and that
country-specific mitigation studies be con-
ducted. Those strategies which are ulti-
mately identified as the best will vary from
country to country depending on several
highly country-specific factors such as its
energy infrastructure, energy resource
base, political and social system, economic
system, level of development, and the gen-
eral health of the economy. As the debate
over greenhouse gases and mitigation strat-
egies continues, the need to conduct rep-
resentative country-specific case studies
which are consistent with these and other
factors will continue to increase. It is this
need which prompted the research de-
scribed in this report.
^69 Printed on Recycled Paper
-------
In September 1989, a study was com-
pleted which focused on evaluating the
feasibility of developing a country-specific
carbon dioxide (CO2) emissions forecast
model. Key objectives of the 1989 study
were to develop a pilot scale emissions
model for energy related sources and to
test or validate the methodology used in
the model to estimate historical and future
CO2 emissions. This study was referred to
as the Pilot Study, and the results and
recommendations for future research were
summarized in a model development and
implementation plan, which described the
methodology used to develop a pilot scale
emission model for Poland, South Korea,
France, and India. The plan also summa-
rized the results of an analysis performed
to validate the country- and sector-specific
emissions estimates developed by the pi-
lot scale model. Finally, the plan laid the
groundwork for a full scale model develop-
ment program based on the results of the
Pilot Study.
The methodology used to develop the
pilot scale model differs from other tech-
niques applied to develop energy related
CO2 emissions models. In general, the
model consists of regression equations that
estimate the demand for specific energy
types (e.g., coal, oil, gas, electricity) in key
energy end-use sectors of a country, in-
cluding the industrial, transportation, resi-
dential, commercial, and agricultural
sectors. The input variables for these re-
gression equations include standard eco-
nomic and demographic variables such as
GNP,* fuel prices, population, and per-
sonal consumption expenditures. The equa-
tions are developed by performing
regression analyses using historical en-
ergy, economic, and other data sets for
each country. To estimate emissions for a
country with the model, the demand for
specific fuel types in each sector is first
estimated by the regression equations for
years for which economic and other input
variables are provided. These energy de-
mand estimates are then converted into
CO, emissions using country-specific emis-
sion factors and information which charac-
terize the performance and makeup of each
country's secondary energy production sys-
tem (e.g., electric utilities and refineries).
A validation of the results from the pilot
model showed that the methodology used
to develop the model represented a poten-
tially viable basis from which a country-
specific global emissions model could be
developed. As discussed earlier, a model
development work plan was prepared to
guide future work at the conclusion of the
pilot study. Based in part on these conclu-
sions, EPA decided to initiate a more com-
prehensive Phase 2 study. This report sum-
marizes the findings from this Phase 2
study.
Scope of the Phase 2 Study
In the Phase 2 study, the basic meth-
odology developed under the Pilot Study
was used to develop and validate CO2
emission models for the 14 countries listed
in Table 1. As for the Pilot Study, countries
selected for the Phase 2 study represented
a mix of differing economic, social, and
developmental characteristics. The coun-
tries included those participating in the
IPCC (Intergovernmental Panel on Climate
Change) process.
A primary objective of the Phase 2
study was to develop and test the CO2
emissions models developed for the 14
countries listed in Table 1. Other key ob-
jectives were to conduct a limited test of
the model's forecasting capability and to
use the models and accompanying global
databases to examine historical energy
and emissions patterns. The equations,
databases, and computer software devel-
oped in support of these objectives are
referred to as the Trace Gas Accounting
System (T-GAS),,
Consistent with the above objectives,
three technical analyses were conducted:
For each of the 14 countries, T-GAS
regression equations were estimated
and the model framework was devel-
oped. The representativeness of the
results from the model was then ex-
amined by comparing T-GAS results
with established energy and emis-
sions databases. Where discrepan-
cies were identified, assessments
were conducted to identify the source
of the discrepancies and, where pos-
sible, the model was revised.
Historical energy use data from the
OECD and historical CO2 emissions
estimates from T-GAS were devel-
oped and summarized for each of
the 14 countries.
• T-GAS emissions forecast capability
was examined by conducting a lim-
ited test forecast in which emissions
were estimated for Poland from 1958
to 2030 with T-GAS and the results
were compared to other model re-
sults developed for the IPCC.
Development/Performance of
Models for the 14 Countries
The quantity and mix of fuels used to
produce outputs vary greatly among sec-
tors and nations. Nevertheless, the eco-
nomic principles that guide a firm's
technological decisions can be used to
analyze fuel consumption and assess the
potential for change. As the price of a fuel
rises, the impetus for technological change
increases and the resultant change can
reduce the amount of fuel used to produce
a unit of output. Similarly, as the price of
one fuel rises relative to another, substitu-
tion allows firms to replace some fraction
of the fuel whose price has risen with the
fuel against which that price has been
registered. The key factors influencing
these changes can be identified and quan-
tified based on an econometric analysis of
historical data. By assuming that such op-
portunities for technological change and
fuel substitution persist, the behavioral re-
sponses or response functions estimated
from historical data can be useful in fore-
casting the effect of future changes.
The T-GAS model uses these response
functions to estimate demand for coal, oil,
gas, heat, and electricity. The results of
the development of these response func-
tions for each country are too complex to
discuss in this summary but are discussed
in detail by country in the full report. In-
stead, this summary examines the perfor-
mance of the equations developed for each
country.
The representativness and performance
characteristics of T-GAS response func-
tions were evaluated by comparing a
"backcast" of CO2 emissbns from the model
with an historical CO2 emissions record
developed for the U.S. Department of En-
ergy (DOE)* for each of the 14 countries.
Rotty. R.M. and G. Marland. Production of CO, from
Fossil Fuel Burning by Fuel Type, 1860-1982. U S
Department of Energy, 1984.
Table 1. Countries in Phase 2 Study
OECD' Countries Developing Countries
Franco
Italy
Japan
United Kingdom
United States
West Germany
Other Non-OECD Countries
Brazil
India
Mexico
South Korea
China
Hungary
Poland
Soviet Union
* Gross national product
Organization for Economic Co-operation and Development
-------
T-GAS backcasts were calculated by us-
ing historical economic and other data as
inputs to the regression equations in the
model to calculate the fuel-specific de-
mand for energy. The amount of CO2 re-
leased was then estimated by multiplying
these energy demand estimates by coun-
try-specific CO2 emission factors.
This model validation exercise attempts
to test the model's endogenous behavior,
such as the regression equations which
are used to estimate energy demand. The
integrity of this validation is maintained in
several ways. First, the nations used in this
validation represent a wide range of eco-
nomic systems, including developed econo-
mies, economies that have undergone rapid
development, developing economies, and
centrally planned economies. The equa-
tions for fuel intensity were estimated with
data from the period 1971 to 1985, but the
backcast extends to the late 1950s (in
some cases) and early 1960s, depending
on the availability of historical economic
data. The timing of this break provides an
additional test of the methodology, tf the
backcast reproduces the DOE historical
CO2 record, it may indicate that the dra-
matic price changes of the 1970s and 1980s
did not change the behavioral relationship
between energy use and economic activ-
ity. In fact, such stability does appear to
occur and indicates that the behavioral
relationship that prevailed in the 1970s
and 1980s can be used to forecast into the
1990s and the beginning of the next cen-
tury.
Figures 1 through 14 compare the emis-
sions estimates developed by T-GAS with
DOE emissions estimates. The figures
show the percent difference between T-
GAS DOE emission estimates for the years
examined in this study. The percent differ-
ence is defined in the following manner:
percent difference -
([T-GAS - DOE]/T-GAS)*100.
The results for the 14 countries are
presented in order of decreasing emis-
sions significance; i.e., those countries with
the greatest emissions are presented first.
As the figures show, T-GAS results are
generally in good agreement with the DOE
emission estimates. For most countries
examined, T-GAS emissions are usually
well within 10% of the DOE emission esti-
mates. The stated accuracy of the DOE
record is ± 10%.
The difference between T-GAS and
DOE emissions is consistently higher than
10% for a few countries, including South
Rotty, R.M. and G. Marland. Production of CO, from
Fossil Fuel Burning by Fuel Type, 1860-1982. U.S.
Department of Energy, 1984.
40
30
20
10
J-/0
$-20
-30
-40
-50
1961
1965
1969
1973
1977
1981
1985
Figure 1.
Percent difference between T-GAS and DOE emission estimates for the United
States from 1961 to 1986.
50
40
30
20
10
0
-30
-40
-50
1970 1972 1974 1976 1978 1980 1982 1984 1986
Figure 2. Percent difference between T-GAS and DOE emission estimates for the Soviet
Union from 1970 to 1986.
Korea, India, West Germany, and Mexico.
For South Korea, the results presented in
Figure 12 are somewhat misleading. Al-
though the differences shown for the late
1950s and early 1960s are high, the abso-
lute value of the difference between these
two methodologies is relatively small. In
fact, a detailed examination of the emis-
sions results for South Korea shows that
the model accurately predicts the rapidly
increasing energy and emissions use trends
which have occurred there since the early
1960s.
For India, agreement between the two
methodologies is poor from about 1976 to
1985. In an effort to identify the reasons for
this poor agreement, the reliability of the
OECD energy data used to develop T-
GAS response functions was examined. A
comparative study was conducted to com-
-------
1969 1971 1973 1975 1977 1979 1981 1983 1985
Figure 3. Percent difference between T-GAS and DOE emission estimates for China from
1969 to 1985.
50
40
20
5 -a,
£ -30
-40
-50
Figure 4.
s
'
y
1973
1975
1977
1979
1981
1983
1985
Percent difference between T-GAS and DOE emission estimates for Japan from
1973 to 1986.
pare OECD energy data with other energy
databases developed for India. These com-
parisons revealed no serious discrepan-
cies with the OECD database and indicated
that the discrepancy between DOE and T-
GAS emission estimates might be attribut-
able to errors in the energy data used to
develop the DOE data. The DOE data
were developed in part from United Na-
tions energy data statistics, and the en-
ergy data comparisons conducted in the
T-GAS study reveal some unexplained data
inconsistencies. A similar comparative
study was conducted for Mexico and dis-
crepancies were found between the OECD
energy data used to develop T-GAS and
the United Nations data sets used in the
DOE record. Conclusions could not be
reached as to which data set for Mexico
was in error.
West German emissions estimates for
oil use, a dominant fossil fuel source of
CO2 emissions in that country, agree rea-
sonably well with DOE for all years (the
difference is less than 10%). Coal use
compares well in the early years but di-
verges significantly from the DOE values
starting about 1970 (T-GAS is higher than
DOE). This divergence in coal use is a
primary cause of the poor agreement be-
tween the DOE and T-GAS total emissions
estimates shown in Figure 5. Closer ex-
amination of the results shows that T-GAS
total coal estimates appear to be over-
stated for many years in the 1970s and
1980s.
Summary of Historical
Emissions and Energy Use
Data
In the process of developing the emis-
sions backcasts discussed earlier, a sub-
stantial volume of historical emissions data
was developed for the 14 countries. These
historical emissions data and the historical
energy use databases used to develop T-
GAS can be used to examine historical
emissions patterns of individual countries
and to identify the factors contributing to
emissions changes within a country. By
examining the factors which have influ-
enced emissions changes in the past, valu-
able information can be developed
concerning how best to reduce emissions
in the future, and what changes might be
expected as a result of implementing spe-
cific emissions mitigation strategies. Such
detailed assessments could not be con-
ducted within the scope of this study. How-
ever, a brief overview of the types of data
developed in this study should provide the
reader with a basis for assessing how
such information could be used in perform-
ing such assessments in the future. A more
comprehensive summary of the historical
emissions and energy data developed for
each country is presented for each country
in the full report.
Figures 15, 16, and 17 present esti-
mates of total CO2 developed by T-GAS
for the 14 countries evaluated. Figure 15
shows historical emissions for the four
countries in this study that produce the
most emissions. Figure 17 presents the
historical emissions associated with sev-
eral developing countries, while Figure 16
presents the historical emissions associ-
ated with several European countries. Al-
though a detailed assessment of these
data was not conducted, some interesting
features in the data are noted. First, the
data show that CO2 emissions from sev-
eral major countries (e.g., France, Brazil,
Japan, West Germany, the United King-
dom) have been declining over the years.
These emission reductions have occurred
in spite of the fact that the economies of
many of these countries continue to grow.
-------
Several different and complex reasons
for these emission reductions can be iden-
tified based on the data developed in this
study, and each reason points to areas
where further mitigation evaluations should
be focused. Figure 17 compares CO2 emis-
sions for developing nations for various
years. As the figure shows, Brazil's emis-
sions trends are an anomaly compared to
those of other developing nations: growth
in emissions has decreased dramatically
since the late 1970s. For many years, the
Brazilian electric utility sector has steadily
increased its use of non-fossil fuels (prima-
rily hydropower) to the point that over 90%
of the total electricity generated in the coun-
try comes from non-fossil fuel power plants.
This, coupled with the fact that during the
1970s many energy end-use sectors (i.e.,
industrial, residential, and agricultural) in
Brazil dramatically increased their use of
electricity, has significantly contributed to
reductions in the use of fossil fuels. Clearly,
emissions mitigation strategies aimed at
reducing electricity consumption in Brazil
would do little to reduce CO2 emissions.
Instead, strategies aimed at reducing the
consumption of fossil fuels in major end-
use sectors seem more appropriate. For
Brazil, this includes the transportation sec-
tor, which accounted for about 50% of the
total fossil fuel consumption there in 1987.
Within this sector, road transportation is
the most significant, accounting for 87% of
the total energy consumed. The second
most significant consumer of fossil fuels is
air transport (7%), followed by ship trans-
port (3%).
Figures 18 and 19 show historical emis-
sions and energy use trends associated
with Japan and the U.S., respectively. For
both countries, the figures can be used to
examine historical changes in the overall
CO2 emission intensity for each country
(i.e., the amount of CO2 produced per unit
of total energy consumed). As Figure 18
shows, total emissions in Japan have been
increasing slowly relative to increases in
total energy consumption. As a conse-
quence, the emissions intensity in Japan
has decreased significantly from a value of
0.73 tons of carbon per TOE* in 1973 to
0.60 tons of carbon per TOE in 1986 (al-
most a 20% reduction). On the other hand,
the emissions intensity for the U.S. has
changed very little since the early 1960s.
Figure 19 shows that both total emissions
and total energy use have steadily in-
creased since the 1960s. In the early 1960s,
the emissions intensity was 0.72 tons of
carbon per TOE. By 1987, the value had
decreased very little to 0.70 tons of carbon
per TOE (less than a 3% decrease). There
are several reasons that Japan has re-
duced its emissions intensity so signifi-
cantly. As for Brazil, Japan has
simultaneously expanded its use of non-
fossil fuels at electric utilities and expanded
the role of electricity use as a major fuel in
several key sectors. Based on the regres-
sion analysis conducted to develop the
response functions, it also appears that
the energy intensity associated with sev-
eral major sectors has been steadily de-
creasing in Japan. This indicates that
energy is being used more efficiently and
that the economy has steadily moved to-
ward providing goods and services which
do not require significant amounts of en-
ergy to produce (e.g., electronics).
50
40 - - -
30 - -
70+~
0
-704-
-30 - -
-40
-50
ZJ
7963
1967
1971
1975
1979
1983
Figure 5. Percent difference between T-GAS and DOE emission estimates for West Germany
from 1963 to 1985.
1961
1964 1967 1970 1973
1976
1979 1982
1985
* Ton of oil equivalent.
Figure 6. Percent difference between T-GAS and DOE emission estimates for the United
Kingdom from 1961 to 1986.
-------
Development/Analysis of an
Emissions Forecast for Poland
Although the model was used in this
study primarily to estimate and assess his-
torical CO2 emissions, it can also be used
to estimate future emissions if values for
future economic and other variables are
provided as inputs. Key objectives of the
Phase 2 study were to examine the model's
forecasting capability and to assess its
ability to perform emissions mitigation
evaluations. The results of such an effort
are presented for Poland. In general, Po-
land was selected for this component of
the study for several reasons: (1) it is an
50
40
30
20
10
-10-
-30.
-40-
-50-
pn
/ / /M
TC2C3
J7ll7lin_i7l_cn
Figure 7.
i^ i ' rm ' * i • ' i • *—i • '—r™1—'—i—1—'—i—«—i—r
1958 1961 1964 1967 1970 1973 1976 1979 1982 1985
Percent difference between T-GAS and DOE emission estimates for Poland from
1958 to 1986.
50
40
30
10
0
-10
-30-
-40-
-50
X
• ^ i ' i '—i—'—r—i—i—i—i—i—i—i—i—i i i—r—i—i—-
1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984
Figure 8. Percent difference between T-GAS and DOE emission estimates for India from
1962 to 1985.
important case study in the mitigation evalu-
ations conducted under the IPCC; and (2)
the energy, economic, and political struc-
ture in Poland poses challenges to the
development of a representative emissions
model for the country.
The Response Strategies Working
Group of the IPCC conducted an emis-
sions mitigation study for Poland. In the
IPCC study, several future emissions sce-
narios were developed and evaluated us-
ing a different emissions model. Using the
T-GAS model developed for Poland, an
attempt was made to reproduce and com-
pare the model results obtained in the
IPCC study by using scenario assump-
tions from the IPCC study as inputs to T-
GAS.
Two different emission scenarios were
examined: the base case scenario and the
structural change scenario. Both provide
forecasts of emissions of CO2 from 1985 to
2030. In the base case scenario, the IPCC
study assumed the continuation of current
trends; i.e., no important changes in cur-
rent trends in the overall structure of the
economy, patterns of energy end-use de-
mand, or energy efficiency. Depending on
the year, base case economic growth as-
sumptions in the Polish economy range
from 2.0 to 2.6% annual growth in GNP.
The structural change scenario was devel-
oped as an IPCC case study in part to
assess the impact that restructuring the
Polish industrial sector would have on emis-
sions. Structural change is simulated by
adjusting the growth rates of five industrial
subsectors included in both the T-GAS
model and the IPCC study model (e.g.,
iron and steel production, chemical manu-
facturing). Since the most energy intensive
industries are assumed to grow more slowly
than in the base case, the structural change
scenario generally simulates a shift toward
less energy intensive industries (e.g., away
from steel production). In general, growth
rates for the five industrial subsectors are
lower here than in the base case scenario.
The structural change scenario takes into
consideration improvements in living con-
ditions, increases in the production of con-
sumer goods, and reduced reliance on
heavy industries. The overall level of eco-
nomic activity in the structural change sce-
nario is greater than in the base case.
The emissions forecast for the base
case scenario is illustrated in Figure 20.
Emissions from 1958 to 1985 (i.e., the
non-forecast years) are also shown in the
figure to provide the reader with an histori-
cal context from which to examine the
trends in future emissions. Total base case
emissions in Poland were estimated by T-
GAS to be 194,100,000 metric tons of
-------
carbon by 2030. This represents an overall
60% increase in emissions from 1985, or
an annual average increase of 1.2% per
year. This increase is on the low side of
the historical annual average increases
seen in Poland since the late 1950s. About
84% of the emissions are associated with
the use of coal, 11% with the use of oil,
and the remaining 5% with the use of gas.
This is very similar to the current distribu-
tion of emissions by fuel type and is con-
sistent with historical fuel mix patterns seen
in Poland. This fuel mix forecast can be
considered reasonable if it can be assumed
that Poland has sufficient energy reserves
over the next 40 years to satisfy the level
of fossil fuel demand estimated by the
model. For coal, this assumption may be
optimistic.
The T-GAS estimate of total emissions
for the base case is significantly lower than
the estimates developed for the IPCC case
study. In the IPCC base case, total emis-
sions were estimated to be 263,700,000
metric tons of carbon by 2030. This esti-
mate is 35% higher than the T-GAS esti-
mate for 2030. An investigation was
conducted to identify the possible source
of this discrepancy. Several potential rea-
sons were identified. First, in the IPCC
model it was assumed that no improve-
ment in energy efficiency would occur over
the efficiencies which existed in 1985 for
each sector. In T-GAS, energy efficiency
improvement is allowed to occur in a man-
ner which is consistent with changes in the
factors that have influenced Polish energy
efficiency change in the past. That is, the
model allows a "business as usual" or
base case efficiency improvement to oc-
cur. This is a key difference with the IPCC
methodology and could result in T-GAS
estimating lower energy consumption and
subsequent emissions. It seems reason-
able to assume that some improvement in
energy efficiency will occur in Poland by
2030.
A second potential source of discrep-
ancy is that the IPCC model is not capable
of projecting the future mix of primary en-
ergy (e.g., coal, oil, gas, nuclear, hydro).
Instead, the mix is assumed. In T-GAS,
fuel mix is estimated based on the regres-
sion equations developed from historical
data sets. This is a key difference in the
way IPCC and T-GAS models work and
could cause significant variations in fuel
mix and subsequent emissions estimates.
Figure 21 shows the emissions esti-
mated for the structural change scenario.
A comparison of the base case results with
the structural change results shows that
emissions decrease only slightly under the
structural change scenario. For example,
I
8
1
a
1
40-
30-
20-
10-
0 -
-10-
-20 -
-30 -
-40-
-50-
™n*™m F3r^LR3™_nn
A
"v i--\-'i>'\iiiiiii -i \ 1 1 1 r-
1963 1966 1969 1972 1975 1978 1981 1984
Figure 9. Percent difference between T-GAS and DOE emission estimates for France from
1963 to 1985.
Percent Difference from DOE Record
40 -i
30 -
20 -
10 -
0-
-10 -
-20 -
-30 -
-40 -
-50 -
71 _f7l_ ™
>-^ [23 I/I l/j "-^ r?~1 L-^J I—" L/l •— " •— • LZJ «-*T^rrvi *~^ — "•
LdM 121
^v l • • 1 • • I • • | • • | • • | T- -i 1 1 1 j— ^i r-
1962 1965 1968 1971 1974 1977 1980 1983
Figure 10. Percent difference between T-GAS and DOE emission estimates for Italy from 1962
to 1985.
total emissions in 2030 under the base
case scenario were estimated by T-GAS to
be 194,100,000 metric tons of carbon. To-
tal emissions for 2030 under the structural
change scenario were estimated by T-GAS
to be 182,314,000 metric tons of carbon;
about 6% less than the base case. An
analysis of the results indicates that, al-
though some actions were taken to reduce
emissions under the structural change sce-
nario (i.e., efficiency gains in industry and
transportation and reduced growth in en-
ergy intensive industrial subsectors), high
growth in the overall economy resulted in
significant energy increases that in some
cases overshadow the energy reductions
associated with the efficiency improve-
ments specified for structural change. For
example, the consumption of total energy
(solid, liquid, and electricity) in the trans-
portation sector in 2030 was estimated to
-------
50
40
20
10
I """*
2 -20
-30
-40
-50
1972
1974
1976
1978
1980
1982
1984
1986
Figure 11. Percent difference between T-GAS and DOE emission estimates for Mexico from
1972 to 1986.
O*=r
/
1958 1961 1964 1967 1970 1973 1976 1979 1982 1985
Figure 12. Percent difference between T-GAS and DOE emission estimates for Soutf) Korea
from 1958 to 1985.
be 17.8 MTOE* for the base case sce-
nario. Total energy consumption in the
transportation sector was estimated to be
21.4 MTOE for the structural change sce-
nario even though greater energy efficiency
improvements were assumed to occur in
road transportation compared to the base
case. In general, T-GAS estimates that, as
the Polish people become more wealthy
relative to the base case (as indicated by
the rapid increases in GDP** relative to the
base case), the demand for increased mo-
bility and other consumer related energy
activities will increase significantly in Po-
land and could overshadow the benefits
associated with improved energy efficiency.
A comparison of the T-GAS and IPCC
results for the structural change scenario
shows that T-GAS estimates are again
tower than the results from the IPCC case
study. In the IPCC case study, total emis-
sions were estimated to be 231,000,000
tons of carbon. Although this is about 20%
higher than the T-GAS results, it is less of
a discrepancy than existed when compar-
ing the base case results, where a 35%
difference was identified.
Summary/Conclusions
The forecasting exercise for Poland
described above suggests that T-GAS is
capable of developing representative and
credible emission estimates up to at least
2030. Although the results obtained from
T-GAS were consistently tower than the
results from the IPCC model, the likely
reasons for the discrepancy were ex-
plained. In at least one case, the T-GAS
estimates may be more representative than
the IPCC case study assumptions (i.e., in
the base case, IPCC assumes no effi-
ciency improvements will occur, while T-
GAS allows a "business as usual" efficiency
improvement to occur). These results also
show that T-GAS is a useful analytic tool.
For example, an analysis of the results
from the structural change scenario showed
that even moderate increases in the over-
all economic activity of the country (i.e., as
measured by GDP) can overshadow the
emission reductions associated with a va-
riety of major technological improvements.
The results above also demonstrate
that T-GAS is capable of performing de-
tailed emissions mitigation evaluations for
individual countries. Recall that, in the struc-
tural change scenario, a complex set of
scenario assumptions were represented in
T-GAS which were intended to simulate
the effects of simultaneous changes in
technology efficiency, economic activity,
and industrial restructuring. Specifically,
technology change aimed at improving en-
ergy efficiency was simulated for several
sectors and subsectors (e.g., road trans-
portation, iron and steel production, chemi-
cal manufacturing, pulp and paper
production, light industry). Industrial restruc-
turing was simulated by adjusting the eco-
nomic growth rates of energy intensive
industries downward (e.g., iron and steel)
and adjusting the economic growth rates
of less energy intensive industries upward.
Million tons of oil equivalent.
Gross domestic product.
8
-------
40 -
Is 30 -
f* 20-
Uj
8 «,
1
*~ n .
g U
5 -20 -
1
c -30 -
-40 -
-50 -
y/vA'/,^^ — —^ —
'/j L^-£_n
1971
1973
1975
1977
1979
1981
1983
Figure 13. Percent difference between T-GAS and DOE emission estimates for Brazil from 1971 to 1984.
o:
UJ
50
40
30
20
-10
-20
-30
-40
-50
1962
/
/
1965
1968
1971
1974
1977
1980
1983
Figure 14. Percent difference between T-GAS and DOE emission estimates for Hungary from 1962 to 1985.
9
-------
1960
1965
1970
1975
1980
1985
Figure 15. Comparison of CO, emissions for various years from the four largest emission sources evaluated in the study.
260
^ 240-
•I 220-
.o 160-
| 140-
I 120-
I 1°°~
| *)-
? 60-
I ^>-
20-
0
West Germany
1960
1965
1970
1975
-t—i—i—i
1980
1985
Figure 16. Comparison of CO, emissions for various years from selected European countries.
10
-------
I
s
I
I
7955
7962
7966
1970
Figure 17. Comparison of CO2 emissions for various years from the four developing nations.
Total Energy Use in Million TOE
Total Emissions in Million Metric Tons of Carbon
1975
1977
1979
1981
1983
1985
Figure 18. Comparison of emissions and energy use data for Japan from 1973 to 1986.
11
-------
Q>
5
X
Ul
v
-— Total Energy Use in Billion TOE
Total Emissions in Billion Metric Tons of Carbon
1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985
Figure 19. Comparison of emissions and energy use data for the United States from 1961 to 1985.
200.0
I
1
I
O 7"GasGas
V TGas Oil
+ TGas Coal
X TGas Total
o.o
1958 1964 1970 1976 1982 1988 1994 2000 2006 2012 2018 2024 2030
Figure 20. T-GAS emissions estimates for Poland from 1958 to 2030 under the base case scenario.
12
-------
187.5
175.0
162.5
150.0
137.5
125.0
112.5
100.0
87.5
75.0
62.5
50.0
<3 37.5
25.0
12.5
0.0
1958 1964 1970 1976 1982 1988 1994 2000 2006 2012 2018 2024 2030
Figure 21. T-GAS emissions estimates for Poland from 1958 to 2030 under the structural change scenario.
13
•&U.S. GOVERNMENT PRINTING OFFICE: 1991 - 548-028/40066
-------
-------
-------
S. Pkxotand T. Lynch are with Alliance Technologies Corp., Chapel Hill, NC 27514:
R. Kaufmann and C. Cleveland are with Boston Univ., Boston, MA 02215; and B.
Moore is the the Univ. of New Hampshire, Durham, NH 03824.
Paul Jeffrey Chappell is the EPA Project Officer, (see below).
The complete report, entitled "Analysis of Historical Radiatively Important Trace Gases
(RITG) Emissions: Development of a Trace Gas Accounting System (T-GAS) for 14
Countries,' (Order No. PB91-216325/AS; Cost: $39.00, subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati, OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA PERMIT NO. G-35
Official Business
Penalty for Private Use $300
EPA/600/S9-91/019
------- |