1
Improving the Carbon Dioxide Emission Estimates from
the Combustion of Fossil Fuels in California
And
Spatial disaggregated estimate of energy- related carbon
dioxide for California
Principal Investigator:
Michael Hanemann
Prepared for the California Air Resources Board and the California Environmental
Protection Agency
Prepared by:
Stephane de la Rue du Can
Tom Wenzel
Lynn Price
Environmental Energy Technologies Division
Lawrence Berkeley National Laboratory
October, 2008
Contract # 05- 310 “ Improving the Carbon Dioxide Emission Estimates from the
Combustion of Fossil Fuels in California” and augmentation to contract number 05- 310
“ Spatial disaggregated estimate of energy- related carbon dioxide for California”
2
Disclaimer
The statements and conclusions in this Report are those of the contractor and not
necessarily those of the California Air Resources Board. The mention of commercial
products, their source, or their use in connection with material reported herein is not
to be construed as actual or implied endorsement of such products.
3
Acknowledgments
This work was supported by the California Air Resources Board through the U. S.
Department of Energy under Contract No. DE- AC02- 05CH11231. We would like to
thank Marc Fisher at Berkeley Lab and Nehzat Motallebi at California Air Resources
Board for their extremely helpful guidance throughout this project. We would like
also to thank a number of people for their assistance in providing and interpreting
data, including Andy Alexis, Kevin Eslinger, Glenn Gallagher, Ying Hsu, Larry
Hunsaker, Webster Tasat and Walter Wong ( California Air Resources Board);
Andrea Gough ( California Energy Commission); and Hendrik G. van Oss ( U. S.
Geological Survey). This Report was submitted in fulfillment of Contract # 05- 310
“ Improving the Carbon Dioxide Emission Estimates from the Combustion of Fossil
Fuels in California” and augmentation to contract number 05- 310 “ Spatial
disaggregated estimate of energy- related carbon dioxide for California” by UC
Berkeley under the sponsorship of the California Air Resources Board. Work was
completed as of October 2008.
4
Table of Contents
Abstract....................................................................................................................... ...... 8
Executive Summary.......................................................................................................... 9
A. Improving the Carbon Dioxide Emission Estimates from the Combustion of
Fossil Fuels in California................................................................................................ 12
1. Introduction ......................................................................................................... 12
2. Uncertainties by Sector ....................................................................................... 13
2.1 Electricity and CHP Sector .............................................................................. 13
2.2 Refinery Sector................................................................................................. 17
2.3 Oil and Gas Extraction Industries .................................................................... 21
2.4 Industry Feedstocks.......................................................................................... 23
2.5 Transportation .................................................................................................. 27
3. Uncertainties by Fuel .......................................................................................... 40
3.1 Reference versus Sectoral Approach................................................................ 40
3.2 Calorific Values and Carbon Emission Factors Uncertainties ......................... 43
4. Conclusion........................................................................................................... 49
5. Recommendations................................................................................................ 50
B. Spatial Disaggregation of CO2 Emissions for the State of California.............. 55
1. Introduction ......................................................................................................... 55
2. Methodology ........................................................................................................ 56
2.1 CO2 Emissions.................................................................................................. 56
2.2 Bottom- up versus Top- down Approach........................................................... 57
2.3 Geographical Boundary.................................................................................... 58
3. Overview.............................................................................................................. 59
4. Stationary source emissions ................................................................................ 64
4.1 Overview .......................................................................................................... 64
4.2 Natural Gas....................................................................................................... 66
4.3 Petroleum ......................................................................................................... 68
4.4 Coal .................................................................................................................. 71
5. Mobile Sources .................................................................................................... 72
5.1 On- road vehicles .............................................................................................. 72
5.2 Aviation............................................................................................................ 74
5.3 Rail ................................................................................................................... 78
5.4 Marine .............................................................................................................. 80
6. CO2 emissions in the South Coast Air Basin ....................................................... 81
6.1 Stationary sources ............................................................................................ 82
6.2 Mobile Sources................................................................................................. 83
7. CO2 emissions from electricity generation versus end- use ................................. 87
8. Conclusion........................................................................................................... 94
References..................................................................................................................... .. 95
List of Abbreviations and Acronyms .......................................................................... 101
Appendices..................................................................................................................... 103
5
List of Tables and Figures
Tables
Table ES 1. 2004 CO2 emissions from CALEB and percent uncertainty, by sector .......... 7
Table A- 1. Fossil Fuel Consumption for Electricity and Heat Generation by Industry
Type, 2004.................................................................................................................. 11
Table A- 2. Natural Gas Used for Useful Thermal Output................................................ 12
Table A- 3. Input to California Refineries in 2005 ( kbbl) ................................................. 14
Table A- 4. CEC Form M13 Report, 2005 ........................................................................ 14
Table A- 5. Natural Gas Consumption in Refineries......................................................... 17
Table A- 6. Oil and Gas Extraction Energy Use as Estimated in CALEB ........................ 19
Table A- 7. Use of Natural Gas in Oil and Gas Extraction ( Mcf) ..................................... 19
Table A- 8. Chemical Manufacturing Value of Shipments in California ( in millions of
dollars)....................................................................................................................... 21
Table A- 9. 2004 Natural Gas Consumption in Chemicals Plants in California ( Mcf) ..... 21
Table A- 10. Non- Energy Use of Fuel in 2000 ( TBtu)...................................................... 22
Table A- 11. Comparison of CARB CO2 emission estimates and SEDS fuel sales, for
water craft................................................................................................................... 36
Table A- 12. Reconciliation Errors by Energy Source in Trillion Btu .............................. 38
Table A- 13. California Coal Supply and Consumption ( kst) ........................................... 39
Table A- 14. 2000 CO2 Emissions from CALEB ( Mt CO2).............................................. 40
Table A- 15. Carbon Content Factors, Storage Factors and Fraction of Oxidation used in
CALEB....................................................................................................................... 41
Table A- 16. Ranking of CO2 Emissions from Fuel Combustion in 2004 ........................ 42
Table A- 17. Fuel use, CO2 emissions, and CO2 emission factors of ten largest California
electricity generating facilities in U. S. EPA CEM database ...................................... 45
Table A- 18. Percentage Uncertainties .............................................................................. 47
Table B- 1. IPCC main source categories.......................................................................... 53
Table B- 2. Methods used to allocate CO2 emissions to counties, by sector and fuel....... 55
Table B- 3. Comparison of CO2 emissions from CARB inventory and LBNL estimate, by
sector .......................................................................................................................... 57
Table B- 4. Impact of including domestic and international flights on California 2004 CO2
emission inventory ..................................................................................................... 71
Table B- 5. California airports by county .......................................................................... 72
Table B- 6. Allocation of 2004 California aircraft CO2 emissions to counties, by type of
flight ........................................................................................................................... 73
Table B- 7. California airports by air basin and county..................................................... 81
Table B- 8. CO2 emissions by aircraft, by air basins and type of flight ............................ 82
6
Figures
Figure A- 1. 2004 Carbon Dioxide Emissions from Fuel Combustion in California,
Million Metric Tons ( Mt) CO2 ................................................................................... 10
Figure A- 2. Other Hydrocarbons, Hydrogen and Oxygenates from U. S. EIA 810.......... 16
Figure A- 3. Comparison of gasoline use ( 2004 CARB) and sales ( 2007- 08 CalTrans) by
county, millions of gallons ......................................................................................... 27
Figure A- 4. Comparison of diesel fuel use ( 2004 CARB) and sales ( 2007- 08 CalTrans)
by county, millions of gallons .................................................................................... 28
Figure A- 5. Percent difference in gasoline use ( 2004 CARB) and sales ( 2007- 08
CalTrans) by county ................................................................................................... 29
Figure A- 6. Percent difference in diesel fuel use ( 2004 CARB) and sales ( 2007- 08
CalTrans) by county ................................................................................................... 30
Figure A- 7. Passenger- miles and CO2 emission rate of flights originating in California. 31
Figure A- 8. Fuel use of intrastate flights originating in California .................................. 32
Figure A- 9. Fuel use of interstate flights originating in California .................................. 32
Figure A- 10. Fuel use of international flights originating in California........................... 33
Figure A- 11. Comparison of bottom- up emissions inventory with California total jet fuel
sales ............................................................................................................................ 34
Figure A- 12. Comparison of 2004 US commercial aviation fuel use, from four sources 35
Figure A- 13. Trends in California transportation fuel sales and use, estimated by U. S.
EIA SEDS and reported by California Board of Equalization................................... 37
Figure A- 14. Distribution of passenger- miles on international flights, by originating state
and international destination ...................................................................................... 50
Figure B- 1. 2004 California CO2 emissions ( Mt) by fuel and sector ............................... 57
Figure B- 2. 2004 CO2 emissions by sector and county ................................................... 58
Figure B- 3. Sectoral distribution of 2004 CO2 emissions, by county.............................. 59
Figure B- 4. Per capita CO2 emissions ( tonnes per capita) by county.............................. 60
Figure B- 5. Per capita CO2 emissions from fossil fuel combustion, by county ............... 61
Figure B- 6. Absolute and per capita CO2 emissions by stationary sources, by fuel type
and county .................................................................................................................. 62
Figure B- 7. Absolute and per capita CO2 emissions from natural gas combustion by
stationary sources, by sector and county.................................................................... 64
Figure B- 8. CO2 emissions from petroleum product combustion by stationary sources, by
sector .......................................................................................................................... 65
Figure B- 9. Absolute and per capita CO2 emissions from petroleum product combustion
by stationary sources, by sector and county............................................................... 67
Figure B- 10. Absolute and per capita CO2 emissions from coal combustion by stationary
sources, by sector and county..................................................................................... 69
Figure B- 11. CO2 emissions from on- road vehicles, by county and vehicle type ........... 70
Figure B- 12. California CO2 emissions from on- road vehicles, by vehicle type ............ 70
Figure B- 13. CO2 emissions from aircraft, by county of origin and type of flight.......... 73
Figure B- 14. CO2 emissions from railroad activity, by county ....................................... 75
Figure B- 15. CO2 emissions from marine activity, by air basin...................................... 76
Figure B- 16. 2004 South Coast Air Basin CO2 emissions by fuel and sector.................. 78
7
Figure B- 17. SCAB CO2 emissions by stationary sources, by sector and fuel type......... 79
Figure B- 18. SCAB CO2 emissions from on- road vehicles, by vehicle type ................... 80
Figure B- 19. CO2 emissions by aircraft, by air basin of origin and type of flight............ 82
Figure B- 20. 2004 CO2 emissions from electricity generation, by county....................... 84
Figure B- 21. 2004 electricity generation, by fuel type and county .................................. 85
Figure B- 22. 2005 electricity consumption, by sector and county ................................... 86
Figure B- 23. 2004 fossil fuel electricity generation per capita, by fossil fuel type and
county......................................................................................................................... 89
Figure B- 24. 2005 electricity consumption per capita, by sector and county................... 90
8
Abstract
Central to any study of climate change is the development of an emission inventory that
identifies and quantifies the State’s primary anthropogenic sources and sinks of
greenhouse gas ( GHG) emissions. CO2 emissions from fossil fuel combustion accounted
for 80 percent of California GHG emissions ( CARB, 2007a). Even though these CO2
emissions are well characterized in the existing state inventory, there still exist significant
sources of uncertainties regarding their accuracy.
The first part of this report evaluates accounting for CO2 emissions based on the
California Energy Balance database ( CALEB) developed by Lawrence Berkeley National
Laboratory ( LBNL), in terms of what improvements are needed and where uncertainties
lie. The estimated uncertainty for total CO2 emissions ranges between - 21 and + 37
million metric tons ( Mt), or - 6% and + 11% of total CO2 emissions. The report also
identifies where improvements are needed for the upcoming updates of CALEB.
However, it is worth noting that the California Air Resources Board ( CARB) GHG
inventory did not use CALEB data for all combustion estimates. Therefore the range in
uncertainty estimated in this report does not apply to the CARB’s GHG inventory. As
much as possible, additional data sources used by CARB in the development of its GHG
inventory are summarized in this report for consideration in future updates to CALEB.
The second part of this report allocates California’s 2004 statewide CO2 emissions from
fuel combustion to the 58 counties in the state. The total emissions are allocated to
counties using several different methods, based on the availability of data for each sector.
The CO2 emissions data by county and source are described through figures, maps, and
graphs in this report.
9
Executive Summary
Central to any study of climate change is the development of an emission inventory that
identifies and quantifies the State’s primary anthropogenic sources and sinks of greenhouse
gas ( GHG) emissions. The accounting of carbon dioxide ( CO2) emissions from fossil
combustion, which represents the majority of GHG emissions in California, requires
having access to reliable and concise energy statistics. In 2005, Lawrence Berkeley
National Laboratory ( LBNL) evaluated several sources of California energy data,
primarily from the California Energy Commission and the U. S. Energy Information
Administration, to develop the California Energy Balance Database ( CALEB). This
database manages highly disaggregated data on energy supply, transformation, and end-use
consumption for each type of energy commodity from 1990 to the most recent year
available ( generally 2004) in the form of an energy balance. CARB used this database in
the development of its latest official inventory of greenhouse gas ( GHG) emissions for
the state of California ( CARB, 2007a). For some sources, CARB directly used estimates
on fuel use from CALEB; however, for other sources, CARB used their own estimates of
fuel use and CO2 emissions. CARB requested that LBNL undertake an assessment of
CALEB to highlight uncertainties and areas of future development of the database.
Futhermore, at CARB’s request, the original research contract for improving the
characterization of California’s CO2 emissions was augmented to develop a disaggregated
estimate of energy- related CO2 emissions. CO2 emissions are relatively well characterized
at the State level; however no estimates were available at a more disaggregated spatial
level. Understanding the CO2 emission profile, finding ways of validating these on a sector-by-
sector basis, and providing a validation approach to the statewide greenhouse gas
emission inventory ( EI) through disaggregation is an important service for building AB32
GHG EI baselines and projections.
Hence, two main research areas are investigated in this report. The first part of the report
focuses at the State level and describes uncertainties in using CALEB as a source for the
GHG State emissions inventory. The second part of the report describes a first attempt to
account for California CO2 emissions from fossil fuel combustion at the county level.
ES A. Improving the Carbon Dioxide Emission Estimates from the Combustion of
Fossil Fuels in California
The first part of this report evaluates accounting for CO2 emissions using the California
Energy Balance database ( CALEB), in terms of what improvements are needed and where
uncertainties lie. The key areas of uncertainty related to CO2 emissions include differences
between various data sets, estimates of bunker fuel consumption for international transport,
estimates of petroleum products used as feedstocks in refineries and chemical plants, and
estimates of the carbon content of the various fossil fuels combusted in California.
An attempt was made to quantify some of the uncertainties where a secondary data set
was available for comparison with data used in CALEB. Table ES 1 shows the
distribution of state CO2 emissions and rough estimates of their uncertainty by sector, for
the year 2004. In this report only in- state CO2 emissions from fuel combustion are
considered; other GHG and CO2 from electricity imports are excluded. CO2 emissions
from in- state electricity generation represent about 75% of total GHG emissions. A
10
positive percentage in the table indicates that the current estimate of CALEB CO2
emissions may be too low, while a negative percentage indicates that the current estimate
may be too high. The estimated uncertainty for total CO2 emissions ranges between - 19
and + 37 Mt, or - 5% and + 11% of total CO2 emissions.
Table ES 1. 2004 CO2 emissions from CALEB and percent uncertainty, by sector
2004 emissions
Category Estimated uncertainty
CO2 ( Mt) % CO2 ( Mt) % over each
category total
% over total
inventory
Electricity/ CHP* 62 18% 0.40 1% 0.1%
coal 4 1% 0.47 12% 0.1%
petroleum products 9 3% - 0.07 - 1% -
natural gas 49 14% - - -
Refining** 29 8% - - -
Oil/ gas extraction 14 4% 4.00 28% 1.1%
Industry feedstocks 1.8 1% ± 1.77 ± 100% ± 0.5%
Transportation 177 51% - 8.04 - 5% - 2.2%
On- road vehicles 167 48% - 7.17 - 4%
Gasoline 138 39% - 8.52 - 6% - 2.4 %
Diesel 29 8% 1.35 5% 0.4 %
Aviation 3 1% - 0.84 - 28% - 0.2 %
Marine 3 1% - 6% -
Rail 3 1% - 0.03 - 1% -
Other*** 66 19% - - -
Reconciliation errors - - - 6.2 to 13.0 - 2% to 4%
Emission Factors - - - 2.7 to 17.6 - 1% to 5%
Total 350 100% - 18.7 to 36.8 - 5% to 11%
* Combined Heat and Power ( CHP)
** Uncertainties with hydrogen production are not estimated
*** includes emissions from other sectors such as other industry, residential, commercial/ institutional,
agriculture/ forestry/ fishing/ fish farms and non- specified.
The table indicates that the largest uncertainties come from unresolved reconciliation
errors between supply and consumption data (- 2% to + 4%), carbon emission factor
uncertainties (- 1% to + 5%), gasoline use by motor vehicles ( 2%), and fuel use in
upstream (+ 1.1%) oil and gas operations. There also are small uncertainties in emissions
from fuel used as feedstock in chemical plants, fuel used in electric and Combined Heat
and Power ( CHP) plants, diesel used by motor vehicles, and fuel used for commercial
aviation.
The largest uncertainty lies in reconciling statistics on fuel supply and consumption;
available data do not match for most fuels. Many data gaps remain in accounting for total
energy flows in California, especially for petroleum products such as natural gas liquids
( NGLs), liquefied petroleum gas ( LPG), or still gas. The second largest uncertainty comes
from the use of national carbon emission factors as default factors, as no specific factors are
available for the state of California. In terms of sectors, the transport sector represents a
large source of uncertainty. Uncertainty in gasoline used by vehicles is estimated by
comparing results from a bottom- up emissions inventory model ( EMFAC) with total
11
gasoline sales. The representation of combined heat and power ( CHP) in the energy
balance needs to be improved by allocating all energy used for commercial and industrial
CHP to the sector where the generated electricity is used; all CHP energy use by facilities
whose primary business is to sell electricity and heat should be allocated to the electricity
generation sector. Finally, reported data on energy use in upstream oil and gas operations is
lacking, as reflected in the uncertainties in Table ES- 1.
Clearly understanding these uncertainties and developing new methodologies or data
collection activities to reduce them can significantly improve the characterization of
California’s CO2 emissions. We recommend that the California Air Resources Board
( CARB) conduct surveys on key industries where data are missing or unreliable, mostly
the refinery sector, the oil and gas industries and the chemical industries. Development of
bottom- up models to estimate CO2 emissions by sector would also help understand where
energy is ultimately used. We recommend collaboration with the U. S. Energy
Information Administration ( U. S. EIA) and U. S. Environmental Protection Agency ( U. S.
EPA), who collect data and develop methodologies at the national level, in order to
benefit from their work and experience. Finally, as the transport sector is such a large
source of CO2 emissions in California, further data collection is needed to better
understand the trends in activity in this sector.
ES B. Spatial Disaggregation of CO2 Emissions for the State of California
The second part of this report allocates California’s 2004 statewide CO2 emissions from
fuel combustion to the 58 counties in the state. Again, only in- state CO2 emissions from
fuel combustion are considered; other GHG and CO2 from electricity imports ( which
represent about one- quarter of total emissions from electricity generation) are excluded.
The total emissions are allocated to counties using several different methods, based on
the availability of data for each sector. Data on natural gas use in all sectors are available
by county. Fuel consumption by power and combined heat and power generation plants
is available for individual plants. Bottom- up models were used to distribute statewide
fuel sales- based CO2 emissions by county for on- road vehicles, aircraft, and watercraft.
All other sources of CO2 emissions were allocated to counties based on surrogates for
activity. CO2 emissions by sector were estimated for each county, as well as for the South
Coast Air Basin. It is important to note that emissions from some sources, notably
electricity generation, were allocated to counties based on where the emissions were
generated, rather than where the electricity was actually consumed. In addition, several
sources of CO2 emissions, such as electricity generated in and imported from other states
and international marine bunker fuels, were not included in the analysis. CARB does not
include CO2 emissions from interstate and international air travel in the official California
GHG inventory, so those emissions were allocated to counties for informational purposes
only. Los Angeles County is responsible for by far the largest CO2 emissions from
combustion in the state: 83 Mt, or 24% of total CO2 emissions in California, more than
twice that of the next county ( Kern, with 38 Mt, or 11% of statewide emissions). The
South Coast Air Basin accounts for 122 MtCO2, or 35% of all emissions from fuel
combustion in the state. The distribution of emissions by sector varies considerably by
county, with on- road motor vehicles dominating most counties, but large stationary
sources and rail travel dominating in other counties.
The CO2 emissions data by county and source are available in an excel workbook.
12
A. Improving the Carbon Dioxide Emission Estimates from the
Combustion of Fossil Fuels in California
1. Introduction
Analysts assessing energy policies and energy modelers forecasting future trends need to
have access to reliable and concise energy statistics. Lawrence Berkeley National
Laboratory ( LBNL) evaluated several sources of California energy data, primarily from
the California Energy Commission ( CEC) and the U. S. Energy Information
Administration ( U. S. EIA), to develop the California Energy Balance Database
( CALEB). This database manages highly disaggregated data on energy supply,
transformation, and end- use consumption for each type of energy commodity from 1990
to the most recent year available ( generally 2004) in the form of an energy balance,
following the methodology used by the International Energy Agency ( IEA). In addition
to displaying energy data, CALEB also calculates state- level energy- related carbon
dioxide ( CO2) emissions using the methodology of the Intergovernmental Panel on
Climate Change ( IPCC) ( Murtishaw et al., 2005).
The California Air Resource Board ( CARB) used the initial version of CALEB to
construct its official inventory of greenhouse gas ( GHG) emissions, published on line in
November 2007 ( CARB, 2007a). This report evaluates the areas where improvement to
CALEB is needed and assesses uncertainties associated with CO2 emissions accounting
from the CALEB database. The key areas of uncertainty related to CO2 emissions in
CALEB include differences between various data sets, estimates of bunker fuel
consumption for international transport, estimates of petroleum products used as
feedstocks in refineries and chemical plants, and estimates of the carbon content of the
various fossil fuels combusted in California. Clearly understanding these uncertainties
and developing new methodologies or data collection activities to reduce these
uncertainties can significantly improve the characterization of California’s fuel
consumption and CO2 emissions.
This report qualitatively estimates the level of uncertainty related to emissions from fuel
consumption in the CO2 emissions estimates based on the CALEB database, investigates
the development of new or improved methodologies for estimating the consumption of
specific fuels for which data are scarce or unreliable, and provides recommendations
regarding new data collection activities to improve the accuracy of fuel consumption and
CO2 emissions in California.
CO2 emissions from fuel combustion are the principal GHG emitted in California. In
2004, CO2 emissions from fuel combustion in California accounted for 80% of total
emissions ( CARB, 2007a). As fossil fuel is combusted, CO2 is emitted as a result of
oxidation of the carbon in the fuel. Figure A- 1 shows CO2 resulting from fuel combustion
in California from the California Inventory ( CARB, 2007a).
13
Figure A- 1. 2004 Carbon Dioxide Emissions from Fuel Combustion in California,
Million Metric Tons ( Mt) CO2
0 50 100 150 200
Electricity Generation ( 1A1ai)
Combined Heat and Power
Generation ( CHP) ( 1A1aii)
Other Energy Industries ( 1A1cii)
Petroleum Refining ( 1A1b)
Manufacturing Industries and
Construction ( 1A2)
Transport ( 1A3)
Commercial/ Institutional ( 1A4a)
Residential ( 1A4b)
Agriculture/ Forestry/ Fishing/ Fish
Farms ( 1A4c)
Non- Specified ( 1A5)
Mt
Nat Gas
Petroleum
Coal
Source: CARB, 2007a
Note: Code indicated in parentheses refers to IPCC category associated with the source of
emissions
Three energy commodities consumed in the economy produce CO2 emissions: natural
gas, oil, and coal. Figure A- 1 shows the relative importance of CO2 emissions by product
and sector. In California, the transport sector is by far the main source of CO2 emissions
resulting from fuel ( petroleum) combustion, followed by the electric and CHP sector.
However, it is worth noting that CO2 emissions related to electricity imports ( roughly
27% of supply) are not accounted for in this figure.
2. Uncertainties by Sector
2.1 Electricity and CHP Sector
The main purpose of an energy balance such as CALEB is to reconcile the supply and
eventual use of each energy product. The transformation sector, which includes the
energy used during the conversion of primary energy into secondary energy products,
represents one of the largest sectors in the energy balance. Electricity generation is
included in the transformation sector, where inputs of fuel are shown as negative values
and outputs of electricity are shown as positive values. In the case of combined heat and
power ( CHP) facilities, the quantity of fuel to produce electricity is shown in the
transformation sector while the quantity of fuel used to produce heat is shown in the
sectors where the heat is ultimately used, and not in the transformation sector. Therefore,
no data on heat output is shown in the transformation sector.
14
The electricity sector is disaggregated into five types of energy providers, following the
U. S. EIA classifications currently used in the Electric Power Annual publications and
data sets: utilities; integrated power producers ( IPPs); combined heat and power ( CHP),
electric power sector; CHP, industrial sector; and CHP, commercial sector. The category
“ CHP, electric power sector” includes facilities whose primary business is to sell
electricity, or electricity and heat, to the public; i. e. North American Industry
Classification System ( NAICS) category 22 plants. The data is shown by four fuel input
categories: coal, natural gas, other gases and total petroleum products.
2.1.1 Data Sources
In the CALEB database, data on fuel consumption by provider type come from the U. S.
EIA’s Electric Power Annual ( U. S. EIA, 2007). The U. S. EIA collects the information
through questionnaire EIA- 906 for electric power plants and EIA- 920 for CHP facilities.
Prior to 2004, the EIA- 906 form was also used to collect data from CHP plants. In
January 2004, a new form, the EIA- 920, was introduced to collect data from CHP plants
only. The reporting is mandatory for all power plants with a nameplate rating of 1 MW
and above that are connected to the electric grid1. Table A- 1 shows the data reported in
U. S. EIA’s Electric Power Annual and used in the CALEB database for 2004.
Table A- 1. Fossil Fuel Consumption for Electricity and Heat Generation by Industry
Type, 2004
( TBtu) Coal Petroleum Natural Gas Other Gases
Total Electric Power Industry 27 24 887 21
Electric Utilities 1 102
Independent Power Producers 13 455
CHP, Electric Power 22 8 173 1
CHP, Commercial Power 0 16
CHP, Industrial Power 5 2 142 20
Source: U. S. EIA, 2007
2.1.2 Uncertainties
There are mainly two shortcomings in the representation of the power sector and CHP in
the CALEB database.
Fuel Input Breakdown
One of the shortcomings of the current CALEB database is that it does not provide a
breakdown of fuel inputs beyond the four categories that are directly available from the
U. S. EIA’s Electric Power Annual ( i. e. coal, natural gas, other gases and petroleum
products). Disaggregated data by petroleum product ( distillate fuel oil, residual fuel oil,
petroleum coke, and waste and other oil) are available at the facility level for non- utility
plants on the U. S. EIA website, starting in 1998 only. This disaggregation could be
1 Beginning for reporting year 2007, the EIA- 906 and EIA- 920 forms were replaced by combined form
EIA- 923 “ Power Plant Operations Report.
15
integrated in future versions of CALEB. In the case of “ other” gases, defined as “ blast
furnace gas, propane gas, and other manufactured and waste gases derived from fossil
fuels”, no more detail is available. This lack of detail reduces the accuracy of calculating
CO2 on a product basis and also reduces the ability to balance each energy product
between supply and consumption, which is the essence of an energy balance. We propose
to disaggregate petroleum used by electricity generation/ CHP facilities by distillate fuel
oil, residual fuel oil, petroleum coke, and waste and other oil in future versions of
CALEB.
CHP representation
The second weakness of the CALEB database concerns the treatment of energy used
solely to produce heat in CHP plants. In CALEB, fuel used to generate electricity is
shown in the transformation sector, while fuel used to produce heat is shown in the end-use
sector where the heat is ultimately used ( commercial and industrial sectors).
In the case of natural gas, end- use data were taken from the CEC ( CEC, 2005) which do
not include input of natural gas for heat production from CHP plants. In order to adjust
for these quantities of natural gas consumed for the useful thermal output of CHP in the
end- use sectors, the amounts of natural gas used by individual CHP facilities solely to
generate heat were gathered from the U. S. EIA Form 906/ 920 Databases ( U. S. EIA,
2007b). However, these data are only available for non- utility facilities starting in 1998
( Table A- 2). Therefore, in CALEB, data for natural gas for useful thermal output ( UTO)
from CHP facilities from 1990 to 1997 are not included in the end- use sectors in which
the heat was ultimately used. This represents an omission of 4 to 9 Mt CO2, based on data
from the period 1998 to 2004 when data are available.
Table A- 2. Natural Gas Used for Useful Thermal Output
Unit 1998 1999 2000 2001 2002 2003 2004
MMcf 119,735 88,535 154,321 158,794 165,561 142,317 71,698
Mt CO2 6.63 4.90 8.54 8.79 9.17 7.88 3.97
Source: U. S. EIA, 2007b
Data on coal energy consumption comes from the U. S. EIA Annual Coal Report ( U. S.
EIA, 2005a) which includes all coal used by CHP facilities in three sectors: industrial,
commercial and electric power sectors. The U. S. EIA report does not distinguish whether
fuel inputs are used to generate electricity or heat. In CALEB, coal use to produce
electricity is reported in the transformation sector with data from the U. S. EIA Electric
Power Annual ( U. S. EIA, 2007a). Coal use in the end- use sector comes from the U. S.
EIA Annual Coal Report without adjusting for coal use to produce electricity. Therefore
the data on final consumption includes coal use in industrial CHP facilities to produce
electricity, which is already accounted in the transformation sector, and excludes coal use
in NAICS category 22 CHP facilities to produce heat, which is included in the electric
power sector in the U. S. EIA Annual Coal Report. As coal from industrial CHP to
produce electricity is larger than coal used by NAICS category 22 CHP plants use to
produce heat, CALEB is overestimating coal consumption in the final sector by 206
thousand of short ton of coal, which represents 0.47 Mt CO2 in 2004. Over the year, the
difference ranges by month from 0.14 Mt CO2 to 0.71 Mt CO2.
16
In the case of petroleum products, data for final consumption in CALEB comes from
diverse sources. For distillate fuel oil and residual fuel oil, data come from U. S. EIA’s
“ Sales of Fuel Oil and Kerosene” report ( U. S. EIA, 2007c). Energy use for commercial
and industrial CHP facilities is also reported in the commercial and industrial sectors,
while the electric power sector includes energy used by NAICS category 22 CHP plants.
For petroleum coke, CALEB only reports final energy use consumption from cement
plants ( USGS, 2007), and includes all energy use by CHP plants. Petroleum coke is also
used by refineries for their own use, which is reported in the energy sector in CALEB.
Overall, the reconciliation of many different data sources to represent a full picture of
energy use in the power sector and in the end- use sectors has lead to some uncertainties
in understanding what exactly is included in each sector. Residual fuel oil, distillate oil
and coal used for electricity production from industrial and commercial CHP facilities are
overestimated, as quantities used to produce electricity are accounted for in both the
power sector and the end use sector. On the other hand residual fuel oil, distillate oil and
coal used for heat production by NAICS category 22 CHP facilities are not included in
either the power sector or the end use sectors. Finally, in the case of natural gas, data
before 1998 does not account for energy use for UTO production in the end use sectors.
2.1.3 Alternative Sources/ Methods and Recommendations
The representation of CHP in an energy balance is a complex matter, as attention needs to
be taken to ensure that no double- counting occurs. In the CALEB database, more
evaluation of each data point for each energy product type in each subsector needs to be
carried out. Uncertainties lie in the accounting of CHP as part of the end use sectors or as
part of the power sector for the energy used for heat and for electricity production.
In the future, we recommend that all the energy used by industrial and commercial CHP
facilities be included in the appropriate end use sectors. This is consistent with the 2006
IPCC guidelines on GHG inventories. Moreover, all energy used by CHP NAICS
category 22 facilities will be included in the transformation sector, with fuel input shown
as a negative value, and electricity and heat output shown as a positive value. This
adjustment to CALEB will also require that data on heat output by end use be collected,
to indicate where the heat produced by CHP NAICS category 22 plants is ultimately
consumed.
Furthermore, we recommend collaborating with the U. S. EIA team that processes the
U. S. EIA Annual Power database. Several attempts were made to obtain data before
1998 on natural gas consumption by individual non- utility facilities, but with no success.
Also, data by fuel type can potentially be obtained by the U. S. EIA. For its latest
inventory, CARB obtained the most detailed data from U. S. EIA, via a special data
request. We hope that to obtain the same detailed data in the future to update the CALEB
database.
Overall, we estimated that the uncertainties with data used in CALEB may underestimate
CO2 emissions from coal used by 0.47Mt of CO2 ( 0.1% of total CO2 emissions) and
17
overestimate CO2 emissions from oil by 0.07Mt of CO2 ( negligible compared to total
CO2)
2.2 Refinery Sector
CO2 emissions from refineries originate from three main sources: fuel combustion,
fugitive sources and industrial processes. Fugitive emissions are broadly defined as all
GHG emissions from oil and gas systems except from fuel combustion ( IPCC, 2006).
Industrial process emissions occur from production processes where CO2 is a by- product
of chemical reactions. Estimates of the uncertainty of fugitive and industrial process
emissions are outside the scope of this report.
2.2.1 Data Sources
Fuels used in refineries are shown in the transformation and energy sectors of CALEB.
The transformation sector shows inputs of crude oil, unfinished oil and additives2 as
negative numbers, and outputs of each petroleum product as positive numbers. Input and
output data are from the CEC ( Yearly Input and Output at Refineries, CEC 2006a)
reported through form U. S. EIA 810. Table A- 3 shows fuel inputs to refineries. When
calculating CO2 emissions, the transformation of crude oil and feedstocks into petroleum
products does not involve combustion, so no CO2 emissions from fuel input are
accounted for in CALEB. However, this process does result in fugitive CO2 emissions.
Table A- 3. Input to California Refineries in 2005 ( kbbl)
Inputs kbbl
Crude Oil 672,032
Butane 1,729
Isobutane 2,380
Other Hydrocarbons, Hydrogen and Oxygenates 10,718
Unfinished Oils 27,191
Source: CEC 2006a
The energy sector shows the consumption of energy needed to operate refineries. In
CALEB, this is shown in the sub- category “ Energy Sector: Own Use” and data for
refineries come from the CEC Ca Petroleum Industry Information Reporting Act lifornia
Refinery Monthly Fuel Use Report Form M13 ( CEC, 2006b).
Table A- 4 shows data reported in M13 for 2005. Fuels used in this category were
assumed to be entirely combusted.
Table A- 4. CEC Form M13 Report, 2005
2 Additives includes the category called “ Other hydrocarbons, hydrogen and Oxygenates” from EIA 810.
18
Description
Distillate Fuel Oil, Used As Refinery Fuel kbbl 155
Liquefied Petroleum Gases, Used As Refinery Fuel kbbl 1,706
Natural Gas, Used As Refinery Fuel MCf. 132,707
Still Gas, Used As Refinery Fuel kbbl 40,795
Marketable Petroleum Coke, Used As Refinery Fuel kbbl 1,660
Catalyst Petroleum Coke, Used As Refinery Fuel kbbl 11,675
Purchased Electricity GWh 3,107
Purchased Steam k LBS 12,508
Other Fuel Used at Refinery 1 Varies 4
Source: CEC, 2006b
2.2.2 Uncertainties
One of the main uncertainties when collecting energy use for the refinery sector is the
determination of how much energy is used for different purposes. CO2 emissions are
estimated differently if the quantity of fuel used is consumed for its heating value or for
its chemical proprieties, i. e. whether it is burned or used as a feedstock for the production
of other products.
Refinery Fuel Input
Crude oil intake into California refineries was taken from aggregated numbers from the
Petroleum Industry Information Reporting Act ( PIIRA) database provided by the CEC
( Yearly Input and Output at Refineries, CEC 2006a). Another Energy Commission data
set ( Oil Supply Sources to California Refineries, CEC 2006c) provides alternate figures
for crude oil receipts by source. Those figures tend to be from 1% to 3% higher than the
figures reported in the Yearly Input and Output at Refineries report. For the year 2005 for
example, the Yearly Input and Output at Refineries report shows 672,032 kbbl of crude
oil intake while the Oil Supply Sources to California Refineries report shows 674,276
kbbl.
Data on butane, isobutene, other hydrocarbons and unfinished oils ( see Table A- 3), as
well as specific petroleum products, are provided by the Energy Commission based on
the U. S. EIA report 810 submissions ( Yearly Input and Output at Refineries, CEC
2006a). Due to the complexity of the refining industry, some products are reported as
both input and output. In order to avoid double counting, LBNL subtracted the reported
outputs from inputs so that only net inputs are shown. However, no specific information
is available to differentiate inputs that are used in the refining process from feedstocks
used to produce hydrogen ( see next section). Also, no conversion factor or carbon content
is provided or detailed information that described these inputs to allow the use of precise
energy conversion and carbon content factors.
Fuel Use for Industrial Process - Hydrogen Feedstocks
The production of hydrogen in California is growing rapidly as it allows oil refineries to
meet limits on sulfur content in refined fuels. Because most of the refineries are
switching to heavier crude oil, increasing amounts of hydrogen are needed to strip the
19
sulfur and to crack the hydrocarbons. Demand is met by own production from refineries
and also by independent industrial hydrogen plants ( Ritchey, 2006). The production of
hydrogen results in CO2 emissions from a chemical reaction. Feedstocks used in
California to produce hydrogen include natural gas, LPG, naphtha, and refinery fuel gas.
Emissions associated with hydrogen production for use in refining activities needs to be
included in refinery activities and not in the petrochemical manufacture sector. Care
should be taken to ensure that the feedstock for the hydrogen plant is not also reported as
fuel combustion, and vice versa.
Inputs of fuel in refineries, reported by the CEC ( CEC 2006a) includes a category called
“ Other Hydrocarbons, Hydrogen and Oxygenates” which is defined as followed:
“ Other Hydrocarbons, Hydrogen and Oxygenates: Materials received by a
refinery and consumed as a raw material. Includes hydrogen, coal tar derivatives,
gilsonite, oxygenates and natural gas received by the refinery for reforming into
hydrogen. Natural gas to be used as fuel is excluded.” ( U. S. EIA Form 810)
These quantities are reported as input to refineries in CALEB and are shown under the
product category “ Additives”. However, data reported over time in this category is
decreasing, which is going against the observed trend of increasing hydrogen production.
Figure A- 2 shows the time series for the category Other Hydrocarbons, Hydrogen and
Oxygenates.
Figure A- 2. Other Hydrocarbons, Hydrogen and Oxygenates from U. S. EIA 810
Other Hydrocarbons, Hydrogen and Oxygenates
0
10,000
20,000
30,000
40,000
50,000
60,000
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Thousands of Barrels
More information is needed to differentiate the type of feedstock used in the refinery
sector. Hydrogen feedstock and production needs to be clearly stated, as estimation of
CO2 emissions will differ depending on whether natural gas, refinery fuel gas, LPG or
naphtha is used as the feedstock.
Fuel Combusted
20
A significant portion of the energy products in a refinery is used for process energy. Fuel
use reported by refineries to the CEC in form M13 ( CEC, 2006b) was assumed to
represent the fuel used for the energy production process and entirely combusted.
The instructions for the M13 refinery questionnaire are limited3 and a better
understanding of the coverage of fuel reported in this data set is needed. The accounting
of fuel use in the production of hydrogen is a major uncertainty. It is not clear if form
M13 includes fuel use by hydrogen plants for energy purposes. Moreover, a growing
number of independent hydrogen merchants are producing hydrogen outside refinery
facilities. The amount of energy used by these industries is unknown.
Uncertainties concerning fuel used by refineries also includes the use of conversion
factors. Since refinery fuel gas is a highly variable source of CO2 emissions across
refineries, a conversion factor specific to California refineries needs to be calculated.
Similarly, petroleum coke is provided under two different items: marketable petroleum
coke and catalyst petroleum coke; however no specific energy and carbon factors are
available to better account for these products.
Finally, consumption of natural gas by refineries is also available from a different source:
the CEC collects data from utilities on natural consumption disaggregated by SIC/ NAICS
codes ( CEC, 2005). Table A- 5 shows data from the CEC M13 and the CEC SIC/ NAICS
code. Data from the two sources differ over time. According to experts, some of the
difference is explained by the fact that the CEC M13 not only includes pipeline quality
natural gas, but also lease fuel gas or associated gas. A better understanding of what each
category accounts for is needed. In CALEB, data from M13 is reported in the energy
sector and the difference, when data from the CEC SIC/ NAICS are higher, is reported as
input to refineries.
Table A- 5. Natural Gas Consumption in Refineries
Mcf Source 1990 1995 2000 2004
Petroleum and Coal
Products Manufac.
CEC SIC-NAICS
80,035 103,475 148,134 136,061
Refinery Fuel M13 91,972 89,402 121,401 129,338
Combined Heat and Power ( CHP) Plants
As mentioned earlier, little is known on the fuel use reported by CEC M13 from the
instruction form that complements the data collection. Hence, concerns were raised that
CALEB was double- counting fuel consumption in refinery CHP facilities in cases where
CEC M13 forms were including this energy use. CALEB already reports energy use for
electricity production in CHP in the electricity sub- sector with data reported by the U. S.
EIA Annual Power database ( U. S. EIA, 2007a).
3 CEC- M13 Instructions:
“ The CEC Form M13 is used to collect data on fuel, electricity, and steam consumed for all purposes at the
refinery. Refiners in the state of California are required to file this report.”
21
However, during their work on the inventory, CARB staff determined that the CEC M13
form does not include fuel used in CHP.
2.2.3 Alternative Sources/ Methods and Recommendations
In its latest inventory, ARB used data obtained from the Journal of Oil & Gas to estimate
the amount of hydrogen generated by refineries each year. From this, they back-calculated
the CO2 released and estimated the fuel input needed ( natural gas, refinery gas,
naphtha or residual oil) to generate this hydrogen. Access to these data would help LBNL
would improve their estimate; LBNL intends to follow the same methodology when it
updates the CALEB database.
However, the issue remains as some refineries report natural gas used in hydrogen
production in the CEC M13 data set. With increasing production and use of hydrogen, it
is becoming necessary to collect data that allow for the accounting of process emissions
associated with hydrogen production, as well as to make sure that energy used for energy
purposes are included in CALEB. In the future, mandatory reporting from refineries will
resolve these issues.
In this report, we did not estimate uncertainties with hydrogen production as too little
information is available. In future versions of CALEB, the potential of using data from
the Journal of Oil & Gas will be assessed4 as well as the possibility of using mandatory
reporting from refineries in future years,
2.3 Oil and Gas Extraction Industries
2.3.1 Data Sources
Oil and gas extraction energy use covers the energy used for pumping and processing
crude oil as well as extraction of natural gas and natural gas liquids ( NGL). In California,
the quantities of energy used for oil and gas extraction tend to be exceptionally high due
to the use of thermally enhanced oil recovery process ( TEOR). TEOR uses large amounts
of natural gas to heat crude oil to render it less viscous. Natural gas use for oil and gas
extraction grew from 190 Bcf in 1990 to 295 Bcf in 2001 ( Murtishaw, 2005).
Main data sources in CALEB:
Natural gas consumption is taken from the CEC disaggregated data on natural gas
consumption by SIC/ NAICS code ( NAICS category 211 and 213) ( CEC, 2005) to which
was added data on CHP fuel input to produce heat5 from U. S. EIA 906/ 920 compiled at
the facility level for the years 1996 to 2004 ( U. S. EIA, 2007b).
4 We have inquired in the past about the possibility of obtaining data from the Journal of Oil & Gas, but
were refrained by the cost. However, as it seems to be the only publicly available source of data on
hydrogen production, we will work with CARB and the journal staff to get these data for future CALEB
updates.
5 In CALEB, the energy use for electricity production in CHP is shown under the electricity sub- sector in
the transformation sector while the energy use for heat production appears in the end use sectors directly.
22
Petroleum Products: data from the U. S. EIA Annual Fuel Oil and Kerosene Report6
( U. S. EIA, 2007c) were used, subtracting the value obtained by the M13 form on refinery
fuel use already accounted for under the category “ refinery”. The U. S. EIA Annual Fuel
Oil and Kerosene Report publishes statistics on distillate fuel, residual fuel and kerosene
fuel oil used by each oil company, defined as the company's own use for operations in
drilling equipment, use at the refinery, exploration company, oil drilling company, and
pipeline company, but excluding feedstocks.
Table A- 6 shows the energy used in oil and gas extraction sector as estimated in CALEB.
Table A- 6. Oil and Gas Extraction Energy Use as Estimated in CALEB
Unit 1990 2000 2004
Distillate Fuel Oil kbbl 493 233 297
Fuel Oil kbbl 27 0 0
Natural Gas Bcf 191 297 267
Note: 1990 do not include natural gas for producing heat from CHP, in 2000 and 2004, these
amounts to 19 and 13 Bcf respectively.
2.3.2 Uncertainties
No comprehensive data set showing all fuel types used for oil and gas extraction is
collected at the state or national level. Hence CALEB gathers data from several different
sources, increasing the risk of coverage issues. This is a particularly important issue as a
considerable amount of energy is used for TEOR in California. A review of the CALEB
data for oil and gas operations in a Western States Petroleum Association ( WSPA) Memo
to CARB ( Lev- On, 2007) indicates omissions of crude oil and associated gas consumed
at upstream operations for steam generation and other combustion needs. According to
this memo, emissions from the use of crude oil not captured in the CALEB database
contributed up to 4 Mt CO2 in 1990, but appear negligible for 2000 and 2005. Emissions
from the combustion of associated gases not captured in the CALEB database may
contribute up to 4 Mt CO2 for 2004.
2.3.3 Alternative Sources/ Methods and Recommendations
Natural Gas
Alternative data on natural gas consumption is available from the U. S. EIA Natural Gas
Navigator database ( 2008). Table A- 7 shows natural gas used for processing oil and gas
in California from the U. S. EIA Natural Gas Navigator database. These data were not
included in CALEB to avoid double- counting with CEC disaggregated data on natural
gas consumption by SIC/ NAICS code ( code category 211 and 213), which provides
much higher numbers. In 2004, the CEC data shows 267 Bcf natural gas used in oil and
gas extraction, while the U. S. EIA shows only 62.5 Bcf ( Table A- 7).
Table A- 7. Use of Natural Gas in Oil and Gas Extraction ( Mcf)
6 Energy Information Administration, Form EIA- 821, " Annual Fuel Oil and Kerosene Sales Report"
23
2004 2005 2006 Definitions
Re- pressuring 22,405 29,134 29,001 Injection of gas into oil or gas reservoir
Lease Fuel
Consumption
37,337 37,865 33,211 Natural gas used in well, field, and lease operations, such
as gas used in drilling operations, heaters, dehydrators,
and field compressors.
Gas Plant Fuel
Consumption
2,760 2,875 2,475 Natural gas used as fuel in natural gas processing plants.
Total 62,502 69,874 64,687
Source: U. S. EIA Natural Gas Navigator ( U. S. EIA, 2008a)
More information is needed to understand how natural gas use by oil and gas companies
is reported in the CEC data set. In the case of oil and extraction, consumption of natural
gas can be injected to re- pressure oil or gas reservoir formations, or burned to produce
steam that will serve to liquefy the heavy crude oil extracted. This implies different CO2
emissions accounting.
Associated Gas, Crude Oil and Distillates
NGLs consumption in CALEB includes input to refineries under the transformation
sector, based on data from the CEC ( CEC 2006a) and data on industrial consumption
from API ( API, 2002). However, considerable statistical difference exists between NGL
supply and demand, with consumption and/ or exports totaling much less than production.
This was noted in the 2005 CALEB report as an area for future improvement ( Murtishaw,
2005). One possible source of NGL consumption is the use of NGL directly by oil
companies in their oil and gas extraction processes.
In its inventory, CARB uses data from the Division of Oil, Gas, and Geothermal
Resources ( DOGGR) of the California Department of Conservation to determine how
much crude oil, lease fuel and distillate are used in this sector. For years prior to 2001,
when DOGGR data were not available for lease fuel use, U. S. EIA data were used, as
recommended by the DOGGR.
Emissions from the combustion of associated gases not captured in the CALEB database
may contribute up to an additional 4 Mt CO2 for 2004.
2.4 Industry Feedstocks
Some of the fuel supplied to an economy is used as raw material ( or feedstock) for the
manufacture of products such as plastics and fertilizer. In some cases, the carbon from the
fuels is oxidized quickly to CO2; in other cases, the carbon is stored ( or sequestered) in
the product, sometimes for as long as centuries. Hence, this use of energy products has a
different accounting methodology in terms of carbon emissions. The carbon balance for
non- energy use is complex. The amount of carbon stored is calculated by multiplying the
potential emissions of each fuel used as a feedstock by a fuel specific storage factor. This
requires collecting information on both the energy use and non- energy use of fuel in the
chemical industry, as well as collecting data on the type of chemicals produced to
determine the storage factors.
24
The chemical industry is an important part of the California economy that has increased
at an annual average growth rate ( AAGR) of 7.5% from 1997 to 2006 ( Table A- 8). The
California chemical industry includes a very wide mix of products. The dominant
chemical sub- sector in California is pharmaceuticals, representing 62% of shipments in
the California chemical industry in 2006, with an average annual growth rate of nearly
13% since 1997.
Table A- 8. Chemical Manufacturing Value of Shipments in California ( in millions
of dollars)
NAICS 1997 2006 AAGR
325 Chemical mfg 19,303 36,922 7.5%
3251 Basic chemical mfg 2,664 2,621 - 0.2%
3252 Resin, syn rubber, & artificial syn fibers &
filaments mfg
1,100 1,414 2.8%
3253 Pesticide, fertilizer, & other agricultural
chemical mfg
502 840 5.9%
3254 Pharmaceutical & medicine mfg 8,006 23,075 12.5%
3255 Paint, coating, & adhesive mfg 2,272 3,218 3.9%
3256 Soap, cleaning compound, & toilet preparation
mfg
2,965 3,733 2.6%
3259 Other chemical product & preparation mfg 1,794 2,019 1.3%
Source: US Census, 2006;
Most of the chemical manufacturing in California consists of industrial gas production
( hydrogen, nitrogen, oxygen, argon), dyes and pigments, and other basic inorganic
chemical manufacturing, which includes products such as bleach, borax, sulfuric acid,
plating materials, high temperature carbons and graphite products and catalysts ( Galitsky
and Worrell, 2005).
2.4.1 Data Sources
Natural Gas
• Energy Use Chemical Industries: the CEC maintains a database on natural gas
consumption at three different levels of aggregation ( CEC, 2005). The most detailed data
are at the 3- to 4- digit NAICS category level. These values do not include the shares of
natural gas used for CHP- generated heat, which were added from the U. S. EIA 906/ 920
database ( U. S. EIA, 2007b) as explained in Section 2.1. Table A- 9 shows natural gas
consumption in the chemical industry at the 4th digit level.
Table A- 9. 2004 Natural Gas Consumption in Chemicals Plants in California ( Mcf)
Category NAICS 4 digit Category Source Mcf
3251 Basic Chemical Manufacturing CEC 4,617
25
3252 Resin, Synthetic Rubber, and Artificial Synthetic Fibers CEC 1,023
3253 Pesticide, Fertilizer, and Other Agricultural Chemical Manufacturing CEC 752
3254 Pharmaceutical and Medicine Manufacturing CEC 3,700
3255 Paint, Coating, and Adhesive Manufacturing CEC 324
3256 Soap, Cleaning Compound, and Toilet Preparation Manufacturing CEC 391
3259 Other Chemical Product and Preparation Manufacturing CEC 384
NS Heat production in CHP U. S.
EIA
1,495
NS: Not Specified
Source: CEC, 2005; U. S. EIA, 2007b
Non- Energy Use: The portion of natural gas that is used as feedstock is unknown.
However, these data are available at the national level from U. S. EPA National US
Inventory ( U. S. EPA, 2008). In order to estimate the portion that was used in California,
we calculated that California accounts for 3% of the total US shipments of basic chemical
and fertilizer products in 2001, and applied this share to the total natural gas used for
non- energy use in the US chemical industry. As a result we estimate that 10.2 TBtu of
natural gas were used as feedstocks in producing basic chemical and fertilizer products in
California in 2001. The share of natural gas used as feedstock to total natural gas used in
the chemical industry was then calculated ( 47%) and applied to other years. Table A- 10
summarizes our estimates for non- energy use of fuel in the chemical industry for
California for 2000.
Table A- 10. Non- Energy Use of Fuel in 2000 ( TBtu)
Natural Gas LPG Petrochemical
feedstocks
Chemicals and Allied Products 25 13 11
of which used as feedstoks 12 13 11
Storage Factors 91% 91% 54%
• Carbon Stored: the storage factor for natural gas ( 91%) comes from the inventory
of California greenhouse gases and sinks ( CEC, 2002), which is higher than the national
storage factor ( 67%).
Petroleum Product
• Energy Use in Chemical Industries: data for LPG and petrochemical feedstock
consumption by end- use sector were taken largely from State Energy Data System
( SEDS, U. S. EIA, 2007d), since it provides a comprehensive set of data for ten categories
of petroleum products. However no breakdown by sub- sector is available. Moreover, as
SEDS only provides data with a four- year delay, different sources were used for more
recent years. For LPG, consumption estimates were provided by the U. S. EIA
( Lindstrom, 2008) which are based on data from the American Petroleum Institute ( API).
Data on petrochemical feedstock consumption were taken from SEDS and assumed to be
entirely consumed in the chemical industry sub- sectors. When data were not available for
recent years, we estimated consumption based on the same principle used in SEDS:
allocating the total US consumption to the states according to the value- added of their
organic chemical industries.
26
• Non- energy Use: we assumed LPG and petrochemical feedstocks to be entirely
consumed for non- energy purposes.
• Carbon Stored: the storage factor for LPG ( 91%) and for petrochemical
feedstocks ( 54%) came from the inventory of California greenhouse gases and sinks
( CEC, 2002). The storage factor for LPG is higher from the national storage factor
( 66%), while the storage factor for the petrochemical feedstock is lower than the national
storage factor ( 66%).
2.4.2 Uncertainties
CO2 emissions from the chemical industry represent 0.5% of the total CO2 emissions in
California. However, the chemical industry in California accounted for 8.2% of industry
natural gas consumption and 17% of industry petroleum product consumption.
Complex Accounting
There is no easy method to estimate CO2 emissions for the chemical industry. The
chemical industry is a very complex industry that produces a wide range of products. It is
divided into seven broad categories under NAICS category 325, which are further broken
down into multiple subcategories that include over 1,000 products. The basic chemical
industry is the most energy- intensive segment, and also the most diverse, within the
chemical industry. This industry sector alone accounts for nearly 50% of the chemical
sector's total energy use in California. In many instances basic chemicals are utilized as
inputs in the production process of other industries.
The difficulties in gathering data are many. First, data on energy consumption by fuel
type need to be available by industrial subsector. This is the case for natural gas, but not
for other petroleum products. Second, data on the share of this energy use needs to be
broken down further to define the quantity used as feedstock to the chemical process, as
opposed to the quantity of fuel combusted. Finally, depending on the type of chemical
produced, a percentage of the fuel used as feedstock will be stored in the product or
emitted. This percentage also needs to be estimated.
Lack of Information
Uncertainties relating to the CO2 emissions from energy use in the chemical industry
come principally from a lack of available data. First, data on energy use by industrial
subsectors is only available for natural gas. Second, the share of the energy use for non-energy
purposes, i. e. as feedstock, is not available. Finally, production of the different
chemical outputs produced is not available, which makes it difficult to estimate the
storage factors. Import and export of feedstocks to the state are also crucial.
At the national level, the Manufacturing Energy Consumption Survey ( MECS, U. S. EIA,
2005b) collects data on energy use at the sub- sectoral level. The survey also specifically
requests participants to report on energy used for purposes other than for heat, power, and
electricity generation ( feedstocks). MECS provides this information only for four
27
regions7, and not at the state level, and with an increasing level of data withheld for
confidentiality reasons.
The Annual Survey of Manufacturers ( U. S. Census, 2005) provides information about the
quantities of chemicals produced, but only at the national level. This allows the
assessment of the types of chemicals produced in the US, for which carbon storage is
calculated.
Storage factors
The CEC calculated storage factors for California in 1999; however, neither the time nor
the resources were available to conduct a thorough survey. Moreover, this was the first
attempt to conduct an inventory for the state and many other issues were also at stake.
The U. S. EPA national inventory calculates annually a single aggregate storage factor for
eight fuel feedstocks. For 2006, the storage factor was 62%, meaning that 62% of the net
non- fuel use was destined for long- term storage in products, while 38% was emitted to
the atmosphere directly as CO2 ( U. S. EPA, 2008). The approach to estimate this factor is
based on identifying the commodities derived from petrochemical feedstocks, and
calculating the net import/ export for each.
A similar approach needs to be done for California in order to improve CO2 emissions
accounting for the state. However, this requires access to data that currently are not
collected.
2.4.3 Alternative Sources/ Methods and Recommendations
The need for data on energy use in the chemical industry, on energy use as feedstock, on
quantity of chemical output produced, and on feedstock trade movement, is essential to
improve the accounting of CO2 emissions for the chemical industry.
A survey of the major chemical plants in California involved in the production of
chemical material that require feedstocks would be a beneficial input. It would help
provide data on the quantity of energy used as feedstock and the major chemical outputs
produced.
We estimate the uncertainty of all feedstocks combined as 1.8Mt of CO2, or 0.5% of total
CO2. This number corresponds to the total CO2 emissions from natural gas. LPG and
petroleum feedstocks used in the chemical industry, without including energy use for
CHP. Data are not available to estimate California specific energy use and storage factors
for individual feedstocks.
2.5 Transportation
Transportation is the major source of CO2 emissions in California, with on- road vehicles
representing 94% of all transportation emissions. The estimation of CO2 emissions from
mobile sources is challenging, as fuel sales are very decentralized and end users are
mobile rather than stationary sources.
7 Northeast, Midwest, South and West; the West region includes California.
28
2.5.1 Data Sources
We used U. S. EIA State Energy Data System ( U. S. EIA, 2007d) data for California fuel
sales by fuel type. U. S. EIA uses several state- level data series to allocate total national
product supplied, reported in Petroleum Supply Annual ( U. S. EIA, 2008b), to the states.
U. S. EIA conducts three surveys to track the monthly sale of petroleum- based fuels: EIA-
782A, a survey of all ( 100) refiners and gas plant operators; EIA- 782B, a survey of a
sample ( 27,000) of fuel resellers and retailers; and EIA- 782C, a survey of all ( 170) prime
suppliers that produce, import or transport a refined petroleum product across state
borders. Data from all three surveys are reported at the state level in U. S. EIA’s
Petroleum Market Annual series ( U. S. EIA, 2008c).
The volumes reported nationally and for each state vary among the three surveys for
several reasons: EIA- 782A reports sales at the point of production, whereas EIA- 783C
reflects sales at the point of likely consumption. Therefore, states with major refining
operations, such as California, have higher reported sales in EIA- 782A ( at the point of
production) than in EIA- 782C ( at the point of eventual use). In addition, EIA- 782C also
includes fuel imports by firms that are neither refiners nor gas plant operators; such
imports are not included in volumes reported in EIA- 782A ( U. S. EIA, 2008c).
The fuel sales reported by prime suppliers ( EIA- 782C) is substantially lower than total
product supplied ( EIA- 782A), for a variety of reasons. For example, the prime supplier
data does not include sales of bonded jet fuel for international flights. Also, to the extent
that airlines import their own jet fuel, the prime supplier sales would not capture those
sales since an airline is not considered a prime supplier. In addition, diesel fuel may get
' winterized' by adding jet fuel later down the supply chain before a sale. As a result, the
product supplied data would classify the product as jet fuel whereas the prime supplier
would report it as diesel fuel ( Heppner, 2008). In SEDS the total national product
supplied ( EIA- 782A) is allocated to states using the detailed state level data from fuel
resellers and retailers ( EIA- 782B) and prime suppliers ( EIA- 782C).
U. S. EIA further disaggregates total annual sales by end use. In SEDS, motor gasoline
and distillate ( diesel) fuel used for on- road vehicles is allocated to states using Highway
Statistics Table MF- 21 ( FHWA, 2007), which is based on state reported fuel tax receipts.
Jet fuel is allocated to the states using Petroleum Marketing Annual ( PMA) sales by
prime suppliers ( EIA- 782C), which is reported by state. Diesel fuel used for railroads
and vessel bunkering, and residual fuel used for vessel bunkering, are allocated to states
using EIA- 821 " Annual Fuel Oil and Kerosene Sales Report”. EIA 821 is a mandatory
reporting questionnaire sent to companies that sell fuel oil and kerosene to gather
information on quantity sold to the different end uses.
According to IPCC guidelines, fuels consumed for international maritime shipping as
well as international aviation should be excluded from national inventories ( IPCC, 1996).
However, in the IEA energy balance format, aviation fuels consumed for both
international flights and domestic flights are also reported as separate items. Murtishaw
et al. ( 2005) describes the methodology used to estimate this breakdown of marine and
air transportation to intrastate, interstate, and international destinations. About 95% of
California’s 2000 transport- sector residual fuel consumption is allocated as international
marine bunker fuel. For the remaining 5% of 2000 transport- sector residual fuel, 3.5%
29
was used by interstate marine shipping, while only 1.5% was consumed by intrastate
marine shipping. Distillate fuel use by ocean- going vessels was estimated by applying a
ratio of 0.06 gallons of distillate fuel for every gallon of residual fuel used, resulting in an
estimate of 2.2 million barrels of distillate used by ocean- going vessels. We applied the
same interstate and intrastate breakdown for ocean- going vessels that we used for
residual fuel, resulting in 2.1 million barrels distillate fuel for international, 0.07 million
barrels for interstate, and 0.03 million barrels for intrastate shipping by ocean- going
vessels. Based on U. S. EIA data, there were an additional 1.6 million barrels of distillate
fuel used by non- ocean- going ( i. e. commercial harbor craft and personal recreational)
vessels, which we allocated to intrastate shipping. Of the distillate fuel consumed by all
marine vessels, we estimated that 55% were consumed by international marine activity,
43% by intrastate activity, and the remaining 2% by interstate activity.
Concerning air transport, CALEB estimated that 39.9% of California’s 2000 jet fuel
consumption was for international flights, 52.7% was for interstate flights, and 7.4% was
for intrastate flights, using the EEA’s aircraft movement methodology ( Murtishaw et al.,
2005; EEA, 2004).
2.5.2 Uncertainties
One method to assess the accuracy of the estimates of fuel use by transport sector is to
estimate fuel use using a sectoral, or bottom- up approach, where the number of vehicles
and miles traveled are multiplied by a CO2 emission factor to obtain total CO2 emissions.
CARB has already developed such models for on- road vehicles and watercraft; we
developed a similar simple bottom- up model for aviation fuel use. In this section we
compare fuel use reported in SEDS with bottom- up estimates of fuel use by each major
transport mode.
On- road vehicles
CARB’s EMFAC mobile source emission modeling system combines tailpipe emission
rates, activity data, and vehicle population data to estimate CO2 emissions from on- road
vehicles by vehicle type and county ( Eslinger, 2008). CARB used these model outputs to
allocate CO2 emissions from total fuel sales reported to the Bureau of Equalization in the
official GHG inventory. The 2004 reported sales of gasoline for use by on- road vehicles
in 2004 were 5.8% lower than modeled using EMFAC, while sales of diesel fuel were
5.3% higher than modeled using EMFAC.
CARB staff recently compared EMFAC’s estimate of statewide CO2 emissions and
gasoline use with that from the CalCARS model developed by the CEC ( CARB, 2004).
The analysis found that, for the entire light- duty vehicle fleet, the EMFAC model
estimated 6% greater gasoline use in 2000 and 4% greater in 2002 than the CalCARS
model. While the two models are in good agreement for the entire vehicle fleet, fuel use
by individual model years can vary greatly. For instance, the EMFAC model estimated
17% lower gasoline use for model year 2000 vehicles in 2002 than the CalCARS model.
CARB should update this analysis using more recent output from the revised EMFAC
and CalCARS models.
30
The California Department of Transportation ( CalTrans) also has developed estimates of
vehicle gasoline and diesel fuel use by county, using the Motor Vehicle Stock, Travel and
Fuel Forecast ( MVSTAFF) model. MVSTAFF allocates estimated vehicle miles traveled
and fuel consumption to counties based on VMT on state highways from the Traffic
Accident Surveillance and Analysis System ( TASAS) file, and VMT on all other public
roads from the Highway Performance Monitoring System ( HPMS, CalTrans 2006).
Figure A- 3 through Figure A- 6 compare 2005 data on fuel sales by county from
CalTrans’ MVSTAFF model with 2004 fuel use by county from CARB’s EMFAC
model. Figure A- 3 and Figure A- 4 show the absolute fuel use and sales, where each
point represents a county; Figure A- 5 and Figure A- 6 show the percent difference
between the two estimates, by county. Statewide gasoline sales estimated by CalTrans are
8% lower than statewide gasoline use estimated by CARB; on the other hand, statewide
diesel sales estimated by CalTrans are 10% higher than diesel use estimated by CARB.
Figure A- 3. Comparison of gasoline use ( 2004 CARB) and sales ( 2007- 08 CalTrans)
by county, millions of gallons
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500
2004 gasoline use ( CARB EMFAC)
2005 gasoline sales ( CalTrans MVSTAFF)
Statewide, CalTrans sales are 8% lower than CARB use
Los Angeles County
31
Figure A- 4. Comparison of diesel fuel use ( 2004 CARB) and sales ( 2007- 08
CalTrans) by county, millions of gallons
0
100
200
300
400
500
600
0 100 200 300 400 500 600
2004 diesel fuel use ( CARB EMFAC)
2005 diesel fuel sales ( CalTrans MVSTAFF)
Statewide, CalTrans sales are 10% higher than CARB use
Los Angeles County
Note in Figure A- 5 and Figure A- 6 that the four counties with the greatest gasoline use
( according to CARB; Los Angeles, San Diego, Orange, Riverside, shown in pink in
Figure A- 5), which account for half of all gasoline use, all have lower gasoline sales
estimated by CalTransthan gasoline use estimated by CARB. Six of the ten counties with
the greatest diesel use ( according to CARB; Los Angeles, San Bernardino, Riverside, San
Diego, Orange, San Joaquin, shown in pink in Figure A- 6), which account for half of all
diesel use, all have higher diesel sales estimated by CalTrans than use estimated by
CARB.
32
Figure A- 5. Percent difference in gasoline use ( 2004 CARB) and sales ( 2007- 08
CalTrans) by county
- 40% - 20% 0% 20% 40% 60% 80% 100%
Alameda
Alpine
Amador
Butte
Calaveras
Colusa
Contra Costa
Del Norte
El Dorado
Fresno
Glenn
Humboldt
Imperial
Inyo
Kern
Kings
Lake
Lassen
Los Angeles
Madera
Marin
Mariposa
Mendocino
Merced
Modoc
Mono
Monterey
Napa
Nevada
Orange
Placer
Plumas
Riverside
Sacramento
San Benito
San
San Diego
San
San Joaquin
San Luis
San Mateo
Santa
Santa Clara
Santa Cruz
Shasta
Sierra
Siskiyou
Solano
Sonoma
Stanislaus
Sutter
Tehama
Trinity
Tulare
Tuolumne
Ventura
Yolo
Yuba
CalTrans sales CalTrans sales greater than CARB use
less than CARB use
335%
211%
100%
120%
Statewide CalTrans
sales are 8% less
than CARB use
33
Figure A- 6. Percent difference in diesel fuel use ( 2004 CARB) and sales ( 2007- 08
CalTrans) by county
- 60% - 40% - 20% 0% 20% 40% 60% 80% 100% 120% 140% 160% 180%
Alameda
Alpine
Amador
Butte
Calaveras
Colusa
Contra Costa
Del Norte
El Dorado
Fresno
Glenn
Humboldt
Imperial
Inyo
Kern
Kings
Lake
Lassen
Los Angeles
Madera
Marin
Mariposa
Mendocino
Merced
Modoc
Mono
Monterey
Napa
Nevada
Orange
Placer
Plumas
Riverside
Sacramento
San Benito
San
San Diego
San
San Joaquin
San Luis
San Mateo
Santa
Santa Clara
Santa Cruz
Shasta
Sierra
Siskiyou
Solano
Sonoma
Stanislaus
Sutter
Tehama
Trinity
Tulare
Tuolumne
Ventura
Yolo
Yuba
CalTrans sales CalTrans sales greater than CARB use
less than CARB use
379%
338%
392%
219%
Statewide CalTrans
sales are 10% more
than CARB use
Aviation
LBNL has developed a bottom- up model of the fuel used by commercial aircraft taking
off from California airports for the year 2000 ( Murtishaw et al., 2005). In this report, we
extended the calculation for the period 1990 to 2006. The model uses the U. S. Bureau of
Transportation Statistics Air Carriers: T- 100 Segment data sets from 1990 to 2006 for
detailed information on flights and passenger- miles by origin/ destination and aircraft
type, and average fuel intensity by aircraft type and flight distance from European
Environment Agency’s EMEP/ CORINAIR Emission Inventory Guidebook ( EEA 2006).
34
The model was used in Murtishaw et al., 2005 to allocate total jet fuel sales to intrastate,
interstate, and international flights originating in California.
Figure A- 7 shows the trend in passenger- miles reported by the U. S. Bureau of
Transportation Statistics ( BTS) and CO2 emission rate ( per passenger- mile) calculated by
LBNL, of all flights originating in California from 1990 to 2006. Passenger- miles
increased dramatically between 1990 and 2000, nearly doubling in that ten- year period.
Passenger- miles declined in 2001 through 2003, likely due to the aftermath of the
terrorist attacks on September 11, 2001. However, passenger- miles began to increase
again in 2004. In general the CO2 emission rate has decreased during this period, with
the exception of 2001 to 2003. Note that passenger- miles are used to calculate the
emission rate, even though 13% of all California aviation CO2 emissions in 2003 are
attributable to flights with no passengers ( rather they are flights for transporting freight
and mail).
Figure A- 7. Passenger- miles and CO2 emission rate of flights originating in
California
0
50
100
150
200
250
300
1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
Year
Passenger- miles ( billions)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
lbs CO2 per passenger- mile
Passenger- miles
lbs CO2 per p- m
Figure A- 8 through Figure A- 10 show the trend in fuel use for intrastate ( California),
interstate ( US domestic), and international flights originating in California. Note that for
earlier years the EEA report does not have fuel factors for some older aircraft types; the
fraction of all passenger- miles flown by aircraft for which fuel factors are not provided
are shown in red in each figure. Historically, fuel use grew fastest for international
flights; however, international flights were also most affected by the terrorist attacks in
2001. Since 2001, fuel use has grown at a similar rate for intrastate, domestic and
international flights.
35
Figure A- 8. Fuel use of intrastate flights originating in California
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
Year
Fuel use ( million metric tonnes)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Percent of miles where fuel use of flight is
unknown
Fuel use
Fuel info missing
Figure A- 9. Fuel use of interstate flights originating in California
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
Year
Fuel use ( million metric tonnes)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Percent of passenger- miles where fuel use of flight is
unknown
Fuel use
Fuel info missing
36
Figure A- 10. Fuel use of international flights originating in California
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
Year
Fuel use ( million metric tonnes)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Percent of passenger- miles where fuel use of
flight is unknown Fuel use
Fuel info missing
Figure A- 11 compares the LBNL bottom- up inventory of fuel use from aviation in
California with reported jet fuel sales in California, from SEDS 2007. The figure
indicates that our bottom- up inventory substantially under- estimates jet fuel use, by 34%
in 2004 and up to 50% in earlier years. The figure also indicates that jet fuel sales ( in
red) waver from year to year, while estimated fuel use ( in blue) increased consistently in
most years ( except for 2001 and 2002, following the terrorist attacks).
One source of error in our estimate is the miles by flight segment reported in the BTS air
travel data; these are clearly air route distances between airports, rather than the distances
actually flown. One study has found that route changes and aircraft circling because of
delays ( referred to as “ uplift”) can add an additional 9% to 10% to flight distances
( EUROCONTROL 1992). Assuming an additional 10% of fuel use from uplift in our
bottom- up inventory reduces the gap between our inventory and SEDS to 28%.
37
Figure A- 11. Comparison of bottom- up emissions inventory with California total jet
fuel sales
0
2
4
6
8
10
12
14
16
1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
Year
Fuel use ( million metric tonnes)
Bottom- up inventory commercial aviation
Total jet fuel sales
The US Federal Aviation Administration has developed SAGE, a more sophisticated
model to estimate fuel use by commercial aircraft ( FAA 2005a). SAGE has been used to
estimate fuel consumption by the country in which the flight originated ( FAA 2005b).
Figure A- 12 compares 2004 commercial aviation fuel use for the US from SAGE and
from the LBNL model. The figure indicates that the LBNL model estimates 5% more
total jet fuel use than SAGE, even though fuel use is not estimated for aircraft accounting
for 10% of the flight miles in the LBNL model, and SAGE accounts for uplift and the
LBNL model does not. Correcting both of these factors would increase the LBNL
estimate, possibly by as much as 20%. The figure also indicates that the LBNL model
understates the fraction of fuel use from domestic flights ( in blue), and overstates the
fraction from international flights ( in green), relative to the SAGE estimate. Finally,
Figure A- 12 compares the two bottom- up estimates with U. S. EIA prime supplier and
total supplied jet fuel use in SEDS ( in pink). SEDS reports 25 million gallons of national
jet fuel sales in 2004, 17% higher than the SAGE estimate and 11% higher than the
LBNL estimate. The SEDS estimate to total jet fuel supplied is 20% higher than the
prime supplier fuel sales, which excludes jet fuel imported by airlines.
38
Figure A- 12. Comparison of 2004 US commercial aviation fuel use, from four
sources
14,796
12,347
20,909
25,074
6,576 10,068
0
5,000
10,000
15,000
20,000
25,000
FAA inventory
( SAGE)
LBNL inventory Jet fuel sales ( EIA
PMA)
Jet fuel sales ( EIA
SEDS)
Source
Jet fuel ( million gallons)
International flights
Domestic flights
Marine
CARB has developed bottom- up inventories of CO2 emissions from ocean- going vessels
( 3.1 Mt CO2, CARB, 2005) and harbor craft ( 1.2 Mt CO2, CARB, 2007b). Emissions
from ocean- going vessels are estimated from 0 to 24 nautical miles ( 2.3 Mt CO2), and 24
to 100 nautical miles ( 0.8 Mt CO2), off the coast of California; in its official inventory
CARB includes only emissions up to 100 nautical miles, but reports an additional 11.1
Mt CO2 from international bunker fuels used beyond 100 nautical miles.
We compared the CO2 emissions from the combustion of residual fuel oil and distillate
fuel in ocean vessels and harbor craft, as estimated in the CARB inventory, with the 2004
fuel sales, as estimated in SEDS. Table A- 11 indicates that the CARB inventory
estimates greater CO2 emissions from water craft using distillate fuel than SEDS. The
table also suggests that the CARB inventory estimates less CO2 emissions than SEDS
from combustion of residual fuel oil from international marine travel. However, this
could be an accounting issue, as the CARB inventory includes 1.1 million metric tonnes
of CO2 emissions from international marine vessel port activitives and transit while in
California waters in its “ other” category and total emissions from combustion of residual
fuel oil are identical in the inventory and in SEDS.
The CARB inventory reports CO2 emissions from international ships traveling beyond
100 nautical miles of California’s coast, based on the SEDS estimate of sales of
international bunker fuels. However, it is clear that these numbers do not account for the
total CO2 emissions from international ships using California’s ports. CARB plans to
39
develop in the future an estimate of all CO2 emissions from interstate and international
marine traffic using California ports ( Alexis, 2008).
Table A- 11. Comparison of CARB CO2 emission estimates and SEDS fuel sales, for
water craft
Trip type 2004 CO2 emissions( Mt)
( included/ excluded
in CARB inventory)
Fuel
CARB
inventory
SEDS
fuel use
Difference
International Residual fuel oil 11.1 12.5 12%
( excluded) Distillate fuel 0.0 0.6 NA
Other* Residual fuel oil 2.0 0.7 - 67%
( included) Distillate fuel 1.3 0.5 - 60%
Total Residual fuel oil 13.1 13.1 0%
Distillate fuel 1.3 1.0 - 23%
Combined 14.4 13.6 - 6%
* includes port activities and transit in California waters of intrastate, interstate, and
international marine travel, as well as harbor craft.
Rail
In 1991 Booz- Allen & Hamilton developed a 1987 bottom- up inventory of criteria
pollutant emissions for CARB ( CARB, 1991). This inventory estimated 141 million
gallons of diesel fuel use by locomotives for five different service types: intermodal
freight, mixed freight, short haul, yard operations, and passenger transport. CARB
updated this inventory in 2006 ( CARB, 2006); the updated inventory estimates 306
million gallons of diesel fuel used by locomotives ( CARB, 2007a).
The official CARB greenhouse gas inventory uses SEDS estimates of 226 million gallons
of diesel fuel ( and 348 million scf of natural gas) for locomotives in 1990, and 310
million gallons of diesel ( 280 million scf of natural gas) in 2004. Therefore CARB’s
bottom- up inventory estimates 1% less diesel fuel use for locomotives in 2004 than the
official inventory based on SEDS estimates.
2.5.3 Alternative Sources/ Methods and Recommendations
We contacted the California Energy Commission and inquired about the PIIRA database.
PIIRA requires qualifying petroleum industry companies to submit weekly, monthly, and
annual data to the California Energy Commission. Data collection began in 1982. In
2006, the PIIRA regulations were amended to increase the frequency and level of detail
in the information reported by the industry. Specifically, the A15 survey collects data on
fuel sales by retail outlet. About 80% of outlets have provided data in the first year of the
survey; however, these data are not yet available for analysis ( Schremp, 2008).
We also contacted the Board of Equalization and downloaded data from their website
( CBE, 2008). However, two problems were identified with the fuel tax data. First, gallons
sold are reported by fiscal, not calendar year. Data for some of the later years are
reported by month, so we could recreate calendar year sales; however, monthly data are
not available for years before 2000. Another issue is total vs. taxable gallons; while all
40
motor gasoline sold is taxed, only about 90% of diesel fuel, and a small percentage of jet
fuel, is taxed, and therefore included in the Board of Equalization estimates ( see Figure
A- 13; it is not clear why SEDS reports much higher diesel fuel sales in 2003).
Figure A- 13. Trends in California transportation fuel sales and use, estimated by
U. S. EIA SEDS and reported by California Board of Equalization
0
2000
4000
6000
8000
10000
12000
14000
16000
1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
Motor gasoline ( millions of gallons)
0
1000
2000
3000
4000
5000
6000
Jet fuel and diesel ( millions of gallons) Motor gasoline
Diesel fuel
Jet fuel
filled symbols and solid lines are EIA SEDS;
open symbols/ dashed lines are CA BOE
Figure A- 13 indicates that California’s estimates of motor gasoline sales from tax receipts
closely match those estimated in SEDS. Trends for diesel fuel sales also track SEDS
estimates fairly well, although a portion of diesel fuel sales are exempt from taxation.
However, because most jet fuel sold in California is exempt from tax, California data on
jet fuel tax receipts cannot be used to estimate total jet fuel use in the state.
3. Uncertainties by Fuel
3.1 Reference versus Sectoral Approach
The CO2 emissions from fuel combustion can be calculated by one of two methods: the
reference approach or Tier 1 and the sectoral approach or Tier 2 ( IPCC, 1996; Murtishaw
et al., 2005). The reference approach is a “ top- down” which focuses on estimating the
emissions from the carbon content of fuels supplied to or sold in a jurisdiction. The
reference approach assumes that all fuel reported as “ supplied to the economy” is
combusted ( adjusting for known non- energy uses). The sectoral is a “ bottom- up”,
approach that calculates CO2 emissions at the source where fuel is ultimately combusted,
using actual end- use consumption data or estimates of activity multiplied by energy
intensity factors. For verification purposes, IPCC recommends reporting results of their
calculations using both approaches, and to explain differences between estimates under
the two approaches.
41
CALEB displays a “ total consumption” energy flow, which for each fuel type is the sum
of all end- use consumption of energy, use for transformation, own use of energy in the
energy sector, transformation losses, and distribution losses. In theory, these totals should
match the total amount supplied, but since supply, transformation, and end use data are
collected and reported separately, the totals rarely balance precisely. Thus, reconciliation
errors, which the International Energy Agency ( IEA) calls “ statistical differences”, refer
to the difference between total supply of any given fuel and the total consumption of that
fuel for transformative and end use consumption. This expresses the unresolved
discrepancies between the supply, transformation, and end use consumption figures.
The energy balance constructed in 2005 for the year 2000 shows the reconciliation error
for every energy product supplied and consumed in California ( Murtishaw et al., 2005).
Table A- 12 shows in Tbtu the reconciliation errors for every fuel. The table also shows
the percent of total consumption that the fuel represents in total fuel consumed and the
percent reconciliation error between the quantity supplied and consumed to the total
amount of fuel consumed. For example, in 2000, natural gas consumption represents
40.3% of total fuel consumption, the reconciliation error between consumption and
supply is 225 TBtu ( consumption is 225 TBtu greater than supply), which represents
8.9% of total natural gas consumption. The net reconciliation error in CALEB is 21 TBtu,
which represents about 0.3% of total energy consumption ( 6,227 TBtu).
Table A- 12. Reconciliation Errors by Energy Source in Trillion Btu
Product Percent of
total
consumption
Difference between
supplied and consumption
( reconciliation error)
Difference as percent
of total product
consumption
Nat Gas 40.3% 225 8.9%
NGL 0.4% - 6 - 22.7%
Additives 2.5% 21 13.6%
Crude - - 17 - 0.5%
Tot Pet. Products 55.7% - 86 - 2.5%
Still Gas 3.3% - 42 - 20.3%
LPG 0.9% - 18 - 31.5%
Motor Gas 27.8% - 61 - 3.5%
Aviation Gas 0.1% - 1 - 27.4%
Jet Fuel 9.3% 0 0.0%
Kerosene 0.0% - 1 - 49.3%
Dist Fuel 8.8% 0 0.0%
Res Fuel 0.2% 0 0.0%
Pet Coke 1.7% 0 0.0%
Lubricants 0.5% - 22 - 70.8%
Asphalt 2.0% 25 20.0%
Waxes 0.1% - 2 - 54.6%
Special Naphtha 0.1% 3 34.4%
Petrochem feedstocks 0.2% - 4 - 34.9%
Other Petro Prods 0.8% - 14 - 27.4%
Coal 1.1% - 65 - 90.2%
Net reconciliation error 21 0.3%
Total Consumption 100% 6,227
42
3.1.1 Data Sources
Tracking energy consumption for all end uses and fuel types used in California is a
difficult task. It requires collecting information from multiple sources and assessing data
gaps. The report Development of Energy Balances for the State of California ( Murtishaw
et al., 2005) describes in detail the different sources of data used to construct the energy
balance table above.
3.1.2 Uncertainties
Overall, the reconciliation errors are comparable to those found for many countries in the
IEA data ( IEA, 2003a; IEA, 2003b). However, individual reconciliation errors by fuel
can be substantial.
Coal: Prior to 2003, substantial reconciliation errors exist, where supply is much higher
than end use consumption. The reconciliation error ranges from 4% in 2003 to - 64% in
2001 ( Table A- 13). At this point, it is unclear what explains such large differences, as all
data come from the same source, U. S. EIA.
Table A- 13. California Coal Supply and Consumption ( kst)
Source 2000 2001 2002 2003 2004 2005
Import U. S. EIA, 2006 5,691 7,881 6,543 2,762 3,001 2,726
Stock U. S. EIA, 2006 61 - 54 - 1 46 - 33 NA
Total Consumption SEDS, 2007d 2,954 2,834 2,943 2,866 2,847 2,849
Statistical differences Cons- Supply - 2,737 - 5,047 - 3,600 104 - 154 123
Reconciliation Error % - 48% - 64% - 55% 4% - 5% 4%
Natural Gas: reconciliation errors of natural gas range from - 199 Bscf in 2004 to 238
Bscf in 2000, which represent - 9% to 10% of total natural gas supplied to California. The
smallest reconciliation error, for 2002, is 4 Bscf, representing only 0.2% of total natural
gas supplied to California. The use of several sources of data to account for natural gas
supplied and used in California could account for these differences. The primary source
for all natural gas supply data is the U. S. EIA’s Natural Gas Navigator ( U. S. EIA, 2008),
while consumption mainly comes from the CEC ( CEC, 2005). Consumption of natural
gas data are also available through the U. S. EIA’s Natural Gas Navigator database, but
with less detail. Moreover, U. S. EIA’s consumption data are 2% lower than CALEB
consumption data in 2001, and 11% higher than CALEB data in 2004.
Petroleum Products: data on consumption of petroleum products in the state is the most
challenging to gather, because there are about 20 different types of products in use and
the distribution system is managed by many operators, rather than a few large utilities.
Table A- 12 shows the 2000 statistical differences for every petroleum product. Kerosene,
lubricants, asphalt, waxes, special naphtha, and petrochemical feedstocks all have
substantial statistical differences but each product only represents a small share of the
total energy consumption in California.
Comparing supply with consumption is a meaningful way of assessing data coverage.
However, neither the supply data nor the consumption data are complete for all fuel
consumed in California. For example, no data were available on trade of some petroleum
43
products, such as LPG, NGL, jet kerosene, etc. Statistics on movement of petroleum
products between states does not exist for every product and may be cumbersome to
collect. This highlights the difficulty of tracking energy flows in California.
3.1.3 Alternative Source/ Methods and Recommendations
Improved accounting of fuel supplied and used in California is needed to narrow the
differences shown in Table A- 12. This is a challenging task as many fuel products enter and
exit the state without being reported. The recent amendment of the PIIRA database to increase
the frequency and level of detail in the information reported by the industry will help in
improving the reconciliation between supply and consumption. The U. S. EIA conducts about
76 surveys with different time frames, from weekly to every four years. A list of such surveys
is provided in Appendix A. Some of the data gathered through these surveys are available at the
state level, such as Annual Refinery Report ( U. S. EIA- 820) which is also processed by the
CEC. These data were used in CALEB. The CEC has ongoing work with staff at the U. S. EIA
to gather more of the information collected through these surveys. A next step would be to
collaborate further with the U. S. EIA and assess if more data could be obtained from the data
reported to the state or estimated to the state level by U. S. EIA.
We estimated that uncertainties associated with reconciliation errors due to data gap range
from - 6Mt CO2 to 13Mt CO2 ( Table A- 14). These results are based on CALEB database
for 2000 data.
Table A- 14. 2000 CO2 Emissions from CALEB ( Mt CO2)
Nat Gas Petroleum Coal Total
Reference Approach 119 219 13 350
difference 13 4 - 6 11
Sectoral Approach 132 223 7 361
3.2 Calorific Values and Carbon Emission Factors Uncertainties
3.2.1 Data Sources
Energy balances use a common energy unit to allow comparison and balancing between
flows and products. However, data are usually collected in physical units, such as volume
or mass. Conversion from physical units to energy units is determined by the quality of a
product, and can vary between regions, over time, and by uses. SEDS ( U. S. EIA, 2007d)
provides detailed annual conversion factors for California for natural gas and coal, and
distinguishes between their heating value depending upon whether the fuel is used in the
electricity sector, the industry sector, or in other sectors. Conversion factors for
petroleum products are generally considered constant over time and uses. The U. S. EIA’s
annual U. S. average conversion factor for liquefied petroleum gas ( LPG), which reflects
the quantity- weighted average of their components that may fluctuate over time, is used
in CALEB. For motor gasoline, CALEB uses an annual California- specific conversion
factor calculated by the Energy Commission ( Bemis, 2004).
Once an energy balance has been constructed, CO2 emissions resulting from fossil fuel
combustion can be calculated. CALEB has been designed to calculate CO2 emissions from
44
energy consumption. According to IPCC, conversion of fuel combustion to CO2 emissions
requires three types of carbon factors: ( 1) emission factors, ( 2) storage factors, and ( 3)
oxidation factors ( IPCC, 1996). Carbon emission factors convert the fuel consumed into the
maximum amount of carbon that can be released in the atmosphere during combustion. U. S.
average emission factors are used in CALEB ( U. S. EPA, 2005). Carbon storage factors are
applied to the share of carbon stored when consuming fuel for non- energy purposes, as
explained in Section 2.4. Non- energy uses also include asphalt and road oil use for road
construction, as well as waxes and lubricants that are used directly for their chemical
proprieties and are not combusted. The storage factors for asphalt, waxes and lubricants were
taken from the California GHG inventory ( CEC, 2002). Finally, carbon oxidation factors are
the proportion of carbon in fuel that is oxidized to CO2 during combustion. A small
proportion of carbon is stored in solids such as ash and soot arising from incomplete
combustion of carbon in fuel. Average international values from the IPCC are used for those
factors ( IPCC, 1996).
The first column in Table A- 15 shows the carbon factors that have been used in the
calculation of carbon emissions from fuel combustion in CALEB. All of these factors were
taken from U. S. EPA ( U. S. EPA, 2008) except for the energy commodity “ additives” and
“ petrochemical feedstocks”. For the former, we used the same emissions factor and
oxidized fraction as crude oil. Petrochemical feedstocks are composed of two products:
naphtha and other oils, which have different emission factors. The production of each of
these products is available from the annual CEC reports on refinery operations ( CEC,
2005). Hence, the share of each product was used to calculate an average emission factor.
Table A- 15. Carbon Content Factors, Storage Factors and Fraction of Oxidation
used in CALEB
Carbon Coefficient Storage Factor Fraction Oxidized
Unit kgC/ MMBtu % %
Natural Gas 14.47 91% 99.5%
Still Gas 17.51 - 99.5%
LPG 16.98 * 91% 99%
Motor Gas 19.34 * - 99%
Aviation Gas 18.87 - 99%
Jet Fuel 19.33 * - 99%
Kerosene 19.73 - 99%
Distillate Fuel 19.96 - 99%
Residual Fuel 21.50 - 99%
Pet Coke 27.85 - 99%
Lubricants 20.23 50% 99%
Asphalt 20.64 100% 99%
Waxes 19.81 100% 99%
Special Naphtha 19.86 0% 99%
Petrochemical feedstocks 19.87 * 51% * 99%
Other Petro Prods 20.23 * 10% 99%
NGL 18.24 80% 99.5%
Coal 25.76 - 98%
Crude Oil 20.23 * - 99%
Mustishaw et al., 2005; * vary annually ( factors presented are for 2000)
45
3.2.2 Uncertainties
The heating value and carbon content of some fuels varies across time and across region.
Uncertainties with the carbon content of gasoline are discussed first because
approximately half of all California CO2 emissions from fossil fuel combustion are
associated with motor gasoline consumption ( Table A- 16). Uncertainties with carbon
content of natural gas are provided next, as about 40% of California greenhouse gas
emissions from fossil fuel combustion are attributable to natural gas consumption.
Finally, carbon contents of coal and petroleum products are discussed. However, it
should be noted that California energy consumption statistics include more than 20
different petroleum products.
Table A- 16. Ranking of CO2 Emissions from Fuel Combustion in 2004
( million metric tonne ( Mt) of CO2)
Fuel Mt CO2 %
Motor Gasoline 140.2 32.8%
Natural Gas 112.6 26.3%
Distillate 40.8 9.5%
Coal 37.5 8.8%
Imported Electricity 27.6 6.5%
Refinery Gas 19.6 4.6%
Associated gas 15.8 3.7%
Other 6.7 1.6%
Catalyst Coke 6.1 1.4%
Petroleum Coke 4.1 1.0%
Bituminous Coal 4.0 0.9%
Jet Fuel 2.8 0.7%
LPG 2.4 0.6%
Residual Fuel Oil 2.1 0.5%
Lubricants 1.0 0.5%
Naphtha 0.6 0.2%
Petroleum feedstocks 0.5 0.1%
Natural Gas Liquids 0.3 0.1%
Municipal Solid Waste 0.2 0.1%
Aviation Gasoline 0.2 0.1%
Tires 0.2 0.0%
Kerosene 0.2 0.0%
Source: CARB, 2007
• Motor gasoline consumption is the largest source of CO2 emissions from fuel
combustion in California. Uncertainties linked to the heating value and carbon factors of
motor gasoline are directly transferred to the total emissions of motor gasoline. For
example, if these factors increase by 1%, emissions increase accordingly. The
composition of California reformulated gasoline, designed to meet CARB regulations,
differs from that of average US gasoline. For the conversion of motor gasoline from
physical ( barrels) to energy ( Btu) units, CALEB uses a California- specific conversion
factor calculated by CARB ( Bemis, 2004). However, this was available only for 1995 and
46
1997. Concerning carbon content, a national average estimate was used. Calculation of
annual heating values and carbon contents of gasoline used in California will improve the
precision of California emission inventory. Moreover, the increased use of ethanol, as
opposed to MTBE, as a blending component of gasoline needs to be clearly specified, as
no carbon is associated with ethanol use. Ethanol is produced from the fermentation of
biomass and is considered carbon neutral by the IPCC. In the energy balance, it is
accounted as an input to the refineries under the product category biomass and it is
subtracted before calculating carbon emissions emitted from motor gasoline
consumption.
• Natural gas is a major fuel used in California, representing 39% of total CO2
emissions from fuel combustion in 2004. California relies heavily on imported natural
gas. In 2002, only about 15% of the natural gas supply is from in- state sources, while
almost half is imported from the Southwest U. S., a little over one- quarter from Canada,
and the remainder from the Rocky Mountain states, which began supplying natural gas to
California in 1992 ( Murtishaw et al., 2005). Heating value and carbon content values
vary according to the natural provenance. Data on the heating value used in CALEB
comes from SEDS that provides a conversion factor for California annually and for its
different use. However, a US average factor for the carbon content of natural gas was
used.
• Coal burned in California8 is imported from Colorado, Kentucky, New Mexico,
Utah, West Virginia, and Wyoming. Coal imports were relatively steady from 1990 to
1997, at which point they jumped from 2,794 thousand short tons ( kst) ( 65 TBtu) to
7,881kst ( 179 TBtu) in 2001 and started to decrease to reach 2,726 in 2005. Similarly to
natural gas, the heating value used in CALEB comes from SEDS, which varies annually
and by use. However, the carbon content of coal is an US average. This is a shortcoming,
since the carbon content of coal varies by the state in which it was mined and by coal
rank, and because the sources of coal for each consuming sector vary year by year.
• Other Fuel: California- specific carbon factors must be estimated for other fuels.
The fuels that are most likely to deviate from the US average are LPG, NGL, still gas and
petrochemical feedstocks. As mentioned in Section 2.2 data on petroleum coke
consumption in refineries are available under two distinct categories: marketable
petroleum coke and catalyst petroleum coke. However, the same energy conversion and
carbon emission factors were applied to both types of coke in CALEB. In a memo to
CARB from the Western States Petroleum Association ( Lev- On, 2007), a survey of some
WSPA members indicated that the 27.85 kg ( 61.4 lb) C/ MBtu factor used in CALEB may
overestimate the carbon content for catalyst coke by about 10 to 15%. The heating value
used by WSPA members may also be different from the one used in CALEB. WSPA
reports that the heating value varies significantly by time and across refineries.
We estimated that uncertainties associated with the carbon content values used in
CALEB are in the range of - 1% to + 5%. This range was calculated by using lower and
upper carbon content factors given in the IPCC guidelines ( IPCC, 2006) in the 2000
CALEB database.
8 Excluding coal used to produce imported electricity
47
3.2.3 Alternative Source/ Methods and Recommendations
Testing procedure
U. S. EPA’s Acid Rain Program requires that the emissions of electricity generation
facilities throughout the country be measured with continuous emissions monitoring
( CEM) systems. The program requires the reporting of hourly emissions measurements
of CO2, SO2, and NOx emissions from all facilities over 25 megawatts, and new facilities
under 25 megawatts that do not use low- sulfur fuel ( sulfur content less than 0.05% by
weight). Utilities can report CO2 emissions either by measuring them using a CO2 CEM,
or through estimation using an O2 CEM or a mass balance estimation ( U. S. EPA 2008b).
We obtained CEM CO2 measurements from 68 generation facilities in California, and
matched their 2004 CO2 emissions with fuel consumption estimates from U. S. EIA’s
906/ 920 and 860 time series data. We were able to match 64 of the 68 facilities in the
CEM database with their counterpart in the U. S. EIA database, accounting for virtually
100% of the measured CO2 emissions. These facilities account for 27% of the total fuel
consumption reported in the U. S. EIA database; it is not clear why the remaining four
facilities are not included in the CEMS database. We then calculated the actual 2004
CO2 emission factor per Btu for each facility matched in both databases. The average
CO2 emission factor for all matched facilities is 0.060 grams of CO2 per Btu of fuel; this
factor is 13% higher than the 0.053 g/ Btu emission factor for natural gas, but lower than
the 0.073 g/ Btu emission factor for diesel fuel ( CARB 2007a; 95% of all fossil fuel used
for in- state electricity generation is natural gas). Table A- 17 shows the fuel use,
emissions, and emissions factor for the ten largest facilities in both the U. S. EIA and
CEM datasets ( all of these facilities used only natural gas); these facilities account for
15% of the total reported fuel use ( U. S. EIA), and 55% of the total measured CO2
emissions ( CEM), from electricity generation in California. As shown in the table, the
emission factors of the ten largest individual plants vary from 0.057 to 0.090 g/ Btu, or
5% less than to 50% more than the statewide average of 0.060 g/ Btu, and 8% to 70%
more than the statewide average of 0.053 g/ Btu for energy generation facilities burning
natural gas.
48
Table A- 17. Fuel use, CO2 emissions, and CO2 emission factors of ten largest
California electricity generating facilities in U. S. EPA CEM database
Facility U. S. EIA fuel use
( TBtu)
CEM CO2
emissions
( Mt)
CO2 emission
factor
( grams/ Btu)
Moss Landing Power Plant 46.3 2.8 0.061
La Paloma Generating LLC 41.1 2.7 0.066
Delta Energy Center 41.1 2.4 0.057
Encina 34.3 2.1 0.061
AES Alamitos LLC 35.0 2.1 0.059
Elk Hills Power LLC 26.9 1.7 0.062
High Desert Power Project LLC 27.8 1.6 0.058
Los Medanos Energy Center 26.4 1.6 0.060
AES Huntington Beach LLC 16.1 1.4 0.090
Ormond Beach 24.0 1.4 0.059
Total 319.1 19.8 0.062
National Inventory
For the Inventory of U. S. Greenhouse Gas Emissions and Sinks, U. S. EPA estimates CO2
emissions from fuel combustion based on the heat content of the fuel and carbon content
coefficients in terms of carbon content per quadrillion Btu ( QBtu), using data from the
U. S. EIA. Carbon content factors are similar to the carbon content coefficients contained
in the IPCC's default methodology ( IPCC, 2006), with modifications reflecting fuel
qualities specific to the United States. Carbon content factors are derived from fuel
sample data, using descriptive statistics to estimate the carbon share of the fuel by weight.
The heat content of the fuel is also estimated based on the sample data, or where sample
data are unavailable or unrepresentative, by default values that reflect the characteristics
of the fuel as defined by market requirements.
The U. S. EPA provides a complete description of the method and data sources used in
Annex 2- Methodology and Data for Estimating CO2 Emissions from Fossil Fuel
Combustion from the US Inventory of U. S. Greenhouse Gas Emissions and Sinks ( U. S.
EPA, 2008). It is possible to replicate the methodology used, but data that are available at
the national level may not always be available at the state level.
Other Sources
For coal, the U. S. EIA provides a description of the coal used in California by electric
utility, industrial plant or other use; for each subsector, it provides the quantity of fuel by
its source. These data enable the calculation of a coal carbon factor specific to California
( Distribution of U. S Coal by Destination, U. S. EIA, 2008) 9
Calculation of specific carbon factors for all energy products consumed in California
should be carried over to allow for a more precise estimate of the CO2 emissions in the
state. We estimate that uncertainties associated with the carbon content values used in
9 http:// www. eia. doe. gov/ cneaf/ coal/ page/ coaldistrib/ d_ ca. html
49
CALEB are in the range of - 1% to + 5%. This range was calculated by using lower and
upper carbon content factors given in the IPCC guidelines ( IPCC, 2006) in the CALEB
database.
4. Conclusion
There are several important improvements to the energy balance that can be made to
better account for CO2 emissions from fuel combustion in California. This is mainly
because CALEB is built on data from many different sources. Care needs to be taken that
energy supply and consumption are properly matched, to eliminate or minimize any
double- counting. A difficulty is that surveys and questionnaires gathering the data across
the US are centralized through a federal agency, the U. S. EIA. Data are not always
reported at the state level, and when they are, they are often allocated to states using
proxies for actual supply and consumption. Finally, energy is used through a multitude
of different products and across many different end use activities. Gathering all the data
necessary to have a complete picture of all energy flows is a challenging task and data are
not always available.
This report focuses mainly on evaluating the areas where improvement is needed and
assessing uncertainties associated with CO2 emissions accounting. An attempt was made
to quantify uncertainties using alternative data, when such data were available. We
estimate a low and high uncertainty relative to current total CO2 emission estimates.
However, for some sectors these uncertainties are underestimated, as alternative data
were not available for all sectors or processes. For example, we did not estimate a range
of uncertainty for hydrogen production, as no alternative data were found. Moreover,
when alternative data was available, the range chosen for each sector was intentionally
large, to include all possible errors that could be identified and quantified with the
category considered. Table A- 18 shows the resulting range in percent uncertainty by
category and for the total state CO2 emissions, for the year 2004. A positive percentage
indicates that the current estimate of CO2 emissions is too low, while a negative
percentage indicates that the current estimate is too high. The table indicates that the
largest uncertainties come from unresolved reconciliation errors between supply and
consumption data (- 2% to + 4%), carbon emission factor uncertainties (- 1% to + 5%),
gasoline use by motor vehicles ( 2%), and fuel use in upstream (+ 1.1%) oil and gas
operations. There also are small uncertainties in emissions from fuel used as feedstock in
chemical plants fuel used in electric and CHP plants, diesel used by motor vehicles, and
fuel used for commercial aviation. The estimated uncertainty for all sectors ranges from -
19 and + 37 Mt, or - 5% and + 11% of total CO2 emissions.
50
Table A- 18. Percentage Uncertainties
2004 emissions
Category Estimated uncertainty
CO2 ( Mt) % CO2 ( Mt) % over each
category total
% over total
inventory
Electricity/ CHP* 62 18% 0.40 1% 0.1%
coal 4 1% 0.47 12% 0.1%
petroleum products 9 3% - 0.07 - 1% -
natural gas 49 14% - - -
Refining** 29 8% - - -
Oil/ gas extraction 14 4% 4.00 28% 1.1%
Industry feedstocks 1.8 1% ± 1.77 ± 100% ± 0.5%
Transportation 177 51% - 8.04 - 5% - 2.2 %
On- road vehicles 167 48% - 7.17 - 4%
Gasoline 138 39% - 8.52 - 6% - 2.4 %
Diesel 29 8% 1.35 5% 0.4 %
Aviation 3 1% - 0.84 - 28% - 0.2 %
Marine 3 1% - 6% -
Rail 3 1% - 0.03 - 1% -
Other*** 66 19% - - -
Reconciliation errors - - - 6.2 to 13.0 - 2% to 4%
Emission Factors - - - 2.7 to 17.6 - 1% to 5%
Total 350 100% - 18.7 to 36.8 - 5% to 11%
* Combined Heat and Power ( CHP)
** Uncertainties with hydrogen production are not estimated
*** ncludes emissions from other sectors such as other industry, residential, commercial/ institutional,
agriculture/ forestry/ fishing/ fish farms and non- specified.
The largest uncertainty lies in reconciling statistics on fuel supply and consumption;
available data do not match for most fuels. Many data gaps remain in accounting for total
energy flows in California, especially for petroleum products such as natural gas liquids
( NGL), liquefied petroleum products ( LPG), or still gas. The second largest uncertainty
comes from the use of national carbon factors which do not reflect California factors.
The largest uncertainty in the transport sector, gasoline used by vehicles, is estimated by
comparing results from a bottom- up emissions inventory model ( EMFAC) with total
gasoline sales. The representation of combined heat and power ( CHP) in the energy
balance needs to be improved by allocating all energy used for commercial and industrial
CHP to the sector where the generated electricity is used; all CHP energy use by facilities
whose primary business is to sell electricity and heat should be allocated to the electricity
generation sector. Finally, reported data on energy use in upstream oil and gas operations
is lacking.
5. Recommendations
5.1.1 Improve CALEB
There are a few areas where the CALEB database can be updated with new data
identified in this report. This mainly includes the energy used in CHP and the
51
disaggregation of individual petroleum product inputs to the electricity generation sector.
To the extent possible, improvements identified in this report will be included in the
update of CALEB to 2006, which will be funded by CEC.
In addition to the new sources identified in this report, there are several other
improvements that can be made to CALEB. Those improvements, and the data required
to make them, are discussed below.
5.1.2 Conduct Surveys
For these industries where the accounting of CO2 emissions requires more data on energy
use, such as refineries, oil companies and chemical industries, surveys that collect the
additional data needed would help to fill the gaps in the CALEB database. This could be
done on the basis of the national MECS survey ( U. S. EIA, 2005b), or more specifically
directed to the accounting of CO2 emissions in these industries. This will also allow the
industry to have a better representation of their CO2 emissions trends over time and give
CARB the opportunity to monitor progress in reducing emissions.
5.1.3 Bottom- Up Models
Bottom- up models are a very helpful tool to assess the energy use in end use sectors and
to corroborate top- down sales data. CARB has developed a few bottom- up models to
account for particulate and other criteria pollutant emissions. An adaptation of these
models to account for CO2 emissions would be very valuable for the GHG emissions
inventory. For example, little is known regarding the quantity of diesel fuel used by the
agriculture sector. It would help to develop an estimation based on equipment penetration
and time of use to compare with available data on fuel sales. This type of analysis would
be most valuable for petroleum products used, where sales data do not always indicate
the breakdown of consumers by sector.
5.1.4 Collaboration with the U. S. EIA and U. S. EPA
The U. S. EIA gathers a wealth of information on fuel production and supply through
multiple questionnaires and surveys. CEC and/ or CARB should obtain dedicated access
to these data to improve data collection for the state. For example, data disaggregated at
the petroleum product level representing inputs to non- utility electricity generation
facilities are only available from 1998. We requested that U. S. EIA provide these data
prior to 1998; however, these data are confidential and were not provided to us.
Collaboration with the U. S. EPA could also help assess what information is necessary to
develop specific carbon factors for California. Consultation with U. S. EPA would be
beneficial for CARB to develop specific carbon factors and feedstock carbon storage
factors for California.
5.1.5 Compare measured and calculated CO2 emissions from electric
utilities
U. S. EPA’s Continuous Emission Monitoring program measures hourly CO2 emissions
from electricity gen