Science Advisory Board (SAB) Draft Report (7/26/12) for Quality SAB Review of EPA’s Accounting Framework for Biogenic CO2 Emissions from 10 Stationary Sources (September 2011)


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
EPA-SAB-12 -xxx
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
Subject: SAB Review of EPA's Accounting Framework for Biogenic CO2 Emissions from
Stationary Sources (September 2011)
Dear Administrator Jackson:
EPA's Science Advisory Board (SAB) was asked to review and comment on the EPA's Accounting
Framework for Biogenic CO2 Emissions from Stationary Sources (Framework, September 2011). The
Framework considers the scientific and technical issues associated with accounting for emissions of
biogenic carbon dioxide (CO2) from stationary sources and develops a method to adjust the stack
emissions from stationary sources using biological material based on the induced changes in carbon
stocks on land (in soils, plants and forests). A panel of experts was formed under the auspices of the
SAB to conduct the review. Advice is provided through the chartered SAB. The panel, comprised of
experts in forestry, agriculture, greenhouse gas measurement and inventories, land use economics,
ecology, climate change and engineering met in a face-to-face meeting and held four teleconferences.
Assessing the greenhouse gas implications of using biomass to produce energy is a daunting task and the
EPA is to be commended for its effort. The context for the Framework arose when the EPA established
thresholds for greenhouse gas emissions from stationary sources for the purposes of Clean Air Act
permits under the New Source Review (Prevention of Significant Deterioration program) and Title V
operations program. The Agency had to consider how to include biogenic emissions in determining
whether thresholds for regulation have been met. In July 2011, the EPA deferred for a period of three
years the application of permitting requirements to biogenic carbon dioxide emissions from bioenergy
and other biogenic stationary sources, while committing to a detailed examination of the issues
associated with biogenic CO2.
The Agency sought a method of "adjusting" biogenic carbon emissions from stationary sources to credit
those emissions with carbon uptake during sequestration or, alternatively, avoided emissions from
natural decay (e.g., from residues and waste materials). Without a way of adjusting those emissions, the
Agency's options would be either a categorical inclusion (treating biogenic feedstocks as equivalent to
fossil fuels) or a categorical exclusion (excluding biogenic emissions from determining applicability
thresholds for regulation). The purpose of the Framework was to propose a method for calculating the
adjustment or Biogenic Accounting Factor (BAF) for biogenic feedstocks based on their interaction with
the carbon cycle.
In general, the SAB found that the science and technical issues relevant to accounting for biogenic CO2
emissions are different for each feedstock category and sometimes differ within a category. For instance,

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
forest-derived woody biomass has a much longer rotation period than agricultural feedstocks. The
Framework includes most of the elements that would be needed to gauge changes in CO2 emissions;
however, the reference year approach employed does not provide an estimate of the additional emissions
and the sequestration changes in response to biomass feedstock demand. Estimating additionality is
essential, as it is the crux of the question at hand. To do so requires an anticipated baseline approach.
Because forest-derived woody biomass is a long-rotation feedstock, the Framework would need to
model a "business as usual" scenario along some time scale and compare that carbon trajectory with a
scenario of increased demand for biomass. Although this would not be an easy task, it would be
necessary to estimate carbon cycle changes associated with the biogenic feedstock. In addition, an
anticipated baseline would be needed to estimate additional changes in soil carbon stock over time. In
general the Framework should provide a means to estimate the effect of stationary source biogenic
feedstock demand, on the atmosphere over time.
In the attached report, the SAB provides some suggestions for an "anticipated baseline" approach while
acknowledging the uncertainty and difficulty associated with modeling future scenarios. It would be
particularly important to capture market and landscape level effects, specifically the complex interaction
between electricity generating facilities and forestry markets; market driven shifts in planting,
management and harvests; induced displacement of existing uses of biomass; land use changes; and the
relative contribution of different feedstock source categories (logging residue, pulpwood or roundwood
harvest). A landscape, versus stand or plot, perspective is important because simultaneous management
decisions may affect net carbon emissions over time.
For agricultural feedstocks, the variables in the Framework capture most of the factors necessary for
estimating the carbon change associated with the feedstock use, including a factor to represent the
carbon embodied in products leaving the stationary source, the proportion of feedstock lost in
conveyance, the offset represented by sequestration, the site-level difference in net carbon flux and the
emissions that would occur "anyway" from removal or diversion of nongrowing feedstock (e.g., corn
stover) and other variables. For short rotation agricultural feedstocks where carbon accumulation occurs
within one to a few years, the Framework can, with some adjustments to address estimation problems
(including an anticipated baseline for soil carbon changes) and careful consideration of data and
implementation, represent direct carbon changes in a particular region. As recognized by the Agency,
for many waste feedstocks (municipal solid waste, construction and demolition waste, industrial wastes,
manure, tire-derived wastes and wastewater), combustion to produce energy releases CO2 that would
have otherwise been returned to the atmosphere from the natural decay of waste. The Agency chose not
to model natural decomposition in the Framework but modeling the decay of agricultural and forest
residues based on their alternate fate (e.g., whether the materials would have been disposed in a
controlled or uncontrolled landfill or left on site, or subject to open burning) could be incorporated to
improve scientific accuracy.
The Framework does not discuss the different time scales inherent in the carbon cycle nor does it
characterize potential intertemporal tradeoffs associated with the use of biogenic feedstocks. However it
is important that intertemporal tradeoffs be made transparent for policymakers. For forest-derived
roundwood, carbon debts and credits can be created in the short run with increased harvesting and
planting respectively but in the long run, net climate benefits can accrue with net forest growth. While it
is clear that the Agency can only regulate emissions, its policy choices about regulating emissions will

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
be better informed with consideration of the temporal distribution of biogenic emissions and associated
carbon sequestration or avoided emissions.
Overall, the SAB found that quantification of most components of the Framework has uncertainties,
technical difficulties, data deficiencies and implementation challenges. These issues received little
attention in the Framework, but are important considerations relevant to scientific integrity and
operational efficiency. Moreover, the Agency should consider consistency between biogenic carbon
accounting and fossil fuel emissions accounting. Ideally both fossil fuels and biogenic feedstocks should
be subject to the same emissions accounting. While there are no easy answers to accounting for the
greenhouse gas implications of bioenergy, further consideration of the issues raised by the SAB and
revisions to the Framework could result in more scientific rigor in accounting for biogenic emissions. A
dissenting opinion, offered in Appendix E, recommends that the Agency abandon the Framework
altogether and instead choose to exempt biogenic CO2 emissions from greenhouse gas regulation so long
as aggregate measures of land-based carbon stocks are steady or increasing.
Given the challenges associated with improving and implementing the Framework, the SAB encourages
EPA to consider developing default BAFs by feedstock category and region. Under EPA's Framework,
facility-specific BAFs would be calculated to reflect the incremental carbon cycle and net emissions
effects of a facility's use of a biogenic feedstock. With default BAFs, biogenic emissions from a facility
would be based on the weighted combination of default BAFs relevant to a facility's feedstock
consumption and location. The defaults might vary by region, prior land use and current land
management practices due to the differences these might cause in the interaction between feedstock
production and the carbon cycle. They are likely to be more scientifically robust in that they could rely
on readily available data. The defaults would also have administrative advantages in that they would be
easier to implement and update. Facilities could also be given the option of demonstrating a lower BAF
for their feedstocks.
The SAB acknowledges that practical considerations will weigh heavily in the Agency's calculus. In
fact, any method that might be adopted or considered, including methods proposed by the SAB, should
be subject to an evaluation of the costs of compliance and the carbon emissions savings likely to be
achieved as compared to both a categorical inclusion and a categorical exclusion. Uncertainties in the
assessment of both the costs and the emissions savings should be analyzed and used to inform the choice
of policy.
The SAB appreciates the opportunity to provide advice on the Framework and looks forward to your
response.
Sincerely,
Enclosure

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
U.S. Environmental Protection Agency
Science Advisory Board
Biogenic Carbon Emissions Panel
CHAIR
Dr. Madhu Khanna, Professor, Department of Agricultural and Consumer Economics, University of
Illinois at Urbana-Champaign, Urbana, IL
MEMBERS
Dr. Robert Abt, Professor of Forestry, Department of Forestry and Environmental Resources, College
of Natural Resources, North Carolina State University, Raleigh, NC
Dr. Morton Barlaz, Professor, Civil, Construction, and Environmental Engineering, Engineering, North
Carolina State University, Raleigh, NC
Dr. Richard Birdsey, Program Manager, Climate, Fire, and Carbon Cycle Sciences, Northern Research
Station, USDA Forest Service, Newtown Square, PA
Dr. Marilyn Buford, National Program Leader, Silviculture Research, Research & Development,
USDA Forest Service, Washington, DC
Dr. Mark Harmon, Professor and Richardson Chair, College of Forestry, Oregon State University,
Corvallis, OR
Dr. Jason Hill, Assistant Professor, Bioproducts and Biosystems Engineering, College of Food,
Agricultural and Natural Resource Sciences, University of Minnesota, St. Paul, MN
Dr. Stephen Kelley, Professor and Department Head, Forest Biomaterials, College of Natural
Resources, North Carolina State University, Raleigh, NC
Dr. Richard Nelson, Director and Department Head, Engineering Extension Programs, Kansas State
University Center for Sustainable Energy, Manhattan, KS
Dr. Lydia Olander, Director, Ecosystem Services Program, Nicholas Institute for Environmental Policy
Solutions, Duke University, Durham, NC
Dr. John Reilly, Senior Lecturer and Co-Director, Joint Program on the Science and Policy of Global
Change, Center for Environmental Policy Research, E19-439L, Massachusetts Institute of Technology,
Cambridge, MA
i

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Dr. Charles Rice, Distinguished Professor, Department of Agronomy, Soil Microbiology, Kansas State
University, Manhattan, KS
Dr. Steven Rose, Senior Research Economist, Energy and Environmental Analysis Research Group,
Electric Power Research Institute, Palo Alto, CA
Dr. Daniel Schrag, Professor of Earth and Planetary Sciences, Harvard University, Cambridge, MA
Dr. Roger Sedjo, Senior Fellow and Director of the Center for Forest Economics and Policy Program,
Resources for the Future, Washington, DC
Dr. Ken Skog, Supervisory Research Forester, Economics and Statistics Research, Forest Products
Laboratory, USDA Forest Service, Madison, WI
Dr. Tristram West, Ecosystem Scientist, Joint Global Change Research Institute, University of
Maryland, College Park, MD
Dr. Peter Woodbury, Senior Research Associate, Department of Crop and Soil Sciences, College of
Agriculture and Life Sciences, Cornell University, Ithaca, NY
SCIENCE ADVISORY BOARD STAFF
Dr. Holly Stallworth, Designated Federal Officer, U.S. Environmental Protection Agency, Science
Advisory Board (MOOR), 1200 Pennsylvania Avenue, NW, Washington, DC 20460
* / Dr. Sedjo's dissenting opinion is offered in Appendix E.
ii

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
U.S. Environmental Protection Agency
Science Advisory Board
BOARD
CHAIR
Dr. Deborah L. Swackhamer, Professor and Charles M. Denny, Jr. Chair in Science, Technology and
Public Policy, Hubert H. Humphrey School of Public Affairs and Co-Director of the Water Resources
Center, University of Minnesota, St. Paul, MN
SAB MEMBERS
Dr. George Alexeeff, Director, Office of Environmental Health Hazard Assessment, California
Environmental Protection Agency, Oakland, CA
Dr. David T. Allen, Professor, Department of Chemical Engineering, University of Texas, Austin, TX
Dr. Pedro Alvarez, Department Chair and George R. Brown Professor of Engineering, Department of
Civil & Environmental Engineering, Rice University, Houston, TX
Dr. Joseph Arvai, Svare Chair in Applied Decision Research, Institute for Sustainable Energy,
Environment, & Economy, Haskayne School of Business, University of Calgary, Calgary, Alberta,
Canada
Dr. Claudia Benitez-Nelson, Full Professor and Director of the Marine Science Program, Department
of Earth and Ocean Sciences, University of South Carolina, Columbia, SC
Dr. Patricia Buffler, Professor of Epidemiology and Dean Emerita, Department of Epidemiology,
School of Public Health, University of California, Berkeley, CA
Dr. Ingrid Burke, Director, Haub School and Ruckelshaus Institute of Environment and Natural
Resources, University of Wyoming, Laramie, WY
Dr. Thomas Burke, Professor and Jacob I. and Irene B. Fabrikant Chair in Health, Risk and Society
Associate Dean for Public Health Practice, Johns Hopkins Bloomberg School of Public Health, Johns
Hopkins University, Baltimore, MD
Dr. Terry Daniel, Professor of Psychology and Natural Resources, Department of Psychology, School
of Natural Resources, University of Arizona, Tucson, AZ
Dr. George Daston, Victor Mills Society Research Fellow, Product Safety and Regulatory Affairs,
Procter & Gamble, Cincinnati, OH
in

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Dr. Costel Denson, Managing Member, Costech Technologies, LLC, Newark, DE
Dr. Otto C. Doering III, Professor, Department of Agricultural Economics, Purdue University, W.
Lafayette, IN
Dr. Michael Dourson, President, Toxicology Excellence for Risk Assessment, Cincinnati, OH
Dr. David A. Dzombak, Walter J. Blenko, Sr. University Professor of Environmental Engineering,
Department of Civil and Environmental Engineering, College of Engineering, Carnegie Mellon
University, Pittsburgh, PA
Dr. T. Taylor Eighmy, Senior Vice President for Research, Office of the Vice President for Research,
Texas Tech University, Lubbock, TX
Dr. Elaine Faustman, Professor and Director, Institute for Risk Analysis and Risk Communication,
School of Public Health, University of Washington, Seattle, WA
Dr. John P. Giesy, Professor and Canada Research Chair, Veterinary Biomedical Sciences and
Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
Dr. Jeffrey K. Griffiths, Professor, Department of Public Health and Community Medicine, School of
Medicine, Tufts University, Boston, MA
Dr. James K. Hammitt, Professor, Center for Risk Analysis, Harvard University, Boston, MA
Dr. Barbara L. Harper, Risk Assessor and Environmental-Public Health Toxicologist, and Division
Leader, Hanford Projects, and Program Manager, Environmental Health, Department of Science and
Engineering, Confederated Tribes of the Umatilla Indian Reservation (CTUIR), West Richland, WA
Dr. Kimberly L. Jones, Professor and Chair, Department of Civil Engineering, Howard University,
Washington, DC
Dr. Bernd Kahn, Professor Emeritus and Associate Director, Environmental Radiation Center, Georgia
Institute of Technology, Atlanta, GA
Dr. Agnes Kane, Professor and Chair, Department of Pathology and Laboratory Medicine, Brown
University, Providence, RI
Dr. Madhu Khanna, Professor, Department of Agricultural and Consumer Economics, University of
Illinois at Urbana-Champaign, Urbana, IL
Dr. Nancy K. Kim, Senior Executive, Health Research, Inc., Troy, NY
Dr. Cecil Lue-Hing, President, Cecil Lue-Hing & Assoc. Inc., Burr Ridge, IL
iv

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Dr. Judith L. Meyer, Professor Emeritus, Odum School of Ecology, University of Georgia, Lopez
Island, WA
Dr. James R. Mihelcic, Professor, Civil and Environmental Engineering, University of South Florida,
Tampa, FL
Dr. Christine Moe, Eugene J. Gangarosa Professor, Hubert Department of Global Health, Rollins
School of Public Health, Emory University, Atlanta, GA
Dr. Horace Moo-Young, Dean and Professor, College of Engineering, Computer Science, and
Technology, California State University, Los Angeles, CA
Dr. Eileen Murphy, Director of Research and Grants, Ernest Mario School of Pharmacy, Rutgers
University, Piscataway, NJ
Dr. James Opaluch, Professor and Chair, Department of Environmental and Natural Resource
Economics, College of the Environment and Life Sciences, University of Rhode Island, Kingston, RI
Dr. Duncan Patten, Director, Montana Water Center, and Research Professor, Hydroecology Research
Program, Department of Land Resources and Environmental Sciences, Montana State University,
Bozeman, MT
Dr. Stephen Polasky, Fesler-Lampert Professor of Ecological/Environmental Economics, Department
of Applied Economics, University of Minnesota, St. Paul, MN
Dr. C. Arden Pope, III, Professor, Department of Economics, Brigham Young University, Provo, UT
Dr. Stephen M. Roberts, Professor, Department of Physiological Sciences, Director, Center for
Environmental and Human Toxicology, University of Florida, Gainesville, FL
Dr. Amanda Rodewald, Professor of Wildlife Ecology, School of Environment and Natural Resources,
The Ohio State University, Columbus, OH
Dr. Jonathan M. Samet, Professor and Flora L. Thornton Chair, Department of Preventive Medicine,
Keck School of Medicine, University of Southern California, Los Angeles, CA
Dr. James Sanders, Director and Professor, Skidaway Institute of Oceanography, Savannah, GA
Dr. Jerald Schnoor, Allen S. Henry Chair Professor, Department of Civil and Environmental
Engineering, Co-Director, Center for Global and Regional Environmental Research, University of Iowa,
Iowa City, IA
Dr. Gina Solomon, Deputy Secretary for Science and Health, Office of the Secretary, California
Environmental Protection Agency, Sacramento, CA
v

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Dr. Daniel O. Stram, Professor, Department of Preventive Medicine, Division of Biostatistics,
University of Southern California, Los Angeles, CA
Dr. Peter S. Thorne, Director, Environmental Health Sciences Research Center and Professor and
Head, Departmment of Occupational and Environmental Health, College of Public Health, University of
Iowa, Iowa City, IA
Dr. Paige Tolbert, Professor and Chair, Department of Environmental Health, Rollins School of Public
Health, Emory University, Atlanta, GA
Dr. John Vena, University of Georgia Foundation Professor in Public Health and Head, Department of
Epidemiology and Biostatistics, Georgia Cancer Coalition Distinguished Scholar, College of Public
Health , University of Georgia, Athens, GA
Dr. Robert Watts, Professor of Mechanical Engineering Emeritus, Tulane University, Annapolis, MD
Dr. R. Thomas Zoeller, Professor, Department of Biology, University of Massachusetts, Amherst, MA
SCIENCE ADVISORY BOARD STAFF
Dr. Angela Nugent, Designated Federal Officer, U.S. Environmental Protection Agency, Science
Advisory Board (1400R), Washington, DC
vi

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
NOTICE
This report has been written as part of the activities of the EPA Science Advisory Board (SAB), a
public advisory group providing extramural scientific information and advice to the Administrator and
other officials of the Environmental Protection Agency. The SAB is structured to provide balanced,
expert assessment of scientific matters related to problems facing the agency. This report has not been
reviewed for approval by the agency and, hence, the contents of this report do not necessarily represent
the views and policies of the Environmental Protection Agency, nor of other agencies in the Executive
Branch of the Federal government, nor does mention of trade names of commercial products constitute a
recommendation for use. Reports of the SAB are posted on the EPA website at http://www.epa.gov/sab.
vii

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
2
3	List of Figures	ii
4	Acronyms and Abbreviations	iii
5	1. EXECUTIVE SUMMARY	4
6	2. INTRODUCTION	13
7	3. RESPONSES TO EPA's CHARGE QUESTIONS	15
8	3.1. The Science of Biogenic CO: Emissions	15
9	3.2. Biogenic CO: Accounting Approaches	18
10	3.3. Methodological Issues	20
11	3.4. Accounting Framework	29
12	3.5. Case Studies	40
13	3.6. Overall Evaluation	42
14	4. DEFAULT BAFs BASED ON FEEDSTOCK CATEGORIES	47
15	REFERENCES	48
16	APPENDIX A: Charge to the Panel	A-l
17	APPENDIX B: Temporal Changes in Stand Level Biogenic Emissions Versus Fossil Emissions	B-l
18	APPENDIX C: Fate of Landscape Residue after Harvest and System Storage of Carbon	C-l
19	APPENDIX D: Carbon Balances over Time in an Existing Forest System	D-l
20	APPENDIX E: Dissenting Opinion from Dr. Roger Sedjo	E-l
21
l

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Figure B-l:
Figure C-l:
Figure C-2:
Figure D-l
Figure D-2
Figure D-3
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
List of Figures
Surface temperature change from biogenic emissions versus fossil fuel over time. Adapted
from Cherubini et al. (2012) and reprinted with copyright permission	B-2
Fate of residue/slash left after harvest as function of k and time since harvest	C-l
Landscape average store of residue/slash as function of k and harvest interval	C-2
Changes in carbon stores of major forest ecosystem pools when a 50 year clear-cut harvest
system is established and continued. The result is a continued carbon balance	D-2
Changes in carbon stores of major forest ecosystem pools when a 50 year clear-cut harvest
system is replaced by a 25 year clear-cut harvest system in 2010. The result is a carbon
debt	D-2
Changes in carbon stores of major forest ecosystem pools when a 50 year clear-cut harvest
system is replaced by a 100 year clear-cut harvest system in 2010. The result is a carbon
gain	D-3
li

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Acronyms and Abbreviations
AVOIDEMIT
Avoided Emissions
BAF
Biogenic Accounting Factor
ch4
Methane
C02
Carbon Dioxide
C02e
Carbon Dioxide Equivalent
DOE
Department of Energy
EPA
Environmental Protection Agency
FASOM
Forestry and Agricultural Sector Optimization Model
GHG
Greenhouse gases
GROW
Growth
GtC/y
Gigatons of carbon per year
GTMM
Global Timber Market Model
GTP
Global Temperature Potential
GWP
Global Warming Potential
GWPbio
Global Warming Potential of biomass
I
Carbon Input
K
Proportion of Carbon Lost per unit of time
LAR
Level of Atmospheric Reduction
LEAK
Leakage
n2o
Nitrous Oxide
NSR
New Source Review
PRODC
Carbon in Products
PSD
Prevention of Significant Deterioration
RPA
Resources Planning Act
SAB
Science Advisory Board
SEQP
Sequestered Fraction
SITE TNC
Total Net Change in Site Emissions
SRTS
Sub-regional Timber Supply Model
USD A
United States Department of Agriculture
111

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
1. EXECUTIVE SUMMARY
Biogenic CO2 emissions from bioenergy are generated during the combustion or decomposition of
biologically-based material. Biogenic feedstocks differ from fossil fuels in that they may be replenished
in a continuous cycle of planting, harvesting and regrowth. The same plants that provide combustable
feedstocks for electricity generation also sequester carbon from the atmosphere. Plants convert raw
materials present in the ecosystem such as carbon from the atmosphere and inorganic minerals and
compounds from the soil (including nitrogen, potassium, and iron) and make these elemental nutrients
available to other life forms. Carbon is returned to the atmosphere by plants and animals through
decomposition and respiration and by industrial processes, including combustion. Biogenic CO2 is
emitted from stationary sources through a variety of energy-related and industrial processes. Thus, the
use of biogenic feedstocks results in both carbon emissions and carbon sequestration.
EPA's Accounting Framework for Biogenic CO 2 Emissions from Stationary Sources (Framework,
September 2011) explores the scientific and technical issues associated with accounting for emissions of
biogenic carbon dioxide (CO2) from stationary sources and develops a method to adjust the stack
emissions from bioenergy based on the induced changes in carbon stocks on land (in soils, plants and
forests). The context for the Framework is the treatment of biogenic CO2 emissions in stationary source
regulation given the unique feature of plant biomass in providing uptake of carbon dioxide (CO2) from
the atmosphere during the photosynthesis. Under the Clean Air Act, major new sources of certain air
pollutants, defined as "regulated New Source Review (NSR) pollutants" and major modifications to
existing major sources are required to obtain a permit. The set of conditions that determine which
sources and modifications are subject to the Agency's permitting requirements are referred to as
"applicability" requirements. Since greenhouse gases are included in the definition of a "regulated NSR
pollutant," EPA has to make a determination about whether a source meets the "applicability threshold"
to trigger permitting requirements. As of January 2011, for facilities already covered by the PSD or Title
V programs, greenhouse gas emission increases of 75,000 tons per year (tpy) or more, on a carbon
dioxide equivalent (CC^e) basis, would be subject to technology requirements under the PSD program.
As of July 1, 2011, more facilities became subject to regulation based on their greenhouse gas
emissions. Specifically new and existing stationary sources (that are not already covered by the PSD or
Title V programs) that emit greenhouse gas emissions of at least 100,000 tpy became subject to
greenhouse gas regulation even if they do not exceed the permitting thresholds for any other pollutant.
The question before the Agency, and hence, the motivation for the Framework, is whether and how to
consider biogenic greenhouse gas emissions in determining these thresholds for permitting. The SAB's
consensus advice is highlighted in this Executive Summary with more details in the attached report. A
dissenting opinion is found in Attachment E.
Categorical inclusion or exclusion
In approaching these questions, EPA considered whether it could a priori categorically include or
exclude biogenic emissions in its greenhouse gas accounting. Thus, before doing its own analysis, EPA
considered whether it could make a categorical determination. A categorical inclusion would count all
biogenic carbon emissions at the combustion source, similar to fossil fuel emissions while a categorical
exclusion would exempt biogenic carbon emissions from greenhouse gas regulation. The Agency
4

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
rejected both extremes and asked the SAB whether it supported their conclusion that a priori categorical
approaches are inappropriate for the treatment of biogenic carbon emissions.
The decision about a categorical inclusion or exclusion will likely involve many considerations that fall
outside the SAB's scientific purview such as legality, feasibility and, possibly, political will. The SAB
cannot speak to the legal or regulatory complexities that could accompany any policy on biogenic
carbon emissions but this Advisory offers some scientific observations that may inform the
Administrator's policy decision.
Carbon neutrality cannot be assumed for all biomass energy a priori. There are circumstances in which
biomass is grown, harvested and combusted in a carbon neutral fashion but carbon neutrality is not an
appropriate a priori assumption; it is a conclusion that should be reached only after considering a
particular feedstock's production and consumption cycle. There is considerable heterogeneity in
feedstock types, sources and production methods and thus net biogenic carbon emissions will vary
considerably. Of course, biogenic feedstocks that displace fossil fuels do not have to be carbon neutral to
be better than fossil fuels in terms of their climate impact.
Given that some biomass could have positive net emissions, a categorical exclusion would remove any
responsibility on the stationary source for CO2 emissions from its use of biogenic material and provide
no incentive for the development and use of best management practices. Conversely, a categorical
inclusion would provide no incentive for using biogenic sources that compare favorably to fossil energy
in terms of greenhouse gas emissions. The dissenting opinion in Attachment E offers support for a
categorical exclusion so long as aggregate measures of land-based carbon stocks are steady or
expanding.
Biogenic Accounting Factor (BAF) Calculation
The Framework presents an alternative to a categorical inclusion or exclusion by offering an equation
for calculating a Biogenic Accounting Factor (BAF) would be used to adjust the onsite biogenic
emissions at the stationary source emitting biogenic CO2 on the basis of information about growth of the
feedstock and/or avoidance of biogenic emissions and more generally the carbon cycle.
Forest-Derived Woody Biomass
The EPA's stated objective was to accurately reflect the carbon outcome of biomass use by stationary
sources. For forest-derived woody biomass, the Framework did not achieve this objective. To calculate
BAF for biomass from roundwood trees, the Agency proposed the concept of regional carbon stocks
(with the regions unspecified) and posed a "rule" whereby any bioenergy usage that takes place in a
region where carbon stocks are increasing would be assigned a BAF of 0. This decouples the BAF from
a particular facility's biogenic emissions and the sequestration (offset) associated with its particular
feedstock. Emissions from a stationary facility would be included or excluded from greenhouse gas
regulation depending on a host of factors in the region far beyond the facility's control.
To accurately capture the carbon outcome, an anticipated baseline approach is needed, and a landscape
level perspective. An anticipated baseline requires selecting a time period and determining what would
have happened anyway without the harvesting and comparing that impact with the carbon trajectory
5

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
associated with harvesting of biomass for bioenergy. Although any "business as usual" projection would
be uncertain, it is the only means by which to gauge the incremental impact of woody biomass
harvesting. The Framework discusses this anticipated future baseline approach but does not attempt it.
Instead a fixed reference point and an assumption of geographic regions were chosen to determine the
baseline for whether biomass harvesting for bioenergy facilities is having a negative impact on the
carbon cycle. The choice of a fixed reference point may be the simplest to execute, but it does not
properly address the additionality question, i.e., the extent to which forest stocks would have been
growing or declining over time in the absence of bioenergy. The Agency's use of a fixed reference point
baseline coupled with a division of the country into regions implies that forest biomass emissions could
be granted an exemption simply because the location of a stationary facility is in an area where forest
stocks are increasing. The reference point estimate of regionwide net emissions or net sequestration does
not indicate, or estimate, the difference in greenhouse gas emissions (the actual carbon gains and losses)
over time that stem from biomass use. As a result, the Framework fails to capture the causal connection
between forest biomass growth and harvesting and atmospheric impacts and thus may incorrectly assess
net CO2 emissions of a facility's use of a biogenic feedstock.
A landscape, versus stand or plot, perspective is important because land-management decisions are
simultaneous, e.g., harvesting, planting, silvacultural treatments. Thus, there are concurrent carbon stock
gains and losses that together define the net implications over time. A landscape level analysis, and BAF
calculation, will capture these.
Agricultural and Waste Feedstocks
For faster growing biomass like agricultural crops, the anticipated future baseline approach is still
necessary to reflect changes in dynamic processes, e.g., soil carbon, "anyway" emissions, and landscape
changes. For agricultural feedstocks in general, the Framework captures many of the factors necessary
for estimating the offsite carbon change associated with use of short rotation (agricultural) feedstocks.
These include factors to represent the carbon embodied in products leaving a stationary source, the
proportion of feedstock lost in conveyance, the offset represented by sequestration, the site-level
difference in net carbon flux as a result of harvesting, the emissions that would occur "anyway" from
removal or diversion of nongrowing feedstocks (e.g., corn stover) and other variables. In addition to the
anticipated baseline, a noticeable omission is the absence of consideration of nitrous oxide (N2O)
emissions from fertilizer use, potentially a major onsite greenhouse gas loss that could be induced by a
growing bioenergy market.
For short rotation feedstocks where carbon accumulation and "anyway" emissions are within one to a
few years (i.e., agricultural residues, perennial herbaceous crops, mill wood wastes, other wastes), the
Framework may, with some adjustments to address estimation problems (including an anticipated
baseline for soil carbon changes, residue disposition and land management) and careful consideration of
data and implementation, accurately represent direct carbon changes in a particular region. For logging
residues and other feedstocks that decay over longer periods, decomposition cannot be assumed to be
instantaneous and the Framework could be modified to incorporate the time path of decay of these
residues if they are not used for bioenergy. This time path should consider the alternative fate of these
residues, which in some cases may involve removal and burning to reduce risks of fire or maintain forest
health.
6

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
For waste materials (municipal solid waste), the Framework should consider the alternate disposition of
waste material (what would happen if not used as feedstock) in an anticipated baseline (counterfactual)
framework. This anticipated baseline should include emissions and partial capture of methane (CH4)
emissions from landfills. In general, when accounting for emissions from wood mill waste and pulping
liquor, the EPA should recognize these emissions are part of a larger system that includes forests, solid
wood mills, pulp mills and stationary energy sources. Accounting for greenhouse gases in the larger
system should track all emissions or forest stock changes over time across the outputs from the system
so as to account for all fluxes. Within the larger system, the allocation of fluxes to wood/paper products
or to a stationary source is a policy decision. The Agency should consider how its Framework meets the
scientific requirement to account (allocate) all emissions across the larger system of forests, mills and
stationary sources over time.
Leakage
Leakage is a phenomenon by which efforts to reduce emissions in one place affect market prices that
shift emissions to another location or sector. The Framework's equation for BAF includes a term for
leakage, however the Agency decided that calculating values for leakage was outside the scope of the
Framework. While that decision was expedient, it should be recognized that incorporating leakage,
however difficult, may change the BAF results radically. "Bad" leakage (called "positive" leakage in the
literature) occurs when the use of biogenic feedstocks causes price changes which, in turn, drive changes
in consumption and production outside the boundary of the stationary source, even globally, that lead to
increased carbon emissions. One type of positive leakage could occur if land is diverted from food/feed
production to bioenergy production which increases the price of conventional agricultural and forest
products in world markets and leads to conversion of carbon -rich lands to crop production and the
release of carbon stored in soils and vegetation. The use of biogenic feedstocks can also affect the price
of fossil fuels by lowering demand for them and thereby increasing their consumption elsewhere.
"Good" leakage (called "negative" leakage in the literature) could occur if the use of biomass leads to
carbon -offsetting activities elsewhere. The latter could arise for example, if increased demand for
biomass and higher prices generate incentives for investment in forest management, beyond the level
needed directly for bioenergy production, which increases net forest carbon sequestration.
The existing literature in the social sciences shows that the overall magnitude of leakage, associated with
the use of bioenergy for fuel is highly uncertain and differs considerably across studies and within a
study, depending on underlying assumptions. It will also differ by feedstock and location. Rather than
eschewing the calculation of leakage altogether, the Agency might instead, try to ascertain the
directionality of net leakage — whether it is positive (leading to increased carbon emissions elsewhere)
or negative (leading to carbon offsetting activities) — and incorporate that information in its decision
making. In some cases even net directionality may be hard to establish. In cases where prior research has
indicated directionality, if not magnitude, such information should be used to explore supplementary
policy approaches to prevent positive leakage at the source or to control it where it occurs. In addition,
the Agency should be alert to leakage that may occur in other media (e.g., fertilizer runoff into
waterways) and the need for targeted policies to prevent or abate it. However, consistency with the
treatment of fossil fuels is something to consider (discussed below).
7

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Time scale
The Framework seeks to determine annual changes in emissions and sequestration rather than assessing
the manner in which these changes will impact the climate over longer periods of time. In so doing, it
does not consider the different ways in which use of bioenergy impacts the carbon cycle and global
temperature over different time scales. Nor does it consider temporal differences of climate effects on
the environment. Some recent studies have shown some intertemporal tradeoffs that should be
highlighted for policymakers. In the short/medium run, at the stand level, there can be a lag time
between emissions (through combustion) and sequestration (through regrowth) with the use of forest
biomass. At the landscape level, there can be concurrent debts and credits with harvesting and planting.
The impacts of the temporal pattern on climate response depend on the framework used. Some modeling
exercises have shown that the probability of limiting warming to or below 2°C in the twenty-first
century is dependent upon cumulative emissions by 2050 (Meinshausen et al. 2009). This suggests that
an early phase of elevated emissions from forest biomass could reduce the odds of limiting climate
warming if warming is limited to 2°C. Other modeling exercises by the same research team have shown
that in long run scenarios (100 years or more) in which total emissions were fixed, climate response is
relatively insensitive to the emissions pathway (Allen et al. 2009). Other studies have shown that
harvesting of biomass for bioenergy may have minimal effect on peak warming if regrowth is sufficient
to compensate for carbon losses that accompany harvest on a cumulative basis (NRC 2011; Cherubini et
al. 2012). This suggests that an intervention in forests or farming that results in an increase in storage of
carbon or emissions reductions that endures longer than 100 years (or be "permanent") may reduce the
peak climate response. Conversely, interventions that reduce storage of carbon or increase emissions for
longer than approximately 100 years may have a negative effect on peak warming response. The
accumulation of live plant, dead matter, and soil carbon should not be assumed to occur automatically or
be permanent; rather growth and accumulation should be monitored and evaluated for changes resulting
from management, market forces or natural causes.
If the climate effect of biogenic feedstocks is explored, the degree to which biogenic feedstocks curtails
fossil fuel use should be assessed and quantified. Given the slow response of the carbon and climate
system, if biogenic feedstocks displace the use of fossil fuels for longer than 100 years, then there may
be a beneficial climate effect. In contrast, if the use of biogenic feedstocks does not displace fossil fuels,
then any presumed beneficial climate consequences of biogenic carbon may be overestimated.
Spatial Scale
The use of unspecified "regions" as fuelsheds in combination with a reference year baseline is a central
weakness of the Framework with respect to forest-derived feedstocks. The EPA used a variable for the
Level of Atmospheric Reduction (LAR) to capture the proportion of potential gross emissions that are
offset by sequestration during feedstock growth, however the calculation of LAR captures landscape
wide changes rather than facility-specific carbon emissions associated with actual fuelsheds. As a result,
the estimates of the BAFs are sensitive to the choice of the spatial region as shown in the Agency's own
case study.
Implementation details
8

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
The EPA's Framework was lacking in implementation details. Implementation is crucial and some of
the Agency's current proposals will be difficult to implement. Data availability and quality, as well as
procedural details (e.g., application process, calculation frequency) are important considerations for
assessing the feasibility of implementation and scientific accuracy of results.
Consistency with fossil fuel emissions accounting
For comparability, there should be consistency between fossil fuel and biogenic emissions accounting.
Fossil fuel feedstock emissions accounting from stationary sources under the Clean Air Act are not
adjusted for offsite greenhouse gas emissions and carbon stock changes. Unlike fossil fuels, however,
biogenic feedstocks have carbon sequestration that occurs within a relevant timeframe. While EPA's
primary goal is to account for this offsetting sequestration, its biogenic emissions accounting should be
consistent with emissions accounting for fossil fuels for other emissions accounting categories—
including losses, international leakage, and fossil fuel use during feedstock extraction, production and
transport. Including some accounting elements for biomass and not for fossil fuels would be a policy
decision without the underlying science to support it.
Recommendations for Revising BAF
The Agency faces daunting technical challenges if it wishes to implement the Framework's facility-
specific BAF approach. If the EPA decides to retain and revise a facility-specific Framework, the SAB
recommends consideration of the following improvements.
• Develop a separate BAF equation for each feedstock category as broadly categorized by type,
region, prior land use and current management practices. Feedstocks could be categorized into short
rotation dedicated energy crops, crop residues, forest residues, municipal solid waste, trees/forests
with short accumulation times, trees/forests with long accumulation times and agricultural residue,
wood mill residue and pulping liquor.
o For long-accumulation feedstocks like roundwood, use an anticipated baseline approach to
compare emissions from increased biomass harvesting against a baseline without increased
biomass demand. For long rotation woody biomass, sophisticated modeling is needed to capture
the complex interaction between electricity generating facilities and forest markets and landscape
level effects, in particular: market driven shifts in planting, management and harvests; induced
displacement of existing users of biomass; land use changes, including interactions between
agriculture and forests; and the relative contribution of different feedstock source categories
(logging residuals, pulpwood or roundwood harvest).
o For residues, consider alternate fates (e.g., some forest residues may be burned if not used for
bioenergy) and information about decay. An appropriate analysis using decay functions would
yield information on the storage of ecosystem carbon in forest residues.
o For materials diverted from the waste stream, consider their alternate fate, whether they might
decompose over a long period of time, whether they would be deposited in anaerobic landfills,
whether they are diverted from recycling and reuse, etc. For feedstocks that are found to have
relatively minor impacts, the Agency may need to weigh ease of implementation against
scientific accuracy. After calculating decay rates and considering alternate fates, including
9

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
avoided methane emissions, the Agency may wish to declare certain categories of feedstocks
with relatively low impacts as having a very low BAF or setting it to 0 or possibly negative
BAFs in the case where methane emissions are avoided.
Incorporate various time scales and consider the tradeoffs in choosing between different time scales.
For all feedstocks, consider information about carbon leakage to determine its directionality as well
as leakage into other media.
10

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Default BAFs
Given the conceptual and scientific deficiencies of the Framework described above, and the prospective
difficulties with implementation, the SAB recommends consideration of default BAFs by feedstock
category and region. Under EPA's Framework, facilities would use individual BAFs designed to capture
the incremental carbon cycle and net emissions effects of their use of a biogenic feedstock. With default
BAFs by feedstock category, facilities would use a weighted combination of default BAFs based on
their particular bundle of feedstocks. The defaults are likely to be more scientifically robust in that they
could rely on readily available data and reflect landscape and aggregate demand effects, including
previous land use. Default BAFs might also vary by region and current land management practices due
to differences these might cause in the interaction between feedstock production and the carbon cycle.
The defaults would also have administrative advantages in that they would be easier to implement and
update. Default BAFs for each category of feedstocks would differentiate among feedstocks using
general information on their role in the carbon cycle. An anticipated baseline would allow for
consideration of prior land use, management, alternate fate (what would happen to the feedstock if not
combusted for energy) and regional differences. They would be applied by stationary facilities to
determine their quantity of biogenic emissions that would be subject to the EPA's Tailoring Rule.
Facilities could also be given the option of demonstrating a lower BAF for the feedstock they are using.
This would be facilitated by making the BAF calculation transparent and based on data readily available
to facilities. Properly designed, a default BAF approach could provide incentives to facilities to choose
feedstocks with the lower greenhouse gas impacts.
The SAB also explored certification systems as a possible way to obviate the need to quantify a specific
net change in greenhouse gases associated with a particular stationary facility. Carbon accounting
registries have been developed to account for and certify CO2 emissions reductions and sequestration
from changes in forest management. Theoretically, for the EPA's purposes, a certification system could
be tailored to account for emissions of a stationary facility after a comprehensive evaluation. Ultimately,
the SAB concluded that it could not recommend certification without further evaluation because such
systems could also encounter many of the same data, scientific and implementation problems that
bedevil the Framework.
Conclusion
With the increasing threat of global climate change, it is important to have scientifically sound methods
to account for greenhouse gas emissions caused by human activities. As the Agency has recognized, the
greenhouse gas implications of bioenergy are more complex and subtle than the greenhouse gas impacts
of fossil fuels. Unlike fossil fuels, forests and other biological feedstocks can grow back and sequester
CO2 from the atmosphere. Given the complicated role that bioenergy plays in the carbon cycle, the
Framework was written to provide a structure to account for net CO2 emissions. The Framework is a
step forward in considering biogenic carbon emissions.
The focus of the Framework is on point source emissions from stationary facilities with the goal of
accounting for any offsetting carbon sequestration that may be attributed to the facility's use of a
biogenic feedstock. To create an accounting structure, the Agency drew boundaries narrowly in
accordance with its regulatory domain. These narrow regulatory boundaries are intended to account for
biogenic carbon uptake and release associated with biomass that is combusted for energy purposes. As
11

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
such, this Framework does not consider, nor is it intended to consider, all greenhouse gas emissions
associated with the production and use of biomass energy. Comprehensive accounting for both biogenic
and fossil fuels would extend through time and space to estimate the long-term impacts on net
greenhouse gas emissions. To estimate net impact that can be attributed to bioenergy, the EPA would
need to calculate the net change in global emissions over time resulting from increased use of biomass
feedstocks as compared to a future without increased use of biogenic feedstocks. To capture this
difference, the boundaries of analysis would need to include all factors in the life cycle of the feedstock
and its products although computing global emissions changes for individual facilities has its own
daunting challenges.
The boundaries imposed by EPA's regulatory authority necessarily restrict its policy choices, however
economic research has shown that the most cost-effective way to reduce greenhouse gas emissions (or
any other pollution) is to regulate or tax across all sources until they face equal marginal costs. Given
the Agency's authority under the Clean Air Act, the most cost-effective economy-wide solution is not
within its menu of choices. The Agency's regulation of stationary sources does not include other users
of biomass (e.g. consumers of ethanol) that also have impacts on the carbon cycle as well as downstream
consumers of products produced by these facilities. Note that EPA can only regulate end-of-stack
emissions and thus has to design a system that fits within its regulatory authority.
The Agency has taken on a difficult but worthy task and forced important questions. Practical
considerations will, no doubt, weigh heavily in the Agency's calculus. In fact, any method that might be
adopted or considered, including methods proposed by the SAB, should be subject to an evaluation of
the costs of compliance and the carbon emissions savings likely to be achieved as compared to both a
categorical inclusion and a categorical exclusion. Uncertainties in the assessment of both the costs and
the emissions savings should be analyzed and used to inform the choice of policy. In this Advisory, the
SAB offers suggestions for how to improve the Framework while encouraging the Agency to think
about options outside its current policy menu. While the task of accounting for biogenic carbon
emissions defies easy solutions, it is important to assess the strengths and limitations of each option so
that a more accurate carbon footprint can be ascribed to the various forms of bioenergy.
12

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
2. INTRODUCTION
Greenhouse gas (GHG) emissions from the largest stationary sources became subject to regulation under
the Prevention of Significant Deterioration (PSD) and Title V Operating Permit Programs of the Clean
Air Act in January 2011. To target these regulations, EPA enumerated specific conditions under which
these Clean Air Act permitting requirements would apply. Initially, only sources currently subject to the
PSD permitting program or Title V (i.e., those that are newly-constructed or modified in a way that
significantly increases emissions of a pollutant other than greenhouse gases) would be subject to
permitting requirements for their greenhouse gas emissions. For these projects, only greenhouse gas
emission increases of 75,000 tons per year (tpy) or more, on a carbon dioxide equivalent (C02e) basis,
would be subject to technology requirements under the PSD program. As of July 1, 2011, more facilities
became subject to regulation based on their greenhouse gas emissions. Specifically, new and existing
stationary sources (that are not already covered by the PSD or Title V programs) that emit greenhouse
gas emissions of at least 100,000 tpy are subject to greenhouse gas regulation even if they do not exceed
the permitting thresholds for any other pollutant. For these facilities, the PSD and Title V requirements
would be triggered. The PSD program imposes "best available control technology" requirements to
control greenhouse gas emissions. Title V generally does not impose technology requirements but rather
requires covered facilities to report an overall compliance plan for meeting the requirements of the
Clean Air Act.
EPA's staged-approach to regulating greenhouse gases from stationary sources sought to focus on the
nation's largest greenhouse gas emitters and hence "tailored" the requirements of these Clean Air Act
permitting programs to cover power plants, refineries, and cement production facilities that meet certain
conditions while exempting smaller sources like farms, restaurants, schools and other facilities. The
question before the agency, and hence, the motivation for this SAB review, is whether and how to
consider biogenic greenhouse gas emissions in determining whether facilities meet certain thresholds (as
defined above) for Clean Air Act permitting. Biogenic CO2 emissions from bioenergy are generated
during the combustion or decomposition of biologically based material.
It is in this context that the EPA Office of Air and Radiation requested the EPA's Science Advisory
Board (SAB) to review and comment on its Accounting Framework for Biogenic CO 2 Emissions from
Stationary Sources (Framework, September 2011). The Framework considers the scientific and
technical issues associated with accounting for emissions of biogenic carbon dioxide (CO2) from
stationary sources and develops a framework to adjust the stack emissions from stationary sources using
bioenergy based on the induced changes in carbon stocks on land (in soils, plants and forests). Because
of the unique role of biogenic feedstocks in the overall carbon cycle, EPA deferred for a period of three
years the application of permitting requirements to biogenic CO2 emissions from bioenergy and other
biogenic stationary sources. In its deferral, EPA committed to conduct a detailed examination of the
science and technical issues associated with biogenic CO2 emissions and submit its study for review by
the Science Advisory Board. To conduct the review, the SAB Staff Office formed the Biogenic Carbon
Emissions Panel with experts in forestry, agriculture, greenhouse gas measurement and inventories, land
use economics, ecology, climate change and engineering.
13

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
The SAB was asked to review and comment on (1) the agency's characterization of the science and
technical issues relevant to accounting for biogenic CO2 emissions from stationary sources; (2) the
agency's framework, overall approach, and methodological choices for accounting for these emissions;
and (3) options for improving upon the framework for accounting for biogenic CO2 emissions (See
Appendix A: Charge to the SAB Panel).
The Biogenic Carbon Emissions Panel held a face-to-face meeting in October 25 - 27, 2011, and
teleconferenced four times during 2012. During the course of deliberations, the SAB Panel reviewed
background materials provided by the Office of Air and Radiation and heard from numerous public
commenters. This Advisory provides the SAB's main conclusions. Highlights are provided in the
Executive Summary and detailed responses to the individual charge questions are provided in the body
of the report.
14

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
3. RESPONSES TO EPA's CHARGE QUESTIONS
3.1. The Science of Biogenic CO? Emissions
Charge Question 1: In reviewing the scientific literature on biogenic CO2 emissions, EPA
assessed the underlying science of the carbon cycle, characterizedfossil and biogenic carbon
reservoirs, and discussed the implications for biogenic C02 accounting.
Does the SAB support EPA's assessment and characterization of the underlying science and the
implications for biogenic C02 accounting?
EPA has done an admirable job of reviewing the science behind the carbon cycle and greenhouse gas
emissions and their relationship to climate change, extracting some of the critical points that are needed
to create the proposed Framework. At the same time, there are several important scientific issues that are
not addressed in the EPA document, as well as scientific issues that are briefly discussed but not
sufficiently explored in terms of how they relate to the Framework. In the following section, the SAB
describes a series of deficiencies with the EPA characterization of the science behind biogenic CO2
accounting and suggests some areas where the science could be strengthened.
Time scale
One fundamental deficiency in the EPA report is the lack of discussion of the different time scales
inherent in the carbon cycle and the climate system that are critical for establishing an accounting
system. This is a complicated subject because there are many different time scales that are important for
the issues associated with biogenic carbon emissions. At the global scale, there are multiple time scales
associated with mixing of carbon throughout the different reservoirs on the Earth's surface. When
carbon dioxide is released into the air from burning fossil fuels, roughly 45% stays in the air over the
course of the following year. Of the 55% that is removed, roughly half is taken up by the ocean, mostly
in the form of bicarbonate ion, and the other half is taken up by the terrestrial biosphere, primarily
through reforestation and enhanced photosynthesis. The airborne fraction (defined as the fraction of
emissions that remains in the air) has been remarkably constant over the last two decades.
There is considerable uncertainty over how the magnitude of ocean and terrestrial uptake will change as
the climate warms during this century. If the entire ocean were to instantly reach chemical equilibrium
with the atmosphere, the airborne fraction would be reduced to 20 to 40% of cumulative emissions, with
a higher fraction remaining in scenarios with higher cumulative emissions. In other words, the ocean
chemical system by itself cannot remove all the CO2 released in the atmosphere. Because carbon uptake
by the ocean is limited by the rate of mixing between the shallow and deeper waters, this complete
equilibration is expected to take thousands of years. Over this century, if global CO2 emissions continue
to rise, most models predict that ocean uptake will stabilize between 3 to 5 gigatons per year (GtC/y),
implying that the fraction of emissions taken up by the ocean will decrease. For the terrestrial biosphere,
there is a much wider envelope of uncertainty; some models predict that CO2 uptake will continue to
keep pace with the growth in emissions, while other models suggest that CO2 uptake will decline, even
becoming a net source of CO2 to the atmosphere if processes such as release of carbon from the tundra
or aridification of the tropics were to occur.
15

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Over the time scale of several thousand years, once ocean equilibration is complete and only 20 to 40%
of cumulative emissions remains in the atmosphere, dissolution of carbonate rocks on land and on the
ocean floor will further reduce the airborne fraction to 10 to 25% over several thousand years to ten
thousand years. Excess anthropogenic CO2 emissions will stay in the atmosphere for more than 100,000
years, slowly drawn down by silicate weathering that converts the CO2 to calcium carbonate, as well as
slow burial of organic carbon on the ocean floor. The size of this "tail" of anthropogenic CO2 depends
on the cumulative emissions of CO2, with higher cumulative emissions resulting in a higher fraction
remaining in the atmosphere.
Another important time scale for considering accounting systems for biogenic carbon emissions is the
period over which the climate responds to carbon dioxide and other greenhouse gases. The importance
of the timing of emissions depends on whether one uses a global warming limit or a cumulative
emissions limit. Some modeling exercises have shown that the probability of limiting warming to 2 °C
or below in the twenty-first century is dependent upon cumulative emissions by 2050 (Meinshausen et
al. 2009). This suggests that an early phase of elevated emissions from forest biomass could reduce the
odds of limiting climate warming if warming is limited to 2 °C. Another climate modeling study has
demonstrated that peak warming in response to greenhouse gas emissions is primarily sensitive to
cumulative greenhouse gas emissions over a period of roughly 100 years, and, so long as cumulative
emissions are held constant, is relatively insensitive to the emissions pathway within that time frame
(Allen et al. 2009). What this means is that an intervention in forests or farming that results in either an
increase or decrease in storage of carbon or emissions reductions must endure longer than 100 years to
have an influence on the peak climate response as long as cumulative emissions from all sources are
constant. Conversely, if these changes last less than 100 years, harvesting of biomass for bioenergy
resulting in release of carbon dioxide will have a relatively small effect on peak warming. While the
harvesting of trees for bioenergy can result in a carbon debt even at the landscape level (Mitchell et al.
2012), this may not reflect potential climate benefits at longer time scales if biomass is regrown
repeatedly and substituted for coal over successive harvest cycles (Galik and Abt 2012).
Time scales are also important for individual feedstocks and their regeneration at a more local scale.
Given that the EPA's objective is to account for the atmospheric impact of biogenic emissions, it is
important to consider the turnover times of different biogenic feedstocks in justifying how they are
incorporated into the Framework. The fundamental differences in stocks and their turnover times as they
relate to impacts on the atmosphere are not well discussed or linked. If a carbon stock is cycling quickly
on land and regrowth is sufficient to compensate for carbon losses from harvesting, it may have a
beneficial impact when it displaces fossil fuel over successive cycles of growth and harvest (assuming
this temporal displacement exceeds 100 years). If the carbon stock, or some part of it, turns over more
slowly, if regrowth is not assured or if feedstocks are not being used to continuously displace fossil
fuels, the impact on climate worsens.
There is a continuum of carbon stock size and turnover among the biogenic feedstock sources included
in the Framework, but there is little background discussion of the variation in stock and turnover and
how that informs the accounting method. The Framework sets up categories of feedstocks based on their
source, but these groupings do not translate into differential treatment in the Framework. The science
section could walk through the carbon stocks covered by the scope of the Framework and their relevant
turnover times.
16

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Appendix B discusses a set of studies by Cherubini and co-authors (Cherubini et al. 2011; 2012) that
provide examples for estimating the temporal distribution of atmospheric impacts from biomass
harvesting by framing the analysis in terms of global warming potentials (GWPs) and global
temperature potentials (GTPs) for harvested biomass. Figure B-l in Appendix B, adapted from
Cherubini et al. (2012), depicts mean surface temperature changes for a simple contrived comparison of
biogenic emissions from a single stand over hundreds of years as compared to comparable fossil
emissions. While much is assumed regarding global activity (emissions, landscape responses,
investment behavior), Figure B-l demonstrates the importance of the time horizon and the weight to
place on temperature increases that occur in the short term versus temperature increases that occur later.
As shown in Figure B-l, a 50-year time horizon (or less) would obscure the longer-term climate
consequences of bioenergy. The GTPbio value would continue to decline for time horizons beyond 100
years since there is no net temperature increase after 100 years. The choice of weighting of temperature
effects at different time horizons could be influenced by the estimated damages associated with the
temperature increases as well as the social rate of time preference for avoiding damages. The discussion
by Kirschbaum (2003, 2006) of the impact of temporary carbon storage (the inverse of temporary carbon
release from biomass harvesting for bioenergy) points out that the exact climate impact of temporary
CO2 storage (or emissions) depends on the type of impact, as some depend on peak temperature,
whereas others, such as melting of polar ice sheets, depend more on time-averaged global temperature.
There is no scientifically correct answer when choosing a time horizon, although the Framework should
be clear about what time horizon it uses, and what that choice means in terms of valuing long term
versus shorter term climate impacts.
Disturbance
Because ecosystems respond in complicated ways to disturbances (e.g., harvesting, fire) over long
periods of time, and with a high degree of spatial heterogeneity, the state of knowledge about
disturbance and impacts on carbon stocks and turnover should be reviewed within the context of
relevant time scales and spatial extents. This is highly relevant to producing accurate estimates of
biogenic emissions from the land. There is also insufficient treatment given to the existing literature on
the impact of different land management strategies on soil carbon, which is important for understanding
how carbon stocks may change over many decades.
Non-CO2 Greenhouse Gases
The Framework does not incorporate greenhouse gases other than CO2. Ideally both fossil fuels and
biogenic fuels should be subject to the same emissions accounting to fully capture the difference
between the two types of fuels in terms of their greenhouse gas emissions. For biogenic feedstocks, the
most important source of non-CC>2 emissions is likely to be N2O produced by the application of fertilizer
(Crutzen, et al. 2007). In particular, if the biomass feedstock is from an energy crop that results in
different N2O emissions vis-a-vis other crops, should this be counted? Is it negligible? This issue is not
introduced in the science section. N2O is relatively long-lived (unlike methane) and therefore the climate
impacts of heavily fertilized biomass (whether in forests or farms) are greater than non-fertilized
biomass. There is a substantial literature on N2O from fertilizer use that was not discussed in the
Framework. If the decision to not count non-CC>2 greenhouse gases stems from a need to render the
carbon accounting for biogenic sources parallel with fossil fuels, this needs to be explicitly discussed.
17

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
3.2. Biogenic CO? Accounting Approaches
Charge Question 2: In this report, EPA considered existing accounting approaches in terms of
their ability to reflect the underlying science of the carbon cycle and also evaluated these
approaches on whether or not they could be readily and rigorously applied in a stationary source
context in which onsite emissions are the primary focus. On the basis of these considerations,
EPA concluded that a new accounting framework is neededfor stationary sources.
2(a). Does the SAB agree with EPA's concerns about applying the IPCC national approach to
biogenic CO2 emissions at individual stationary sources?
Yes. The IPCC national approach is an inventory of global greenhouse emissions (i.e., all emissions are
counted). It is comprehensive in quantifying all emissions sources and sinks, but does not describe
linkages among supply chains. In other words, it is essentially a "production-based inventory" or
"geographic inventory" rather than a "consumption-based inventory" (Stanton et al. 2011). Moreover, it
offers a static snapshot of emissions at any given time, but it does not expressly show changes in
emissions over time. As such, the IPCC national approach does not explicitly link biogenic CO2
emission sources and sinks to stationary sources, nor does it provide a mechanism for measuring
changes in emissions as a result of changes in the building and operation of stationary sources using
biomass.
2(b). Does the SAB support the conclusion that the categorical approaches (inclusion and
exclusion) are inappropriate for this purpose, based on the characteristics of the carbon cycle?
A decision about a categorical inclusion or exclusion will likely involve many considerations that fall
outside the SAB's scientific purview, such as legality, feasibility and, possibly, political will. The SAB
cannot speak to the legal or full implementation difficulties that could accompany any policy on
biogenic carbon emissions but some scientific observations that may inform the Administrator's policy
decision are offered below.
The notion that biomass is carbon neutral arises from the fact that the carbon released as C02upon
combustion was previously removed from the atmosphere as CO2 during plant growth. Thus, the
physical flow of carbon in the biomass combusted for bioenergy represents a closed loop that passes
through a stationary source. Under an accounting framework where life cycle emissions associated with
the production and use of biomass are attributed to a stationary source, assuming carbon neutrality of
biomass implies that the net sum of carbon emissions from all sources and sinks is zero, including all
supply chain and market-mediated effects. Carbon neutrality cannot be assumed for all biomass energy a
priori (Rabl et al. 2007; E. Johnson 2009; Searchinger et al. 2009). There are circumstances in which
biomass is grown, harvested and combusted in a carbon neutral fashion but carbon neutrality is not an
appropriate a priori assumption; it is a conclusion that should be reached only after considering a
particular feedstock production and consumption cycle. There is considerable heterogeneity in feedstock
types, sources, production methods and leakage effects; thus net biogenic carbon emissions will vary
considerably.
18

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Given that some biomass combustion could have positive net emissions, a categorical exclusion would
remove any responsibility on the stationary source for CO2 emissions from its use of biogenic material
from the entire system (i.e., the global economy) and provide no incentive for the development and use
of best management practices. Conversely, a categorical inclusion would provide no incentive for using
biogenic sources that compare favorably to fossil energy in terms of greenhouse gas emissions.
The commentary above merely reflects some scientific considerations. The SAB recognizes that, in
reality, the EPA may face difficult tradeoffs between ease of implementation and other goals (e.g.,
maximizing scientific accuracy by modeling the decomposition of logging residues). While an
alternative approach of default Biogenic Accounting Factors (BAFs) is offered for the agency's
consideration (see Section 4), the SAB cannot advise the Agencyon the legal feasibility of any approach.
2(c). Does the SAB support EPA's conclusion that a new framework is neededfor situations in
which only onsite emissions are consideredfor non-biologically-based (i.e., fossil) feedstocks?
Through discussions with the Panel at the public meeting, the EPA agreed that this question is redundant
with other charge questions and therefore does not require a separate response.
2(d). Are there additional accounting approaches that could be applied in the context of biogenic
CO2 emissions from stationary sources that should have been evaluated but were not?
Several other agencies are developing methods for assessing greenhouse gas emissions by facilities that
could inform the approach developed by the EPA. These include the DOE 1605(b) voluntary greenhouse
gas registry targeted to entities, which has many similar characteristics to the approach proposed by EPA
for stationary sources. There is also the Climate Action Registry developed in California that uses a
regional approach to calculate baselines based on inventory data and may inform the delineation of
geographic regions and choice of baselines in the EPA approach. USDA also is developing in parallel an
accounting approach for forestry and agricultural landowners. It would be beneficial if the EPA and
USDA approaches could be harmonized to avoid conflicts and take advantage of opportunities for
synergy.
19

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
3.3. Methodological Issues
Charge Question 3: EPA identified and evaluated a series of factors in addition to direct biogenic
CO2 emissions from a stationary source that may influence the changes in carbon stocks that
occur offsite, beyond the stationary source (e.g., changes in carbon stocks, emissions due to land-
use and land management change, temporal and spatial scales, feedstock categorization) that are
related to the carbon cycle and should be considered when developing a framework to adjust total
onsite emissions from a stationary source.
3(a). Does SAB support EPA's conclusions on how these factors should be included in accounting
for biogenic CO2 emissions, taking into consideration recent advances and studies relevant to
biogenic CO2 accounting?
The SAB's response to this question differs by feedstock. On balance, the Framework includes many
important factors but some factors suffer from significant estimation and implementation problems.
For agricultural feedstocks, the factors identified by EPA to adjust the CO2 emissions from a stationary
source for direct off-site changes in carbon stocks are appropriate but suffer from significant estimation
and implementation problems. The Framework includes factors to represent the carbon embodied in
products leaving a stationary source, the proportion of feedstock lost in conveyance, the offset
represented by sequestration, the site-level difference in net carbon flux as a result of harvesting, the
emissions that would occur "anyway" from removal or diversion of non-growing feedstocks (e.g., corn
stover) and other variables. In some cases, energy crops like miscanthus and switchgrass have
significant potential to sequester carbon in the soil and be sinks for carbon rather than a source
(Anderson-Teixeira et al. 2009). In other cases, the production of bioenergy could result in by-products
like biochar which sequester significant amounts of carbon. A large value of the SITE TNC and/or
SEQP variables in the accounting equation could result in a negative BAF for such feedstocks. The
Framework should clarify how a negative BAF would be used and whether it could be used by a facility
to offset fossil fuel emissions. Restricting BAF to be non-negative would reduce incentives to use
feedstocks with a large sequestration potential.
For waste materials (municipal solid waste, manure, wastewater, construction debris, etc.), the
Framework assigns a BAF equal to 0 for biogenic CO2 released from waste decay at waste management
systems, waste combustion at waste incinerators or combustion of captured waste-derived CH4. The
Framework further states that for any portion of materials entering a waste incinerator that is harvested
for the purpose of energy production at that incinerator, biogenic CO2 emissions from that material
would need to be accounted according to the Framework calculations. Municipal solid waste biomass is
either disposed of in a landfill or combusted in facilities at which energy is recovered. Smaller amounts
of certain waste components (food and yard waste) may be processed by anaerobic digestion and
composting. The SAB concurs with the Framework that the CO2 released from the decomposition of
biogenic waste in landfills, compost facilities or anaerobic digesters could reasonably be assigned a BAF
of 0. In addition, given that methane (CH4) is a more potent greenhouse gas than CO2, the Framework
should account for CH4 emissions from landfills in cases where the methane is not captured. The SAB
recognizes that EPA may address methane in other regulatory contexts.
20

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
When accounting for emissions from waste sources including logging residue, wood mill waste and
pulping liquor, the EPA should recognize that these emissions are part of a larger system where they can
be co-products with commercial products. For logging residue, wood mill waste and pulping liquor the
larger system includes forests, solid wood mills, pulp mills and stationary energy sources. Accounting
for greenhouse gases in the larger system needs to track all biomass emissions or forest stock changes
and needs to assure they are allocated over time across the outputs (product and co-products) from the
system so as to account for all fluxes. Within the larger system, the allocation of fluxes to wood/paper
products or to emissions from a stationary source can be supported by scientific reasoning but is
ultimately a policy decision. The Agency should consider how the Framework meets the scientific
requirement to account for (allocate) all emissions to products and co-products across the larger system
of forest, mills and stationary sources over time.
For roundwood, the calculation of BAF would need to account for the time path of carbon accumulation
and emissions from logging residue and apply a landscape perspective. The landscape perspective is
important because of simultaneous management decisions that emit and sequester greenhouse gases
concurrently and therefore define the net implications over time. The Framework recognizes some of the
challenges associated with defining the spatial and temporal time scale and in choosing the appropriate
baseline. Ultimately, however, the Framework chooses an approach that disregards any consideration of
the time scales over which biogenic carbon stocks are accumulated or depleted and does not actually
estimate carbon stock changes associated with biomass use. Instead the Framework attempts to
substitute a spatial dimension for time and creates an accounting system that generates outcomes
sensitive to the regional scale at which carbon emissions attributed to a stationary source are evaluated.
Below are some comments on particular factors.
Level of Atmospheric Reduction (LAR): The term refers to the proportional atmospheric carbon
reduction from sequestration during feedstock regrowth (GROW) or avoided emissions (AVOIDEMIT)
from the use of residues that would have been decomposed and released carbon emissions "anyway."
The scientific justification for constraining the range of LAR to be greater than 0 but less than 1 is not
evident since it is possible for feedstock production to exceed feedstock consumption. These two terms
are not applicable together for a particular feedstock and representing them as additive terms in the
accounting equation can be confusing. Additionally, the value of LAR for forest biomass is sensitive to
the size of the region for which growth is compared to harvest.
Loss (L): This term is included in the Framework to explicitly adjust the area needed to provide the total
feedstock for the stationary facility. It is a term used to include the emissions generated by the feedstock
lost during storage, handling and transit based on the strong assumption that most of the carbon in the
feedstock lost during transit is immediately decomposed. It is therefore important to separate the use of
this Loss term for estimating the area needed to provide the feedstock and for estimating the carbon
emissions released by the operation of the stationary source. To more accurately estimate the actual loss
of carbon due to these losses, one would need to model the carbon storage and fluxes associated with the
feedstock lost, which are likely to be a function of time. The number of years considered would be a
policy decision; the longer the period, the larger the proportion of loss that would be counted. The
Framework tacitly assumes an infinitely long horizon that results in the release of all the carbon stored
in the lost feedstock.
21

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Products (PRODC): The removal of products from potential gross emissions is justified scientifically;
however, the scientific justification for treating all products equally in terms of their impact on
emissions is not clear. For some products (e.g., ethanol and paper), the stored carbon will be released
rapidly while for other products, such as furniture, it might be released over a longer period of time. The
Framework implicitly assumes that all products have infinite life-spans, an assumption without
justification or scientific foundation. For products that release their stored carbon rapidly, the
consequences for the atmosphere are the same as for combustion of the feedstock. To precisely estimate
the stores of products so as to estimate the amount of carbon released, one would need to track the stores
as well as the fluxes associated with product pools. The stores of products could be approximated by
modeling the amount stored over a specified period of time.
A second way in which PRODC is used is as a means of prorating all area-based terms such as LAR,
SITE-TNC and Leakage. This is potentially problematic because it makes the emissions embodied in co-
products dependent on the choice of regional scale at which LAR is estimated. As the size of the region
contracts, LAR tends towards zero and the amount of gross emissions embodied in PRODC increases
and exacerbates the implications of the scale sensitivity of the LAR value.
Avoided Emissions (AVOIDEMIT): This term refers to transfers of emissions that would occur
"anyway" from removal or diversion of non-growing feedstocks like corn stover and logging residues.
In the Framework, feedstocks may be mathematically credited with avoided emissions if the residues
would have decayed "anyway." Specifically, AVOIDEMIT is added to GROW in the numerator in
determining the LAR or proportion of emissions that are offset by sequestration or avoided emissions.
As with the Loss term, there is an implicit assumption of instantaneous decomposition that appears to be
a simplifying assumption. While this may be a convenient assumption, it should be explained and
justified. To improve scientific accuracy, the EPA could explore some sample calculations (as described
below), taking into account regional differences in decay rates. Once this information is gathered and
analyzed, the EPA may then need to make a decision that weighs scientific accuracy against
administrative expediency and other factors.
Since the concept reflected in "avoided emissions" is actually "equivalent field-site emissions," it would
be clearer to refer to it this way since emissions are not so much avoided as they are shifted to another
venue. With residues left in the forest, some of the materials might take decades to fully decompose. For
accuracy, the hypothetical store of carbon would have to be tracked. To approximate these stores, one
could compute the average amount of carbon remaining after a period of years.
The scientific theory behind losses and stores of ecosystem carbon was developed by Olson (1963) and
could be applied to the fate of residues and slash in both forest and agricultural systems. The store of
carbon in an ecosystem depends upon the amount of carbon being input (I) and the proportion of carbon
lost per time unit, referred to as the rate-constant of loss (k). Specifically the relationship is I/k. In the
case of residues or slash that are burned in the field or in a bioenergy facility, the store of carbon is
essentially zero because most of the input is lost within a year (k> 4.6 per year assuming at least 99% of
the material is combusted within a year). On the other hand, if the residue or slash does not lose its
carbon within a year, the store of carbon would be greater than zero and, depending on the interval of
residue or slash creation, could be greater than the initial input. Appendix C provides more information
22

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
on the fate of residue after harvest and landscape storage of carbon. For example, if slash is generated
every 25 years (1=100 per harvest area/25=4 per year) and the slash is 95% decomposed within 25 years
(k=0.12 per year), one cannot assume a store of zero because the average ecosystem store in this case
would actually be 33% of the initial input (4/0.12=33.3). If the input occurred every 5 years (1=100 per
harvest/5=20 per year) for the same decay rate-constant, then the average store would be 167% of the
initial input (20/0.12=167). Moreover, it cannot be assumed that because the rate-constant of loss (k) is
high, that the stores will always be low. That is because the input (I) is a function of the interval of
residue or slash generation; the shorter the interval of generation, the higher the effective input because a
higher proportion of the forest or agricultural system is contributing inputs. For example, if there is 1
unit of residue/slash generation per harvest, then an annual harvest on a system basis creates 1 unit of
material; if there is 1 unit of residue/slash generation per harvest, then a harvest every 10 years creates
an average harvest of 0.1 units (1 unit/10 years = 0.1 unit per year). This relationship means that if
residue or slash is generated annually and 95% is lost to decomposition in that period, then the forest
system could store 33% of the initial input (I/k=l/3). For the values of k usually observed in agricultural
setting (50%) per year), an annual input would lead to a store in excess of 145% of the initial input
(I/k=l/0.69). Burning of this material would cause a decrease in carbon stores analogous to that of
reducing mineral soil stores as accounted for in SITE TNC, but this loss is not accounted for in the
proposed Framework.
There are several ways in which losses from residue/slash decomposition could be used in the
Framework. One is to track the annual loss of carbon from decomposition. This would be analogous to
tracking the regrowth of feedstock annually, but in this case it would be the annual decomposition loss.
The annual decomposition loss would then be credited as equivalent to combustion as fuel. The
advantage of this system is that it would track the time course of release. The disadvantage is that it
increases transaction costs. An alternative based on a fuelshed (or other larger area) would be to
calculate the average fraction of residue or slash that would remain over the harvest interval and subtract
that from the amount harvested. The difference between the amount harvested and the amount that
would have remained is an index of the equivalent amount of release via decomposition. For example, if
10 metric tons of either residue or slash is created per year in a fuelshed and 65% of the slash would
have decomposed on average over a given harvest interval, then decomposition would have been
equivalent to a release of 65% of the amount of fuel used (6.5 metric tons). This would mean that 3.5
metric tons that would have been stored was lost by combustion; hence 6.5 metric tons would be
credited in the current calculation of LAR. However, if 35% of the slash would have decomposed on
average over the harvest interval, then use of 10 metric tons as fuel would reduce carbon stores of
residues and slash by 6.5 metric tons. This would result in a so-called "avoided emissions" credit of 3.5
metric tons.
In addition to considering actual decomposition losses, the Framework needs to consider the starting
point of residue and slash harvest. The carbon released by combustion will be a function of the starting
point, with systems that start with residues and slash having a different timeline of release than those
that newly create residue and slash. The former will have the release rate linearly related to the harvest
interval, whereas the latter will likely have a curvilinear relationship that is a function of the rate-
constant of loss (k).
23

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Instead of a simplifying assumption of instantaneous decomposition, a more accurate calculation could
be developed that determines a loss rate-constant appropriate to the material and climate to estimate the
amount of carbon that could have been stored had the material not been burned. This amount could be
approximated by using the relationships developed by Olson (1963) and reducing the number of
calculations involved. When approximations are used, they should be checked against more precise
methods to determine the magnitude of possible approximation errors. Several mechanisms could be
used to simplify the estimation of these numbers, ranging from calculators that require entry of a few
parameters (e.g., average amount of residue or slash generated, the area of source material, the interval
of harvest) to look-up tables that are organized around the parameters used to generate them. While
there is some uncertainty regarding the loss rate-constants, these sorts of parameters are routinely used
in scientific assessments of the carbon cycle and their uncertainty is not much greater than any other
parameter required by the Framework.
The Framework should provide guidance on how logging residue will be distinguished from forest
feedstock since that will influence the BAF for that biomass and create incentives to classify as much
material as possible as residue and slash despite the fact that some of the "residue/slash" material such
as cull trees would be "regenerated" via feedstock regrowth.
Total Net Change in Site Emissions (SITE TNC): This term is the annualized difference in the stock of
land-based carbon (above and below ground, including changes in standing biomass and soil carbon)
that results on the site where the feedstock is produced.
The estimates of SITE TNC will be site-specific and will depend on the knowledge about previous
history of land use at that site, the specific agricultural or forestry management practices utilized and the
length of time over which they have been practiced. To the extent that the use of bioenergy leads to a
change in these practices relative to what would have been the case otherwise, it will be important to use
an anticipated baseline approach to determine the stock of land based carbon in the absence of bioenergy
and to compare that to the stock with the use of bioenergy. As discussed below in response to charge
question 4(f), this anticipated baseline could be developed at a regional or national scale and include
behavioral responses to market incentives. Alternatively, look-up tables could be developed based on
estimates provided by existing large scale models such as CENTURY or FASOM for feedstock based
and region specific SITC TNC estimates.
It should be noted that soil carbon sequestration is not a permanent reduction in CO2 emissions. The
Framework, however, treats permanent reductions in emissions, for example, due to a reduction in the
LOSS of biomass to be equivalent to reductions due to an increase in soil carbon sequestration which
could be temporary. Since soil carbon sequestration is easily reversible with a change in land
management practices, the implementation of this Framework will need to be accompanied by frequent
monitoring to determine any changes in soil carbon stocks and to update the BAF value for a facility.
Sequestration (SEQP): This term refers to the proportion of feedstock carbon embodied in post-
combustion residuals such as ash or biochar. Including sequestration in the Framework is appropriate;
however, the approach taken is subject to the same problems as those described for Products. There is no
scientific literature cited to support the idea that all the materials produced by biogenic fuel use do not
decompose. This is the subject of ongoing research, but it seems clear that these materials do
24

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
decompose. The solutions to creating a more realistic and scientifically justified estimate are the same as
for the Products term (see above).
Leakage (LEAK): The Framework includes this term for leakage but is silent on the types of leakage
that would be included and how leakage would be measured. EPA representatives said the Framework
did not provide a quantification methodology for leakage because assessing leakage requires policy- and
program-specific details that are beyond the scope of the report. However, there are several conceptual
and implementation issues that merit further discussion in the Framework.
The use of biogenic feedstocks could lead to leakage by diverting feedstocks and land from other uses
and affecting the price of conventional forest and agricultural products, which can lead to indirect land
use changes that release or increase carbon stored in soils and vegetation. The use of these feedstocks
could also affect the price of fossil fuels by lowering demand for them and increasing their consumption
elsewhere (also referred to as the rebound effect on fuel consumption); this would offset the greenhouse
gas savings from the initial displacement of fossil fuels by bioenergy (Chen and Khanna 2012). Leakage
effects will vary by feedstock and location and could be positive (if they lead to carbon emissions
elsewhere) or negative (if they lead to carbon uptake activities). As will be discussed in Section 3.4 [in
response to question 4(f)], the latter could arise, for example, if increased demand for biomass and
higher prices generate incentives for investment in forest management that increases forest carbon
sequestration. Some research has shown that when a future demand signal is strong enough, expectations
about biomass demand for energy (and thus revenues) can reasonably be expected to produce
anticipatory feedstock production changes with associated changes in land management and land-use
(e.g., Sedjo and Sohngen, in press, 2012). Thus price changes can lead to changes in consumption and
production decisions outside the boundary of the stationary source, even globally.
While the existence of non-zero leakage is very plausible, the appropriateness of attributing emissions
that are not directly caused by a stationary facility to that facility has been called into question
(Zilberman et al. 2011). While first principles in environmental economics show the efficiency gains
from internalizing externalities by attributing direct environmental damages to responsible parties, they
do not unambiguously show the social efficiency gains from attributing economic or environmental
effects (such as leakage) that occur due to price changes induced by its actions to that facility
(Holcombe and Sobel, 2001). Moreover, leakage caused by the use of fossil fuels is not included in
assessing fossil emissions generated by a stationary facility. Liska and Perrin (2009) show that military
activities to secure oil supplies from the Middle East lead to indirect emissions that could increase the
carbon intensity of gasoline. Thus, the technical basis for attributing leakage to stationary sources and
inherent inconsistency involved in including some types of leakage and for some fuels makes the
inclusion of leakage as a factor in the BAF calculation a subjective decision. Including some types of
leakage (for example, due to agricultural commodity markets) and not others (such as those due to the
rebound effect in fossil fuel markets) and for biomass and not fossil fuels would be a policy decision
without the underlying science to support it.
Empirically, assessing the magnitude of leakage is fraught with uncertainty. Capturing leakage would
entail using complex global economic models that incorporate production, consumption and land use
decisions to compare scenarios of increased demand for biogenic feedstocks with a baseline scenario
without increased demand. Global models that include trade across countries in agricultural and forest
25

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
products can aid in determining the leakage effects on land use in other countries. Global models of the
forestry sector include Sedjo and Sohngen (2012) and Ince et al. (2011). Existing models would need to
be expanded to include the multiple lignocellulosic feedstocks considered in this Framework that can
compete to meet demand for bioenergy to determine net leakage effects. Methods would then need to be
developed to assign leakage factors to individual feedstocks. The existing literature assessing the
magnitude of leakage from one use of a biogenic feedstock (corn ethanol) shows that its overall
magnitude in the case of leakage due to biofuel production is highly uncertain and differs considerably
across studies and within a study depending on underlying assumptions (Khanna et al. 2011; Khanna
and Crago 2012). Other feedstock-use combinations would also need to be evaluated. If the magnitude
of leakage is plagued with too much uncertainty, if possible, its direction should at least be stated and
recognized in making policy choices. Depending on the level of uncertainty, supplementary policies
might be possible to reduce leakage due to changes in land use, such as restrictions on the types of land
that could be used to produce the biogenic feedstocks and the types of biogenic feedstocks that could be
used to qualify for a BAF less than 1. Some of these implementation issues with estimating BAF and
leakage will be discussed further in Section 3.4.
3(b). Does SAB support EPA '.s distinction between policy and technical considerations
concerning the treatment of specific factors in an accounting approach?
A clear line cannot be drawn between policy and technical considerations in an accounting approach. In
fact, the lack of information on EPA's policy context and the menu of options made it more difficult to
fully evaluate the Framework. Because the reasonableness of any accounting system depends on the
regulatory context to which it is applied, the Framework should describe the Clean Air Act motivation
for this proposed accounting system, including how the Agency regulates point sources for greenhouse
gases and other pollutants. The document should make explicit the full gamut of Clean Air Act policy
options for how greenhouses gases could be regulated, including any potential implementation of carbon
offsets or certification of sustainable forestry practices. The Framework also should describe the EPA's
legal boundaries regarding upstream and downstream emissions. Technical considerations can influence
the feasibility of implementing a policy just as policy options can influence the technical discussion. The
two need to go hand in hand rather than be treated as separable.
The Framework explicitly states that it was developed for the policy context where it has been
determined that a stationary source emitting biogenic CO2 requires a means for "adjusting" its total
onsite biogenic emissions estimate on the basis of information about growth of the feedstock and/or
avoidance of biogenic emissions and more generally the carbon cycle. However, in the discussion on the
treatment of specific factors it states in several places that this treatment could depend on the program or
policy requirements and objectives. Certain open questions described as "policy" decisions (e.g., the
selection of regional boundaries, marginal versus average accounting, inclusion of working or non-
working lands, inclusion of leakage) made the evaluation of the Framework difficult. Clearly, the policy
context matters and the EPA's reticence in describing the policy context and in taking positions on open
questions (as well as lack of implementation details) meant that the Framework was inadequately
defined for proper review and evaluation.
Specifically, if the policy context is changed — for example, if carbon accounting is needed to support a
carbon cap and trade or carbon tax policy — then the appropriateness of the Framework would need to
26

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
be evaluated relative to alternative approaches such as life cycle analysis for different fuel streams.
Modifying how certain factors are measured or included may not be sufficient. In fact, a different
Framework would likely be needed if a national or international greenhouse gas reduction commitment
exists. Furthermore, the BAFs developed for regulating the emissions from stationary sources would
likely conflict with measures of greenhouse gas emissions from bioenergy used in other regulations such
as California's cap and trade system for regulating greenhouse gases.
Economic research has shown that the most cost-effective way to reduce greenhouse gas emissions (or
any other pollution) is to regulate or tax across all sources until they face equal marginal costs. The most
cost-effective solution would involve setting carbon limits (or prices) on an economy-wide basis and not
selectively for particular sources or sectors. Given the EPA's limited authority under the Clean Air Act,
the most efficient economy-wide solution is not within its menu of policy choices. EPA's regulation of
stationary sources will exclude other users of biomass also have equivalent impacts on the carbon cycle
as well as downstream emissions from consuming the products produced by these facilities. Note that
biogenic emissions accounting would still be an issue even under an economy-wide emissions policy.
3(c). Are there additional factors that EPA should include in its assessment? If so, please specify
those factors.
As stated above, for agricultural biomass from energy crops and crop residues, the factors included in
the Framework capture most of the direct off-site adjustments needed to account for the changes in
carbon stocks caused by a facility using agricultural feedstocks although they do not account for
leakage. However, an anticipated baseline is needed for soil carbon, residue disposition and land
management changes. For forest biomass, the Framework needs to incorporate the time path of carbon
accumulation in forests (after energy emissions from harvested roundwood) and forest investment and
multi-stand decisions. As discussed in Section 3.1, EPA should consider the time path of the "anyway"
emissions that would have occurred on the land if logging residue were not used for energy production
and weigh the benefits of scientific accuracy against the administrative simplicity of assuming
instantaneous decomposition. For municipal solid waste biomass, the Framework needs to consider
other gases and CH4 emissions from landfills. Given that methane emissions from landfills are
sometimes not captured, crediting waste material for avoided emissions of methane may be
inappropriate. As the Framework states, the carbon impact of using waste for energy production in
combustion facilities should nonetheless be subjected to a biogenic accounting framework. It should be
gauged relative to the CH4 emissions, if any, that would be released during decomposition in a landfill.
N2O emissions, especially from fertilizer use, should also be considered. Furthermore, the inclusion of
non-CC>2 greenhouse gases in general should be consistent between biogenic and fossil fuel accounting.
For instance, there are also transportation -related emissions losses in the delivery of natural gas.
3(d). Should any factors be modified or eliminated?
For reasons discussed above, factors such as PRODC, AVOIDEMIT and SEQP could be improved by
incorporating the time scale over which biomass is decomposed or carbon is released back to the
atmosphere. LAR needs to be modified to be scale insensitive and to address additionality. Factors can
be separated by feedstocks according to their relevance for accounting for the carbon emissions from
27

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
1	using those feedstocks. For example, GROW and leakage may not be relevant for crop and forest
2	residues.
3
28

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
3.4. Accounting Framework
Charge Question 4: EPA's Accounting Framework is intended to be broadly applicable to
situations in which there is a need to represent the changes in carbon stocks that occur offsite,
beyond the stationary source, or in other words, to develop a "biogenic accounting factor" (BAF)
for biogenic CO2 emissions from stationary sources.
Question 4(a). Does the Framework accurately represent the changes in carbon stocks that occur
offsite, beyond the stationary source (i.e., the BAF)?
For agricultural biomass, the variables in EPA's proposed equation for BAF represent the basic factors
necessary for estimating the offsite carbon change associated with stationary source biomass emissions,
including changes in storage of carbon at the harvest site. For short accumulation feedstocks, where
carbon accumulation and "anyway" emissions are within one to a few years (i.e., agricultural residues,
perennial herbaceous crops, mill wood wastes, other wastes), with some adjustments to address
estimation problems (including an anticipated baseline for soil carbon changes, residue disposition and
land management) and careful consideration of data and implementation, the Framework can accurately
represent carbon changes offsite. However, for long accumulation feedstocks where carbon
accumulation and "anyway" emissions occur over decades [i.e., wood harvested specifically for energy
use (roundwood) and logging residue], the Framework does not accurately account for changes in
carbon stocks offsite for several reasons discussed below in response to charge question 4(b).
The Framework also does not consider other greenhouse gases (e.g., N2O from fertilizer use and CH4
emissions from landfills). Excluding CH4 because it is not "CO2" is not a legitimate rationale. It would
need to be included to estimate the "difference in C02e the atmosphere sees." In addition, excluding
CH4 emissions from landfills is inconsistent with the Framework's desire to account for displaced on-
site changes in CO2. For the same reasons, the basis for excluding N2O emissions from biomass
production is unclear. It also needs to be included to estimate the net changes in atmospheric greenhouse
gases. Accounting for N2O from fertilization would be consistent with tracking changes in soil carbon
which are a response to agricultural management systems, which includes fertilizer decisions.
Question 4(b). Is the Framework scientifically rigorous?
The SAB did not find the Framework to be scientifically rigorous. Specifically, the SAB identified a
number of deficiencies that need to be addressed.
Time scale: As discussed previously, one deficiency in the Framework is the lack of proper
consideration of the different time scales inherent in the carbon cycle and the climate system that are
critical for establishing an accounting system. This is a complicated subject because there are many
different time scales that are important for the issues associated with biogenic carbon emissions.
Scientific understanding of the time scale over which the climate system responds to cumulative
emissions implies that the carbon release caused by harvesting and combusting biomass at stationary
sources is a serious problem if carbon storage, on average, is reduced over long periods of time. So long
as rates of growth across the landscape are sufficient to compensate for carbon losses from harvesting
29

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
over the long run, the climate system is less sensitive to the imbalance in the carbon cycle that might
occur in the short run from harvesting of biomass for bioenergy facilities. A scientifically rigorous
evaluation of the impact of biomass harvest on the carbon cycle should consider the temporal
characteristics of the cycling as well as the spatial simultaneous decisions made across stands and plots.
Annual accounting of carbon stocks, while helpful in tracking net carbon emissions, is likely to give an
inaccurate assessment of the overall climate and atmospheric carbon cycle impacts.
The Framework also does not consider the length of time it takes ecosystems to respond to disturbances,
such as those due to the harvesting of biomass, nor does it consider the spatial heterogeneity in this
response. This has implications for the accuracy with which the impact of different land management
strategies on carbon stocks in soil and vegetation is estimated.
The Framework subtracts the emissions associated with products — including ethanol, paper, and timber
— from the calculation of emissions from a stationary source, through the PRODC term. While the EPA
may not have the discretion to treat all emissions equally, distinguishing between immediate emissions
from the facility and downstream emissions (as these products will inevitably be consumed within a
short period of time) does not make sense scientifically. From the perspective of the carbon cycle and
the climate system, all these facilities extract biomass from the land and the vast majority of that
biomass is converted to carbon dioxide, adding to cumulative emissions and, hence, a climate response.
Spatial scale: There is no peer reviewed literature cited to support the delineation of spatial scales for
biogenic CO2 accounting and different carbon pools to be accounted for at different spatial scales. For
example, the atmospheric impact of feedstocks is gauged on a regional basis in terms of its impact on
forest carbon stocks (except for case study 5) while impacts due to land use change are accounted for at
the site level.
The Framework's use of a regional scale for accounting for the net changes to the atmosphere is an
artificial construct developed to (a) avoid the need for site-specific chain of custody carbon accounting
with separate streams for each feedstock and (b) as an alternative to capturing changes in carbon stocks
over time. The calculation of LAR uses regional landscape wide carbon changes but does not actually
estimate changes attributable to biomass demand (see next discussion). This approach attempts to
simplify implementation using available forest inventory data and circumvents the need for accounting
for changes in carbon stocks specific to the site or feedstock sourcing region (fuelshed) which may be
more complex, costly and difficult to verify. However, as noted, it doesn't provide an actual estimate of
carbon changes due to stationary source biomass demand, and it makes the estimate of the BAFs
sensitive to the choice of the spatial region chosen for accounting purposes. As shown by case study 1,
there are significant implications of this choice for the emissions attributed to a facility.
Additionality: A key question is whether the harvesting of biomass for bioenergy facilities is having a
negative impact on the carbon cycle relative to emissions that would have occurred in the absence of
biomass usage. This requires determining what would have happened anyway without the harvesting
and comparing the impact with the increased harvesting of biomass for bioenergy in order to isolate the
incremental or additional impact of the bioenergy facility. However, while the Framework discusses the
"business as usual" or "anticipated future baseline" approach, it implements a reference point approach
30

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
that assesses carbon stocks on a regional basis at a given point in time relative to a historic reference
carbon stock.
For forest carbon stocks, the choice of a fixed reference point may be the simplest to execute, but it does
not actually address the question of the extent to which forest stocks would have been growing/declining
over time in the absence of a particular bioenergy facility. The use of a fixed reference point baseline
implies that forest biomass emissions could be considered carbon neutral if forest stocks are increasing.
This is simply an artifact based on the choice of the baseline that will be used. The problem is thus: a
region with decreasing carbon stocks may in actuality have more carbon than what would have
happened without the increased harvesting of biomass. Similarly, a region with increasing carbon stocks
may have less stores carbon than would be the case without the facility using biomass. By default, this
approach creates "sourcing" and "non-sourcing" regions. Thus, a carbon accumulating region is a
"source" of in situ carbon that can be given to support biomass use, and a carbon losing region is a "non-
source" of carbon and cannot support biomass use. The reference year approach provides no assurances
at all that a "source" region is gaining carbon due to biomass use, or that a "non-source" region is losing
carbon due to biomass use.
For example, for roundwood use under the Framework, a region may have carbon accumulation with
respect to the reference year (and be assigned LAR=1 according to the Framework); however, harvest of
a 150+ year old forest in the region for energy production would not be counted in a facility's
greenhouse gas emissions even though there is less carbon storage than there would have been otherwise
and only a portion of the forest's carbon would be recovered within the next 100 years. To estimate the
"difference in atmospheric greenhouse gases" over some period, one must estimate how carbon
accumulation differs between a biomass use case and a case without biomass use (business as usual
case).
Assessing uncertainty: The Framework acknowledges uncertainty but does not discuss how it will be
characterized and incorporated to assess the potential uncertainty in the estimate of the BAF value.
Selecting an acceptable risk level is a policy decision but characterizing uncertainty and risks is a
scientific question. There are numerous drivers that can change biogenic carbon stocks, even in the
absence of biomass harvesting for energy. These include changes in economic conditions, domestic and
international policy and trade decisions, commodity prices, and climate change impacts. There is
considerable uncertainty about the patterns of future land use, for example, whether land cleared for
bioenergy production will stay in production for decades to come. The potential impact of these forces
on biogenic carbon stocks and the uncertainty of accounting need to be considered further. Ideally, the
EPA should put its BAF estimates into context by characterizing the uncertainties associated with BAF
calculations and estimating uncertainty ranges. This information can be used to give an indication of the
likelihood that the BAFs will achieve the stated objective. The uncertainty within and among variables
for any estimate may vary widely between feedstocks and across regions. Finally, it should be pointed
out that while parameter uncertainty is important to consider throughout the Framework, alternative
policy options (e.g., categorical inclusion and exclusion) do not have parameter uncertainty yet their
effect on atmospheric carbon is also uncertain.
Leakage: The Framework states that the likelihood of leakage and the inclusion of a leakage term will
be based on a qualitative decision. There is essentially no guidance in the document about how leakage
31

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
might be quantified and no examination of the literature regarding possible leakage scenarios (consider
Murray et al. 2004). A number of statements/assumptions were made regarding the area and intensity of
wood harvest increases to accommodate biomass access. There was no examination of the scientific
literature on wood markets and therefore no science-based justification for these
statements/as sumpti ons.
Other areas: Other areas that require more scientific justification include assumptions regarding
biomass losses during transport and their carbon implications, the choice of a 5-year time horizon
instead of one that considered carbon cycling, and the decision to include only CO2 emissions and
exclude other greenhouse gas emissions. Additionally, assumptions about the impacts of harvests on soil
carbon and land use changes on carbon sequestration need to be more rigorously supported.
Inconsistencies: Below are some inconsistencies within the Framework that should be resolved or
justified:
(1) Consistency with fossil fuel emissions accounting: Fossil fuel feedstock emissions accounting
from stationary sources under the Clean Air Act are not adjusted for offsite GREENHOUSE
GAS emissions and carbon stock changes. Does that imply that by default BAFs should be zero
as well? No, because, unlike fossil fuels, biogenic feedstocks have carbon sequestration that
occurs within a timeframe relevant for offsetting CO2 emissions from the biomass' combustion.
For comparability, however, biomass and fossil fuels emissions accounting should be similar for
other emissions categories. These include non-C02 GREENHOUSE GAS emissions, losses,
leakage, and fossil fuel use during feedstock extraction, production and transport. This issue is
also discussed in Section 3.3.1.
(2) Biogenic and fossil fuel emissions accounting for losses: The Framework's handling of carbon
losses during handling, transport, and storage introduces an inconsistency between how fossil
emissions are counted at a stationary source and how biomass emissions are counted. For
biomass emissions the Framework includes emissions associated with loss of feedstock between
the land and the stationary source. For natural gas the emissions attributed to the stationary
source do not include fugitive greenhouse gas emissions from gas pipelines. Why would loss
emissions be included for biomass when they are not included for natural gas?
(3) Inconsistency in the consideration of land management and the associated greenhouse gas flux
accounting: The Framework accounts for soil carbon stock changes, which are a function of the
land management system, soil, and climatic conditions. However, it does not account for the
non-C02 greenhouse gas changes like N2O that are jointly produced with the soil carbon
changes. Soil carbon changes influence both the below and above ground carbon stock changes
associated with changes in the land management system.
(4) Reference year and B AU baseline use: The Framework proposes using a reference year
approach: however, it implicitly assumes projected behavior in the proposed approach for
accounting for soil carbon changes and municipal waste decomposition.
32

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
(5) Definition of soil. There is a good deal of variation in the Framework as to the definition of
"soil." At one point it appears to be defined as all non-feedstock carbon such as slash, surface
litter, and dead roots as well as carbon associated with mineral soil. In other places, the
Framework seems only to consider the carbon associated with mineral soil. Unfortunately this
inconsistency in the use of the term "soil" creates confusion regarding interpretation and
implementation. When soil is defined as non-feedstock carbon (that is all forms of dead carbon)
and then implemented as mineral soil carbon (one form of dead carbon), it is impossible to
ensure a mass balance as dead material above- and below ground is accounted for in one place,
but then not elsewhere. Inconsistent definitions of soil carbon mean that statements regarding the
impact of management cannot be unequivocally assessed. For example, if the broader definition
of soil is being invoked, then the statement that management of forests can reduce soil carbon
could be justified (Harmon et al. 1990; Johnson and Curtis 2001). However, if the narrower
definition of mineral soil carbon is being invoked, then there is very little empirical evidence to
justify this statement (Johnson and Curtis 2001); and in fact there is evidence that forest
management can at least temporarily increase mineral soil carbon.
Soil carbon should be defined and used consistently throughout the document. If defined
broadly, then consistent use of subcategories would eliminate much confusion. For example, if
organic horizons such as litter are part of the soil, then consistently referring to total soil, organic
soil horizons, and mineral horizons would be essential. Had that been done, the confusion about
the impact of forest management on soil carbon would have been eliminated as management can
greatly influence organic horizons, but have little effect on mineral horizons. If defined narrowly
to only include mineral soil, then the EPA should develop a terminology for the other carbon
pools (e.g., organic horizons, aboveground dead wood, and belowground dead wood) that
ensures that mass balance is possible.
To define soil carbon, EPA should consider the merits of an aggregated soil term versus
subcategories based on source of the carbon, the controlling processes, and their time dynamics.
While the aggregated term "soil" is simple, it potentially combines materials with very different
sources, controlling processes, and time dynamics, creating an entity that will have extremely
complex behavior. It also creates the temptation of a broad term being used for a subcategory.
Separating into woody versus leafy materials would account for different sources and to some
degree time dynamics. In contrast, separating into feedstock versus non-feedstock material (as
appears to be done in the Framework) creates a poorly defined boundary as woody branches
would be soil if they are not used, but could be viewed as not being soil if they are. A feedstock-
based system also does not separate materials into more uniform time dynamics (if leaves and
wood are not harvested, then materials with lifespans that differ an order of magnitude are
combined). Controlling processes, be they management or natural in nature, differ substantially
for above- versus belowground carbon; hence they should be divided.
Underlying the need for a clear definition of soil in the document is the complexity of soil
outcomes that differ based on conditions. Some noteworthy omissions from forest soil science
might have informed the Framework's treatment of soil carbon in forest ecosystems (Alban and
Perala 1992; Mattson and Swank 1989; Binkley and Resh 1999; Black and Harden 1995;
33

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Edwards and Ross-Todd 1983; Gilmore and Boggess 1963; Goodale et al. 2002; Grigal and
Berguson 1998; Homann et al. 2001; Huntington 1995; Johnson and Curtis 2001; Laiho et al.
2003; Mroz et al. 1985; Nave et al. 2010; Richter et al. 1999; Sanchez et al. 2007; Schiffman and
Johnson 1989; Selig et al. 2008; Tang et al. 2005; Tolbert et al. 2000).
Question 4(c). Does the Framework utilize existing data sources?
First, and most importantly, the Framework does not provide implementation specifics. Therefore, it is
difficult to assess data availability and use. These issues are discussed here and in the sections that
follow.
A more meaningful question is "Are the proposed data sets adequate to account for the effects of
biogenic carbon cycling on CO2 emissions from a facility?" The Framework does use existing data, but
the data are not adequate to attribute emissions to a facility. For example, the Framework mentions the
use of the USDA Forest Service's Forest Inventory and Analysis (FIA) data at some unspecified scale.
However, carbon stock change data are likely not very accurate at the scale of the agricultural or forest
feedstock source area for a facility.
The Framework requires data and/or modeling of land management activities and their effects on CO2
emissions and stock changes. For example for agricultural systems, data are required on the type of
tillage and the effect of such tillage on soil carbon stocks for different soil types and climatic conditions.
Such data are not likely to be available at the required scales. In one of the case studies, for example, the
Century model is used to model soil carbon stocks. Is the use of this particular model proposed as a
general approach to implement the Framework? Since this model generally addresses soil carbon only to
a depth of 20 centimeters, does that represent a boundary for the Framework? Recent work has shown
that such incomplete sampling can grossly misestimate changes in soil carbon for agricultural practices
such as conservation tillage (Baker et al. 2007; Kravchenko and Robertson 2011). Which version of the
model? Would EPA run this model and select parameters appropriate for each feedstock production area
for each facility? How robust are the predictions of this model for the range of soils, climatic conditions,
and management practices expected to be covered by the Framework? Could some other model be used
that produces different results for a given facility?
The Framework implies that data are required from individual feedstock producers. Collecting such data
would be costly and burdensome. Additionally, to the extent that feedstocks are part of commodity
production and distribution systems that mix material from many sources, it is not likely to be feasible to
determine the source of all feedstock materials for a facility.
The Framework includes a term for leakage but eschews the need to provide any methodology for its
quantification. Example calculations are carried out for leakage in one of the case studies without any
explanation for their source. However, leakage can be positive or negative, and while many publications
speculate about certain types of leakage, no data are presented, nor are data sources for different types of
leakage suggested or discussed. The Framework does provide an example calculation of leakage in the
footnote to a case study, but this does not a substitute for a legitimate discussion of the literature and
justification and discussion of implications of choices. In addition, such data are unlikely to be available
at the scales required. The implications and uncertainties caused by using some indicator or proxy to
34

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
estimate leakage need to be discussed. If leakage cannot be estimated well, is it possible to put an error
range on the leakage value (e.g., a uniform distribution) and assess the impact of this uncertainty on the
overall uncertainty in the BAF value? For some cases, such as the conversion of agricultural land to
biomass production from perennial crops, leakage may be described as likely increasing net emissions.
In cases such as this where prior research has indicated directionality, if not magnitude, such
information should be used. As previously noted, there is also a consistency issue with the reference
year approach because leakage estimation will require an anticipated baseline approach of some sort.
In summary, it is not clear that all of the data requirements of the Framework can be met. Furthermore,
even if the data are acquired, they may not be adequate to attribute emissions to a facility.
Question 4(d). Is it easily updated as new data become available?
In principal it would be feasible to update the calculations as new data become available. Some kinds of
data, such as those from FIA, are updated periodically, could be used to update the analysis. However,
as discussed for other sub-questions, it is not clear exactly what data and resolution are required and
whether all the required data are readily available.
The Framework uses an annual or five-year interval for updating calculations. For some kinds of data,
such as soil and forest carbon stocks, this interval is too short to detect significant changes based on
current or feasible data collection methodologies. This implies that statistical or process models would
be used to estimate short-term changes for reporting purposes.
Lastly, if BAF is not under the control of the facility, frequent calculation of the BAF would introduce
considerable uncertainty for the facility. This would particularly be the case if a leakage factor were
included in the BAF and would need to be updated frequently with changes in market conditions.
However, if the accounting is infrequent, shifts in the net greenhouse gas impact may not be captured.
Clearly, the EPA will have to weigh tradeoffs between the accuracy of greenhouse gas accounting and
ease of implementation and other transactions costs.
Question 4(e). Is it simple to implement and understand?
It is neither. While the approach of making deductions from the actual emissions to account for
biologically based uptake/accumulation is conceptually sound, it is not intuitive to understand because it
involves tracking emissions from the stationary source backwards to the land that provides the feedstock
rather than tracking the disposition of carbon from the feedstock and land forwards to combustion and
products. The Framework also appears to be difficult to implement, and possibly unworkable, especially
due to the many kinds of data required to make calculations for individual facilities. Additionally, the
factors (variable names) in the Framework do not match those used in the scientific literature and may
be misunderstood. Lastly, many elements of the Framework are implicit rather than explicit. For
example, the time frame during which changes in atmospheric greenhouse gases will be assessed is not
explicit. The time frame for specific processes is often implicit, such as the emissions of CO2 from
biomass that is lost in transit from the production area to the facility; this loss is assumed to be
instantaneous.
35

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Much more detailed information is required about how the Framework would be implemented. It would
be helpful to know the specific data sources and/or models to be used. To assess the adequacy of data,
more information is needed on implementation and the degree of uncertainty acceptable for
policymakers to assign BAF values.
Question 4(f). Can the SAB recommend improvements to the framework to address the issue of
attribution of changes in land-based carbon stocks?
The Framework uses a reference year baseline approach to determining BAF in combination with a
regional spatial scale. As mentioned in response to charge question 4(b), this approach is not adequate
in cases where feedstocks accumulate over long time periods because it does not allow for the estimation
of the incremental effect on greenhouse gas emissions over time of feedstock use. To gauge the
incremental effect on forest carbon stocks due to the use of forest-derived woody biomass (specifically,
the value of the LAR), an anticipated baseline approach is needed. This involves estimating a "business
as usual" trajectory of emissions and forest stocks and comparing it with alternate trajectories that
incorporate increased demand for forest biomass over time. The anticipated baseline approach should
also be applied to determine soil carbon for all types of feedstocks for soils, residue, waste disposition
and land management.
An anticipated baseline approach must incorporate market effects even when direct effects of the use of
biogenic feedstocks on carbon emissions are being estimated. The projected baseline level of forest
carbon stocks will need to be compared with the level in the case when there is demand for roundwood
for bioenergy to assess the change in forest stocks due to the demand for bioenergy. The case with
demand for bioenergy should consider the possibility that investment in long-lived trees could be driven
by expectations about wood product prices and biomass prices, leading landowners to expand or retain
land in forests, plant trees, shift species composition, change management intensity and adjust the timing
of harvests. The role of demand and price expectations/anticipation is well developed in the economics
literature (e.g., see Muth 1992) and also in the forest modeling literature (Sedjo and Lyon 1990; Adams
et al. 1996; Sohngen and Sedjo 1998), which includes anticipatory behavior in response to future forest
carbon prices and markets (Sohngen and Sedjo 2006; Rose and Sohngen 2011). The U.S. Energy
Information Administration (EIA) has projected rising energy demands for biogenic feedstock based on
market and policy assumptions, which could be met from a variety of sources, including energy crops
and residues, but also short rotation woody biomass and roundwood (EIA 2012; Sedjo 2010; Sedjo and
Sohngen 2012). The extent to which price expectations and anticipation of future demand for bioenergy
are going to drive forest management decisions, and regional variations in them, would need to be
empirically validated. One study shows forest carbon change in a decade (and thereafter) that exceeds
the modeled increased cumulative wood energy emissions over the decade (Sedjo and Tian, in press,
2012). This would be the case if demand is anticipated to increase in the future. Some other modeling
studies suggest more limited responses to increased wood energy demand that differ across regions. One
such model for the United States indicates a large response in the South, in the form of less forest
conversion to non-forest use, but much less response in the North and West (USDA FS 2012; Wear
2011).
To capture both the market, landscape and biological responses to increased biomass demand, a
bioeconomic modeling approach is needed with sufficient biological detail to capture inventory
36

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
dynamics of regional species and management differences as well as market resolution that captures
economic response at both the intensive (e.g., changing harvest patterns, utilization or management
intensity) and extensive margins (e.g., land use changes). While several models have these features
[USDA Forest Service Resources Planning Act (RPA) models in Wear 2011; Sub-regional Timber
Supply in Abt et al. in press 2012; Forest and Agricultural Sector Optimization Model (FASOM) in
Adams et al. 2005; and the Global Timber Market Model (GTMM) in Sohngen and Sedjo 1998], they
differ in scope, ecological and market resolution, and how future expectations are formed. FASOM and
GTMM employ dynamic long term equilibria that adopt the rational expectations philosophy that
decisions incorporate expectations about future prices and market opportunities. In the RPA and SRTS
models, agents respond to current supply, demand, price signals so that expectations are assumed to be
driven by current market conditions. While the rational expectations approach has internal logical
consistency and can better simulate long-term structural change, it is not designed for prediction but
instead to evaluate potential futures and deviations between futures. These models should incorporate
the multiple feedstocks (including crop and logging residues) from the agricultural and forest sectors
that would compete to meet the increased demand for bioenergy.
Energy policies can influence the mix of feedstocks used, such as the use of logging residues and the
level of projected traditional wood demand, and thus the impact of woody bioenergy demand on timber
markets (Daigneault et al. in press 2012). A lower level of timber demand from pulp and paper mills and
sawmills, for example, will lead to lower harvest levels and fewer available logging residues. If only
residues are allowed to qualify as renewable, then the woody bioenergy industry is explicitly tied to the
future of the traditional wood industries. However, if roundwood is used for bioenergy, then the market
outcome is more complicated. A lower level of traditional harvest could lead to fewer available residues
(which could raise the price of residues and set a physical upper limit on residue supply), but could also
lead to higher inventory levels and lower roundwood prices, which would favor increased roundwood
utilization for bioenergy. Modeling the interaction across traditional wood consumers, bioenergy
consumers, changes in the utilization and mix of products and the displacement of one wood consumer
by another as markets evolve will be difficult, but could have a significant impact on the estimate of the
carbon consequences of bioenergy use.
As with any modeling, uncertainties will need to be assessed. Models that include price expectations
effects or the impact of current year prices would need to be validated. However, validation means
different things for different kinds of models. For an econometric model, reproducing history is a form
of validation, as is evaluating errors in near-term forecasts. Simulation models are not forecast models.
They are designed to entertain scenarios. Validation for simulation models is evaluating parameters and
judging the reasonableness of model responses—both theoretically and numerically—given
assumptions. Evaluation will help improve representation of average forest and agricultural land
management behavior. Evidence affirming or indicating limitations of the effect of prices on investment
in retaining or expanding forest area across various U.S. regions may be found by a review of empirical
studies of land use change.
Selection of an appropriate model requires judgment and understanding of the structure and assumptions
of alternative models and their strengths and weaknesses. This could be supplemented with one or more
approaches to choosing a model. These include validation of existing models at the relevant temporal
and spatial scale by a means appropriate to the model type, as well as using more than one model to
37

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
compare and triangulate outcomes. Note that models of different types (e.g., projections vs. forecasting
models) require different types of evaluation.
The anticipated baseline approach could be based on a national/global scale model or a regional scale
after weighing the strengths and weaknesses of the two approaches. An example of a regional scale
model is that by Galik and Abt (2012) where they tested the effects of various scales on greenhouse gas
outcomes and found that in the southern United States, market impacts (negative leakage) had a
significant impact on forest carbon impacts, but the results were dependent on time period evaluated and
were particularly sensitive to scale. The authors evaluated carbon consequences of bioenergy impacts
from stand level to state level and found that as scale increased, market responses mitigated forest
carbon impacts. In addition to being sensitive to scale, another disadvantage of the regional scale models
is that they would not account for leakage across different regions. However, regional models can
incorporate greater heterogeneity in forest growth rates, their carbon impacts and in the price
responsiveness of forest management decisions. The SAB has not conducted a detailed review of these
models to suggest which model and which scale would be the most appropriate.
While market effects are important, there is value in making separate estimates of biological land carbon
changes alone (without market effects). Specifically, biophysical process response modeling results are a
critical input to economic modeling. Ecosystem modeling is not a substitute for economic modeling,
which is necessary to estimate behavioral changes driven by biomass feedstock demand that produces
the estimates emissions and sequestration changes. Ecosystem modeling would establish carbon storage
in the absence of positive or negative leakage and may have lower uncertainty - especially for logging
residue - than the estimate with leakage. Appendix D depicts three biological scenarios for the total
carbon storage in a forest system, including live, dead, and soil stores of carbon. Graphically, Figure D-2
in Appendix D shows how the storage of carbon in a forest system could respond to a shorter harvest
interval. Note that all graphs in Appendix D show the biological response and do not account for
management changes that could be induced through markets or policies.
Modeling physical land carbon responses over time (without market effects) would show how carbon
storage varies by such factors as length of harvest rotations, initial stand age and density, thinning
fraction, and growth rates. These carbon responses to management decisions are important inputs for
economic modeling of management changes and their carbon consequences. Such modeling could also
include the effect of avoided fire emissions on forest land due to biomass removal. This information
could indicate what forest conditions and practices could provide higher rates of accumulation,
information that might be helpful for EPA in designing its policy response so that incentives could be
provided to favor harvest in areas with a higher likelihood of carbon accumulation.
Question 4(g). Are there additional limitations of the accounting framework itself that should be
considered?
A number of important limitations of the Framework are discussed below:
Framework ambiguity: Key Framework features were left unresolved, such as the selection of regional
boundaries (the methods for determining as well as implications), marginal versus average accounting,
inclusion of working or non-working lands in the region when measuring changes in forest carbon
38

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
stocks, inclusion/exclusion of leakage, and specific data sources for implementation. As a result, the
Framework's implementation remains ambiguous. The ambiguity and uncertainty in the text regarding
what are stable elements versus actual proposals also clouded the evaluation. If the EPA is entertaining
alternatives and would like the SAB to comment on alternatives, then the alternatives should be clearly
articulated and the proposed Framework and case studies should be presented with alternative
formulations to illustrate the implementation and implications of alternatives.
Feedstock groups: The proposal designates three feedstock groupings. However, it is not clear what
these mean for B AF calculations, if anything. The Framework does not incorporate the groupings into
the details of the methodology or the case studies. As a result, it is currently impossible to evaluate their
implications.
Potential for Unintended consequences: The proposed Framework is likely to create perverse incentives
for investors and land-owners and result in unintended consequences. For investors, the regional
baseline reference year approach will create regions that are one of two types — either able to support
bioenergy from forest roundwood (up to the gain in carbon stock relative to the reference year), or not.
As a result, a stationary source investor will only entertain keeping, improving, and building facilities
using biomass from regions designated as able to support bioenergy. However, as noted previously,
regions losing carbon relative to the reference year could actually gain carbon stock in relative terms due
to improved biomass use and management to meet market demands. In addition, the definitions of
regions would need to change over time. The designation of regions (and their corresponding LARs) that
comes from the reference year approach will create economic rents and therefore financial stakes in the
determination of regions and management of forests in those regions.
The proposed Framework could also create perverse incentives for landowners. For instance,
landowners may be inclined to clear forest land a year or more in advance of growing and using energy
crops. Similarly, landowners may be more inclined to use nitrogen fertilizers on feedstocks or other
lands in conjunction with biomass production. Such fertilization practices have non-CC>2 greenhouse gas
consequences (specifically N2O emissions) that are not presently captured by the Framework. It should
be noted that agricultural intensification of production via fertilization is a possible response to increased
demand for biomass for energy. If onsite N2O emissions are not accounted for, the carbon footprint of
agricultural feedstocks could be significantly underestimated.
Assessment of Monitoring and Estimation Approaches: The Framework lacks a scientific assessment of
different monitoring/estimation approaches and their uncertainty. This is a critical omission as it is
essential to have a good understanding of the technical basis and uncertainty underlying the use of
existing data, models and look-up tables. A review of monitoring and verification for carbon emissions
from different countries, both from fossil and biogenic sources, was recently released by the National
Research Council that may provide some guidance (National Research Council 2010).
39

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
3.5. Case Studies
Charge Question 5: EPA presents a series of case studies in the Appendix of the report to
demonstrate how the accounting framework addresses a diverse set of circumstances in which
stationary sources emit biogenic CO2 emissions. Three charge questions are proposed by EPA.
Overall Comments
In general, case studies are extremely valuable for informing the reader with examples of how the
Framework would apply for specific cases. While they illustrate the manner in which a BAF is
calculated, the data inputs are illustrative only and may or may not be the appropriate values for an
actual biomass-to-energy project. Moreover, the case studies are simplistic relative to the manner in
which biomass is converted to energy in the real world. For all case studies in the Framework, additional
definition of the context is needed, along with examples of how the data are collected or measured, and a
discussion of the impacts of data uncertainty. Overall, the case studies did not fully cover the relevant
variation in feedstocks, facilities, regions, etc. of potential BAFs that is required to evaluate the
methodology. For clarity, it might be useful to start with a specific forestry or agricultural feedstock
example as the base case, then add the impacts of the more detailed cases, e.g., additional losses,
products, land use changes. This may be more useful than a series of completely separate examples,
each illustrating different pieces of the Framework.
Question 5(a). Does the SAB consider these case studies to be appropriate and realistic?
The case studies did not incorporate "real-world" scenarios which would have served as models for
other situations that may involve biogenic carbon emissions. More would have been learned about the
proposed Framework by testing it in multiple, unique case studies with more realistic data development
and inclusion. Additional case studies for landfills and waste combustion, switchgrass, waste, and other
regions would be useful, as well as illustrations of the implementation of feedstock groups, and
Framework alternatives.
For example, Case Study 4 considers a scenario where corn stover is used for generating electricity.
While it is possible that this scenario could be implemented, this particular case study is not realistic
because very few electrical generation facilities would combust corn stover or agricultural crop residues
only. A more likely scenario might be supplementing a co-firing facility with a low percentage of corn
stover. Additionally, the assumption of uniform corn stover yields across the region is not realistic.
Variation should be expected in the yield of corn stover across the region.
In another example, Case Study 5 calculates the net biogenic emissions from converting agricultural
land in row crops to poplar for electricity production. This case study is also not representative of "real
world" agricultural conditions as switching from one energy crop to another is uncommon. The formula
provided for estimating the standing stock of carbon in the aboveground biomass in the poplar system is
not intuitive. The methods for determining biomass yield and measuring changes in soil carbon (which
will depend on current use of the land) are not described.
40

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Question 5(b). Does the EPA provide sufficient information to support how EPA has applied the
accounting framework in each case?
There remained considerable uncertainty in many of the inputs. In addition, some sensitivity/uncertainty
analysis would be useful. The results of this analysis may guide the EPA in further model development.
For example, if the BAF is determined to be zero, or not statistically different from zero in most case
studies, then this could pave the way for a simpler framework. As discussed in Section 4 below, a
simpler approach could be designed to develop default BAFs for categories of feedstocks based on how
their management and use interacts with the carbon cycle.
Question 5(c). Are there alternative approaches or case studies that EPA should consider to
illustrate more effectively how the framework is applied to stationary sources?
Additional case studies should be designed based on actual or proposed biomass to energy projects to
capture realistic situations of biomass development, production, and utilization. For example, Case
Study 1 describes the construction of one new plant. What would happen if 10 new plants were to be
proposed for a region? And how would the introduction of multiple facilities at the same time impact the
accounting for each facility?
All terms/values used to determine the BAF need to be referenced to actual conditions throughout the
growth/production/generation processes that would occur in each case study including how these values
would actually be implemented by one or more involved parties. Regional look-up tables could be
valuable and EPA could learn a great deal by trying to develop look -up tables.
Additional case studies could be developed for perennial herbaceous energy crops, annual
energy/biomass sorghums, rotations with food and energy crops, cropping systems on different land and
soil types, municipal solid waste and internal reuse of process materials. Each of these feedstocks should
be assessed across alternative regions so that the variation in carbon changes across regions could be
gauged.
For example it would be very useful to consider the application of the Framework to a cellulosic ethanol
plant fueled with coal or gas, and consider the emissions of CO2 from fermentation (not combustion)
and the production of ethanol which is rapidly combusted to CO2 in a non-stationary engine. While such
an operation is associated with three major sources of CO2 emissions (listed here), only one is included
in the Framework; only two may be considered under EPA's regulatory authority, yet all three are
emissions to the atmosphere. It would be useful for EPA to at least describe the emissions that are
excluded from consideration so that biogenic carbon emissions from stationary sources can be viewed in
context.
At least two case studies are needed on municipal solid waste. One case study should be on waste
combustion with electrical energy recovery. EPA should also perform a case study on landfill disposal
of municipal solid waste. Here it is important to recognize that landfills are repositories of biogenic
organic carbon in the form of lignocellulosic substrates (e.g., paper made from mechanical pulp, yard
waste, food waste). There is literature to document carbon storage and the EPA has recognized carbon
storage in previous greenhouse gas assessments of municipal solid waste management.
41

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
In Case Study 3 the data used in Table 3 to describe the 'paper co-product' will vary with the grade of
paper. The 'carbon content of product' may vary between 30 to 50% depending on the grade and the
amount of fillers and additives. Also, some significant carbon streams in a mill can go to landfills and
waste water treatment. The submitted comments from NCASI include a useful example of the
detail/clarity that could be used to enhance the value of the Case Studies.
After completion of the case studies, there should be a formal evaluation of (1) the ease with which data
were developed and the model implemented, (2) whether the results are robust and useful in recognition
of the uncertainty in the various input parameters, and (3) whether the model results lead to unintended
consequences.
Case studies could be developed to assess and develop a list of feedstocks or applications that could be
excluded from accounting requirements as "anyway" emissions. A sensitivity analysis using case studies
could be used to develop reasonable offset adjustment factors if they are needed to adjust anyway
feedstocks for impact on long term stocks like soil if needed.
3.6. Overall Evaluation
Charge Question 6: Overall, this report is the outcome of EPA '.s analysis of the science and
technical issues associated with accounting for biogenic CO2 emissions from stationary sources.
Question 6(a). Does the report-in total-contribute usefully to advancement of understanding of
accounting for biogenic CO2 emissions from stationary sources?
Yes, the Framework contributes to advancing the understanding of accounting for biogenic emissions
and addresses many issues that arise in such an accounting system. It is thoughtful and far reaching in
the questions it tackles. Its main contribution is to force important questions and offer some ways to deal
with these. It covers many of the complicated issues associated with the accounting of biogenic CO2
emissions from stationary sources and acknowledges that its choices will have implications for the
estimates of CO2 emissions obtained. These include those raised by SAB and discussed above, related to
the choice of baseline, region selection and the averaging of emissions/stocks over space and time.
However, the solutions offered in many cases, particularly those related to the use of harvested wood for
bioenergy, lack transparency or a scientific justification.
Question 6(b). Does it provide a mechanism for stationary sources to adjust their total onsite
emissions on the basis of the carbon cycle?
Clearly the Framework offers a mechanism to adjust total on-site emissions. For short accumulation
feedstocks (i.e., agricultural residues, perennial herbaceous crops, mill wood wastes, other wastes), the
Framework could, with some modifications and careful consideration of data and implementation,
accurately represent the direct carbon changes offsite. Leakage, however, both positive and negative,
remains a troublesome matter if left unresolved. Moreover, the Framework offers no scientifically sound
way to define a region. The definition of the regional scale can make a large difference to the estimate of
emissions from a facility using wood as a biomass. Moreover, if there is no connection between actions
42

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
of the point source and what happens in the region, there is no foundation for using regional changes in
carbon stocks to assign a BAF to the source.
The Framework also does not make a clear scientific case for use of waste or what is called "anyway"
emissions. Scientifically speaking, all biogenic emissions are "anyway" emissions. Even most woody
biomass harvested from old growth forests, would, if left undisturbed, eventually die and decompose,
returning carbon to the atmosphere. The appropriate distinction is not whether the product is waste or
will eventually end up in the atmosphere anyway, but whether the stationary source is leading to an
increase or a decrease in biogenic carbon stocks and associated change in GWP. To do this, the
Framework must consider an anticipated baseline and the time period for "anyway" emissions and that
this may vary across different types of waste feedstocks.
An important limitation of the proposed Framework is that the accounting system replaces space for
time and applies responsibility for things that happen on the land to a point source, for which the agent
who owns that point source has no direct control. The proposed approach, which attempts to estimate
facility-feedstock specific BAFs, would estimate an individual point source's BAF based on average
data in a region in which it is located. Any biogenic carbon accounting system that attempts to create
responsibility or give credit at a point source for carbon changes upstream or downstream from the point
source must relate those responsibilities and credits to actions under control of the point source.
However, the Framework does not clearly specify a cause and effect relationship between a facility and
the biogenic CO2 emissions attributed to it. In particular, if the BAF is assigned to a plant when it is
approved for construction, as the BAF is currently designed, those emissions related to land use change
will have nothing to do with the actual effect of the point source on land use emissions because the data
on which it is based would predate the operation of the plant.
The dynamics of carbon accumulation in vegetation and soils and carbon and methane release through
decomposition present a challenge for any accounting system because anticipated future changes in
vegetation should, in principle, be factored into BAF. These future changes depend on natural processes
such as fires and pest outbreaks that are not easily foreseen, and because of climate change and broader
environmental change, we face a system that is hard to predict. Projecting forward based on current or
historical patterns is subject to biases of unknown direction and magnitude. More importantly, land use
decisions are under the control of landowners, who will be responding to unknown future events. The
Framework recognizes this issue and chooses to use a Reference Point Baseline, the serious limitations
of which have been discussed previously.
Overall, the EPA's regulatory boundaries, and hence the Framework, are in conflict with a more
comprehensive carbon accounting that considers the entire carbon cycle and the possibility of gains from
trade between sources, among sources or between sources and sinks to offset fossil fuel combustion
emissions. Scientifically, a comprehensive greenhouse gas accounting would extend downstream - to
emissions from by-products, co-products or products such as ethanol combustion or ethanol by-products
such as distillers dried grains that are sold as livestock feed that ultimately becomes CO2 (or CH4).
However, doing so would need to consider consistency with fossil fuel emissions accounting and
emissions currently regulated (such as by EPA with vehicle greenhouse emissions standards). As for
gains from trade, by restricting its attention to the regulation of point source emissions, EPA's analysis
does not allow for the possibility that a fossil CO2 emitter could contract with land owners to offset their
43

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
emissions through forest protection and regrowth or carbon accumulation in soils. Bioenergy would still
need to confront the issue of crediting offset carbon accumulation however. By staying within
boundaries drawn narrowly around the stationary source, the Framework eclipses a more comprehensive
approach to greenhouse gas reductions that would address all sources and sinks and take advantage of
gains from trade.
Question 6(c). Does the SAB have any advice regarding potential revisions that might enhance the
final document?
Overall, the Framework would be enhanced by including a description of its regulatory context and
specifying the boundaries for regulating upstream and downstream emissions while implementing the
regulation. The motivation for the Framework should be explained as it relates to Clean Air Act
requirements and any recent court rulings. The Framework should also make explicit the constraints
within which greenhouse gases can be regulated under the Clean Air Act. In doing this, the EPA could
be clear that these issues have not been settled but that some assumptions were necessary to make a
decision about the Framework. The EPA could also stipulate that further development of a regulatory
structure might require changes to the accounting system. While the SAB understands the EPA's interest
in describing an accounting system as a first step and potentially independent of the regulatory structure,
the reader needs this background in order to understand the boundaries and context for the accounting
structure and to evaluate the scientific integrity of the approach.
Similarly, the Framework is mostly silent on how possible regulatory measures under the Clean Air Act
may relate to other policies that affect land use changes or the combustion/oxidation of products from
the point sources that will release carbon or other greenhouse gases. For example if a regulatory or
incentive system exists to provide credits for carbon offsets through land use management then under
some conditions it would be appropriate to assign a BAF of 1 to biogenic emissions given that the
carbon consequences were addressed through other policies.
The Framework does not make explicit how it does or does not address emissions downstream from a
point source such as in the case of a biofuels or paper production facility where the product (biofuels,
paper) may lead to CO2 emissions when the biofuels are combusted or the paper disposed of and
possibly incinerated. For example, if paper products are incinerated the incinerator may well be a point
source that comes under Clean Air Act regulation. However, biofuels used in vehicles would not be
subject to regulation as a point source. Though biofuel combustion emissions are already regulated,
along with combustion of gasoline, via EPA's vehicle greenhouse gas emissions standards, the EPA
needs to make clear the implicit assumptions on how biogenic carbon will be treated upstream and
downstream from the point source if this Framework is used to regulate CO2 emissions under the
constraints imposed by the Clean Air Act for regulating stationary sources.
The Framework is lacking in implementation details. Implementation is crucial and some of the EPA's
current proposals will be difficult to implement. Data availability and quality, as well as procedural
details (e.g., application process, calculation frequency) are important considerations for assessing the
feasibility of implementation and scientific accuracy of results. Implementation details (e.g., data,
technical processes, administrative procedures, timing) need to be laid out, discussed and justified.
44

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Among other things, the discussion should note alternatives, uncertainty and implications via case
studies.
Recommendations for Revising BAF
In response to the charge to the SAB, recommendations are offered here for revising the Framework. In
the next section, the SAB suggests an alternative - default BAFs. If EPA decides to revise the
Framework, the following recommendations for specific improvements to the document (and
methodology) are summarized here. Many of the issues raised in previous responses regarding the
treatment of specific factors included in the Framework are specific to particular feedstocks. The clarity
of the Framework would be improved by differentiating among feedstocks based on how their
management and use interacts with the carbon cycle. Feedstocks could be categorized into short rotation
dedicated energy crops, crop residues, forest residues and long rotation trees, as grown in different
regions, on different prior land-use types and with different management practices. A BAF equation
could be developed for each of these categories of feedstocks.
If EPA decides to revise the Framework, the following recommendations for specific improvements are
summarized below.
• Develop a separate BAF equation for each feedstock category as broadly categorized by type,
region, prior land use and current management practices. Feedstocks could be categorized into short
rotation dedicated energy crops, crop residues, forest residues, perennial crops, municipal solid
waste, long rotation trees and waste materials including wood mill residue and pulping liquor.
o For long-accumulation feedstocks like woody biomass, use an anticipated baseline and landscape
approach to compare emissions from increased biomass harvesting against a baseline without
increased biomass demand. For long rotation woody biomass, sophisticated modeling is needed
to capture the complex interaction between electricity generating facilities and forest markets, in
particular, market driven shifts in planting, management and harvests, induced displacement of
existing uses of biomass, land use changes, including interactions between agriculture and forests
and the relative contribution of different feedstock source categories (logging residuals,
pulpwood or roundwood harvest).
o For residues, consider incorporating information about decay after an appropriate analysis in
which storage of ecosystem carbon is calculated based on decay functions.
o For materials diverted from the waste stream, consider their alternate fate, whether they might
decompose over a long period of time, whether they would be deposited in anaerobic landfills,
whether they are diverted from recycling and reuse, etc. Implementation complexity, cost and
scientific accuracy should be considered. For feedstocks that are found to have relatively minor
impacts, the EPA may need to weigh ease of implementation against scientific accuracy. After
calculating decay rates and considering alternate fates, EPA may wish to declare certain
categories of feedstocks with relatively low impacts as having a very low BAF or setting it to 0.
• Incorporate various time scales and consider the tradeoffs in choosing between different time scales.
45

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
1	• For all feedstocks, consider information about carbon leakage to determine its directionality as well
2	as leakage into other media.
3
46

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
4. DEFAULT BAFs BASED ON FEEDSTOCK CATEGORIES
There are no easy answers to accounting for the greenhouse gas implications of bioenergy. Given the
uncertainties, technical difficulties and implementation challenges associated with implementing the
facility-specific BAF approach embodied in the Framework, the SAB encourages the EPA to "think
outside the box" and look at alternatives to the Framework and its implementation as proposed. One
promising alternative is default BAFs for each feedstock category. Given the conceptual and scientific
deficiencies of the Framework, and the prospective difficulties with implementation, the SAB
recommends consideration of default BAFs by feedstock type, region, land management and prior land
use. Under EPA's Framework, facilities would use individual BAFs designed to capture the incremental
carbon cycle and net emissions effects of their use of a biogenic feedstock. With default BAFs, facilities
would use a weighted combination of default BAFs relevant to their feedstock consumption and
location.
The defaults BAFs would rely on readily available data and reflect landscape and aggregate demand
effects, including previous land use. The defaults would also have administrative advantages in that they
would be easier to implement and update. Default BAFs for each category of feedstocks would
differentiate among feedstocks using general information on their role in the carbon cycle. An
anticipated baseline would allow for consideration of prior land use, management, alternate fate (what
would happen to the feedstock if not combusted for energy) and regional differences. Default BAFs
might vary by region, prior land use and current land management practices due to differences these
might cause in the interaction between feedstock production and the carbon cycle. They would be
applied by stationary facilities to determine their quantity of biogenic emissions that would be subject to
the Agency's Tailoring Rule. Case studies should be used to evaluate default BAFs applicability to
heterogeneous facilities. Facilities could also be given the option of demonstrating a lower BAF for the
feedstock they are using. This would be facilitated by making the BAF calculation transparent and based
on data readily available to facilities. Default BAFs should be carefully designed to provide incentives to
facilities to choose feedstocks with the lower greenhouse gas impacts.
The SAB also explored certification systems as a possible way to obviate the need to quantify a specific
net change in greenhouse gases associated with a particular stationary facility. Carbon accounting
registries have been developed to account for and certify CO2 emissions reductions and sequestration
from changes in forest management. Ultimately, however, the SAB concluded that it could not
recommend certification without further evaluation. Moreover, such systems could encounter many of
the same data, scientific and implementation problems that bedevil the Framework.
47

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
REFERENCES
Abt, K.L., Abt, R.C. and Galik, C.G. (2012, In Press). Effect of bioenergy demands and supply response
on markets, carbon and land use. Forest Science.
Adams, D., Alig, R., Callaway, J., Winnett, S. and McCarl, B. (1996). The Forest and Agricultural
Sector Optimization Model (FASOM): Model Structure, Policy and Applications. Portland, OR:
USDA Forest Service, Pacific Northwest Experiment Station.
Adams, D.; Alig, R.; McCarl, B. and Murray, B. (2005, February). Bruce McCarl Website. Retrieved
April 2012, from Texas A & M Agricultural Economics:
http://agecon2.tamu.edu/people/facultv/mccarl-bruce/papers/1212FASOMGHG doc.pdf.
Alban, D.H. and Perala, D.A. (1992). Carbon Storage in Lake States Aspen Ecosystems. Canadian
Journal of Forest Research, 1107-1110.
Allen, M. R., Frame, D. J., Huntingford, C., Jones, C. D., Lowe, J. and Meinshausen, M. (2009).
Warming Caused by Cumulative Carbon Emissions toward the Trillionth Tonne. Nature, 1163 -
1166.
Anderson-Teixeira, K.; Davis, S.; Masters, M.; Delucia, E. (2009). Changes in soil organic carbon under
biofuel crops. Global Change Biology Bioenergy, 75-96.
Baker, J., Ochsner, T., Venterea, R. and Griffis, T. (2007). Tillage and soil carbon sequestration—What
do we really know? Agriculture, Ecosystems and Environment, 118.
Binkley, D. and Resh, S.C. (1999). Rapid changes in soils following Eucalyptus afforestation in Hawaii.
Soil Science Society of America Journal, 222-225.
Black, T. and Harden, J.W. (1995). Effect of timber harvest on soil carbon storage at Blodgett Exp
Forest, California. Canadian Journal of Forest Research, 1385 - 1396.
Chen, X., and Khanna, M. (2012). The Market Mediated Effects of Biofuel Policies. Agbioforum. 15
(1), 11-
Cherubini, F., Peters, G., Berntsen, T., Stromman, A. and Hertwich, E. (2011). C02 Emissions from
Biomass Combustion for Bioenergy: Atmospheric Decay and Contribution to Global Warming.
Global Change Biology Bioenergy, 413 - 426.
Cherubini, F., Guest, G. and Stromman, A. (2012). Application of Probability Distributions to the
Modeling of Biogenic C02 Fluxes in Life cycle Assessment. Global Change Biology Bioenergy,
1 - 15.
48

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Crutzen, P., Mosier, A., Smith, K. and Winiwarter, W. (2007). N20 Release from Agro-biofuel
Production Negates Global Warming Reduction by Replacing Fossil Fuels. Atmos. Chem. Phys.
Discussion, 11191 - 11205.
Daigneault, A.; Sohngen, B. and Sedjo, R. (In press, 2012). An Economic Approach to Assess the Forest
Carbon Implications of Biomass Energy. Environmental Science & Technology, Forthcoming.
Edwards, N.T. and Ross-Todd, B.M. (1983). Soil carbon dynamics in a mixed deciduous forest
following clear cutting with and without residue. Soil Science Society of America Journal, 1014-
1021.
Energy Information Administration (EIA). (2012). Annual Energy Outlook 2012 Early Release
Overview. Washington, D.C.: U.S. Department of Energy.
Galik, C.S. and Abt, R.C. (2012). The effect of assessment scale and metric selection on the greenhouse
gas benefits of woody biomass. Biomass and Bioenergy, In Press.
Gilmore, A.R.; Boggess, W.R. (1963). Effects of past agricultural practices on the survival and growth
of planted trees. Soil Science Society of America Proceedings, 98-101.
Goodale, C.L.; Apps, M.J.; Birdsey, R.A.; Field, C.B.; Heath, L.S.; Houghton, R.A.; Jenkins, J.C.;
Kohlmaier, G.H.; Kurz, W.; Liu, S.R.; Nabuurs, G.J.; Nilsson, S. and Shvidenko, A.Z. (2002).
Forest carbon sinks in the Northern Hemisphere. Ecological Applications, 891 - 899.
Grigal, D.F.and Berguson, W.E. (1998). Soil carbon changes associated with short-rotation systems.
Biological Biogengineering, 311-311.
Harmon, M., Ferrell, W., and Franklin, J. (1990). Effects on carbon storage of conversion of old-growth
to young forests. Science, 699 - 702.
Holcombe, R., and Sobel, R. (2001). Public Policy Toward Pecuniary Externalities. Public Finance
Review, 29.
Homann, P.S.; Bormann, B.T. and Boyle, J.R. (2001). Detecting treatment differences in soil carbon and
nitrogen resulting from forest manipulations. Soil Science Society of America Journal, 463-469.
Huntington, T. (1995). Carbon sequestration in an aggrading forest ecosystem in the southeastern USA.
Soil Science Society of America Journal, 1459-1467.
Ince, P., Kramp, A., Skog, K., Spelter, H. and Wear, D. (2011). U.S. Forest Products Module: A
Technical Document Supporting the Forest Service 2010 RPA Assessment. Madison, WI: U.S.
Forest Service, Forest Products Laboratory.
Johnson, D.W. and Curtis, P.S. (2001). Effects of forest management on soil C and N Storage: meta
analysis. Forest Ecology and Management, 227-238.
49

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Johnson, E. (2009). Goodbye to carbon neutral: Getting biomass footprints right. Environmental Impact
Assessment Review, 29.
Khanna, M.; Crago, C. and Black, M. (2011). Can biofuels be a solution to climate change? The
implications of land use change-related emissions for policy. Interface Focus, 233-247.
Khanna, Madhu; and Crago, C.L. (2012). Measuring Indirect land Use Change with Biofuels:
Implications for Policy. Annual Review of Resource Economics. DOI: 10.1146/annurev-resource-
110811-114523.
Kirschbaum, M.U.F. (2003). Can Trees Buy Time? An Assessment of the Role of Vegetation Sinks as
Part of the Global Carbon Cycle. Climatic Change, 58: 47-71.
Kirschbaum, M.U.F. (2006). Temporary Carbon Sequestration Cannot Prevent Climate Change.
Mitigation and Adaptation Strategies for Global Change. 11: 1151-1164.
Kravchenko, A. and Robertson, G. (2011). Whole Profile Soil Carbon Stocks: The Danger of Assuming
Too Much from Analyses of Too Little. Soil Science Society of America Journal, 75.
Laiho, R.; Sanchez, F.; Tiarks, A.; Dougherty, P.M. and Trettin, C.C. (2003). Impacts of intensive
forestry on early rotation trends in site carbon pools in the southeastern U.S. Forest Ecology and
Management, 177-189.
Liska, A. and Perrin, R. (2009). Indirect land use emissions in the life cycle of biofuels: regulations vs
science. Biofuels, Bioproducts and Biorefining, 318-328.
Mattson, K.G. and Swank, W.T. (1989). Soil and detrital carbon dynamics following forest cutting in the
Southern Appalachians. Biology and Fertility of Soils, 247-253.
Meinshausen, M.; Meinshausen, N.; Hare, W.; Raper, S.C.B.; Frieler, K.; Knutti, R.; Frame, D.J.and
Allen, M.R (2009). Greeenhouse-gas emission targets for limiting global warming to 2°C.
Nature 458:1158-1162.
Mitchell, S.R., Harmon, M.E. and O'Connell, K.E.B. (2012) Carbon debt and carbon sequestration
parity in forest bioenergy production. Global Climate Change Bioenergy (2012), doi:
10.1111/j.1757-1707.2012.01173.x
Mroz, G.D.; Jurgensen, M.F.and Frederick, D.J. (1985). Soil nutrient changes following whole tree
harvesting on 3 Northern Hardwood Sites. Soil Science Society of America Journal, 1552-1557.
Murray, B.C.; McCarl, B.A.; Lee, H. (2004). Estimating Leakage from Forest Carbon Sequestration
Programs. Land Economics, 109-124.
50

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Muth, J. (1992). Rational Expectations and the Theory of Price Movements. In International Library of
Critical Writings in Economics, Volume 1 (pp. 3-23). Aldershot, UK: Elgar.
National Research Council. (2010). Verifying Greenhouse Gas Emissions: Methods to Support
International Climate Agreements. Washington, D.C.: The National Academies Press.
Nave, L.E.; Vance, E.D.; Swanston, C.W. and Curtis, P.S. (2010). Harvest impacts on soil carbon
storage in temperate forests. Forest Ecology and Management, 857-866.
Olson, J. S. (1963). Energy Storage and the Balance of Producers and Decomposers in Ecological
Systems. Ecology, 322 - 331.
Rabl, A., Benoist, A., Dron, D., Peuportier, B., Spadaro, J. and Zoughaib, A. (2007). How to Account
for C02 Emissions from Biomass in a Life Cycle Analysis. International Journal of Life Cycle
Analysis.
Richter, D.D.; Markewitz, D.; Trumbore, S.W. and Wells, C.G. (1999). Rapid accumulation and
turnover of soil carbon in a re-establishing forest. Nature, 56-58.
Rose, S. and B. Sohngen. 2011. Global Forest Carbon Sequestration and Climate Policy Design.
Journal of Environment and Development Economics. Vol. 16, no. 4. : 429-454.
Sanchez, F.G.; Coleman, M.; Garten Jr., C.T.; Luxmoore, R.J.; Stanturf, J.W.; Trettin, C. and
Wullschleger, S.D. (2007). Soil carbon, after 3 years, under short-rotation woody crops grown
under varying nutrient and water availability. Biomass andBioenergy, 793-801.
Schiffman, P.M.and Johnson, W.C. (1989). Phytomass and detrital carbon storage during forest
regrowth in the southeatsern U.S. Piedmont. Canadian Journal of Forest Research, 69-78.
Searchinger, T., Hamburg, S., Melillo, J., and Chameides, W. (2009). Fixing a Critical Climate
Accounting Error. Science, 326.
Sedjo, R. (2010). The Biomass Crop Assistance Program: Some Implications for the Forest Industry.
Washington, D.C. Resources for the Future.
Sedjo, R., and Lyon, K. (1990). The Long Term Adequacy of World Timber Supply. Washington, D.C.
Resources for the Future.
Sedjo, R., and Sohngen, B. (In Press, 2012). Wood as a Major Feedstock for Biofuel Production in the
U.S.: Impacts on Forests and International Trade. Journal of Sustainable Forests.
Sedjo, R. and Tian, X. (2012). What is the Carbon Footprint of Wood Biomass Energy. Journal of
Forestry, Forthcoming.
51

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Selig, M.F.; Seiler, J.R.and Tyree, M.C. (2008). Soil carbon and C02 efflux as influenced by the
thinning of loblolly pine. Forest Science, 58-66.
Sohngen, B., and Sedjo, R. (1998). A Comparison of Timber Market Models: Static Simulation and
Optimal Control Approaches. Forest Science, 24-36.
Sohngen, B. and Sedjo, R. (2006). Carbon Sequestration in Global Forests Under Different Carbon Price
Regimes. The Energy Journal: Special issue, Multi-Greenhouse Gas Mitigation and Climate
Policy, 109-162.
Stanton, E., Bueno, R., Ackerman, F., Erickson, P.; Hammerschlag, R. and and Cegan, J. (2011).
Consumption-Based Greenhouse Gas Emissions Inventory for Oregon - 2005. Technical Report.
Somerville, MA: Stockholm Environment Institute-U.S. Center.
Tang, J.; Qi, Y.; Xu, M.; Misson, L.; Goldstein, A.H. (2005). Forest thinning and soil respiration in a
ponderosa pine plantation in the sierra Nevada. Tree Physiology, 57-66.
Tolbert, V.R.; Thornton, F.C.; Joslin, J.D.; Bock, B.R.; Bandaranayake, W.; Hoiuston, A.; Tyler, D.;
Pettry, D.W. and Green, T.H. (2000). Increasing belowground carbon sequestration with
conversion of agricultural lands to production of bioenergy crops. New Zealand Journal of
Forest Science, 138-149.
USDA Forest Service. (2012). The Future of America's Forests and Rangelands - The 2010 Resources
Planning Act (RPA) Assessment. Washington, D.C.: USDA, In Press.
Verified Carbon Standard Association, (n.d.). Verified Carbon Standard: A Global Benchmark for
Carbon. Retrieved from http://www.v-c-s.org/methodologies/find
Vokoun, M., Wear, D., and Abt, R. (2009). Testing for Change in Structural Elements of Forest
Inventories. Forest Science, 455 - 466.
Wear, D. (2011). Forecasts of county-level land uses under three future scenarios: a technical document
supporting the Forest Service 2010 RPA Assessment. Ashville, NC: U.S. Department of
Agriculture Forest Service, Southern Research Station.
Zilberman, D.; Hocman, G. and Rajagopal, D. (2011). On the Inclusion of Indirect Land Use in Biofuel
Regulations. University of Illinois Law Review, 413-434.
52

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
APPENDIX A: Charge to the Panel
MEMORANDUM
To:	Holly Stall worth, DFO
Science Advisory Board Staff Office
From:	Paul Gunning, Acting Director
Climate Change Division
Subject: Accounting Framework for Biogenic Carbon Dioxide (CO2) Emissions from
Stationary Sources and Charge Questions for SAB peer review
The purpose of this memorandum is to transmit the draft Accounting Framework for Biogenic
CO2 Emissions study and the charge questions for consideration by the Science Advisory Board
(SAB) during your upcoming peer review in fall 2011.
In January 2011, the U.S. Environmental Protection Agency (EPA) announced a series of steps it
would take to address biogenic CO2 emissions from stationary sources. In addition to specific
regulatory action, EPA committed to conduct a detailed examination of the science and technical
issues related to accounting for biogenic CO2 emissions and to develop an accounting framework
for those emissions. The study transmitted today is that examination.
The study identifies key scientific and technical factors that should be considered when
constructing any framework for accounting for the impact of utilizing biologically-based
feedstocks at stationary sources. It then provides EPA's recommendations on those issues and
presents a framework for "adjusting" estimates of onsite biogenic CO2 emissions (i.e., a
"biogenic accounting factor" or BAF) on the basis of information about the carbon cycle.
As indicated in the accompanying materials, advice on these issues will be important as EPA
moves through the steps to address biogenic CO2 emissions from stationary sources. We look
forward to the SAB's review.
Please contact me if you have any questions about the attached study and charge.
A-l

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Charge Questions
EPA is providing this study, Accounting Framework for Biogenic CO 2 Emissions from
Stationary Sources (September 15, 2011), to the Science Advisory Board (SAB) to review
EPA's approach on accounting for biogenic CO2 emissions from stationary sources, including
the scientific basis and methodological components necessary to complete that accounting.
Objective
EPA is charging the SAB to review and comment on (1) EPA's characterization of the science
and technical issues relevant to accounting for biogenic CO2 emissions from stationary sources;
(2) EPA's framework, overall approach, and methodological choices for accounting for these
emissions; and (3) options for improving upon the framework for accounting for biogenic CO2
emissions.
This charge does not ask the SAB for regulatory recommendations or legal interpretation of the
Clean Air Act statutes related to stationary sources.
Charge Questions
1.	Evaluation of the science of biogenic CO2 emissions
In reviewing the scientific literature on biogenic CO2 emissions, EPA assessed the underlying
science of the carbon cycle, characterized fossil and biogenic carbon reservoirs, and discussed
the implications for biogenic CO2 accounting. Does the SAB support EPA's assessment and
characterization of the underlying science and the implications for biogenic CO2 accounting?
2.	Evaluation of biogenic CO2 accounting approaches
In this report, EPA considered existing accounting approaches in terms of their ability to reflect
the underlying science of the carbon cycle and also evaluated these approaches on whether or not
they could be readily and rigorously applied in a stationary source context in which onsite
emissions are the primary focus. On the basis of these considerations, EPA concluded that a new
accounting framework is needed for stationary sources.
2(a). Does the SAB agree with EPA's concerns about applying the IPCC national approach
to biogenic C02 emissions at individual stationary sources?
2(b). Does the SAB support the conclusion that the categorical approaches (inclusion and
exclusion) are inappropriate for this purpose, based on the characteristics of the carbon
cycle?
2(c). Does the SAB support EPA's conclusion that a new framework is needed for situations
in which only onsite emissions are considered for non-biologically-based (i.e., fossil)
feedstocks?
A-2

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
5(b). Does the EPA provide sufficient information to support how EPA has applied the
accounting framework in each case?
5(c). Are there alternative approaches or case studies that EPA should consider to illustrate
more effectively how the framework is applied to stationary sources?
6. Overall evaluation
Overall, this report is the outcome of EPA's analysis of the science and technical issues
associated with accounting for biogenic CO2 emissions from stationary sources.
6(a). Does the report - in total - contribute usefully to the advancement of understanding on
accounting for biogenic C02 emissions from stationary source?
6(b). Does it provide a mechanism for stationary sources to adjust their total onsite emissions
on the basis of the carbon cycle?
6(c). Does the SAB have advice regarding potential revisions to this draft study that might
enhance the utility of the final document?
A-4

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
APPENDIX B: Temporal Changes in Stand Level Biogenic Emissions Versus Fossil
Emissions
Cherubini et al. (2011) analyzes temperature increases on the basis of GWP (global warming potential)
whereas Cherubini et al. (2012) analyzes climate impacts using GTP (global temperature potential).
GWP is the time integral of the change in radiative forcing from a pulse emission of CO2 (in this case,
from harvested biomass) and subsequent sequestration by biomass growth, whereas GTP is the integral
of actual temperature response to a pulse emission of CO2 and subsequent sequestration by biomass
growth. Both studies use a simple contrived comparison of biogenic emissions from a single stand over
hundreds of years to comparable fossil emissions. Much is assumed regarding for instance global
activity and emissions, and climate and carbon cycle dynamics. Also, importantly, landscape responses
and investment behavior are not reflected which represent concurrent and related emissions and
sequestration that affect net global emissions changes.
Both studies incorporate a suite of carbon uptake mechanisms (such as oceanic uptake) in addition to
regrowth in forest stands. In this context, the GTPbio, discussed by Cherubini (2012), is a more accurate
metric for the actual climate response. The idea of the GTPbio is simple: it represents the increase in
global average temperature over a given period due to a transient increase in carbon dioxide in the
atmosphere (between the initial biomass combustion or respiration and the ultimate regrowth of the
carbon stock) relative to the temperature response to a release of an equivalent amount of fossil CO2 at
time 0 (expressed as a fraction between 0 and 1). To calculate a GTPbio value, a time scale must be
specified. The calculation for GTPbio is the ratio of the average temperature increase with biogenic
emissions followed by reabsorbtion by biomass regrowth over, say, 100 years divided by the average
temperature increase from the initial emission alone over 100 years. For short accumulation feedstocks,
such as perennial grasses, GTPbio would be a very small fraction due to fast carbon accumulation times
(ignoring leakage effects). For feedstocks with long accumulation times, one must compute the change
in global temperature over time, accounting for the decline in temperature change as carbon is
reabsorbed.
Cherubini et al. (2011, 2012) provide an artificial simplified example for a single forest stand. The same
type of metric could be used to compare temperature changes or changes in radiative forcing associated
with increased biomass energy use for one year or more for a landscape or nation - taking into account
the land carbon change over time associated with increased biomass energy use. This would involve
comparison of a business as usual case to an increased biomass use case. A simpler metric that compares
the cumulative radiative forcing of biogenic feedstocks to the cumulative radiative forcing of fossil fuels
over time could also be used, e.g.. Cherubini's GWPbio. However the broader literature should be
considered regarding the climate implications of alternative emissions pathways (see charge question 1
response) while considering uncertainty in global emissions, climate response and the carbon cycle.
Figure B-l demonstrates the importance of the time horizon or, more specifically, the weight to place on
temperature increases that occur in the short term versus temperature increases that occur later. Consider
a scenario in which biomass is harvested, but the carbon stock is replaced within a 100 year time scale.
The GTPbio for a 100-year regrowth and a 100 year time horizon is roughly 0.5, meaning that the time-
integrated global average temperature increase within that 100 year period is 50% of the temperature
B-l

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
increase caused by an equivalent amount of fossil carbon (or straight CO2 release without regrowth of
biomass). However, using the average temperature increase for the biogenic case over 100 years masks
the fact that although there will be an initial increase in temperature near the beginning of the 100 year
period the reabsorption of carbon in the forest will bring the effect on ground temperature to nearly zero
by year 100, giving an average temperature that was 50% of the average fossil temperature increase over
100 years. In fact the instantaneous temperature change for the biogenic case falls below zero slightly
before 100 years because oceans initial absorb extra CO2 in response to the initial biogenic emission (see
Figure B-l, adapted from Cherubini 2012, Figure 5a). The temperature effect equilibrates to zero as the
ocean CO2 is balanced. A more precise picture of intertemporal effects is shown in Figure B-l, adapted
from Cherubini et al. (2012).
Figure B-l: Surface temperature change from biogenic emissions versus fossil fuel over time. Adapted from
Cherubini et al. (2012) and reprinted with copyright permission.
Cherubini et al. (2012) have shown that if biomass is harvested and the carbon is reabsorbed within a
100 year time scale, the global average temperature increase over that 100 year period is 50% of the
temperature increase caused by an equivalent amount of fossil carbon. We might conclude that biogenic
emissions are roughly 50% as damaging as fossil fuels, however the high point of temperature increase
created by biogenic emissions occurs early in the 100 year cycle and is back to zero by the time the
carbon is reabsorbed. For the case where carbon is recovered within 100 years Cherubini et al. (2012)
have shown that at 20 years, the average temperature increase (over 20 years) from biogenic fuel is 97%
B-2

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
1	of the temperature increase caused by an equivalent amount of fossil carbon; for years 21 to 100 years,
2	the average increased is 0.37 and for years 101 to 500, the increase is 0.02.
3	A current practice for international reporting under IPCC guidelines and international treaty negotiations
4	is to use greenhouse gas emissions and sink values that represent the cumulative radiative forcing for
5	greenhouse gases over a 100 year period with uniform weighting over 100 years. Greenhouse gas values
6	are reported in tons CO2 equivalent where one ton of CO2 equivalent is an index for the cumulative
7	radiative forcing for a pulse emission of one ton of CO2 over 100 years. The CO2 equivalent for a ton of
8	other greenhouse gases is given by how many times more radiative forcing it produces over 100 years
9	compared to CO2 (e.g., 21 times for CH4) (EPA 2012).
B-3

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
APPENDIX C:
Fate of Landscape Residue after Harvest and System Storage of
Carbon
The decomposition of materials left after harvest can be estimated from the negative exponential decay
equation (Olson 1963): Ct=Co exp[-kt] where Ct=is the amount at any time t, Co is the initial amount, k
is the rate-constant of loss, and t is time. Solving this function for a range of rate-loss constants results in
the relationship shown in Figure C-l for a range of k that covers the most likely range for decomposition
rates of leafy to woody material in North America. In no case does the store instantaneously drop to zero
as assumed in the Framework.
CD
c
'c
03
E
CD
-i—•
=3
CL
C
4—
O
c
o
•c
o
CL
o
100
Time since harvest (years)
Figure C-l: Fate of residue/slash left after harvest as function of k and time since harvest.
C-l

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
The amount of carbon stored on average in a forest system or fuel-shed comprised of units or stands that
generate equal amounts of residue or slash is given by: I/k, where I is the average forest input of residue
or slash. To create a relative function independent of the amount of residue or slash created, the input of
each harvest unit or stand can be set to either 1 (to give the proportion of the input) or 100 (to give a
percent of the input). The average forest input (I) would therefore be equal to 1/Rr or 100/ RH where Rr
is the harvest return interval. Using this relationship to solve the average store relative to the input is
presented in Figure C-2 for the most likely range of decomposition rates for leafy to woody material in
North America. This indicates that there are a wide range of possible cases in which the store of residue
or slash can exceed the initial input (shown by the horizontal line indicating storage of 1). This means
that combusting this material will cause the store to drop by the amount indicated, and this amounts to
the net flux of carbon to the atmosphere. To a large degree there is a negative relationship between the
harvest interval and k; materials with high values of k (i.e., leafy) are typically harvested with short
intervals between harvests and material with low values of k (i.e., large wood) are typically harvested
with long interval between harvests. This suggests that the effect of harvesting residues and slash is
largely independent of the loss rate-constant.
CD
c
'c
03
E
CD
L_
-t—'
=3
CL
C
c
o
'-4—'
L_
o
CL
o
100
Harvest interval-years
Figure C-2: Landscape average store of residue/slash as function of k and harvest interval.
C-2

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
APPENDIX D: Carbon Balances over Time in an Existing Forest System
To determine whether a forest harvest system for existing forest acreage creates a carbon debt, or
alternatively, a gain it is appropriate to examine this problem at the landscape-level (or in the context of
biogenic carbon a fuel-shed basis). Note the discussion that follows refers only to existing managed
forests (and their stored carbon) and not broader landscape effects such as the expansion or contraction
of forest area. At the forest system level there are three possible cases: (1) a relatively constant, steady-
state store of carbon if the harvest system is continued unchanged, (2) an increase of carbon stores to a
higher steady state if the intensity of harvest declines, and (3) a decrease of carbon stores to a higher
steady-state if the intensity of harvest increases. These cases are illustrated in Figures 4-6 which are
based on the online Forest Sector Carbon Calculator used in the forest system landscape mode
(http://landcarb.forestry.oregonstate.edu/default.aspx).
In Figure D-l, a 50-year clear-cut harvest rotation was practiced until 2010 and then continued for 500
years. This resulted in no carbon debt. If tracked at the stand scale one would see carbon levels rising
and falling, but over time the net balance is zero. In contrast, if one converted the 50-year clear-cut
harvest rotation system to a 25-year clear-cut harvest rotation system as in Figure D-2 there would have
been a decline in carbon stores in the ecosystem. This decline would be considered a carbon debt and
while not permanent (i.e., forever), it would remain as long as the 25-year management system persists.
If the 50-year clear-cut harvest rotation was replaced by a 100-year clear-cut system at year 2010, then
there would have been a gain carbon stores (Figure D-3). That gain would remain as long as that 100-
year clear-cut system of management was maintained. All these simulations all assumed that soil
productivity is maintained regardless of harvest interval.
At the existing forest level (as opposed to the stand level), live, dead, and soil stores all acted the same.
Each of these pools either remained in balance (i.e., no net gain) or could increase or decrease depending
on how the interval of harvest changes. The steady-state store of all three pools is controlled by the I/k
relationship developed by Olson (1963), where / is the input of carbon to the pools and k is the
proportion lost from the system in respiration and harvest (the live also has a loss related to mortality of
trees). As the harvest interval decreases the input to the pool (I) decreases and the proportion lost via
harvest (k) increases. This explains why the ecosystem stores decrease when the harvest interval is
shortened and why they increase when the harvest interval is increased. A similar response happens
when one takes a larger share of the carbon stores away when there is a harvest.
These dynamics have several important implications that need to be considered in the context of
biogenic carbon: (1) long-term carbon debts, gains, and balances are best examined at the forest system-
level (not to mention the broader agriculture-forest landscape level), (2) all forest carbon pools can
exhibit either debts, gains, or remain relatively constant, (3) most systems of forest management will
reach a steady-state if maintained over a long enough period and this steady-state can be maintained as
long as the management system is continued, and (4) ultimately reaching a steady-state does not
determine if there has been a loss or gain in carbon as this depends on how harvest management changes
from one steady-state to the next.
D-l

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Total Stores for Forest Ecosystem (Live + Dead + Stable Soils)
Stable (Soil) Stores ¦¦ Dead Ecosystem Stores	Live Ecosystem Stores
2	Figure D-l: Changes in carbon stores of major forest ecosystem pools when a 50 year clear-cut harvest system is
3	established and continued. The result is a continued carbon balance.
Total Stores for Forest Ecosystem (Live + Dead + Stable Soils)
Stable (Soil) Stores	Dead Ecosystem Stores	Live Ecosystem Stores
350
300
'J?
5 250
0
oj
i. 200
V)
®
1	150
H 100
cu
o
50
0
1910 1960 2010 2060 2110 2160 2210 2260 2310 2360 2410 2460 2510
Year of Simulation
meh - landscape level run - 4/3/2012 2:25:36 PM PST
4
5	Figure D-2: Changes in carbon stores of major forest ecosystem pools when a 50 year clear-cut harvest system is
6	replaced by a 25 year clear-cut harvest system in 2010. The result is a carbon debt.
7
D-2

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
1
2
3
600—,
Total Stores for Forest Ecosystem (Live + Dead + Stable Soils)
I Stable (Soil) Stores	Dead Ecosystem Stores	Live Ecosystem Stores
1910 1960 2010 2060 2110 2160 2210 2260 2310 2360 2410 2460 2510
Year of Simulation
meh - landscape level run - 4/3/2012 2:37:09 PM PST
Figure D-3: Changes in carbon stores of major forest ecosystem pools when a 50 year clear-cut harvest system is
replaced by a 100 year clear-cut harvest system in 2010. The result is a carbon gain.
D-3

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
APPENDIX E: Dissenting Opinion from Dr. Roger Sedjo
Introduction
EPA's Science Advisory Board (SAB) was asked to review and comment on the EPA's Accounting
Framework for Biogenic C02 Emissions from Stationary Sources (Framework September 2011). The
motivation for the Accounting Framework "is whether and how to consider biogenic greenhouse gas
emission in determining thresholds ... for Clean Air Act permitting" (p. 4). To my knowledge the SAB
Report has been complete and is being submitted to the broader SAB process. The comments below
(and page numbers cited) relate to the SAB Report draft of 6-15-12 (SAB 2012).
I take fundamental issue with many of the elements of the SAB Report. Although I largely agree with
the Report's criticisms of the absence of supporting science for many of the Framework's suggested
approaches, I find unconvincing and unscientific much of the Report's attempt to salvage large elements
Framework's approach. My comments focus largely, but not entirely, to forest issues in the Report not
only because that is the area of my greatest expertize but also because the defects in the Framework
approach are most egregious in forestry.
The EPA considered whether to categorically include biogenic emission in its greenhouse gas
accounting or whether to categorically exclude biogenic emissions (p 6-7). The Agency rejected both
extremes and asked the SAB whether it supported their conclusion that categorical approaches are
inappropriate for treatment of biogenic carbon emissions. However, I do not believe that this issue was
properly vetted within the SAB process. Although the statement that "carbon neutrality cannot be
assumed for all biomass energy a priori" (p 7) is correct, it misrepresents the serious position developed
by the Intergovernmental Panel on Climate Change (IPCC 2006) and commonly used included a critical
qualification regarding the condition of land cover generally and forest stock specifically. This
requirement is missing from the simplistic evaluation statement. This position is supported in the
Appendix to this piece, (USDA appendix by Hohenstein, 2012), which notes that the major IPCC
rationale does not claim "a priori" neutrality. The IPCC, which suggested this approach, makes carbon
neutrality contingent on an aggregate monitoring approach that focuses on the changes in aggregate land
use and forests. Thus, the definitive development of the wide spread exclusion of biogenic and wood
does not, in fact, involve an a priori assumption of neutrality. Rather it involves a qualification (for
wood) that the forest stock be constant or expanding. I should note here that consideration of that
important qualification was largely absence from the evaluation by the SAB and, in my judgment,
aggressively discouraged by the organizers from the SAB discussion.
Finally, if the proposed Accounting Framework were capable of providing reliable accounting, one
might give it serious consideration as an alternative to the IPCC approach in achieving the EPA
objectives. However, as is acknowledged by the Report (e.g., p. 15), the proposed Accounting
Framework is replete with problems as are the calculations of the elements necessary for calculating the
Biological Accounting Factor (BAF). The acknowledged scientific weaknesses in the EPA document are
identified throughout the SAB Report.
E-l

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
This paper demonstrates below that the SAB Report has not adequately addressed some of these issues
and has not found ways to estimate in a scientifically acceptable way the values of some of the requisite
components of the BAF.
Defects in the Accounting Framework
Questions raised in the Report about the Framework run from the appropriateness of the proposed use of
the same accounting framework for the various feedstocks, which are different, to issues dealing with
the appropriate baseline and questions concerning the relevant timescale. The SAB Report essentially
embraces a variant of the BAF approach, which was developed in the Framework, even though the
Report points to numerous important weaknesses of the BAF approach. The BAF is a simple accounting
model that tries to identify and measure the various components and impacts of carbon emissions and
accumulations from biomass energy sources. Ultimately, the Report essentially embraces the general
BAF approach but applies it differently to individual biogenic feedstocks. However, the Report
acknowledges throughout that a number of the components of the BAF cannot be adequately measured.
For example, the Report acknowledges that for important major elements of the Framework, e.g.,
leakage, there is no satisfactory monitoring or measurement system. Leakage, which can be either
positive of negative, may involve the deflection of deforestation and associated emission out of
woodshed under consideration or it may involve sequestration associated with offsetting forest
management outside of that woodshed. Thus, the values of these major elements are essentially
empirical, could be either positive or negative, but have their impacts outside of the area of direct
observation. But, without accurate leakage values, the BAF approach proposed cannot accurately
estimated for carbon changes. It cannot even determine the sign of the changes with any great accuracy.
Thus, although the Reports states that "it is important to have scientifically sound methods to account
for greenhouse gas emission caused by human activities" (p 13), it acknowledges that the it is widely
acknowledged in the literature that leakage cannot to be readily measured with any accuracy (Murray et
al. 2004; Macauley et al. 2009). Nevertheless, in contradiction of this finding the Report suggests that
"the Agency ... try to ascertain the directionality of net leakage ... and incorporate that information into
decision making." (p 9-10). This suggestion flies in the face of the concept of "scientifically sound
methods."
Indeed, the application of the proposed framework would either need to leave these elements of the BAF
empty, as suggested in the USDA letter posted on the SAB website, or nonscientific guesses would need
to be imposed, as suggested in parts of the Report. In either case large errors in measurement appear
almost inevitable and, rather than providing the regulators with accurate information, would provide
misinformation to regulators and would likely redound to errors in the application of regulations. The
idea introduced in the Report of default BAFs does not do anything to address their fundamental lack of
scientific rigor.
Other thorny issues involve questions of the boundaries of a woodshed and/or a region, which relate to
the leakage question, the intermixing of industrial wood and biomass so that significant portions of any
harvest are used for each, and the export of biomass for energy, e.g., the large flow of wood pellets to
Europe, where their emissions for the production of bioenergy will not be captured in the accounting.
Finally, any accounting approach that tries to monitor each biomass using unit is surely going to be time
E-2

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
consuming and expensive, perhaps too expensive to justify the use of the biomass for energy (Sedjo and
Sohngen 2012).
An important defect is that the Report embraces a carbon-debt framework. However, this framework is
an artifact of an arbitrary decision of how the accounting system is applied. If the forest is sustainability
managed, then there is no carbon-debt. Withdrawals equal growth for both biomass and carbon.
Accounting debts can occur in some circumstances, however. For a mature forest stand, if the
accounting period begins with the harvest of the stand, as in the Manomet Study, a debt is incurred for
that stand. Note that net carbon sequestration could be occurring in that forest but on different stands.
Most forests are multi-aged and hence will have net growth occurring on some stands while stock
reductions occur on other stands.
An additional source of confusion regarding carbon debt is related to the accounting period. If the
accounting focuses on a stand and the accounting period begins with the harvest, a debt will be
establishment for the forest stand. However, if the accounting begins with the forest establishment, e.g.,
at tree planting, then the initial post planting growth is building up a stock of carbon that will be released
at harvest. Thus, any future debt from that stand will have been offset in advance of the harvest and no
intertemporal net carbon debt is incurred.
Thus, although an accounting debt can be found for mature stands, the debt is an artifact of the time
period selected and the choice of how narrowly to define the relevant forest stands. Furthermore, a
carbon debt will not be occurred for sustainably managed forests. In the aggregate, the U.S. forest
system is more than sustainable as demonstrated by the FIA's data going back to a least 1952. Thus, a
fully accounting of the entire managed US forest does not find a carbon-debt.
In summary, the Report identifies a host of problems with the proposed Accounting Framework, and
reports that "the SAB did not find the Framework to be scientifically rigorous" (p 30). Indeed, although
the Framework is said to "include most of the elements that would be needed to gauge changes in C02
emissions," the problems with the effective of monitoring, measurement and verification of several of
the components are daunting.
Alternative Approaches for Accounting for Biogenic Carbon
One wonders why the SAB exerted so much effort to try to save the Accounting Framework, containing
as it does, such fundamental defects. It is my understanding that the SAB was asked to review and
comment on the Framework, but not necessarily to save it. Indeed, as noted above, EPA's change
included the question of "whether ... to consider biogenic greenhouse gas emission in determining
thresholds ... for Clean Air Act permitting" (p. 4).
Nevertheless, despite the identification of very serious defects in the approach, there is a considerable
attempt in the SAB process to downplay the problems and ignore the lack of scientific bases for
measuring some of the elements, apparently in order to preserve a variant of the approach, no matter
how defective.
There are at least two basic ways that one might approach the problem of estimating the net emissions
associated with biogenic energy. The highly regarded scientific organization, Intergovernmental Panel
on Climate Change (IPCC) has suggested an aggregate approach that would focus on the changes in
E-3

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
aggregate land use and forests to determine whether, for example, aggregate forest stocks are expanding
or contracting. This approach has been supported by the USDA (Hohenstein 2012) in a response to an
earlier draft report by the SAB.
In the context of measuring the total aggregate forest the issue of leakage and anticipatory management
within the US does not arise since to total system is evaluated. Where the aggregate is subdivided into a
few large international regions, these issues are more easily captured since flows in forest biomass are
measured in the international trade statistics and individual woodshed monitoring is not necessary.
Indeed, for the US this approach can easily be put in place at low cost since the Forest Service has been
undertaking Forest Inventory Assessments (FIA) for over fifty years.
The alternative to the IPCC approach, suggested by the Accounting Framework, involves the individual
audit of each separate woodshed associated with a facility and an attempt to estimate the impact of each
individual operation on net emissions. Such an approach would be a monitoring nightmare complicated
by the fact that wood feedstock could, and likely would on occasion, be brought into one region from
other small regions as required, this situation would involve leakage. Leakage could be replete since
more regions would almost surely involve more leakage. Not only is the individual wood shed audit
approach much more expensive, it also is inadequate since wood sheds are not always well defined and
wood will undoubtedly flow across various woodsheds and leakage will occur. However, such detail is
entirely unnecessary for purposes of the broad monitoring of biogenic facilities and their effects on
atmospheric carbon. The relevant consideration is not the infinitesimal impact of each individual
facility. Rather, the concern is with the grand aggregate impact of the bioenergy system on net
emissions. If this approach does not properly account for the effects of leakage and anticipatory forest
management (reverse leakage), the BAF estimates will have basic errors.
The Framework approach and the SAB Report appear to accept the notion that the Framework
Accounting approach is superior to the IPCC approach. However, no evidence of this is provided either
in argumentation or in analytical studies. Nevertheless, it is probably indisputable that the costs of the
Accounting Framework approach with its estimated BAFs are far higher than those associated with the
IPCC approach.
Five Summarizing Points
First, the guidelines provided by the EPA for the SAB Report essentially accept the Framework view
and dismisses the IPCC suggested approach with regard to biogenic feedstocks within the land use
sector, including forests. This was done despite that fact that there was no serious discussion by our
SAB group of the adequacy or viability of the IPCC approach. Indeed the IPCC approach was dismissed
by the EPA as inadequate on rather flimsy grounds. I note that my position is supported in the letter by
William Hohenstein, Director of the Climate Change Program Office posted at the SAB website. The
letter states that USDA "prefers the IPCC accounting framework" approach and takes issue with the
rationale used by the SAB Report and its dismissal of the IPCC approached. USDA differs with the
assertion of the SAB Report and maintains "the IPCC approach is not equivalent to an a priori
assumption that these feedstocks are produced in a carbon neutral manner or an assertion that land use
activities contributing feedstocks to the energy sector can be managed without consideration of
atmospheric outcome."
E-4

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
Second, an attempt to assess the carbon debt of individual stands fundamentally misses the point since it
is the entire forest, not individual stands that are relevant to the carbon footprint as seen by the
atmosphere. As such, the attempt to imperfectly apply the BAF to individual forests is costly and
irrelevant to the aggregate U.S. carbon footprint.
Third, although the Report acknowledges the dynamic nature of market driven supply systems that
would be providing the biogenic energy feedstock, it essentially uses a static approach that largely
ignores various market responses and adaptations to changing circumstances. Although the Report
acknowledges that investment decisions for trees must predate their utilization by years and indeed
decades, this reality is not incorporated into any BAF calculation. Indeed, while investment decisions
must be driven by the anticipation of the existence and size of future markets, these considerations are
acknowledged for wood biomass in parts of the Report and then disregarded in the application of the
approach for regulatory purposes. Thus, the actual approach suggested is essentially static, missing the
essential dynamic nature of the supply process. Despite these basic defects, the Report recommendations
are treated as if they are scientifically sound.
Fourth, the Report erroneously states that incentives for producing replacement bioenergy crops are
absence. Such a result would occur in viable markets only if there were no anticipation of increasing
future demand. However, a variety of signals, including requirements of renewal portfolio standards and
forecasts of dramatic biomass energy demand increases over the next couple of decades by various
authoritative organizations, e.g., EIA.
Fifth, the Report tends to support a very expensive and onerous regulatory accounting system rather than
a much more efficient system such as suggested by the IPCC. This support is given without any apparent
serious assessment or rationale that the regulatory results of the BAF system will be equal to or superior
to those that would result from a much less expensive and less onerous IPCC type approach.
In summary, I find that although the SAB Report provides a useful critique of the Accounting
Framework and the BAF approach. However the Report falls into the trap of trying to make a basically
defective system functional and tends to support many aspects of that flawed system. In the end the
Report largely ignores its own criticisms and supports a fundamentally flawed approach. Thus, since the
motivation for the Accounting Framework "is whether and how to consider biogenic greenhouse gas
emission in determining thresholds ... for Clean Air Act permitting" (p. 4), it can rationally be
concluded that biogenic greenhouse gas emission are best not considered in determining thresholds or
perhaps considered only of the forest and land use conditions as such that they do not meet minimal
IPCC conditions.
References:
Hohenstein, William. 2012. Comments submitted to the EPA SRB regarding the Environmental
Protection Agency's draft Accounting Framework for Biogenic C02 Emissions from Stationary
Sources on behalf of USD A. May 25.
IPCC. 2006. Guidelines for National Greenhouse Gas Inventories: Agriculture, Forestry and Other Land
Use Volume 4.
E-5

-------
Science Advisory Board (SAB) Draft Report (7/26/12) for Quality
Review — Do Not Cite or Quote —
This draft has not been reviewed or approved by the chartered SAB and
does not represent EPA policy.
1	Macauley, Molly K., Daniel F. Morris, Roger A. Sedjo, Kate Farley, Brent L. Sohngen "Forest
2	Measurement and Monitoring: Technical Capacity and "How Good Is Good Enough?" RFF
3	Report | December 2009.
4	Murray, B.C.; McCarl, B.A.; Lee, H. (2004). Estimating Leakage from Forest Carbon Sequestration
5	Programs. Land Economics, 109-124.
6	Sedjo, Roger and Brent Sohngen. 2012. "Carbon Sequestration in Forests and Soils," in Annual Review
7	of Resources Economics.
8	SAB. 2012. 6-15-12 Deliberative Draft report of the Biogenic Carbon Emission Panel.
9
E-6

-------