S Cilia
And <
r- -* OHO Emissions . Si
^ * -*£t%-^«F- * ** •» v "^"i, _ "*ji
! * m Jk^klVA ^
Sinks & Emission Offsets
-------
-------
SOLID WASTE MANAGEMENT AND GREENHOUSE GASES
A Life-Cycle Assessment of Emissions and Sinks
2nd EDITION
May 2002
-------
-------
TABLE OF CONTENTS
EXECUTIVE SUMMARY: BACKGROUND AND FINDINGS ES-1
ES.l GREENHOUSE GASES AND CLIMATE CHANGE ES-1
ES.2 WHAT IS THE UNITED STATES DOING ABOUT CLIMATE CHANGE? ES-2
ES.3 WHAT IS THE RELATIONSHIP OF MUNICIPAL SOLID WASTE TO
GREENHOUSE GAS EMISSIONS? ES-3
ES.4 WHY EPA PREPARED THIS REPORT AND HOW IT HAS BEEN USED ES-4
ES.5 HOW WE ANALYZED THE IMPACT OF MUNICIPAL SOLID WASTE ON
GREENHOUSE GAS EMISSIONS ES-5
ES.6 RESULTS OF THE ANALYSIS ES-9
ES.7 OTHER LIFE-CYCLE GHG ANALYSES AND TOOLS ES-18
ES.8 LIMITATIONS OF THE ANALYSIS ES-20
1. METHODOLOGY 1
1.1 THE OVERALL FRAMEWORK: A STREAMLINED LIFE-CYCLE
INVENTORY 2
1.2 MSW MATERIALS CONSIDERED IN THE STREAMLINED LIFE-CYCLE
INVENTORY 2
1.3 KEY INPUTS AND BASELINES FOR THE STREAMLINED LIFE-CYCLE
INVENTORY 4
1.4 SUMMARY OF THE LIFE-CYCLE STAGES 7
1.5 ESTIMATING AND COMPARING NET GHG EMISSIONS 13
2. RAW MATERIALS ACQUISITION AND MANUFACTURING 15
2.1 GHG EMISSIONS FROM ENERGY USE IN RAW MATERIALS
ACQUISITION AND MANUFACTURING 15
2.2 NON-ENERGY GHG EMISSIONS FROM MANUFACTURING AND RAW
MATERIALS ACQUISITION 18
2.3 RESULTS 19
2.4 LIMITATIONS 19
3. FOREST CARBON SEQUESTRATION 29
3.1 MODELING FRAMEWORK 30
3.2 THE NORTH AMERICAN PULP AND PAPER (NAPAP) MODEL 32
3.3 THE TIMBER ASSESSMENT MARKET MODEL (TAMM) AND THE
AGGREGATE TTMBERLAND ASSESSMENT SYSTEM (ATLAS) 38
3.4 THE FOREST CARBON MODEL (FORCARB) 42
3.5 THE HARVESTED CARBON MODEL (WOODCARB) 44
3.6 APPLYING THE MODELS FOR WOOD PRODUCTS 46
3.7 RESULTS 48
3.8 LIMITATIONS 52
4. SOURCE REDUCTION AND RECYCLING 55
4.1 GHG IMPLICATIONS OF SOURCE REDUCTION 55
4.2 GHG IMPLICATIONS OF RECYCLING 57
4.3 SOURCE REDUCTION WITH MATERIAL SUBSTITUTION 60
4.4 LIMITATIONS • • 61
5. COMPOSTING • • 65
5.1 POTENTIAL GREENHOUSE GAS EMISSIONS 66
5.2 POTENTIAL CARBON STORAGE 66
5.3 NET GHG EMISSIONS FROM COMPOSTING 76
5.4 LIMITATIONS 78
6. COMBUSTION • • 81
-------
6.1 METHODOLOGY 83
6.2 RESULTS \""""".".93
6.3 LIMITATIONS OF THE ANALYSIS '. 93
7. LANDFILLING "Z.'""!"!!!!!!!"."."!97
7.1 EXPERIMENTAL VALUES FOR METHANE GENERATION AND CARBON
STORAGE 98
7.2 FATES OF LANDFILL METHANE: CONVERSION TO CO2, EMISSIONS,
AND FLARING OR COMBUSTION WITH ENERGY RECOVERY 101
7.3 UTILITY CO2 EMISSIONS AVOIDED 103
7.4 NET GHG EMISSIONS FROM LANDFILLING 103
7.5 LIMITATIONS 103
8. ACCOUNTING FOR EMISSION REDUCTIONS .'.".'""."."".'"'.".'.'113
8.1 NET GHG EMISSIONS FOR EACH WASTE MANAGEMENT OPTION 113
8.2 APPLYING EMISSION FACTORS 115
8.3 OTHER LIFE-CYCLE GHG ANALYSES AND TOOLS 118
8.4 OPPORTUNITIES FOR GHG REDUCTIONS 119
-------
EXECUTIVE SUMMARY: BACKGROUND AND FINDINGS
In the 21st century, management of municipal solid waste (MSW) continues to be an
important environmental challenge facing the United States. In 2000, the United States generated
232 million tons of MSW, an increase of 13 percent over 1990 generation levels and 53 percent
over 1980 levels.1 Climate change is also a serious issue, and the United States is embarking on a
number of voluntary actions to reduce the emissions of greenhouse gases (GHGs) that can
intensify climate change. By presenting material-specific GHG emission factors for various waste
management options, this report examines how the two issues—MSW management and climate
change—are related.
Among the efforts to slow the potential for climate change are measures to reduce
emissions of carbon dioxide (COa) from energy use, decrease emissions of methane (CHO and
other non-carbon dioxide GHGs, and promote long-term storage of carbon in forests and soil.
Management options for MSW provide many opportunities to affect these processes, directly or
indirectly. This report integrates information on the GHG implications of various management
options for some of the most common materials in MSW. To our knowledge, this work represents
the most complete national study on climate change emissions and sinks from solid waste
management practices. The report's findings may be used to support a variety of programs and
activities, including voluntary reporting of emission reductions from waste management
practices.
ES.1 GREENHOUSE GASES AND CLIMATE CHANGE
Climate change is a serious international environmental concern and the subject of much
research and debate. Many, if not most, of the readers of this report will have a general
understanding of the greenhouse effect and climate change. However, for those who are not
familiar with the topic, a brief explanation follows.2
A naturally occurring shield of "greenhouse gases" (primarily water vapor, carbon
dioxide, methane, and, nitrous oxide), comprising 1 to 2 percent of the Earth's atmosphere,
absorbs some of the solar radiation that would otherwise be radiated to space and helps warm the
planet to a comfortable, livable temperature range. Without this natural "greenhouse effect," the
average temperature on Earth would be approximately -2 degrees Fahrenheit, rather than the
current 57 degrees Fahrenheit.3
Many scientists are alarmed by a significant increase in the concentration of CC-2 and
other GHGs in the atmosphere. Since the pre-industrial era, atmospheric concentrations of CO2
have increased by nearly 30 percent and CH4 concentrations have more than doubled. There is a
growing international scientific consensus that this increase has been caused, at least in part, by
1 U.S. EPA Office of Solid Waste, Municipal Solid Waste in the United States: 2000 Facts and
Figures, EPA (2002), p. 2.
2 For more detailed information on climate change, please see The 2001 Inventory of U.S.
Greenhouse Gas Emissions and Sinks: 1990-1999,
(http://www.epa.gov/globalwarming/publications/emissions/us2001/index.html) (April 2001); and Climate
Change 2001: The Scientific Basis (J.T. Houghton, et al., eds. Intergovernmental Panel on Climate Change
[IPCC]; published by Cambridge University Press, 2001). To obtain a list of additional documents
addressing climate change, access EPA's global warming Web site at www.epa.gov/globalwarming.
3 Climate Change 2001: The Scientific Basis, op. cit., pp. 89-90.
ES-1
-------
human activity, primarily the burning of fossil fuels (coal, oil, and natural gas) for such activities
as generating electricity and driving cars.4
Moreover, in international scientific circles a consensus is growing that the buildup of
CC>2 and other GHGs in the atmosphere will lead to major environmental changes such as (1)
rising sea levels that may flood coastal and river delta communities; (2) shrinking mountain
glaciers and reduced snow cover that may diminish fresh water resources; (3) the spread of
infectious diseases and increased heat-related mortality; (4) possible loss in biological diversity
and other impacts on ecosystems; and (5) agricultural shifts such as impacts on crop yields and
productivity.5 Although reliably detecting the trends in climate due to natural variability is
difficult, the most accepted current projections suggest that the rate of climate change attributable
to GHGs will far exceed any natural climate changes that have occurred during the last 1,000
years.6
Many of these changes appear to be occurring already. Global mean surface temperatures
already have increased by about 1 degree Fahrenheit over the past century. A reduction in the
northern hemisphere's snow cover, a decrease in Arctic sea ice, a rise in sea level, and an increase
in the frequency of extreme rainfall events all have been documented.7
Such important environmental changes pose potentially significant risks to humans,
social systems, and the natural world. Many uncertainties remain regarding the precise timing,
magnitude, and regional patterns of climate change and the extent to which mankind and nature
can adapt to any changes. It is clear, however, that changes will not be easily reversed for many
decades or even centuries because of the long atmospheric lifetimes of GHGs and the inertia of
the climate system.
ES.2 WHAT IS THE UNITED STATES DOING ABOUT CLIMATE CHANGE?
In 1992, world leaders and citizens from some 200 countries met in Rio de Janeiro,
Brazil, to confront global ecological concerns. At this "Earth Summit," 154 nations, including the
United States, signed the Framework Convention on Climate Change, an international agreement
to address the danger of global climate change. The objective of the Convention was to, stabilize
GHG concentrations in the atmosphere over time at a level at which man-made climate
disruptions would be minimized.
By signing the Convention, countries made a voluntary commitment to reduce GHGs or
take other actions to stabilize emissions of GHGs. All Parties to the Convention were required to
develop and periodically update national inventories of their GHG emissions. The United States
ratified the Convention in October 1992. One year later, the United States issued its Climate
Change Action Plan (CCAP), which calls for cost-effective domestic actions and voluntary,
cooperation with states, local governments, industry, and citizens to reduce GHG emissions.
In order to achieve the goals outlined in the Climate Change Action Plan, EPA initiated
several new voluntary programs to realize the most cost-effective opportunities for reducing
emissions. For example, in 1994 EPA created the Landfill Methane Outreach Program, which
aims to reduce landfill CEU emissions by facilitating the development of projects that use landfill
4Ibid.,p.l. ,
5 J.J. McCarthy, et al., eds. 2001. Climate Change 2001: Impacts, Adaptation, and Vulnerability.
IPCC. Cambridge University Press, pp. 9-13.
6 Climate Change 2001: The Scientific Basis, op. cit, p. 2.
7 Ibid., p. 4.
ES-2
-------
gas to produce energy.8 In that same year, EPA introduced the Climate and Waste Program, with
its focus on a broader set of waste management practices and climate protection.
To date, EPA's voluntary partnership programs for climate protection have achieved
substantial environmental results. In 2000 alone, these programs reduced GHG emissions by, 35
million metric tons of carbon equivalent (MMTCE)—the equivalent of eliminating the emissions
from approximately 25 million cars. In addition, substantial CHu emission reductions—estimated
at more than 1 MMTCE for the period from 1999-2000—are being obtained as an ancillary
benefit of Clean Air Act (CAA) regulatory requirements that were promulgated in 1996. These
reductions,are expected to rise to nearly 47 MMTCE by 2004.
Meanwhile, an increasing number of states have been instituting their own voluntary
actions to reduce emissions. Thirty-nine states and Puerto Rico have created GHG Inventories for
their own emissions. Twenty-five states and Puerto Rico have completed or initiated state action
plans, which list steps to reduce emissions. At least six of these states—^Delaware, Iowa,
Minnesota, Montana, New Jersey, and Oregon—have incorporated the reduction of waste into
their, GHG mitigation strategies. Finally, a few states—including California, Maine, New
Hampshire, and Wisconsin—are in the process of establishing GHG registries, which enable
companies and other entities to report voluntary emission reductions.
ES.3 WHAT IS THE RELATIONSHIP OF MUNICIPAL SOLID WASTE TO
GREENHOUSE GAS EMISSIONS?
What does MSW have to do with rising sea levels, higher temperatures, and GHG
emissions? For many wastes, the materials in MSW represent what is left over after a long series
of steps: (1) extraction and processing of raw materials; (2) manufacture of products; (3)
transportation of materials and products to markets; (4) use by consumers; and (5) waste
management.
Virtually every step along this "life cycle" impacts GHG emissions. Waste management
decisions can reduce GHGs by affecting one or more of the following:
(1)' Energy consumption (specifically, combustion of fossil fuels) associated with making,
transporting, using, and disposing the product or material that becomes a waste.
(2) Non-energy-related manufacturing emissions, such as the CO2 released when
limestone is converted to lime (which is needed for use in aluminum and steel
manufacturing).
(3) CHU emissions from landfills where the waste is disposed.
(4) Carbon sequestration, which refers to natural or man-made processes that remove
carbon from the atmosphere and store it for long periods or permanently.
The first three mechanisms add GHGs to the atmosphere and contribute to global
warming. The fourth—carbon sequestration—reduces GHG concentrations by removing CO2
from the atmosphere. Forest growth is one mechanism for sequestering carbon; if more biomass
is grown than is removed (through harvest or decay), the amount of carbon stored in trees
increases, and thus carbon is sequestered.
8 The Landfill Methane Outreach Program (LMOP) is a voluntary assistance and partnership
program that helps facilitate and promote the use of landfill gas as a renewable energy source. By
controlling landfill gas instead of allowing it to migrate into the air, the LMOP helps businesses, states, and
communities protect the environment and build a sustainable future. The program has an Internet home
page (http://www.epa.gov/landfill.html) and can be reached via a toll-free hotline number (800-782-7937).
ES-3
-------
Different wastes and waste management options have different implications for energy
consumption, CHL^ emissions, and carbon sequestration. Source reduction and recycling of paper
products, for example, reduce energy consumption, decrease combustion and landfill emissions,
and increase forest carbon sequestration.
ES.4 WHY EPA PREPARED THIS REPORT AND HOW IT HAS BEEN USED
Recognizing the potential for source reduction and recycling of municipal solid waste to
reduce GHG emissions, EPA included a source reduction and recycling initiative in the original
1994 Climate Change Action Plan and set an emission reduction goal based on a preliminary
analysis of the potential benefits of these activities. It was clear that a rigorous analysis would be
needed to gauge more accurately the total GHG emission reductions achievable through source
reduction and recycling. That all of the options for managing MSW should be considered also
became clear. By addressing a broader set of MSW management options, a more comprehensive
picture of the GHG benefits of voluntary actions in the waste sector could be determined and the
relative GHG impacts of various waste management approaches could be assessed. To this end,
EPA launched a major research effort, the results of which were published in the first edition of
this report in September 1998. This edition of the report includes additional materials and
incorporates updated data affecting many of the material-specific results. The emission factors
presented will continue to be updated and improved as more data become available. The latest
emission factors, reflecting these ongoing revisions, can be found on the EPA Global Warming
Web site .
The primary application of the GHG emission factors in this report is to support
mitigation accounting for waste management practices that mitigate climate change. In recent
years, the emission factors have been applied for this purpose in a number of ways. In
conjunction with the U.S. Department of Energy, EPA has used these estimates to develop
guidance for voluntary reporting of GHG reductions, as authorized by Congress in Section
1605(b) of the Energy Policy Act of 1992.
Other applications have included quantifying the GHG reductions from voluntary
programs aimed at source reduction and recycling, such as EPA's WasteWise and Pay-As-You-
Throw programs. EPA also has worked with the Climate Neutral Network to develop company-
specific GHG "footprints" for the network's member companies, who have pledged to become
GHG "neutral" through emission reductions or offset activities.
The international community has shown considerable interest in using the emission
factors—or adapted versions—to develop GHG emissions estimates for non-U.S. solid waste
streams.9 For example, Environment Canada recently employed our life-cycle methodology and
components of our analysis to develop a set of Canada-specific GHG emission factors to support
analysis of waste-related mitigation opportunities.10
Additionally, EPA worked with the International Council for Local Environmental
Initiatives (ICLEI) to incorporate GHG emission factors into its municipal GHG accounting
software. Currently, 350 communities participate in ICLEFs Cities for Climate Protection
Campaign, which helps them establish a GHG emission reduction target and implement a
Note that waste composition and product life cycles vary significantly among countries. This
report may assist other countries by providing a methodologic framework and benchmark data for
developing GHG emission estimates for their solid waste streams.
Environment Canada. 2001. Determination of the Impact of Waste Management Activities on
Greenhouse Gas Emissions. Prepared by ICF Consulting, Tome-Smith Associates, and Enviros-RIS.
ES-4
-------
comprehensive local action plan designed to achieve that target/Currently, we are exploring other
options for broadening the use of our research internationally.
To make it easier for organizations to use these emission factors, EPA created the Waste
Reduction Model (WARM) spreadsheet tool.11 WARM enables waste managers and other users
to calculate changes in total GHG emissions quickly by entering in information on baseline and
alternative waste management practices. By applying the appropriate material-specific emission
factors for each practice, the tool generates an estimate of the net GHG impact from
implementing the alternative waste management practice as compared to the baseline practice.
ES.5 HOW WE ANALYZED THE IMPACT OF MUNICIPAL SOLID WASTE ON
GREENHOUSE GAS EMISSIONS
To measure the GHG impacts of MSW, one must first decide which wastes to analyze.
We surveyed the universe of materials and products found in MSW and determined those that are
most likely to have the greatest impact on GHGs. These determinations were based on (1) the
quantity generated; (2) the differences in energy use for manufacturing a product from virgin
versus recycled inputs; and (3) the potential contribution of materials to CHU generation in
landfills. By this process, we limited the analysis to the following 16 items:
• Aluminum Cans;
• Steel Cans;
• Glass;
• HDPE (high-density polyethylene) Plastic;
• LDPE (low-density polyethylene) Plastic;
• PET (polyethylene terephthalate) Plastic;
• Corrugated Cardboard;
• Magazines/Third-class Mail;
« Newspaper;
• Office Paper;
• Phonebooks;
• Textbooks;
• Dimensional Lumber;
• Medium-density Fiberboard;
• Food Discards; and 'r
• Yard Trimmings.
The foregoing materials constitute 64.4 percent, by weight, of MSW, as shown in Exhibit
ES-1.12
11 WARM is available on the EPA Web site:
http://www.epa.gov/globalwarming/actions/waste/warm.htm.
12 Note that these data are based on national averages. The composition of solid waste varies
locally and regionally; local or state-level data should be used when available.
ES-5
-------
We also examined the GHG implications of managing mixed paper, mixed plastics,
mixed organics, mixed recyclables, and mixed MSW.
• Mixed paper is recycled in large quantities
and is an important class of scrap material in
many recycling programs. Presenting a single
definition of mixed paper is difficult,
however, because recovered paper varies
considerably, depending on the source. For
purposes of this report, we identified three
categories of mixed paper according to the
dominant source—broad (general sources),
office, and residential.
• Mixed plastics is comprised of HDPE, LDPE,
and PET, and is estimated by taking a
weighted average of the 2000 recovery rates
for these three plastic types.
• Mixed organics is a weighted average of food
discards and yard trimmings, using generation
rates for 2000.
• Mixed recyclables are materials that are
typically recycled. As used in this report, the
term includes the items listed in Exhibit ES-1,
except food discards and yard trimmings. The
emission factors reported for mixed
recyclables represent the average GHG
emissions for these materials, weighted by the
tonnages at which they were recycled in 2000.
• Mixed MSW is comprised of the waste
material typically discarded by households
and collected by curbside collection vehicles;
it does not include white goods (e.g.,
refrigerators, toasters) or industrial waste.
This report analyzes mixed MSW on an "as-
disposed" (rather than "as-generated") basis.
Exhibit ES-1
Percentage of 2000 U.S. Generation of MSW for
Materials in This Report
Material
Aluminum Cans
Steel Cans
Glass
HDPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber (listed
as "Wood - Containers and
Packaging)
Medium-density Fiberboard
Food Discards
Yard Trimmings
TOTAL
Percentage of
MSW Generation
(by Weight)
0.7%
1.1%
5.5%
1.6%
1.3%
0.8%
13.0%
3.3% .
6.5%
3.2%
0.3% '
0.5%
3.4%
NA
11.2%
12.0% ;
64.4%
Source: U.S. EPA. 2002. Municipal Solid Waste in
the United States: 2000 Facts and Figures, EPA
530-R-02-001.
We developed a streamlined life-cycle inventory for each of the selected materials. Our
analysis is streamlined in the sense that it examines GHG emissions only and is not a more
comprehensive environmental analysis of all emissions from municipal solid waste management
options."
13
13 EPA's Office of Research and Development (ORD) performed a more extensive application of
life-cycle assessment for various waste management options for MSW. A decision support tool (DST) and
life-cycle inventory (LCI) database for North America have been developed with funding by ORD through
a cooperative agreement with the Research Triangle Institute (RTI) (CR823052). This methodology is
based on a multi-media, multi-pollutant approach and includes analysis of GHG emissions as well as a
broader set of emissions (air, water, and waste) associated with MSW operations. At the time of publication
of this report, the MSW-DST is available for site-specific applications. For further information, contact
Keith Weitz at rti.org or (919) 541-6973. The LCI database is expected to be released in 2002. The Web
site address for further information is: .
ES-6
-------
We focused on those aspects of the life cycle that have the potential to emit GHGs as
materials change from their raw states to products and then to waste. Exhibit ES-2 shows the
steps in the life cycle at which GHGs are emitted, carbon sequestration is affected, and utility
energy is displaced. As shown, we examined the potential for these effects at the following points
in a product's life cycle: ...'..
• Raw material acquisition (fossil fuel energy and other emissions, and changes in forest
carbon sequestration); ;
• Manufacturing (fossil fuel energy emissions); and
• Waste management (CO2 emissions associated with composting, non-biogenic CO2 and
nitrous oxide (N2O) emissions from combustion, and CH-t emissions from landfills); these
emissions are offset to some degree by carbon storage in soil and landfills, as well as
avoided utility emissions from energy recovery at combustors and landfills.
At each of these points, we also
considered transportation-related energy
emissions. Estimates of GHG emissions
associated with electricity used in the raw
materials acquisition and manufacturing steps
are based on the nation's current mix of energy
sources,14 including fossil fuels, hydropower, and
nuclear power. Estimates of GHG emission
reductions attributable to utility emissions
avoided from waste management practices,
however, are based solely on the reduction of
fossil fuel use.15
We did not analyze the GHG emissions
associated with consumer use of products
because energy use for the selected materials is
small (or zero) at this point in the life cycle. In
addition, the energy consumed during use would
be approximately the same whether the product
was made from virgin or recycled inputs.
To apply the GHG estimates developed
in this report, one must compare a baseline
scenario with an alternative scenario, on a life-
cycle basis. For example, we could compare a
baseline scenario, where 10 tons of office paper
are manufactured, used, and landfilled, to an
alternative scenario, where 10 tons are
manufactured, used, and recycled.
Improvements to the First Edition
This report is the second edition of Greenhouse Gas
Emissions from Management of Selected Materials in
Municipal Solid Waste. This edition includes the following
improvements:
• Incorporates new data on energy and recycling loss rates
from EPA's Office of Research and Development;
• Expands the analysis of the GHG benefits of composting,
including results of CENTURY model runs;
• Develops emission factors for five new material types:
magazines/third-class mail, phonebooks, textbooks,
dimensional lumber, and medium-density fiberboard;
• Develops emission factors for two new categories of
mixed materials: mixed plastics and mixed organics;
• Incorporates new energy data into calculations of utility
offsets;
•. Revises carbon coefficients and fuel use for national
average electricity generation;
• Updates information on landfill gas recovery rates;
• Adds a discussion of emerging issues in the area of
climate change and waste management; and
• Provides a list of suggested proxy values for voluntary
reporting of GHG emission reductions.
These changes and/or revisions are described in more detail
throughout the report.
14 The emissions are based on the current national grid mix, as opposed to regional grids.
15-.
We adopted this approach based on suggestions from several reviewers who argued that fossil
fuels should be regarded as the marginal fuel displaced by waste-to-energy and landfill gas recovery
systems. -
ES-7
-------
Virgin inputs
.'*»'
Materials Extracted:
Trees, Ore, Oil, etc.
ZOO®
,^;«vr**.m,»&
terf3r\
Energy
Energy
Exhibit ES-2 Greenhouse Gas Sources and Sinks Associated with the Material Life Cycle
life Cycle Stage GHG Emissions Sinks & Emission Offsets
Raw Materials
Acquisition
Energy and
Non-Energy-
Related Emissions
^»*%
f |J i
Reduced Carbon
Sequestration
in Forests
CO:
Energy and Non-
Energy-Related Emissions
Recycling
, . _ ."--r •increased, „";•:
::. AvoidKlFosaLv,;: , :~ ForestCarBon—--'
l ':-*
^ Waste. '
Management
Energy-Related
Emissions
*• ' • *J . W^ /•, <, -t^"'- "•« —v.'
2V i
Emissions Emissions
Uncontrolled CH4
Emissions or CH4 Flared
and Recovered Energy
ES-8
-------
Exhibit ES-3 shows how GHG sources and sinks are affected by each waste management
strategy. For example, the top row of the exhibit shows that source reduction16 (1) reduces GHG
emissions from raw materials acquisition and manufacturing; (2) results in an increase in forest
carbon sequestration; and (3) does not result in GHG emissions from waste management. The
sum of emissions (and sinks) across all steps in the life cycle represents net emissions.
Exhibit ES-3 Components of Net Emissions for Various MSW Management Strategies
MSW
Management
Strategy
Source Reduction
Recycling
Composting (food
discards, yard
trimmings)
Combustion
Landfllling
GHG Sources and Sinks
Raw Materials Acquisition and
Manufacturing
Decrease in GHG emissions,
relative to the baseline of
manufacturing
Decrease in GHG emissions due to
lower energy requirements
(compared to manufacture from
virgin inputs) and avoided process
non-energy GHGs
No emissions/sinks
No change
No change
Changes in Forest or
Soil Carbon Storage
Increase in forest carbon
sequestration (for organic
materials)
Increase in forest carbon
sequestration (for organic
materials)
Increase in soil carbon
storage
No change
No change
Waste Management
No emissions/sinks
Process and transportation
emissions associated with
recycling are counted in the
manufacturing stage
Compost machinery emissions
and transportation emissions
Non-biogenic CO2, N2O
emissions, avoided utility
emissions, and transportation
emissions
QHLt emissions, long-term
carbon storage, avoided utility
emissions, and transportation
emissions
ES.6 RESULTS OF THE ANALYSIS
Management of municipal solid waste presents many opportunities for GHG emission
reductions. Source reduction and recycling can reduce GHG emissions at the manufacturing
stage, increase forest carbon sequestration, and avoid landfill CELt emissions. When waste is
combusted, energy recovery displaces electricity generated by utilities by burning fossil fuels
(thus reducing GHG emissions from the utility sector), and landfill CJHU emissions are avoided.
Landfill CHt emissions can be reduced by using gas recovery systems and by diverting organic
materials from landfills. Landfill CHt can be flared or utilized for its energy potential. When used
for its energy potential, landfill CHt displaces fossil fuels, as with MSW combustion.
16-
In this analysis, the source reduction techniques we analyze involve using less of a given
product without using more of some other product—e.g., making aluminum cans with less aluminum
("lightweighting"); double-sided rather than single-sided photocopying; or reuse of a product. We did not
consider source reduction of one product that would be associated with substitution by another product—
e.g., substituting plastic boxes for corrugated paper boxes. Nor did we estimate the potential for source
reduction of chemical fertilizers and pesticides with increased production and use of compost. For a
discussion of source reduction with material substitution, see Section 4.3.
ES-9
-------
In order to support a broad portfolio of climate change mitigation activities covering a
range of GHGs, many different methodologies for estimating emissions will be needed. The
primary result of this research is the development of material-specific GHG emission factors that
can be used to account for the climate change benefits of waste management practices.
Exhibits ES-4 and ES-5 present the GHG impacts of source reduction, recycling,
composting, combustion, and landfilling. The impacts are presented on a per-ton managed basis
for the individual and mixed materials, using the waste generation reference point. Exhibit ES-4
presents these values in MTCE/ton, and Exhibit ES-5 presents the values in metric tons of carbon
dioxide equivalent/ton (MTCO2E/ton). For comparison, Exhibits ES-6 and ES-7 show the same
results (in MTCE/ton and MTCO2E/ton, respectively) using the raw material extraction reference
point. In these tables, emissions for 1 ton of a given material are presented across different
management options.17 The life-cycle GHG emissions for each of the first four waste
management strategies—source reduction, recycling, composting, and combustion—are
compared to the GHG emissions from landfilling in Exhibits ES-8 and ES-9. These exhibits show
the GHG values for each of the first four management strategies, minus the GHG values for
landfilling. With these exhibits, one may compare the GHG emissions of changing management
of 1 ton of each material from landfilling (often viewed as the baseline waste management
strategy) to one of the other waste management options.
All values shown in Exhibits ES-4 through ES-9 are for national average conditions (e.g.,
average fuel mix for raw material acquisition and manufacturing using recycled inputs; typical
efficiency of a mass burn combustion unit; national average landfill gas collection rates). GHG
emissions are sensitive to some factors that vary on a local basis, and thus site-specific emissions
will differ from those summarized here.
Following is a discussion of the principal GHG emissions and sinks for each waste
management practice and the effect that they have on the emission factors:
• Source reduction, in general, represents an opportunity to reduce GHG emissions in a
significant way.18 For many materials, the reduction in energy-related CO2 emissions from the
raw material acquisition and manufacturing process, and the absence of emissions from waste
management, combine to reduce GHG emissions more than other options.
• For most materials, recycling has the second lowest GHG emissions. For these materials,
recycling reduces energy-related CO2 emissions in the manufacturing process (although not
as dramatically as source reduction) and avoids emissions from waste management. Paper
recycling increases the sequestration of forest carbon.
• Composting is a management option for food discards and yard trimmings. The net GHG
emissions from composting are lower than landfilling for food discards (composting avoids
CEU emissions), and higher than landfilling for yard trimmings (landfilling is credited with
the carbon storage that results from incomplete decomposition of yard trimmings). Overall,
given the uncertainty in the analysis, the emission factors for composting or combusting these
materials are similar.
17 Note that the difference between any two values for a given material in Exhibit ES-4 (i.e.,
emissions for the same material in two waste management options) is the same as the difference between
the two corresponding values in Exhibit ES-5.
18 As noted above, the only source reduction strategy analyzed in this study is lightweighting.
Consequently, the results shown here do not directly apply to material substitution.
ES-10
-------
Exhibit ES-4
Net GHG Emissions from Source Reduction and MSW Management Options - Emissions Counted from a Waste
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mxed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW as Disposed
Generation Reference Point (MTCE/Ton)1
Source
Reduction2
-2.49
-0.79
-0.14
-0.49
-0.61
-0.49
-0.51
-1.04
-0.81
-0.80
-1.28
-1 .23
-0.55
-0.60
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling
-4.11
-0.49
-0.08
-0.38
-0.47
-0.42
-0.71
-0.74
-0.95
-0.68
-0.91
-0.75
-0.67
-0.67
NA
NA
-0.67
-0.67
-0.83
-0.41
-0.76
NA
NA
Composting3
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.05
-0.05
NA
NA
NA
NA
NA
-0.05
NA
Combustion4
0.02
-0.42
0.01
0.23
0.23
0.28
-0.19
-0.13
-0.21
-0.18
-0.21
-0.18
-0.22
-0.22
-0.05
-0.06
-0.19
-0.18
-0.17
0.25
-0.17
-0.06
Landfilling5
0.01
0.01
0.01
0.01
0.01
0.01
0.08
-0.12
-0.21
0.62
-0.21
0.62
-0.10
-0.10
0.17
-0.09
0.10
0.07
0.15
0.01
0.05
0.03
NA: Not applicable, or in the case of composting of paper, not analyzed.
'MTCE/ton: Metric tons of carbon equivalent per short ton of material. Material tonnages are on an as-managed
Source reduction assumes initial production using the current mix of virgin and recycled inputs.
''There is considerable uncertainty in our estimate of net GHG emissions from composting;
the values of zero are plausible values based on assumptions and a bounding analysis.
"Values are for mass bum facilities with national average rate of ferrous recovery.
Values reflect estimated national average methane recovery in year 2000.
ES-11
(wet weight) basis.
-------
Exhibit ES-5
Net GHG Emissions from Source Reduction and MSW Management Options - Emissions Counted from a Waste Generation
Reference Point (MTC02En"on)1
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSVtf as Disposed
Source Reduction2
-9.15
-2.89
-0.50
-1.79
-2.25
-1.78
-1.89
' -3.80
-2.97
-2.95
-4.70
-4.49
-2.01
-2.20
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling
-15.07
-1.79
-0.28
-1.40
-1.71
-1.55
-2.60
-2.70
-3.48
-2.48
-3.34
-2.74
-2.45
-2.47
NA
NA
-2.47
-2.47
-3.05
-1.51
-2.80
NA
NA
Composting3
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.20
-0.20
NA
NA
NA
NA
NA
-0.20
NA
Combustion4
0.06
-1.53
0.05
0.85
0.85
1.04
-0.68
-0.49
-0.77
-0.65
-0.77
-0.65
-0.81
-0.81
-0.19
-0.23
-0.68
-0.68
-0.62
0.93
-0.61
-0.21
-0.13
Landfiliinq5
0.04
0.04
0.04
0.04
0.04
0.04
0.28
-0.44
-0.76
2.28
-0.76
2.28
-0.38
-0.38
0.62
-0.34
0.37
0.25
0.56
0.04
0.19
0.12
0.24
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1MTCO2E/ton: Metric tons of carbon dioxide equivalent per short ton of material. Material tonnages are on an as-managed (wet weight) basis.
2Source reduction assumes initial production using the current mix of virgin and recycled inputs.
^here is considerable uncertainty in our estimate of net GHG emissions from composting;
the values of zero are plausible values based on assumptions and a bounding analysis.
Values are for mass burn facilities with national average rate of ferrous recovery.
5Values reflect estimated national average methane recovery in year 2000.
ES-12
-------
Exhibit ES-6
Net GHG Emissions from Source Reduction and WISW Management Options - Emissions Counted from a Raw
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
rood Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW as Disposed
Materials Extraction Reference Point (MTCE/Ton)
Source
Reduction1
0.00
0.00
0.00
o.oo
0.00
0.00
-0.28
-0,58
-0.3,5
-0.50
-0.65
-0.64
-0.50
-0.50
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling2
-1.61
0.30
0.06
0.10
0.15
0.06
-0.47
-0.28
-0.49
-0.37
-0.27
-0.16
-0.6?
-0.58
NA
NA
-0.30
-0.3Q
0.02
O.Q9
-0.40
NA
NA
Composting2
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0,05
-0.05
NA
NA
NA
NA
NA
-0.05
NA
Combustion2
2.51
0.37
0.15
0.72
0.85
0.77
0.05
0.33
0.25
0.13
0.42
0.41
-0.17
-0.12
-0.05
-0.06
0.19
0.19
0.68
0.76
0.19
-0.06
Note ihat totals may not add due to rounding, and more digits may be displayed than are significant.
Landfilling2
2.50
0.80
0.15
0.50
0.63
0.50
0.32
0.34
0.25
0.93
0.43
1.21
-0.06
-0.01
0.17
-O.Q9
0.48
0.45
1.01
0.52
0.41
0.03
NA: Not applicable, or in the case of composting of paper, not analyzed.
'Source reduction assumes initial production using the current mix of virgin and recycled inputs.
Includes emissions from the initial production of the material being managed, except for foodwaste, yard waste, and mixed MSW.
ES-13
-------
Exhibit ES-7
Net GHG Emissions from Source Reduction and MSW Management Options - Emissions Counted from a Raw
Materials Extraction Reference Point (MTC02E/Ton)
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW as Disposed
Source
Reduction1
0.00
0.00
0.00
0.00
0.00
0.00
-1.01
-2.11
-1.29
-1.82
-2.37
-2.35
-1.84
-1.84
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling2
-5.92
1.09
0.22
0.38
0.54
0:23
-1.72
-1.02
-1.79
-1.36
-1.01
-0.60
-2.28
-2.11
NA
NA
-1.09
-1.08
0.07
0.34
-1.48
NA
NA
Composting2
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.20
-0.20
NA
NA
NA
NA
NA
-0.20
NA
Combustion2
9.21
1.35
0.55
2.64
3.11
2.82
0.20
1.20
0.91
0.47
1.56
1.49
-0.64
-0.45
-0.19
-0.23
0.70
0.71
2.50
2.79
0.71
-0.21
-0.13
Landfilling2
9.18
2.92
0.54
1.82
2.29
1.82
1.16
1.25
0.92
3.41
1.57
4.43
-0.21
-0.03
0.62
-0.34
1.76
1.64
3.69
1.89
1.51
0.12
0.24
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
'Source reduction assumes initial production using the current mix of virgin and recycled inputs.
Includes emissions from the initial production of the material being managed, exceptfor foodwaste, yard waste, and mixed MSW.
ES-14
-------
Exhibit ES-8
GHG Emissions of MSW Management Options Compared to Landfilling1 (MTCEVTon)
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Dhonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
rood Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW as Disposed
Source Reduction2
Net Emissions Minus
Landfilling Net
Emissions (Current
Mix)
-2.50
-0.80
-0.15
-0.50
-0.63
-0.50
-0.59
-0.92
-0.60
-1.43
-1.07
-1.85
-0.44
-0.50
NA
NA
NA
NA
NA
NA
NA
NA
NA
Net Emissions
Minus Landfilling
Net Emissions
(100% Virgin
Inputs)
-4.68
-1.02
-0.17
-0.54
-0.65
-0.59
-1.03
-1.07
-1.11
-1.63
-1.19
-1.94
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling Net
Emissions Minus
Landfilling Net
Emissions
-4.12
-0.50
-0.09
-0.39
-0.48
-0.43
-0.79
-0.62
-0.74
-1.30
-0.70
-1.37
-0.56
-0.57
NA
: NA
-0.78
-0.74
-0.99
-0.42
: -0.82
NA
NA
Composting3 Net
Emissions Minus
Landfilling Net
Emissions
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.22
0.04
NA
NA
NA
NA
NA
-0.09
NA
Combustion4 Net
Emissions Minus
Landfilling Net
Emissions
0.01
-0.43
0.00
0.22
0.22
0.27
-0.26
-0.01
0.00
-0.80
0.00
-0.80
-0.12
-0.12
-0.22
0.03
-0.29
-0.25
-0.32
0.24
-0.22
-0.09
-0.10
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
'Values for landfilling reflect projected national average methane recovery in year 2000.
2Source reduction assumes initial production using the current mix of virgin and recycled inputs.
Calculation is based on assuming zero net emissions for composting.
Values are for mass burn facilities with national average rate of ferrous recovery.
ES-15
-------
Exhibit ES-9
GHG Emissions of MSW Management Options Compared to Landfilling1 (MTCO2E/Ton)
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW as Disposed
Source Reduction2
Net Emissions Minus
Landfilling Net
Emissions (Current
Mix)
-9.18
-2.92
-0.54
-1.82
-2.29
-1.82
-2.17
-3.36
-2.21
-5.23
-3.94
-6.78
-1.63
-1.82
NA
NA
NA
NA
NA
NA
NA
NA
NA
source Reduction
Net Emissions
Minus Landfilling
Net Emissions
(100% Virgin
Inputs)
-17.15
-3.72
-0.61
-1.99
-2.38
-2.18
-3.79
-3.94
-4.07
-5.99
-4.37
-7.13
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling Net
Emissions Minus
Landfilling Net
Emissions
-15.11
-1.83
-0.32
-1.44
-1.75
-1.59
-2.88
-2.26
-2.72
-4.77
-2.57
-5.03
-2.07
-2.09
NA
NA
-2.84
-2.72
-3.62
-1.55
-2.99
NA
NA
Composting3 Net
Emissions Minus
Landfilling Net
Emissions
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.82
0.15
NA
NA
NA
NA
NA
-0.32
NA
Combustion4 Net
Emissions Minus
Landfilling Net
Emissions
0.02
-1.57
0.01
0.81
0.81
1.00
-0.96
-0.05
-0.01
-2.94
-0.01
-2.94
-0.43
-0.43
-0.81
0.11
-1.06
-0.93
-1.18
0.90
-0.80
-0.33
-0.38
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
'Values for landfilling reflect projected national average methane recovery in year 2000.
2Source reduction assumes initial production using the current mix of virgin and recycled inputs.
'Calculation is based on assuming zero net emissions for composting.
"Values are for mass bum facilities with national average rate of ferrous recovery.
ES-16
-------
• The net GHG emissions from combustion of mixed MSW are lower than landfilling
mixed MSW (under national average conditions for landfill gas recovery). Because, in
practice, combustors and landfills manage a mixed waste stream, net emissions are
determined more by technology factors (e.g., the efficiency of landfill gas collection
systems and combustion energy conversion) than by material specificity. Material-
specific emissions for landfills and combustors provide a basis for comparing these
options with source reduction, recycling, and composting.
The ordering of combustion, landfilling, and composting is affected by (1) the GHG
inventory accounting methods, which do not count CO2 emissions from sustainable biogenic
sources,19 but do count emissions from sources such as plastics; and (2) a series of assumptions
on sequestration, future use of CH4 recovery systems, system efficiency for landfill gas recovery,
ferrous metals recovery, and avoided utility fossil fuels. On a site-specific basis, the ordering of
results between a combustor and a landfill could be different from the ordering provided here,
which is based on national average conditions.
We conducted sensitivity analyses to examine the GHG emissions from landfilling under
varying assumptions about (1) the percentage of landfilled waste sent to landfills with gas
recovery, and (2) CtLt oxidation rate and gas collection system efficiency. The sensitivity
analyses demonstrate that the results for landfills are very sensitive to these factors, which are
site-specific.20 Thus, using a national average value when making generalizations about emissions
from landfills masks some of the variability that exists from site to site.
The scope of this report is limited to developing emission factors that can be used to
evaluate GHG implications of solid waste decisions. We do not analyze policy options in this
report. Nevertheless, the differences in emission factors across various waste management
options are sufficiently large as to imply that GHG mitigation policies in the waste sector can
make a significant contribution to U.S. emission reductions. A number of examples, using the
emission factors in this report, bear this out.
• At the firm level, targeted recycling programs can reduce GHGs. For example, a
commercial facility that shifts from (a) a baseline practice of landfilling (in a landfill with
no gas collection system) 50 tons office paper and 4 tons of aluminum cans to (b)
recycling the same materials can reduce GHG emissions by more than 100 MTCE.
• At the community level, a city of 100,000 with average waste generation (4.5 Ibs/day per
capita), recycling (30 percent), and baseline disposal in a landfill with no gas collection
system could increase its recycling rate to 40 percent-^or example, by implementing a
pay-as-you-throw program—and reduce emissions by about 10,000 MTCE per year.
(Note that further growth in recycling would be possible; some communities already are
exceeding recycling rates of 50 percent).
• A city of 1 million, disposing of 650,000 tons per year in a landfill without gas collection,
could reduce its GHG emissions by nearly 138,000 MTCE per year by managing waste in
a mass burn combustor unit.
19 ,
Sustainable biogenic sources include paper and wood products from sustainably managed
forests. When these materials are burned or aerobically decomposed to CO2, the CO2 emissions are not
counted. Our approach to measuring GHG emissions from biogenic sources is described in detail in
Chapter 1.
20
For details on the sensitivity analyses, see section 7.5 and Exhibits 7-7 and 7-8.
ES-17
-------
• A town of 50,000 landfilling 30,000 tons per year could install a landfill gas recovery
system and reduce emissions by about 6,000 MTCE per year.
• At the national level, if the United States attains the goal of a 35 percent recycling rate by
2005, emissions will be reduced by nearly 10 million MTCE per year compared to a
baseline where we maintain the current 30 percent recycling rate and use the "national
average" landfill for disposal.
ES.7 OTHER LIFE-CYCLE GHG ANALYSES AND TOOLS
Life-cycle analysis is being used increasingly to quantify the GHG impacts of private and
public sector decisions. In addition to the life-cycle analyses that underpin the emission factors in
this report, Environmental Defense,21 ICLEI, Ecobilan, and others have analyzed the life-cycle
environmental impacts of various industry processes (e.g., manufacturing) and private and public
sector practices (e.g., waste management). In many cases, the results of life-cycle analyses are
packaged into software tools that distill the information according to a specific user's needs.
As mentioned earlier, the WARM model was designed as a tool for waste managers to
weigh the GHG impacts of their waste management practices. As a result, the model focuses
exclusively on waste sector GHG emissions, and the methodology used to estimate emissions is
consistent with international and domestic GHG accounting guidelines. Life-cycle tools designed
for broader audiences necessarily include other sectors and/or other environmental impacts, and
are not necessarily tied to the Intergovernmental Panel on Climate Change (IPCC) guidelines for
GHG accounting or the methods used in the Inventory of U.S. Greenhouse Gas Emissions and
Sinks.
• WARM, developed by ICF Consulting for EPA, allows users to input several key
variables (e.g., landfill gas collection system information, electric utility fuel mix,
transportation distances).22 The model covers 21 types of materials and 5 waste
management options: source reduction, recycling, combustion, composting, and
landfilling. WARM accounts for upstream energy and non-energy emissions,
transportation distances to disposal and recycling facilities, carbon sequestration, and
utility offsets that result from landfill gas collection and combustion. The tool provides
participants in the U.S. Department of Energy's 1605b program with the option to report
results by year, by gas, and by year and gas. WARM software is available free of charge
in both a Web-based calculator format and a Microsoft Excel® spreadsheet. The tool is
ideal for waste planners interested in tracking and reporting voluntary GHG emission
reductions from waste management practices and for comparing the climate change
impacts of different approaches. To access the tool, visit:
.
• ICLEI Cities for Climate Protection (CCP) Campaign Greenhouse Gas Emission
Software was developed by Torrie Smith Associates for ICLEI. This Windows-based
tool, targeted for use by local governments, can analyze emissions and emission
reductions on a community-wide basis and for municipal operations alone. The
21 Blum, L., Denison, R.A., and Ruston, V.F. 1997. A Life-Cycle Approach to Purchasing and
Using Environmentally Preferable Paper: A Summary of the Paper Task Force Report," Journal of
Industrial Ecology. 1:3:15-46. Denison, R.A. 1996. "Environmental Life-Cycle Comparison of Recycling,
Landfilling, and Incineration: A Review of Recent Studies;" Annual Review of Energy and the
Environment 21:6:191-237.
22 Microsoft Excel and Web-based versions of this tool are available online at the following Web
site: http://www.epa.gov/globalwarming/actions/waste/tools.html.
ES-18
-------
community-wide, module looks at residential, commercial, and industrial buildings,
transportation activity, and community-generated waste. The municipal operations
module looks at municipal buildings, municipal fleets, and waste from municipal in-
house operations. In addition to computing GHG emissions, the CCP software estimates
reductions in criteria air pollutants, changes in energy consumption, and financial costs
' and savings associated with energy use and other emission reduction initiatives. A
version of the software program was made available for use by private businesses and
institutions during the summer of 2001. CCP software subscriptions, including technical
support, are available to governments participating in ICLEI for a subsidized price of
$240. The full retail price of the software in the United States is $2,000. For more
information, visit: or contact the U.S. ICLEI
officeat(510)-540-8843, iclei_usa@iclei.org. . ,
•' The MSW Decision Support Tool (DST) and life-cycle inventory database for North
America have been developed through funding by ORD through a cooperative agreement
with the Research Triangle Institute (CR823052). The methodology is based on a multi-
media, multi-pollutant approach and includes analysis of GHG emissions as well as a
broader set of emissions (air, water, and waste) associated with MSW operations. The
MSW-DST is available for site-specific applications and has been used to conduct
analyses in several states and 15 communities including use by the U.S. Navy in the
Pacific Northwest. The tool is intended for use by solid waste planners at state and local
levels to analyze and compare alternative MSW management strategies with respect to
cost, energy consumption, and environmental releases to the air, land, and water. The '-
costs are based on full cost accounting principles and account for capital and operating
costs using an engineering economics analysis. The MSW-DST calculates not only
projected emissions of GHGs and criteria air pollutants, but also emissions of more than
30 air- and water-borne pollutants. The DST models emissions associated with all MSW
management activities, including waste collection and transportation, transfer stations,
materials recovery facilities, compost facilities, landfills, combustion and refuse-derived
fuel facilities, utility offsets, material offsets, and source reduction. The differences in
residential, multi-family, and commercial sectors can be evaluated individually. The
software has optimization capabilities that enable one to identify options that evaluate
minimum costs as well as solutions that can maximize environmental benefits, including
energy conservation and GHG reductions.
At the time of the publication of this report, the life-cycle inventory (LCI) database for
North America was to be released in 2002. Plans to develop a Web-based version are
being considered. The MSW-DST provides extensive default data for the full range of
MSW process models and requires minimum input data. However, these defaults can be
tailored to the specific communities using site-specific information. The MSW-DST also
includes a calculator for source reduction and carbon sequestration using a methodology
that is consistent with the IPCC in terms of the treatment of biogenic CO2 emissions. For
more information, refer to the project Web site:
or contact Susan Thornloe, U.S. EPA,
(919)-541-2709, thornloe.susan@epamail.epa.gov, or Keith Weitz, Research Triangle
Institute, (919)-541-6973, kaw@rti.org.
• The Tool for Environmental Analysis and Management (TEAM), developed by Ecobilan,
simulates operations associated with product design, processes, and activities associated
with several industrial sectors. The model considers energy consumption, material
consumption, transportation, waste management, and other factors in its evaluation of
environmental impacts. Many private firms and some government agencies have used the
model. Users pay a licensing fee of $3,000 and an annual maintenance contract of $3,000.
ES-19
-------
This model is intended for use in Europe and was not developed for use in North
America. For more information, visit:
.
ES.8 LIMITATIONS OF THE ANALYSIS
When conducting this analysis, we used a number of analytical approaches and numerous
data sources, each with its own limitations. In addition, we made and applied assumptions
throughout the analysis. Although these limitations would be troublesome if used in the context
of a regulatory framework, we believe that the results are sufficiently accurate to support their use
in voluntary programs. Some of the major limitations include the following:
• The manufacturing GHG analysis is based on estimated industry averages for energy
usage, and in some cases the estimates are based on limited data. In addition, we used
values for the average GHG emissions per ton of material produced, not the marginal
emission rates per incremental ton produced. In some cases, the marginal emission rates
may be significantly different.
• The forest carbon sequestration analysis deals with a very complicated set of interrelated
ecological and economic processes. Although the models used represent the state-of-the-
art in forest resource planning, their geographic scope is limited. Because of the global
market for forest products, the actual effects of paper recycling would occur not only in
the United States but in Canada and other countries. Other important limitations include:
(1) the estimate does not include changes in carbon storage in forest soils and forest
floors; (2) the model assumes that no forested lands will be converted to non-forest uses
as a result of increased paper recycling; and (3) we use a point estimate for forest carbon
sequestration, whereas the system of models predicts changing net sequestration over
time.
• The composting analysis considers a small sampling of feedstocks and a single compost
application (i.e., agricultural soil). The analysis did not consider the full range of soil
conservation and management practices that could be used in combination with compost
and their impacts on carbon storage.
• The combustion analysis uses national average values for several parameters; variability
from site to site is not reflected in our estimate.
• The landfill analysis (1) incorporates considerable uncertainty on CELj generation and
carbon sequestration, due to Limited data availability; and (2) uses landfill estimated CHU
recovery levels for the year 2000 as a baseline.
Finally, throughout most of the report, we express analytical inputs and outputs as point
estimates. We recognize that a rigorous treatment of uncertainty and variability would be useful,
but in most cases the information needed to treat these in statistical terms is not available. The
report includes some sensitivity analyses to illustrate the importance of selected parameters and
expresses ranges for a few other factors such as GHG emissions from manufacturing. We
encourage readers to provide more accurate information where it is available; perhaps with
additional information, future versions of this report will be able to shed more light on uncertainty
and variability. Meanwhile, we caution that the emission factors reported here should be
evaluated and applied with an appreciation for the limitations in the data and methods, as
described at the end of each chapter.
ES-20
-------
PI. METHODOLOGY
This report is the second edition of Greenhouse Gas Emissions from Management of Selected
Materials in Municipal Solid Waste. We made the following improvements to the first edition of the
report:
• Incorporated new energy data and information on recycling loss rates from EPA's Office of
Research and Development (ORD);
• Expanded analysis of the GHG benefits of composting, including results of CENTURY
model runs;
• Developed emission factors for five new material types: magazines/third-class mail,
phonebooks, textbooks, dimensional lumber, and medium-density fiberboard;
• Developed emission factors for two new categories of mixed materials: mixed plastics and
mixed organics;
• Updated national recovery and generation rates to include 2000 data;
• Incorporated new energy data into calculations of utility offsets;
• Revised carbon coefficients and fuel use for national average electricity generation;
• Updated information on landfill gas recovery rates;
• Added a discussion of emerging issues in the area of climate change and waste management;
and
• Provided a list of suggested proxy values for voluntary reporting of GHG emission
reductions.
All of these changes and/or revisions are described in more detail throughout the body of the report.
Because this is the second edition, we have moved some of the background information from the
body of the report to background documents, which are available in the docket in the Resource
Conservation and Recovery Act (RCRA) Information Center. Background Document A provides data on
life-cycle energy intensity and fuel mix, provided by Franklin Associates, Ltd. (PAL) (All other
background documents, and this report, were written by ICF Consulting.) Background Document B
provides a discussion of the review cycles leading up to the first and second editions of the report.
Background Document C includes a discussion of how we screened materials for the first edition of the
report.
The remainder of this chapter provides an overview of the methodology used to calculate the
GHG emissions associated with various management strategies for MSW. The first section briefly
describes the life-cycle framework used for the analysis. Next is a discussion of the materials included in
the analysis. The final three sections present a description of key inputs and baselines, a summary of the
life-cycle stages, and an explanation of how to estimate and compare net GHG emissions and sinks.
-------
1.1 THE OVERALL FRAMEWORK: A STREAMLINED LIFE-CYCLE INVENTORY
Early in this analysis of the GHG benefits of specific waste management practices, it became
clear that all waste management options provide opportunities for reducing GHG emissions, depending
on individual circumstances. Although source reduction and recycling are often the most advantageous
waste management practices from a GHG perspective, a material-specific comparison of all available
waste management options would clarify where the greatest GHG benefits can be obtained for particular
materials in MSW. A material-specific comparison can help policymakers identify the best options for
GHG reductions.
This study determined that the best way to conduct such a comparative analysis is a streamlined
application of a life-cycle assessment (LCA). A full LCA is an analytical framework for understanding
the material inputs, energy inputs, and environmental releases associated with manufacturing, using, and
disposing of a given material. A full LCA generally consists of four parts: (1) goal definition and
scoping; (2) an inventory of the materials and energy used during all stages in the life of a product or
process, and an inventory of environmental releases throughout the product life cycle; (3) an impact
assessment that examines potential and actual human health effects related to the use of resources and
environmental releases; and (4) an assessment of the change that is needed to bring about environmental
improvements in the product or processes.
A full LCA is beyond the scope of this analysis. Rather, the streamlined LCA described in this
report is limited to an inventory of the emissions and other environmental impacts related to global
wanning. This study did not assess human health impacts, necessary environmental improvements, and
air, water, or environmental impacts that do not have a direct bearing on climate change.
1.2 MSW MATERIALS CONSIDERED IN THE STREAMLINED LIFE-CYCLE
INVENTORY
Each material in MSW has different GHG impacts depending on how it is manufactured and
disposed of. We began our research by performing a screening analysis of 37 of the most common
materials and products found in MSW.1 The materials included in screening analysis then were ranked by
their potential for GHG reductions.2 The first edition of the report included 12 materials: aluminum cans,
steel cans,3 glass, high-density polyethylene (HOPE) plastic blow-molded containers, low-density
polyethylene (LDPE) plastic blow-molded containers, polyethylene terephthalate (PET) plastic blow-
molded containers, corrugated cardboard, newspaper, office paper,4 and three grades of mixed paper
(broad, residential, and office). In addition to these materials, we examined the GHG implications of
various management strategies for food discards, yard trimmings, mixed MSW, and mixed recyclables.
1 In addition to the materials and products covered in the first edition of the report, the screening analysis
included the following materials and products: other paper materials (bags and sacks, other paper packaging, books,
other paperboard packaging, wrapping papers, paper plates and cups, folding cartons, other nonpackaging paper, and
tissue paper and towels), other plastic materials (plastic wraps, plastic bags and sacks, other plastic containers, and
other plastic packing), other metal materials (aluminum foil/closures, other steel packaging), and other miscellaneous
materials (miscellaneous durable goods, wood packaging, furniture and furnishings, carpet and rags, and other
miscellaneous packaging).
2 For more information on the screening analysis used to identify materials for the first edition of the report,
see Background Document C.
3 Other steel materials also may be recycled, but this analysis was limited to steel cans from households.
4 Office paper refers to the type of paper used in computer printers and photocopiers.
-------
Comparison of EPA/ORD and EPA/OSW Emission Factors
Efforts to harmonize our previous life-cycle emission factors with the results of recent work by EPA's
Office of Research and Development (ORD) began in October 2000. Noticing significant differences in our
bottom line emission factors, we compared a range of assumptions, including energy consumption, fuel mix, loss
rates, landfill oxidation rate, timing of landfill methane emissions, fraction of landfill gas collected, electricity
mix, transportation distances, and carbon storage. Our comparison of energy intensities and fuel mixes included
process and transportation energy for virgin and recycled production of each material type. Because the previous
Office of Solid Waste (OSW) energy values were based on an average of Franklin Associates, Ltd. (FAL) and
Tellus data, we compared the ORD values to the FAL data, Tellus data, and average of FAL and TeUus data.
This comparison revealed that the differences between the OSW and ORD emission factors are mostly
attributable to the different assumptions about energy consumption (i.e., the sum of precombustion, process, and
transportation energy), fuel mix, and loss rates. In general, we found that ORD's total energy values are lower
than OSW's energy values for both virgin and recycled materials. Comparing fuel mix, we found the most
significant differences occurring for electricity, coal, natural gas, and "other" fuel types comprising process
energy. The fractions of diesel fuel, residual fuel, and natural gas exhibited the greatest disparities for
transportation energy. Our comparison of loss rates, which are used to develop the recycling emission factors,
showed significant variation for office paper, steel cans, and, to a lesser extent, newspaper.
In an effort to reconcile the remaining differences between ORD and OSW estimates of GHG emissions
from the acquisition of raw materials and then: manufacture into products, we identified additional methodological
differences that could be affecting the recycling numbers. In particular, we found that ORD simulates closed-loop
recycling for all materials, while OSW assumes open-loop recycling for office paper and corrugated cardboard.
We also found mat ORD's estimates do not include non-energy process emissions from perfluorocarbons (PFCs).
To isolate any remaining differences between the two analyses, we substituted ORD energy intensities, fuel
mixes, and loss rates into the OSW model.
Once we had identified and resolved all methodological differences between ORD and OSW estimates
for raw materials acquisition and manufacturing, we selected the material types for which we could substitute
ORD data for the existing OSW data: aluminum, glass, HOPE, LDPE, PET, corrugated cardboard,
magazines/third-class mail, newspaper, office paper, phonebooks, and textbooks. For wood products, ORD did
not develop emission factors, while for steel its data was not sufficiently disaggregated to replace the existing
OSW data.
Most of the changes from the first edition of this report reflect additions of new or updated data.
This second edition features an expanded list of material types, including magazines and third-class mail,
phonebooks, textbooks, dimensional lumber, medium-density fiberboard, and several additional
categories of mixed recycled materials (e.g. mixed plastics, mixed organics). This edition also
incorporates updated data developed by ORD through its work on life-cycle management of MSW.
ORD's data set on energy, fuel mix, and loss rates has been thoroughly reviewed by industry and other
stakeholders, and is likely to be more up-to-date than some of the information in the first edition of this
report. Thus, where a complete set of energy intensity and fuel mix data was available from ORD, that
information was incorporated in this report. For other materials—steel cans and mixed paper (broad,
residential, and office definitions)^we retained the original data set developed by FAL. This edition
includes new data (also developed by FAL) on dimensional lumber and medium-density fiberboard.
Exhibit 1-1 lists the materials that were analyzed for this report and the energy-related data sources
underlying the estimates. All of the material types listed in Exhibit 1-1 are discussed in subsequent
chapters and included in exhibits throughout the report, with the exception of three mixed waste
categories. Mixed plastics, mixed recyclables, and mixed organics are included only in Chapter 8 because
emission factors for these materials simply reflect the weighted average emissions of other material
types.
-------
Exhibit 1-1 Materials Analyzed and Energy-related Data Sources
Material
Aluminum Cans
Steel Cans
Glass
Corrugated Cardboard
Magazines/Third-class
Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Energy Data Source
ORD
FAL
ORD
ORD
ORD
ORD
ORD
ORD
ORD
FAL
Material
Medium-Density
Fiberboard
Food Discards
Yard Trimmings
Mixed Paper - Broad
Definition
Mixed Paper -
Residential Definition
Mixed Paper - Office
Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW
Energy Data Source
FAL
NA
NA
FAL
FAL
FAL
Weighted Average
Weighted Average
NA
NA
NA = Not applicable (data not energy-related)
1.3 KEY INPUTS AND BASELINES FOR THE STREAMLINED LIFE-CYCLE
INVENTORY
Evaluating the GHG emissions of waste
management requires analysis of three factors:
(1) GHG emissions throughout the life cycle of
the material (including the chosen disposal
option); (2) the extent to which carbon sinks are
affected by manufacturing and disposing of the
material; and (3) the extent to which the
management option recovers energy that can be
used to replace electric utility energy, thus
reducing utility GHG emissions. In addition, to
provide a consistent basis for comparison, we
made several choices in our GHG accounting
framework in terms of timing and levels of
production. Each of these factors warrants
further discussion.
GHG Emissions Relevant to Waste:
The most important GHGs for purposes of
analyzing MSW management options are CO2,
CHt, N2O, and perfluorocarbons (PFCs). Of
these, CO2is by far the most common GHG
emitted in the United States. Most CO2
emissions result from energy use, particularly fossil fuel combustion. A great deal of energy is consumed
when a product is manufactured and then discarded. This energy is used in the following stages: (1)
extracting and processing raw materials; (2) manufacturing products; (3) managing products at the end of
their useful lives; and (4) transporting materials and products between each stage of their life cycles. This
study estimated energy-related GHG emissions during all of these stages, except for transportation of
products to consumers (because GHG emissions resulting from transportation to consumers will vary
little among the options considered). Much of this report is devoted to explaining the methodology
Comparing GHGs
Carbon dioxide (CO2), methane (CH4), and nitrous
oxide (N2O) are very different gases in terms of their heat-
trapping potential. An international protocol has established
CO2 as the reference gas for measurement of heat-trapping
potential (also known as global warming potential or GWP).
By definition, the GWP of Ikilogram (kg) of CO2 is 1.
CHU has a GWP of 21, which means that 1 kg of
methane has the same heat-trapping potential as 21 kg of CO2.
N20 has a GWP of 310.
PFCs are the most potent GHG included in this
analysis; GWPs are 6,500 for CF4 and 9,200 for C2F6.
In this report, emissions of CO2, CELt, N2O, and PFCs
have been converted to their "carbon equivalents." Because
CO2 is 12/44 carbon by weight, 1 metric ton of CO2 is equal to
12/44 or 0.27 metric tons of carbon equivalent (MTCE). The
MTCE value for 1 metric ton of each of the other gases is
determined by multiplying its GWP by a factor of 12/44. (All
data provided here are from The Intergovernmental Panel on
Climate Change (IPCC), Climate Change 1995: The Science of
Climate Chanee. 1996. o. 121.1
-------
employed for quantifying the energy used— and the resulting CO2 emissions— at each stage in the life
cycle of any given material in MSW. Energy consumed in connection with consumer use of products is
not evaluated, because it is assumed that energy use for the selected materials is small (or zero) at this
point in the life cycle. In addition, the energy consumed during use would be about the same whether the
product is made from virgin or recycled inputs.
a more potent GHG, is produced when organic waste decomposes in an oxygen-free
(anaerobic) environment, such as a landfill. CBU from landfills is the largest source of CEU in the United
States;5 these emissions are addressed in Chapter 7. CH4 is also emitted when natural gas is released to
the atmosphere during production of coal or oil, production or use of natural gas, and agricultural
activities.
N2O results from the use of commercial and organic fertilizers and fossil fuel combustion, as well
as other sources. This analysis estimated N2O emissions from waste combustion.
Perfluorocarbons (CF4 and C2F6) are emitted during the reduction of alumina to aluminum in the
primary smelting process. The source of fluorine for CF4 and C2F6 is the molten cryolite (Na3AlF6) where
the reduction of alumina occurs. Perfluorocarbons are formed when the fluorine in cryolite reacts with
the carbon in the anode (a carbon mass of paste, coke briquettes, or prebaked carbon blocks) and in the
carbon lining that serves as the cathode. Although the quantities of perfluorocarbons emitted are small,
these gases are significant because of their high global warming potential.
Carbon Stocks. Carbon Storage, and Carbon Sequestration: This analysis includes carbon storage
to the extent that it is due to waste management practices. For example, landfilled organic materials
result in landfill carbon storage, as carbon is moved from a product pool (e.g., furniture) to the landfill
pool. The same is true for composted organics that lead to carbon storage in soil.
Carbon sequestration differs from carbon storage because it represents a transfer of carbon from
the atmosphere to a carbon pool. For example, trees in a forest undergo photosynthesis, converting CO2
in the atmosphere to carbon in biomass. In this analysis, we consider the impact of waste management on
forest carbon sequestration.
The baseline against which changes in carbon stocks are measured is a projection by the U.S.
Forest Service of forest growth, mortality, harvests, and other removals under anticipated market
conditions for forest products. One of the assumptions for the projections is that U.S. forests will be
harvested on a sustainable basis (i.e., trees will be grown at a rate at least equal to the rate at which they
are cut).6 Thus, the baseline assumes that harvesting trees at current levels results in no diminution of the
forest carbon stock and no additional CO2 in the atmosphere. On the other hand, forest carbon
sequestration increases as a result of source reduction or recycling of paper products because both source
reduction and recycling cause annual tree harvests to drop below otherwise anticipated levels (resulting
in additional accumulation of carbon in forests). Consequently, source reduction and recycling "get
credit" for increasing the forest carbon stock, whereas other waste management options (combustion and
landfilling) do not.
5 U.S. EPA. 2001.Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-1999. U.S. Environmental
Protection Agency, Office of Policy, Planning and Evaluation, Washington, DC. EPA-236-R-01-001.
6 Assuming a sustainable harvest in the United States is reasonable because from 1952 to 1997 U.S. forest
carbon stocks steadily increased. In the early part of this period, the increases were mostly due to reversion of
agricultural land to forest land. More recently, improved forest management practices and the regeneration of
previously cleared forest areas have resulted in a net annual uptake (sequestration) of carbon. The steady increase in
forest carbon stocks implies sustainable harvests, and it is reasonable to assume that the trend of sustainable harvests
will continue.
-------
Although source reduction and recycling are associated with forest carbon sequestration,
composting—in particular, application of compost to degraded soils—enhances soil carbon storage. Four
mechanisms of increased carbon storage are hypothesized in Chapter 5; a modeling approach is used to
estimate the magnitude of carbon storage associated with three of these mechanisms.
Finally, landfills are another means by which carbon is removed from the atmosphere. Landfill
carbon stocks increase over time because much of the organic matter placed in landfills does not
decompose, especially if the landfill is located in an arid area. However, not all carbon in landfills is
counted in determining the extent to which landfills are carbon stocks. For example, the analysis does not
count plastic in landfills toward carbon storage. Plastic in a landfill represents simply a transfer from one
carbon stock (the oil field containing the petroleum or natural gas from which the plastic was made) to
another carbon stock (the landfill); thus, no change has occurred in the overall amount of carbon stored.
On the other hand, the portion of organic matter (such as yard trimmings) that does not decompose in a
landfill represents an addition to a carbon stock, because it would have largely decomposed into CO2 if
left to deteriorate on the ground.
Although changes in fossil fuel carbon stocks (i.e., reductions in oil field stores that result from
the extraction and burning of oil resources) are not measured directly in this analysis, the reduction in
fossil fuel carbon stocks is indirectly captured by counting the CO2 emissions from fossil fuel combustion
in calculating GHG emissions.
Avoided Electric Utility GHG Emissions Related to Waste: Waste that is used to generate
electricity (either through waste combustion or recovery of CH4 from landfills) displaces fossil fuels that
utilities would otherwise use to produce electricity. Fossil fuel combustion is the single largest source of
GHG emissions in the United States. When waste is substituted for fossil fuel to generate electricity, the
GHG emissions from burning the waste are offset by the avoided electric utility GHG emissions.
Baseline Year: The baseline year selected for most parts of the analysis is the most recent year
for which data are available. However, for the system efficiency and ferrous recovery rate at waste
combustors, this study uses values previously projected for the year 2000. For paper recycling, annual
projections through 2010 were used to develop an average forest carbon storage value for the period from
1996 through 2010.7 The compost analysis relied on model simulations of compost application,
beginning in 1996 and ending in 2005. The carbon storage estimates resulting from these model runs
correspond to model outputs in 2010 in order to maintain consistency with forest carbon storage
estimates. We developed "future"8 scenarios for paper recycling, composting, and carbon storage
analyses because some of the underlying factors that affect GHG emissions are changing rapidly, and this
study seeks to define relationships (e.g., between tonnage of waste landfilled and CELt emissions) that.
represent an average over the next several years.
• Although the existing U.S. municipal waste combustors include a few small facilities that do
not recover energy, the study assumes that those facilities will be closed in the near future.
Thus, the report assumes that all combustors recover energy. The study used an estimate
provided by the combustion industry for anticipated levels of ferrous recovery.
• For paper recycling, earlier analyses indicated that the marginal impact of increased paper
recycling on forest carbon sequestration changes over time. The impact also differs
7 The models we used simulated carbon sequestration through 2040, but we selected a value based on
average conditions through 2010.
8 In the case of system efficiency and ferrous recovery at waste combustors, the year 2000 represented a
future value when the first edition of this report was published. This edition of the report does not reflect these
updated values because more recent data are not available.
-------
depending on the initial paper recycling rate and how that rate changes over time. To
estimate the impact of increased paper recycling on forest carbon sequestration, the study
needed to account for these influences. First, we used the American Forest and Paper
Association's baseline projection that paper recycling rates will continue to increase from
about 35 percent in 1994 to 50 percent by 2000.9 The trajectory for a baseline scenario for
paper recycling passes through 50 percent in 2000, with continued modest increases in the
following years. Because of the need to estimate the effect of efforts (e.g., by EPA) to
enhance recycling beyond the baseline projected rate, we developed a plausible scenario for
enhanced paper recycling rates and then compared the projected forest carbon sequestration
under the baseline and increased recycling scenarios.10 (This approach is fully described in
Chapter 3.)
• The baseline for our landfill recovery scenario is based on estimated recovery rates and
percentages of waste disposed in landfills with no recovery, landfills with flaring, and
landfills with landfill-gas-to-energy projects for the year 2000. According to our estimates,
49 percent of all landfill CIL. was generated at landfills with recovery systems, and the
remaining 51 percent was generated at landfills without landfill gas (LFG) recovery.11 Of the
49 percent of all OHU generated at landfills with LFG recovery, 49 percent (or 24 percent of
all CHO was generated at landfills that use LFG to generate electricity, and 51 percent (or 25
percent of all CKt) at landfills that flare LFG.I2-13
1.4 SUMMARY OF THE LIFE-CYCLE STAGES
Exhibit 1-2 shows the GHG sources and carbon sinks associated with the manufacture of various
materials and the post-consumer management of these materials as wastes. As shown in the exhibit,
GHGs are emitted from (1) the pre-consumer stages of raw materials acquisition and manufacturing, and
(2) the post-consumer stage of waste management. No GHG emissions are attributed to the consumer's
use of any product.
9 Actual paper recovery in 2000 (taken from EPA's Municipal Solid Waste in the United States: 2000 Facts
and Figures) averaged about 53%, confirming that 50 percent is a reasonable estimate for 2000.
Note that this estimate is necessary for analyzing the scenarios; however, it does not represent a plan of
action by EPA.
11 Based on data on (1) year 2000 MSW landfill methane generation of 72.7 million MTCE (from draft U.S.
Climate Action Report - 2001), (2) year 2000 landfill methane recovery of 26.7 million MTCE (also from draft U.S.
Climate Action Report - 2001), and (3) estimated landfill methane recovery efficiency of 75 percent (from U.S.
Methane Emissions 1990-2020: Inventories, Projections, and Opportunities for Reductions).
12 Draft U.S. Climate Action Report - 2001.
The assumption that 49 percent of landfills recovering methane will use it to generate electricity is subject
to change over time based upon changes in the cost of recovery, and the potential payback. Additionally, new
technologies may arise that use recovered methane for purposes other than generating electricity.
-------
Exhibit 1-2 Greenhouse Gas Sources and Sinks Associated with the Material Life Cycle
Sinks & Emission Offsets
GHG Emissions
Life Cycle Stage
i
Raw Materials
Acquisition I
Virgin inputs
Reduced Carbon
Sequestration
in Forests
Energy and
Non-Ertergy-
Related Emissions
Materials Extracted:
Trees, Ore, Oil, etc.
Manufacturing
Energy and Non-
Energy-Related Emissions
Increased
Avoided Fossil, s- forest Carbon
^ Fuej Use > Sequestration*
Waste, , -
JHW Management
, ^^N ^ « \
Carbon
I, Storage in s
theBoii - '
Energy-Related
Emissions
Avoided Fossil,
Fuel Use""
Carbon in Long-Temt
Storage in Landfill
- Combustion ,
Uncontrolled CH4
Emissions or CH4 Flared
J, and Recovered Energy
-------
The remainder of this chapter describes how this study analyzed each of the upstream (raw
materials acquisition, manufacturing, and forest carbon sequestration) and downstream (source reduction,
recycling, composting, combustion, and landfilling) stages in the life cycle. The following sections
explain stages of the life cycle and the corresponding emission factor components presented in Exhibit 1-
3, and outline the GHG emissions and carbon sinks at each stage of the product life cycle. These GHG
emissions and carbon sinks are described in detail and quantified for each material in Chapters 2 through
Exhibit 1-3 Components of Net Emissions for Various MSW Management Strategies
Management
Strategy
Source Reduction
Recycling
Composting
Combustion
Landfilling
GHG Sources and Sinks
Process and Transportation
GHGs from Raw Materials
Acquisition and Manufacturing
Decrease in GHG emissions,
relative to the baseline of
manufacturing
Decrease in GHG emissions due to
lower energy requirements
(compared to manufacture from
virgin inputs) and avoided process
non-energy GHGs
No emissions/sinks
Baseline process and transportation
emissions due to manufacture from
the current mix of virgin and
recycled inputs
Baseline process and transportation
emissions due to manufacture from
the current mix of virgin and
recycled inputs
Forest Carbon
Sequestration or Soil
Carbon Storage
Increase in forest carbon
sequestration
Increase in forest carbon
sequestration
Increase in soil carbon
storage
No change
No change
Waste Management GHGs
No emissions/sinks
Process and transportation
emissions are counted in the
manufacturing stage
Compost machinery
emissions and transportation
emissions
Nonbiogenic CO2, N2O
emissions, avoided utility
emissions, and
transportation emissions
CHf emissions, long-term
carbon storage, avoided
utility emissions, and
* No manufacturing transportation GHG emissions are considered for composting of food discards and yard
trimmings because these materials are not considered to be manufactured.
1.4.1 GHG Emissions and Carbon Sinks Associated with Raw Materials Acquisition and
Manufacturing
The,top left of Exhibit 1-2 shows inputs for raw materials acquisition. These virgin inputs are
used to make various materials, including ore for manufacturing metal products, trees for making paper
products, and petroleum or natural gas for producing plastic products. Fuel energy also is used to obtain
or extract these material inputs.
The inputs used in manufacturing are (1) energy and (2) either virgin raw materials or recycled
materials. In the exhibit these inputs are identified with arrows that point to the icon labeled
"Manufacturing."
-------
For source reduction, the "baseline" GHG emissions from raw materials acquisition and
manufacturing are avoided. This analysis thus estimates, for source reduction, the GHG reductions
(relative to a baseline of initial manufacture) at the raw materials acquisition and manufacturing stages.
Source reduction is assumed to entail more efficient use of a given material. Examples are lightweighting
(reducing the quantity of raw material in a product), double-sided photocopying, and extension of a
product's useful life. No other material substitutions are assumed for source reduction; therefore, this
report does not analyze any corresponding increases in production and disposal of other materials (which
could result in GHG emissions).14
The GHG emissions associated with raw materials acquisition and manufacturing are (1) GHG
emissions from energy used during the acquisition and manufacturing processes, (2) GHG emissions
from energy used to transport materials,15 and (3) non-energy GHG emissions resulting from
manufacturing processes (for aluminum, steel, plastics, office paper, and medium-density fiberboard).
Each type of emission is described below. Changes in carbon sequestration in forests also are associated
with raw materials acquisition for paper products.
Process Energy GHG Emissions: Process energy GHG emissions consist primarily of CO2
emissions from the combustion of fuels used in raw materials acquisition and manufacturing. CO2
emissions from combustion of biomass are not counted as GHG emissions. (See "CO2 Emissions from
Biogenic Sources" on page 12.)
The majority of process energy CO2 emissions are from combustion of fuels used directly, e.g., to
operate ore mining equipment or to fuel a blast furnace. Fuel also is needed to extract the oil or mine the
coal that is ultimately used to produce energy and transport these fuels to the place where they are used.
Thus CO2 emissions from this "pre-combustion energy" are counted in this category as well. When
electricity generated by combustion of fossil fuels is used in manufacturing, the CO2 emissions from the
fossil fuels also are counted.
To estimate process energy GHG emissions, the study first obtained estimates of both the total
amount of process energy used per ton of product (measured in British thermal units or Bra's), and the
fuel mix (e.g., diesel oil, natural gas, fuel oil, etc.). Next, emissions factors for each type of fuel were
used to convert the amount of each type of fuel used to GHG emissions. As noted earlier, making a
material from recycled inputs generally requires less process energy (and uses a different fuel mix) than
making the material from virgin inputs.
Details of the methodology for estimating process energy GHG emissions are provided in
Chapter 2.
Transportation Energy GHG Emissions: Transportation energy GHG emissions consist of CO2
emissions from the combustion of fuels used to transport raw materials and intermediate products to the
final manufacturing or fabrication facility. The estimates of transportation energy emissions are based on:
(1) the amounts of raw material inputs and intermediate products used in manufacturing 1 ton of each
material; (2) the average distance that each raw material input or intermediate product is transported; and
(3) the transportation modes and fuels used. For the amounts of fuel used, the study used data on the
average fuel consumption per ton-mile for each mode of transportation.16 Then an emission factor for
14 Although material substitution is not quantitatively addressed in the report, it is discussed from a
methodological standpoint in Chapter 2 and also is discussed briefly in Chapter 4, Section 4.3.
15 For some materials (plastics, magazines/third-class mail, office paper, phonebooks, and textbooks), the
transportation data we received were included in the process energy data. For these materials, we report total GHG
emissions associated with process and transportation in the "process energy" estimate.
Ifi These data are found in Background Document A.
10
-------
each type of fuel was used to convert the amount of each type of fuel consumed to the GHG emissions
produced. , ,
More detail on the methodology to estimate transportation energy GHG emissions is provided in
Chapter 2.
r -ft: . .
Process Non-Energy GHG Emissions: Some GHG emissions occur directly in the manufacture of
certain materials and are not associated with energy consumption. In this analysis, these emissions are
referred to as process non-energy emissions. For example, the production of steel or aluminum requires
lime (calcium oxide, or CaO), which is produced from limestone (calcium carbonate, or CaCO3), and the
manufacture of lime results in CO2 emissions. Other process non-energy GHG emissions are associated
with the manufacture of plastics, office paper, and medium-density fiberboard. In some cases, process
non-energy GHG emissions are associated only with production using virgin inputs; in other cases, these
emissions result when either virgin or recycled inputs are used. These emissions are described in Chapter
2. • ' ' •
Carbon Sinks: The only carbon sink during the stages of raw materials acquisition and
manufacturing is the additional carbon sequestration in trees associated with source reduction or
recycling of paper products. The methodology for estimating forest carbon sequestration is described in
Chapters.
1.4.2 GHG Emissions and Carbon Sinks Associated with Waste Management
As shown in Exhibit 1-3, there are up to five post-consumer waste management options,
depending on the material: recycling, composting, combustion, and landfilling. This section describes the
GHG emissions and carbon sinks associated with these five options.
Source Reduction: In this analysis, source reduction is measured by the amount of material that
would otherwise be produced but is not generated due to a program promoting source reduction. Thus,
with source reduction there are no emissions from MSW management.
Recycling: When a material is recycled, it is used in place of virgin inputs in the manufacturing
process. The avoided GHG emissions from remanufacture using recycled inputs is calculated as the
difference between (1) the GHG emissions from manufacturing a material from 100 percent recycled
inputs, and (2) the GHG emissions from manufacturing an equivalent amount of the material (accounting
for loss rates) from 100 percent virgin inputs (including the process of collecting and transporting the
recyclables). No GHG emissions occur at the MSW management stage because the recycled material is
diverted from waste management facilities.17 (If the product made from the recycled material is later
composted, combusted, or landfilled, the GHG emissions at that point would be attributed to the product
that was made from the recycled material.) Chapter 4 details GHG emissions from recycling.
All of the materials considered in this analysis are modeled as being recycled in a "closed loop"
(e.g., newspaper is recycled into new newspaper). However, a variety of paper types are recycled under
the general heading of "mixed paper." Mixed paper can be remanufactured, via an open loop, into
boxboard or paper towels. Other materials are recycled in open-loop processes, but due to limited
resources, this study could not analyze all open-loop processes.18
We do not include GHG emissions from managing residues (e.g., wastewater treatment sludges) from the
manufacturing process for either virgin or recycled inputs.
For example, not all steel cans are recycled into more steel cans; not all aluminum cans are recycled into
more aluminum cans.
11
-------
CO2 Emissions from Biogenic Sources
The United States and all other parties to the U.N. Framework Convention on Climate Change
(UNFCCC) agreed to develop inventories of GHGs for purposes of (1) developing mitigation strategies and (2)
monitoring the progress of those strategies. The Intergovernmental Panel on Climate Change (IPCC) developed a
set of inventory methods to be used as the international standard. (IPCC 1997. IPCC Guidelines for National
Greenhouse Gas Inventories, three volumes.) The methodologies used in this report to evaluate emissions and
sinks of GHGs are consistent with the IPCC guidance.
One of the elements of the IPCC guidance that deserves special mention is the approach used to address
COa emissions from biogenic sources. For many countries, the treatment of CO2 releases from biogenic sources is
most important when addressing releases from energy derived from biomass (e.g., burning wood), but this
element is also Important when evaluating waste management emissions (for example, the decomposition or
combustion of grass clippings or paper). The carbon in paper and grass trimmings was originally removed from
the atmosphere by photosynthesis, and under natural conditions, it would cycle back to the atmosphere eventually
as COa due to degradation processes. The quantity of carbon that these natural processes cycle through the Earth's
atmosphere, waters, soils, and biota is much greater than the quantity added by anthropogenic GHG sources. But
the focus of the UNFCCC is on anthropogenic emissions—those resulting from human activities and subject to
human control. Those emissions have the potential to alter the climate by disrupting the natural balances in
carbon's biogeochemical cycle and altering the atmosphere's heat-trapping ability. For processes with CO2
emissions, if the emissions are from biogenic materials and the materials are grown on a sustainable basis, then
those emissions are considered simply to close the loop in the natural carbon cycle. They return to the atmosphere
COa that was originally removed by photosynthesis. In this case, the CO2 emissions are not counted. (For
purposes of this analysis, biogenic materials are paper, yard trimmings, and food discards.) On the other hand,
COa emissions from burning fossil fuels are counted because these emissions would not enter the cycle were it not
for human activity. Likewise, CH4 emissions from landfills are counted. Even though the source of carbon is
primarily biogenic, CBU would not be emitted were it not for the human activity of landfilling the waste, which
creates anaerobic conditions conducive to CfLt formation. Note that this approach does not distinguish between
the timing of CO2 emissions, provided that they occur in a reasonably short time scale relative to the speed of the
processes that affect global climate change. In other words, as long as the biogenic carbon would eventually be
released as C02, it does not matter whether it is released virtually instantaneously (e.g., from combustion) or over
a period of a few decades (e.g., decomposition on the forest floor).
Composting: When organic materials are composted, the anaerobic decomposition of materials
produces CHj. Similarly, the collection and transportation of organics produces non-biogenic emissions.
During the composting process and after the compost is added to the soil, the decomposition of plants
produces biogenic CO2 emissions. Carbon compounds that do not decompose, however, result in long-
term carbon storage. All of the materials that may be composted (e.g., leaves, brush, grass, food waste,
newspaper) originally are produced by trees or other plants. As described in the above in "CO2 Emissions
from Biogenic Sources," the biogenic CO2 emitted from these materials during composting is not
counted in GHG emissions. However, composting does result in increased soil carbon storage due to
increased production of humic material (natural organic polymers, which degrade at a slow rate) and
several other factors, which are described in Chapter 5.
Although composting may result in some production of QHU (due to anaerobic decomposition in
the center of the compost pile), compost researchers believe that the CHU is almost always oxidized to
CO2 before it escapes from the compost pile.
Because the CO2 emissions from composting are biogenic—generally producing no CBU
emissions—the only GHG emissions from composting result from transportation of compostable
materials to composting facilities and mechanical turning of the compost piles. GHG emissions .
associated with compost application are discussed in Chapter 5.
12
-------
Combustion: When waste is combusted, two GHGs are emitted: CO2 and N2O. Non-biogenic
CO2 emitted during combustion (i.e., CO2 from plastics) is counted toward the GHG emissions associated
with combustion, but biogenic CO2 is not. Because most waste combustors produce electricity that
substitutes for utility-generated electricity, the net GHG emissions are calculated by subtracting the
utility GHG emissions avoided from the gross GHG emissions. GHG emissions from combustion are
described in Chapter 6.
Landfilling: When organic matter is landfilled, some of this matter decomposes anaerobically
and releases CBU, a GHG. Some of the organic matter never decomposes at all; instead, it becomes stored
carbon. (Landfilling of metals and plastics does not result in either CEU emissions or carbon storage.)
At some landfills, virtually all of the CBU produced is released to the atmosphere. At others,
is captured for flaring or combustion with energy recovery (i.e., electricity production). Most of the
captured CBU is converted to CO2, but that CO2 is not counted in this study as a GHG because it is
biogenic. With combustion of CB^ for energy recovery, credit is given for the electric utility GHG
emissions avoided. Regardless of the fate of the Cttj, credit is given for the landfill carbon storage
associated with landfilling of some organic materials. GHG emissions and carbon sinks from landfilling
are described in Chapter 7.
1.5 ESTIMATING AND COMPARING NET GHG EMISSIONS
To calculate the net GHG implications of a waste management strategy for a given material, a
baseline and alternative scenarios must be established. For example, a baseline scenario in which 10 tons
of office paper are manufactured, used, and landfilled could be compared with an alternative scenario in
which 10 tons are manufactured, used, and recycled. For each scenario, net GHG emissions are estimated
based on (1) the GHG emissions associated with that material, and (2) any increases in carbon stocks
and/or displaced fossil fuel combustion that offset those emissions. The formula for net GHG emissions
is as follows:
Net GHG emissions = Gross GHG emissions - (Increase in carbon stocks + Avoided utility GHG
emissions)
Comparing net GHG emissions for the two scenarios enables the lowest net GHG emissions to be
identified. For example, when a material is source reduced (i.e., some or all of it is not produced), GHG
emissions throughout the life cycle are avoided. In addition, when paper products are source reduced,
additional carbon may be sequestered in forests.
Similarly, when a material is recycled, the GHG emissions from making an equivalent amount of
material from virgin inputs are reduced. In most cases, recycling reduces GHG emissions because
manufacturing a product from recycled inputs requires less fossil energy than making the product from
virgin inputs and thus reduces energy-related GHG emissions.
If a waste is not source reduced or recycled, it may be either composted (if it is organic matter),
combusted, or landfilled. In any of these cases, GHG emissions are produced during acquisition and
manufacture. These GHG emissions may be augmented by OHU emissions from landfills (which
themselves may be offset to some degree by energy recovery at landfills or landfill carbon storage). If the
wastes are combusted, there may be an offset for avoided utility emissions.
In calculating emissions for the life-cycle scenarios, we can use the following two reference
points:
• In a "raw material extraction" approach (i.e., cradle-to-grave perspective), raw material
acquisition can be used as the "zero point" for emissions, with all emissions being added
(and sinks deducted) from that point on through the life cycle.
13
-------
• In a "waste-generation" approach (solid waste manager's perspective), accounting for GHG
emissions can start at the point of waste generation. All subsequent emissions and sinks from
waste management practices then are accounted for. Changes in emissions and sinks from
raw material acquisition and manufacturing processes are captured to the extent that certain
waste management practices (i.e., source reduction and recycling) affect these processes.
Because it is the difference in emissions between the baseline and alternate scenarios that is
meaningful, using either of these reference points yields the same results. The March 1997 draft working
paper used the cradle-to-grave method to display GHG emissions because it is most consistent with
standard accounting techniques for life-cycle inventories. Several reviewers pointed out that solid waste
decision-makers tend to view raw materials acquisition and manufacturing as beyond their control and
suggested that a waste generation GHG accounting approach would provide increased clarity for
evaluating waste management options. Thus, both editions of this report use the waste generation
approach and define the "standard" raw material acquisition and manufacturing step for each material as
consisting of average GHG emissions based on the current mix of virgin and recycled inputs.19
Exhibit 1-3 indicates how GHG sources and sinks have been counted for each MSW
management strategy in order to estimate net GHG emissions using the post-consumer reference point.
For example, the top row of the exhibit shows that source reduction (1) reduces GHG emissions from raw
materials acquisition and manufacturing, (2) results in an increase in forest carbon sequestration, and (3)
does not result in GHG emissions from waste management. The sum of emissions (and sinks) across all
steps in the life cycle represents net emissions. Section 8.2, "Applying Emission Factors," describes how
waste managers and companies have used these emission factors to estimate GHG emissions and
potential GHG emission reductions associated with integrated waste management. In addition, EPA used
these emission factors to develop the Waste Reduction Model (WARM). WARM enables users to
analyze the GHG savings associated with changing their waste management practices. WARM is
available on EPA's Web site at .
19 Changes in the mix of production (i.e., higher proportions of either virgin or recycled inputs) result in
incremental emissions (or reductions) with respect to this reference point.
14
-------
2. RAW MATERIALS ACQUISITION AND MANUFACTURING
The GHG emissions associated with raw materials acquisition and manufacturing are a key
element of a life-cycle GHG analysis. This chapter describes how we estimated these emissions for 15
materials: aluminum cans, steel cans, glass, three types of plastic (HOPE, LDPE, and PET), corrugated
cardboard, magazines/third-class mail, newspaper, office paper, phonebooks, textbooks, dimensional
lumber, medium-density fiberboard, and mixed paper.
In manufacturing, substantial amounts of energy are used both in the acquisition of raw materials
and in the manufacturing process itself. In general, the majority of energy used for these activities is
derived from fossil fuels. Combustion of fossil fuels results in emissions Of CO2, a GHG. In addition,
manufacturing of some materials also results in GHG emissions that are not associated with energy
consumption. Section 2.1 addresses energy-related CO2 emissions, and Section 2.2 covers non-energy
GHG emissions. Sections 2.3 and 2.4 discuss results and limitations of the analysis, respectively.
2.1 GHG EMISSIONS FROM ENERGY USE IN RAW MATERIALS ACQUISITION AND
MANUFACTURING
To begin our analysis, we estimated the GHG emissions from fossil fuel combustion for both (1)
raw materials acquisition and manufacturing (referred to here as "process energy"), and (2)
transportation (referred to as "transportation energy").
In this analysis, process energy GHG emissions consist primarily of CO2.1 The majority of CO2
emissions are from combustion of fuels used directly, e.g., to operate mining equipment or fuel a blast
furnace. CO2 emissions from fuels used to generate electricity during the manufacturing stage also are
included in process energy emissions. In addition, process energy GHG emissions include indirect
emissions from "pre-combustion" activities, such as oil exploration and extraction, coal mining and
beneficiation, and natural gas production.
Transportation energy GHG emissions consist of CO2 emissions from combustion of fuels used
to transport raw materials and intermediate products to the final manufacturing or fabrication facility. For
transportation of recycled inputs, this analysis considers transportation (1) from the curbside to the
materials recovery facility (MRF), (2) from the MRF to a broker, and (3) from a broker to the plant or
mill where the recycled inputs are used. The transportation values for recycled inputs generally include
the energy used to process the inputs at an MRF. Transportation of finished manufactured goods to
consumers is not included in the analysis. We did not estimate transportation emissions of CH4 or N2O;
these emissions are considerably less significant than CO2 emissions.2 This omission would tend to
understate the GHG impacts from transportation slightly.
Emissions from raw materials acquisition and manufacturing also include CEU associated with
producing, processing, and transporting coal, oil, and natural gas. CHU is emitted during the various
stages of fossil fuel production because CBU is trapped within coal and oil deposits, and is released when
they are mined. Natural gas, of course, consists largely of CBU.
Note, however, that CO2 emissions from combustion of biomass (e.g., in paper manufacturing) are not
counted as GHG emissions (as described in Chapter 1).
2 The Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-1999 estimates 1999 emissions from
transportation to be 468.1 MMTCE for CO2 and 18.5 MMTCE for CBU and N2O combined.
15
-------
We developed separate estimates for GHG emissions from process and transportation energy for
virgin inputs and recycled inputs, generating a total of four separate GHG emissions estimates for each
material: (1) process energy with virgin inputs, (2) process energy with recycled inputs, (3) transportation
energy with virgin inputs, and (4) transportation energy with recycled inputs.
2.1.1 Methodology
We developed GHG emission estimates for each material based on two sets of data: (1) the
amount of each type of fuel used to make 1 ton of the material, and (2) the "carbon coefficient" for each
fuel (a factor that translates the energy value of fuel combusted into the mass of GHGs emitted).
Our methodology in using these two sets of data to estimate process and transportation energy
GHG emissions is best illustrated by an example: To estimate process energy GHG emissions from the
production of 1 ton of newspaper from virgin inputs, we multiplied the amount of each type of fuel
consumed (as measured in million Btu) by the carbon coefficient for that type of fuel (as measured in
metric tons of carbon equivalent, or MTCE, per million Btu). The result was an estimate of the GHG
emissions (in MTCE) from the combustion of each type of fuel required to make 1 ton of newspaper.
Total process energy GHG emissions from making 1 ton of newspaper are simply the sum of the GHG
emissions across all of the fuel types. To estimate the GHG emissions when electricity is used, we used
the national average mix of fuels used to generate electricity.
We estimated GHGs from the energy used to transport raw materials for making 1 ton of a given
product (e.g., newspaper) in the same way. The amount of each fuel used was multiplied by its carbon
coefficient, and the resulting values for each of the fuels were summed to yield total transportation
energy GHG emissions.
In this way, GHG estimates for raw materials acquisition and manufacturing were developed for
each of the manufactured materials considered. As noted in Chapter 1, much of the energy information in
this edition of the report is drawn from an effort conducted by EPA's Office of Research and
Development (ORD) to construct a Decision Support Tool for solid waste managers. The remaining
energy data was developed by Franklin Associates, Ltd. (FAL) as part of the original effort or subsequent
updates.
Most of the materials included in this analysis are assumed to undergo closed-loop recycling
(i.e., materials are remanufactured into a similar product). However, mixed paper is recycled in an open
loop into boxboard and paper towels.3 Thus, the exhibits in this chapter show data not only for the 15
materials of interest, but also for boxboard and paper towels. Because recycling processes data are
similar for HOPE, LDPE, and PET, we adopted the approach used by ORD of using a single energy
profile (fuel mix and energy intensity) for all recycled plastics. For steel cans, we developed GHG
estimates for virgin production using the basic oxygen furnace process,4 and for recycled production, we
used the electric arc furnace process.5
3 FAL provided virgin and recycled manufacturing and transportation data for boxboard and paper towels.
For virgin boxboard, only one type of product is manufactured, as shown in Exhibits 2-3 and 2-4. For recycled
boxboard, there are two types of products, and we obtained two different sets of manufacturing and transportation
data as shown in Exhibits 2-5 and 2-6. We have labeled the two types of boxboard as boxboard "A" and boxboard
"B." These two products differ only with respect to their recycled material inputs (i.e., the proportion of newspaper,
coirugated cardboard, office paper, and coated paper used to manufacture either boxboard "A" or boxboard "B");
both products share the same manufacturing and transportation values for virgin inputs.
4 Note that the basic oxygen furnace process can utilize approximately 25 percent recycled inputs.
5 Note that when recovered steel cans are used as inputs to an electric arc furnace, the resulting steel is not
suited for milling to the thinness of steel sheet needed for use in making new steel cans. Thus, a more precise
16
-------
We used carbon coefficients from the U.S. Department of Energy's Energy Information
Administration for all fuels except electricity.6 The carbon coefficient for electricity was based on the
weighted average carbon coefficients for all fuels used to generate electricity in the United States.7
Because the carbon coefficients from these sources accounted for only the CO2 emissions from
combustion of each type of fuel, we added to these carbon coefficients (1) the average amount of CHU
emitted during the production, processing, and transportation of fossil fuels, and (2) the average CO2
emissions from oil production, due to the flaring of natural gas. We calculated the average fugitive GHG
emissions associated with U.S. production of coal, oil, and natural gas. The resulting average estimates
for fugitive GHG emissions from fossil fuel production were 0.92 kilograms of carbon equivalent per
million Btu (kg CE/million Btu) for coal, 0.10 kg CE/millioh Btu for oil, and 0.70 kg CE/million Btu for
natural gas.8
The carbon coefficients that reflect both CO2 and CHL, emissions are supplied in Exhibit 2-1. (All
exhibits are provided at the end of this chapter.)
The process and transportation GHG values are shown in summary form in Exhibit 2-2. For each
product and each type of input (virgin or recycled), we summed the estimates for process and
transportation GHG emissions, as shown in columns "b" (for virgin inputs) and "c" (for recycled inputs)
of Exhibit 2-2. We also estimated the energy-related GHG emissions from manufacturing each material
from the current mix of virgin and recycled inputs. These values are shown in column "e." (The
remaining two columns of Exhibit 2-2 are discussed later in this chapter.)
The energy intensity and fuel mix data are provided in Exhibits 2-3 through 2-6. For most
materials, the data in the exhibits are for manufacturing processes that either use (1) 100 percent virgin
inputs or (2) 100 percent recycled inputs.9
To estimate the types and amounts of fuels used for process and transportation energy, ORD and
FAL relied on published data (such as engineering handbooks and published production data), contacts
with industry experts, and review by stakeholders and trade organizations. ORD and FAL counted all
energy, no matter where it was used. For example, much aluminum produced in the United States is made
approach would have been to model recovery of steel cans as an open-loop process, in which recovered steel cans
are made into some other steel product. By modeling recovery of steel cans as a closed-loop process, we implicitly
assumed that 1 ton of steel produced from recovered steel cans in an electric arc furnace displaces 1 ton of steel
produced from virgin inputs in a basic oxygen furnace. We believe this is a reasonable assumption. (For the
fabrication energy required to make steel cans from steel, we used the values for fabrication of steel cans from steel
produced in a basic oxygen furnace.)
U.S. Department of Energy, Energy Information Administration. 2000. Annual Energy Review: 1999.
FAL reported the Btu value for electricity in terms of the Btu of fuel combusted to generate the electricity
used at the factory, rather than the (much lower) Btu value of the electricity that is delivered to the manufacturer.
Thus, FAL had already accounted for the efficiency of converting fuels to electricity, and the losses in transmission
and distribution of electricity. We therefore did not need to account for these factors in the carbon coefficient for
electricity.
8ICF Consulting. 1995. Memorandum, "Fugitive Methane Emissions from Production of Coal, Natural
Gas, and Oil," August 8, updated to use global wanning potential for CHj of 21.
9 In the FAL data set, the one exception is the data for steel cans made from virgin inputs, for which FAL
provided data for manufacture from 80 percent virgin inputs and 20 percent recycled inputs. We extrapolated from
this data (and the corresponding values for production using 100 percent recycled inputs) to obtain estimates of the
energy inputs for manufacturing these materials from 100 percent virgin inputs. Similarly, for corrugated cardboard,
ORD assumed that a virgin corrugated box contains a minimum of 14.7 percent total recycled content.
17
-------
from bauxite that is mined and processed into alumina in other countries. The energy required for
overseas bauxite mining and processing is counted in the analysis.
Neither the ORD nor the FAL transportation data reflect transportation of the finished
manufactured product to the retailer and consumer. This omission is only important in estimating the
GHG reductions associated with source reduction (i.e., source reduction reduces transportation energy).
It is not relevant in analyzing GHG implications of recycling compared to other post-consumer
management options, because the amount of transportation energy from the factory to the consumer is
about the same whether the product is manufactured from virgin inputs or recycled inputs. Even for the
source reduction analysis, we expect that the transportation energy from factory to consumer would
represent a very small fraction of the total process and transportation energy.
Finally, it should be noted that during our extensive review of ORD and FAL data, we examined
the most critical assumptions and data elements that each model used to ensure that they accurately
reflect the energy requirements of the raw materials acquisition and manufacturing for the material types
considered. Nevertheless, we recognize that different manufacturers making the same product use
somewhat different processes with different energy requirements and fuel mixes, and that there are
limited data on the extent to which various processes are used. Thus, although our goal was to estimate as
accurately as possible the national average GHG emissions for the manufacture of each material from
virgin and recycled inputs, it is quite likely that individual companies will have GHG emissions that vary
significantly from those estimated here.
2.2 NON-ENERGY GHG EMISSIONS FROM MANUFACTURING AND RAW
MATERIALS ACQUISITION
In addition to GHG emissions from energy use, we also accounted for three additional sources of
GHGs in manufacturing processes:
• When limestone (calcium carbonate, or CaCO3) is converted to lime (calcium oxide, or
CaO), CO2 is emitted. Significant quantities of lime are used in the production of steel,
aluminum, and, to a much lesser extent, office paper.
emissions from natural gas pipelines and processing of natural gas are associated with
the manufacture of plastic products.
• Perfluorocarbons (CF4 and C2F6) are emitted during aluminum smelting.
For plastics and office paper, process non-energy GHG emissions are associated only with
production using virgin inputs. In the case of steel, however, these emissions result when either virgin or
recycled inputs are used (because lime is used in the production of steel from recycled as well as virgin
inputs).
The process non-energy GHGs for each material are shown in the second-to-last column of
Exhibits 2-3 and 2-5 (for manufacture from virgin inputs and recycled inputs, respectively), and are
repeated in column "f ' of Exhibit 2-2. ORD supplied the non-energy CO2 emissions for aluminum, glass,
corrugated cardboard, and newspaper. We based our calculation for PFC emissions from aluminum on
the Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-1998.10
10 To estimate aluminum PFC emissions on a per-ton basis, we divided the inventory estimates for CF4 and
C2Ffi emissions from aluminum by total primary aluminum production, yielding units in MTCE/ton.
18
-------
Non-energy CO2 emissions for the other materials, as well as CELj emissions, are based on the
original analysis supporting the first edition of this report.11
2.3 RESULTS ' ' .' " .
Our estimates of the total GHG emissions from raw materials acquisition and manufacturing for
each material are shown in Exhibit 2-2, column "g." In order to obtain these estimates, we summed the
energy-related GHG emissions (column " e" ) and the nbtt-energy GHG emissions (column " f).
The process energy and transportation GHG values that were developed as described earlier in
this chapter are shown in the third-to-last columns of Exhibits 2-3. and 2-5, and in the last columns of
Exhibits 2-4 and 2-6 (the last columns of Exhibits 2-3 and 2-5 show the total process energy GHG
emissions). , - .•: •-.--.
Total GHG emissions associated with the raw materials acquisition and manufacturing stage of
the product life cycle are shown in the three righthand columns of Exhibit 2-2. These three columns
correspond to the type of inputs that occur during the recycling process: virgin inputs, recycled inputs, or
the current mix of virgin and recycled inputs.
2.4 LIMITATIONS
There are several limitations to the analysis of the GHG emissions associated with raw materials
acquisition and manufacturing, as described below.
The approach used in this analysis provides values for the average GHG emission rates per ton of
material produced, not the marginal emission rates per incremental ton produced. In some cases, the
marginal emission rates may be significantly different. For example, reducing production of plastic
products from virgin inputs may not result in a proportional decrease in CHU emissions from natural gas
pipelines and natural gas processing. The operating pressure in natural gas pipelines and the number and
size of leaks in the pipeline determine CH* emissions from natural gas pipelines. Consequently, the
amount of natural gas consumed at one end of the pipeline (e.g., to make plastic) does not affect the level
of pipeline CKU emissions in a direct, linear way. As another example, long-term reductions in electricity
demand could selectively reduce demand for specific fuels, rather than'reducing demand for all fuels in
proportion to their representation in the current average fuel mix. This analysis estimates average carbon
conversion rates largely because the marginal rates are much more difficult to estimate. Nevertheless, we
believe the average values provide a reasonable approximation of the GHG emissions.
In addition, the analysis assumes that the GHG emissions from manufacturing a given product
change in a linear fashion as the percentage of recycled inputs moves from 0 to 100 percent. In other
words, the analysis assumes that both the energy intensity arid the fuel mix change in linear paths over
this range. However, it could be that GHG emissions from manufacturing move in a non-linear path,
(e.g., some form of step function) when the percentage of recycled inputs changes, due to capacity limits
in manufacturing or due to the economics of manufacturing processes. , •
The transportation energy required for the final stage of transportation (to the consumer) was not
considered. Consequently, some carbon emissions reductions for "lightweighted" products for these
transportation stages were not considered; these savings are likely to be small.
ICF Consulting. 1994. Memorandum, "Detailed Analysis of Greenhouse Gas Emissions Reductions from
Increased Recycling and Source Reduction of Municipal Solid Waste," July 29, p. 48 of the Appendix prepared by
Franklin Associates, Ltd., dated July 14, 1994. '• : • ••
19
-------
The information used in this analysis represents the best available data from published and
unpublished industry sources, some of it quite dated. Therefore, the data may not necessarily reflect
recent trends in industrial energy efficiency or changes in the fuel mix.
Finally, this static analysis does not consider potential future changes in energy usage per unit of
output or alternative energy (e.g., non-fossil) sources. Reductions in energy inputs due to efficiency
improvements could occur in either virgin input processes or recycled input processes. Efficiency
improvements and switching to alternative energy sources will result directly in GHG emissions
reductions and may change the reductions possible through increased recycling or source reduction.
20
-------
Exhibit 2-1
Carbon Coefficients For Selected Fuels
Fuel Type
basolme
LPfi
Distillate Fuel
Residual Fuel
DipQpl
Oil/Lubricants
Steam (non-paper products)
Steam (paper products)
Mational Average Fuel Mix for Electricity
National Average Fossil Fuel Mix for
Electricity
Coal Used for Electricity
Coal Used by Industry (Non-Coking
Coal)
Natural Gas
Nuclear
Other (Petroleum Coke)
Metric Tons of C02
from Combustion Per
Million Btu
0.07
0.06
0.07
0.08
0.07
0.07
0.07
0.05
0.06
0.08
0.09
0.09
0.05
0.00
0.10
kg Carbon
Equivalent (CE)
from
Combustion Per
Million Btu
19.25
16.91
19.87
21.41
19.87
20.16
18.21
12.92
15.79
23.18
24.86
25.10
13.78
084
27.78
Metric Tons of
Fugitive CH4
Emissions Per
Million Btu
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00011
0.00004
0.00010
0.00015
0.00016
0^00016
0.00012
0.00002
kg CE from
Fugitive
Methane
Emissions Per
Million Btu
0.098
0.10
0.10
0.10
0.10
0.10
0.61
0.25
0.59
0.86
0.92
0.92
0.70
0.10
kg CE Emitted
Per Million Btu
Consumed
19.35
17.01
19.97
21.51
19.97
20.26
18.81
13.17
16.25
24.04
25.78
26.02
14.48
0.84
27.87
21
-------
Exhibit 2-2
GHG Emissions from the Manufacture of Selected Materials
(Metric Tons of Carbon Equivalent (MICE) per Ton of Product)
(a)
Type of Product
Aluminum Cans
Steel Cans
Glass Containers
HOPE
.DPE
DET
Corrugated Boxes
Magazines/Third-class Mail
Newspaper
Office Paper
Phonsbooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Mixed Paper
Broad Def'n (= Boxboard "A")
Residential Def'n (= Boxboard "B")
Office Def'n (= Paper Towels)
(b)
Virgin Input Combined
Process and
Transportation Energy
MTCE per Ton of Product
Made With Virgin Inputs)
3.52
n 77
(J,if
0.11
0.48
0.59
0.55
0,22
0.46
0.59
0.27
0.67
0.59
0.05
0.10
0.32
0.32
0.91
(c)
Recycled Input
Combined Process and
Transportation Energy
Emissions
(MTCE per Ton of
Product Made With
Virgin Inputs)
0.25
n 97
U.£(
0.07
0.04
0.04
0.04
0.46
n Q.7
U.O/
n /n
0.41
n K7
U.O/
0.07
0.12
0.4c
0.43
0.75
W
Percent Recycled Inputs
in the Current Mix of
Virgin and Recycled
Inputs
49%
44%
9%
4%
18%
62%
22%
52%
32%
12%
13%
fWn
U /o
C-JO/
O I /
coo/
OO/
38°/
(e)
Current Mix Combined
Process and Transportation
Energy Emissions
(MTCE per Ton of Product
Made With the Current Mix of
Virgin and Recycled Inputs)
1.90
0.55
0.10
0.44
0.56
0.46
0.24
0.46
0.46
0.30
0.64
0.59
0.05
n m
U. I'
0.3
0.3
0.85
(0
Process Non-Energy Emissions (MTCE
Virgin Inputs
1.15
0.24
0.04
0.05
0.05
0.03
0.00
0.00
0.00
0.01
0,00
0.00
0.00
0.00
0.00
0.00
0.00
Recycled
Inputs
0.02
0.24
0.00
0.00
0.00
0.00
0.00
0.00
0,00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Current Mix
0.59
0.24
0.03
0.05
0.05
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
(g)
Average Combined Process and
Transportation Energy and Process
Non-Energy Emissions (MTCE per
Ton of Product)
Virgin
Inputs
4.67
1.01
0.16
0.53
0.64
0.58
0.22
0.46
0,59
0.28
0.67
0.59
0.05
0.10
0.32
0.32
0.91
Recycled
Inputs
0.27
0.51
0.07
0.04
0.04
0.04
0.25
0.46
0.34
0.37
0.41
0.57
0.07
0.12
0.43
0.43
0.75
Current
Mix
2.49
0.79
0.14
0.49
0.61
0.49
0.24
0.46
0.46
0.31
0.64
0.59
0.05
0.10
0.38
0.38
0.85
Explanatory notes: To estimate the GHG emissions from manufacturing, we first estimated the process and transportation GHG emissions when 100
percent vkgin inputs, or 100 percent recycled inputs, are used. For each product and each type of input (virgin or recycled), we summed the estimates for process
and transportation GHG emissions. Next we estimated the GHG emissions from manufacturing each material from the current mix of virgin and recycled inputs.
We began with estimates of the percentage of recycled inputs currently used in the manufacture of each material, as shown in column d. We used these
percentages to develop a weighted average value for the GHG emissions associated with the manufacture of each material from the current mix of virgin and
recycled inputs. Specifically, we used the estimate of the percentage of recycled inputs in the current mix, together with the estimates for GHG emissions from
manufacture using vkgin or recycled inputs, to develop estimates of GHG emissions from manufacture using the current mix of virgin and recycled inputs
(column "e").
22
-------
Explanatory notes for Exhibit 2-2 (continued):
C° "f
1 <
' ^
n0n-ener^ GHG emissions from manufacturing. First, this column shows the process non-energy GHG
< ^ emiSSiODS When reCyded lnpUtS « USed (these Values «» ^^ c°Pied fr°m the final cois of
/' ^ u ^^ n°n'energy °HG ^^ ^ manufacturing each product from the current mix of virgin and
^
23
-------
Exhibit 2-3 ..
GHG Emissions Per Ton of Product Manufactured from Virgin Inputs
Process GHGs Only
Aluminum Cans
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Dimensional Lumber
Medium-density Fiberboard
Process
Energy
(Million Btu
Per Ton of
205.80
31.58
6.49
28.69
35.26
32.82
25.13
32.99
39.92
37.01
39.61
35.07
2.53
10.18
32.26
7344
Average Fuel Mix (in Percent)
Gasoline
0.16
0.21
0.55
0.00
0.00
0.00
0.01
0.15
0.25
0.08
0.18
0.18
1.57
0.14
0.00
0.00
LPG
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.01
0.01
0.00
0.00
0.00
0.00
Distillate
Fuel
0.82
5.06
1.45
0.00
0.00
0.00
0.02
0.32
0.52
0.17
0.38
0.38
0.00
0.38
0.00
0.01
Residual
Fuel
4.06
0.35
0.47
33.14
32.59
36.67
0.54
8.30
0.75
; 4.33
9.99
9.96
0.00
0.05
0.94
1.80
Biomass/H
ydro
0.03
0.00
0.03
1.16
1.56
1.62
61.33
24.27
9.09
60.53
8.86
9.14
32.81
51.90
59.34
24.89
Diesel
0.50
0.00
0.00
0.00
0.00
0.00
1.20
o.b'b'
0.68
0.00
0.00
0.00
15.99
1.26
1.36
0.45
Electricity
80.36
21.02
10.12
5.64
7.66
7.10
14.06
•vs 25.40
54.2,1
13.24
30.56
30.47
43.09
27.61
5.32
28.15
Coal
0.83
53.90
7.18
4.59
6.15
6.42
15.52
17.11
1.75
, ,,8.92
2059
20.52
0.00
0.00
24.01
2.93
Natural
Gas
13.04
19.45
79.95
51.35
46.52
42.41
7.31
24.33
32.43
12.68
29.28
V' 29.19
6.53
18.68
9.02
41.78
Nuclear
0.17
0.00
0.23
4.00
5.36
5.59
0.01
0.11
0.27
0.06
0.13
0.13
0.00
0.00
0.00
0.0(
Other
0.02
0.00
0.02
0.13
0.17
0.18
0.00
0.01
0.04
0.01
0.02
0.02
0.00
0.00
0.00
0.0(
Total
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
10(
Process Energy
GHG Emissions
"-(WITCE/Ton of
Product)
3.36
0.68
0.10
0.48
0.59
0.55
0.19
0.46
0.58
0.27
0.67
0.59
0.03
0.08
0.29
0.87
Process Non-
Energy GHG
Emissions
(MTCEfTon of
Product)
1.15
0.24
0.04
0.05
0.05
0.03
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
Total Process
GHG
Emissions
(MTCE/Ton of
Product)
4.51
0.91
0.15
0.53
0.64
0.58
0.20
0.46
0.58
0.28
0.67
0.03
0.08
0.29
0.87
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
24
-------
Exhibit 2-4
GHG Emissions Per Ton of Product Manufactured from Virgin Inputs
Transportation GHGs Only
Type of Product
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/TTrird-class Mail
Newspaper
Office Paper
'honebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Boxboard
Paper Towels
Transportation
Energy
(Million Btu Per
Ton of Product)
7.47
4.60
0.58
NA
NA
NA
1.31
NA
0.50
NA
NA
NA
0.88
1.01
1.79
2.07
Gasoline
0.10
0.00
0.10
0.05
0.10
0.00
0.00
0.00
0.00
LPG
0.08
0.00
0.08
0.00
0.08
0.00
0.00
0.00
0.00
Distillate
Fuel
0.39
0.00
0.40
0.05
0.39
0.00
0.00
0.00
0.00
Average Fuel Mix (in Percent)
Residual
Oil
79.88
1.76
2.64
0.27
3.63
0.00
0.06
0.05
0.52
Biomass/Hy
dro
0.05
0.00
0.04
0.01
0.05
0.00
0.00
0.00
0.00
Diesel
11.34
98.24
88.95
98.51
87.97
100.00
98.10
99.93
99.46
Electricity
0.34
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
Coal
0.86
0.00
0.89
0.00
0.86
0.00
0.00
0.00
0.00
Natural
Gas
6.58
0.00
6.51
1.07
6.53
0.00
1.84
0.01
0.02
Nuclear
0.33
0.00
0.36
0.04
0.34
0.00
0.00
0.00
0.00
Note that totals may not add due to rounding, and more digits may be displayed than are significant. ' '
Other
0.05
0.00
0.03
001
0.05
0.00
0.00
0.00
0.00
Total
100
100
100
100
100
100
100
100
100
Transportation
Energy GHG
Emissions
(MTCE/Ton of
Product)
0.16
0.09
0.01
NA
NA
NA
n n"3
NA
001
NA
NA
NA
0.02
002
0 04
0 04
Note that for some materials, transportation data was included in the process energy estimates and not provided separately, denoted by "NA" in this table.
25
-------
Exhibit 2-5
GHG Emissions Per Ton of Product Manufactured from Recycled Inputs
Process GHGs Only
Type of Product
Aluminum Cans
Steel Cans
3lass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Made from Reused Dimensiona
Lumber
Boxboard Made from Broad
Definition of Mixed Paper
Boxboard Made from
Residential Definition of Mixed
Paper.
Paper Towels Made from
Recoverd File Stock
Process
Energy
(Million Btu
Per Ton of
Product)
14.85
11.78
4.32
4.17
4.17
4.17
11.73
31.97
21.98
20.12
22.02
33.51
3.17
10.99
22.53
22.53
51.69
Averag
Gasoline
0.27
0.01
0.55
0.03
0.03
0.03
0,01
0.16
0.30
Oi20
0^20
0.21
0.00
0.13
0.00
0.00
0.00
LPG
0.01
0.17
0.00
0.03
0.03
0.03
0.05
0.01
0.00
0.01
0.01
0.01
0.00
0.00
0.03
0.03
0.00
Distillate
Fuel
5.12
0.07
0.39
1.05
1.05
1.05
0.05
0.32
0.58
0,42
0.42
0.60
0.00
0.35
0.02
0.02
0.00
Residual
Fuel
0.44
0.03
0.26
1.24
1.24
1.24
0.66
8.45
0.30
10,96
10.96
10.02
0.00
0.04
0.36
0.36
0.45
3iomass/H
ydro
0.05
0.00
0.03
12.48
12.48
12.48
0.00
22.85
0.05
0.02
0.02
8.38
0.00
48,05
0.00
0.00
6.94
Fuel Mix (in Percent)
Diesel
• 0.52
0.00
0.00
0.05
0.05
0.05
0.31
0.00
0.00
0.00
0.00
0.00
23.61
8.56
0.49
0.49
0.15
Electricity
_. 63.74
77.28
5.10
33.21
33.21
33.21
51.11
25.87
57.75
^33.53
33.53
30.40
76.39
25.56
67.46
67.46
36.32
Coal
0.75
0.65
0.54
0.02
0.02
0.02
38.40
17.43
1.09
22.58
22.58
20.61
0.00
0.00
24.36
24.36
0.98
Natural
Gas
28.78
21.80
92.91
20.34
20.34
20.34
9.40
24.79
39.59
. 32,12
32.12
29.57
0.00
17.29
7.25
7.25
55.14
Nuclear
0.29
0.00
0.21
0.09
0.09
0.09
0.01
0.11
0.30
0.14
0.14
0.17
0.00
0.00
0.00
0.00
0.00
Other
0.04
0.00
0.02
31.44
31.44
31.44
0.00
0.01
0.04
0.02
0.02
0.02
0.00
0.00
0.00
0.00
0.00
Total
100
100
100
100
100
100
. 100
100
100
100
100
100
100
100
100
100
100
Process Energy
GHG Emissions
(MTCE/Ton of
Product)
0.24
0.19
0.06
0.04
0.04
0.04
0.23
0.46
0.34
0.37
0.41
0.57
0.05
0,09
0.42
0.42
0.74
Energy GHG
Emissions
(MTCEnVjn of
Product)
0.02
0.24
0.00
0.00
0,00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
GHG
Emissions
(MTCE/Ton of
Product)
0.26
0.43
0.06
0.04
0.04
0.04
0.23
0.46
' 0.34
0.37
0.41
0.57
0.05
0.09
0.42
0.42
0.74
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
* Recycled boxboard using a "broad" definition of mixed paper comprised of 24 percent newsprint, 48 percent corrugated cardboard, 20 percent office paper, and 8 percent coated paper.
** Recycled boxboard using a "residential" definition of mixed paper comprised of 23 percent newsprint, 53 percent corrugated cardboard, 14 percent office paper, and 10 percent coated paper.
*** Recycled boxboard using an "office paper" definition of mixed paper comprised of 21 percent newsprint, 5 percent corrugated cardboard, 38 percent office paper, and 36 percent coated paper.
26
-------
Exhibit 2-6
GHG Emissions Per Ton of Product Manufactured from Recycled Inputs
Transportation GHGs Only
r
i.
f
f
:ii
t
jr
h-
f
f"'
i-
•
,-
•
,.
b
Type of Product
Aluminum Cans
Steel Cans
Glass
HOPE
L>DEE
B&i: - : .-•
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper,
3honebooks
Textbooks
Recycled Lumber from
Medium-density Fiberboard
Made from Reused
Dimensional Lumber
3oxboard Using the "Broad
Definition" for Mixed Paper .
Boxboard Using the "Single-
:amily Residential Definition"
or Mixed Paper
3aper Towels Using "Office
Paper" for Mixed Paper
Transportation
Energy
(Million Btu Per
Ton of Product)
0.40
4.03
0.34
0.08
0.08
0.08
0.80
NA
0.03
NA
IMA
NA
0.97
1.27
0.01
0.01
0.01
Gasoline
0.08
0.00
0.10
0.00
o.oo
0.00
0.05
0.10
0.00
0.00
6.03
0.03
0.03
Average Fuel Mix (in Percent)
LPG
0.06
0.00
0.08
0.00
0.00
0.00
0.00
0.08
0.00
0.00
O.QO
0.00
0.00
Distillate
Fuel
0.32
0.00
0.41
0.00
0.00
0.00
0.05
0.39
0.00
0.00
0.00
0.00
0.00
Residua!
Oil
3.07
0.01
2.59
56.50
56.50
56.50
0.22
3.75
.0.00
0.05
.' d.oo
0.00
0.00
Biomass/Hy
dro
0.04
0.00
0.04
0.00
0.00
0.00
0.01
0.05
0.00
0.00
(
0.00
0.00
0.00
Diesel
90.11
99.99
89.01
5.96
5.96
5.96
98.55
87.87
100.06
98.46
0.00
0.00
0.00
Electricity
0.00
0.00
0.00
2.53
2.53
2.53
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Coal
0.68
0.00
0.89
0.00
0.00
0.00
0.00
0.86
0.00
0.00
0.00
0.00
..
0.00
Natural
Gas
5.32
0.00
6.51
35.01
35.01
35.01
1.07
6.53
.. 0.00
., i-47
0.00
0.00
0.00
Nuclear
0.27
0.00
0.35
0.00
0.00
0.00
0.04
0.32
0.00
0.00
0.00
0.00
0.00
Other
0.04
0.00
0.03
0.00
0.00
0.00
0.00
0.05
0.00
0.00
0.00
0.00
0.00
Total
100
100
100
100
100
100
ioo
100
• 100
100
' 0
• 0
0
Transportation
Energy GHG
Emissions
(MTCEAon of
Product)
0.01
0.08
0.01
0.00
0.00
0.00
0.02
' 0.00
• o.oo
0.00
0.00
0.00
0.02
0.03
" 0.00
0.00
0.00
Note that for some materials, transportation data was included in the process energy estimates and not provided separately, denoted by "NA" in this table.
"* Recycled boxboard using a "broad" definition of mixed paper comprised of 24 percent newsprint, 48 percent corrugated cardboard, 20 percent office paper,
and 8 percent coated paper.
** Recycled boxboard using a "residential" definition of mixed paper comprised of 23 percent newsprint, 53 percent corrugated cardboard, 14 percent office paper,
and 10 percent coated paper. : •.*•*. ••.,-•.
*** Recycled boxboard using an "office paper" definition of mixed paper comprised of 21 percent newsprint, 5 percent corrugated cardboard, 38 percent office paper,
and 36 percent coated paper.
27
-------
This page intentionally left blank
28
-------
3. FOREST CARBON SEQUESTRATION
This chapter presents estimates of the forest carbon sequestration that results from recycling or
source reducing corrugated cardboard, magazines and third-class mail, newspaper, office paper,
phonebooks, textbooks, dimensional lumber, and medium-density fiberboard.
One of the large-scale processes that influences the cycling of carbon is the uptake or release of
carbon from forests. When trees are cleared for agriculture or other activities, carbon is released
(generally in the form of CO2). On the other hand, when forests are planted and allowed to continue
growing, they absorb atmospheric CO2 and store it in the form of cellulose and other materials. When the
rate of uptake exceeds the rate of release, carbon is said to be sequestered.
In the United States, uptake by forests has exceeded release since about 1977, primarily due to
forest management activities and the reforestation of previously cleared areas. This net sequestration of
carbon in forests represents a large and important process. EPA estimates that the annual net CO2 flux
(i.e., the excess of uptake minus release) in U.S. forests was about 107 million metric tons of carbon
equivalent (MMTCE) in 1999,1 offsetting about 7 percent of U.S. energy-related CO2 emissions. In
addition, about 17 million metric tons of carbon was stored in wood products.
When paper and wood products are recycled or source reduced, trees that would otherwise be
harvested are left standing. In the short term, this reduction in harvesting results in a larger quantity of
carbon remaining sequestered, because the standing trees continue to store carbon, whereas paper and
wood product manufacture and use tends to release carbon.2 In the long term, some of the short-term
benefits disappear as market forces result in less planting of new managed forests than would otherwise
occur, so that there is comparatively less forest acreage in trees that are growing rapidly (and thus
sequestering carbon rapidly).
Considering the effect of forest carbon sequestration on U.S. net GHG emissions, it was clear
that a thorough examination was warranted for this study. The complexity and long time frame of carbon
sequestration in forests, coupled with the importance of market dynamics that determine land use,
dictated the use of best available models. This chapter describes our method for applying models to
estimate the effect of forest carbon sequestration associated with paper and wood product recycling and
source reduction.
We worked with the U.S. Department of Agriculture Forest Service (USDA-FS) to use models of
the U.S. forest sector to estimate the amount of forest carbon sequestration per incremental ton of paper
and wood reduced and recycled. We used the USDA-FS system of models because (1) they are the best
models available in modeling the species composition, inventory, and growth of forests; and (2) these
models had been used previously to analyze climate change mitigation options for the Climate Change
1 U.S. EPA. 2001. Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-1999. U.S. Environmental
Protection Agency, Office of Policy, Planning and Evaluation, Washington, DC. EPA-236-R-01-001. Note that the
estimate cited (110 MMTCE) includes only carbon storage in trees and understory, which is consistent with the
forest components included in this report. If forest floor and soils were included as well, the total would be 171
MMTCE.
2 The forest carbon inventory in any year equals the carbon inventory the year before, plus net growth, less
harvests, less decay. Thus, when harvests are reduced, the inventory increases. However when inventories become
high relative to the carrying capacity of the land, the rate of growth decreases because net growth (the rate at which
growth exceeds decay) declines.
29
-------
Performance of the USDA-FS Forest Models
Researchers have never formally assessed the
accuracy of the USDA-FS models of the forest sector. In
analyses that compare the forest impacts of a policy
scenario with those of a baseline scenario (such as the
analysis described in this chapter), the USDA-FS model
results are probably reasonable. Much of the uncertainty in
the model results is due to assumptions that apply to both
the baseline and policy scenarios—assumptions about
population growth, economic growth, tree growth, and land
use changes. Any error in these assumptions would tend to
bias the results in the baseline and policy scenarios in the
same direction. Thus, when the outcomes of the baseline
and policy scenarios are compared, errors in the
assumptions tend to cancel each other out.
Action Plan. Because the models did not
enable us to estimate the forest carbon
sequestration associated with recycling of
each type of paper separately, we obtained a
single estimate of the sequestration from
recycling any type of paper.
The methodology described in this
chapter finds that increased recycling of paper
or wood products results in incremental forest
carbon sequestration of 0.73 MTCE/ton and
0.50 MTCE/ton, respectively.3 The USDA-FS
models do not directly estimate the effect of
source reduction. To derive these estimates
we evaluated the mix of virgin and recycled
inputs used to manufacture each material. As
described later, this mix is different for each
product. The resulting carbon sequestration rates range from 0.30 MTCE/ton (for corrugated cardboard)
to 0.66 MTCE/ton (for phone books).
The remainder of this chapter is divided into seven parts. Section 3.1 provides an overview of the
linkages between the five models used in the paper and wood analysis. Sections 3.2 through 3.5 describe
the models in greater detail and briefly discuss the inputs, assumptions, and outputs for each model,
focussing on the paper analysis. Section 3.6 describes the approach used to analyze wood products.
Section 3.7 presents the results of the analysis, and Section 3.8 discusses the limitations of the individual
models and of the analysis as a whole.
3.1 MODELING FRAMEWORK
Working with the USDA-FS, we used six models to estimate the impacts of increased recovery
and source reduction of paper and wood products on forest carbon sequestration.
For paper and wood products, we used five linked models to arrive at forest carbon sequestration
estimates. The first model projects the decline in U.S. pulpwood harvests when paper recovery increases.
The second and third models use the outputs of the first model, together with other inputs and
assumptions, to estimate the extent to which reduced pulpwood harvests due to paper recovery result in
lower U.S. timber harvests and increased timber inventories.4 The fourth and fifth models use the outputs
of the second and third models, and estimate how the increased timber inventories and decreased timber
harvests due to paper recovery translate into (1) increased forest carbon sequestration and (2) changes in
carbon sequestration in wood-in-use carbon sinks (e.g., wood used in home construction). Exhibit 3-1
shows how the models are linked.
The paper analysis proceeded as follows:
(1) We developed two recovery scenarios - an estimated baseline paper recovery rate for the year
2000 of 50 percent and a hypothetical year 2000 paper recovery rate of 55 percent as inputs to
3 Although a relationship exists, it is not directly measurable. Moreover, for the relationship to remain valid,
there must be continued investment in tree planting and growth. We believe this continued investment will occur,
because projections of forest product use consistently point to increases in demand.
4 The USDA-FS projections of forest product demand- account for continued high demand for all types of
forest products. •
30
-------
Exhibit 3-1
USFS Models of the Forest Sector
Macro
Economic
Data
1
NAPAP
(pulp and paper)
AREA
MODELS
(forest acreage)
TAMM
(timber harvests)
Forest
Inventory
Data
Timberland .
„„_. .|
Prices
Inventory
Growth
Removals
forest Inventory,
forest growth
parameters
I
ATLAS
(forest growth)
Carbon
accounting
Acres, inventory
growth, removals
FORCARB
(forest carbon)
All harvest
HARVCARB
(harvested carbon)
Denotes 'hand1 linkages between models/modelers requiring manipulation of data.
the North American Pulp and Paper (NAPAP) model (the model is described in Section 3.2).
The 50 percent recovery rate used for the baseline scenario was based on previous paper
industry projections.5 We used a 55 percent recovery rate for the high recovery scenario
because (1) we considered this to be a plausible recovery rate with additional government
programs to promote recycling, and (2) this recovery rate corresponded to EPA's goal of
increasing recovery of MSW in the original (1993) Climate Change Action Plan. We then
assumed that over the next 15 years, the recovery rates under both scenarios would continue
to rise and would converge in the year 2016 at 57 percent. (We assumed convergence so that
we could isolate the long-term carbon sequestration effects that might result from increasing
paper recovery in the near term.) The paper recovery rates for both scenarios then were
5 Actual paper recovery in 2000 (taken from EPA's.Municipal Solid Waste in the United States: 2000 Facts
and Figures) averaged about 53%, confirming that 50 percent is a reasonable estimate for 2000.
31
-------
projected to rise slowly from 57 percent in 2016 to 61 percent in 2040. This adjustment to the
model incorporated our assumption that the current trend of increasing paper recovery rates
would continue into the future. The NAPAP model then was run to model the pulpwood
harvests from 1985 to 2040 that would be associated with (1) the baseline paper recovery rate
and (2) the high paper recovery rate.
(2) The outputs from NAPAP for projected pulpwood harvests in the two scenarios were used as
inputs to the Timber Assessment Market Model (TAMM), which projects U.S. timber
harvests, and the Aggregate Timberland Assessment System (ATLAS) model, which projects
timber growth and changes in the U.S. forest inventory (where inventory is a function of both
growth and harvests). The TAMM and ATLAS models are described more fully in Section
3.3. The TAMM and ATLAS models were run, using the NAPAP inputs, to generate
estimates of U.S. harvest levels and forest inventories for each year through 2040, for both the
baseline and high recovery scenarios.
(3) The outputs from TAMM and ATLAS for forest harvest levels and forest inventories in the
two scenarios were used as inputs to the Forest Carbon (FORCARB) model, described in
Section 3.4, which projects forest carbon sequestration. The FORCARB model produced, as
outputs, estimates of U.S. forest carbon sequestration for each year through 2040, for both the
baseline and high recovery scenarios.
(4) FORCARB outputs also were used as inputs to the WOODCARB (also known as
HARVCARB, or Harvested Carbon) model, which tracks the flow of carbon in wood products
(see Section 3.5).
For wood products, we used essentially the same process, but bypassed step 1 by creating a
scenario involving increased recycling of wood, which causes a corresponding reduction in softwood
harvest. This harvest forecast provided the basis for inputs to ATLAS, which in turn was linked to
FORCARB and WOODCARB to evaluate carbon flows. As with paper, the increment in carbon storage
between the base case scenario and the higher recycling scenario is calculated. This increment is divided
by total tons of wood recycled to estimate a carbon storage rate (MTCE per ton of wood recycled).
3.2 THE NORTH AMERICAN PULP AND PAPER (NAPAP) MODEL
The NAPAP model is a linear optimization model6 that uses forecasts of the U.S. economy (e.g.,
growth in population and the economy) to estimate the quantity of hardwood and softwood trees
harvested for pulpwood in North America each year.7 The model predicts the quantity of pulpwood
harvested each year based on estimated demand and supply curves. The quantity harvested is the quantity
at which these curves intersect.
6 A linear optimization model begins with a set of constraints (e.g., profits = revenues - costs; costs = labor
costs + equipment costs + administrative costs + overhead costs) and an objective function (e.g., maximize profits).
The model uses principles of matrix algebra to find the solution (e.g., the total level of output) at which the objective
function is optimized (e.g., profits are maximized).
7 A number of analyses have been conducted using results from the NAPAP models. These analyses include
(1) USDA Forest Service. 1994. RPA Assessment of the Forest and Rangeland Situation in the United States -1993
Update, USDA Forest Service Forest Resource Report No. 27 (Washington, DC: USDA Forest Service), 75 pp.; (2)
Haynes, Richard W., Darius M. Adams, and John R. Mills. 1995. The 1993 RPA Timber Assessment Update, USDA
Forest Service General Technical Report RM-GTR-259 (Fort Collins, CO: Rocky Mountain Forest and Range
Experiment Station), 66 pp.; (3) Ince, Peter J. 1995. What Won't Get Harvested Where and When: The Effects of
Increased Paper Recycling on Timber Harvest, Yale School of Forestry and Environmental Studies Program on
Solid Waste Policy, Working Paper #3 (New Haven, CT: Yale University) 75 pp.; and (4) Environmental Defense
32
-------
3.2.1
•
3.2.2
Inputs to the NAPAP Model
The NAPAP model includes four major inputs:
Macroeconomic forecast data (e.g., estimates of U.S. population growth and growth in per-
capita gross domestic product);
Paper manufacturing capacity as of a baseline year;8
Manufacturing costs for each different paper manufacturing process; and
Assumed levels of future harvests from public forests.
Equations and Assumptions Used in the NAPAP Model
The NAPAP model incorporates equations for the following functions and assumptions:
Estimated pulpwood supply functions (reflecting an increasing supply of pulpwood at
increasing market prices) for three U.S. regions (West, South, and North) and two regions in
Canada;
Estimated supply functions for four principal categories of recovered paper-ntiewspaper,
corrugated boxes, mixed papers, and the aggregate of pulp substitutes and high-grade de-
inking—in each supply region (the supply functions reflect an increasing supply of recovered
paper at increasing market prices);
An unlimited supply of labor and energy at the market price in each supply region;
A fixed-quantity supply function for residues from manufacture of solid wood products, such
as lumber and plywood, mostly in the form of "pulp chips";
Demand functions9 for all 13 principal categories of paper and paperboard products produced
in North America10 (the demand functions reflect increasing demand due to population
growth and growth in the gross domestic product, and decreasing demand due to increasing
market prices);
Functions for changes in paper manufacturing capacity (including capacity for both virgin
and recycled inputs), assuming that when demand for paper increases, the investment in
paper manufacturing capacity that is needed to meet demand will be made in those types of
capacity where the ratio of profitability to capital cost is the highest; and
The ratio of the weight of paper recovered to the weight of the fiber actually used in
manufacturing new paper, after accounting for discards during processing and losses during
manufacturing.
Fund. 1995. Paper Task Force Recommendations for Purchasing and Using Environmentally Preferable Paper:
Final Report of the Paper Task Force (New York, NY: Environmental Defense Fund), 245 pp.
The baseline year for paper manufacturing capacity is 1986. The model predicts how capacity for each
paper manufacturing process changes each year from 1986 onward. The model's predictions for paper manufacturing
capacity in 1995, based on the 1986 baseline as updated, were within 5 percent of actual 1995 paper manufacturing
capacity.
9 Separate demand functions are incorporated for U.S. domestic demand, Canadian domestic demand, and
demand from various trading regions for exported paper products from the United States and Canada.
These paper grades include newspaper, coated and uncoated free sheet, coated and uncoated
groundwood, linerboard, and corrugating medium.
33
-------
The major assumptions of the NAPAP model include basic assumptions of economic analysis—
i.e., that markets are perfectly competitive and that paper manufacturers seek to maximize their profits.
Because owners of private forests may not always act to maximize their profits, NAPAP assumes that
they will continue historical patterns of economic behavior (which USDA-FS has modeled through
econometric methods). In addition, the model assumes (1) specific levels of harvests from public forests;
and (2) specific future technology options.11 Finally, the NAPAP pulpwood supply functions are the same
for both the baseline and the high recycling scenario. In other words, the supply functions do not
incorporate market feedbacks to account for changes in the age structure of forests or the acreage of
forested land. The age structure of forests could change as increased paper recovery reduces tree.
harvests, so that on average trees grow longer. Forested acreage could change if higher paper recovery
leads to decreased demand for pulpwood and lower pulpwood prices, leading some landowners to
convert forested land to farmland or ranchland.12
For this analysis, the USDA-FS simulated different recovery rates for the two scenarios^or the
year 2000,50 percent in the baseline scenario and 55 percent in the high recovery scenario. The
cumulative amounts of paper recovered under the baseline and high recovery scenarios are shown in
Exhibit 3-2.
11 The model assumes that certain technologies that existed in 1995 but were not yet commercialized (e.g.,
two newspaper processes with higher yields) would enter the commercial marketplace in the period from 1995-2000.
12 The NAPAP pulpwood supply functions incorporate projections of timber inventories over time from a
prior run of the linked TAMM and ATLAS models. Ideally, the NAPAP portion of this analysis would have used
two separate projections of timber inventories over time: one projection based on the baseline paper recovery
scenario and another based on the high paper recovery scenario. NAPAP recently has been revised so that it may
now be run iteratively with TAMM and ATLAS; however, NAPAP did not have that capability at the time this
analysis was conducted.
34
-------
Exhibit 3-2
Cumulative Paper Recovery
Under the Baseline and High Recovery Scenarios
(Million Short Tons)
Year
A. Baseline Scenario
B. High Recovery
Scenario
C. Incremental Paper
Recovery Under the
High Recovery
Scenario
-------
NAPAP scenarios generally are specified in terms of recovered fiber utilization rates, which
differ somewhat from paper recovery rates. To assure that the model inputs for fiber utilization are
consistent with paper production, recovery, and consumption projections prepared by the American
Forest and Paper Association (AFPA), Franklin Associates, Ltd. developed a set of conversion factors.
USDA-FS used these conversion factors to adjust the demand functions for paper products. The effect
was to increase the projections of paper demand and increase the estimates of the equilibrium quantity of
paper produced.13
Trade in forest products between the United States and Canada was assumed to be fixed at levels
projected in recent USDA-FS studies. As a result, any change in North American pulpwood harvests due
to increased U.S. paper recovery would be shown in the NAPAP outputs as a change in U.S. pulpwood
harvests. Thus, the forest carbon effects of increased paper recovery in the United States were modeled
as if those effects occur entirely in the United States.
3.2.3 Outputs of the NAPAP Model
The principal outputs of the NAPAP model, for each of the two scenarios modeled, are annual
U.S. pulpwood harvests from the present to the year 2040. These harvests are broken down into four
categories of pulpwood: (1) softwood roundwood; (2) softwood residues; (3) hardwood roundwood; and
(4) hardwood residues. The NAPAP estimates of pulpwood harvests for each scenario—for selected
years from 1995 to 2040—are shown in Exhibit 3-3. As the exhibit shows, the NAPAP model projected
that higher paper recovery rates until the year 2016 would result in pulpwood harvests that would be
substantially below the baseline from 1995 to 2000 (because of the recovered paper substitutes for pulp
that would otherwise be made from trees). From 2005 to 2010, the higher recovery scenario would result
in slightly higher pulpwood harvests than under the baseline.14 From 2020 onward, annual pulpwood
harvests would be the same under the baseline and high recovery scenarios (because after 2016 the paper
recovery rates would be the same in both scenarios).
13 Specifically, the USDA-FS adjusted the NAPAP model by increasing the elasticity of demand for paper
products so that it reflected the historical relationship between (1) paper demand and (2) population and per capita
gross domestic product. "Elasticity of demand" is the extent to which a change in the price of goods will affect the
quantity of the goods demanded and is defined as the percentage change in quantity divided by the percentage
change in price that induced the change in quantity. For example, if the quantity demanded goes down by 2 percent
when the price goes up by 1 percent, the elasticity of demand is -2. Specifically, this is the "own-price elasticity" of
demand because it is measured with respect to the price of the goods in question, as distinct from "cross-elasticity"
of demand, which would be measured with respect to the price of different goods.
14 Pulpwood harvests are projected to be higher between 2005 and 2010 under the high recycling scenario
These harvests are expected to be higher due to the modeled consequences of reduced pulpwood harvests before
2005. Because pulpwood harvests before 2005 are projected to be lower under the high recycling scenario, more
pulpwood remains to be harvested in later years. The increasing supply of pulpwood ready for harvest reduces
pulpwood prices, leading to modeled increases in industry demand for non-paper uses. The increased industry
demand results in slightly higher pulpwood harvests after 2005. However, it is important to note that under the
baseline scenario, pulpwood harvests are projected to decline between 2000 and 2005. This decline is because the
increase in paper recycling during this period is projected to be greater than the increase in paper consumption.
36
-------
Exhibit 3-3
U.S. Pulpwood Harvests as Predicted by the NAPAP
Model for Selected Years (Million Cubic Feet)
Year
Baseline
Scenario
High Paper
Recovery
Scenario
1995
7.152
6.982
2000
7.230
6.858
2010
7.328
7.362
2020
7.808
7.808
2030
7.989
7.989
2040
8.173
_8J73
U.S. Pulpwood Harvests as Predicted
by the NAPAP Model
9,000
s%-1/:«wS*M\ mi »•£\J»M-^
"•rj? T Wr fl* *V$f - M? fti - |W vt^P - *•*•*'
" "^^t^^1^^^^^^
6,000
Baseline Scenario
UJH viy'»ji'!- *ifl ""— ^^ PaPer Recovery Scenario
1 Mmf»,^lw^Hi\^\mH tm
»Au^uuu i\\ tB* 4*1' "^"- 4* A -*, vr
if if in rr i ihfn tfnm 1111 nftri ifn n fr
Year
37
-------
3.3 THE TIMBER ASSESSMENT MARKET MODEL (TAMM) AND THE AGGREGATE
TIMBERLAND ASSESSMENT SYSTEM (ATLAS)
TAMM and ATLAS are spatial equilibrium models.15 TAMM models U.S. timber harvests
through 2040, and ATLAS models changes in U.S. forest growth, and inventory of growing stock
volume through 2040.16 The two models are interrelated, because timber harvests depend in part on
timber inventory, and timber inventory depends in part on prior harvest levels. (This interrelationship is
shown graphically in Exhibit 3-1 with arrows going in both directions between the two models.) To
obtain consistency in the projections of the two models, an iterative process is used. TAMM outputs for
timber removals are used as inputs to ATLAS, and the resulting ATLAS outputs for forest growth and
inventory are used as inputs to TAMM. This cycle is continued until the difference in projections
between one cycle and the next has been reduced to an acceptably small amount. To reduce the costs of
modeling in this analysis, no hand linkages were made to transfer price estimates from TAMM back to
the Area Models (see Exhibit 3-1), nor to transfer timberland area estimates from the Area Models back
to ATLAS. Implicitly, the forested area was modeled as being unaffected by increased paper recovery
rates.
TAMM's estimates of timber harvests are based on four factors: (1) estimated demand for solid
forest products (such as softwood and hardwood lumber and panel products such as plywood) based on
projected macroeconomic data (e.g., growth in population and in the economy); (2) estimates of
pulpwood harvests from the NAPAP model; (3) estimates of fuelwood harvests (held constant at recent
levels); and (4) estimates of annual forest growth from ATLAS.
The ATLAS estimates of forest growth and inventory are based on (1) the previous year's
inventory, (2) timber harvests from TAMM, and (3) estimated forest growth parameters.
3.3.1 Inputs to the TAMM Model
The TAMM model is based on eight major inputs:17
• U.S. pulpwood harvests, from the NAPAP model;
15 A spatial equilibrium model is an optimization model (see footnote 6 in this chapter) that accounts for
costs of transportation of products'from producing regions to consuming regions.
16 The descriptions of the TAMM and ATLAS models are drawn from Richard W. Haynes et al. 1993.
Alternative Simulations of Forestry Scenarios Involving Carbon Sequestration Options: Investigation of Impacts on
Regional and National Timber Markets, U.S. Department of Agriculture Forest Service, Pacific Northwest Station,
August 5. Two articles which give a more detailed description of the TAMM model are (1) Adams, D.M. and R. W.
Haynes 1980 The 1980 Softwood Timber Assessment Market Model: Structure, Projections, and Policy
Simulations, Forest Science Monograph No. 22 (Washington, DC: USDA Forest Service), 62 pp., and (2) Adams,
D M and R W Haynes, A Spatial Equilibrium Model of U.S. Forest Products Markets for Long-Range Projection
and Policy Analysis. In Andersson et al., eds., "Systems Analysis in Forestry and Forest Industries," TIMS Studies
in the Management Sciences 21(1986)73-87. Two journal articles which describe analyses based on the TAMM
model are (1) Adams, D.M. and R.W. Haynes, Softwood Timber Supply and the Future of the Southern Forest
Economy, Southern Journal of Applied Forestry 15(1991):31-37, and (2) Adams, D.M and R.W. Haynes. 1991.
"Estimating the Economic Impacts of Preserving Old-Growth on Public Lands in the Pacific Northwest, The
Northwest Environmental Journal 6(2):439-441.
17 Inputs to the TAMM model are documented in: Haynes, R.W. 1990. An Analysis of the Timber Situation
in the United States: 1989-2040, Gen. Tech. Rep. RM-199 (Ft. Collins, Colorado: USDA Forest Service, Rocky
Mountain Forest and Range Experiment Station), 286 pp.
38
-------
• U.S. fuelwood harvests, from a fuelwood model; >
• Assumed levels of future timber harvests from public forests, from USDA-FS harvest plans;
• U.S. net imports of forest products, from a trade model; ! •
• Changes in U.S. forested acreage over time, from a prior run of forest AreaModels;18
• Growth in forest inventory, from the ATLAS model; ;
• Macroeconomic forecast data, e.g., on U.S. housing starts, housing repairs,!and remodeling;
and :
• Installed capacity as of 1990 for producing timber products, such as lumber or plywood,
from harvested trees.
3.3.2 Equations and Assumptions Used in the TAMM Model
The TAMM model incorporates equations for the following:
• Estimated timber product supply functions (reflecting an increasing supply .of timber
products at increasing market prices) for eight U.S. regions; and ;
• Estimated demand functions for U.S. demand for all major uses of lumber and plywood
(reflecting decreasing demand for such products at increasing market prices).
Also, changes in supply capacity for timber products are predicted by the model, based on
anticipated changes in relative regional profitability or rate of return from capital investment.19
The TAMM model includes several major assumptions:20 i
• General assumptions of competitive markets, increasing demand for wood products with
increasing economic activity, profit maximization by owners of lumber and plywood mills,
and continued historical patterns of economic behavior by owners of forest ,land (these
behavior patterns may not be strictly profit maximizing); and |
• Specific assumptions regarding particular levels of public harvests and projiected changes in
technology. . ;
In addition, TAMM and ATLAS assume (1) specified levels for net imports of softwood
products, and (2) no net imports of hardwood lumber. !
18 In the NAPAP portion of this analysis, timber inventories over time were not affected by the different
paper recovery rates in the two different scenarios analyzed, but in the TAMM and ATLAS models, timber
inventories were estimated independently for the two different scenarios. |
19 Specifically, TAMM uses an assumption that changes hi capital investment are a function of past changes
in output (i.e., that manufacturers' expectations about the profitability of capital investment are based on past
changes in output). :
20 Assumptions of the TAMM model are documented in the following two reports: (1) Haynes, R.W. 1990.
An Analysis of the Timber Situation in the United States: 1989-2040, Gen. Tech. Rep. RM-199. (Fort Collins,
Colorado: USDA Forest Service, Rocky Mountain Forest and Range Experiment Station), 286 pp'.; and (2) Haynes,
R.W., D.M. Adams, and J.R. Mills. 1995. The 1993 RPA Timber Assessment Update, Gen. Techi Rep. RM-GTR-
259 (Fort Collins, Colorado: USDA Forest Service, Rocky Mountain Forest and Range Experiment Station), 66 pp.
39
-------
3.3.3 Inputs to the ATLAS Model
The ATLAS model, for each simulation year, relies on four major inputs:
Forest inventory at the beginning of the previous period, from a prior ATLAS model run;
Forest removals during the previous period, from the TAMM model;
Changes in forest acreage, from a prior run of a modified version of the Southern Area
Model; and
State-by-state data on the number of forested acres and the volume of timber per forested
acre (shown as "Forest Inventory Data" in Exhibit 3-1).
3.3.4
Equations and Assumptions of the ATLAS Model
The ATLAS model incorporates equations that allow the model to simulate shifts in forest
management intensities and consequent changes in yields. Projected shifts in forest management
intensities are based on (1) the modeled prices of forest products, (2) the costs of various management
practices, and (3) the timber yields associated with each management practice.
The only major assumption in the ATLAS model is that owners of private forests manage their
forests at the level of intensity indicated by recent average forest planting rates. Otherwise, the model is
very simple, relying on a basic mathematical proposition that forest inventory in any period equals forest
inventory in the previous period, plus net growth, minus harvests. Net growth is gross growth less
mortality from fire, storm, insects, and disease.
3.3.5 Outputs of the TAMM/ATLAS Models
The outputs of the linked TAMM and ATLAS models are projections, through 2040, of U.S.
inventories of forest growing stock volumes (i.e., the volume of trees growing in forests), annual U.S.
sawtimber harvests, and forest growth.
We used the TAMM/ATLAS data on forest growing stock inventories as inputs to FORCARB.
Exhibit 3-4 shows the growing stock inventories of privately owned forest lands in the United States as
projected by the TAMM/ATLAS models. As the exhibit shows, forest growing stock inventories range
from 1 to 2 billion cubic feet higher under the high recovery scenario than under the baseline scenario for
the entire simulation period.
40
-------
Exhibit 3-4
Growing Stock Inventories of
Privately Owned Forest Lands in the US
As Projected by the TAMM/ATLAS Models
(Billion Cubic Feet)
Year
Baseline
Scenario
High Paper
Recovery
Scenario
1995
478
478
2000
488
489
2010
515
517
2020
532
534
2030
541
544
2040
545
548
Growing Stock Inventories of Privately
Owned Forest Lands in the US as
Projected by the TAMM/ATLAS Models
550 -rrZT*
540 -
530 '
520 - £.;'
510
500 -
490-
480
470
•- rf *t f '4^f.^^i •--'.*-?:lTO«)**v^
Hl4iJJ4^^n^^^Uf^5S*P * "
•S-i .!>s.-4ri, •c&r-i-J-IV^rJiP^'- '"• s'.
•*4 it-1; ^~ 'Kt; i -„ *; r J • tBP^' "" ' -v -
-|4fa» & i,F. L - --i/: r i?"-> '"
^|^«t-f ^sstif'FHkfP'^ 7;:,:•,>»»'"-"'V^
¥ r I*!- t-^er-'g^pr^" * -";-' r.- - :. *
«i«>«.t'HA-»^.£'.anSr-~ ^. >r-^s#-jf*7i--v
-•-High Paper
Recovery Scenario
1995 2000 2005 2010 2015 2020 2025 2030 2035 2040
Year
41
-------
3.4 THE FOREST CARBON MODEL (FORCARB)
The Forest Carbon Model (FORCARB) projects U.S. forest carbon sequestration (including soil,
forest floor, and understory carbon) each year through 2040, based on outputs from the TAMM/ATLAS
linked models.21
3.4.1 Inputs to the FORCARB Model
The major inputs to the FORCARB model are the following:
• Forest growing stock inventories—by tree species, age, and region—from the linked
TAMM/ATLAS models; and
• The percentage carbon composition for different species of trees, as grown in different forest
regions.
3.4.2 Assumptions of the FORCARB Model
The USDA-FS tracks information in TAMM/ATLAS in terms of growing stock volume, i.e., the
merchantable portion of trees. Tree volume is larger than growing stock volume, due to additional
volume in non-merchantable portions of the tree such as roots and branches. The FORCARB model uses
the simplifying assumption that tree volume is a constant multiple of growing stock volume. Carbon in
the tree volume in the U.S. forest industry then is estimated based on the percentage carbon content of
different species of trees.
When a tree is harvested, FORCARB no longer counts the carbon remaining in the non-
merchantable portion of the tree (e.g., tree roots) following harvest. In other words, FORCARB uses the
simplifying modeling assumption that the carbon in the non-merchantable portion of the tree is no longer
sequestered and is converted to CO2 emissions.
3.4.3 Outputs of the FORCARB Model
As outputs, the FORCARB model produces estimates of total U.S. forest carbon inventories and
estimates of sawtunber and pulpwood harvests for each year through 2040. The amount of forest carbon
sequestration in a given year equals the increase in forest carbon inventories during that year. If forest
carbon inventories decrease, net emissions, i.e., negative forest carbon sequestration, would occur.
Exhibit 3-5 shows the projected carbon inventories of U.S. forests, as projected by the
FORCARB model, for the baseline and high paper recovery scenarios. The forest carbon inventories that
served as the basis for these annual changes counted carbon in trees and understory (e.g., small trees), but
not carbon in the soil and forest floor. These carbon stocks were not included because of the high level of
uncertainty in estimating and modeling their carbon content.
21 The description of the FORCARB model here is drawn from Birdsey, Richard A., and Linda S. Heath.
1993. Carbon Sequestration Impacts of Alternative Forestry Scenarios - Draft (Radnor, PA: U.S. Department of
Agriculture Forest Service, Global Change Research Program), pp. 47-51. A number of studies analyzing forest
issues using the FORCARB and HARVCARB models have been published in journal articles. Among these are three
that also explain the FORCARB and HARVCARB models. These three articles are (1) Plantinga, A.J. and R.A.
Birdsey. 1993. "Carbon Fluxes Resulting from U.S. Private Timberland Management," Climatic Change 23:37-53;
(2) Heath, L.S. and R.A. Bkdsey. 1993. "Carbon Trends of Productive Temperate Forests of the Coterminous
United States," Water, Air, and Soil Pollution 70:279-293; and (3) Heath, L.S. and R.A. Bkdsey. 1993. "Impacts of
Alternative Forest Management Policies on Carbon Sequestration on U.S. Timberlands," World Resource Review
5:171-179.
42
-------
Exhibit 3-5
U.S. Forest Carbon Inventory, Trees, Understory, Soil, and Forest Floor
As Predicted by the FORCARB Model
(Million Metric Tons of Carbon)
Year
A Baseline Scenario
B. Hicih Paoer Recovery Scenario
C. Incremental Carbon Stored
Under the High Paper Recovery
Scenario CB-At
2000
8641
8.665
24
2010
9.076
9.118
42
2020
9.322
9.364
42
2030
9.442
9.480
38
2040
9.497
9.537
40
U.S. Forest Carbon Inventory
As Predicted by the FORCARB Model
(Trees, Understory, Soil, and Forest Floor)
9,800
9,600
9,400
9,200 -
9,000
8,800
8,600
8,400
8,200
8,000
I'&frfc-^iwLjjy^JSyk^jj/fl^
B Baseline Scenario
HHigh Paper Recovery
^> *I I *.f.]?>..*i.^.l..t^.^....r..^..f*1fe*.^.r..is>.A...i!.j:;S*9^s.
2000
2040
43
-------
Exhibit 3-6 shows the change in U.S. forest carbon inventories, expressed as an annual average
for decades from 2000 to 2040. Inventories increase more quickly under the high recycling scenario than
under the baseline recycling scenario, through the decade ending 2010. After 2010, the rate of increase in
forest carbon inventories is essentially the same for both scenarios. This consistency is because the paper
recovery rate is modeled as converging in 2016 to the same rate in both scenarios.
3.5 THE HARVESTED CARBON MODEL (WOODCARB)
WOODCARB (also known as the Harvested Carbon Model, or HARVCARB) can be thought of
as a spreadsheet model that projects the disposition of harvested wood across four different potential
scenarios, for 50 years into the future.22 The spreadsheet would include estimates of the percentage of
four categories of wood that will be found in four potential fates at 10-year intervals: (1) products (a
"wood-in-use" sink); (2) landfills; (3) combustion for energy; and (4) aerobic decomposition. Some
change in the fate of a wood product occurs over time: wood products that are in use in the early years
are likely to be landfilled or combusted in later years. The four different categories of wood considered in
the model are softwood and hardwood pulpwood, and softwood and hardwood sawtimber. The model has
separate fate estimates for three regions of the United States: west, south, and north.
We combined the average annual sawtimber and pulpwood harvest estimates from FORCARB
with the fate estimates in the WOODCARB spreadsheet, to obtain estimates of the amount of carbon
from harvested wood that would be found in each of the four potential fates for 50 years into the future.
3.5.1 Inputs to the WOODCARB Model
As the only input to the WOODCARB model, the USDA-FS used the annual sawtimber and
pulpwood harvests from the FORCARB model.
3.5.2 Assumptions of the WOODCARB Model
The WOODCARB model assumes that the material management patterns for the four categories
of wood over a 50-year period do not change (e.g., the model does not assume any change in the
proportion of waste or disposed wood burned for energy).
22 This USDA-FS model is an adaptation of the HARVCARB model developed earlier (C. Row, and R.B.
Phelps, 1990, "Tracing the flow of carbon through the U.S. forest products sector," Presentation at the 19th World
Congress, International Union of Forestry Organizations, 5-11 August 1190, Montreal, Quebec), and described more
fully in Row and Phelps, 1996, "Wood Carbon Flows and Storage after Timber Harvest," in Forests and Global
Change. Vol 2, R. Neil Sampson and Dwight Hair, eds. American Forests, Washington, DC, p 27-58. This
description of the USDA-FS implementation of the model is based on R.A. Birdsey and L.S. Heath, op cit, pp. 50-
51.
44
-------
Exhibit 3-6
Average Annual Change
In U.S. Forest Carbon Inventories
As Predicted by the FORCARB Model
(Million Metric Tons of Carbon)
Time Period
A. Baseline
Scenario
B. High Paper
Recovery
Scenario
C. Incremental
Annual Forest
Carbon
Sequestration in
the High Paper
Recovery
Scenario IB-A1
Decade
Ending
2000
45.48
47.89
2.40
Decade
Ending
2010
43.47
45.25
1.79
Decade
Ending
2020
24.56
24.59
0.03
Decade
Ending
2030
11.96
11.70
-0.26
Decade
Ending
2040
5.52
5.74
0.22
Average Annual Change in U.S.
Forest Carbon Inventories
2000 2010 2020 2030 2040
Decade Ending
45
-------
3.5.3 Outputs of the WOODCARB Model
In this analysis, WOODCARB provided outputs for the amount of carbon (1) retained in wood-
in-use sinks; (2) landfilled; (3) combusted for energy; and (4) aerobically decomposed, for each year
from 1995 to 2040. Because other parts of our analysis address landfills and combustion, and aerobic
decomposition has no GHG effects, we used only the estimates of the amount of carbon retained in
wood-in-use sinks (a form of carbon sequestration). We included this amount in our estimate of total
"forest carbon," even though this carbon is stored in locations outside of forests.
Exhibit 3-7 shows the wood-in-use sinks for the baseline and high recovery scenarios from 1990
to 2040, as predicted by the WOODCARB model. As shown in the exhibit, the wood-in-use sinks are
slightly less under the high recovery scenario than under the baseline scenario. The WOODCARB model
predicts this result because under the high recovery scenario, tree harvests are reduced. Under the fixed
proportions of the fates of wood assumed in WOODCARB, less wood is available for each of the fates
for wood products, including wood-in-use sinks. As noted above, WOODCARB uses fixed proportions
for the disposition of harvested wood (e.g., paper, housing, and furniture). With increased paper
recovery, wood prices would be expected to decline (due to reduced demand), and more wood probably
would be used for housing and furniture. Because WOODCARB does not account for any change in the
price of wood and its impacts on wood-in-use sinks, the values in Exhibit 3-7 are probably a slight
underestimate of the amount of carbon in wood-in-use sinks under the high recovery scenario.
3.6 APPLYING THE MODELS FOR WOOD PRODUCTS
As the preceding discussion indicates, the USDA-FS modeling system is quite complex and
requires extensive coordination between model components. The modeling of the effects of paper
recycling and source reduction was conducted over a 2-year period and involved efforts of several
experts. After publication of the first edition of this report, EPA received several requests to evaluate the
effect of recycling and source reduction of solid wood products, especially dimensional lumber and
medium-density fiberboard. For these products, the USDA-FS, EPA, and ICF Consulting conducted a
more streamlined analysis to characterize forest carbon flows.
The streamlined analysis bypassed the use of NAPAP and TAMM. Rather than creating a
market-based harvest scenario by using these models, a harvest scenario was developed based on the
expert judgment of two USDA-FS experts in forest products and carbon flows: Dr. Ken Skog of the
Forest Products Laboratory and Dr. Linda Heath of the Northeast Research Station. Dr. Skog indicated
that the majority of solid wood products are derived from softwood, and a large-scale wood recycling
program might result in a corresponding reduction in softwood harvest of about 1.7 percent. This harvest
forecast provided the basis for inputs to ATLAS, which in turn was linked to FORCARB and
WOODCARB to evaluate carbon flows.
The reductions were distributed throughout the USDA-FS regions in proportion to baseline
harvest for the period 1998-2007. The cumulative reduction in softwood harvest was 26.4 million short
tons.
The effect of this reduction in harvest is to increase carbon sequestration in forests. To be
consistent with the approach for paper recycling, effects were analyzed only for the tree and understory
components (and excluded forest floor and soils). The total carbon sequestration was converted to a rate
per increment of (a) recycling or (b) source reduction. Dr. Skog provided the following rough estimates
of the system efficiencies, on a mass basis, for producing wood products from virgin inputs or recycled
inputs:
46
-------
Exhibit 3-7
U.S. Cumulative (Since 1990) Wood-in-Use Sinks
as Predicted by the HARVCARB Model
(Million Metric Tons of Carbon)
Year
A. Baseline Scenario
B. High Paper Recovery
Scenario
C. Change in Carbon
Storage in Wood-in-Use
Sinks FB-Al
2000
733
726
-7
2010
1.216
1.208
-8
2020
1.634
1.630
-4
2030
2.Q28
2 026
-2
2040
2 381
2 379
-2
Change in Carbon Storage in Wood-in-Use Sinks
Between the Baseline and High Paper Recovery Scenarios,
as Predicted by the HARVCARB Model
(Million Metric Tons of Carbon)
2000
o I?
2010
2020
2030
2040
O
•5
o
15
c
.0
I
47
-------
• 1.1 tons roundwood input per ton finished product; and
• 1.25 tons recycled wood input per ton of finished product.
Based on these factors:
• For every ton of solid wood product that is source reduced, the reduction in timber harvest is
1.1 tons; and
• Assuming that overall demand for wood products is constant, increases in recycling reduce
timber harvest so that for every ton of solid wood product recycled, the reduction in timber
harvest is 0.88 tons (=1.1/1.25).
To develop estimates of the incremental forest carbon sequestration rates, we divided the change
in forest carbon sequestration by the rates of recycling or source reduction that correspond to the reduced
tonnages of softwood harvest. •
The final step was to estimate effects on the productpool. For wood products, we assumed a
carbon density of 0.531 tons C/ton wood (or 0.48 MTCE per short ton wood), corresponding to
softwoods in a Southeast and South Central pine forest (one of the principal sources of softwood
harvests), based on Birdsey 1992.23 Other key assumptions were the following:
• For source reduction, every ton of wood product removed from the product pool results in a
corresponding decline in carbon mass in that pool; and
• For increased recycling (i.e., at levels above the current rate), every 1 ton of wood recycled
yields 0.8 ton of product (=1/1.25). According to Dr. Skog, it is reasonable to assume that the
mass lost in the recycling process is burned. Thus the carbon loss from the product pool is (1
ton wood recycled - 0.8 ton wood retained) * 0.48 MTCE/ton wood = 0.10 MTCE/ton.
Note that the effect on the product pool from both source reduction and recycling is to decrease carbon
sequestration. This decrease offsets some of the benefit of increasing sequestration in the forest pool.
3.7 RESULTS
As noted at the beginning of this chapter, we first obtained estimates of the forest carbon
sequestration24 from paper recycling, and then used those estimates to develop estimates of the forest
carbon sequestration from source reduction of paper.
We estimated the forest carbon sequestration per ton of paper recycled at various points in the
future by dividing the cumulative difference in forest carbon between the high recovery and baseline
scenarios by the cumulative difference in the amount of paper recovered between the two scenarios. To
estimate the forest carbon sequestration in each scenario, we summed the forest carbon sequestration
estimates generated by the FORCARB model and the wood-in-use sink estimates generated by the
WOODCARB model.
23
Birdsey, Richard A. 1992. Carbon Storage and Accumulation in the United States Forest Ecosystems.
USDA Forest Service. General Technical Report WO-59. Table 1.2'
2 As noted earlier, the term forest carbon sequestration is intended to include both the carbon stored in
forests and the carbon stored in wood-in-use sinks.
48
-------
The USDA-FS projected forest carbon inventories under the baseline and high recovery
scenarios at several points in time (i.e., 2000, 2010, 2020, 2030, and 2040). The estimates of incremental
forest carbon sequestration per ton of paper recovered vary across time, as shown in Exhibit 3-8. Note
that the estimates of incremental forest carbon sequestration decline from 2000 to 2020, and then
stabilize.
We chose the forest carboasequestration factor fqr the period ending in 2010 as the best
approximation of the forest carbon benefits from increasing source reduction and recycling over the near
term. This value—0.73 1VITCE per short ton of paper recovered—^alls between the higher value for 2000
and the lower values for later years in the simulation period. We selected this value to approximate the
short-term carbon sequestration benefits of source reduction and recycling because it balances the
following: (1) relatively high carbon sequestration benefits will be achievable in the near term; (2) forest
carbon sequestration benefits drop somewhat over time; and (3) more uncertainty is associated with the
long-term carbon sequestration effects and market response (because model predictions far into the future
are more uncertain than near-term predictions). In sum, we believe that the value for the year 2010 strikes
the best balance in capturing the relatively higher short-term benefits of forest carbon sequestration and
recognizing that these benefits decline over time.25
Using the forest carbon sequestration estimate for paper recovery, we developed estimates for
forest carbon sequestration associated with source reduction of paper, as shown in Exhibit 3-9. We
estimated source reduction values under two assumptions: that source reduction displaces only virgin
inputs, and that it displaces the current mix of virgin and recycled inputs.26 We estimated that forest
carbon sequestration for source reduction, assuming displacement of vkgin inputs, is the same as for
paper recovery. Although this approach for estimating the effects of source reduction does not consider
the loss rates associated with paper recovery, we believe it is a reasonable first approximation. To
estimate the forest carbon sequestration for source reduction assuming displacement of the current mix of
inputs, we used an additional factor, i.e., the percentage of vkgin inputs in the current mix of inputs. For
this calculation (column "d" in Exhibit 3-9), we account for the fact that displacement of recycled inputs
does not have any impact on forest carbon sequestration.
25 The impact of increased paper recycling and source reduction on forest growing stock inventories (3
billion cubic feet in addition to the baseline scenario of 541 cubic feet in 2030, as shown in Exhibit 3-4) is only 0.5
percent. This amount is less than the likely statistical error in measuring the inventories. Although the estimated
effect is a small proportion of the total inventory, the relationship between recycling and stocks is clear, and the
magnitude of the effect is plausible and is significant on a per-ton basis.
26 Source reduction may conceivably displace 100 percent vkgin inputs if the quantity of paper recovered
does not change with source reduction, and all recovered paper is used to make new paper. In that case, if the
quantity of paper manufactured is reduced through source reduction, all of the reduction in inputs would come from
vkgin inputs. It is more likely, however, that source reduction reduces both vkgin and recycled inputs. ,-. ,-.;
49
-------
Exhibit 3-8
Increased Forest Carbon Storage Per Ton of Paper Recovered
Cumulative Change Between the Baseline and High
Paper Recovery Scenarios for:
A. Forest Carbon Stocks* Cmillion MTCE^
B. Wood-in-Use Stocks Cmillion MTCE)
C. Incremental Carbon Stored Cmillion MTCE) [A+B]
D. Incremental Paper Recovery Cmillion short tons')
E. Incremental Carbon Sequestration CMTCE/ton) [C/D]
2000
24.0
-7.0
17.0
19.7
0.9
2010
41.9
-8.0
33.9
46.2
0.7
2020
422
-4.0
_3JL2.
81 4
0.5
2030
397
-20
37.7
81.4
0.5
2040
41.9
-2.0
39.9
81.4
0.5
Includes trees and understory.
Increased Forest Carbon Sequestration Per Ton of Paper
Recovered
2000
2040
50
-------
Exhibit 3-9
Forest Carbon Sequestration Per Ton of Paper Product Recycled or Source Reduced
(a)
Material
Corrugated
Cardboard
Magazines/
Third-class
Mail
Newspaper
Office Paper
Phonebooks
Textbooks
(b)
Recycling,
Recovering 1
Incremental
Ton of Paper
(MTCE).
0.73
0.73
0.73
0.73
0.73
0.73
(c)
Source
Reduction,
Assuming
Displacement of 1
Ton of Paper
Made from the
Virgin Inputs
(MTCE)
0.73
0.73
0.73
0.73
0.73
0.73 •
(d)
Percent
Virgin
Inputs in
the Current
Mix of
Inputs
41%
84%
49%
67%
89%
87%
(e)
(e = b * d)
Source Reduction,
Assuming Displacement of
One Ton of Paper Made
From the Current Mix of
Virgin and Recycled Inputs
(MTCE)
-0.30
-0.61
-0.36
-0.49
-0.66
-0.64
Exhibit 3-10 displays the results of the analysis for dimensional lumber and medium-density
fiberboard (the results are the same for both of these wood products). As shown in the top of the exhibit,
the ratio of carbon stored per ton of reduced harvest is 0.99 MTCE/metric ton wood for 2010. Using the
system efficiencies for wood products conversion rates and expressing emission factors in MTCE per
short ton, the effects on the forest pool as of 2010 are the following: .
• Recycling: 0.79 MTCE/ton '
• Source reduction: 0.98 MTCE/ton
As noted earlier, recycling and source reduction would reduce the amount of carbon in the wood products
pool; this effect is shown in the middle section of Exhibit 3-10. The bottom section shows the effect of
summing the increase in forest carbon and the decrease in product pool carbon. Using 2010 as the most
relevant period, the results are the following:
• Recycling: 0.69 MTCE/ton
• Source reduction: 0.50 MTCE/ton
Recycling has higher net carbon storage. Although it has a lower rate of forest carbon sequestration than
source reduction, it also has a much smaller decrement in carbon storage in the product pool.
51
-------
Exhibit 3-10
Increase in Forest Carbon per Unit Chanae in Harvest
Emission factor — increase in forest carbon (trees + understorv) per unit change in harvest
Based on density and carbon content of softwoods in Southeast & South Central Region, Pine
; Forest Type
2000 2010 2020 2030 2040
Per ton reduced wood harvest (MT/MT)
Per ton reduced carbon harvest (MT/MT)
Per ton increased recvclina (MT/short ton)
Per ton reduced prodn of solid wood product
(source reduction) (MT/short -ton) ,
0.96
1.81
0.77
0.96
0.99
1.86
0.79
0.98
0.99
1.87
0.79
0.99
0.99
1.86
0.79
0.98
0.97
1.83
0.78
0.97
Emission factor — change in product pool carbon per unit change in recycling or source
Based on carbon content of softwoods in Southeast & South Central Region. Pine Forest Type
2000 2010 2020 2030 2040
Per ton increased recvclina (MT/short ton)
Per ton reduced prodn of solid wood product
(source reduction) (MT/short ton)
-0.10
-0.48
-0.10
-0.48
-0.10
-0.48
-0.10
-0.48
-0.10
-0.48
Emission factor — increase in forest + product pool carbon per unit change in recycling or
Based on carbon content of softwoods in Southeast & South Central Region, Pine Forest Type
2000 2010 2020 2030 2040
Per ton increased recyclina (MT/short ton)
Per ton reduced prodrj of solid wood product
(source reduction) (MT/short ton)
0.67
0.48
0.69
0.50
0.70
0.51
0.69
0.50
0.68
. 0.49
3.8 LIMITATIONS
Any analysis based on a complex system of models is subject to the limitations introduced by
each model in the system. The limitations of each component model derive from (1) the assumptions
made in developing the model; (2)'the input equations used in the model; and (3) the potential impact of
factors not included in the model. Because of these.limitations, the actual behavior of markets for paper,
wood, and other forest products (and the actual choices made by owners of private forestland) could
differ from those predicted by the system of forest models. We believe that most of these limitations
would tend to bias estimates under the baseline and high recycling scenarios in the same direction, so that
the estimated differences between the two scenarios should be relatively accurate. Some limitations could
result in unequal bias in the estimates, however, leading to biased estimates of the differences.
This section first discusses limitations associated with the geographic scope of the analysis.
Secondly, we discuss limitations that could bias the estimates. Limitations that could bias both scenarios
in the same direction are listed next. This section concludes with a brief discussion of the uncertainties
introduced by the choice of a time period over which incremental forest carbon sequestration is
estimated. . .
52
-------
3.8.1 Limitations of Geographic Scope of Analysis and Results
Although the goal of this analysis is to estimate the impact of paper recycling and source
reduction on GHG emissions in the United States, the actual effects would occur in Canada and other
countries as well. . . ' . « . ' • . -
• The USDA-FS models treat forest product markets in the United States and Canada as a
single integrated economic and biological system. But they do not treat Canadian forest
inventories in the same way and degree of detail as U.S. forest inventories. The estimated
impacts of increases in recycling and source reduction were treated as impacts on U.S.
forests. Because much of the economically marginal paper production is from Canadian pulp
sources, source reduction, in particular, would lower demand for Canadian timber. In any
ease, U.S. and Canadian forests actually would share the effects.
• More than 20 percent of the paper currently recovered in the two countries is exported. Some
proportion of the increased, amounts of recycled paper—probably more than 20 percent—
would undoubtedly be exported. Current exports comprise 43 percent of the world trade in
recovered paper. The major buyers of this paper are developing countries in Asia and Latiti
America, with Korea, Taiwan, and Mexico being major destinations. The alternative sources
of fiber for the paper industry in these countries are pulp and fiber from non-forest sources
(agricultural refuse, hemp, bamboo, and rubber and palm oil trees). Very little comes from
forest harvests. Forests in these countries, however, are not necessarily managed on a
sustainable basis. It'is difficult to determine which of these effects would dominate—
displacement of non-forest fiber (with no forest carbon impact) or displacement of
unsustainably managed forest fiber (with a benefit larger than that in U.S. forests).27
• NAPAP does not account for any effects of lower pulpwood prices (due to higher paper
recycling rates) on net exports of U.S. pulpwood to non-Canadian markets. Lower pulpwood
prices would be expected to result in increased exports and possibly changes in foreign
timber inventories. Though U.S. pulpwood exports are currently less than 1 percent of U.S.
pulpwood production, some virgin pulp fiber is now being exported from southern and
western ports in the form of pulp chips. The future potential for pulp chip exports is difficult
to estimate.
• The competition to U.S. and Canadian exports of both recovered and newly manufactured
paper is likely to come from two sources. First, all other developed countries are also likely
to intensify recycling and source reduction programs, with additional recovery of paper fiber.
Second, a major developing source of fiber for paper is the establishment of intensive forest
plantations in tropical and southern hemisphere countries, particularly Australia, Brazil,
Chile, Indonesia, New Zealand, and South Africa. The effect of additional world sources of
paper fiber from developed countries on these forest plantation programs is difficult to
estimate.
3.8.2 Limitations Expected to Bias the Results
Two limitations in the system of forest sector models could result in biased estimates of the
incremental forest carbon sequestration from increased paper recycling. The limitations are as follows:
27 A comprehensive description of the world paper industry, its fiber sources, and environmental concerns
can be found in International Institute for Environment and Development (IIED), 1996, Towards a Sustainable
Paper Cycle, IIED: London, 258 pp. This study, prepared for the World Business Council for Sustainable
Development, treats many of the issues covered in this chapter, but on a global basis.
53
-------
• The modeling system does not account for any conversion of U.S. forestland to farmland or
rangeland that might occur in response to lower prices for pulpwood due to higher paper
recycling rates. The NAPAP model does not account for potential changes in timber
inventory in the near term due to lower harvests associated with higher paper recovery. Nor
does it account for potential changes in forest acreage in the longer term if higher paper
recovery depresses pulpwood prices enough to induce landowners to convert forested
acreage to other uses. The TAMM and ATLAS models likewise do not allow for long-term
changes in forested acreage due to increased paper recovery. These effects, however, may be
small. Converting forestland to agriculture or to industrial, commercial, or residential uses is
far more likely to result from much higher land values for crops or development, if the land
is suitable or in a favorable location.
• This analysis did not consider carbon storage in forest soils and forest floors, because of the
high level of uncertainty in projecting changes in carbon storage. Nonetheless, projections of
carbon storage in forest soils and floors under the baseline and high recycling scenarios, as
generated by the FORCARB model, suggest that incremental carbon storage under the high
recycling scenario could be slightly higher than shown here, if storage in soils and the forest
floor were included.
3.8.3 Limitations Not Expected to Bias the Results
We expect that several limitations in the system of forest models would bias—to about the same
extent—the estimates of forest carbon sequestration in the baseline and high recycling scenarios. The
limitations thus would not result in significant bias in the estimate of the difference in forest carbon
sequestration between the two scenarios. These limitations are as follows:
• The macroeconomic forecasts used in the models (e.g., for population growth and growth in
per-capita gross domestic product) are simply forecasts, and may turn out to be inaccurate;
• The historical supply and demand functions used in the models may change in the future. For
example, (1) demand for newspaper may drpp sharply due to competition from electronic
news media, or (2) improved technologies or tree diseases not anticipated in the models may
significantly change the cost of producing forest products; and
• Future harvests from public forestland may be different from those projected.
3.8.4 The Use of a Point Estimate for Forest Carbon Sequestration
As shown in Exhibit 3-8, estimates of forest carbon sequestration due to increased paper
recycling vary over time. As noted above, in choosing a single point estimate, we selected the time period
that best balances the competing criteria of (1) capturing the long-term forest carbon sequestration
effects, and (2) limiting the uncertainty inherent in projections made well into the future. The range of
forest carbon sequestration estimates over time, and the limitations of the analysis discussed above,
indicate that there is considerable uncertainty in the point estimate selected. In comparison to the
estimates of other types of GHG emissions and sinks developed in other parts of this analysis, the
magnitude of forest carbon sequestration is relatively high. Based on these forest carbon sequestration
estimates, source reduction and recycling of paper are found to have substantial net GHG reductions.
Because paper products comprise the largest share of municipal waste generation (and the largest
volumes of waste managed through recycling, landfilling, and combustion), it is important to bear in
mind the uncertainty in the forest carbon sequestration values when evaluating the results of this report.
54
-------
4. SOURCE REDUCTION AND RECYCLING
This chapter presents estimates of GHG emissions and carbon sequestration resulting from
source reduction and recycling of 15 manufactured materials: aluminum cans, steel cans, glass
containers, plastic containers (LDPE, HDPE, and PET), corrugated boxes, magazines'/third-class mail,
newspaper, office paper, phonebooks, textbooks, dimensional lumber, medium-density fiberboard, and
mixed paper. '
To estimate GHG emissions associated with source reduction and recycling (and other MSW
management options), we used a baseline scenario in which the material is manufactured from the current
mix of virgin and recycled inputs, but has not yet been disposed of or recycled. Thus, the baseline for
each material.already incorporates some emissions from raw materials acquisition and manufacturing
using the current mix of virgin and recycled inputs. Using this measurement convention, it follows that
source reduction1 reduces GHG emissions from the raw material acquisition and manufacturing phase of
the life cycle for all materials. Moreover, source reduction of paper results in forest carbon sequestration
(as discussed in Chapter 3).
Manufacturing from recycled inputs generally requires less energy, and thus lower GHG
emissions, than manufacturing from virgin inputs. Our recycling analysis indicates that recycling reduces
GHG emissions for each of the materials studied. , •
4.1 GHG IMPLICATIONS OF SOURCE REDUCTION
When a material is source reduced (i.e., less of the material is made), the GHG emissions
associated with making the material and managing the post-consumer waste are avoided. As discussed
above, under the measurement convention used in this analysis, source reduction has (1) negative raw
material and manufacturing GHG emissions (i.e., it avoids baseline emissions attributable to current
production); (2) forest carbon sequestration benefits for" paper products (also treated as negative
emissions, as estimated in Chapter 3); and (3) zero waste management GHG emissions. Exhibit 4-1
presents the GHG implications of source reduction. The values for forest carbon sequestration were
copied from Exhibit 3-8.
In order to compare source reduction to other solid waste management alternatives, we compared
the GHG reductions from source reduction to the life-cycle GHG emissions of another solid waste
management option (e.g., landfilling). This approach enables policy makers to evaluate, on a per-ton
basis, the overall difference in GHG emissions between (1) source reducing 1 ton of material and (2)
manufacturing and then managing (post-consumer) 1 ton of the same material. Such comparisons are
made in the Executive Summary and in Chapter 8 of this report. For most materials, source reduction has
lower GHG emissions than the other waste management options.
In this analysis, the values reported for source reduction apply to material lightweighting or extension of a
product's useful life. We assume no substitution by another material or product, and thus we assume no offsetting
GHG emissions from another material or product. Thus, the data do not directly indicate GHG effects of source
reduction that involves material substitution. Considerations for estimating the GHG effects of material substitution
are presented in Section 4.3 below.
55
-------
Exhibit 4-1
GHG Emissions for Source Reduction
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed MSW
Avoided GHG Emissions from Raw?
Materials Acquistion and Manufacturing
For Current
Mix of Inputs
-2.49
-0.79
-0.14
-0.49
-0.61
-0.49
-0.24
-0.46
-0.46
-0.31
-0.64
-0.59
-0.05
-0.10
NA
NA
NA
NA
For 1 00%
Virgin Inputs
-4.67
-1.01
-0.16
-0.53
-0:64
-0.58
-0.22
-0.46
-0,59
-0.28
-0.67
-0'.59
-0.05
-0.10
NA
NA
NA
NA
Post-consumer
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
NA
NA
NA
Changes in Forest Carbon Storage
For Current
Mix of Inputs
0.00
0.00
0.00
O'.OO
0.00
0.00
-0.28
-0.58
-0.35
-0.50
-0.65
-0.64
-0.50
-0.50
NA
NA
NA
NA
For 100%
Virgin Inputs
0.00
0.00
0.00
0.00
0.00
0.00
-0.73
-0.73
-0.73
-0.73
-0.73
-0.73
-0.50
-0.50
NA
NA
NA
NA
Net Emissions
For Current
Mix of Inputs
-2.49
-0.79
-0.14
-0.49
-0.61
-0.49
-0.5t
-1.04
-0.81
-0.80
-1.28
-1.23
-0.55
-0.60
NA
NA
NA
NA
Net Emissions
For 100%
Virgin Inputs
-4.67
-1.01
-0.16
-0.53
-0.64
-0.58
-0.96
-1.19
-1.32
-1.01
-1.40
-1.32
-0.55
-0.60
NA
NA
NA
NA
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
56
-------
4.2 GHG IMPLICATIONS OF RECYCLING
When a material is recycled, it is used in place of virgin inputs in the manufacturing process,
rather than being disposed of and managed as waste.2 As with source reduction of paper products,
recycling of paper also results in forest carbon sequestration.
Most of the materials considered in this analysis are modeled as being recycled in a, closed loop
(e.g., newspaper is recycled into new newspaper). However, several paper types are recycled in an open
loop (i.e., they are recycled into more than one product) under the general heading of mixed paper.3
Mixed paper is included because it is recycled in large quantities, and is an important class of scrap
material in many recycling programs. However, presenting a single definition of mixed paper is difficult
because each mill using recovered paper defines its own supply, which varies with the availability and
price of different grades of paper.
For the purpose of this report, we identified three definitions for mixed paper: broad, office, and
residential. To assist recyclers in determining which definition corresponds most closely to mixed paper
streams they manage, the composition of each is presented in Exhibit 4-2. The broad definition of mixed
paper includes almost all printing-writing paper, folding boxes, and most paper packaging. Mixed paper
from offices includes copier and printer paper, stationary and envelopes, and commercial printing. The
typical mix of papers from residential curbside pick-up includes high-grade office paper, magazines,
catalogs, commercial printing, folding cartons, and a small amount of old corrugated containers. Mixed
paper as characterized by the broad and residential definitions can be remanufactured via an open loop
into recycled box-board. Mixed paper from offices is typically used to manufacture commercial paper
towels.
When any material is recovered for recycling, some portion of the recovered material is
unsuitable for use as a recycled input. This portion is, discarded either in the recovery stage or in the
remanufacturing stage. Consequently, less than 1 ton of new material generally is made from 1 ton of
recovered materials. Material losses are quantified and translated into loss rates. In this analysis, we used
estimates of loss rates provided by Franklin Associates, Ltd. (FAL) for steel, dimensional lumber, and
medium-density fiberboard (the same materials for which we used FAL's energy data, as described in
Chapter 2). EPA's Office of Research and Development (ORD) provided loss rates for the other
materials. These values are shown in Exhibit 4-3.
GHG emission reductions associated with remanufacture using recycled inputs are calculated by
taking the difference between (1) the GHG emissions from manufacturing'a material from 100 percent
recycled inputs, and (2) the GHG emissions from manufacturing an equivalent amount of the material
(accounting for loss rates) from 100 percent virgin inputs.
2 Note that when paper is manufactured from recycled inputs, the amount of paper sludge produced is
greater than when paper is made from virgin inputs. This difference is because recycled paper has more short fibers,
which must be screened out. We made a preliminary estimate of the GHG emissions from paper sludge managed in
landfills; our results indicated that net GHG emissions (i.e., CELt emissions minus carbon sequestration) were close to
zero. Because the emissions are small and highly uncertain, no quantitative estimate is included in this report.
3 This report also includes estimates for mixed MSW, mixed plastics, mixed organics, and mixed
recyclables, i.e., a mixture of the principal paper, metal, and plastic materials that are recycled. These other mixed
materials are discussed in Chapter 8.
57
-------
Exhibit 4-2
Summary of Mixed Paper Scenarios
(Composition as a percentage of total)
Paper Grade
Uncoated groundwood paper
Coated free sheet paper
Coated groundwood paper
Uncoated free sheet paper
Cotton fiber paper
Bleached bristols
Newspaper
Virgin corrugated boxes
Recycled corrugated boxes
Unbleached kraft folding boxes
Bleached kraft folding boxes
Recycled folding boxes
Bleached bags and sacks
Unbleached bags and sacks
Unbleached wrapping paper
Converting paper
Special industrial paper
Other paperboard
Paper plates and cups
Tissue, towels
Set-up boxes
Other paper packaging
Totals
All Paper and
Paperboard in
MSW(l)
4.9%
5.0%
4.3%
14.3%
0.1%
1.5%
13.3%
29.6%
6.8%
1.5%
2.8%
3.0%
0.4%
2.1%
0.1%
0.3% ,
1.3%
2.5%
1.2%
3.9%
0.3%
0.8%
100.0%
Mixed Paper:
Broad Definition (2)
4.9%
12.0%
11.5%
37.6%
0.4%
3.9%
2.9%
5.7%
5.7%
7.9%
1.0%
5.6%
0.2%
0.7%
100.0%
Mixed Paper:
Offices (3)
7.9%
13.9%
30.7%
41.6%
1.8%
4.1%
100.0%
Mixed Paper:
Single-Family
Residential (4)
2.2%
11.5%
17.7%
18.4%
0.2%
2.8%
2.9%
12.2%
2.8%
4.1%
5.8%
8.0%
1.6%
9.0%
0.6%
100.0%
(1) All grades of paper and paperboard in MSW.
(2) Excludes newspaper, old corrugated containers, tissue produce, paper plates and cups, converting and special
industrial papers, non-packaging paperboard such as album covers and posterboard, and paper labels.
(3) Includes the high-grade papers (ledger and computer printout) as well as stationery, mail, magazines, and
manila folders. Could be recovered as "File Stock."
(4) Represents a typical collection of mixed paper from a single-family curbside program. Includes printing-
writing papers, corrugated boxes, folding cartons, and bags and sacks.
Source: Working papers prepared by Franklin Associates, Ltd., October 1997.
58
-------
Exhibit 4-3
Loss Rates For Recovered Materials
(a)
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
(b)
Data Source
FAL & ORD *
FAL
FAL & ORD *
FAL & ORD *
FAL & ORD *
FAL & ORD *
FAL & ORD*
FAL & ORD *
FAL & ORD *
FAL & ORD *
FAL & ORD *
FAL & ORD *
FAL
FAL
(c)
Percent of
Recovered
Materials
Retained in the
Recovery Stage
100
100
90
90
90
90
100
, 95
95
91
95
95
88
88
(d)
Product Made
per Ton of
Recycled
Inputs In the
Manufacturing
Stage
0.93
0.98
0.98
0.86
0.86
0.86
0.93
0.71
0.94
0.66
0.71
0.69
0.91
0.91
(e)
(e = c * d)
Tons of Product
Made Per Ton
Recovered
Materials
0.93
0.98
0.88
0.78
0.78
0.78
0.93
0.67
0.90
0.60
0.68
0.66
0.80
0.80
* FAL provided data for column (c), while ORD provided data for column (d). •.,...
Explanatory notes: The value in column "b" accounts for losses such as recovered newspapers that were
unsuitable for recycling because they were too wet. Column "c" reflects process waste losses at the manufacturing
plant or mill. Column "d" is the product of the values in Columns "b" and "c."
The results of our analysis are shown in Exhibit 4-4. In this exhibit, for each material we present
the differences between manufacture from virgin and recycled inputs for (1) energy-related GHG
emissions (both in manufacturing processes and transportation), (2) process non-energy-related GHG
emissions, and (3) forest carbon sequestration. Our method of accounting for loss rates yields estimates
of GHG emissions on the basis of metric tons of carbon equivalent (MTCE) per short ton of material
collected for recycling (rather than emissions per ton of material made with recycled inputs).
We recognize that some readers may find it more useful to evaluate recycling in terms of tons of
recyclables as marketed rather than tons of materials collected. To adjust the emission factors reported in
Exhibit 4-4 for that purpose, one would scale up the recycled input credits shown in columns "b" and
"d" of that exhibit by the ratio of manufacturing loss rate to total loss rate (i.e., Exhibit 4-3 column "c"
divided by column "d").
Another way that recycling projects can be measured is in terms of changes in recycled content
of products. To evaluate the effects of such projects, one could use the following algorithm:4
4 This approach would apply only where the products with recycled content involve the same "recycling
loop" as the ones on which the values in this report are based (e.g., aluminum cans are recycled in a closed loop into
more aluminum cans).
59
-------
Trecyo = Tprod*(RCp-RCi)/L, where
Trecyc = tons of material recycled, as collected
Tprod = tons of the product with recycled content
RCP = recycled content (in percent) after implementation of the project
RQ = recycled content (in percent) initially
L = loss rate (from Exhibit 4-3, column " d")
Then, one could use the emission factors in this report directly with the tons of material recycled
(as collected) to estimate GHG emissions.
In order to compare GHG emissions from recycling to those attributable to another solid waste
management option such as landfilling, we compared the total GHG emissions from recycling the
material to the GHG emissions from managing the disposal of the same material under another waste
management option. The baseline for a given material (which includes GHG emissions from raw
materials acquisition and manufacturing for the current mix of virgin and recycled inputs) for both
options is the same. Overall, because recycling reduces the amount of energy required to manufacture
materials (as compared to manufacture with virgin inputs) and leads to avoided process non-energy GHG
emissions, recycling has lower GHG emissions than all other waste management options except for
source reduction.
4.3 SOURCE REDUCTION WITH MATERIAL SUBSTITUTION
As noted above, our analysis of source reduction is based on an assumption that source reduction
is achieved by practices such as lightweighting, double-sided copying, and material reuse. However, it is
also possible to source reduce one type of material by substituting another material. Analyzing the GHG
impacts of this type of source reduction becomes more complicated. Essentially, one would need to
estimate the net GHG impacts of (1) source reduction of the original material, and (2) manufacture of the
substitute material and its disposal fate. A quantitative analysis of source reduction with material
substitution was beyond the scope of this report because of the large number of materials that could be
substituted for the materials analyzed in this report (including composite materials, e.g., a composite of
paper and plastic used in juice boxes), and the need for application-specific data. Where both the original
material and the substitute material are addressed in this report, however, the GHG impacts of source
reduction with material substitution naay be estimated.
The estimate would be based on (1) the data provided in this report for the material that is source
reduced; (2) the mass substitution rate for the material that is substituted; and (3) data in this report for
the material substituted. The mass substitution rate is the number of tons of substitute material used per
ton of original material source reduced. Note, however, that in calculating the mass substitution rate, one
should account for any difference in the number of times that a product made from the original material
is used prior to waste management, compared to the number of times a product made from the substitute
material will be used prior to waste management.
To estimate the GHG impacts of source reduction with material substitution (per ton of material
source reduced), one should consider the following: a specific baseline scenario, including waste
management; an alternative scenario, involving the substitute material and a waste management method;
the number of tons of material used in each scenario, using the mass substitution rate; the net GHG
emissions for the baseline; the GHG impacts of source reduction of the original material; the GHG
impacts of manufacturing the substitute material; and the GHG impacts of waste management for the
substitute material. Among other factors, these considerations will allow for a comparison of net GHG
emissions from source reduction with material substitution to the baseline.
60
-------
4.4 LIMITATIONS
Because the data presented in this chapter were developed earlier in Chapters 2 and 3, the
limitations discussed in those chapters also apply to the values presented here. Five other limitations are
as follows:
• There may be GHG impacts from disposal of industrial wastes, particularly paper sludge at
paper mills. Because of the complexity of analyzing these second-order effects and the lack
of data, we did not include them in our estimates. We did perform a screening analysis for
paper sludge, however, based on (1) data on sludge generation rates and sludge composition
(i.e., percentage of cellulose, hemicellulose, lignin, etc. in sludge),5 and (2) professional
judgment on the CHU generation rates for cellulose, etc. The screening analysis indicated that
net GHG emissions (CHU emissions minus carbon storage) from paper sludge are probably on
the order of 0.00 MTCE per ton of paper made from virgin inputs to 0.01 MTCE per ton for
recycled inputs. Our worst case bounding assumptions indicated maximum possible net GHG
emissions ranging from 0.03 to 0.11 MTCE per ton of paper (depending on the type of paper
and whether virgin or recycled inputs are used).
• The recycling results are reported in terms of GHG emissions per ton of material collected
for recycling. Thus, the emission factors incorporate assumptions on loss of material through
collection, sorting, and remanufacturing. There is uncertainty in the loss rates: some
materials recovery facilities and manufacturing processes may recover or use recycled
materials more or less efficiently than estimated here.
• The models used to evaluate forest carbon sequestration and those used to evaluate energy
and non-energy emissions differ in their methods for accounting for loss rates. Although one
can directly adjust the emission factors reported here for process emissions so that they apply
to tons of materials as marketed (rather than tons as collected), there is no straightforward
way to adjust the forest carbon estimate.
• Because our modeling approach assumes closed-loop recycling for all materials except
mixed paper, it does not fully reflect the prevalence and diversity of open-loop recycling.
Most of the materials in our analysis are recycled into a variety of manufactured products,
not just into the original material. Resource limitations prevent an exhaustive analysis of all
the recycling possibilities for each of the materials analyzed.
• For the purpose of simplicity, we assumed that increased recycling does not change overall
demand for products. In other words, we assumed that each incremental ton of recycled
inputs would displace virgin inputs in the manufacturing sector. In reality, there may be a
relationship between recycling and demand for products with recycled content, since these
products become cheaper as the supply of recycled materials increases.
5ICF Consulting. 1996. Memorandum to EPA Office of Solid Waste, "Methane Generation from Paper
Sludge," December.
61
-------
Exhibit 4-4
GHG Emissions for Recycling
(MTCE/Ton of Material Recovered)
(a)
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
(b)
Recycled Input
Credit*:
Process
Energy
-2.92
-0.48
-0.03
-0.34
-0.43
-0.40
0.04
0.00
-0.21
0.06
-0.18
-0.01
0.02
0.01
0.08
0.08
-0.08
(c)
Recycled Input
Credit*:
Transportation
Energy
-0.14
-0.01
0.00
0.00
0.00
0.00
-0.01
0.00
-0.01
0.00
0.00
0.00
0.00
0.00
-0.02
-0.02
-0.02
(d)
Recycled Input
Credit*:
Process Non-
Energy
-1.05
0.00
-0.04
-0.04
-0.04
-0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
(e)
Forest Carbon
Sequestration
0.00
0.00
0.00
0.00
0.00
0.00
-0.73
-0.73
-0.73
-0.73
-0.73
-0.73
-0.69
-0.69
-0.73
-0.73
-0.73
(0
GHG Reductions
From Using
Recycled Inputs
Instead of
Virgin Inputs
-4.11
-0.49
-0.08
-0.38
-0.47
-0.42
-0.71
-0.74
-0.95
-0.68
-0.91
-0.75
-0.67
-0.67
-0.67
-0.67
-0.83
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
Material that is recycled after use is then substituted for virgin inputs in the production of new products. This credit represents the difference in emissions that
results from using recycled inputs rather than virgin inputs, The credit accounts for loss rates in collection, processing, and remanufacturing. Recycling credit is
based on a weighted average of closed- and open-loop recycling for mixed paper. All other estimates are for closed-loop recycling.
62
-------
Explanatory notes for Exhibit 4-4: Columns "b" and "c" show the reduction in process energy GHGs and transportation energy GHGs from making
each material from recycled inputs, rather than virgin inputs. The values in columns "b" and "c" are based on (1) the difference in energy-related GHG emissions
between making 1 ton of the material from 100% virgin inputs and from 100% recycled inputs, multiplied by (2) the estimated tons of material manufactured from
1 ton of material recovered, after accounting for loss rates in the recovery and remanufacturing stages. We first estimated the values in columns "b" and "c"
based on data provided by FAL and ORD, as shown hi Exhibits 2-2 through 2-5. Note that for two of the mixed paper definitions, the process energy GHG
emissions are higher when using recycled inputs than when using virgin inputs (as shown by positive values in column "b"). This differemce is because the
manufacture of boxboard (the product of open-loop recycling of these types of mixed paper) from virgin inputs uses a high proportion of biomass fuels, and the
biogenic CO2 emissions from biomass fuels are not counted as GHG emissions (see the discussion of biogenic CO2 emissions in Chapter 1). Still, because of
forest carbon sequestration, the net GHG emissions from recycling these two mixed paper definitions are negative.
For column " d, " which presents the process non-energy GHG emissions from recycling, we used (1) data showing the difference in process non-energy
GHG emissions between'making 1 ton of the material from 100% virgin inputs and from 100% recycled inputs (as shown in the second-to-last column of Exhibits
2-2 and 2-4) multiplied by (2) the estimated amount of material manufactured (hi tons) from 1 ton of material recovered, after accounting for loss rates in the
recovery and remanufacturing steps.
Next, column "e " shows the estimated forest carbon sequestration from recycling of paper products, as estimated in Chapter 3. The last column (column
"f') sums columns "b" through "e" to show the GHG implications of recycling each material.
63
-------
-------
5. COMPOSTING
This chapter has been extensively revised since the first edition of this report. The revised
chapter presents the results of an in-depth analysis to determine the net GHG impacts of composting yard
trimmings and food discards. As research in the areas of erosion control, soil fertility, and bio-based
products continues, we are likely to uncover additional GHG and other benefits of composting.
This chapter presents estimates of GHG emissions and sinks from composting yard trimmings
and food discards.1 The chapter is organized as follows:
Section 5.1 presents an estimate of potential CO2 and CF^ emissions from composting;
Section 5.2 quantifies the potential carbon storage benefits of applying compost to soils;
Section 5.3 presents net GHG emissions from composting; and
Section 5.4 discusses the limitations of this analysis.
Composting may result in (1) CEL* emissions from anaerobic decomposition; (2) long-term
carbon storage in the form of undecomposed carbon compounds; and (3) non-biogenic CO2 emissions
from collection and transportation of the organic materials to the central composting site, and from
mechanical turning of the compost pile.2 Composting also results in biogenic CO2 emissions associated
with decomposition, both during the composting process and after the compost is added to the soil.
Because this CO2 is biogenic in origin, however, it is not counted as a GHG in the Inventory of U.S.
Greenhouse Gas Emissions and Sinks3 (as explained in Section 1.4.2) and is not included in our
accounting of emissions and sinks.
Our analysis suggests that composting, when managed properly, does not generate
emissions, but it does result in some carbon storage (associated with application of compost to soils), as
well as minimal CO2 emissions from transportation and mechanical turning of the compost piles. In order
to maintain consistency with other chapters in this report, we selected point estimates from the range of
emission factors — covering various compost application rates and time periods — developed in our
analysis. The point estimates were chosen based on a "typical" compost application rate of 20 tons of
compost per acre, averaged over three soil-crop scenarios. In terms of timing, the carbon storage values
for the year 2010 were selected to be consistent with forest carbon storage estimates presented in Chapter
4 of this report. Overall, we estimate that centralized composting of organics results in net GHG storage
of 0.05 MTCE/wet ton of organic inputs composted and applied to agricultural soil.
1 Although paper and mixed MSW can be composted, we did not analyze the GHG implications of
composting them because of time and resource constraints.
2 CO2 emissions from delivery of compost to its final destination were not counted because compost is a
marketable product, and CO2 emissions from transportation of other marketable, finished goods to consumers have
not been counted in other parts of this analysis.
3 U.S. EPA. 2001. Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-1999. U.S. Environmental
Protection Agency, Office of Policy, Planning and Evaluation, Washington, DC. EPA-236-R-01-001.
65
-------
5.1 POTENTIAL GREENHOUSE GAS EMISSIONS
Two potential types of GHG emissions are associated with composting: (1) CHU from anaerobic
decomposition; and (2) non-biogenic CO2 from transportation of compostable materials, and turning of
the compost piles.
5.1.1 CHLt
To research the issue of CKU emissions, we fkst conducted a literature search for articles on CBU
generation from composting. We found very few articles specifically addressing CHU emissions from
composting published between 1991 and 1999,4 and thus decided not to continue searching for earlier
articles. Because CEU emissions from composting are addressed only occasionally in the literature, we
contacted several composting experts from universities and the U.S. Department of Agriculture to discuss
the potential for CHLt generation, based on the nature of carbon flows during composting. Our CH4
analysis is based on their expert opinions.
The researchers we contacted stated that well-managed compost operations usually do not
generate CBLt because they typically maintain an aerobic environment with proper moisture content to
encourage aerobic decomposition of the materials. The researchers also noted that even if CHU is
generated in anaerobic pockets in the center of the compost pile, the CH^ is most likely oxidized when it
reaches the oxygen-rich surface of the pile, where it is converted to CO2. Several of the researchers
commented that anaerobic pockets are most apt to develop when too much water is added to the compost
pile. They noted that this problem rarely occurs because compost piles are much more likely to be
watered too little rather than too much.
We concluded from the available information that CH4 generation from centralized compost piles
is essentially zero.
5.1.2 CO2 from Transportation of Materials and Turning of Compost
This study estimated the indirect CO2 emissions associated with collecting and transporting
organics to centralized compost facilities, and turning the compost piles. We began with estimates
developed by Franklin Associates, Ltd. (FAL) for the amount of diesel fuel required to (1) collect and
transport 1 ton of organics5 to a central composting facility (363,000 Btu) and (2) turn the compost pile
(221,000 Btu).6 We converted these estimates to units of metric tons of carbon equivalent (MTCE) per
ton of organics, based on a carbon coefficient of 0.02 MTCE per million Btu of diesel fuel. This resulted
in an estimate of 0.01 MTCE of indirect CO2 emissions per ton of material composted in a centralized
facility.
5.2 POTENTIAL CARBON STORAGE
We also evaluated the effect of compost application on soil carbon storage. We did not find
information on carbon storage associated with compost derived specifically from yard trimmings or food
discards. Nevertheless, it is reasonable to expect that these materials are basically homogeneous with
respect to the fate of their stored carbon, even though their initial moisture and carbon content differs.
4 Among the papers with pertinent information is that of HJ. Hellebrand, 1998, Emission of Nitrous Oxide
and other Trace Gases during Composting of Grass and Green Waste, J. Agric. Engineering Research 69:365-375.
5 Measured on a wet weight basis, as MSW is typically measured.
6 Franklin Associates, Ltd. 1994. The Role of Recycling in Integrated Solid Waste Management to the Year
2000 (Stamford, CT: Keep America Beautiful), pp. 1-27, 30, and 31.
66
-------
To develop carbon storage estimates for composted organics, we researched the processes that
affect soil carbon storage, reviewed the results of experiments on the soil carbon impacts of applying
organic amendments (e.g., compost, manure, biosolids, and crop residues), and interviewed experts on ,
the potential carbon storage benefits of composting organics as compared to other methods of disposal.
During this process, four hypotheses were proposed regarding the benefits of applying organics compost
to soil:
(1) Many soils have been depleted in organic matter through cultivation and other practices.
Adding compost can raise soil carbon levels by increasing organic matter inputs. Soils
degraded by intensive crop production, construction, mining, and other activities lose organic
matter when decomposition rates and removals of carbon in harvests exceed the rate of new
inputs of organic materials. Adding compost shifts the balance so that soil organic carbon
levels are restored to higher levels. Some of the compost carbon is retained by the system.
(2) Nitrogen in compost can stimulate higher productivity, thus generating more crop residues.
This "fertilization effect" would increase soil carbon due to the larger volume of crop
residues, which serve as organic matter inputs.
.... . _, . . , ^ ; , ^ , -{ _, .
(3) The composting process leads to increased formation of stable carbon compounds (e.g., humic
substances, aggregates) that then can be stored in the soil for long (>50 years) periods of time.
Humic substances comprise 60-80 percent of soil organic matter and are made .up of complex
compounds that render them resistant to microbial attack.7 In addition to humic substances,
soil organic carbon may be held in aggregates (i.e., stable organo-mineral complexes in which
carbon is bonded with clay colloids and metallic elements) and protected against microbial
attack.8 '"' '
(4) The application of compost produces a multiplier effect by qualitatively changing the
dynamics of the carbon cycling system and increasing the retention of carbon from non-
compost sources. Some studies of other, compost feedstocks (e.g., farmyard manure, legumes)
have indicated that the addition of organic matter to soil plots can increase the potential for
storage of soil organic carbon. The carbon increase apparently comes not only from the
organic matter directly, but also from retention of a higher proportion of carbon from residues
of crops grown on the soil. This multiplier effect could enable compost to increase carbon
storage by more than its own direct contribution to carbon mass accumulation.
Our research efforts did not yield any primary data that could be used to develop quantitative
estimates of the soil carbon storage benefits of compost. Therefore, we developed modeling approaches
to investigate the possible effects of compost application on soil carbon storage. Section 5.2.2 describes
application of the CENTURY model to quantify soil carbon restoration and nitrogen fertilization
associated with compost application to carbon-depleted soils. We conducted a bounding analysis,
described in Section 5.2.6, to address the third hypothesis, incremental humus formation. Although
several of the experts we spoke with cited persuasive qualitative evidence of the existence of a multiplier
effect, we were unable to develop an approach to quantify this process. In that sense, our carbon storage
estimates are likely to be conservative (i.e., understate carbon storage rates), at least for soils with high
silt and/or clay content where this process is most likely to apply.
7 N. Brady and R. Weil. 1999. The Nature and Properties of Soils (Upper Saddle River, NJ: Prentice Hall).
8 R. Lai et al. 1998. The Potential of U.S. Cropland to Sequester Carbon and Mitigate the Greenhouse
Effect (Ann Arbor, MI: Sleeping Bear Press, Inc).
67
-------
Our analyses of soil carbon restoration, nitrogen fertilization, and incremental humus formation
apply relatively simple models of very complex processes. These processes probably are controlled by a
number of biological, physicochemical, and compost management factors, such as application (i.e.,
silviculture, horticulture, agriculture, and landscaping); application rate; regional and local climatic
factors; soil type; and, to a lesser extent, compost feedstock (e.g., grass, leaves, branches, yard trimmings,
food discards). In addition, the results are time-dependent, so the year in which benefits are assessed has
an effect on the magnitude of carbon storage.
Note that the framework used here describes the soil carbon benefits of composting relative to
landfilling and combustion. In all three management methods, yard trimmings are collected and removed
from soils in residential or commercial settings. This removal may result in some loss of organic carbon
from the "home soil." An estimate of the "absolute" soil carbon storage value would net out whatever
loss occurs due to the removal of the yard trimmings. This effect is probably a negligible one, however,
and we were unable to find empirical data on it. Because the decrement in carbon in "home soil" applies
equally to all three management practices, and emission factors are intended to be viewed relative to
other management practices (see Chapter 8), neglecting the carbon loss from the home soil does not
compromise the validity of the results.
5.2.1 Modeling Soil Carbon Restoration and Nitrogen Fertilization
As mentioned above, this analysis included an extensive literature review and interviews with
experts to consider whether the application of compost leads to long-term storage of carbon in soils.
After determining that neither the literature review nor discussions with experts would yield a basis for a
quantitative estimate of soil carbon storage, we evaluated the feasibility of a simulation modeling
approach. We initially identified two simulation models with the potential to be applied to the issue of
soil carbon storage from compost application: CENTURY9 and the Rothamsted C (ROTHC-26.3)10
model. Both are peer-reviewed models whose structure and application have been described in scores of
publications. They share several features:
« Ability to run multi-year simulations;
• Capability to construct multiple scenarios covering various climate and soil conditions and
loading rates: and
• Ability to handle interaction of several soil processes, environmental factors, and
management scenarios such as carbon: nitrogen (C:N) ratios, aggregate formation, soil
texture (e.g., clay content), and cropping regime.
Given the extensive application of CENTURY in the United States, its availability on the Internet, and its
ability to address many of the processes important to compost application, we decided to use CENTURY
rather than ROTHC-26.3.
5.2.2 CENTURY Model Framework
CENTURY is a Fortran model of plant-soil ecosystems that simulates long-term dynamics of
carbon, nitrogen, phosphorus, and sulfur. It tracks the movement of carbon through soil pools—active,
slow, and passive—and can show changes in carbon levels due to the addition of compost.
9 Metherell, A., L. Harding, C. Cole, W. Parton. 1993. CENTURY Agroecosystem Version 4.0, Great
Plains System Research Unit Technical Report No. 4, USDA-ARS Global Climate Change Research Program,
(Colorado State University: Fort Collins, CO).
10 This model was developed based on long-term observations of soil carbon at Rothamsted, an estate in the
United Kingdom where organic amendments have been added to soils since the 19th century.
68
-------
In addition to soil organic matter pools, carbon can be found in surface (microbial) pools and in
above- and below-ground litter pools. The above-ground and below-ground litter pools are divided into
metabolic and structural pools based on the ratio of lignin to nitrogen in the litter. The structural pools
contain all of the lignin and have much slower decay rates than the metabolic pools. Carbon additions to
the system flow through the various pools and can exit the system (e.g., as CO2, dissolved carbon, or
through crop removals).
The above-ground and below-ground litter pools are split into metabolic and structural pools
based on the ratio of lignin to nitrogen in the litter. The structural pools contain all of the lignin and have
much slower decay rates than the metabolic pools. The active pool of soil organic matter includes living
biomass, some of the fine particulate detritus,11 most of the non-humic material, and some of the more
easily decomposed fulvic acids. The active pool is estimated to have a mean residence time (MRT)12 of a
few months to 10 years.13 The slow pool includes resistant plant material (i.e., high lignin content)
derived from the structural pool and other slowly decomposable and chemically resistant components. It
has an MRT of 15-100 years.14 The passive pool of soil organic matter includes very stable materials
remaining in the soil for hundreds to thousands of years.15
CENTURY does not simulate increased formation of humic substances associated with organic
matter additions, nor does it allow for organic matter additions with high humus content to increase the
magnitude of the passive pool directly. (Because CENTURY does not account for these processes, we
developed a separate analysis, described in Section 5.2.6.)
CENTURY contains a submodel to simulate soil organic matter pools. Additional submodels
address nitrogen, phosphorus, sulfur, the water budget, leaching, soil temperature, and plant production,
as well as individual submodels for various ecosystems (e.g., grassland, cropland). The nitrogen
submodel addresses inputs of fertilizer and other sources of nitrogen, mineralization of organic nitrogen,
and uptake of nitrogen by plants. •..•••
5.2.3 Inputs
The CENTURY model simulates the long-term dynamics of. various plant-soil ecosystems (e.g.,
grassland, agricultural land, forest, and savanna). The model uses a series of input files to specify
modeling conditions: Crop, Harvest, Fertilization, Cultivation, Organic Matter Addition, Irrigation,
Grazing, Fire, Tree Type, Tree Removal, Site, and Weather Statistics. A' schedule file is used to specify
the timing of events.
For this analysis, we developed a basic agricultural scenario where land was converted from
prairie to farmland (growing corn) in 1921 and remains growing corn through 2030. We then evaluated
more than 30 scenarios to examine the effect of several variables on soil carbon storage:
11 Detritus refers to debris from dead plants and animals.
12 The term "mean residence time (MRT)" is used interchangeably with "turnover time" and is the average
time in which a unit (e.g., a carbon atom) resides within a "state'' .where there is Jaoih an input and an output. MRT is
only strictly defined at steady-state (i.e., inputs = outputs), but as most soils systems have a continuing input of
carbon and an approximately equal output through decomposition and transfer to other pools, MRT is often used to
describe carbon dynamics in soils. Mathematically, it is the ratio of (a) mass in the pool to (b) throughput of carbon.
For example, if a given carbon pool has a mass of 1,000 kg and the inflow is 1 kg/yr, the MRT is 1,000 kg / (1 kg/yr)
= 1,000 yr.
13 Metherell et al. 1993, Brady and Weil 1999.
14
15
Ibid.
Ibid.
69
-------
• Compost application rate and frequency;
• Site characteristics (rainfall, soil type, irrigation regime);
• Fertilization rate; and
• Crop residue management.
Compost application rates were adjusted using the organic matter (compost) files for each
compost application rate included in our analysis. We compared the effect of applying compost annually
for 10 years (1996-2005) at seven different application rates: 1.3, 3.2, 6.5,10,15, 20, and 40 wet tons
compost/acre (corresponding to 60-1,850 grams of carbon per square meter).16 We also investigated the
effect of compost application frequency on the soil carbon storage rate and total carbon levels. We ran
the model to simulate compost applications of 1.3 wet tons compost/acre and 3.2 wet tons compost/acre
every year for 10 years (1996-2005) and applications of 1.3 wet tons compost/acre and 3.2 wet tons
compost/acre applied every five years (in 1996, 2001, and 2006). The simulated compost was specified
as having 33 percent lignin,1717:1 carbon-to-nitrogen (C:N) ratio,18 60:1 carbon-to-phosphorus ratio, and
75:1 carbon-to-sulfur ratio.19 We also ran a scenario with no compost application for each combination of
site-fertilization-crop residue management. This scenario allowed us to control for compost application,
i.e., to calculate the change in carbon storage attributable only to the addition of compost.
The majority of inputs needed to specify a scenario reside in the site file. The input variables in
this file include the following:
• Monthly average maximum and minimum air temperature;
» Monthly precipitation;
• Lignin content of plant material;
• Plant nitrogen, phosphorus, and sulfur content;
• Soil texture;
• Atmospheric and soil nitrogen inputs; and
16 The model requires inputs in terms of the carbon application rate in grams per square meter. The
relationship between the carbon application rate and compost application rate depends on three factors: the moisture
content of compost, the organic matter content (as a fraction of dry weight), and the carbon content (as a fraction of
organic matter). Our inputs are based on values provided by Dr. Harold Keener of Ohio State University, who
estimates that compost has a moisture content of 50 percent, an organic matter fraction (as dry weight) of 88 percent,
and & carbon content of 48 percent (as a fraction of organic matter). Thus, on a wet weight basis, 21 percent of
compost is carbon.
17 Percent lignin was estimated based on the lignin fractions for grass, leaves, and branches specified by
compost experts (particularly Dr. Gregory Evanylo at Virginia Polytechnic Institute and State University, and lignin
fractions reported hi M.A. Barlaz, "Biodegradative Analysis of Municipal Solid Waste in Laboratory-Scale
landfills," EPA 600/R-97-071,1997. FAL provided an estimate of the fraction of grass, leaves, and branches in yard
trimmings in a personal communication with ICF Consulting, November 14,1995. Subsequently, FAL obtained and
provided data showing that the composition of yard trimmings varies widely in different states. The percentage
composition used here (50 percent grass, 25 percent leaves, and 25 percent branches on a wet weight basis) is within
the reported range.
18 The C:N ratio was taken from Brady and Weil, 1999, The Nature and Property of Soils: Twelfth Edition,
Prentice Hall. '
19 C:P and C:S,ratios were based on the literature and conversations with composting experts, including Dr.
Gregory Evanylo at Virginia Polytechnic Institute and State University.
70
-------
• Initial soil carbon, nitrogen, phosphorus, and sulfur levels.
Several sets of detailed site characteristics from past modeling applications are available to users.
We chose two settings: an eastern Colorado site with clay loam soil and a southwestern Iowa site with
silty clay loam soil. Both settings represent fairly typical Midwestern corn belt situations where
agricultural activities have depleted soil organic carbon levels. The Colorado scenario is available as a
site file on the CENTURY Web site;20 Dr. Keith Paustian, an expert in the development, and application
of CENTURY, provided the specifications for the Iowa site (as well as other input specifications and
results for several of the runs described here).
wj -'••'.
We also varied fertilization rate. As discussed earlier, one of our hypotheses was that the
mineralization of nitrogeri in compost could stimulate crop growth, leading to production of more organic
residues, which in turn would increase soil organic carbon levels. The strength of this effect would vary
depending on the availability of other sources of nitrogen. To investigate this hypothesis, we analyzed
different rates of synthetic fertilizer addition ranging from zero up to a typical rate to attain average crop
yield (90 Ibs. N/acre for the Colorado site, 124 Ibs. N/per acre for the Iowa site). We also analyzed
fertilizer application at half of these typical rates. ,-.•-. : • '
Finally, we simulated two harvest regimes, one where the corn is harvested for silage (where 95
percent of the above-ground biomass is removed) and the other where corn is harvested for grain (where
the "stover" is left behind to decompose on the field). These simulations enabled us to isolate the effect
of the carbon added directly to the system in the form of compost, as opposed to total carbon inputs
(which include crop residues).
5.2.4 Outputs
CENTURY is capable of providing a variety of output data, including carbon storage in soils,
CO2 emissions due to microbial respiration, and monthly potential evapotranspiration. The outputs we
chose were carbon levels for each of the eight soil pools: structural carbon in surface litter, metabolic
carbon in surface litter, structural carbon in soil litter, metabolic carbon in soil litter, surface pool, active
pool, slow pool, and passive pool. Our output data cover the period from 1900 through 2030. In general,
we focussed on the difference in carbon storage between a baseline scenario, where no compost was
applied, and a with-compost scenario. We calculated the delta between the two scenarios to isolate the
effect of compost application. Output data in grams of carbon per square meter were converted to MTCE
by multiplying by area (in square meters).
To express results in units comparable to those for other sources and sinks, we divided the
increase in carbon storage by the short tons of organics required to produce the compost.21 That is, we
express the factors as a carbon storage rate in units of MTCE per wet short ton of organic inputs (not
MTCE per short ton of compost).
5.2.5 Results
The carbon storage rate declines with time after initial application. The rate is similar across application
rates and frequencies, and across the site conditions we simulated. Exhibit 5-1 displays results for the
Colorado and Iowa sites, for the 10-, 20-, and 40-ton per acre application rates. As indicated on the graph,
the soil carbon storage rate varies from about 0.08 MTCE per wet ton organics immediately after
20
http://www.nrel.colostate.edu/PROGILALMS/MODELING/CENTURY/CENTURY.html
21 We assume 2.1 tons of yard trimmings are required to generate 1 ton of composted yard trimmings. Thus,
to convert the results in this report (in MTCE per wet ton yard trimmings) to MTCE per wet ton of compost, multiply
by 2.1. To convert to MTCE per dry ton compost, multiply values in this report by 4.2 (assuming 50 percent
moisture content).
v*,
71
-------
Exhibit 5-1 Soil C Storage-CO and IA sites; 10, 20, and 40 tpa application rates
0.09
0.08
O
.--•-.. C010 tpa
..-A--- CO20tpa
---*--- CO 40 tpa
IA 10 tpa
(A 20 tpa
lA 40 tpa
0.01
8
-------
Exhibit 5-2 shows the carbon storage rate for the Iowa site and the effect of nitrogen fertilization.
The two curves in the exhibit both represent the difference in carbon storage between (a) a with-compost
scenario (20 tons per acre) and (b) a baseline where compost is not applied. The nitrogen application
rates differ in the following ways:
• The curve labeled "Typical N application" represents application of 124 Ibs. per acre, for
both the compost and baseline scenario. Because the nitrogen added via compost has little
effect when nitrogen is already in abundant supply, this curve portrays a situation where the
carbon storage is attributable solely to the organic matter additions in the compost.
• The curve labeled "Half N application" represents application of 62 Ibs. per acre. In this
scenario, mineralization of nitrogen added by the compost has an incremental effect on crop
productivity compared to the baseline. The difference between the baseline and compost
application runs reflects both organic matter added by the compost and additional biomass
produced in response to the nitrogen contributed by the compost.
0.09
0.08
Exhibit 5-2
Incremental Carbon Storage as a function of Nitrogen Application Rate
Iowa site, corn harvested for grain
0.00
O3 T— CO LO
03 O O O
O3 O3 O O O
•<- T- CM CM CM
1^- O3 T- CO LO h- O3
O O O
CM CM CM
Year
O
CM
T- co in
s s s
CM CM CM
h- O3
s s
CM CM
-Typical N
application
Half N application
73
-------
The difference in incremental carbon storage rates between the two fertilization scenarios is less than
0.01 MTCE per ton, indicating that the nitrogen fertilization effect is small. Note that this finding is
based on the assumption that farmers applying compost also will apply sufficient synthetic fertilizer to
maintain economic crop yields. If this assumption is not well-founded, or in situations where compost is
applied as a soil amendment for road construction, landfill cover, or similar situations, the effect would
be larger.
When viewed from the perspective of total carbon, rather than as a storage rate per ton of inputs
to the composting process, both soil organic carbon concentrations and total carbon stored per acre
increase with increasing application rates (see Exhibit 5-3). Soil organic carbon concentrations increase
throughout the period of compost application, peak in 2006 (the last year of application), and decline
thereafter due to decomposition of the imported carbon. Exhibit 5-3 displays total carbon storage
(including baseline carbon) in soils on the order of 40 to 65 metric tons per acre (the range would be
higher with higher compost application rates or applications with a longer term).
Exhibit 5-3 Total Soil C
Iowa site, corn harvested for grain
£ 10,000
o
E 8,000
O)
6,000
4,000
2,000
0
70
;• 60
a«*****«r»«tt«H«f#*l
•-50
-40
H—
&—
— 1 .3 tons/acre
-20 tons/acre
8*
o
(0
O
M
o>
c
> . • 30 -5
-•20 E
10
o/ cv
Year
74
-------
5.2.6 Incremental Humus Formation
The third of the four hypotheses describing the benefits of composting, as compared to
alternative management methods, is predicated on incremental formation of stable carbon compounds
that can be stored in the soil for long periods of time. CENTURY does not simulate this process, i.e., it
does not allow for organic matter additions with high humus content to directly increase the magnitude of
the passive pool. Therefore, we used a bounding analysis to estimate the upper and lower limits of the
magnitude of this effect. In this analysis, we evaluated the amount of long-term soil carbon storage when
organics are composted and applied to soil.
During the process of decomposition, organic materials typically go through a series of steps
before finally being converted to CO2, water, and other reaction products. The intermediate compounds
that are formed, and the lifetime of these compounds, can vary widely depending on a number of factors,
including the chemical composition of the parent compound. Parent compounds range from readily
degradable molecules such as cellulose and hemicellulose to molecules more resistant to degradation,
such as lignin, waxes, and tannins.
Composting is designed to promote rapid decomposition of organics, thus reducing their volume.
Some evidence suggests that composting produces a greater proportion of humus than that typically
formed when organics are left directly on the ground. The conditions in the two phases are different. The
heat generated within compost piles favors 'thermophilic' (heat-loving) bacteria, which tend to produce a
greater proportion of stable, long-chain carbon compounds (e.g., humic substances) than do bacteria and
fungi that predominate at ambient soil temperatures.
Increased humus formation associated with compost application is a function of two principal
factors:
(1) The fraction of carbon in compost that is considered "passive" (i.e., very stable); and
(2) The rate at which passive carbon is degraded to CO2.
Estimates for the first factor are based on experimental data compiled by Dr. Michael Cole of the
University of Illinois. Dr. Cole found literature values indicating that between 4 and 20 percent of the
carbon in finished compost degrades quickly.23 Dr. Cole averaged the values he found in the literature
and estimated that 10 percent of the carbon in compost can be considered "fast" (i.e., readily
degradable). The remaining 90 percent of carbon in compost can be classified as either slow or passive.
We were not able to locate experimental data that delineates the fractions of slow and passive carbon in
compost; therefore, we developed upper and lower bound estimates based on Dr. Cole's professional
judgement. He suggested values of 30 percent slow and 60 percent passive, and 45 percent slow and 45
percent passive for the upper and lower bounds on passive content, respectively.24
For the second factor, we chose a mean residence time for passive carbon of 400 years based on
the range of values specified in the literature.25
23 Very little information is available on the characteristics of compost derived from yard trimmings or food
discards. However, Dr. Cole found that the composition of composts derived from other materials is broadly
consistent, suggesting that his estimates may be reasonably applied to yard trimmings or food scrap compost.
24 We focussed only on the passive pool because (1) the CENTURY model does not allow for direct input
of organic carbon into the passive pool, and (2) the model runs resulted in very little indirect (i.e., via other pools)
formation of passive carbon. Although the first factor is also true for the slow pool, the second is not. Had we
analyzed slow carbon in the same way as passive carbon, there would be potential for double-counting (see
discussion in Section 5.3).
25
Metherell et al. 1993, Brady and Weil 1999.
75
-------
Combining the two bounds for incremental humus formation (60 percent passive and 45 percent
passive), we estimated the incremental carbon storage implied by each scenario (see Exhibit 5-4).
Exhibit 5-4 Incremental Carbon Storage:
MTCE/wet ton vs time
0.06
- - Upper bound
Low er bound
0.00
20
30 40 50 60 70
Time since compost application (years)
80
90
100
The upper bound on the incremental carbon storage from composting is more than 0.05 MTCE
per ton of organics (shown in the top left of the graph); the lower bound is approximately 0.03 MTCE per
ton (shown in the bottom right of the graph) after about 100 years. Incremental storage is sensitive to the
fraction of carbon in compost that is passive but is not very sensitive to the degradation rate (within a
100-year time horizon, over the range of rate constants appropriate for passive carbon).
To select a point estimate for the effect of incremental humus formation, we took the average
storage value across the two bounding scenarios, when time equals 10 years (i.e., approximately 2010).
The resulting value is 0.05 MTCE/ton. The 2010 time frame was chosen for this analysis because the
forest carbon estimates presented in Chapter 3 of this report are for the period ending in 2010.
5.3 NET GHG EMISSIONS FROM COMPOSTING
The approaches described in Section 5.2 were adopted to capture the range of carbon storage
benefits associated with compost application. However, this dual approach creates the possibility of
double counting. In an effort to eliminate double counting, we evaluated the way that CENTURY
partitions compost carbon once it is applied to the soil.
To do so, we ran a CENTURY model simulation of compost addition during a single year and
compared the results to a corresponding reference case (without compost). We calculated the difference
in carbon in each of the CENTURY pools for the two simulations and found that the change in the
passive pool represented less than 0.01 percent of the change in total carbon. Therefore, CENTURY is
not adding recalcitrant carbon directly to the passive pool. Next, we graphed the change in the passive
pool over time to ensure that the recalcitrant compost carbon was not being cycled from the faster pools
into the passive pool several years after the compost is applied. As Exhibit 5-5 shows, CENTURY does
not introduce significant increments (over the base case) of recalcitrant carbon into the passive pool at
anytime.
76
-------
Exhibit 5-5 Difference in Carbon Storage Between Compost Addition and Base Case
yearly application with 20 tons compost
1000
900
800
700
JS- 600
500
CO
O 400
300
200
100
0
)fe;XfXT>fr^^
—X— Passive
—-—Total Carbon
CD CO O CM •* CO
O> O5 O O O O
cn en o o o o
1- -i- CM CM CM CM
a
O
CM
O
CM
CO O CM
T- CM CM
O O O
CM CM CM
Year
Based on our analysis, it appears that CENTURY is appropriately simulating carbon cycling and
storage for all but the passive carbon introduced by compost application. Because passive carbon
represents approximately 52 percent of carbon in compost (the midpoint of 45 percent and 60 percent),
we scaled the CENTURY results by 48 percent to reflect the proportion of carbon that can be classified
as fast or slow (i.e., not passive).
Exhibit 5-6 shows the soil carbon storage and transportation-related emissions and sinks, and
sums these to derive estimates of a net GHG emission factor, using the same sign convention as our
broader analysis. A negative value denotes carbon storage; a positive value denotes emissions.
Summing the values corresponding to typical application rate and the 2010 time frame for soil
carbon restoration (-0.02 MTCE/ton), increased humus formation (-0.05 MTCE/ton), and transportation
emissions (0.01 MTCE/ton), the result is-0.05 MTCE/ton.26
26
The addends do not sum to the total, due to rounding.
77
-------
Exhibit 5-6
Net GHG Emissions from Composting
(In MTCE Per Short Ton of Yard Trimmings Composted)
Emission/ Storage Factor (for 2010)
Soil Carbon Restoration
Unweighted
-0.04
Proportion of
C that is not
passive
48%
Weighted
estimate
-0.02
Increased
Humus
Formation
-0.05
Transportation
Emissions
0.01
Net Carbon Flux
-0.05
5.4 LIMITATIONS
Due to data and resource constraints, this chapter does not explore the full range of conditions
under which compost is managed and applied, and how these conditions would affect the results of this
analysis. Instead, this study attempts to provide an analysis of GHG emissions and sinks associated with
centralized composting of yard trimmings and food discards (henceforth, organics) under a limited set of
scenarios. Our analysis was limited by the lack of primary research on carbon storage and CHU generation
associated with composting. The limited availability of data forced us to rely on two modeling
approaches, each with its own set of limitations. In addition, our analysis was limited by the scope of the
report, which is intended to present life-cycle GHG emissions of waste management practices for
selected material types, including food discards and yard trimmings.
5.4.1 Limitations of Modeling Approaches
Due to data and resource constraints, we were unable to use CENTURY to evaluate the variation
in carbon storage impacts for a wide range of compost feedstocks (e.g., yard trimmings mixed with food
discards, food discards alone). As noted earlier, resource constraints limited the number of soil types,
climates, and compost applications simulated. The CENTURY results also incorporate the limitations of
the model itself, which have been well documented elsewhere. Perhaps most importantly, the model's
predictions of soil organic matter levels are driven by four variables: annual precipitation, temperature,
soil texture, and plant lignin content. Beyond these, the model is limited by its sensitivity to several
factors for which data are difficult or impossible to obtain (e.g., presettlement grazing intensity, nitrogen
input during soil development).27 The model's monthly simulation intervals limit its ability to fully
address potential interactions between nitrogen supply, plant growth, soil moisture, and decomposition
rates, which may be sensitive to conditions that vary on a shorter time scale.28 In addition, the model is
not designed to capture the hypothesis that, due to compost application, soil ecosystem dynamics change
so that more carbon is stored than is actually being added to the soil (i.e., the multiplier effect).
27 Parton, W., D.Schimel, C. Cole, and D. Ojima. 1987. "Analysis of Factors Controlling Soil Organic
Matter Levels in Great Plains Grasslands." Soil Sci. Soc. Am.J. Vol. 51 (1173-1179).
28 Paustian, K., W. Parton, and Jan Persson. 1992. "Modeling Soil Organic Matter in Organic-Amended
and Nitrogen-Fertilized Long-Term Plots." Soil Sci. Soc. Am. J. Vol. 56 (476-488).
78
-------
CENTURY simulates carbon movement through organic matter pools. Although the model is
designed to evaluate additions of organic matter in general, to our knowledge it has not been applied in
the past to evaluate the application of organics compost. CENTURY is parameterized to partition carbon
to the various pools based on ratios of lignin to nitrogen and lignin to total carbon, not on the amount of
organic material that has been converted to humus already. We addressed this limitation by developing
an "add-on" analysis to evaluate humus formation in the passive pool, scaling the CENTURY results,
and summing the soil carbon storage values. There is some potential for double-counting, to the extent
that CENTURY is routing some carbon to various pools that is also accounted for in the incremental
humus analysis. We believe that this effect is likely to be minor.
The bounding analysis used to analyze increased humus formation is limited by the lack of data
specifically dealing with composts composed of yard trimmings or food discards. This analysis is also
limited by the lack of data on carbon in compost that is passive. The approach of taking the average value
from the two scenarios is simplistic but appears to be the best available option.
5.4.2 Limitations Related to the Scope of the Report
As indicated above, this chapter presents our estimates of the GHG-related impacts of
composting food discards and yard trimmings. These estimates were developed within the framework of
the larger report; therefore, our presentation of results, estimation of emissions and sinks, and description
of ancillary benefits was not comprehensive. The remainder of this section describes specific limitations
of our compost analysis.
As in the other chapters of this report, the GHG impacts of composting reported in this chapter
are relative to other possible disposal options for yard trimmings (i.e., landfilling and combustion). In
order to present absolute GHG emission factors for composted yard trimmings that could be used to
compare composting to a baseline of leaving yard trimmings on the ground where they fall, we would
need to analyze the home soil. In particular, the carbon storage benefits of composting would need to be
compared to the impact that removal of yard trimmings has on the home soil.
As mentioned in Section 5.4.1, due to data and resource constraints, our analysis considers a
small sampling of feedstocks and a single compost application (i.e., agricultural soil). We analyzed two
types of compost feedstocks—yard trimmings and food discards—although sewage sludge, animal
manure, and several other compost feedstocks also may have significant GHG implications. Similarly, we
assumed that compost was applied to degraded agricultural soils, despite widespread use of compost in
land reclamation, silviculture, horticulture, and landscaping.
This analysis did not consider the full range of soil conservation and management practices that
could be used in combination with compost and the impacts of those practices on carbon storage. Some
research indicates that adding compost to agricultural soils in conjunction with various conservation
practices enhances the generation of soil organic matter to a much greater degree than applying compost
alone. Examples of these conservation practices include conservation tillage, no tillage, residue
management, crop rotation, wintering, and summer fallow elimination. Research suggests that allowing
crop residues to remain on the soil rather than turning them over helps to protect and sustain the soil
while simultaneously enriching it. Alternatively, conventional tillage techniques accelerate soil erosion,
increase soil aeration, and hence lead to greater GHG emissions.29
29 R. Lai et al. 1998. The Potential of U.S. Cropland to Sequester Carbon and Mitigate the Greenhouse
Effect (Ann Arbor, MI: Sleeping Bear Press, Inc).
79
-------
As is the case in other chapters, the methodology used to estimate GHG emissions from
composting did not allow for variations in transportation distances. We recognize that the density of
landfills versus composting sites in any given area would have an effect on the extent of transportation
emissions derived from composting. For example, in states that have a higher density of composting sites,
the hauling distance to such a site would be less and would require less fuel than transportation to a
landfill. Alternatively, transporting compost from urban areas, where compost feedstocks may be
collected, to farmlands, where compost is typically applied, potentially would require more fuel because
of the large distance separating the sites.
Emission factors presented in this chapter do not capture the full range of possible GHG
emissions from compost. Some of the nitrogen in compost is volatilized and released into the atmosphere
as N2O shortly after application of the compost. Based on a screening analysis, we estimated N2O
emissions to be less than 0.01 MTCE per wet ton of compost inputs and thus considered this effect to be
negligible.
Addressing the possible GHG emission reductions and other environmental benefits achievable
by applying compost instead of chemical fertilizers, fungicides, and pesticides was beyond the scope of
this report. Manufacturing these agricultural products requires energy. To the extent that compost may
replace or reduce the need for these substances, composting may result in reduced energy-related GHG
emissions. Although we understand that compost is generally applied for its soil amendment properties
rather than for pest control, compost has been effective in reducing the need for harmful or toxic
pesticides and fungicides.30
In addition to the carbon storage benefits of adding compost to agricultural soils, composting can
lead to improved soil quality, improved productivity, and cost savings. As discussed earlier, nutrients in
compost tend to foster soil fertility.31 In fact, composts have been used to establish plant growth on land
previously unable to support vegetation. In addition to these biological improvements, compost also may
lead to cost savings associated with avoided waste disposal, particularly for feedstocks such as sewage
sludge and animal manure.
For example, the use of compost may reduce or eliminate the need for soil fumigation with methyl
bromide (an ozone-depleting substance) to kill plant pests and pathogens.
31 N. Brady and R. Weil. 1999. The Nature and Properties of Soils (Upper Saddle River, NJ: Prentice Hall).
80
-------
6. COMBUSTION
This chapter presents estimates of the net GHG emissions from combustion of each of the
materials considered in this analysis and several categories of mixed waste streams (e.g., mixed paper,
mixed recyclables, and mixed MSW). Combustion of MSW results in emissions of CO2 (because nearly
all of the carbon in MSW is converted to CO2) and N2O. Note that CO2 from burning biomass sources
(such as paper products and yard trimmings) is not counted as a GHG because it is biogenic (as explained
in Section 1.4). ,
Combustion of MSW with energy recovery in a waste-to-energy (WTE) plant also results in
avoided CO2 emissions at utility and metals production facilities. First, the electricity produced by a
WTE plant displaces electricity that would otherwise be provided by an electric utility power plant.
Because most utility power plants burn fossil fuels and thus emit CO2, the electricity produced by a WTE
plant reduces utility CO2 emissions. These avoided GHG emissions must be subtracted from the GHG
emissions associated with combustion of MSW. Second, most MSW combusted with energy recovery in
the United States is combusted in WTE plants that recover ferrous metals (e.g., steel) and non-ferrous
materials (e.g., non-ferrous metals and glass).1 The recovered ferrous metals and non-ferrous materials
then are recycled.2 As discussed in Chapter 4, processes using recycled inputs requke less energy than
processes using virgin inputs. In measuring GHG implications of combustion, one also must account for
the change in energy use due to recycling associated with metals recovery.
WTE facilities can be divided into three categories: (1) mass burn, (2) modular, or (3) refuse-
derived fuel (RDF). A mass burn facility generates electricity and/or steam from the combustion of
mixed MSW. In the United States, about 70 mass burn facilities process approximately 21 million tons of
MSW annually.3 Modular WTE plants generally are smaller than mass burn plants and are prefabricated
off-site so that they can be assembled quickly where they are needed. Because of their similarity to mass
burn facilities, modular facilities are treated as part of the mass burn category for the purposes of this
analysis.
An RDF facility combusts MSW that has undergone varying degrees of processing, from simple
removal of bulky and noncombustible items to more complex processes (shredding and material
recovery), which result in a finely divided fuel. Processing MSW into RDF yields a more uniform fuel
that has a higher heating value than is produced by mass burn or modular WTE.4 In the United States,
1 We did not consider any recovery of materials from the MSW stream that may occur before MSW is
delivered to the combustor. We considered such prior recovery to be unrelated to the combustion operation—unlike
recovery of steel from combustor ash, an activity that is an integral part of the operation of many combustors.
2 Note that material recovery at WTE facilities has increased in recent years, and this trend may continue as
more facilities install material recovery systems. According to the Integrated Waste Services Association's 2000
Waste-to-Energy Directory of United States Facilities (www.wte.org), ferrous metal recovery at WTE facilities
increased from more than 773,000 tons in 1999 to more than 788,000 tons in 2000. During the same period, on-site
recycling more than doubled, from approximately 462,000 tons to 939,000 tons.
3 Integrated Waste Services Association, The 2000IWSA Waste-To-Energy Directory of United States
Facilities, Table 1. This estimate assumes that 92 percent of combustion system capacity gets utilized, per e-mail
correspondence with Maria Zannes of IWSA (June 12, 2001).
4 MSW processing into RDF involves both manual and mechanical separation to remove materials such as
glass and metals that have little or no fuel value. ' •
81
-------
approximately 12 facilities process and combust RDF, 7 facilities combust RDF using off-site
processing, and 7 facilities process RDF for combustion off-site. These 26 facilities process
approximately 8 million tons of MSW annually.5
This study analyzed the net GHG emissions from combustion of mixed waste streams, and the
following individual materials at mass burn and RDF facilities:
« Aluminum Cans;
• Steel Cans;
• Glass Containers;
• HOPE Plastic;
• LDPE Plastic;
• PET Plastic;
• Corrugated Cardboard;
• Magazines and Third-class Mail;
• Newspaper;
• Office Paper;
• Phonebooks;6
• Textbooks;7
• Dimensional Lumber;
• Medium-density Fiberboard;
• Food Discards; and
• Yard Trimmings.
Net emissions consist of (1) emissions of non-biogenic CO2 and N2O minus (2) avoided GHG
emissions from the electric utility sector and from processing with recycled inputs (e.g., steel produced
from recycled inputs requires less energy than steel from virgin inputs). There is some evidence that as
combustor ash ages, it absorbs CO2 from the atmosphere. We did not count absorbed CO2, however,
because we estimated the quantity to be less than 0.01 MTCE per ton of MSW combusted.8 Similarly, the
residual waste from processing MSW into RDF is typically landfilled. Some potential exists for the
organic fraction of this residual waste to yield GHG emissions when landfilled. We did not count these
emissions, however, because the quantity emitted is estimated to be less than 0.01 MTCE per ton of
MSW processed into RDF.9 -
5 Integrated Waste Services Association, The 2000IWSA Waste-To-Energy Directory of United States
Facilities, Table 1.
6 Newspaper used as proxy, as material-specific data were unavailable.
7 Office paper used as proxy, as material-specific data were unavailable.
8 Based on data provided by Dr. Jurgen Vehlow, of the Institut fur Technische Chemie in Karlsruhe,
Germany, we estimated that the ash from 1 ton of MSW would absorb roughly 0.004 MTCE of CO2.
9 Based on data provided by Karen Harrington, principal planner for the Minnesota Office of Environmental
Assistance, we estimated that landfilling the residual waste would emit roughly 0.003 MTCE of CO2per ton of MSW
82
-------
The results showed that combustion of mixed MSW has small negative net GHG emissions (in
absolute terms). Combustion of paper products, dimensional lumber, medium-density fiberboard, food
discards, and yard trimmings results in negative net GHG emissions. Processing steel cans at a
combustor, followed by recycling the ferrous metal, likewise results in negative net GHG emissions.
Combustion of plastic produces positive net GHG emissions, and combustion of aluminum cans and
glass results in small positive net GHG emissions. The reasons for each of these results are discussed in
the remainder of this chapter.10
6.1 METHODOLOGY
The study's general approach was to estimate the (1) gross emissions of CO2 and N2O from
MSW and RDF combustion (including emissions from transportation of waste to the combustor and ash
from the combustor to a landfill) and (2) CO2 emissions avoided due to displaced electric utility
generation and decreased energy requirements for production processes using recycled inputs.11 To
obtain an estimate of the net GHG emissions from MSW and RDF combustion, we subtracted the GHG
emissions avoided from the direct GHG emissions. We estimated the net GHG emissions from waste
combustion per ton of mixed MSW and per ton of each selected material in MSW. The remainder of this
section describes how we developed these estimates. ;
6.1.1 Estimating Direct CO2 Emissions from MSW Combustion
The carbon in MSW has two distinct origins. Some of it is derived from sustainably harvested
biomass (i.e., carbon in plant and animal matter that was converted from CO2 in the atmosphere through
photosynthesis). The remaining carbon in MSW is from non-biomass sources, e.g., plastic and synthetic
rubber derived from petroleum.
For reasons described in Section 1.4, this study did not count the biogenic CO2 emissions from
combustion of biomass. On the other hand, we did count CO2 emissions from combustion of non-biomass
components of MSW—plastic, textiles, and rubber. Overall, only a small portion of the total CO2
emissions from combustion are counted as GHG emissions.
For mixed MSW, we used the simplifying assumptions that (1) all carbon in textiles is non-
biomass carbon, i.e., petrochemical-based plastic fibers such as polyester (this is a worst-case
assumption); and (2) the category of "rubber and leather" in EPA's MSW characterization report12 is
composed almost entirely of rubber. Based on these assumptions, this study estimated that there are 0.11
pounds of non-biogenic carbon in the plastic, textiles, rubber, and leather contained in 1 pound of mixed
MSW.13 We assumed that 98 percent of this carbon would be converted to CO2 when the waste is
processed into RDF. Facsimile from Karen Harrington, Minnesota Office of Environmental Assistance to ICF
Consulting, October 1997. . • :
10 Note that Exhibits 6-1, 6-2, and 6-5 do not show mixed paper. Mixed paper is shown in the summary
exhibit (Exhibit 6-6). The summary values for mixed paper are based on the proportions of the four paper types
(newspaper, office paper, corrugated cardboard, and magazines/third-class mail) that comprise the different "mixed
paper" definitions.
11A comprehensive evaluation also would consider the fate of carbon remaining in combustor ash.
Depending on its chemical form, carbon may be aerobically degraded to CO2, anaerobically degraded to CHf, or
remain in a relatively inert form and be stored. Unless the ash carbon is converted to CHt (which we considered to be
unlikely), the effect on the net GHG emissions would be very small.
12 U.S. EPA Office of Solid Waste. 2002. Municipal Solid Waste in the United States: 2000 Facts and
Figures. EPA 530-R-02-001. ,
13 JCF Consulting. 1995. Memorandum. "Work Assignment 239, Task 2: Carbon Sequestration in
Landfills," April 28, Exhibit 2-A, column "o." ,.
83
-------
combusted, with the balance going to the ash. Then we converted the 0.11 pounds of non-biomass carbon
per pound of mixed MSW to units of MTCE per ton of mixed MSW combusted. The resulting value for
mixed MSW is 0.10 MTCE per ton of mixed MSW combusted,14 as shown in Exhibit 6-1.
The study estimated that HDPE and LDPE are 84 percent carbon, while PET is 57 percent
carbon (based on a moisture content of 2 percent). We assumed that 98 percent of the carbon in the
plastic is converted to CO2 during combustion. The values for CO2 emissions, converted to units of
MTCE per ton of plastic combusted, are shown in column "b" of Exhibit 6-1.
6.1.2 Estimating N2O Emissions from Combustion of Waste
Studies compiled by the Intergovernmental Panel on Climate Change (IPCC) show that MSW
combustion results in measurable emissions of N2O, a GHG with a high global warming potential
(GWP).15 The IPCC compiled reported ranges of N2O emissions, per metric ton of waste combusted,
from six classifications of MSW combustors. This study averaged the midpoints of each range and
converted the units to MTCE of N2O per short ton of MSW. The resulting estimate is 0.01 MTCE of N2O
emissions per ton of mixed MSW combusted. Because the IPCC did not report N2O values for
combustion of individual components of MSW, we used the 0.01 value not only for mixed MSW, but
also as a proxy for all components of MSW, except for aluminum cans, steel cans, glass, HDPE, LDPE,
and PET.16
6.1.3 Estimating Indirect CO2 Emissions from Transportation of Waste to the WTE Plant
Next, this study estimated the indirect CO2 emissions from the transportation of waste. For the
indirect CO2 emissions from transporting waste to the WTE plant, and ash from the WTE plant to a
landfill, we used an estimate for mixed MSW developed by Franklin Associates, Ltd. (FAL).17 We then
converted the FAL estimate from pounds of CO2 per ton of mixed MSW to MTCE per ton of mixed
MSW. This resulted in an estimate of 0.01 MTCE of CO2 emissions from transporting 1 ton of mixed
MSW and the resulting ash. We assumed that transportation of any individual material in MSW would
use the same amount of energy as transportation of mixed MSW.
6.1.4 Estimating Gross GHG Emissions from Combustion
To estimate the gross GHG emissions per ton of waste combusted, we summed the values for
emissions from combustion CO2, combustion N2O, and transportation CO2. The gross GHG emission
estimates, for mixed MSW and for each individual material, are shown in column "e" of Exhibit 6-1.
6.1.5 Estimating Utility CO2 Emissions Avoided
Most WTE plants in the United States produce electricity. Only a few cogenerate electricity and steam.
In this analysis, we assumed that the energy recovered with MSW combustion would be in the form of
electricity. This analysis is shown in Exhibit 6-2. We used three data elements to estimate the avoided
electric utility CO2 emissions associated with combustion of waste in a WTE plant: (1) the energy
14 Note that if we had used a best-case assumption for textiles, i.e., assuming they have no petrochemical-
based fibers, the resulting value for mixed MSW would have been 0.09 MTCE per ton of mixed MSW combusted.
15 Intergovernmental Panel on Climate Change, Greenhouse Gas Inventory Reference Manual, Volume 3,
(undated) p. 6-33. The GWP of N2O is 310 times mat of CO2.
16 This exception was made because at the relatively low combustion temperatures found in MSW
combustors, most of the nitrogen in N2O emissions is derived from the waste, not from the combustion air. Because
aluminum and steel cans do not contain nitrogen, we concluded that running these metals through an MSW
combustor would not result in N2O emissions.
17 Franklin Associates, Ltd. 1994. The Role of Recycling in Integrated Solid Waste Management to the Year
2000 (Stamford, CT: Keep America Beautiful, Inc.), p. 1-24.
84
-------
Exhibit 6-1
Gross Emissions of GHGs from MSW Combustion (MTCE/Ton)
(a)
Material Combusted
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed MSW
Carpet
Personal Computers
(b)
Combustion CO2
Emissions From
Non-Biomass
Per Ton
Combusted
0.00
0.00
0.00
0.76
0.76
0.56
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.10
0.47
0.75
(c)
Combustion
N20 Emissions
Per Ton
Combusted
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
6.01
0.01
0.00
0.00
(d)
Transportation
C02 Emissions
Per Ton
Combusted
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
(e)
(e = b + c + d)
Gross GHG
Emissions Per
Ton Combusted
0.01
0.01
0.01
0.77
0.77
0.56
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.12
0.48
0.76
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
Note that Exhibits 6-1, 6-2, and 6-5 show coated paper but not mixed paper;
mixed paper is shown in the summary exhibit (Exhibit 6-6).
The summary values for mixed paper are based on the proportions of the four paper types (newspaper,
office paper, corrugated cardboard, and coated paper) that comprise the different "mixed paper" definitions.
I he values tor phone books ana textbooks are proxies, oasea on newspaper ana ottice paper, respectively.
85
-------
Exhibit 6-2
Avoided Utility GHG Emissions from Combustion at Mass Burn and RDF Facilities
(a)
Material Combusted
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed MSW**
(b)
Energy Content (Btu
Per Pound)
-335 a
-210 a
-235 a
18,687 b
18,687 b
9,702 c,d
7,043 b
5,258 d
7,950 b
6,800 b,f
7,950 g
6,800 h
8,300 i
8,300 i
2,370 b
2,800 j
5,000 k
(c)
Energy
Content
(Million Btu Per
Ton)
-0.7
-0.4
-0.5
37.4
37.4
19.4
14.1
10.5
15.9
13.6
15.9
13.6
16.6
16.6
4.7
5.6
10.0
(d)
Mass Burn
Combustion
System
Efficiency
(Percent)
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
17.8%
(e)
RDF
Combustion
System
Efficiency
(Percent)
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
16.3%
(0
Emission Factor for
Utility-Generated
Electricity (MTCE/
Million Btu of
Electricity
Delivered)
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
0.081
(g)
(g = c*d*f)
Avoided Utility CO2
Per Ton Combusted
at Mass Burn
Facilities (MTCE)
-0.01 *
-0.01 *
-0.01 *
0.54
0.54
0.28
0.20
0.15
0.23
0.20
0.23
0.20
0.24
0.24
0.07
0.08
0.14
(h)
(h = c*e*f)
Avoided Utility
COzPerTon
Combusted at
RDF Facilities
(MTCE)
-0.01 *
-0.01 *
-0.01 *
0.49
0.49
0.25
0.18
0.14
0.21
0.18
0.21
0.18
0.22
0.22
0.06
0.07
0.13
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
* The amount of energy absorbed by 1 ton of steel, aluminum cans, or glass in an MSW combustor would, if not absorbed,
result in less than 0.01 MTCEof avoided utility CO2.
** Mixed MSW represents the entire waste stream as disposed of.
a We developed these estimates based on data on the specific heat of aluminum, steel, and glass and calculated the energy required to raise the temperature
of aluminum, steel, and glass from ambient temperature to the temperature found in a combustor (about 750° Celsius). We obtained the
specific heat data from Incropera, Frank P.and David P. DeWitt, Introduction to Heat Transfer, Second Edition (New York: John Wiley & Sons) 1990, pp. A3-A4.
b MSW Fact Book.
c Gaines and Stodolsky.
d For PET plastic, we converted the value of 9,900 Btu/pound dry weight, to 9,702 Btu/pound wet weight, to account for a moisture content of 2 percent.
e We used Franklin Associates, Ltd.'s value for magazines as a proxy for the value for coated paper.
f We used the MSW Fact Book's value for mixed paper as a proxy for the value for office paper.
g. We used newspapers as a proxy for phonebooks.
h We used office paper as a proxy for textbooks.
i We used the higher end of the Btu factor for Basswood from the USFS. Basswood is a relatively soft wood so its high end Btu content should be most ,
similar to an average factor for all wood types. Fons, W. L; Clements, H. B.; Elliott, E. Ft.; George, P. M. 1962. Project Fire Model. Summary Progress
Report-ll. Period May 1,1960, to April 30,1962. Macon, GA: U.S. Department of Agriculture, Forest Service, Southeastern Forest Experiment Station, Southern
Forest Fire Laboratory. 58 p. [16824]
j Procter and Redfem, Ltd. and ORTECH International.
k Telephone conversation among IWSA, American Ref-Fuel, and ICF Consulting, October 28,1997.
86
-------
content of mixed MSW and of each separate waste material considered, (2) the combustion system
efficiency in converting energy in MSW to delivered electricity, and (3) the electric utility CO2 emissions
avoided per kilowatt-hour of electricity delivered by WTE plants.
Energy content: For the energy content of mixed MSW, we used a value of 5,000 Btu per pound
of mixed MSW combusted, which is a value commonly used in the WTE industry.18 This estimate is
within the range of values (4,500 to 6,500 Btu per pound) reported by FAL19 and is slightly higher than
the 4,800 Btu per pound value reported in EPA's MSW Fact Book.20 For the energy content of RDF, we
used a value of 5,700 Btu per pound of RDF combusted.21 This estimate is within the range of values
(4,800 to 6,400 Btu per pound) reported by the DOE's National Renewable Energy Laboratory (NREL).22
For the energy content of specific materials in MSW, we consulted three sources: (1) EPA's MSW Fact
Book (a compilation of data from primary sources), (2) a report by Environment Canada,23 and (3) a
report by Argonne National Laboratories.24 We assume that the energy contents reported in the first two
of these sources were for materials with moisture contents typically found for the materials in MSW (the
sources implied this but did not explicitly state it). The Argonne study reported energy content on a dry
weight basis.
Combustion system efficiency: To estimate the combustion system efficiency of mass burn
plants, we used a net value of 550 kilowatt-hours (kwh) generated by mass burn plants per ton of mixed
MSW combusted.25 To estimate the combustion system efficiency of RDF plants, we evaluated three
sources: (1) data supplied by an RDF processing facility located in Newport, Minnesota; (2) the
Integrated Waste Services Association (IWSA) report Waste-to-Energy Directory: Year 2000; and (3) the
U.S. Department of Energy's (DOE) National Renewable Energy Laboratory. We used the Newport
Processing Facility's reported net value of 572 kwh generated per ton of RDF for two reasons.26 First,
18 Telephone conversation among representatives of Integrated Waste Services Association, American Ref-
Fuel, and IGF Consulting, October 28,1997.
19 Franklin Associates, Ltd. 1994. The Role of Recycling in Integrated Solid Waste Management to the Year
2000 (Stamford, CT: Keep America Beautiful, Inc.), pp. 1-16.
20 U.S. Environmental Protection Agency, Office of Solid Waste. 1995. MSW Fact Book, Version 2.0
(Washington, D.C.: U.S. Environmental Protection Agency).
21 Note that this is a value reported by an RDF facility located in Newport, Minnesota; the data were
provided by the Minnesota Office of Environmental Assistance. Facsimile from Karen Harrington, Minnesota Office
of Environmental Assistance to ICF Consulting, October 1997.
22 U.S. Department of Energy, National Renewable Energy Laboratory. 1992. Data Summary of Municipal
Solid Waste Management Alternatives Volume IV: Appendix B - RDF Technologies (Springfield, VA: National
Technical Information Service, NREL/TP-431-4988D), p. B-5. ;
23 Procter and Redfern, Ltd. and ORTECH International. 1993. Estimation of the Effects of Various
Municipal Waste Management Strategies on Greenhouse Gas Emissions, Part II (Ottawa, Canada: Environment
Canada, Solid Waste Management Division,, and Natural Resources Canada, Alternative Energy Division).
24 Gaines, Linda, and Frank Stodolsky. 1993. Mandated Recycling Rates: Impacts on Energy Consumption
and Municipal Solid Waste Volume (Argonne, IL: Argonne National Laboratory), pp. 11 and 85.
25 Note that this is the value reported by Integrated Waste Services Association in its comments to the draft
version of the first edition of this report. This value is within the range of values reported by others in response to
this draft. Letter received from Maria Zannes, Integrated Waste Services Association, Washington, DC, August 25,
1997.
26
The net energy value reported accounts for the-estimated energy required to process MSW into RDF and
the estimated energy consumed by the RDF combustion facility.
87
-------
this value is within the range of values reported by the other sources. Second, the Newport Processing
Facility provided a complete set of data for evaluating the overall system efficiency of RDF plants.27
Next, we considered losses in transmission and distribution of electricity. Using a transmission
and distribution loss rate of 5 percent,28 we estimated that 523 kwh are delivered per ton of waste
combusted at mass burn facilities, and 544 kwh are delivered per ton of waste input at RDF facilities
We then used the value for the delivered kwhs per ton of waste combusted to derive the implicit
combustion system efficiency (i.e., the percentage of energy in the waste that is ultimately delivered in
the form of electricity). To determine this efficiency, we first estimated the Btu of MSW needed to
deliver 1 kwh of electricity. We divided the Btu per ton of waste by the delivered kwh per ton of waste to
obtain the Btu of waste per delivered kwh. The result is 19,200 Btu per kwh for mass burn and 21,000
Btu per kwh for RDF. Next we divided the physical constant for the energy in 1 kwh (3,412 Btu) by the
Btu of MSW and RDF needed to deliver 1 kwh, to estimate the total system efficiency at 17.8 percent for
mass burn and 16.3 percent for RDF (Exhibit 6-2, columns "d" and "e").29
Electric utility carbon emissions avoided: To estimate the avoided utility CO2 from waste
combustion, we used the results in columns "c" and "d," together with a "carbon coefficient" of 0.081
MTCE emitted per million Btu of utility-generated electricity (delivered), based on the national average
fossil fuel mix used by utilities30 as shown in Exhibits 6-3 and 6-4. This approach uses the average fossil
fuel mix as a proxy for the fuels displaced at the margin when utility-generated electricity is displaced by
electricity from WTE plants. In other words, we assume that nuclear, hydropower, and other non-fossil
sources generate electricity at essentially fixed rates; marginal demand is met by fossil sources.31 The
actual carbon reductions could vary depending on which type of fuel used to generate electricity is
displaced at the margin. The resulting estimates for utility carbon emissions avoided for each material are
shown in columns "g" and "h" of Exhibit 6-2.
6.1.6 Approach to Estimating CO2 Emissions Avoided Due to Increased Steel Recycling
Next, the study estimated the avoided CO2 emissions from increased steel recycling made
possible by steel recovery from WTE plants for (1) mixed MSW and (2) steel cans. Note that we did not
credit increased recycling of non-ferrous materials, because of lack of data on the proportions of those
materials being recovered. The result tends to overestimate net GHG emissions from combustion.
For mixed MSW, we estimated the amount of steel recovered per ton of mixed MSW combusted,
based on (1) the amount of MSW combusted in the United States, and (2) the amount of steel recovered,
post-combustion. Ferrous metals are recovered at approximately 83 WTE facilities in the United States
27 The data set included estimates on the composition and amount of MSW delivered to the processing
facility, as well as estimates for the heat value of RDF, the amount of energy required to process MSW into RDF,
and the amount of energy used to operate the RDF facility.
28 Personal communication among representatives of Integrated Waste Services Association, American Ref-
Fuel, and ICF Consulting, October 28,1997.
29 Note that the total system efficiency is the efficiency of translating the energy content of the fuel into the
energy content of delivered electricity. The estimated system efficiencies of 17.8 and 16.3 percent reflect losses in
(1) converting energy in the fuel into steam, (2) converting energy in steam into electricity, and (3) delivering
electricity. The losses in delivering electricity are the transmission and distribution losses, estimated at 5 percent.
30 Value estimated using data from the Energy Information Administration, Annual Energy Review 2000
(Washington, DC: U.S. Government Printing Office, DOE/EIA-0384(2000)) August 2001.
31 Non-fossil sources are expected to meet baseload energy requirements because of the financial incentive
for these energy sources to generate at capacity. In general, the marginal cost of producing more power from these
sources is minimal compared to the capital costs associated with establishing the facility.
88
-------
and at seven RDF processing facilities that do not generate power on-site. These facilities recovered a
total of nearly 789,000 tons per year of ferrous metals in 2000.32 By dividing 789,000 tons (total U.S.
steel recovery at combustors) by total U.S. combustion of MSW, which is approximately 30 million tons,
we estimated that 0.03 tons of steel are recovered per ton of mixed MSW combusted (as a national
average)..
For steel cans, we first estimated the national average proportion of steel cans entering WTE
plants that would be recovered. As noted above, approximately 90 percent of MSW destined for
combustion goes to facilities with a ferrous recovery system. At these plants, approximately 98 percent of
the steel cans would be recovered. We multiplied these percentages to estimate the weight of steel cans
recovered per ton of steel cans combusted—about 0.88 tons per ton.
Finally, to estimate the avoided CO2 emissions due to increased recycling of steel, we multiplied
(1) the weight of steel recovered by (2) the avoided CO2 emissions per ton of steel recovered. The result
was an estimated avoided CO2 emissions of approximately 0.43 MTCE per ton for steel cans and 0.01
MTCE per ton for mixed MSW, as shown in column "d" of Exhibit 6-5.
32
Facilities.
Integrated Waste Services Association, The 2000IWSA Waste-To-Energy Directory of United States
89
-------
Exhibit 6-3
Estimating the Emission Factor for Utility-Generated Electricity
Electric Utility Consumption of Fossil Fuels to Generate Electricity
Coal (Quadrillion Btu)
Natural Gas (Quadrillion Btu)
Petroleum (Quadrillion Btu)
Total (Quadrillion Btu)
Energy Value of one Quadrillion Btu
(measured in Kilowatt-hours)
Total (Billion kwh)
Net Generation: Before Transmission and Distribution Losses (Fossil Fuels Only)
Coal (Billion kwh)
Natural Gas (Billion kwh)
Petroleum (Billion kwh)
Total (Billion kwh)
Generation Efficiency (Fossil Fuels Only)
Generation Output (Billion kwh)
Consumption (Billion kwh)
Efficiency (Percent)
Efficiency of Energy Conversion From Fossil Fuels to Delivered Electricity
Transmission and Distribution Losses (TDL) (Percent)
Delivered Electricity Efficiency (Percent)
Efficiency of Energy Conversion and Delivery for Fossil Fuels (Percent)
Estimated Emission Factor for Delivered Electricity
(MTCE/MBtu of Electricity Delivered)
Weighted Average Emission Factor of the U.S. Mix of Fuels Used to Generate Electricity
(Kilograms of Carbon in CO2 per Million Btu Consumed)
Weighted Average Emission Factor (MTCE/million Btu)
Efficiency of Energy Conversion and Delivery (Percent)
Emission Factor for Delivered Electricity (MTCE/million Btu)
Value
17.5
3.1
0.8
21.4
2.9E+11
6,268
1,692
290
72
2,054
2,054
6,268
33%
9%
91%
30%
All Fuels
16.38
0.01638
30%
0.05493
Value
Fossil Fuels Only
24.04
0.02404
i 30%
0.08060
Source
DOE, EIA, 'Annual Energy Review: 2000," July 2001, Diagram 5.
DOE, EIA, 'Annual Energy Review: 2000,'July 2001, Diagram 5.
DOE, EIA, 'Annual Energy Review. 2000,' July 2001, Diagram 5.
The sum of coal, natural gas, and petroleum.
DOE, EIA, "Form EIA 1605 (1997)," Appendix E.
(21.44 Quad Btu) x (2.92875x1011 kWh/Quad Btu) / (109 kwh/Billion kwh)
DOE, EIA, 'Annual Energy Review: 2000," August 2001, Table 8.3.
DOE, EIA, 'Annual Energy Review: 2000," August 2001, Table 8.3.
DOE, EIA, "Annual Energy Review: 2000," August 2001, Table 8.3.
The sum of coal, natural gas, and petroleum.
Calculated above.
Calculated above.
Generation Output / Consumption, i.e. 2,067 / 6,279.
DOE, EIA, "Annual Energy Review: 2000," August 2001, "Electricity Notes."
Calculated as 100 percent (Deliverable Electricity) - 9 percent (TDL)
Generation Efficiency x Delivered Electricity Efficiency, i.e., 0.33 x 0.91.
Exhibit 6-4 of this report.
Converting kilograms of carbon (kg C) to metric tons of carbon (MTC).
Calculated above.
Weighted Average Emission Factor (MTCE/million Btu) / Conversion
Efficiency.
90
-------
Exhibit 6-4
Estimating the Weighted Average Carbon Coefficient of the
U.S. Average Mix of Fuels Used to Generate Electricity
(MTCE/Million Btu)
Fuel
Coal
Natural Gas
Petroleum***
Nuclear
Hydroelectric
Other
Total
Weighted Average - All Fuels
Weighted Average - Fossil Fuels
Net Generation*
(Billion kwh)
1,692
290
72
705
253
2
3,015
Percentage of Generation:
All Fuels (%)
56.1%
9.6%
2.4%
23.4%
8.4%
0.1%
100%
Percentage of
Generation: Fossil
Fuels (%)
82%
14%
4%
100%
Carbon
Coeificents**
(Kg CE Emitted
Per Million Btu
Consumed)
25.78
14.48
21.51
0
0
0
NA
16.38
24.04
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
* Source: ElA's Annual Energy Review: 2000, Table 8.3 Electricity Net Generation at Electric Utilities, 1949-2000," for 2000.
** Values include fugitive methane emissions (weighted by the GWP of methane).
*** The carbon coefficient for residual fuel is used as a proxy for petroleum.
91
-------
Exhibit 6-5
Avoided GHG Emissions Due to Increased Steel Recovery from
MSW at WTE Facilities
(a)
Material Combusted
Aluminum Cans
Steel Cans
Glass
HOPE
LOPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed MSW
(b)
Tons of Steel
Recovered Per
Ton of Waste
Combusted
(Tons)
0.00
0.88
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.02
(c)
Avoided CO2
Emissions Per
Ton of Steel
Recovered
(MTCE/Ton)
0.00
0.49
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.49
(d)
Avoided CO2
Emissions Per
Ton of Waste
Combusted
(MTCE/Ton)
0.00
0.43
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
*The value in column "d" is a national average and is weighted to reflect 98 percent recovery at the
90 percent of facilities that recover ferrous metals.
92
-------
6.2 RESULTS
The results of this analysis are shown in Exhibit 6-6. The results from the last columns of
Exhibits 6-1, the last two columns of Exhibit 6-2, and the last column of Exhibit 6-3 are shown in
columns "b" through "e" in Exhibit 6-6. The net GHG emissions from combustion of each material at
mass burn and RDF facilities are shown in columns "f' and "g," respectively. These net values represent
the gross GHG emissions (column "b"), minus the avoided GHG emissions (columns "c," "d," and
"e"). As stated earlier, these estimates of net GHG emissions are expressed for combustion in absolute
terms. They are not values relative to some other waste management option. They are expressed in terms
of short tons of waste input (i.e., tons of waste prior to processing).
We estimate that combustion of mixed MSW at mass burn and RDF facilities reduces net post-
consumer GHG emissions to -0.04 and -0.03 MTCE per ton, respectively. Combustion of paper products
has negative net post-consumer GHG emissions ranging from -0.14 to -0.22 MTCE per ton at mass burn
facilities and from -0.13 to -0.20 MTCE per ton at RDF facilities. Net GHG emissions are negative
because CO2 emissions from burning paper are not counted (because they are biogenic) and fossil fuel
burning by utilities to generate electricity is avoided. Likewise, combustion of medium-density
fiberboard and dimensional lumber also results in negative net GHG emissions, with both equaling -0.23
MTCE at mass burn facilities and -0.21 at RDF facilities. Finally, net GHG emissions for food discards
and yard trimmings.(two other forms of biomass) are also negative, but of a smaller magnitude (-0.05 and
-0.07 MTCE per ton of material, respectively, for mass burn and -0.05 and -0.06 MTCE per ton of
material, respectively, for RDF). .
Combustion of plastics results in substantial net GHG emissions, estimated from 0.21 to 0.27
MTCE per ton of material combusted for mass burn facilities, and from 0.25 to 0.30 MTCE per ton of
material input to RDF facilities. This result is primarily because of the high content of non-biomass
carbon in plastics. Also, when combustion of plastic results in electricity generation, the utility carbon
emissions avoided (due to displaced utility fossil fuel combustion) are much lower than the carbon
emissions from the combustion of plastic. This result is largely due to the lower system efficiency of
WTE plants, compared with electric utility plants. Recovery of ferrous metals at combustors results in
negative net GHG emissions, estimated at -0.42 MTCE per ton of steel cans, due to the increased steel
recycling made possible by ferrous metal recovery at WTE plants.
6.3 LIMITATIONS OF THE ANALYSIS
The certainty of the analysis presented in this chapter is limited by the reliability of the various
data elements used. The, most significant limitations are as follows: .:••
• Combustion system efficiency of WTE plants may be improving. If efficiency improves,
more utility CO2 will be displaced per ton of waste combusted (assuming no change in utility
emissions per kwh), and the net GHG emissions from combustion of MSW will decrease.
• Data for the RDF analysis were provided by the Minnesota Office of Environmental
Assistance and were obtained from a single RDF processing facility and a separate RDF
combustion facility. Research indicates that each RDF processing and combustion facility is
different. For example, some RDF combustion facilities may generate steam for sale off-site,
which can affect overall system efficiency. In addition, the amount of energy required to
process MSW into RDF and the amount of energy used to operate RDF combustion facilities
can be difficult to quantify and can vary among facilities on a daily, seasonal, and annual
basis. Thus, the values used for the RDF analysis should be interpreted as approximate
values.
93
-------
Exhibit 6-6
Net GHG Emissions from Combustion at WTE Facilities
(a)
Material Combusted
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Yard Trimmings
Food Discards
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed MSW
(b)
Gross GHG
Emissions Per
Ton Combusted
(MTCE/Ton)
0.01
0.01
0.01
0.77
0.77
0.56
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.12
(c)
Avoided Utility
C02 Per Ton
Combusted at
Mass Burn
Facilities
(MTCE/Ton)
-0.01
-0.01
-0.01
0.54
0.54
0.28
0.20
0.15
0.23
0.20
0.23
0.20
0.24
0.24
0.08
0.07
0.20
0.20
0.19
0.14
(d)
Avoided Utility
C02 Per Ton
Combusted at
RDF Facilities
(MTCE/Ton)
-0.01
-0.01
-0.01
0.49
0.49
0.25
0.18
0.14
0.21
0.18
0.21
0.18
0.22
0.22
0.07
0.06
0.19
0.18
0.17
0.13
(e)
Avoided CO2
Emissions Per
Ton Combusted
Due to Steel
Recovery
(MTCE/Ton)
0.00
0.43
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
NA
NA
0.01
(f)
(f = b-c-e)
Net GHG
Emissions from
Combustion at
Mass Burn
Facilities
(MTCE/Ton)
0.02
-0.42
0.01
0.23
0.23
0.28
-0-19
-0.13
-0.21
-0.18
-0.21
-0.18
-0.22
-0.22
-0.06
-0.05
-0.19
-0.18
-0.17
-0.04
(g)
(g = b-d-e)
Net GHG
Emissions from
Combustion at
RDF Facilities
(MTCEm>n)
0.02
-0.42
0.01
0.28
0.28
0.31
-0.17
-0.12
-0.19
-0.16
-0.19
-0.16
-0.20
-0.20
-0.06
-0.04
-0.17
-0.17
-0.15
-0.02
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
94
-------
The reported ranges for N2O emissions were broad. In some cases the high end of the range was
10 times the low end of the range. Research has indicated that N2O emissions vary with the type of waste
burned. Thus, the average value used for mixed MSW and for all MSW components should be
interpreted as an approximate value.
• For mixed MSW, the study assumed that all carbon in textiles is from synthetic fibers
derived from petrochemicals (whereas, in fact, some textiles are made from cotton, wool, and
other natural fibers). Because we assumed that all carbon in textiles is non-biogenic, we
counted all of the CO2 emissions from combustion of textiles as GHG emissions. This
assumption will slightly overstate the net GHG emissions from combustion of mixed MSW,
but the magnitude of the error is small because textiles represent only a small fraction of the
MSW stream. Similarly, the MSW category of "rubber and leather" contains some biogenic
carbon from leather. By not considering this small amount of biogenic carbon, the analysis
slightly overstates the GHG emissions from MSW. combustion.
• Because the makeup of a given community's mixed MSW may vary from the national
average, the energy content also may vary from the national average energy content used in
this analysis. For example, MSW from communities with a higher- or lower-than-average
recycling rate may have a different energy content, and MSW with more than the average
proportion of dry leaves and branches will have a higher energy content.
• In this analysis, we used the national average recoyery rate for steel. Where waste is sent to a
WTE plant with steel recovery, the net GHG emissions for steel cans will be slightly lower
(i.e., more negative). Where waste is sent to a WTE plant without steel recovery, the net
GHG emissions for steel cans will be the same as for aluminum cans (i.e., close to zero). We
did not credit increased recycling of non-ferrous materials, because of a lack of information
on the proportions of those materials. This assumption tends to result in overstated net GHG
emissions from combustion.
• This analysis used the national average fossil fuel mix for electricity as the proxy for fuel
displaced at the margin when WTE plants displace utility electricity. If some other fuel or
mix of fuels is displaced at the margin (e.g., coal), the avoided utility CO2 would be different
(e.g., for coal, the avoided utility CO2 would be about 0.01 MTCE per ton higher for mixed
MSW, and the net GHG emissions would be -0.05 MTCE instead of -0.04 MTCE per ton).
95
-------
-------
7. LANDFILLING
This chapter presents estimates of GHG emissions and carbon storage from landfilling the
materials considered in this analysis. For this study, we estimated the CEU emissions, transportation-
related CO2 emissions, and carbon storage that will result from landfilling each type of organic waste and
mixed MSW. The GHG accounting principles used in the analysis follow.1
• When food discards, yard trimmings, paper, and wood are landfilled, anaerobic bacteria
degrade the materials, producing CH4 and CO2. CH4 is counted as an anthropogenic GHG,
because even though it is derived from sustainably harvested biogenic sources, degradation
would not result in CEU emissions if not for deposition in landfills. The CO2 is not counted
as a GHG in this context because if it were not emitted from landfills, it would be produced
through natural decomposition. Because metals do not contain carbon, they do not generate
CH4 when landfilled. Plastics do not biodegrade, and therefore do not generate any
• Transportation of waste materials to a landfill results in anthropogenic CO2 emissions, due to
the combustion of fossil fuels in the vehicles used to haul the wastes.
Because food discards, yard trimmings, and paper are not completely decomposed by
anaerobic bacteria, some of the carbon in these materials is stored in the landfill. Because
this carbon storage would not normally occur under natural conditions (virtually all of the
organic material would degrade to CO2, completing the photosynthesis/respiration cycle),
this is counted as an anthropogenic sink.2
We developed separate estimates of emissions from landfills without gas recovery systems, those
that flare CBU, those that combust CKLt for energy recovery, and from the national average mix of these
three categories. Our national average emission estimate accounts for the extent to which CHt will be
flared at some landfills and combusted for energy recovery at others.3
From the standpoint of post-consumer GHG emissions, landfilling some materials — including
magazines/third-class mail, newspaper, phonebooks, dimensional lumber, medium-density fiberboard,
leaves, and branches-results in net storage (i.e., carbon storage exceeds CHt plus transportation energy
emissions) at all landfills, regardless of whether gas recovery is present. At the other extreme, office
paper, textbooks, and food discards result in net emissions regardless of landfill gas collection and
recovery practices. The remaining materials have net post-consumer emissions that are either very low
(aluminum, steel cans, and plastics have transportation-related emissions of 0.01 MTCE per ton,
regardless of whether gas collection is present) or borderline, depending on whether the landfill has gas
1 These principles are described in broad terms in Section 1.5 of this report.
2 However, carbon in plastic that remains in the landfill is not counted as stored carbon, because it is of
fossil origin.
3 Currently, most landfill CHt recovery in the United States—both for flaring and electricity—is occurring
in response to a 1996 EPA rule that requires a well-designed and well-operated landfill gas collection system at
landfills that (1) have a design capacity of at least 2.5 million metric tons or 2.5 million cubic meters; (2) are
calculated to emit more than 50 metric tons of non-CHj organic compounds per year; and (3) received waste on or
after November 11, 1987 (Federal Register, Vol. 61, No. 49, p. 9905, March 12,1996). For the year 2000, an
estimated 43 percent of landfill CHt was generated at landfills with landfill gas recovery systems subject to these
requirements or installed on a voluntary basis.
97
-------
recovery (e.g., mixed MSW has net emissions at landfills without gas recovery, but net carbon storage at
landfills with gas recovery). .<
7.1 EXPERIMENTAL VALUES FOR CH4 GENERATION AND CARBON STORAGE
To estimate CHU emissions and carbon storage from landfilling of specific materials, we used
data from laboratory experiments conducted by Dr. Morton Barlaz.4 The experiments provided data on
(1) the amount of CBU generated by each type of organic material, when digested by bacteria in anaerobic
conditions simulating those in a landfill; and (2) the amount of carbon remaining, undecomposed (i.e.,
stored) at the end of the experiment.
7.1.1 Experimental Design
Dr. Barlaz placed each type of organic waste and mixed MSW in separate reactor vessels, in
which he maintained anaerobic conditions similar to those in a landfill, but controlled to favor maximum
CH} generation. Dr. Barlaz measured the amount of CEU generated in each reactor and the amount of
undecomposed carbon remaining in each reactor at the end of the experiment. Each material was tested in
four reactors, and the results from each reactor were averaged.5
At the start of the experiment, Dr. Barlaz dried a sample of each material and analyzed the
amount of cellulose, hemicellulose, and lignin (and, for food discards, protein) in each material.
Cellulose, hemicellulose, and protein partly decompose in a landfill, resulting in CHU generation. Lignin
is relatively stable and non-decomposable under anaerobic conditions.
Portions of each material were weighed, placed in two-liter plastic containers (i.e., reactors), and
allowed to decompose anaerobically under warm, moist conditions designed to accelerate decomposition.
The reactors were seeded with a small amount of well-decomposed refuse containing an active
population of CBLrproducing microorganisms (the "seed"), to ensure that CHU generation was not
limited due to an insufficient population of microorganisms. To promote degradation, water was cycled
through each reactor. Nitrogen and phosphorus were added so that CHU generation would not be limited
by a lack of these nutrients.
The reactors were allowed to run for periods varying from three months to two years. The
experiment ended for each reactor when one of two conditions was met: (1) no measurable CH4 was
being emitted (i.e., any CBU that was being emitted was below the detection limits of the analytical
equipment); or (2) a curve generated mathematically from an analysis of the reactor's prior CHU
generation indicated that the reactor had produced at least 95 percent of the CBU that it would produce if
allowed to run indefinitely.
Dr. Barlaz measured the amount of CHU generated during the experimental period and subtracted
the amount of CBU attributable to the seed. At the end of the experiment, he opened the reactors, drained
the leachate, dried and weighed the contents, and analyzed the percentage composition of cellulose,
hemicellulose, and lignin (and, for food discards, protein) in the remaining contents. He then measured
the percentage of total volatile solids in the remaining contents. This amount included the cellulose,
hemicellulose, lignin, and protein, and any other carbon-containing components such as waxes and
tannins.
4 Barlaz, M.A. 1997. "Biodegradative Analysis of Municipal Solid Waste in Laboratory-Scale Landfills,"
EPA 600/R-97-071. Dr. Barlaz's work was funded by EPA's Air and Energy Engineering Research Laboratory.
5 Barlaz, op. cit.
98
-------
The experimental results were used to estimate the amount of carbon remaining in the reactor
that was attributable to the seed6 and the amount attributable to the material. The experiment was
assumed to reflect actual landfill conditions, and the organic carbon remaining undegraded in the reactors
was assumed to remain undegraded over the long term in landfills, i.e., it would be stored.
Dr. Barlaz's experiment did not specifically test all of the paper grades described in this report.
He did evaluate four specific grades: newspaper, corrugated boxes, office paper, and coated paper. We
identified proxies for five additional material types for which we had no experimental data. We assumed
that magazines placed in a landfill will have characteristics similar to those observed for coated paper.
Similarly, we assumed that phonebooks and textbooks would behave in the same way as newspaper and
office paper, respectively. Experimental results for branches were used as a proxy for dimensional
lumber and medium-density fiberboard.
As discussed in Section 4.2, we included the following three definitions of mixed paper among
the materials analyzed in this report:
• Broadly defined mixed paper, which includes almost all printing-writing paper, folding
boxes, and most paper packaging;
• Residential mixed paper, which includes the typical mix of papers from residential curbside
pick-up (e.g., high-grade office paper, magazines, catalogs, commercial printing, folding
cartons, and a small amount of old corrugated containers); and
• Mixed paper from offices, which includes copy and printer paper, stationary and envelopes,
and commercial printing.
To develop estimates of CH4 emissions and carbon storage for these three categories of mixed paper, we
used the detailed characterization of mixed paper (shown in Exhibit 4-2) developed by Franklin
Associates, Ltd., and we assigned analogs among the four paper grades tested by Dr. Barlaz. Exhibit 7-1
characterizes the composition of the two products made from mixed paper: boxboard (made using either
a broad or a residential mix of recycled paper) and paper towels (made from recycled office paper).
Emissions were calculated using these characterizations of the mixed paper grades and the values
obtained from Dr. Barlaz's experiment for newspaper, corrugated boxes, office paper, and coated paper.7
Dr. Barlaz tested seed alone to be able to control for the amount of CH4 generation and carbon storage
that was attributable to the seed.
7 Note that Exhibits 7-2 through 7-4 do not show mixed paper; however, mixed paper is shown in Exhibits
7-6 through 7-8. Exhibits 7-2 through 7-8 appear at the end of the chapter.
99
-------
Exhibit 7-1
Proxies for Composition Mixed Paper (Percent)
Paper Grade
Newspaper1
Corrugated Boxes 2
Office Paper 3
Coated Paper 4
Total
Broad Definition for
Mixed Paper
24
48
20
8
100
Mixed Paper from
Residential Sources
23 .
'•' -'•• ' "•• 53
14
10
100
Mixed Paper from
Offices
21
'5
38
36
100
Explanatory Notes:
1 Includes newspaper, uncoated groundwood paper, recycled folding boxes, and set-up boxes.
2 Includes virgin and recycled corrugated boxes.
3 Includes uncoated free sheet paper, cotton fiber paper, bleached bristols, unbleached kraft folding boxes, bleached
kraft folding boxes, bleached bags and sacks, unbleached bags and sacks, and unbleached wrapping paper.
4 Includes coated free sheet paper and coated groundwood paper.
7.1.2 CEU Generation: Experimental Data and Adjusted Values
The amount of CHU generated by each type of organic material (after deducting the CBU
attributable to the seed), is shown in column "b" of Exhibit 7-2.
As a check on bis experimental results, Dr. Barlaz estimated the amount of CEU that would have
been produced if all of the cellulose, hemicellulose, and protein from the waste material that was
decomposed during the experiment had been converted to equal parts of CHU and COi (CELrproducing
microorganisms generate equal amounts, by volume, of CBU and CO2 gas).8 Dr. Barlaz referred to this
amount as the material's "CHU potential." He then calculated the percentage of the CHU potential for
each material accounted for by the sum of (1) the measured CHU generation, and (2) the amount of CHU
that could be formed from the carbon in the leachate that was removed from the reactor and from the
carbon in the refuse that remained in the reactor at the end of the experiment.9 The resulting percentages
of the CH4 potential are shown in column "c" of Exhibit 7-2. CHU potential not accounted for could be
due to either (1) leaks of CHU; (2) measurement error; or (3) carbon in the cell mass of microorganisms
(which was not measured). .
CBU recovery was below 85 percent of the CHU potential for five materials: coated paper, office
paper, food discards, leaves, and branches..In using Dr. Barlaz's data, we needed to make a choice
regarding how to allocate this missing carbon. We chose to assume that some of it had been converted to .
microorganism cell mass, and the remainder had been degraded. Dr. Barlaz postulated a higher CH4 yield
based on assumptions that (1) 5 percent of the carbon in cellulose and hemicellulose (and protein in the
case of food discards) that was degraded was converted into the cell mass of the microbial population;
and (2) 90 percent of the carbon-containing compounds that were degraded but not converted to cell mass
were converted to equal parts of CHU and CO2. The "corrected yields," based on these assumptions, are
shown in column " d" of Exhibit 7-2.
Ibid. Lignin was not considered in this check because cellulose, hemicellulose, and protein account for
nearly all of the CHU generated.
9 Note that any carbon that was converted to cell mass in microorganisms was not considered in this
calculation.
100
-------
We decided, in consultation with Dr. Barlaz, to use the "corrected yields" for leaves, branches,
and office paper because we believed that these values were more realistic than the measured yields.10'u
The CHU values that we used for each material (either the measured yield, or the "corrected"
yield) are shown again in column "f' of Exhibit 7-2. In order to maintain consistent units with the other
parts of our analysis, we converted the units for CBLt generation from milliliters per dry gram of waste, to
MTCE per wet ton of waste.12 The resulting values are shown in column "g" of Exhibit 7-2. The value
for yard trimmings is a weighted average of the values for grass, leaves, and branches, based on an
assumption that yard trimmings are composed of 50 percent grass, 25 percent leaves, and 25 percent
branches (on a wet weight basis).
7.1.3 Carbon Storage: Experimental Data and Calculations
Carbon storage was estimated by calculating the amount of carbon remaining in each reactor at
the end of the experiment and then subtracting the amount of carbon remaining that was attributable to
the seed. The difference between the two values is the amount of carbon from the waste material that
remained in the reactor, undecomposed, at the end of the experiment. Because the conditions in the
reactor simulated landfill conditions (favorable to optimized decomposition), approximately this amount
of carbon would be stored if the material were landfilled. Carbon storage for each material is presented in
Exhibit 7-3.13
7.2 FATES OF LANDFILL CHL,: CONVERSION TO CO2, EMISSIONS, AND FLARING OR
COMBUSTION WITH ENERGY RECOVERY
In this analysis, we accounted for (1) the conversion in the landfill of some portion of landfill
CELt to CO2 and (2) the capture of CELt, either for flaring or for combustion with energy recovery (in
either case, the captured CBU is converted to CO2).14 Exhibit 7-4 presents this analysis.
The exhibit begins with the CELt generation per wet ton of each material, which is shown in
column "b" (the values were simply copied from the last column of Exhibit 7-2). Columns "c" through
"k" calculate net GHG emissions from CELt generation for each of three categories of landfills: (1)
landfills without landfillgas (LFG) recovery; (2) landfills with LFG recovery that flare LFG; and (3)
landfills with LFG recovery, which generate electricity from the LFG. Columns "1" through "n" show
10 The corrected yield was not available for coated paper/magazines. For food discards, even though the
CELt potential recovery percentage was lower than 85 percent, we used the measured yield, as shown in column "b."
We made this choice for food discards because the "corrected yield" for food discards was greater than the
maximum possible yield (shown in column "e" of the exhibit). Dr. Barlaz had calculated the maximum possible
yield for each material based on the CELt yield if all of the cellulose, hemicellulose, and protein in the material (1)
decomposed and (2) was converted to equal parts of CELf and CO2.
11 Note that EPA's Office of Research and Development (ORD) uses the same data as the basis for its
estimation of CELt yields. In that analysis, ORD does not use "corrected" values for materials with low CELt
recovery, but rather uses observed experimental values fbr all materials.
12 To make the conversion, we used the ratio of dry weight to wet weight for each material and a global
warming potential of 21 for CHLj.
13 The approach for estimating carbon storage is more fully described in, Barlaz, Morton, "Carbon Storage
During Biodegradation of Municipal Solid Waste Components in Laboratory-Scale Landfills," paper submitted for
publication, Department of Civil Engineering, North Carolina State University, Raleigh, NC, 1997.
14 The CO2 that is emitted is not counted as a GHG because it is biogenic in origin (as described in "CO2
Emissions from Biogenic Sources: in Chapter 1).
101
-------
the estimated percentage of landfills in each category in 2000.15'16 The final column shows the weighted
average GHG emissions from CH^ generation across all types of landfills.
To estimate MSW CB^ emissions from each category of landfill, we first estimated the
percentage of landfill CEU that is oxidized near the surface of the landfill. We estimated that 10 percent
of the landfill CHL* generated is either chemically oxidized or converted by bacteria to CO2,17 and the
remaining 90 percent is available for atmospheric CEU emissions.
To estimate MSW CHU emissions from landfills with LFG recovery, we used the assumption that
these landfills have an average LFG recovery efficiency of 75 percent.18 We then calculated avoided
utility GHG emissions from landfills with electricity generation. Because energy recovery systems
experience down time, during which CKU is flared rather than used to generate electricity, we
incorporated a 15 percent system efficiency loss into our estimates for avoided utility emissions.19
We also estimated the percentage of MSW disposed in each category of landfill in 2000.
According to our estimates, 49 percent of all landfill CKj was generated at landfills with recovery
systems, and the remaining 51 percent was generated at landfills without LFG recovery.20 Of the 49
percent of all CBU generated at landfills with LFG recovery, 49 percent (or 24 percent of all CEU) was
generated at landfills that use LFG to generate electricity, and 51 percent (or 25 percent of all CBU) at
landfills that flare LFG.21'22
Our results are shown in the final column of Exhibit 7-4. The materials with the highest rates of
net GHG emissions from CEU generation, as shown in column "o"—corrugated boxes, office paper, and
textbooks—also have the highest gross CELt generation, as shown in column "b." The recovery of CH4 at
landfills reduces the CH* emissions for each material in proportionate amounts but does not change ,the
15 Draft £7.5. Climate Action Report - 2001 (CAR). At the time of publication of this report, the CAR was
still being reviewed; however, EPA expected that these estimates will not change in the final version.
16 Note that estimates of percent CUt generation at landfills with recovery have decreased since the first
edition of this report was published (hi the first edition, we estimated that 54 percent of CH4 would be generated at
landfills with recovery). This difference is because the first edition relied on 1995 projections of year 2000
generation and recovery, whereas this version uses the most recent estimates of conditions in 2000.
17 An oxidation rate of 10 percent is cited by Liptay, K., J. Chanton, P. Czepiel, and B. Mosher, "Use of
Stable Isotopes to Determine Methane Oxidation in Landfill Cover Soils," Journal of Geophysical Research, April
1998,103(D7), pp. 8243-8250; and Czepiel, P.M., B. Mosher, P.M. Grill, and R.C. Harriss. 1996. "Quantifying the
Effects of Oxidation on Landfill Methane Emissions," Journal of Geophysical Research, 101, pp. 16721-16729.
18 Several commenters on the draft version of the first edition of this report suggested a range of values; 75
percent was most often cited as a best estimate. Moreover, EPA has used this figure in its most recent publications
[see, for example, £7.5. Methane Emissions 1990-2020: Inventories, Projections, and Opportunities for Reductions
(Washington, D.C.: U.S. EPA) September 1999]. :
19 EPA. 1999. Landfill Gas-to-Energy Project Opportunities: Background Information on Landfill Profiles,
Office of Air and Radiation, EPA 430-K-99-002, pp. 3-13.
20 Based on data on (1) year 2000 MSW landfill CH* generation of 72.7 million MTCE (from draft U.S.
Climate Action Report - 2001), (2) year 2000 landfill CH4 recovery of 26.7 million MTCE (also from draft U.S.
Climate Action Report - 2001), and (3) estimated landfill CHU recovery efficiency of 75 percent (from U.S. Methane
Emissions 1990-2020: Inventories, Projections, and Opportunities for Reductions).
21 Draft U.S. Climate Action Report - 2001.
22 The assumption that 49 percent of landfills recovering CEU will use it to generate electricity is subject to
change over time based upon changes in the cost of recovery and the potential payback. Additionally, new
technologies may arise that use recovered CH4 for purposes other than generating electricity.
102
-------
ranking of materials by CEU emissions. Leaves, branches, and the two wood products have the lowest
rates of net GHG emissions from CEL* generation.
7.3 UTILITY CO2 EMISSIONS AVOIDED
Exhibit 7-5 presents a list of conversion factors and physical constants used to convert CUt
combusted for electricity production to avoided CO2 emissions. Using data on Btu per cubic feet of CELj,
kwh of electricity generated and delivered per Btu, and kilograms of utility carbon avoided per Btu
delivered, we estimated that 0.18 MTCE is avoided per MTCE of Cftt combusted. This figure then was
incorporated into exhibit 7-4 to estimate net GHG emissions from landfills with electricity generation. As
mentioned earlier in this chapter, our analysis assumes that 24 percent of landfills in the United States
combust landfill CH^ for electricity generation. We also assume a 15 percent system efficiency loss,
reflecting the fact that landfill gas-to-energy facilities incur some system "down time," as shown in
column 1.
7.4 NET GHG EMISSIONS FROM LANDFILLING
To determine the net GHG emissions from landfilling each material, we summed the net GHG
emissions from CBU generation, carbon storage (treated as negative emissions), and transportation CO2
emissions. The results are shown in Exhibit 7-6. The four columns under section "e" of the exhibit may
be used by local MSW planners to estimate GHG emissions from MSW in a given community.
As the exhibit shows, the post-consumer results for organic materials vary widely. For some
materials—in particular, magazines/third-class mail, newspaper, phonebooks, dimensional lumber,
medium-density fiberboard, and yard trimmings (in particular, leaves and branches)—landfilling results
in substantial net GHG reductions. For others—including corrugated cardboard, office paper, textbooks,
and food discards—net emissions are significant. For the rest, net emissions and reductions are relatively
small.
7.5 LIMITATIONS
Perhaps the most important caveat to the analysis of GHG emissions and storage associated with
landfilling is that the results are based on a single set of laboratory experiments, those conducted by Dr.
Morton Barlaz. Although researchers other than Dr. Barlaz have conducted laboratory studies that track
the degradation of mixed MSW, his experiments were the only ones we identified that rigorously tested
materials on an individual basis. Dr. Barlaz is recognized as an expert on the degradation of different
fractions of MSW under anaerobic conditions, and his findings with respect to the CELt potential of
mixed MSW are within the range used by landfill gas developers. Nevertheless, given the sensitivity of
the landfill results to estimated CEU generation and carbon storage, we recognize that more research is
needed in this area.
Another important caveat relates to our estimate that 49 percent of MSW landfill CH4 is
generated at landfills with LFG recovery systems. The net GHG emissions from landfilling each material
are quite sensitive to the LFG recovery rate. Because of the high global warming potential of CHU, small
changes in the LFG recovery rate (for the national average landfill) could have a large effect on the net
GHG impacts of landfilling each material and the ranking of landfilling relative to other MSW
management options. The effects of different rates of LFG recovery are shown in Exhibit 7-7. Column
"b" of the exhibit shows net GHG emissions if 20 percent of waste was disposed of at landfills with
recovery. The remaining columns show net GHG emissions at increasing LFG recovery rates, up to a 60
percent rate. As the exhibit shows, the net post-consumer GHG emissions for landfilling mixed MSW
decline significantly as recovery increases. At the local level, the GHG emissions from landfilling MSW
depend on whether the local landfill has LFG recovery, as shown in Exnibit 7-6.
103
-------
Because the national average estimate of emissions is based on estimated year 2000 LFG
recovery levels, there are several limitations associated with the use of this emission factor. Fkst, because
landfill CHU generation occurs over time and has significant timing delays (i.e., CHU generation may not
begin until a few years after the waste is deposited in the landfill and can continue for many years after
the landfill is closed), the values listed in this chapter represent total CEU generated, over time, per ton of
waste landfilled. To the extent that LFG recovery rates shift dramatically over time, these shifts are not
reflected in the analysis. Second, landfills with LFG recovery may be permitted, under EPA regulations,
to remove the LFG recovery equipment when three conditions are met: (1) the landfill is permanently
closed, (2) LFG has been collected continuously for at least 15 years, and (3) the landfill emits less than
50 metric tons of non-CHU organic compounds per year.23 Although the removal of LFG recovery
equipment will permit CHU from closed landfills to escape into the atmosphere, the amounts of CHU
emitted should be relatively small, because of the relatively long time period required for LFG collection
before LFG recovery equipment is removed. Third, several methodological issues are associated with
applying the CHU generation estimates from the Climate Action Report to develop the national average
emission factors:
(1) The generation estimates in the CAR include closed landfills (generation is modeled as a
function of waste in place), whereas the estimates used in this report apply to ongoing
generation (which is routed to open landfills);
(2) Likewise, both the flaring and landfill gas-to-energy estimates also include closed landfills;
and
(3) The distribution of waste in place is not a perfect proxy for the destination of ongoing waste
generation.
CEU oxidation rate and landfill gas collection system efficiency are also important factors driving
results. We used values of 10 percent and 75 percent, respectively, as best estimates for these factors.
Reviewers who commented on the draft of the first edition of this report and sources in the literature have
reported estimates ranging from about 5 percent to 40 percent for oxidation, and from about 60 to 95
percent for collection system efficiency. We investigated the sensitivity of our results to these
assumptions, and our results are shown in Exhibit 7-8. We portray the sensitivity as a bounding analysis;
i.e., we use the combinations of variables yielding the upper bound emission factor (5 percent oxidation,
60 percent collection efficiency) and the lower bound (40 percent oxidation, 95 percent efficiency).24 As
the exhibit shows, the materials most sensitive to these variables are those with the highest CHU
generation potential, i.e., corrugated cardboard, office paper, textbooks, food discards, and mixed paper.
Sensitivity varies: the difference between upper and lower bounds ranges from 0.06 MTCE/ton for leaves
and branches to 0.43 MTCE/ton for office paper and textbooks. The post-consumer emission factors of
several materials and mixed material combinations—corrugated cardboard, grass, mixed paper, and
mixed MSW—change from having net storage under the lower bound to having net emissions under the
upper bound.
Ongoing shifts in the use of landfill cover and liner systems are likely to influence the rate of
CBU generation and collection. As more landfills install effective covers and implement controls to keep
water and other liquids out, conditions will be less favorable for degradation of organic wastes. Over the
long term these improvements may result in a decrease in CH4 generation and an increase in carbon
storage. Moreover, Dr. Barlaz believes that the CEU yields from his laboratory experiments are likely to
be higher than CHU yields in a landfill, because the laboratory experiments were designed to generate the
23
Federal Register, Vol. 61, No. 49, p. 9907.
24 The table also reports two intermediate combinations, including the best estimate values.
104
-------
maximum amount of CHU possible. If the CH4 yields used in this analysis were higher than yields in a
landfill, the net GHG emissions from landfilling organic materials would be lower than estimated here.
We assumed that once wastes are disposed in a landfill, they are never removed. In other words,
we assumed that landfills are never "mined." A number of communities have mined their landfills—
removing and combusting the waste—in order to create more space for continued disposal of waste in the
landfill. To the extent that landfills are mined in the future, it is incorrect to assume that carbon stored in
a landfill will remain stored. For example, if landfilled wastes are later combusted, the carbon that was
stored in the landfill will be oxidized to CO2 in the combustor.
Our estimate of carbon avoided utility GHG emissions per unit of CHU combusted assumes that
all landfill gas-to-energy projects are electricity producing. In reality, some projects are "direct gas"
projects, in which CHt is piped directly to the end user for use as fuel. In these cases, the CEU essentially
replaces natural gas as a fuel source. Because natural gas use is less GHG-intensive than average
electricity production, direct gas projects will tend to offset fewer GHG emissions than electricity
projects will—a fact not reflected in our analysis. .
For landfilling of yard trimmings (and other organic materials), we assumed that all carbon
storage in a landfill environment is incremental to the storage that occurs in a non-landfill environment.
In other words, we assumed that in a baseline where yard trimmings are returned to the soil (i.e., in a
non-landfill environment), all of the carbon is decomposed relatively rapidly (i.e., within several years) to
CO2, and there is no long-term carbon storage. To the extent that long-term carbon storage occurs hi the
baseline, the estimates of carbon storage reported here are overstated, and the net post-consumer GHG
emissions are understated.
Finally, our spreadsheet analysis is limited by the assumptions that were made at various steps in
the analysis, as described throughout this chapter. The key assumptions that have not already been
discussed as limitations are the assumptions used in developing "corrected" CEU yields for organic
materials in MSW. Because of the high global warming potential of CH^ a small difference between
estimated and actual CEU generation values would have a large effect on the GHG impacts of landfilling
and the ranking of landfilling relative to other MSW management options.
105
-------
Exhibit 7-2
Methane Yield for Solid Waste Components
(a)
Material
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Food Discards
Yard Trimmings
Grass
Leaves
Branches
Mixed MSW
(b)
Average
Measured
Methane Yield
(ml per dry
gm)
152.3
84.4
74.2
217.3
300.7
144.3
30.5
62.6
92.0
(c)
Percentage of
"Methane
Potential"
Accounted for
87.7
83.7
98.0
55.5
77.4
89.3
75.2
82.8
97.6
(d)
"Corrected"
Methane
Yield (ml per
dry gram)
NA
NA
NA
346.0
386.2
NA
56.0
76.3
NA
(e)
Maximum
Possible
Methane
Yield (ml per
dry gram)
279.7
NA
239.4
398.2
357.6
153.2
108.0
224.9
157.6
(f)
Selected
Methane
Yield (ml
per dry
gm)
152.3
84.4
74.2
346.0
300.7
144.3
56.0
76.3
92.0
(9)
Selected
Methane
Yield
(MTCE/Wet
Ton)
0.537
0.294
0.259
1.207
0.335
0.191
0.214
0.166
0.170
0.286
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
Note that Exhibits 7-1 to 7-3 show coated paper but not mixed paper; mixed paper is shown in Exhibits 7-5 and 7-6. The values for the different
types of mixed paper are based on the proportion of the four paper types (newspaper, office paper, corrugated cardboard, and coated paper)
that comprise the different "mixed paper" definitions.
106
-------
Exhibit 7-3
Carbon Storage for Solid Waste Components
(a)
Material
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Food Discards
Yard Trimmings
Grass
Leaves
Branches
Mixed MSW
(b)
Ratio of
Carbon
Storage to
Dry Weight
(gm C/dry
gm)
0.26
0.34
0.42
0.05
0.08
0.32
0.54
0.38
0.13
(c)
Ratio of
Dry
Weight to
Wet
Weight
0.95
0.94
0.94
0.94
0.30
0.40
0.80
0.60
0.84
(d)
(d = b * c)
Ratio of
Carbon
Storage to Wet
Weight (gm
C/wet gm)
0.25
0.32
0.39
0.05
0.02
0.23
0.13
0.43
0.23
0.11
(e)
Amount of
Carbon Stored
(MICE per Wet
Ton)
0.22
0.29
0.36
0.04
0.02
0.21
0.12
0.39
0.21
0.10
Note that more digits may be displayed than are significant. -
Explanatory Notes for Exhibit 7-3:
(1) Because MSW is typically measured in terms of its wet weight, we needed to convert the ratios for carbon stored as a fraction of dry weight to carbon stored
as a fraction of wet weight. To do this, we used the estimated ratio of dry weight to wet weight for each material. These ratios are shown in column "c" of the
exhibit. For most of the materials, we used data from an engineering handbook.25 For grass, leaves, and branches, we used data provided by Dr. Barlaz.
(2) For consistency with the overall analysis, we converted the carbon storage values for each material to units of MTCE stored per short ton of waste material
landfilled. The resulting values are shown in column "e" of the exhibit.
25 Tchobanoglous, George, Hilary Theisen, and Rolf Eliassen. 1977. Solid Wastes: Engineering Principles and Management Issues (New York:
McGraw-Hill Book Co.), pp. 58 and 60.
107
-------
Exhibit 7-4
Net GHG Emissions from CH4 Generation
(a)
Material
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Grass
Leaves
Branches
Mixed MSW
(b)
" ' CH4
Generation
(MTCE/Wet
Ton)
0.537
0.294
0.259
1.207
0.259
1.207
0.170
' 0.170
0.335
0.191
0.214
0.166
0.170
0.28601
Methane from Landfills
Without Methane
Recovery
(c)
Percentage
of CH4 Not
Oxidized to
C02
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
(d)
Net GHG
Emissions
From CH4
Generation
(MTCE/Wet
Ton)
0.48
0.26
0.23
1.09
0.23
1.09
0.15
0.15
0.30
0.17
0.19
0.15
0.15
0.26
Methane from Landfills With LFG Recovery and:
Flaring Electricity Generation
(e)
Percentage of
CH4 Not
Recovered
(100% Minus
LFG
Collection
System
Efficiency)
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
(0
Percentage
ofCH4Not
Recovered
That Is Not
Oxidized to
C02
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
(3)
Net GHG
Emissions
From CH4
Generation
(MTCE/Wet
Ton)
0.12
0.07
0.06
0.27
0.06
0.27
0.04
0.04
0.08
0.04
0.05
0.04
0.04
0.06
(h)
Utility C02
Emissions
Avoided per
MTCE CH4
Combusted
(MTCE)
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
-0.18
(0
Percentage of
CH4 Recovered
for Electricity
Generation Not
Utilized Due to
System "Down
Time"
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
(i)
Utility C02
Emissions
Avoided
(MTCE/Wet
Ton)
-0.06
-0.03
-0.03
-0.14
-0.03
-0.14
-0.02
-0.02
-0.04
-0.02
-0.02
-0.02
-0.02
-0.03
Percentage of Methane from Each Type
of Landfill In 2000
(I)
Percentage
of CH4 From
Landfills
Without LFG
Recovery in
2000
51%
51%
51%
• 51%
51%
51%
51%
51%
51%
51%
51%
51%
51%
51%
(m)
Percentage
of CH4 From
Landfills With
LFG
Recovery
And Flaring
in 2000
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
(n)
CH4 From
Landfills With
LFG
Recovery and
Electricity
Generation in
2000
24%
24%
24%
24%
24%
24%
24%
24%
24%
24%
24%
24%
24%
24%
Net Methane
Generation
(0)
Net CH4
Emissions
from
Landfilling
(MTCE/Wet
Ton)
0.31
0.17
0.15
0.69
0.15
0.69
0.10
0.10
0.19
0.11
0.12
0.09
0.10
0.16
Avoided CO2
from Energy
Recovery
(P)
Net Avoided
C02
Emissions
from
Landfilling
(MTCE/Wet
Ton)
-0.01
-0.01
-0.01
-0.03
-0.01
-0.03
0.00
0.00
-0.01
-0.01
-0.01
0.00
0.00
-0.01
TOTAL
(q)
Net GHG
Emissions
From
Landfilling
(MTCE/Wet
Ton)
0.29
0.16
0.14
0.66
0.14
0.66
0.09
0.09
0.18
0.10
0.12
0.09
0.09
0.16
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
108
-------
Exhibit 7-5
Calculation to Estimate Utility GHGs Avoided through Combustion of Landfill CH4
Step
Metric tons CIVMTCE CH4
Grams CH^metric ton CH4
Cubic ft. CHVgram CH4
Btu/cubic ft. CH4
kwh Electricity generated/Btu
kwh electricity delivered/kwh
electricity generated
Btu/kwh electricity delivered
Kg utility C avoided/Btu delivered
electricity
Metric Tons avoided utility C/kg utility
C
Ratio of MTCE avoided utility C per
MICE CH4
Value
0.17
1.00E+06
0.05
1,000
0.00008
0.95
3,412
8.060E-05
0.001
0.18
Source
1 /((1 2/44)*Global warming potential of CH4)
Physical constant
1/20: 20 grams per cubic foot of methane at standard temperature and pressure
"Opportunity for LF Gas Energy Recovery in Kentucky,'
OAR September 97, pp. 2-12
1/13,000: from "Opportunity" report p. 2-1 1 , assumes use of internal combustion engines
Telephone conversation among IWSA, American Ref-Fuel, and ICF Consulting, October
28, 1997.
Physical constant
0.08349 MTCE/mmBtu delivered electricity, from Exhibit 6-3. This assumes that LFG
energy recovery displaces fossil fuel generation.
1 000 kg per metric ton
Product from multiplying all factors
109
-------
Exhibit 7-6
Net GHG Emissions from Landfilling
(a)
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Grass
Leaves
Branches
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed MSW
(b)
Net GHG Emissions from CH4 Generation
(MTCE/WetTon)
Landfills
Without LFG
Recovery
0.00
0.00
0.00
0.00
0.00
0.00
0.48
0.26
0.23
1.09
0.23
1.09
0.15
0.15
0.30
0.17
0,19
0.15
0.15
0.53
0.49
0.58
0.26
Landfills
With LFG
Recovery
and Flaring
0.00
0.00
0.00
0.00
0.00
0.00
0.12
0.07
0.06
0.27
0.06
0.27
0.04
0.04
0.08
0.04
0.05
0.04
0.04
0.13
0.12
0.15
0.06
Landfills With
LFG Recovery
and Electric
Generation
0.00
0.00
0.00
0.00
0.00
0.00
0.06
0.03
0.03
0.14
0.03
0.14
0.02
0.02
0.04
0.02
0.02
0.02
0.02
0:07
0.06
0.07
0.03
Year 2000
National
Average
0.00
0.00
0.00
0.00
0.00
0.00
0.29
0.16
0.14
0.66
0.14
0.66
0.09
0.09
0.18
0.10
0.12
0.09
0.09
0:32
0.29
0.35
0.16
(c)
Net Carbon
Storage
(MTCE/Wet
Ton)
0.00
0.00
0.00
0.00
0.00
0.00
-0,22
-0.29
-0,36
-0.04
-0.36
-0.04
-0.21
-0.21
-0.02
-0.21
-0.12
-0.39
-0=21
-0.23
-0.24
-0.21
-0.10
(d)
GHG
Emissions
From
Transportati
on
(MTCE/Wet
Ton)
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
(e)(e = b + c+d)
Net GHG Emissions from Landfilling
(MTCE/WetTon)
Landfills
Without LFG
Recovery
0.01
0.01
0.01
0.01
0.01
0.01
0.27
-0.02
-0.12
1.05
-0.12
1.05
-0.04
-0.04
0.29
-0.03
0.09
-0.23
-0.04
0:31
0.26
0.38
0.17
Landfills With
LFG Recovery
and Flaring
0.01
0.01
0.01
0.01
0.01
0.01
-0.09
-0.21
-0.29
0.24
-0.29
0.24
-0.16
-0.16
0.06
-0.15
-0.06
-0.34
-0.16
-0.08
-0.10
-0.05
-0.02
Landfills With
LFG Recovery
and Electric
Generation
0.01
0.01
0.01
0.01
0.01
0.01
-0.15
-0.25
-0.32
0.10
-0.32
0.10
-0.18
-0.18
0.03
-0.18
-0.08
-0.36
-0.18
-0.15
-0.16
-0.12
-0.06
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
Year 2000
National
Average
0.01
0.01
0.01
0.01
0.01
0.01
0.08
-0.12
-0.21
0.62
-0.21
0.62
-0.10
-0.10
0.17
-0.09
0.01
-0.29
-0.10
0.10
0.07
0.15
0.07
110
-------
Exhibit 7-7
Net GHG Emissions from CH4 Generation at Landfills
Sensitivity Analysis: Varying
(a)
Material
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Grass
Leaves
Branches
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed MSW
the Percentage of Waste Disposed at Landfills with Methane Recovery
(b)
17%
of Waste
Disposed At
Landfills
With LFG
Recovery
0.20
-0.05
-0.15
0.89
-0.15
0.89
-0.07
-0.07
0.25
-0.05
0.06
-0.25
-0.07
0.23
0.19
0.30
0.13
(c)
20%
of Waste
Disposed At
Landfills
With LFG
Recovery
0.19
-0.06
-0.16
0.87
-0.16
0.87
-0.07
-0.07
0.24
-0.06
0.05
-0.26
-0.07
0.22
0.18
0.28
0.12
(d)
49%
of Waste
Disposed At
Landfills
With LFG
Recovery
0.06
-0.13
-0.21
0.59
-0.21
0.59
-0.11
-0.11
0.16
-0.10
0.01
-0.30
-0.11
0.09
0.06
0.14
0.06
(e)
55%
of Waste
Disposed At
Landfills
With LFG
Recovery
0.04
-0.14
-0.23
0.54
-0.23
0.54
-0.12
-0.12
0.15
-0.11
0.00
-0.30
-0.12
0.06
0.03
0.11.
0.05
(f)
60%
of Waste
Disposed at
Landfills
With LFG
Recovery
0.02
-0.15
-0.24
0.49
-0.24
0.49
-0.12
-0.12
0.13
-0.11
-0.01
-0.31
-0.12
0.04
0.01
0.08
0.03
111
-------
Exhibit 7-8
Net GHG Emissions from CH4 Generation at Landfills
Sensitivity Analysis: Varying Oxidation and Gas Collection Efficiency Rates. Based on
Oxidation Rate:
Collection Efficiency:
Material
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Grass
Leaves
Branches
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed MSW
40%
95%
Lower-bound
Emissions
0.15
0.08
0.07
0.35
0.07
0.35
0.05
0.05
0.10
0.05
0.06
0.05
0.05
0.17
0.16
0.19
0.08
25%
85%
Conservative
(High)
Emissions
0.22
0.12
0.11
0.49
0.11
0.49
0.07
0.07
0.14
0.08
0.09
0.07
0.07
0.24
0.22
0.26
0.12
10%
75%
Best
Estimate
0.29
0.16
0.14
0.66
0.14
0.66
0.09
0.09
0.18
0.10
0.12
0.09
0.09
0.32
0.29
0.35
0.16
5%
60%
Upper-
bound
Emissions
0.35
0.19
0.17
0.78
0.17
0.78
0.11
0.11
0.22
0.12
0.14
0.11
0.11
0.38
0.35
0.42
0.19
112
-------
8. ACCOUNTING FOR EMISSION REDUCTIONS
Earlier chapters of this report examined GHG emissions from each of five waste management
options. This chapter recapitulates the emission factors for each option, explains the analytic framework
for applying emission factors, and reviews opportunities for GHG emission reductions.
In the discussion that follows, we focus on national average conditions. For example, we
represent landfills as having the national average landfill gas recovery systems, and we represent
combustors based on mass burn units with the national average system efficiency for collection of ferrous
metal. As shown in the previous chapters, GHG emissions are sensitive to many variables, including
several that are site-specific. At specific locations, the GHG emission factors can differ from those
described below. To allow for customizing of emission factors to better reflect local conditions, EPA has
developed a spreadsheet accounting tool, the Waste Reduction Model (WARM), which enables users to
input several key variables (e.g., information on landfill gas collection systems, electric utility fuel mix,
transportation" distances).1 We encourage readers to take advantage of this model when assessing their
waste management options.
8.1 NET GHG EMISSIONS FOR EACH WASTE MANAGEMENT OPTION
This section presents the net life-cycle GHG emissions for each waste management option for
each material considered. These emissions are shown in 12 exhibits that summarize the GHG emissions
and sinks in MTCE/ton and MTCO2E/ton, which are described in detail in earlier chapters. In these
exhibits, emission factors are shown for mixed plastics, mixed recyclables, and mixed organics. We
developed the emission factor for mixed recyclables by calculating the average (weighted by tons
recycled in 2000) of emission factors for aluminum cans, steel cans, HDPE, LDPE, PET, corrugated
cardboard, magazines/third-class mail, newspaper, office paper, phonebooks textbooks, and wood
products. The emission factor for mixed plastics is the average (weighted by tons recycled in 2000) of
emission factors for HDPE, LDPE, and PET. The mixed organics emission factor is the average
(weighted by tons composted in 2000) of emission factors for yard trimmings and food discards.2
As mentioned in Chapter 1, we used a waste generation reference point for measuring GHG
emissions; i.e., we begin accounting for GHG emissions at the point of waste generation. All subsequent
emissions and sinks from waste management practices then are counted. Changes in emissions and sinks
from raw material acquisition and manufacturing processes are captured to the extent that certain waste
management practices (i.e., source reduction and recycling) affect these processes (for reference, GHG
emissions from raw materials acquisition and manufacturing are shown in the first column of several
exhibits in this chapter). Negative emission factors indicate that from the point of waste generation, some
MSW management options can reduce GHG emissions.
1 Microsoft Excel® and Web-based versions of this tool are available online at the following Web site:
http://www.epa.gov/globalwarming/actions/waste/tools.htnil.
2 All data on recycling and compost rates are from U.S. EPA Office of Solid Waste. 2002. Municipal Solid
Waste in the United States: 2000 Facts and Figures, EPA 530-R-02-001.
113
-------
Exhibits 8-1 and 8-2 show the life-cycle GHG reductions associated with source reduction,
presented in MTCE/ton and MTCO2E/ton, respectively. In brief, the exhibits show that, for all of the
manufactured materials evaluated, source reduction results in GHG emission reductions. On a per-ton
basis, aluminum cans and several paper grades have the greatest potential for emission reduction, due •
primarily to reductions in energy use in the raw material acquisition and manufacturing step and (for
paper) forest carbon sequestration.
Exhibits 8-3 and 8-4 show the life-cycle GHG emissions associated with recycling in MTCE/ton
and MTCO2E/ton, respectively. The third through fifth columns in the exhibits show the GHG reductions
associated with using recycled inputs in place of virgin inputs when the material is remanufactured. As
the final column indicates, recycling results in negative emissions (measured from the point of waste
generation) for all the materials considered in this analysis. Emission reductions associated with
recycling are due to several factors, including avoided waste management emissions and reduced process
energy emissions.3 In addition, emission reductions from recycling paper products (when measured at the
point of waste generation) are due in part to the forest carbon sequestration benefits of recycling paper.
Exhibits 8-5 and 8-6 present the life-cycle GHG emissions from composting food discards, yard
trimmings, and mixed organics in MTCE/ton and MTCO2E/ton, respectively. The exhibits show that
composting these materials results in net emissions of -0.05 MTCE/ton, or -0.20 MTCO2E/ton, based on
the difference between the emissions associated with transporting the materials to the composting facility
and the soil carbon sequestration benefits.
Exhibits 8-7 and 8-8 present the life-cycle GHG emissions from combusting each of the materials
considered in MTCE/ton and MTCO2E/ton, respectively. These exhibits show emissions for mass burn
facilities with the national average rate of ferrous recovery. Results for RDF facilities are similar. As the
exhibits show, mixed MSW combustion has net emissions of -0.04 MTCE/ton, or -0.16 MTCCWton.
Net GHG emissions are positive for plastics, aluminum, and glass, and negative for the other materials.
GHG emissions from landfilling each of the materials in MTCE/ton are shown in Exhibit 8-9.
Exhibit 8-10 presents these values in MTCO2E/ton. The values in the final columns indicate that net
GHG emissions from landfilling mixed MSW, under national average conditions in 2000, are positive.
Among individual materials, emissions are lowest for newspaper, phonebooks, magazines/third-class
mail, wood products, and yard trimmings, and highest for office paper, textbooks, and food discards.
As discussed in Chapter 7 and shown in Exhibit 7-6, the results for landfills are very sensitive to
site-specific factors. Landfill gas collection practices significantly influence the net GHG emissions from
landfilling the organic materials. For mixed MSW, net emissions are 0.17 MTCE/ton in landfills without
landfill gas collection, and -0.06 MTCE/ton in landfills with landfill gas collection and energy recovery.
The largest differences attributable to landfill gas recovery are for office paper and textbooks (both have
a range of approximately 1 MTCE/ton), corrugated cardboard, and mixed paper. The CHU oxidation rate
and gas collection system efficiency also have a strong influence on the estimated net emissions for
mixed waste and the organic materials.
3 Process energy emissions for recycled corrugated cardboard, office paper, wood products (i.e.,
dimensional lumber and medium-density fiberboard), and mixed paper (broad and residential definitions) are
actually higher than those for virgin production because production with recycled inputs tends to use fossil fuel-
derived energy, while production with virgin inputs uses higher proportions of biomass fuel (CO2 from such fuel is
not counted in GHG inventories). In the case of dimensional lumber, production with recycled inputs requires more
energy than virgin production.
114
-------
Exhibits 8-11 and 8-12 display the national average emissions for each management option and
each material in MTCE/ton and MTCO2E/ton, respectively. When reviewing the emission factors, it is
important to recall caveats that appear throughout this report. In particular, these estimates do not reflect
site-specific variability, and they are not intended to compare one material to another. Rather, these
estimates are designed to support accounting for GHG emissions and sinks from waste management
practices. A brief recap of how to apply the emission factors appears in the following section.
8.2 APPLYING EMISSION FACTORS
The net GHG emission estimates presented in Exhibits 8-1 through 8-10 (and the more detailed
estimates in the preceding chapters) provide emission factors that may be used by organizations
interested in quantifying and voluntarily reporting emissions reductions associated with waste
management practices. In conjunction with the U.S. Department of Energy (DOE), EPA has used these
estimates as the basis for developing guidance for voluntary reporting of GHG reductions, as authorized
by Congress in Section 1605(b) of the Energy Policy Act of 1992. Other applications have included
evaluating the progress of voluntary programs aimed at source reduction and recycling, such as EPA's
WasteWise and Pay-as-You-Throw programs.
EPA has also assisted the Climate Neutral Network by using the methods and data described in
this report to develop company-specific GHG "footprints." As part of the program, companies develop
GHG footprints, which include "downstream" waste management activities, for their specific product
lines or facilities. These footprints then are used to determine the reductions or offsets that are necessary
to become GHG-neutral. Companies may use changes in waste management practices as part of their
offset portfolio.
Additionally, EPA worked with the International Council for Local Environmental Initiatives
(ICLEI) to incorporate GHG emission factors into its municipal GHG accounting software. Currently,
350 communities participate in ICLEI's Cities for Climate Protection Campaign, which helps cities and
towns establish a GHG emissions reduction target and implement a comprehensive local action plan
designed to achieve that target. The program has resulted in 7.5 million metric tons of annual GHG
emissions reductions.
In order to apply the emission factors presented in this report, one must first establish two
scenarios: (1) a baseline scenario that represents current management practices (e.g., disposing 10 tons
per year of office paper in a landfill with national average characteristics in terms of LEG collection);
and (2) an alternative scenario that represents the alternative management practice (e.g., recycling the
same 10 tons of office paper).4 The emission factors developed in this report then can be used to
calculate emissions under both the baseline and the alternative management practices. Once emissions
for the two scenarios have been determined, the next step is to calculate the difference between the
alternative scenario and the baseline scenario. The result represents the GHG emission reductions or
increases attributable to the alternative waste management practice.
4 The emission factors are expressed in terms of GHG emissions per ton of material managed. In the case
of recycling, we define 1 ton of material managed as 1 ton collected for recycling. As discussed in Chapter 4, the
emission factors can be adjusted to calculate GHG emissions in terms of tons of recycled materials as marketed
(reflecting losses in collection and sorting processes), or changes in the recycled content of products.
115
-------
Applying Emission Factors: Non-linear Relationship between Recycling and Emission Reductions and
Forest Carbon Leakage
Two caveats should be considered when applying the emission factors to analyze large-scale shifts in
waste management. First, increased recycling and GHG emission reductions may have a non-linear relationship,
such that emission reductions increase at a declining rate as recycling increases. This decline may be due to three
factors: (1) energy use in manufacturing processes may be non-linear with respect to recycled content; (2)
manufacturing capacity for recycled materials may be limited in the short term, so that large-scale increases in
recycling would require additional capital investment in capacity; and (3) market penetration of recyclables may
have limits (e.g., due to performance characteristics), such that recyclables cannot completely replace virgin
inputs in the short term.
In terms of the second caveat, the forest carbon sequestration benefits of paper and wood source
reduction and recycling are based on the assumption that reduced demand for a given paper or wood product
translates directly into reduced tree harvesting. Given that pulpwood and roundwood can be used for many
products, some of the forest carbon sequestration benefits may be lost by an increase in harvests for these other
products. This phenomenon is a form of what is sometimes termed "leakage" in the context of GHG mitigation
projects.
Although both of these issues are important considerations in applying the emission factors in this report,
we note that the emission factors are primarily designed for use by local waste managers. The factors are intended
to assess the GHG impacts of waste management decisions at a small-to-moderate scale. Readers should be
cautious when applying the emission factors at a larger scale, however, since the non-linear nature of the factors
and the issue of leakage become most relevant in the larger context.
Exhibits 8-13 and 8-14 illustrate the results of this procedure in a scenario where the baseline
management scenario is disposal in a landfill with national average conditions (i.e., the weighted average
in terms of landfill gas recovery practice). Alternative scenarios involve source reduction, recycling,
composting, or combustion. The values in the cells of the matrix are expressed in MTCE/ton in Exhibit
8-13 and in MTCC>2E/ton in Exhibit 8-14, and represent the incremental change in GHG emissions. For
example, recycling 1 ton of office paper, rather than landfilling it, reduces GHG emissions by 1.30
MTCE, or 4.76 MTCOaE (see the "Recycling" columns of the exhibits). Continuing the example from
the previous paragraph, if a business implements an office paper recycling program and annually diverts
10 tons of office paper (that would otherwise be landfilled) to recycling, the GHG emission reductions
are:
10 tons/yr * -1.30 MTCE/ton = -13.0 MTCE/yr
Under the sign convention used in this report, the negative value indicates that emissions are
reduced.
Due to resource and data limitations, emission factors have not been developed for all material
types reported by WasteWise partners, the Voluntary Reporting of Greenhouse Gas Program—or 1605(b)
as it is commonly called—and other parties interested in reporting voluntary emission reductions.
However, existing emission factors will continue to be updated and improved and new emission factors
will be developed as more data becomes available. The latest emission factors, reflecting these ongoing
revisions, can be found on the EPA Global Warming Web site
.
In cases where parties have been using source reduction or recycling techniques for materials
not specifically analyzed hi this report, it is possible to estimate the GHG emission reductions by
assigning surrogate materials. A List of materials not specifically analyzed, and their corresponding
surrogates, is presented in Exhibit 8-15. Surrogates are assigned based on consideration of similarities in
characteristics likely to drive life-cycle GHG emissions, such as similarities in energy consumption
116
-------
during the raw material acquisition and manufacturing life-cycle stages. Note that the use of these
surrogates involves considerable uncertainty. ' "
Exhibit 8-15 Recommended Surrogates for Voluntary Reporting
Material Source Reduced
Metal (type unknown)
Mixed Metals
Copper
Iron
Other Ferrous Metals
Other Non-Ferrous Metals
Steel
Plastic (resin unknown)
PVC/Vinyl
Polypropylene
Polystyrene
Other plastic (resin known, but not 41-46)
Rubber
Textiles
Boxboard
Kraft Paper
Coated Paper
High Grade Paper
Paper (type unknown)
Wood
Food .
Qrganics (type unknown)
Other Yard Waste
Surrogate Material
Average of Aluminum and Steel
Average of Aluminum and Steel
Steel Cans
Steel Cans
Steel Cans
Steel Cans
Steel Cans
(PET+HDPE+LDPE)/3
(PET+HDPE+LDPE)/3
(PET+HDPE+LDPE)/3
(PET+HDPE+LDPE)/3
(PET+HDPE+LDPE)/3
(PET+HDPE+LDPE)/3
(PET+HDPE+LDPE)/3
Corrugated Cardboard
Corrugated Cardboard
Magazines/Third-class Mail
Office Paper
Mixed Paper - Broad Definition
Dimensional Lumber
Food Discards
Yard Trimmings
Yard Trimmings
In our effort to continually expand and update life-cycle GHG emission factors for MSW
materials, we are in the process of developing emission factors for carpet and personal computers. The
emission factors will be based on data compiled by Franklin Associates, Ltd. These emission factors will
differ from the other emission factors presented in this report because they are for products, each of
which contain a variety of individual materials. In turn, the life-cycle emission factors will need to
account for GHG emissions associated with the life cycle of each component material. Given the
complexity of this task and the relatively limited life-cycle data on components of these products, EPA
welcomes input from industry stakeholders to augment or verify the activity data that will be the basis for
new emission factors for these products.
117
-------
8.3 OTHER LIFE-CYCLE GHG ANALYSES AND TOOLS
Life-cycle analysis is increasingly being used to quantify the GHG impacts of private and public
sector decisions. In addition to the life-cycle analyses that underpin the emission factors in this report,
Environmental Defense,5 ICLEI, Ecobilan, and others have analyzed the life-cycle environmental impacts
of various industry processes (e.g., manufacturing) and private and public sector practices (e.g., waste
management). In many cases, the results of life-cycle analyses are packaged into life-cycle software tools
that distill the information according to a specific user's needs.
As mentioned earlier, the WARM model was designed as a tool for waste managers to weigh the
GHG impacts of their waste management practices. As a result, the model focuses exclusively on waste
sector GHG emissions, and the methodology used to estimate emissions is consistent with international
and domestic GHG accounting guidelines. Life-cycle tools designed for broader audiences necessarily
include other sectors and/or other environmental impacts, and are not necessarily tied to the
Intergovernmental Panel on Climate Change (IPCC) guidelines for GHG accounting or the methods used
in the Inventoiy of U.S. Greenhouse Gas Emissions and Sinks.
• WARM covers 21 types of materials and 5 waste management options: source reduction,
recycling, combustion, composting, and landfilling. WARM accounts for upstream energy
and non-energy emissions, transportation distances to disposal and recycling facilities,
carbon sequestration, and utility offsets that result from landfill gas collection and
combustion. The tool provides participants in DOE's 1605(b) program with the option to
report results by year^ by gas, and by year and by gas. WARM software is available free of
charge in both a Web-based calculator format and a Microsoft Excel® spreadsheet. The tool
is ideal for waste planners interested in tracking and reporting voluntary GHG emission
reductions from waste management practices and comparing the climate change impacts of
different approaches. To access the tool, visit:
. The latest version of
WARM has the additional capacity to calculate energy savings resulting from waste
management decisions.
• The Cities for Climate Protection (CCP) Campaign's Greenhouse Gas Emission Software
was developed by Torrie Smith Associates for ICLEI. This Windows-based tool, targeted for
use by local governments, can analyze emissions and emission reductions on a community-
wide basis and for municipal operations alone. The community-wide module looks at
residential, commercial, and industrial buildings, transportation activity, and community-
generated waste. The municipal operations module considers municipal buildings, municipal
fleets, and waste from municipal in-house operations. In addition to computing GHG
emissions, the CCP software estimates reductions in criteria air pollutants, changes in energy
consumption, and financial costs and savings associated with energy use and other emission
reduction initiatives. A version of the software program was made available for use by
private businesses and institutions during the summer of 2001. CCP software subscriptions,
including technical support, are available to governments participating in ICLEI for a
subsidized price of $240. The full retail price of the software in the United States is $2,000.
5 Blum, L., Denison, R.A., and Ruston, V.F. 1997. "A Life-Cycle Approach to Purchasing and Using
Environmentally Preferable Paper: A Summary of the Paper Task Force Report," Journal of Industrial Ecology;
Volume 1; No. 3; pp, 15-46. Denison, R.A. 1996. "Environmental Life-Cycle Comparison of Recycling,
Landfilling, and Incineration: A Review of Recent Studies;" Annual Review of Energy and the Environment;
Volume 21, Chapter 6, pp.191-237.
118
-------
For more information, visit: or contact the U.S.
ICLEI office at (510)-540-8843, iclei_usa@iclei.org.
• The MSW Decision Support Tool (DST) and life-cycle inventory database for North
America have been developed through funding by EPA's Office of Research and
Development (ORD) through a cooperative agreement with the Research Triangle Institute
(CR823052). The methodology is based on a multi-media, multi-pollutant approach and
includes analysis of GHG emissions as well as a broader set of emissions (air, water, and
waste) associated with MSW operations. The MSW-DST is available for site-specific
applications and has been used to conduct analyses in several states and 15 communities,
including use by the U.S. Navy in the Pacific Northwest. The tool is intended for use by solid
waste planners at state and local levels to analyze and compare alternative MSW
management strategies with respect to cost, energy consumption, and environmental releases
to the air, land, and water. The costs are based on full-cost accounting principles and account
for capital and operating costs using an engineering economics analysis. The MSW-DST
calculates not only projected emissions of GHGs and criteria air pollutants, but also
emissions of more than 30 air- and water-borne pollutants. The DST models emissions
associated with all MSW management activities, including waste collection and
transportation, transfer stations, materials recovery facilities, compost facilities, landfills,
combustion and refuse-derived fuel facilities, utility offsets, material offsets, and source
reduction. The differences in residential, multi-family, and commercial sectors can be
evaluated individually. The software has optimization capabilities that enable one to identify
options that evaluate minimum costs as well as solutions that can maximize environmental
benefits, including energy conservation and GHG reductions.
At the time of the publication of this report, the LCI database for North America was to be
released in the whiter of 2002. All supporting documentation for the MSW-DST and LCI
database is to be released by spring 2002. Plans to develop a Web-based version are being
considered. The MSW-DST provides extensive default data for the full range of MSW
process models and requires minimum input data. The defaults can be tailored to the specific
communities using site-specific information. For further information, refer to the project
Web site at http://www.rti.org/units/ese/p2/lca.cfmtflife. The MSW-DST also includes a
calculator for source reduction and carbon sequestration using a methodology that is
consistent with the IPCC in terms of the treatment of biogenic CC-2 emissions. For more
information, refer to the project Web site: or
contact Susan Thornloe, U.S. EPA, (919)-541-2709, thornloe.susan@epamail.epa.gov, or
Keith Weitz, Research Triangle Institute, (919)-541-6973, kaw@rti.org.
• The Tool for Environmental Analysis and Management (TEAM), developed by Ecobilan,
simulates operations associated with product design, processes and, activities associated with
several industrial sectors. The model considers energy consumption, material consumption,
transportation, waste management, and other factors in its evaluation of environmental
impacts. Many firms and some government agencies have used the model. Users pay a
licensing fee of $3,000 and an annual maintenance contract of $3,000. This model is
intended for use in Europe and was not developed for use in North America. For more
information, visit: .
8.4 OPPORTUNITIES FOR GHG REDUCTIONS
Although this report has focused on the five most common waste management practices—source
reduction, recycling, composting, combustion, and landfilling—^For select materials, future quantification
efforts may include a number of emerging practices:
119
-------
Co-firing waste biomass. For utilities and power generating companies with coal-fired
capacity, co-firing with waste biomass may represent one of the least-cost renewable energy
options. Co-firing involves replacing a portion of the coal with biomass at an existing power
plant boiler. This replacement can be achieved by either mixing biomass with coal before
fuel is introduced into the boiler or by using separate fuel feeds for coal and biomas,s.
Specific biomass feedstocks include agricultural and wood waste, MSW, and industrial
wastes. Given the increasing use of co-firing technology as an energy source, understanding
its GHG benefits will likely be an important future EPA effort.
Compost as landfill cover. Using compost as landfill cover on closed landfills provides an
excellent environment for the bacteria that oxidize CH4. Under optimal conditions, compost
covers can practically eliminate CBU emissions. Furthermore, the covers offer the possibility
of controlling these emissions in a cost-effective manner. This technology is particularly
promising for small landfills, where landfill gas collection is not required and the economics
of landfill gas-to-energy projects are not attractive. Ancillary benefits also might arise in the
compost market from this technique if using compost as a landfill cover becomes a
widespread practice. An increase in composting could reduce the quantity of organic waste
disposed of at MSW landfills, thereby reducing CBU emissions. Given the recent
development of this practice, quantifying its GHG impacts will likely prove useful as landfill
owners consider adopting the technology.
Bioreactors. Bioreactors are a form of controlled landfilling with the potential to provide
reliable energy generation from solid waste, as well as significant environmental and solid
waste management benefits. The concept is to accelerate the decomposition process of
landfill waste through controlled additions of liquid and leachate recirculation, which
enhances the growth of the microbes responsible for solid waste decomposition. The result is
to shorten the time frame for landfill gas generation, thereby rendering projections of landfill
gas generation rates and yields that are much more reliable for landfill gas recovery.
Anaerobic digestion. Several facilities are using this technique to produce CH4 from mixed
waste, which is then used to fuel energy recovery. The approach generates CHt more quickly
and captures it more completely than in a landfill environment, and thus, from a GHG
perspective, offers a potentially attractive waste management option.6
The paperless office. The rise of computer technology for research, communications, and
other everyday workplace functions has presented a major opportunity for source reduction
in the modern office. Today's offices are commonly equipped with all the necessary
technologies to bypass paper entirely and rely instead on electronic communication. This
form of "comprehensive" source reduction comes with significant GHG benefits, as
described in Chapter 4. Therefore, attempting to quantify and communicate these benefits to
the business community will be an important task in the coming years.
Product stewardship. Increasingly, companies are taking responsibility for the environmental
impacts associated with the full life cycle of their products. Two industries in particular—
carpet and electronics—have been on the forefront of product stewardship efforts.
Carpet: Currently, more than 6 billion pounds of carpet are shipped each year, of which
approximately 200 million pounds are recycled. Although carpet is difficult to recycle due to
its varied make-up, any incremental increase in recycling could have significant climate
6 Environment Canada. 2001. Determination of the Impact of Waste Management Activities on Greenhouse
Gas Emissions. Submitted by ICF Consulting, Torrie-Smith Associates, and Enviros-RIS.
120
-------
benefits. As a result, EPA is working with a group of carpet industry representatives, state
environmental agencies, and non-profit recycling organizations to reach voluntary agreement
on a phase-out of carpet disposal. This product stewardship activity has focused on setting
rates and dates for carpet recovery over the next 10 years and encouraging the carpet industry
to develop a new third-party organization to help coordinate industry efforts. The carpet
industry and states have signed a memorandum of understanding that outlines the process
and guiding framework for developing a 10-year plan on carpet disposal phase-out. EPA will
continue to facilitate this effort in the coming years.7
Electronics: Understanding GHG emissions associated with waste management options for
electronics products is important for a number of reasons. First, electronics are among the
most rapidly growing categories of the U.S. waste stream. Sales of electronics have been
increasing dramatically, and, due to the fairly short period between purchase and discard, the
quantity of electronics discarded is expected to grow significantly in the future. Second,
electronics contain valuable materials that can be reused and/or recycled. Third, many
electronics products contain toxic materials that are covered by hazardous waste regulations.
These three factors have motivated interest on the part of electronics manufacturers, waste
managers, and others in recycling. Electronics will therefore become an increasingly
essential addition to the list of materials analyzed in this report.
EPA will continue to evaluate new opportunities to reduce emissions from waste management as
they become known. We also encourage readers to consider creative approaches to waste management,
particularly those with associated life-cycle energy benefits or carbon storage implications.
All of the exhibits presented so far in this report have expressed GHG emissions in units of
MTCE or MTCO2E, calculated as the sum of the individual gases (CO2, CHU, N2O, and PFCs) weighted
by their global warming potential. In the Voluntary Reporting of Greenhouse Gas Program—also known
as the 1605(b) program—established by DOE's Energy Information Administration, reporting companies
are asked to provide emission reductions for each of the individual gases. In addition, the 1605(b)
program requires emission reductions to be reported in the year they are achieved and does not allow
participants to take credit for future emission reductions. Because the GHG emission factors presented in
this report reflect the "present value" of future emissions and sinks as well as emissions and sinks
occurring in the reporting year, our emission factors are not directly transferrable to the 1605(b) program.
For purposes of supporting the program, we developed a revised set of 1605(b) program emission factors
that reflect emissions by gas and by year. These emission factors provide incremental emissions for a
baseline of landfilling and alternative scenarios of source reduction and recycling. Detailed reporting
instructions and forms are available on DOE's Web site at http://www.eia.doe.gov/oiaf/1605/forms.html.
# * * #
7 Additional information on this activity is available on the Minnesota Office of Environmental Assistance
Web site at http://www.moea.state.mn.us/carpet/care.cfni.
121
-------
We close with a final note about the limitations of the GHG emission estimates in this report.
We based our analysis on what we believed to be the best available data; where necessary, we made
assumptions that we believe are reasonable. The accuracy of the estimates is limited, however, by the use
of these assumptions and limitations in the data sources, as discussed throughout this report. Where
possible, the emission factors reported here can be improved by substituting process- or site-specific data
to increase the accuracy of the estimates. For example, a commercial firm with a large aluminum
recycling program may have better data on the specific fuel mix of its source of aluminum and could thus
calculate a more exact value for the emission factor. Despite the uncertainty in the emission factors, they
provide a reasonable first approximation of the GHG impacts of solid waste management, and we believe
that they provide a sound basis for evaluating voluntary actions to reduce GHG emissions in the waste
management arena.
122
-------
Exhibit 8-1
GHG Emissions for Source Reduction
(MTCE/Ton of Material Source Reduced)
Emissions Measured from a Waste Generation Reference Point1
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
(a) Raw Materials Acquisition
and Manufacturing
Source
Reduction
Displaces
Current Mix of
Virgin and
Recycled
Inputs
-2.49
-0.79
-0.14
-0.49
-0.61
-0.49
-0.24
-0.46
-0.46
-0.31
-0.64
-0.59
-0.05
-0.10
NA
NA
NA
NA
NA
NA
NA
NA
NA
Source
Reduction
Displaces
Virgin Inputs
-4.67
-1.01
-0.16
-0.53
-0.64
-0.58
-0.22
-0.46
-0.59
-0.28
-0.67
-0.59
-0.05
-0.10
NA
NA
NA
NA
NA
NA
NA
NA
NA
(b) Forest Carbon Sequestration
Source
Reduction
Displaces
Current Mix of
Virgin and
Recycled
Inputs
0.00
0.00
0.00
0.00
0.00
0.00
-0.28
-0.58
-0.35
-0.50
-0.65
-0.64
-0.50
-0.50
NA
NA
NA
NA
NA
NA
NA
NA
NA
Source
Reduction
Displaces
Virgin Inputs
0.00
0.00
0.00
0.00
0.00
0.00
-0.73
-0.73
-0.73
-0.73
-0.73
-0.73
-0.50
-0.50
NA
NA
NA
NA
NA
NA
NA
NA
NA
(c)
Waste
Management
Emissions
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
NA
NA
NA
NA
NA
NA
NA
NA
(d = a + b + c)
(d) Net Emissions
Source Reduction
Displaces Current
Mix of Virgin and
Recycled Inputs
-2.49
-0.79
-0.14
-0.49
-0.61
-0.49
-0.51
-1.04
-0.81
-0.80
-1.28
-1.23
-0.55
-0.60
NA
NA
NA
NA
NA
NA
NA
NA
NA
Source
Reduction
Displaces Virgin
Inputs
-4.67
-1.01
-0.16
-0.53
-0.64
-0.58
-0.96
-1.19
-1.32
-1.01
-1.40
-1.32
-0.55
-0.60
NA
NA
NA
NA
NA
NA
NA
NA
NA
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Under the accounting convention used in this analysis, emissions are quantified from a waste generation reference point (once the material has already undergone the raw
materials acquisition and manufacturing phase).
123
-------
Exhibit 8-2
GHG Emissions for Source Reduction
(MTCO2E/Ton of Material Source Reduced)
Emissions Measured from a Waste Generation Reference Point1
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
(a) Raw Materials Acquisition
and Manufacturing
Source
Reduction
Displaces
Current Mix of
Virgin and
Recycled
Inputs
-9.15
-2.89
-0.50
-1.79
-2.25
r1.78
-0.88
-1.69
-1.69
-1.13
-2.33
-2.15
-0.17
-0.36
NA
NA
NA
NA
NA
NA
NA
NA
NA
Source
Reduction
Displaces
Virgin Inputs
-17.11
-3.69
-0.57
-1.95
-2.34
-2.14
-0.82
-1.69
-2.15
-1.02
-2.44
-2.16
-0.17
-0.36
NA
NA
NA
NA
NA
NA
NA
NA
NA
(b) Forest Carbon Sequestration
Source
Reduction
Displaces
Current Mix of
Virgin and
Recycled
Inputs
0.00
0.00
0.00
0.00
0.00
0.00
-1.01
-2.11
-1.29
-1.82
-2.37
-2.35
-1.84
-1.84
NA
NA
NA
NA
NA
NA
NA
NA
NA
Source
Reduction
Displaces
Virgin Inputs
0.00
0.00
0.00
0.00
0.00
0.00
-2.69
-2.69
-2.69
-2.69
-2.69
-2.69
., -1.84
-1.84
NA
NA
NA
NA
NA
NA
NA
NA
NA
(c)
Waste
Management
Emissions
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
NA
NA
NA
NA
NA
NA
NA
NA
(d = a + b + c)
(d) Net Emissions
Source Reduction
Displaces Current
Mix of Virgin and
Recycled Inputs
-9.15
-2.89
-0.50
-1.79
-2.25
-1.78
-1.89
-3.80
-2.97
-2.95
-4.70
-4.49
-2.01
-2.20
NA
NA
NA
NA
NA
NA
NA
NA
NA
Source
Reduction
Displaces Virgin
Inputs
-17.11
-3.69
-0.57
-1.95
-2.34
-2.14
-3.50
-4.38
-4.84
-3.71
-5.13
-4.85
-2.01
-2.20
NA
NA
NA
NA
NA
NA
NA
NA
NA
Note that totals may not add due to rounding and more digits may be displayed .than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Under the accounting convention used in this analysis, emissions are quantified from a waste generation reference point
(once the material has already undergone the raw materials acquisition and manufacturing phase). •:' '.,
-------
Exhibit 8-3
Recycling
(GHG Emissions in MTCE/Ton)
Emissions Measured from a Waste Generation Reference Point1
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
Raw Materials Acquisition and
Manufacturing (RMAM)
(a)
RMAM Emissions
Not Included in
Baseline3 (current
mix of inputs)
2.49
0.79
0.14
0.49
: 0.61
0.49
0.24
0.46
0.46
0.31
0.64
0.59
0.05
0.10
NA
NA
•{
0.38
0.38
0.85
0.51
0.36
NA
NA
(b)
Waste
Generation
Baseline
0.00
0.00
0.00
0.00
0,00
o.po
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
NA
Recycled Input Credit2
(c)
Process
Energy
-2.92
-0.48
-0.03
-0.34
-0.43
-0.40
0.04
0.00
-0.21
0.06
-0.18
-0.01
0.02
0.01
NA
NA
0.08
0.08
-0.08
-0.38
-0.10
NA
NA
(d)
Transportation
Energy
-0.14
-0.01
0.00
0.00
0.00
0.00
-0.01
0.00
-0.01
0.00
0.00
0.00
0.00
0.00
NA
NA
-0.02
-0.02
-0.02
0.00
-0.01
NA
NA
(e)
Process
Non-Energy
-1.05
0.00
-0.04
-0.04
-0.04
-0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
" 0.00
NA
NA
0.00
0.00
0.00
-0.03
-0.02
NA
NA
(f)
Forest Carbon
Sequestration
0.00
0.00
0.00
0.00
0.00
0.00
-0.73
-0.73
-0.73
-0.73
-0.73
-0.73
-0.69
-0.69
'• NA
NA
-0.73
-0.73
-0.73
0.00
-0.63
NA
NA
(g)
Waste
Management
Emissions
.0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
NA
(")
(h = b+c+d+e+f+g)
Net Emissions
-4.11
-0.49
-0.08
-0.38
-0.47
-0.42
-0.71
-0.74
-0.95
-0.68
-0.91
-0.75
-0.67
-0.67
NA
NA
-0.67
-0.67
-0.83
-0.41
-0.76
NA
NA
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Under the accounting convention used in this analysis, emissions are quantified from a waste generation reference point (once the material has already undergone the raw materials acquisition
and manufacturing phase).
Material that is recycled after use is then substituted for virgin inputs in the production of new products. This credit represents the difference in emissions that results from using recycled inputs
125
-------
Exhibit 8-4
Recycling
(GHG Emissions in MTCO2E/Ton)
Emissions Measured from a Waste Generation Reference Point1
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third Class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
Raw Materials Acquisition and
Manufacturing (RMAM)
(a)
RMAM Emissions
Not Included in
Baseline3 (Current
Mix of Inputs)
9.15
2.89
0.50
1.79
2.25
1.78
0.88
1:69
1.69
1/13
2.33
2.15
0.17
0.36
NA
NA
1.38
1.39
3.12
1.85
1.32
NA
NA
(b)
Waste
Generation
Baseline
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
- - NA
Recycled Input Credit2
(c)
Process
Energy
-10.70
-1.75
-0.12
-1.26
-1.57
-1.48
0.13
-0.01
'. -0.76
0.22
-0.65
-0.05
0.07
0.05
NA
NA
0.29
0.29
-0.29
-1.40
-0.38
NA
-NA
(d)
Transportation
Energy
-0.51
-0.04
-0.02
0.00
0.00
0.00
-0.04
0.00
"-'- -0.03
0.00
0.00
0.00
0.01
0.02
~ NA
NA
-0.06
-0.06
-0.07
0.00
-0.04
NA
NA
(e)
Process
Non-,
Energy
-3.86
0.00
-0.14
-0.15
-0.15
-0.08
-0.01
0.00
0.00
-0.02
0.00
0.00
0.00
0.00
NA
NA
0.00
0.00
0.00
-0.12
-0.09
NA
NA
(f)
Forest Carbon
Sequestration
0.00
0.00
0.00
0.00
0.00
0.00
-2.69
-2.69
-2.69
-2.69
-2.69
-2.69
-2.53
-2.53
NA
NA
-2.69
-2.69
-2.69
0.00
-2.30
NA
NA
(g)
Waste
Management
Emissions
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
NA
(h)
(h = b+c+d+e+f+g)
Net Emissions
-15.07
-1.79
-0.28
-1.40
-1.71
-1.55
-2.60
-2.70
-3.48
-2.48
-3.34
-2.74
-2.45
-2.47
NA
NA
-2.47
-2.47
-3.05
-1.51
-2.80
NA
NA
Note that totals may not add due to rounding and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Under the accounting convention used in this analysis, emissions are quantified from a waste generation reference point
(once the material has already undergone the raw materials acquisitionand manufacturing phase). '
Material that is recycled after use is then substituted for virgin inputs in the production of new products. This credit represents the difference in emissions that results from using recycled inputs
" ' 126- '
-------
Values £
Aluminum C
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated
Magazines/
Newspaper
Office Papc
Phonebook
Textbooks
Dimension;
Medium-de
Food Disca
Yard Trimnr
Mixed Papc
Broad De
Resident!
Office Pa
Mixed Plas
Mixed Recv
Mixed Orgaj
Mixed MSV\
Note that totals
NA: Not appli
1 Under the a
(once the ma
2 The value fo
Exhibit 8-5
Composting
(GHG Emissions in MTCE/Ton)
Values are for Mass Bum Facilities with National Average Rate of Ferrous Recovery. Emissions Measured from a Waste Generation
Reference Point1
Material
Aluminum Cans
Steel Cans
nlaoo
^liClOo
HOPE
.DPE
3FT
1— 1
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
3honebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
Raw Materials Acquisition and
Manufacturing (RMAM)
(a)
RMAM Emissions
Not Included in
Baseline2
-2.49
-0.79
-0.14
-0.49
-0.61
-0.49
-0;24
-0.46
-0.46
-0.31
-0.64
-0.59
-0.05
-0.10
NA
NA
0.38
0.38
0.85
0.51
0.36
NA
NA
(b)
Waste Generation
Baseline
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
(c)
Transportation to
Composting
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.01
0.01
NA
NA
NA
NA
NA
0.0
NA
(d)
Soil Carbon
Sequestration
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.07
-0.07
NA
NA
NA
NA
NA
-0.07
NA
(afcV
e)
(e = b+c+d)
Net Emissions
(Post-Consumer)
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.05
-0.05
NA
NA
NA
NA
A
-0.05
NA
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Under the accounting convention used in this analysis, emissions are quantified from a waste generation reference point
(once the material has already undergone the raw materials acquisition and manufacturing phase).
2 The value for mixed MSW is the weighted average of the RMAM emissions for those materials we studied.
127
-------
Exhibit 8-10
Landfilling
(GHG Emissions in MTCO2E/Ton)
Values for Landfill Methane and Net Emissions Reflect Projected National Average Methane Recovery in year 2000. Emissions
Measured from a Waste Generation Reference Point1
Material
Aluminum Cans
Steel Cans •
Glass
HOPE
LDPE
PET , ......
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper . .
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards,.,,.
Yard Trimmings /
Mixed Paper ;
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
Raw Materials Acquisition
and Manufacturing (RMAM)
(a)
RMAM
Emissions
Not Included
in Baseline2
9.15
2.89
0.50
" 1.79
2.25
1.78
; 0.88
I 1.69
! 1.69
1.13
2.33
2.15
0.17
0.36
' NA
; NA
' 1.38
1.39
3.12
1.85
1 .32
NA
NA
(b)
Waste
Generation
Baseline
0.00
0.00
0.00
0.00
;; o.oo
, : o.oo
: o.oo
i 0.00
0.00
0.00
0.00
o.oo
0.00
0.00
'"'• 0.00
0.00
• 0.00
0.00
0.00
0.00
0.00
0.00
: 0.00
(c)
Transportation
to Landfill
0.04
0.04
0.04
0.04
;. ; 0.04
0.04
0.04
0.04
; 0.04
0.04
0.04
. ; 0.04
0.04
: 0.04
0.04
0.04
" , • ' 0.04
0.04
0.04
0.04
0.04
0.04
. .0.04
(d)
Net
Landfill
CH4
0.00
0.00
0.00
0.00
0.00
0.00
1.12
0.61
0.54
2.52
0.54
2.52
; 0.35
0.35
0.70
0.40
: 1.22
1.13
1.35
0.00
0.95
0.54
0.60
(e)
Landfill Carbon
Sequestration
0.00
0.00
. 0.00
0.00
0.00
0.00
-0.82
-1.07
-1.32
-0.16
: -1.32
-0.16
-0.76
: -0.76
-0.08
-0.76
-0.83
-0.87
-0.76
0.00
-0.75
-0.43
, -0.37
W
(f=b+c+d+e)
Net Emissions
0.04
0.04
0.04
0.04
0.04
0.04
0.34
-0.41
-0.74
. 2.40
-0.74
2.40
-0.37
-0.37
0.66
-0.33
0.43
0.30
0.63
0.04
0.24
0.15
0.27
Note that totals may not add due to rounding and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Under the accounting convention used in this analysis, emissions are quantified from a waste generation reference point
(once the material has already undergone the raw materials acquisition and manufacturing phase).
2The value for mixed MSW is the weighted average of the RMAM emissions for those materials we studied.
132
-------
Exhibit 8-11
Net GHG Emissions from Source Reduction and MSW Management Options
(MTCE/Ton)
Emissions Measured from a Waste Generation Reference Point1
Material
Aluminum Cans
Steel .Caris , :
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Maga?ines/Third-class Mail
Newspaper, ,
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings ;
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics '
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
Source
Reduction2
-2.49
-0.79
-0.14
' -0:49
-oiei
-0.49
-0.51
. -1k)4
;,:b;8i
-0:80
-1,28
r1.23
-0.55
-0:60
.NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling
-4.11
-0.49
:-0.08
',. -0.38
-0.47
-0.42
-0.71
.-0.74
,,-0.95
..-0.68
rO.91
-0.75
-0.67
-0.67
. NA
NA
'-0.67
-0.67
-0.83
-0.41
-0.76
NA
NA
Composting
NA
NA
NA
NA
NA
JMA
'. NA
NA
. NA
NA
NA
NA
NA
NA
-0.05
-0.05
NA
NA
NA
NA
NA
-0.05
NA
Combustion3
0.02
' -0.42
0.01
0.23
0.23
0.28
-0.19
-0.13
-0.21
-0.18
-0.21
-0.18
-0.22
-0.22
-0.05
-0.06
-0.19
; -0.18
-0.17
0.25
-0.17
-0.06
-0.04
Landfilling4
0.01
0.01
0.01
0.01
0.01
0.01
0.08
-0.12
-0.21
0.62
:• -0.21
0.62
-0.10
: -0.10
0.17
-0.09
0.10
0.07
0.15
0.01
0.05
0.03
0.07
Note that totals may not add due to rounding, and more digits may be displayed than are significant
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Under the accounting convention used in this analysis, emissions are quantified from a waste generation reference point
(once the material has already undergone the raw materials acquisition and manufacturing phase).
2 Source reduction assumes displacement of current mix of virgin and recycled inputs.
3 Values are for mass burn facilities with national average rate of ferrous recovery.
4 Values reflect projected national average methane recovery in year 2000.
133
-------
Exhibit 8-12
Net GHG Emissions from Source Reduction and MSW Management Options
(MTCO2Em>n)
Emissions Measured from a Waste Generation Reference Point
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
Source
Reduction2
-9.15
-2.89
-0.50
-1.79
-2.25
-1.78
-1.89
-3.80
-2.97
-2.95
,-4.70
-4.49
-2.01
-2.20
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling
-15.07
-1.79
-0.28
-1.40
-1.71
-1.55
-2.60
-2.70
-3.48
-2.48
-3.34
-2.74
-2.45
-2.47
NA
NA
-2.47
-2.47
-3.05
-1.51
-2.80
NA
NA
Composting
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.20
-0.20
NA
NA
NA
NA
NA
-0.20
NA
Combustion3
0.06
-1.53
0.05
0.85
0.85
1.04
-0.68
-0.49
-0.77
-0.65
-0.77
-0.65
-0.81
-0.81
-0.19
-0.23
-0.68
-0.68
-0.62
0.93
-0.61
-0.21
-0.13
Landfilling4
0.04
0.04
0.04
0.04
0.04
0.04
0.28
-0.44
-0.76
2.28
-0.76
2.28
-0.38
-0.38
0.62
-0.34
0.37
0.25
0.56
0.04
0.19
0.12
0.24
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Under the accounting convention used in this analysis, emissions are quantified from a waste generation reference point
(once the material has already undergone the raw materials acquisition and manufacturing phase).
2 Source reduction assumes displacement of current mix of virgin and recycled inputs.
3 Values are for mass bum facilities with national average rate of ferrous recovery.
4 Values reflect projected national average methane recovery in year 2000.
134
-------
ExhibitS-IS
Net GHG Emissions of MSW Management Options Compared to Landfilling1
(MTCE/Ton)
Negative values indicate emission reductions.
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition ,
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
Source Reduction
Net Emissions
Minus Landfilling Net Emissions
Current Mix of Inputs
-2.50
-0.80
-0.15
-0.50
-0.63
-0.50
-0.59
; -0.92
-0.60
-1 .43
-1.07
-1 .85
-0.44
-0.50
NA
NA
NA
NA
NA
NA
NA
NA
NA
100% Virgin Inputs
-4.68
-1.02
-0.17
-0.54
-0.65
-0.59
. -1.03
-1.07
-1.11
-1.63
-1.19
-1.94
-0.44
-0.50
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling
Net Emissions
Minus Landfilling
Net Emissions
-4.12
-0.50
-0.09
-0.39
-0.48
-0.43
-0.79
-0.62
-0.74
-1.30
-0.70
-1.37
-0.56
-0.57
NA
NA
-0.78
-0.74
-0.99
-0.42
-0.82
NA
NA
Composting
Net Emissions
Minus Landfilling
Net Emissions
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.22
0.04
NA
NA
NA
NA
NA
-0.09
NA
Combustion2
Net Emissions
Minus Landfilling
Net Emissions
0.01
-0.43
0.00
0.22
0.22
0.27
-0.26
-0.01
0.00
-0.80
0.00
-0.80
-0.12
-0.12
-0.22
0.03
-0.29
-0.25
-0.32
0.24
-0.22
-0.09
-0.10
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Values for landfilling reflect projected national average methane recovery in year 2000.
2 Values are for mass bum facilities with national average rate of ferrous recovery.
135
-------
Exhibit 8-14
Net GHG Emissions of MSW Management Options Compared to Landfilling1
(MTCO2E/Ton)
Negative values indicate emission reductions.
Material
Aluminum Cans
Steel Cans
Glass
HOPE
LDPE
PET
Corrugated Cardboard
Magazines/Third-class Mail
Newspaper
Office Paper
Phonebooks
Textbooks
Dimensional Lumber
Medium-density Fiberboard
Food Discards
Yard Trimmings
Mixed Paper
Broad Definition
Residential Definition
Office Paper Definition
Mixed Plastics
Mixed Recyclables
Mixed Organics
Mixed MSW (as disposed)
Source Reduction
Net Emissions
Minus Landfilling Net Emissions
Current Mix of Inputs
-9.18
-2.92
-0.54
-1.82
-2.29
-1.82
-2.17
-3.36
-2.21
-5.23
-3.94
-6.78
-1.63
-1.82
NA
NA
NA
NA
NA
NA
NA
NA
NA
100% Virgin Inputs
-17.15
-3.72
-0.61
-1.99
-2.38
-2.18
-3.79
,-3.94
-4.07
-5.99
-4.37
-7.13
-1.63
-1.82
NA
NA
NA
NA
NA
NA
NA
NA
NA
Recycling
Net Emissions
Minus Landfilling
Net Emissions
-15.11
-1.83
-0.32
-1.44
-1.75
-1.59
-2.88
-2.26
-2.72
-4.77
-2.57
-5.03
-2.07
-2.09
NA
NA
-2.84
-2.72
-3.62
-2.99
NA
NA
Composting
Net Emissions
Minus Landfilling
Net Emissions
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.82
0.15
NA
NA
NA
NA
NA
-0.32
NA
Combustion2
Net Emissions
Minus Landfilling
Net Emissions
0.02
-1.57
0.01
0.81
0.81
1.00
-0.96
-0.05
-0.01
-2.94
-0.01
-2.94
-0.43
-0.43
-0.81
0.11
-1.06
-0.93
-1.18
0.90
-0.80
-0.33
-0.38
Note that totals may not add due to rounding, and more digits may be displayed than are significant.
NA: Not applicable, or in the case of composting of paper, not analyzed.
1 Values for landfilling reflect projected national average methane recovery in year 2000.
2 Vaiues are for mass burn facilities with national average rate of ferrous recovery.
136
------- |