r
i#
GLOBRL GL
&f - t Xt.
I
r
t '!^)lr [
Draft
Global Climate Change
to
» , !"«»(«! 4T'9 JSTOPVI"-* ^
EP* ;-
i
Global farming
;Glo.bai
Air and ^EneifgJ
^S. En|»
Aprih*1993
-------
S'X
6, L
-------
EPA
3
GLOBRL CLIMRTE CHHNGE
ENGINEERING ROD PROGRRM
Draft Background Report Submitted to
Global Climate Change Engineering Research Subcommittee
EPA Science Advisory Board
Prepared by
Global Warming Control Branch
Global Emissions and Control Division
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC
3*
April 1993
-------
-------
CONTENTS
PageNo.
Preface i
Background Summaries of Primary Program Areas 1
GHG Emissions Estimation/Database Management
Global Emissions Databases and Database Software 2
* Methane Emissions from Coal Mines 7
Methane Emissions from the Natural Gas Industry 10
Development of Greenhouse Gas Emissions Data from Municipal
Solid Waste Landfills and Other Waste Management Facilities 13
* Effects of Atmospheric Emissions and Chemistry on Radiative Forcing:
Development of Global Emissions Inventories for Tropospheric Ozone
Precursors and Aerosols from Anthropogenic Sources 21
Methane Mitigation
Demonstration of Fuel Cells to Recover Energy/Control Methane
Emissions from Waste Methane Gas Sources 25
Enhanced Coalbed Methane Recovery 29
Biomass Utilization
Biomass Utilization 33
Expanded Future Program - Strategic Directions . .40
Appendices
AEERL Papers from Proceedings: 1992 Greenhouse Gas Emissions
and Mitigation Research Symposium
Additional AEERL Journal Articles and Book Chapters Relating to
EPA's Global Climate Change Research Program
-------
-------
PREFACE
Background
Anthropogenic activities are significantly increasing the atmospheric concentrations
of the greenhouse gases: carbon dioxide, methane, chlorofluorocarbons, and nitrous
oxide. This increase is affecting regional and global climates -evidence suggests a
rise of between 0.3°G and 0.6°C over the last century. This seemingly small increase
should it continue could have a devastating effect on sensitive ecosystems, coastal
habitats, and socio-economic effects on humans.
Since its founding twenty years ago, the U.S. Environmental Protection Agency
(EPA) has responded to many threats to the quality of our Nation's air, land, and
water resources. Many of these threats have been the result of decades of human
activities undertaken without regard for, or prior knowledge of, the adverse effects
on the economic, recreational, and life-sustaining value of our natural resources. The
Science Advisory Board's report": Reducing Risk: Setting Priorities and Strategies
for Environmental Protection" has identified global change as one of its highest
research priorities over the next few decades. The EPA Global Change Research
Program (GCRP), within the Office of Research and Development (ORD), is an integral
part of the U.S. GCRP.
The USGCRP was established by a presidential initiative in the FY90 budget and is
driven by priorities that must be addressed to establish national and international
policies related to global environmental issues, particularly global climate change. The .
USGCRP was developed by the interagency Committee on Earth and Environmental
Sciences (CEES) of the Federal Coordinating Council for Science, Engineering, and
Technology (FCCSET). It is linked internationally to other government and
nongovernmental organizations within the U.S.
Much of AEERL's current research under the GCRP is an integral part of the USGCRP
but the AEERL mitigation research and development work has been largely funded by
directed Congressional appropriation and is generally outside the sanctioned scope of
theUSGCRP. -...
the priority
following:
scientific questions addressed by the EPA research program are the
1. How much will the response of the terrestrial and near coastal biosphere
amplify or dampen global climate change associated with greenhouse
gases?
-------
2. How will climate change effect important terrestrial ecosystems?
3. What are the impacts of altered tropospheric loadings of Radiatively
Important Trace Gases (RITGs) and their precursors, and what are their
Global Warming Potentials (GWPs)?
4. What is the net rate of deforestation and biomass burning?
5. What are the current and projected future anthropogenic emissions of
methane and other greenhouse gases and precursors?
6. What can be done to reduce anthropogenic emissions of methane and other
greenhouse gases and precursors? What is the potential for wood and
other renewable sources of energy as an alternative to fossil fuels?
AEERL's current research program is designed to primarily contribute to
providing answers to questions 3, 5, and 6. To help answer these questions, AEERL is
conducting research in a number of areas including (but not limited to) the following:
1. Greenhouse gas emissions estimation/data base management
2. Methane mitigation
* - *.«. ' ' ' ' '
3. Biomass utilization - : .
Research areas 1 and 2 are beginning to achieve some maturity in that
significant progress is being made while research area 3 has received less attention
due to limited funding availability. .
Purpose of this Report
The Global Climate Change Engineering Research Subcommittee of EPA's Science
Advisory Board (SAB) will review portions of AEERL's engineering research program
dealing with global climate change. A review at this point in the program will help
ensure that technical, intellectual, and financial resources are effectively employed for
continued success in existing areas as well as sound beginnings in the emerging areas.
This report, prepared as introductory material for the subcommittee, presents a
narrative overview of the major areas of the program to be reviewed. Additional
details will be presented to the subcommittee at the review on May 26-27, 1993.
11
-------
Several of the topical areas to be reviewed are reasonably well established.
Their review will be considered research in progress. These areas are:
GHG emissions estimation/database management - This area involves the
development of global GHG emissions estimates (with emphasis on methane) for
major sources such as waste management facilities, cookstoves, coal mines, and
natural gas production/distribution. It also includes the design of software systems
(GloED and GloTech) for storing GHG databases. In addition, it provides for the
development of tropospheric ozone precursor inventories for NOX, NMNC, CO, and
aerosols. Key programs are listed below:
Global Emissions Databases and Database Software
* Methane Emissions from Coal Mines
* Methane Emissions from the Natural Gas Industry
Development of Greenhouse Gas Emissions Data from Municipal Solid Waste
Landfills and Other Waste Management Facilities
Development of Global Emissions Inventories for Tropospheric Ozone
Precursors and Aerosols from Anthropogenic Sources
Methane mitigation options evaluation - This area includes the evaluation of
mitigation opportunities for "waste" methane including the use of fuel cells to recover
energy from landfill/anaerobic digester gases. It also involves the evaluation of
enhanced coalbed recovery concept for recovering methane from coal mines. Principal
programs include:
» Demonstration of Fuel Cells to Recover Energy/Control Methane Emissions
from Waste Methane Gas Sources
Enhanced Coalbed Methane Recovery
The third topical area; biomass utilization, is one of emerging emphasis and,
therefore, should be reviewed taking into consideration the limited scope of AEERL's
research to date. This area includes the evaluation of the most promising
technologies for reducing CO2 emissions through conversion of biomass to (a)
electricity for both industrialized and developing countries and (b) liquid
transportation fuels (with emphasis on methanol via the Hydrocarb process) which
could displace petroleum fuels.
The subcommittee is also asked to examine the proposed strategic directions
for an expanded future AEERL program should substantially increased EPA resources
become available in future fiscal years. This future program is briefly described in the
report and includes collaborative activities with industry and other
federal/international agencies in areas including:
iii
-------
Development of Comprehensive Global Emissions Inventories
Mitigation/Utilization of Waste Methane
Enhanced Use of Biomass to Displace Fossil Fuels
Evaluation of CO2 Sequestration and Disposal
Mitigation Opportunities for Important Greenhouse Gases Other than
COa/Methane
Generate Comprehensive Database of Mitigation Technologies
Integration of Renewables into Fossil Energy Production Technologies to
Reduce Greenhouse Gas Emissions
iv
-------
Background Summaries of
Primary Program Areas
-------
GHG Emissions Estimation/Database Management
GLOBAL EMISSIONS DATABASES AND DATABASE SOFTWARE
Introduction/Background
This program was begun in 1989. The initial emphasis was to assemble
existing information on the greenhouse gases in order to evaluate the needs of the
EPA greenhouse gas (GHG) research emissions program. As the EPA/ORD GHG
research program developed, it became apparent that a computerized database
was needed to act as a repository for the data and to enable frequent updates of
data as newer or better quality data became available. Initial in-house efforts were
directed at spreadsheets using existing, commercially available software. It soon
became apparent, however, that software designed specifically for emissions data
would be superior to general-purpose spreadsheets. Professionally-designed software
could handle emission factors and activity data separately and also could link
references to each piece of data to ensure clear data pedigrees.
An initial effort at a professionally-designed software package resulted in
software called the Global Resource and Emissions Model (GREM). GREM, developed
using the CLIPPER software development shell, was found to have several serious
technical flaws. It could not rectify inconsistent units, was slow, and required large
amounts of computer memory. Consequently, a new software tool called the Global
Emissions Database (GloED) was developed using the "C" programming language.
GloED was found to meet the needs of AEERL data repository/calculator. Further, it
has generated substantial interest in the international research community as other
researchers have found the need for GHG emissions data handling tools.
-»
The efforts directed towards international inventories of GHG have been aimed
first towards development of the GloED software and development of emission
factors and activity data for anthropogenic sources of methane and nitrous oxide.
Methane was chosen as a priority because of AEERL's focus on methane emissions
research.
The development of data on emissions of nitrous oxide began initially with the
verification of a published hypothesis that adipic acid manufacturing plants may be
significant contributors of anthropogenic N2O. In the process of reviewing the
literature for information on N2O, it was concluded that there is a strong potential
that uncertainty in the anthropogenic contribution of this important GHG can be
reduced substantially with relatively little financial resources.
-------
Objectives
The ultimate objective of this project is to assemble databases on emission
sources and sinks for all of the GHG on a country and sector-specific basis.
Databases on mitigation technologies will also be assembled. Most of these data will.
be extracted from the published literature, though some data will be generated by
AEERL and some estimates will be made based on our knowledge of emission factors,
activity data, mitigation potential, cost of technology, and other known factors.
Additionally, database management software will be developed as a repository for
the data and to allow simple data processing.
Approach
The requirements and needed features of the software were conceived in-
house. Radian Corporation, an engineering contractor with extensive experience in
development of computer software and emission inventories, has been commissioned
to provide and test the computer code for the GioED software. Radian Corporation
and other engineering contractors are developing emissions data with which to
populate the software.
This project focuses on development of emissions software first, then partial
population of the software with emissions data. Mitigation software will be developed
and populated later. .
The GloED software system is designed as a tool for generating estimates of
global emissions. It generates emission inventories by combining information about
activities with pollutant-specific emission factors for those activities. Activities are
defined in terms of processes that occur in a specific pollutant source category in a
specific country or at a specific latitude and longitude. Activities are grouped into
discrete data sets within the GloED system. The user selects one or more data sets
and then has the option of narrowing the scope of the inventory by selecting a limited
number of countries, source categories, and pollutants. The final set of data selected
is called a scenario. GloED also can accept data provided by the user. Consequently,
the emissions inventories can be updated as new data become available.
The contents of an emissions inventory scenario can be displayed by GloED in a
variety of ways. A text summary of the emissions inventory will produce a tabular
breakdown of the results by country, source category, and/or pollutant. GloED can
develop a pie chart or bar chart showing the top ten pollutants or countries or
source categories in a form that allows easy comparison among them. Finally, GloED
can display the results of an emissions inventory onto a global map, using different
-------
colors to designate the type and distribution of pollutants in the selected scenario. All
of these output formats can be viewed on the screen, saved to a file, or printed as a
hard copy. The data can also be exported to Lotus 1-2-3, dBase, or ASCII.
In order to maintain a clear data pedigree, each piece of data is referenced by
a publication citation and the name of the person(s) entering the data. Space is also
provided on several of the software screens for entry of notes and other information.
Population of the software will begin with available information on GHG
emissions data on a country and source specific basis. This can be thought of as
filling a 3-D matrix. This 3-D matrix can be envisioned as a matrix cube with countries
along the vertical axis, GHG along the horizontal axis, and sources or sectors filling the
third dimension. The software will first be loaded information on methane emissions.
After the data are loaded and quality checked, additional information will be
estimated to enable a global inventory of all GHG. It is recognized that some of the
estimates will be based on very weak information. However, data quality will be
identified throughout the matrix to identify where data quality need to be fortified.
Though GloED will continue to be the primary emphasis of database software
development throughout 1993, development is being considered for a companion
database that will contain information on GHG mitigation technologies. This will be
called the GloTech (global technology) database. It will be an electronic file cabinet
that will house GHG mitigation technology and will report parameters such as
emissions reduction capability, cost, and date of availability of the technology. Once
populated with data, GloTech will allow scenario development and file interaction
similar to GloED. This will enable the user to perform cost effectiveness calculations
for an array of technologies that will be constructed in a scenario. Once the scenario
is constructed the user can determine total cost, total emissions reduction
performance, and other parameters such as secondary impacts (water, solid waste,
etc.) and estimated dates of ability of the technology and limits to market
penetration. Like GloED, GloTech will allow construction of the scenarios based on
information resident within the software or data provided by the user. GloTech will
also have each piece of information linked to its reference to ensure a clear data
pedigree. ->-.-.
Once GloED and GloTech are operational and fully tested, they can be linked to
that the user can perform "what if scenarios. This will be helpful to policymakers
since it can .show how implementation of certain technologies would affect country
specific emissions, and what the cost of those technologies would be.
-------
Resources
FV
$K
. 90
125
91
155
92
. 235
93
236
Preliminary Results
Initial reviews of prototype GloED software have been very positive. The GloED
software (GloED Version 1.0) was presented during a discussion of database
software needs at the Intergovernmental Panel on Climate Change (IPCC) GHG data
methodology workshop in Geneva, Switzerland in December, 1991, and generated a
great deal of interest from the attendees. On May 27, 1992, a prototype of the
software was demonstrated at the request of the Manager of the Organisation for
Economic Co-Operation and Development (OECD) Programme on GHG Emissions
Inventories during his visit to Washington, DC. That demonstration precipitated
discussions regarding the .utility of GloED as a repository for GHG emissions data for
the IPCC/OECD.
*
The GloED software was modified to accommodate the specific needs of the
OECD regarding hierarchical data entry and user-friendly data editing, and the OECD
version (GloED Version 2.0) of GloED was delivered in January, 1993. Though the
OECD has been unable because of other commitments to provide a rigorous
evaluation of the revised GloED, their preliminary reaction has been very positive. It is
anticipated that GloED will be selected as the IPCC/OECD data repository in the Fall of
1993.
GloED is now populated with sufficient data to produce a global inventory of
methane. Databases in the global methane inventory are:
- Rice Fields
- MSW Landfills
- Coal Mines
- Natural Gas Systems
- Enteric Fermentation
- Motor Vehicles
Fuelwood Combustion
- Miscellaneous Sources
A report on N2O from adipic acid manufacturing plants was completed in FY91.
-------
A design document for GloTech was completed in 1991. Because of limited
funding and competing priorities, the design has not been implemented.
Future Activities
GloED Version 1.0 software (the general-use version) will be completed by the
end of FY93, and will be distributed for peer review. An initial population with available
methane and nitrous oxide databases will also be completed by the end of FY93.
Also, a report will be published which quantifies global anthropogenic nitrous oxide by
source.
It is anticipated that GloED Version 2.0 (the OECD version} will be accepted by
the OECD by the end of FY93. It is also anticipated that some modifications to the
software will be requested by the OECD. These modifications will be completed in
FY94.
GloED Version 1.0 software will be refined and populated with available data for
the remaining major GHG and ozone precursors (CO2, CO, CFC's NOx, and VOC) in
FY94. -
An important future enhancement will be a modification to allow for the
addition of gridded data, which will render the software more useful to atmospheric
chemistry modelers. Later, the software will be integrated with GloTech. A temporal
dimension may also be added to show historical changes in country/sector emissions
and to enable projections of emissions into the future based on political, economic, or
technology assumptions. These additions and enhancements will be added as funding
becomes available.
A report on the global sources of anthropogenic N2O will be published in FY94.
-------
GHG Emissions Estimation/Database Management
METHANE EMISSIONS FROM COAL MINES
Introduction/Background .
Methane (CH4) is a radiatively,important trace gas which may account for
about 18 percent of anthropogenic greenhouse.warming. Atmospheric
concentrations of methane are now increasing at the rate of 1 percent year year.1
Although the global CH4 cycle is not fully understood, known sources of emissions
include rice paddies, ruminants, termites, biomass burning, landfills, wetlands, tundra,
natural gas transmission facilities, and coal mines. Improved emission estimates for
these sources will allow their relative contributions to global emissions to be better
understood and will allow potential targets for mitigation to be identified for
significant anthropogenic sources.
AEERL has, for several years, been developing data to refine the estimate of
global CH4 emissions from coal mines. Current global mine emissions estimates range
from 25 to 65 Tg ChVyear, which corresponds to roughly 10 percent of total annual
CH4 emissions from anthropogenic sources.* These estimates are generally developed
from knowledge of global coal production and simplistic emission factors derived from
coalbed CH4 contents. Surface mines may be underrepresented in these estimates
and abandoned mines and handling facilities have been wholly ignored. Research
underway will greatly reduce the uncertainty in these estimates and will include,
through an emissions measurement program, those categories previously unmeasured
or omitted.
Objectives
1.
2.
3.
. . 4.
To develop a method for calculating methane emissions from
underground mines;
To develop a method for measuring emissions from surface mines;
To measure emissions from a representative set of abandoned mines
and coal handling facilities;
To estimate methane emissions from coal mines domestically and
globally.
-------
Approach
The first objective has been accomplished. A large body of data exists on the
characteristics of U.S. coals and, for deep mined coals, their coalbed CH4 content and
the CH4 content of mine ventilation air. A series of regression equations was
developed which relate coal characteristics to coalbed CH4 content, and coalbed CH4
content to mine emissions. It is now possible to estimate mine emissions directly from
a small set of coal characteristics, using regression equations having R2 values
ranging from 0.56 to 0.71. This is regarded as acceptable for an analysis of this
sort.
A method for measuring CH4 emissions from surface mines has been developed
and validated. It involves the use of open-path Fourier transform infrared (FTIR)
spectroscopy to identify and quantify the concentrations of species in the mine plume,
and dispersion modeling to calculate mine emissions. A preliminary sampling trip had
been made to a large Wyoming surface mine in. the Fall of 1991. With the acquisition
and outfitting of a sampling van and the completion of a method validation study,
sampling will begin in earnest in the Spring of 1994.
The sampling of abandoned mines and coal handling facilities has begun. Both
of these sources conveniently are vented through a few stacks or vents which can be
adequately measured using a vane anemometer and a portable infrared CH4 analyzer.
Anecdotal evidence had suggested that some abandoned mines emit CH4 in
economically recoverable quantities and our measurements program suggests that
there are some large sources. This continuing project will attempt to relate
abandoned mines emissions to coal and mine characteristics so that the results can
be extrapolated on some plausible basis. Handling facilities are of interest because it
has been reported that 25-40% of a coal's CH4 content may be released during coal
preparation and handling. This appears to be a reasonable estimate for some coals
depending on the rate at which they desorb CH4. The sampling program should
establish a relationship between facility emissions and the CH4 desorption properties
of the coals.
Resources
Facilities . ^ '* : '~:T" -~r." -.- - - *
The program has recently acquired and outfitted a GMC van for sampling
emissions at coal mines. The prime contractor on this project has purchased an MOA
-------
open-path FTIR to be used in this effort as well. All equipment necessary for sampling
at abandoned mines and handling facilities has been purchased.
Personnel
Approximately 0.5 person year of EPA technical staff time is devoted to this
project. Contractor support is approximately 2.0 person-years.
Budget
FY
$K
90
40
91
456
92
250
93
310
94
300
95
300
Future Activities
Sampling at surface mines, abandoned mines and handling facilities will continue
for the next few years as funding permits. The duration is not yet clear because it will
depend on the identification of factors that will allow results to be extrapolated. It is
hoped that 3-4 surface mines per year can be sampled and that 2-3 campaigns per
year to abandoned mines and handling facilities can be managed.
References
B. Smith and D. Tirpak, The Potential Effects of Global Climate Change on
the U.S.: Report to Congress. EPA/230-05-89-050, U.S. Environmental
Protection Agency, Office of Policy Planning and Evaluation, Washington,
DC, 1989.
R. J. Cicerone and R.S. Oremland, "Biogeochemical Aspects of
Atmospheric Methane," Global Biooeochemical Cycles. 2 (4): 299 (1988).
-------
GHG Emissions Estimation/Database Management
METHANE EMISSIONS FROM THE NATURAL GAS INDUSTRY
Introduction/Background
The natural gas industry is among those anthropogenic sources of methane
(CH4). Emissions have variously been estimated at 25 to 45 Tg/year globally
suggesting a considerable degree of uncertainty in the emission and/or activity
factors.
The combustion of natural gas produces less carbon dioxide per unit of energy
generated than either oil or coal. For this reason, it has been suggested that global
warming could be reduced in the near term by encouraging fuel switching. However,
methane is a more potent greenhouse gas than carbon dioxide. Since natural gas is
approximately 90 percent methane, leakage of natural gas could reduce or even
eliminate the inherent advantage that natural gas has because of its lower carbon
dioxide emissions. The purpose of this project, therefore, is to quantify methane
emissions from U.S. natural gas industry, not only to determine the advisability of fuel
switching, but also to identify profitable targets for mitigation within the industry.
Objectives
1.
2.
3.
4.
Develop methods for measuring steady emissions or those emissions
nearly constant in time;
Develop methods for calculating unsteady emissions or those'emissions
that are highly variable with time;
Estimate emissions from the domestic gas industry to within 0.5 percent
of production or 100 BCF/year;
Expand this effort to estimate global methane emissions from the gas
industry.
Approach
This cooperative program with the Gas Research Institute (GRI) has been
structured as a three phase study consisting of a scoping phase, a methods
10
-------
development phase and an implementation phase. EPA and GRI conducted
independent scoping studies from which three major problems were identified. First it
was recognized that all types of emissions could not be measured. .Emissions that
were reasonably steady over time could be measured, and those which were highly
variable over time could only be calculated. The second problem was that proven
techniques for measuring steady emissions from all types of sources were not
available nor were appropriate methods for calculating unsteady emissions. New
techniques needed to be developed. Lastly, it was clear that, even if methods for
calculating the unsteady emissions and measuring the steady emissions were available,
the emissions from all sources could not be evaluated because the number of sources
was overwhelming. For example, there are a quarter million gas wells, over a million
miles of pipe, and hundreds of thousands of pressure regulators. Because emissions
from all these sources could not be measured, scientifically defensible techniques
needed to be developed that would allow data obtained for a set of sources to be
extrapolated to similar sources throughout the industry.
Phase II of the project, therefore, became a matter of developing methods for:
1) measuring steady emissions; 2) calculating unsteady emissions; and 3)
extrapolating emissions data. This phase of the project has been essentially
completed and has resulted in the identification of five measurement techniques
applicable to various segments of the industry. Methods of calculating emissions from
the largest sources of unsteady emissions have been determined, and extrapolation
techniques are available.
Phase III of the project is underway and consists primarily of gathering,
analyzing, and extrapolating emissions data. The project is expected to conclude in
the summer of 1994.
Resources
Facilities
Since this project has no laboratory component there are no facilities as such
to describe. Individual subcontractors to GRI are responsible for various
measurement and calculation tasks. These groups include Star Environmental,
Aerodyne Research, Inc., Washington State University, University of New Hampshire,
and Southwest Research Institute. The Radian Corporation under contract to EPA is
responsible for measuring leakage from the underground distribution system and for
project oversight.
11
-------
Personnel
Approximately 0.2 person-year of EPA technical staff time is devoted to this
project. Contractor supportfor the EPA portion of the project is approximately 2.0
person-years.
Budget
FY
$K
90
317
91
235
92
250
93
180
94
200
95
200
Future Activities
FY 1994 Complete estimate of CH4 emissions from U.S. gas industry and identify
potential targets for mitigation.
FY 1994 Begin work to estimate global CH4 emissions from the gas industry.
FY 1996 Present preliminary estimates of global CH4 emissions from the gas
industry.
12
-------
GHG Emissions Estimation/Database Management
DEVELOPMENT OF GREENHOUSE GAS EMISSIONS DATA FROM MUNICIPAL
SOLID WASTE LANDFILLS AND OTHER WASTE MANAGEMENT FACILITIES
Introduction/Background
Waste disposal results in emissions of greenhouse gases including methane
(CH4), carbon dioxide (CO2), nitrous oxide (N20), ozone precursors, and
chlorofluorocarbons. The major sources of CH4 from waste management include
landfills, wastewater treatment lagoons, and livestock waste. Current estimates
suggest that this source accounts for up to 125 Tg/yr or -40% of the estimated
total global anthropogenic emissions of 300 Tg/yr (IPCC, 1992). Landfills have been
estimated to contribute as much as 60 Tg/yr of CH4. Policies are being considered to
reduce greenhouse gas emissions to meet the goals of the United Nations Conference
on Environment and Development held in Rio de Janeiro in 1992. Emissions sources
that are amenable to control - such as landfills -- have been given a high priority for
clarification. (EPA, 1989)
"Waste" CH4 results from the anaerobic decomposition of biodegradable waste
found in landfills, open dumps, waste piles, wastewater treatment lagoons, septic
sewage systems, and livestock waste. This waste CH4 can be a source of pollution as
well as a resource. There are 114 landfill gas (LFG)-to-energy projects in the U.S.
(Thorneloe, 3/92) and 200 LFG-to-energy projects worldwide (Richards, 1989).
Landfill gas is utilized (1) as medium-heating-value fuel, (2) to generate electricity
using internal combustion engines, or gas and steam-fed turbines, and (3) as high-
heating-value fuel in which case the gas is upgraded and fed into a nearby natural gas
pipeline. U.S. landfills currently generate 344 MW of electricity (Thorneloe, 3/92), The
gas that is formed from anaerobic decomposition is typically 50 to 55% CH4, 45 to
50% CO2, and <1% trace constituents.
CH4 is a concern because of its global warming effects and explosive potential.
Emissions of nonmethane organic compounds (NMOC) contribute to tropospheric
ozone which aggravates urban smog and is a concern to human health and the
environment. Other LFG constituents such as vinyl chloride, benzene, carbon
tetrachloride, and methylene chloride are a concern for their cancerous and
honcancerous effects. The Agency has proposed CAA regulations for emissions from
municipal solid waste (MSW) landfills (FR, 1991) which will reduce five health and
welfare effects: (1) explosion hazards, (2) global warming effects from CH4
emissions, (3) human health and vegetation effects caused by ozone formed from
13
-------
NMOCs, (4) carcinogenicity and other possible noncancerous health effects associated
with specific landfill emissions constituents, and (5) odor nuisance (U.S.EPA, 3/91).
Estimates from the proposed.regulations indicate that 621 landfills of the 6,000
existing active landfills would be required to collect and control MSW landfill emissions
(p. 24480, FR, 1991).
The proposed Clean Air Act regulations do not require utilization of the gas.
Although increased CO2 emissions are being traded off for reduced CH4 emissions,
there is a net benefit due to the difference in the radiative forcing capacity between
CO2 and CH4. The radiative forcing capacity of CH4 to CC>2 on a molecular basis is 21
times that of C02 (p. 53, IPCC, 1990). It is hoped that the sites affected by these
regulations will consider LFG to energy as opposed to flaring the gas. The use of
energy recovery for the control of MSW landfill air emissions will result in decreased
emissions of CH4l NMOCs, and toxics. Additional benefits include the conservation of
global fossil fuel resources, reduction of emissions at coal-fired power plants, reduced
dependency on imported oil, and cost savings.to public entities that receive royalty
payments (Thorneloe, 6/92). However, there are many barriers in the U.S. associated
with the utilization of waste CH4.
%
The major sources of CH4 from the anaerobic decomposition of waste include
landfills, wastewater treatment lagoons, septic sewage systems, and livestock waste.
Table 1 presents an estimate of the relative contributions of each of these sources
for the U.S. and globally. These estimates suggest that these sources on average
account for 80 Tg/yr or ~30% of the total global anthropogenic emissions of 300
Tg/yr (IPCC, 1992).
Table 1. U.S. and Global Estimates (Tg/yr) of "Waste" Methane Emissions
Landfills
Wastewater
Treatment/Sewage
Treatment
- -
Livestock Waste
U.S.
Avg.
9
?
4
Range
<6-13)a
Reference
U.S. EPA, 7/92
Safle, 1992
Global
Avg.
30
25
25
Range
20-70
?
(20-30)
Reference
IPCC, 1992
IPCC, 1992
IPCC, 1992
a Potential emissions, not corrected for the amount that is flared or utilized. Approximately
1.2 million tonnes of CH4 is being recovered from U.S. landfills (Thorneloe, 3/92).
14
-------
Estimates of global CH4 emissions were summarized by the IPCC and suggest
that landfills contribute -30 Tg/yr with a range from 20 to 70 Tg/yr (p. 35, IPCC,
1992, Khali) and Rasmussen, 1990). Preliminary estimates generated using AEERL's
empirical model indicate that potential landfill CH4 emissions in the U.S. range from 6.3
to 13 Tg/yr, with an average of 10 Tg/yr.. Global estimates suggest a range of 20
to 40 Tg/yr of CH4 emissions with an average of 30 Tg/yr. Estimates generated
using Bingemer and Crutzen's approach - which is currently proposed as the official
IPCC methodology (OECD, 1991) -- indicate that landfill CH4 emissions contribute 60
Tg/yr globally and 23 Tg/yr in the U.S. (U.S. EPA, 7/92).
The estimates generated using the empirical model are thought to more
accurately reflect the amount of CH4 from landfills that is contributing to the global
CH4 flux (Campbell et al., 1991, Peer et al., 3/92, Peer et al., 1992). The estimate
using the empirical model uses data from landfill gas recovery systems and accounts
for CH4 oxidation and gas recovery efficiency. The data that were used to develop
the empirical model were collected from over 100 U.S. landfills. An EPA report is being
published that documents the development of the model and the estimate of CH4
emissions for U.S. landfills. Future refinements of this estimate will adjust for waste
composition using data being developed on the gas potential of different
biodegradable waste streams.
Estimates for wastewater treatment are less reliable primarily due to a lack of
country-specific data needed to characterize the CH4 potential of municipal and
industrial wastewater treatment. There is also a lack of field data characterizing the
CH4 potential from lagoons (Thorneloe, 2/92). Lagoons (or surface impoundments)
are usually earthen pits used to contain and process wastewater. AEERL is initiating
a field test program in 1993 to collect lagoon characterization data such as the
. biological oxygen demand (BOD) loading, flow rates, and retention time. --
CH4 emissions from wastewater treatment lagoons are not expected to be a
major source in the U.S. since many digesters flare and sometimes utilize the gas to
control hydrogen sulfide emissions. However, lagoons may be a more significant
source in developing countries where lagoons are being more frequently used and the
gas is not controlled. Agencies such as the World Bank (Bartone, 1990) recommend
the use of lagoons for wastewater treatment for developing countries since land
space is readily available, operation is relatively simple, cost is low, and energy
requirements are minimal. This represents a potential opportunity to work with
developing countries to demonstrate that the CH4 can be utilized as an alternative
energy source.
15
-------
Individual onsite wastewater treatment systems, such as septic systems, are
used throughout the world. In China, there are an estimated 10 million biogas pits,
which are designed to produce biogas for household use. However, the majority of
the world does not collect the gas from septic systems. A portion of this CH4 will be
oxidized and some will be emitted to the atmosphere. Field test work by EPA/AEERL
is planned in FY94 to collect data that will result in more reliable estimates for this
source and to determine if this source is amenable to cost-effective control.
The only published global estimate for livestock waste suggests that CH4
emissions from this source are about 28 Tg/yr with a range of about 20 to 35 Tg/yr
(Safle et al., 1992). These estimates were made by collecting information from animal
waste management systems and the quantity of animal waste managed by each
system. Information was also collected from government statistics and literature
reviews. The major uncertainty regarding these estimates is due to the assumptions
and data characterizing the CH4 potential from the waste of free-range animals.
Objectives
The objective of this AEERL program is to develop global estimates of CH4,
COa, N2O, and NMOC emissions for waste management facilities including landfills (LF),
wastewater treatment (WWT) lagoons, septic sewage systems, and livestock wastes.
AEERL has been requested by EPA's Office of Policy Planning and Evaluation (OPPE) and
the Office of Air and Radiation (OAR) to (1) develop reliable country-specific and
global estimates for other waste management facilities including WWT lagoons, septic
sewage systems, livestock wastes, and industrial wastes and (2) project future
emissions for these sources and identify mitigation opportunities.
Approach
-.5
Data collected by AEERL in FY91 resulted in the development of an empirical
model which has been used to develop global and country specific estimates of CH4
from landfilled MSW. Field test data characterizing the emission potential will be
collected for WWT lagoons and an empirical model will be developed. Country-specific
data on municipal and industrial lagoons will be collected for developing global and
country-specific estimates. Uncertainty/sensitivity analyses will be conducted that
assess the reliability of the estimates for WWT lagoons. Lab studies and/or field
studies will be conducted to develop more reliable emission factors for livestock
waste. Global and country-specific estimates will be developed including an uncertainty
analysis of the estimate. Using data collected in FY92 on global waste management
trends, estimates will be made of future emissions of landfilled waste. Gas potential
data for specific waste streams is being developed through a North Carolina State
16
-------
University cooperative agreement. This data will be used in FY93 to refine existing
estimates for MSW landfills and to evaluate N2O emissions.
The.AEERL research program is-being coordinated with EPA's Office of Air
Quality Planning and Standards (OAQPS) and the Office of Solid Waste as well as with
the Department of Energy. The research is being coordinated internationally through
the International Energy Agency (IEA). The principal support contractor
responsibilities for this program area as follows:
Radian Corporation - Field testing of waste management facilities
Research Triangle Institute - Development of GHG emission factors for
wastewater treatment and septic sewage systems
E.H. Pechan & Associates - Development of U.S. and global GHG emissions
estimates from landfills and other waste management processes
Resources
Appropriately .6 person-year of AEERL staff time is dedicated to this program
area. Contractor support includes 4.1 person-years.
FY
$K
90
.200
91
225
92
450
93
290
94
360
Preliminary Results
Key results achieved,to date include the following:
1. Developed a methodology for estimating methane emissions from landfills
and wastewater treatment facilities
2. Developed preliminary global and country-specific estimates of methane
emissions from landfills and wastewater treatment facilities
3. Lead-authored two chapters in OPPE: Report to Congress on
International Inventory of Anthropogenic Methane Emissions
4. Served on and developed a database of U.S. landfill to energy projects
for use by the IEA Expert Working Group on Landfill Gas
17
-------
5. Initiated formation of a technology transfer program on greenhouse gas
emissions from waste management facilities
6. Serve on Steering Committee of the Solid Waste Association of North
America
The following documents describe in more detail results achieved to date in this
program:
Papers
Thorneloe, S. A. "Landfill Gas Utilization - Options, Benefits, and Barriers." Presented
at The Second United States Conference on Municipal Solid Waste Management,
Arlington, Virginia, June 4,1992. Published in Conference Proceedings.
Thorneloe, S. A. "United States Research on Enhancing Landfill Gas Production."
Chapter in Landfill Gas Enhancement Test Cell Data Exchange. Final Report of the
Landfill Gas Expert Working Group. April 1992.
Thorneloe, S. A. "Landfill Gas Utilization - Options and Economics." Presented at the
Sixteenth Annual Conference by the Institute of Gas Technology on Energy from
Biomass and Wastes, Orlando, Florida, March 15,1992. Published in Conference
Proceedings.
Peer, R. L., S. A. Thorneloe, D. L. Epperson. "A Comparison of Methods for Estimating
Global Methane Emissions from Landfills." Presented at NATO Conference on Methane
Emissions in October, 1991. Accepted for Publication in Chemosphere.
Thorneloe, S. A. "U.S. EPA's Global Climate Change Program - Landfill Emissions and
Mitigation Research." Presented at the Third International Landfill Symposium in
Cagliari, Italy, October 14,1991. Published in Conference Proceedings.
Thorneloe, S. A. and R. L. Peer "EPA's Global Climate Change Program - Global Landfill
Methane." Presented at 84th Annual Meeting & Exhibition, Vancouver, British Columbia,
Canada, June 1991. Published in Conference Proceedings.
Thomeloe, S. A. "EPA's Global Climate Change Program - Program Plan for Methane
Emissions from Landfills and Other Waste Disposal Facilities." Presented at the Solid
Waste Association of North America's 14th Annual International Landfill Gas
Symposium in San Diego, California, March 26,1991. Published in Conference
Proceedings.
18
-------
Thorneloe, S. A. and R. L Peer "Landfill Gas and the Greenhouse Effect." Presented at
Landfill Gas: Energy and Environment in Bournemouth, England, October, 17,1990.
Published in Conference Proceedings.
EPA Reports
Landfill Gas Energy Utilization: Technology Options and Case Studies,
EPA-600/R-92-116, June 1992.
Development of an Empirical Model of Methane Emissions from Landfills,
EPA-600/R-92-037, March 1992.
Demonstration of Fuel Cells to Recover Energy from Landfill Gas, Phase I Final Report:
Conceptual Study, EPA-600/R-92-007, January 1992.
Analysis of Factors Affecting Methane Gas Recovery from Six Landfills,
EPA-600/2-91-055, September 1991.
Approach for Estimating Global Landfill Methane Emissions, EPA-600/7-91-00* January
1991.
Air Emissions from Municipal Solid Waste Landfills - Background Information for
Proposed Standards and Guidelines, EPA-450/3-90-OII, March 1991.
Landfill Air Emissions Estimation Model - User Friendly Computer Software and User's
Manual, EPA-600/9-90-085a,b, December 1990. (Prepared for EPA's Control
Technology Center)
Book Chapters
Thorneloe, S. A., Barlaz, M. A., et al. Global Methane Emissions from Waste
Management." In NATO book, The Global Methane Cycle: Its Sources, Sinks,
Distributions, and Role in Global Change. Accepted for publication - 1993.
Beck, L. L., Piccot, S. D., and Kirchgessner, D. A. "Methane Emissions from Industrial
Sources." In NATO book, The Global Methane Cycle: Its Sources, Sinks, Distributions,
and Role in Global Change. Accepted for publication - 1993.
1 9
-------
Future Activities
Develop estimates for other GHG emissions (i.e., N2O, NMOC, NOX, C02) from landfills
and project future emissions, for landfills and revise existing estimates based on field
data results - FY94-95
Develop more creditable country-specific and global estimates for wastewater
treatment lagoons, septic sewage systems, and livestock waste - FY94-95
Determine methane oxidation potential for landfills and how this may be enhanced as
mitigation strategy - FY95
Project future GHG emissions for "other" waste management facilities including
wastewater treatment lagoons, septic sewage systems, livestock waste, and open
dumps - FY96
Continue development of technology transfer tools that encourage utilization of waste
methane sources - FY94-96
20
-------
GHG Emissions Estimation/Database Management
DEVELOPMENT OF GLOBAL EMISSIONS INVENTORIES FOR TROPOSPHERIC
OZONE PRECURSORS AND AEROSOLS FROM ANTHROPOGENIC SOURCES
Introduction/Background
The supplementary report to the 1992 scientific assessment of climate change
prepared by the International Panel on Climate Change (IPCC) identifies key areas of
uncertainty which prevent an adequate understanding of future climate change trends
from being attained, thereby creating opportunities for further research. The
supplement calls for a better understanding of the direct and indirect effects of future
anthropogenic emissions of greenhouse gases, tropospheric ozone precursors, and
aerosols on radiative forcing and hence global climate. Of the long-lived greenhouse
gases, one of the most important in terms of positive radiative forcing is methane.
Ozone is thought to be another radiatively important tropospheric gas. As a result of
a variety of complex atmospheric chemical and photochemical reactions involving
short-lived, inhomogeneously distributed ozone precursors such as nitrogen oxides
(NOX), non-methane hydrocarbons (NMHCs), and carbon monoxide (CO), the radiative
budget of the atmosphere is impacted through changes in the atmospheric
concentrations of ozone and methane. Ozone precursors affect methane
concentration indirectly by changing the hydroxyl radical (OH) sink responsible for the
oxidation of methane in the atmosphere. It has also been suggested that aerosols
may have a significant negative effect on radiative forcing.
Attempts to predict future climate change from all the contributions above will
rely on large-scale atmospheric chemistry models. A necessary starting point in any
model which attempts to predict climate change is an estimate of future-afTmospheric
concentrations of the greenhouse gases and aerosols. However, in order to develop
estimates for methane and tropospheric ozone concentrations, one must include the
contributions of NOX, NMHCs, and CO. This requires information on the emissions of
these species and the chemical reactions which affect methane and ozone
concentrations. Improved emissions inventories for NOXI NMHCs, CO, and
aerosols are needed prior to improving predictions of future climate
change. This has been the primary focus of the Air and Energy
Engineering Research . Laboratory's (AEERL) research efforts in this area.
With an increased understanding of the effects of future anthropogenic emissions on
global climate change will come opportunities for prevention/mitigation which will most
likely include prioritized options for controlling emissions for certain pollutant/source
combinations. -
21
-------
Objectives
A joint research program between AEERL and the Atmospheric Research and
Exposure Assessment Laboratory (AREAL) was established with the following
objectives in mind:
1. To establish base-year and future emissions for NOX, NMHCs, CO, and
aerosols on a global scale (AEERL).
2. To develop means for quantifying the effects of these emissions on
radiative forcing (AEERL and AREAL).
3. To establish global tropospheric trends for OH, ozone, and methane via
visualization (AREAL).
4. To develop the means for making theoretical predictions of methane and
ozone concentrations via modelling (AREAL).
The program summary which follows primarily covers activities involving AEERL
which are related to emissions and radiative forcing (objectives 1 and 2). Only limited
information is provided on AREAL's activities which are related primarily to
atmospheric chemistry (objectives 3 and 4 ). The activities of both Labs are
interrelated due to the fact that emissions and radiative forcing are strongly linked by
atmospheric chemistry considerations.
Approach
The approach to establishing emissions data for NOX, NMHCs, CO, and aerosols
has been to review the scientific literature and databases for data on the major
source categories. From this information, gaps in the emissions data are identified.
Knowledge of these shortcomings in the data has led to recommendations for
generating new data and new estimation approaches for filling the gaps. It is
anticipated that future inventory efforts will proceed in a "one at a time" fashion in
terms of anthropogenic source categories. For each category, an attempt will be
made to include all the pollutants of. interest. Source categories receiving the highest
priority will probably include those with the highest rate of growth in their emissions
and the greatest mitigation potential.
By developing and utilizing global chemistry models, the indirect effects of NOX,
NMHCs, and CO on radiative forcing are being quantified. This approach has involved a
22
-------
comprehensive review of existing data on radiatively active substances and
mechanisms and a specialized effort to quantify the effects of direct-injection
emissions of aircraft in the upper troposphere on ozone, OH, and other species. More
advanced global chemistry models are also being developed to make more accurate
theoretical predictions of atmospheric methane and ozone concentrations using
precursor emissions data. A visualization approach has also been developed for
establishing global trends of greenhouse gases and precursors.
Resources
Personnel: Approximately 1 person-year of EPA staff time and 3 person-years of
Contractor time is dedicated the project.
Budget: In FY92-93, $457K was provided for Contractor support.
Preliminary Results
The purpose of an original effort to gather and assess emissions data for NOX,
NMHCs, and CO was to "pave the way" for a plan to develop global emissions
inventories for tropospheric ozone precursors. This effort emphasized the most
important anthropogenic source of NOX, namely, fossil fuel combustion, since NOX is
thought to control the ozone formation reactions involving CO and NMHCs. Of the
fossil fuel combustion sources, the largest sectors - transportation, utility, and
industrial - were selected for more focused research. Using existing sources of activity
data, emission factors, and spatial data, methodologies were developed to generate
gridded emissions inventories for the transportation, utility, and industrial sectors.
These methodologies attempted to spatially allocate emissions into 1 degree latitude
by 1 degree longitude grid cells within the country boundaries. This spatial resolution
for the gridded inventories was chosen so as to meet the needs of atmospheric
chemistry modelers.
Of the transportation sources, motor vehicles received the most attention since
they are the largest contributors to the generation of tropospheric ozone precursors.
The methodology for generating a gridded inventory for motor vehicle emissions
involved the estimation of vehicle emissions of NOX, NMHCs, CO, and other pollutants
for individual countries and then the spatial allocation of these emissions into individual
grid cells. Country-specific emissions were computed by multiplying traffic volume, in
terms of vehicle miles travelled (VMT) for various classes of vehicles, by appropriate
emission factors. The spatial allocation of emissions was based on allocating the
country traffic volume according to the expected traffic density in a given grid cell.
23
-------
This traffic density estimate was derived by calculating the total distance of roads in
each cell using a computerized chart for the globe. Graphical representations of the
spatially distributed inventories were generated for a limited portion.of the globe, for
demonstration and evaluation purposes.
Progress has been made in the identification of several effects that contribute
to radiative forcing which has contributed to ongoing modification of global warming
potentials. A computer code is available for calculating GWPs as a function of primary
factors including future atmospheric background concentrations resulting from various
emission scenarios. A computerized visualization approach for establishing global
trends of greenhouse gases and precursors has been demonstrated for ozone
working with NASA and using data sets obtained from NASA for the years 1985
through 1990. Some of these visualizations were used in the planning of the 1992
TRACE-A program coordinated by NASA. Some precursor visualization has been
initiated for NOX with NMHC, CO, and methane visualizations to follow.
Future Activities
FY93: Assist in the development of global emissions inventories by taking
advantage of scientific efforts already complete, those underway, and those planned
for the future. Enhance quantification of direct and indirect effects on radiative forcing
from methane, ozone and its precursors. Enhance the visualization studies of global
tropospheric ozone.
FY94: Develop enhanced global emissions data for NOX, NMHC, CO, and
aerosols and assess their effect on radiative forcing from resultant changes in
methane and ozone concentrations. Continue global visualization studies for NOX,
NMHC, CO, and methane.
FY95: Quantify direct-injection emissions from aircraft in the upper troposphere.
Evaluate prevention and control strategies. Continue visualization studies.
24
-------
Methane Mitigation
DEMONSTRATION OF FUEL CELLS TO RECOVER ENERGY/CONTROL METHANE
EMISSIONS FROM WASTE METHANE GAS SOURCES
Introduction/Background
This program involves two waste methane recovery projects using phosphoric
acid fuel cell (PAFC) power plants to produce electricity from landfill gas (LFG) and
from anaerobic digester gas (ADG). Proposed EPA regulations will require LFG
collection and control equipment. Sludge, generated from wastewater treatment
plants, continues to be a problem in terms of finding disposal options. Anaerobic
digesters are used for sludge disposal and in the process produce methane gas.
Commonly used control technologies include boilers, internal combustion engines, gas
turbines and flares. Fuel cells are a potentially superior technology because they are
both a highly efficient and environmentally clean means of converting LFG and ADG to
electricity and clean heat. However, technical and economical questions remain as to
their ability to utilize these gases, and these projects will answer those questions. By
continuing to fund these projects, EPA will provide added impetus to the
commercializations of fuel cells and, at the same time, utilize the technology in a mode
that maximizes pollutant control (VOCs and toxic compounds) and energy generation.
This technology can potentially reduce methane emissions of 9 Tg/yr and 5 Tg/yr from
landfills and wastewater digesters, respectively. The value of the electricity generated
should offset the control cost. Thus, a significant reduction in global warming
emissions may be achieved at no net cost to the consumer.
Objectives
To demonstrate that methane control and subsequent energy recovery from
LFG and APG via fuel cells are technically, economically, and environmentally feasible two
key issues must be addressed: to define a gas pretreatment system to render the
waste methane gas suitable for fuel cell uses and to design the modifications
necessary to ensure that rated power is achieved from the dilute methane fuel. Known
engineering modifications, albeit initially costly to implement, are required to ensure
that rated power is achieved from the dilute waste methane gas. The toughest and
most critical problem is the gas cleanup system. Therefore, much of the R&D focus of
this program is on the waste methane gas contamination problem.
LFG constituent compounds reported by USEPA indicate a typical value for the
total NMOCs of 2700 ppmv, ranging from 240 to 14,000 ppmv. The NMOC
concentration in the landfill gas Is an important measure of the total capacity required
25
-------
in the gas pretreatment system, while the specific individual analyses provide a basis
for gas pretreatment subcomponent sizing. The specific contaminants in the LFG, of
interest to the fuel cell, are sulfur and halides (chiefly chlorides and fluorides). The
sulfur level ranges from 1 to 700 ppmv, with a typical-value on the order of 21 ppmv.
Sufficient data are not available to assess the range of the haiides, but a typical value
of 132 ppmv was calculated for this contaminant. The range of contaminant values
varies not only from site to site, but also at any given site with time due to seasonal
weather or moisture content. These characteristics require the pretreatment system
design to be capable of handling these gas quality variations to avoid expensive site
specific engineering of the pretreatment design which would affect the marketability
and economics of the concept.
The major contaminant in ADG which will affect fuel cell operation and
performance is hydrogen sulfide. Levels may be as high as 100 to 200 ppm, but
typical values are below 10 ppm. ADG is distinctly different from LFG by the absence of
heavy unsaturated hydrogen-carbons, and sulfur and halogen containing hydrocarbon
compounds. This will allow simplification in the ADG pretreatment process relative to
the LFG process.
Approach
To achieve the above mentioned objectives, two demonstration projects have
been initiated. They both have three phases. Phase 1 involves a conceptual design,
cost, and evaluation study. Phase 2 activities include the engineering design,
construction, and testing of the waste methane pretreatment modules to be used in
the demonstrations. Its objective will be to determine the effectiveness of the
pretreatment systems to remove critical fuel cell catalyst poisons such as sulfur and
halides. Phase 3 of the projects is a one-year field demonstration of the fuel cell
energy recovery concept at an existing LFG-to-energy facility and a wastewater
treatment plant with anaerobic digesters. The site selected for the LFG/fuel cell
demonstration is at Pen rose Station, an existing LFG-to-energy facility owned by Pacific
Energy Corporation in Sun Valley, CA (an industrial suburb of Los Angeles). The
ADG/fuel cell demonstration will be conducted at the Back River Wastewater
Treatment Plant located in Baltimore, MD.
Results : :",'.- r- .;* : - -
International Fuel Ceils Corporation (IFC) was awarded a contract by EPA in
January 1991 to demonstrate energy recovery from LFG using a commercial
phosphoric acid fuel cell power plant. Phase 1 of the three-phase project has been
completed {EPA Report EPA-600-R-92-007). Phase 2 activities began in September
1991. The overall design strategy incorporated by IFC is to reduce the primary fuel
26
-------
cell contaminants (sulfur and halides) to levels between 1 and 10 ppm. A nominal value
of around 3 ppm is the design goal. The landfill gas pretreatment system design
incorporates two stages of refrigeration combined with three adsorbent steps. The
use of staged refrigeration provides tolerance to varying landfill gas, constituents. The
first stage significantly reduces the water content and .removes the bulk of the heavier
hydrocarbons from the landfill gas. This step provides flexibility to accommodate
varying landfill characteristics by delivering a relatively narrow cut of halogenated ..
hydrocarbons for the downstream beds in the pretreatment system. A molecular sieve
dehydration bed removes the remaining water in the gas to protect downstream
equipment from freezing. The second refrigeration step removes additional lighter
halogenated hydrocarbons by non-steady state adsorption in the liquid film on the
heat exchanger tubes and enhances the effectiveness of the activated carbon and zinc
oxide beds, which remove the remaining volatile organic compounds and hydrogen
sulfide in the landfill gas. This approach is more flexible than utilizing dry bed
adsorbents alone and has built-in flexibility for the wide range of contaminant
concentrations which can exist from site to site and even with a single site varying with
time.
The construction of the skid-mounted gas pretreatment unit is complete.
Factory check-out of the unit is also finished. The check-out operation included: a
manual valve operation and sequencing procedure; verification of the unit to sustain
maximum process flow at the design inlet pressure and within predicted pressure
drops; and verification of electrical control by the process logic controller. Nitrogen
was used as the test gas.
The pretreatment unit has been shipped to the CA test site and tests are
underway to ascertain the effectiveness of the unit to clean landfill gas. The fuel cell is
scheduled to be connected to the clean-up unit around September 1993.
IFC was also awarded a contract in October, 1992 to demonstrate energy
recovery from ADG. To date, most of the effort has been involved in completing a
speciated gas analysis for Baltimore's Back River Plant. The gas analyses reveal that
total sulfur contaminants are in the order of 6 ppm and total organic contaminants
are less than 1 ppm. No detectible halides were present in the gas stream. These
data imply that gas pretreatment to remove contaminants that poison the fuel cell will
be a fairly easy proposition. Most likely, a zinc oxide bed to remove sulfur
contaminants will be all that is required.
27
-------
Resources
The funding levels for both programs are .listed beiow:
LFG Demo (Los
Angeles) F$K1
ADG Demo
(Baltimore) f$Kl
FY91
747
0
FY92
1015
658
FY93
1700
*
FY94
1100
300
FY95
300
1100
FY96
0
1400
FY97
0
1400
FY98
0
1250
* Contract was not awarded until the end of FY92. $658K of FY92 funds
are sufficient to carry project through FY93.
Future
FY94: Conduct LFG demo. Develop the engineering design for the ADG pretreatment
system and determine requirements for engineering modifications to the fuel
cell to achieve maximum power output on ADG. "**
FY95: Finish the LFG demo and produce a report on the one year operating history
of the fuel cell and LFG pretreatment unit. Build the ADG pretreatment unit.
FY95: Incorporate and test the engineering modifications to the fuel cell to achieve
full rated power from dilute methane fuels (ADG). Test the ADG pretreatment
unit.
FY96: Conduct the ADG demo.
FY96: Finish the ADG demo and produce a report on the one year operating history
of the fuel cell and ADG pretreatment unit
28
-------
Methane Mitigation
ENHANCED COALBED METHANE RECOVERY
Introduction/Background
If the estimate of methane (CH4) emissions from coal mines is thought to be
sufficiently large, it may be considered prudent to utilize CH4 emissions from coal
mines to the extent possible rather than release them to the atmosphere. In current
underground coal mining practice the majority of CH4 emissions exit with the mine
ventilation air but are too dilute to be economically recovered using current
technology. Although wells are drilled ahead of and behind the mine workings to
relieve some of the costly ventilation air requirements, a consistent program of
premine seam degasification and gas recovery is not common practice. A program
of this type could significantly reduce ventilation shaft CH4 emissions. If the practice
of premine degasification is to be encouraged a higher rate of CH4 recovery must be
achieved, and the gas must be utilized or sold rather than released to the
atmosphere.
AEERL intends to demonstrate the Amoco Production Company's nitrogen
flooding process to enhance the recovery of CH4 from coal seams prior to mining.
Although Amoco's interest in developing the technology is focused on CH4 as the
saleable resource, the methods involved will translate fully from the coalbed CH4
industry to the coal mining industry. The enhanced recovery, if achieved in a premine
degasification program, will allow a mine to reduce its costly ventilation air
requirements, and to retrieve more CH4 for utilization or sale for a given drilling cost.
. In this fashion a consistent program of premine degasification may become not only
less costly, but an actual economic benefit.
In conventional reservoirs CH4 is contained as a free gas. In contrast, CH4 in
coal seams is stored as a gas adsorbed on the internal micropores of the coal
matrix. The conventional practice of recovering coalbed CH4 is to reduce total
reservoir pressure by pumping water out of the coal. Some CH4 desorbs from the
coal surface, migrates through the micropores to the cleat or fracture system, and
then travels to the recovery well along with the water. Although the system is simple
it is inefficient because at the lower economic limit of pumping, about 150 psi (10.07
kg/sq cm), as much as 50 percent or more of the original CH4 may remain in the coal.
An additional drawback to reducing the total reservoir pressure is that the driving
force for gas expulsion is lost.
29
-------
An alternative to reducing the total reservoir pressure is to reduce the partial
pressure of CH4 by introducing an inert, low-adsorbing gas at a constant pressure.
Partial pressure of a component is equal to the total system pressure multiplied by
the component's mole concentration in the gas phase-. Therefore the-injection of
nitrogen reduces the relative concentration of CH4 and hence its partial pressure while,
in some cases, increasing total reservoir pressure. Laboratory studies have shown
CH4 recoveries of over 80 percent as well as significantly enhanced rates of recovery.
Modeling studies suggest that the cost of nitrogen is more than offset by the
improvement in production.
Objectives
The objectives of this project are twofold:
1. To attempt to identify, through laboratory research, a gas or gases more
effective and/or less expensive than nitrogen for enhanced methane
recovery;
2. To transfer the enhanced coalbed methane recovery technology to the coal
mining industry, primarily through a demonstration at an eastern
underground coal mine site.
Approach
To optimize the performance and cost effectiveness of the injection system
laboratory studies will be conducted on a suite of coals representative of both
eastern and western coal basins. Adsorption/desorption curves for moist-coals at
reservoir temperatures will be determined for binary, tertiary and multicomponent
systems. Among these gases will be analogs for flue gas and fuel cell off gases.
Results of the laboratory studies will serve as input to Amoco's reservoir model to
determine potential performance of various gases at field scale.
In preparation for the field scale demonstration at an eastern underground
mine site, an applicability study will be conducted to determine those coal basins in
which an enhanced CH4 recovery process would be most successful. This will serve
both to guide our selection of a demonstration site and to assist the coal industry in
determining where the process might be used most profitably. An engineering and
cost analysis will also be performed to identify and analyze the technical issues and
the capital and operating costs associated with various advanced degasification
processes at representative mine types. A broader economic analysis will be
30
-------
performed to identify key economic issues associated with using advanced
degasification processes at both model and actual case study mines. It will also
identify the most economically viable mine types and end use combinations.
Resources
Facilities
Laboratory facilities being used for system optimization studies are located at
the Amoco Production Companies research center in Tulsa, Oklahoma.
Personnel
Approximately 0.3 person-year of EPA technical staff time is expended on this
project with the time being expected to expand with the approach of the
demonstration. Contractor support is approximately 4.0 person-years.
Budget
FY
$K
92
450
93
75
FV
450
95
450+
Future Activities
FY 1992: Begin laboratory studies, the applicability studies, an engineering
and cost analysis and case study review.
FY 1994: Perform economic analysis of process and begin demonstration
planning and design.
FY 1995: Begin demonstration.
31
-------
References
1. R. Puri and D. Yee, "Enhanced Coalbed Methane Recovery," Presented at:
65th Annual Technical Conference and Exhibition of the Society of Petroleum
Engineers, New Orleans, LA, September 23-26, 1990.
32
-------
Biomass Utilization
EVALUATION OF TECHNOLOGIES FOR CONVERSION OF BIOMASS
TO ELECTRICITY AND TRANSPORTATION FUELS
Introduction/Background
The report "Reducing Risk: Setting Priorities and Strategies for Environmental
Protection,"prepared by the EPA Science Advisory Board in 1990, identified the risk of
global climate change among the top four issues confronting the Agency. Carbon
dioxide emission from the combustion of fossil fuels is the dominant anthropogenic
activity affecting greenhouse warming and climate change; evaluating the options for
reducing the rapid increase in atmospheric COg concentration is therefore an
important priority for the EPA. The options for reducing COa emissions include: (1)
recovery from combustion sources followed by CO2 sequester, (2) displacement of
combustible fuels with energy derived from renewables such as wind and solar, or (3)
offsetting COa emissions by utilization of biomass.
There are two principal strategies for utilizing biomass to offset COa emissions.
One strategy involves the sequester of carbon in forests; CO? absorbed from the
atmosphere during photosynthesis .is stored as carbon in the biomass. Another
strategy that has been shown to be far more effective is to grow and harvest
biomass sustainably for the specific purpose of displacing fossil fuels. In this case, the
energy stored in the biomass is utilized in conventional combustion systems and the
CO2 is reabsorbed by the regenerated biomass. Among the renewable technologies,
biomass is unique in that it can be a source of either electricity or liquid fuels. The
technology of choice for producing electricity from biomass, either as forest residues
or short-rotation woody crops produced on energy plantations, is likely to be
integrated gasifier/gas turbine systems. These units would offer higher efficiency than
coal-fired steam/electric power generation as well as lower capital costs. For the
modest scales needed in developing countries, they offer greatest potential for
effective displacement of coal as the premium fuel.
Biomass is the only renewable energy source that can displace petroleum in the
transportation sector by conversion to a liquid fuel. It therefore represents the most
practicable approach for the reduction of COg emissions from that source; which
comprises about 30 percent of U.S. emissions (nearly equal to all combustion
emissions from stationary U.S. sources). The need to replace petroleum with an
alternative fuel is recognized in the National Energy Strategy and PL102-486, which
seek to displace 30 percent of petroleum by the year 2010. If the alternative fuel is
derived from biomass, a major impact on greenhouse gas emissions would be
33
-------
achieved; if it is produced from indigenous feedstocks at a cost that is competitive
with petroleum fuels, a significant improvement of the national economy would also
accrue.
Objectives
i
From the above considerations, the initial focus of AEERL's biomass utilization
program has the purpose of selecting and evaluating the most promising approaches
for production of electricity in small-scale gasifier/turbine systems and for the
production of transportation fuel on a large scale. The specific objectives of the
program are:
1. To define the most energy efficient and economical advanced technologies
for conversion of biomass to electricity, including fuel cells and hydrogen
strategies.
2. To catalyze the development and commercialization of biomass fueled
power generation systems applicable to industrialized and developing
countries.
3. To evaluate technologies for conversion of biomass to liquid transportation
fuel; identify and support the development of the one that can have
greatest impact on reduction of greenhouse gas emissions and
displacement of petroleum fuels.
Approach
To achieve the above objectives, AEERL is supporting a fundamental study at
Princeton University that is evaluating renewable energy strategies for reducing
emissions of greenhouse gases. Our support is focused on biomass energy
conversion systems for power generation, including advanced gas turbine cycles and
biomass integrated gasifier/gas turbine (BIG/GT) systems. Operating scale is an
important factor in these economic and efficiency assessments. This information will
be applied in the second objective to select promising technologies for future
demonstration. In anticipation of a BIG/GT demonstration, AEERL is participating with
the State of Vermont in tests of an aeroderivative turbine system to evaluate its
compatibility with gasified biomass wastes. These tests are part of a comprehensive
technical and environmental feasibility study that is being conducted in cooperation
with Vermont and Brazil as a basis for potential demonstrations of gas
turbine/electric generating systems.
34
-------
The Strategic Environmental Research and Development Program (SERDP) will
assist AEERL with the second objective by providing funding to demonstrate an
innovative energy conversion technology fueled with biomass at a DoD installation.
The DoD .operates a large number of small energy conversion systems that burn fossil
fuels and are in need of repair or replacement. Converting or replacing existing
equipment with systems that utilize biomass would eliminate S02, produce zero net
gain of CO2, reduce air toxic emissions, and reduce waste disposal problems. This
project is an opportunity for the DoD, EPA, DOE, USDA, AID, national labs, and
industry to cooperate in demonstrations that will benefit each organization.
The technical approach is to identify the DoD site, select the most viable
technology, identify the partners, and design, build, and test the system. The
DoD/CERL will provide the demonstration site, specific information to aid the
technology selection process, and system operators. The EPA/AEERL will select
technologies to be considered, evaluate environmental and site specific data, and
coordinate project participant's activities. The DOE/NREL will provide expertise in the
technology selection process. The Regional Biomass Programs, USDA, and ORNL will
provide off-site resource information.. Industry would provide system
development/design and hardware depending on the technology selected.
AID/Winrock would examine opportunities for transferring technology to international
markets. A successful demonstration would allow developing countries to get
approval for financing from multi-lateral lenders.
The coordination between DoD and partners will be such that the design of the
project will be in the best interest of the DoD installation. The biomass fuel supply
would be generated by activities on-site and/or in the community. The potential
systems will be comprised of off-the-shelf components or manufacturable-by existing
industries. The technical risks will be minimized by the proper selection of technology
based on the available site, size of system, type of fuel, qualifications of operators,
and lessons learned by all cooperators. The technologies could be modularized to
allow for varying fuel supplies or energy demand. This project would provide the jump
start needed for the development of equipment, design of systems, and creation of
markets. Conversion technologies fueled with biomass could be applied in developed
or developing countries, industrial sites, or rural areas.
SERDP will also assist AEERL with funding for the third objective which consists
of technical and economic studies of processes for production of alcohol fuels. The
emphasis is on the Hydrocarb process which would produce methanol from biomass
and natural gas in three basic steps: hydrogasification of biomass to produce
methane, thermal pyrolysis of methane to hydrogen and CO, and synthesis of
methanol by reaction of hydrogen with CO. The process currently exists only in
35
-------
concept: AEERL's initial evaluation has involved process simulation studies, based on
thermodynamic equilibrium assumptions for each step, which provide initial indications
of the performance of the system with respect to thermal efficiency, alcohol yield per
unit of feedstock, and potential for C02 emission reduction. Using this information,
economic analyses are performed and comparisons are made with alternative
technologies for producing alcohol fuels from biomass in terms of their potential for
petroleum displacement and for reducing COa emissions from mobile sources.
Because of the importance of alternative fuels to the reduction of ozone and
other air toxics in the State of California, the South Coast Air Quality Management
District requested proposals for the demonstration of new domestic technologies for
methanol production. AEERL, together with Acurex Environmental Corporation
successfully responded with a proposal to cosponsor a demonstration of the
Hydrocarb process in a small, 50-lb/hr unit. That project is now underway, with a
conceptual design and preliminary equipment designs completed. We anticipate a 3-4
year development and test program for this process.
Preliminary laboratory studies of the kinetics of the two principal reactions of
the Hydrocarb process, biomass hydrogasification and methane pyrolysis, were
carried out under AEERL sponsorship. Additional kinetic investigations will be
undertaken in an in-house laboratory to provide fundamental research support as
part of the process development activity. These studies will use a thermobalance
reactor to evaluate gasification kinetics of various types of biomass and a pyrolysis
reactor to investigate the feasibility of combining steam reforming of methane during
the pyrolysis step to further simplify the process by eliminating indirect heating of the
pyrolysis reactor.
Resources
Facilities - Off-site facilities for the Hydrocarb project will be located in the Los Angeles
area, probably adjacent to the California Polytechnic Institute at Pamona. The
process simulations and .preliminary experimental evaluations were carried out under
interagency agreements with the Brookhaven National Laboratory where the process
was conceived. Independent process analysis and economic studies were done at
AEERL. The in-house research lab is being set up in nearby Acurex facilities at RTP
where a dedicated Hydrocarb laboratory has been established to house the
thermobalance reactor which is currently being fabricated under contract by the
Institute of Gas Technology.
Personnel - Approximately 0.8 person-year of EPA staff time is dedicated to the
Hydrocarb project and another 1.4 person-years for biomass utilization for electric
- 36
-------
power generation. Contractor support includes 3.5 person-years for the off-site
project and 0.25 person-year for in-house support. The Princeton University study
supports approximately 0.3 person-year.
Budget - The State of Vermont cooperative agreement totalled $365K for FY90-92
and the Princeton University cooperative agreement totalled $70K for FY92-93. Off-
site contractor support budget totalled $190K at Brookhaven for FY 1990-92 and
$400K at Acurex during FY 1991-92, including subcontracts. $75K has been provided
to establish the in-house Hydrocarb lab and $41K for design of the thermobalance
reactor at IGT. FY93 funds from SERDP are $500K for Hydrocarb and $750K for
utilization of biomass technologies on military installations.
Hydrocarb
Vermont
Princeton
Military
FY90
100K
50K
0
0
FY91
90K
175K
0
0
FY92
444K
140K
20K
0
FY93
600K
0
50K
750K
FY94
500K
0
55K
750K
FY95
500K
0
0
750K
Preliminary Results
Wood chips and bagasse pellets gasified easily in the BIG/GT tests conducted
at General Electric under the Vermont cooperative agreement. The composition and
heating value of the biomass product gas were compatible with gas turbine
requirements. However, the particulate removal performance of the pilot facility single
stage cyclone did not meet turbine specifications. In addition, alkali metal compounds
in the particulate matter carried over from the gasifier exceeded turbine limits.
Improved particulate removal technology, designed specifically for biomass feedstock
characteristics, could meet turbine requirements for both particulate and alkali
loading. Fuel bound nitrogen (FBN) compounds were also measured since they can be
converted to NOX during combustion in a gas turbine. Since this conversion is highly
dependent on gas turbine combustor design, no firm conclusions regarding NOX
production can be reached without actual combustion testing.
Funding for biomass utilization research on transportation fuels has
concentrated mainly on the Hydrocarb process because the results of our
assessments indicate a very large potential for environmental and economic impacts.
As a source of alternative fuel, those assessments show that this process can, in
principal, produce more than three times as much fuel energy from a given biomass
supply as other process options. Since the amount of biomass that can be produced
37
-------
from available land suitable for short rotation woody crops will be the limiting factor
that determines the amount of petroleum that can be displaced and the amount of
CO2 reduction that can be achieved, the yield of liquid fuel from that supply is a critical
variable. The economic assessments have concluded that Hydrocarb should be
capable of producing clean liquid fuel at significantly lower cost than the best
alternative technology. Because the price of methanol produced by this route is
potentially less than the equivalent price of gasoline, CO2 reduction from vehicles using
that methanol would be achieved at no effective cost. In addition, significant
reduction of tropospheric ozone and other air toxics associated with gasoline
emissions is expected.
A project peer review held January 13, 1993, consisting of a panel of 8
extramural authorities on the various technical aspects of the Hydrocarb process,
recommended that the project proceed as proposed in a 50-lb/hr research and
development effort aimed at establishing technical feasibility in actual hardware.
Future Activities
FY93: A site will be selected and a planning procurement package will be
prepared and released to contracts for the first military demonstration. Construction
of the gasification reactor will be completed and testing will begin on the Hydrocarb
project. Fabrication of the thermobalance reactor will be completed for in-house
kinetic studies and a second laboratory reactor will be fabricated and installed for
methane decomposition and steam injection studies.
FY94-95: A contract will be awarded for the first military demonstration and
construction should be completed depending on the technology selected. The second
military demonstration should be funded and the site selection and procurement
process initiated. Hydrocarb gasification reactor tests will be completed;
construction and testing of the pyrolysis reactor will begin. Demonstration of a
biomass gasification/turbine system for generating electric power will be initiated.
FY96-97: Testing of the first military demonstration should be completed and
construction of the second military demonstration should be underway. Testing of
uncoupled Hydrocarb reactors will be completed. An independent evaluation of results
will be undertaken to assess performance, cost, and efficiency.
FY-98-99: Testing of the second military demonstration should be underway.
A Process Development Unit with 24 to 36-in diameter reactors and -methanol
converter will be constructed to evaluate Hydrocarb as an integrated system.
38
-------
EXPANDED FUTURE PROGRAM -
STRATEGIC DIRECTIONS
39
-------
EXPANDED FUTURE PROGRAM - STRATEGIC DIRECTIONS
Summary .
With an expanded EPA program ORD has the potential to head off many of the
disruptive environmental and economic impacts associated with global warming via
strategically focused work on intermediate to long-term prevention. EPA is in an ideal,
independent position to undertake and coordinate overarching R&D to assess and
catalyze integration of attractive prevention-oriented changes over the next 5 to 20
years which will reduce conventional pollutants as well as greenhouse gases. EPA,
given a technological base developed through this R&D, will also be in a position to
incorporate solutions into EPA prevention, regulatory development and enforcement
programs. Special emphasis will be placed on maintaining future commercial
advantages for the U.S. This R&D can be accomplished by working in the following
areas collaborating with industry and other federal agencies:
Assure Development of Comprehensive Global Emissions Inventories
Mitigation/Utilization of Waste Methane
Enhanced Use of Biomass to Displace Fossil Fuels
Evaluation of CO2 Sequestration & Disposal
Mitigation Opportunities for Important Greenhouse Gases Other than
CO2/Methane
Generate Comprehensive Database of Mitigation Technologies
Integration of Renewables into Fossil Energy Production Technologies to
Reduce Greenhouse Gas Emissions
Assure Development of Comprehensive Global Emissions Inventories
A key to meaningful analyses of global climate futures and analyses of
prevention/mitigation options are accurate inventories of greenhouse gases. Such
inventories are woefully inadequate and need to be developed for a substantial
number of substances including: aerosols, carbon black, CO2 , CO, CH4, NMHC,
halogenated HCs, NOX, N20, NH3, CFCs, and SO2. EPA ORD is a excellent position to
coordinate such an effort by extending capabilities used for national inventories of
conventional air pollutants. EPA ORD is sensitive to needs for large scale modeling and
has excellent atmospheric chemistry capabilities. Work is already underway to
enhance global emission inventory data for methane and provide inventory software
for OECD data base requirements. However, a large-scale effort is needed to pull
together existing inventories, evaluate their quality, integrate them on a consistent
bases where possible, and coordinate/develop needed additional data, especially as it
relates to future projections. To do this we will extend the work now being initiated
40
-------
by the Global Emissions Inventories Activity (GEIA) of the International Global
Atmospheric Chemistry Project (IGAC). Initial GEIA work has shown that the individual
efforts by many investigators in developing emission inventories has lead to widely
.varying, incomplete, and largely, unsatisfactory inventories. Widely varying aspects
include different units of measure, spatial resolutions, temporal resolutions, source
characterization, and statistical certainty. This project will provide development of
inventories as rapidly as possible taking maximum advantage of the scientific efforts
already completed, those underway, and those that are planned for the future.
Assessment, coordination, synthesis, filling of gaps, and quality assurance will all be
important components of this effort. Whenever new source emission data is collected
in the field, it will be collected as a function of the various process control operations
and modifications that can reduce or prevent emissions. The aim is to have the best
emission inventory data available as possible for the important needs that lie ahead in
projecting the global climate problem and to evaluate mitigation opportunities.
Mitigation/Utilization of Waste Methane
Methane is a major greenhouse gas with good potential for mitigation.
Radiative forcing reductions can be achieved over the short-term. Currently, EPA has a
limited assessment/mitigation program to fill the R&D gap left by other government
agencies. This program could be usefully expanded to research the strategic options
and then bring technology for prevention/mitigation through the demonstration stage
of development. Emphasis would be placed .on utilization options to conserve
methane and displace other fossil energy use. Waste methane can be a significant
resource once it has been cleaned of sulfur and chlorine compounds. EPA's expertise
in gas cleaning will be used to great advantage here. This work would also include
continuing assessments to identify and evaluate the most important sources and
projection of future atmospheric methane concentrations under various scenarios.
Development of mitigation/utilization technologies would be focused on landfills, waste
disposal (such as landfills, wastewater treatment/anerobic digesters, and septic
systems), petroleum and natural gas production, methane hydrate mining, natural gas
pipelines, and coal mining. A substantial portion of this mitigation/utilization effort
would be researching the applications for developing countries.
Enhanced Use of Biomass to Displace Fossil Fuels
Use of biomass as an alternative hydrocarbon fuel is a major mitigation
opportunity. Long-term options for increasing biomass as a way of enhancing COa
uptake from the atmosphere can only succeed if the biomass is used to displace fossil
hydrocarbons used for energy. EPA R&D can catalyze commercial utilization of this
renewable resource. Research will be focused on solutions which will be applicable to
both industrialized and developing countries. Internationally-diverse applications will be
41
-------
necessary to deal with the growing sector of the problem in developing countries, but
many opportunities exist. Efforts will be focused on technological advancements and
innovative applications of existing technologies to introduce biomass into the existing
and emerging energy infrastructure through waste utilization, optimization of use
(more efficient use), leveraging the existing resource availability through new
processes, and expanding biomass production. For example, the Hydrocarb process
currently under development by EPA offers potential for low cost production of
methanol to displace gasoline in the transportation sector. Another example is the
potential use of Hydrocarb or new wood carbonization processes to eliminate a
widespread, tragically wasteful, and highly-polluting use of wood; this being the
practice of conversion of wood to charcoal for fuel and other purposes in developing
countries using primitive, inefficient processes. Such processes can produce clean
fuels, such as methanol, and clean carbon substitutes for charcoal. A further
technology example is the use of simple turbine-based gasification/combustion or
combustion systems, which are showing developmental promise for power generation.
Research is also needed to resolve the generic problems associated with utilization of
agricultural wastes for energy in low-to-medium tech combustion systems such as
conventional stokers or fluid bed combustors.
Evaluation of CO2 Sequestration & Disposal *
It will be critical for EPA to expand its preliminary assessments of the options
for CO2 sequestration and disposal. Japan and European countries are seriously
assessing (and in some cases developing) such technologies, and they have the
potential to allow use of coal to generate electricity with limited net CO2 emissions.
High costs and high additional fossil energy use flag the need for careful evaluation.
Unfortunately, many of the disposal options may not be permanent, the COa
eventually returning to the atmosphere. Thus it will be critical to have unbiased
monitoring and modeling of rates of CO2 re-release to the atmosphere. Such
approaches may also add other environmental burdens because of their disposal
nature and increased energy use. It will be critical to identify those that are the most
environmentally acceptable. The brute force method to global climate control is to
remove CO2 from power plant flue gas and dispose of it in the ocean or on land.
One problem with this is that only a fraction of CO2 is generated in power plants; the
rest of the CO2 is generated in transportation, heat generation, small area sources,
and non-energy sources. Solutions will need to be.developed for these sources. In
addition if the CO2 is recovered from power plant flue gases or from future advanced
integrated combined cycle plants, the CO2 has to be disposed of somewhere without
re-release. There are multiple issues associated with CO2 disposal: technical and cost
feasibility, environmental effects, the actual CO2 sequestration effectiveness achieved,
sustainability with time, perpetual monitoring requirements, acceptance at all levels by
42
-------
the concerned public, and high costs. It is critical that these problems be resolved via
EPA R&D. Finally, if these technologies are not developed with environmental
considerations incorporated, they may only reduce the peak atmospheric COg buildup
and delay.it somewhat. If undertaken improperly by-Japan or the Scandinavian
countries,, for example, they may create new anthropogenic sources of COa which
could emit even if all direct fossil energy emissions were brought to a halt, and for
which there are no know remedial actions. Thus it will be extremely important to
undertake CO2 disposal assessment including technical feasibility, environmental
effects, and long-term costs. Some preliminary work on the feasibility of disposal
remedial actions is also needed. Our suggested program would also include
assessment of other schemes for atmospheric CO2 fixation/utilization aimed at using
terrestrial or marine biomass. Many of these schemes are both environmentally
questionable or unlikely solutions for other reasons but require assessments to avoid
future problems and to seek workable options.
Mitigation Opportunities for Important Greenhouse Gases Other than
CO2/Methane
While CO2 and methane are the largest positive contributors to the greenhouse
effect, other direct and indirect greenhouse gases need mitigation R&O in order to
have the potential technology options assessed for effectiveness and cost. This will
allow their comparison with the (XVmethane options as means for reducing the
global warming threat. EPA is in a prime position to research the required
prevention/mitigation technologies because many of these substances are already of
concern from a criteria air pollutant standpoint. Extensive background in emission
factor development and emission inventories can help to identify the most important
sources. These problems can be put into three categories: 1) substances that affect
methane atmospheric lifetime, 2) substances that effect tropdspheric ozojje (a
greenhouse gas), and 3) long-lived trace greenhouse gases (including nitrous oxide
and non-hydrogenated chlbrofluorocarbons (e.g., CF4), and others. In the case of
category 1, CO, NMHC (nonmethane hydrocarbons), and NOX all have effects on
atmospheric levels of the hydroxyl radical (and therefore methane lifetime).
Controlling CO and NMHC can help to reduce future methane concentrations in the
atmosphere. For category 2, CO; NHMC, and NOX all have influences on tropospheric
ozone production. Reduction of emissions can reduce tropospheric ozone. Category
3 sources are in general fairly specific sources, and R&O on prevention/mitigation for
these (such as adipic acid plants for N2O and aluminum plants for CF4) can be quite
focused. EPA R&O is used to dealing with such prevention problems and finding the
most cost-effective approach.
43
-------
Generate Comprehensive Database of Mitigation Technologies
Because there are a great many options for global warming mitigation and
because these options have significant consequences for other forms of pollution, a
total environmental impact prospective must be used to sort between global warming
mitigation options. Using its knowledge of current environmental regulation and with
anticipation for future environmental regulation needs, EPA will maintain a life cycle
data base of global warming mitigation options along with consequences for
compliance with all environmental regulation and needs. It is anticipated that with
adequate funding a centralized clearinghouse could be maintained, including a Control
Technology Center and a comprehensive guide to assessments. As a result of
undertaking this project, EPA will be in a position to provide critical information on the
cost o_f ,CO2 emissions avoided for various applications throughout the industrialized
and developing countries. EPA has already established the framework for the
database in the form of a computer program called GloTech, and is interfacing with
assessment efforts, such as that of the International Energy Agency, to determine the
best means of gathering, utilizing and distributing such information.
Integration of Renewables into Fossil Energy Production Technologies to
Reduce Greenhouse Gas Emissions
Incorporating the use of renewables into the existing fossil energy production
infrastructure provides the major option for reducing use of fossil hydrocarbon
reserves. Although it is highly speculative, to the best of our knowledge, no other
federal agency is even considering an assessment of this concept. Optimum use of
renewable resources requires that they be used at cost-effective points in commercial
and industrial operations.
-.*
The end result could be both environmentally and economically attractive. EPA
ORD is in a strategically attractive position to help catalyze this evolutionary process
in an orderly way using its engineering expertise in pollution prevention and sensitivity
to environmental considerations. In addition to reducing greenhouse gas emissions,
the other conventional air pollutants of concern to the Agency can be reduced with no
extra economic burden. Reaching a sustainable use of fossil carbon will obviously
require major restructuring of how we and other countries use the fossil hydrocarbon
reserves in conjunction with a growing renewable energy supply sector. The major
objective would be to make renewable energy the backbone of the system while
constraining and preserving the fossil hydrocarbon reserves. The fundamental
understanding for these technologies already exists. Once the assessments are
completed, EPA policy could be important to overcome the political and other
pressures that can be brought to bear (e.g., international fossil energy interests,
44
-------
which would initially feel they have nothing to gain from such initiatives). One key goal
of this project will be catalyze the development of technology for production of
hydrocarbon fuel using a totally renewable, sustainable carbon cycle involving
integrated yse of renewable hydrogen and biomass (the last bullet abpve). Our
suggested approach would include a number of systems studies at the beginning of
this project to pave the way for future R&D work on all feasible and cost-effective
technology applications.
For example, the following possibilities are options that might be assessed:
International utilization of demand-side solar PV applications for pollution
reduction in industrialized and developing countries
Introduction of solar-PV and wind-based H2 into existing refineries and
petrochemical operations for pollution control and CO2 prevention
Introduction of renewable hydrogen into advanced power cycles. This
has shown promise in early cost studies.
Hydrocarbon fuel production based on a totally sustainable carbon
cycle. A specific concept is Hydrocarb-AR (AH Renewable) which would
use H£ from wind or solar instead of methane as a feedstock. Fuel cost
could be roughly $0.65 per gallon higher initially using current technology
for supplying the hydrogen but could drop substantially with targeted
R&D.
45
-------
-------
APPENDIX 1
AEERL Papers from Proceedings
1992 Greenhouse Gas Emissions and Mitigation Research Symposium
The Air and Energy Engineering Research Laboratory of the U.S. Environmental
Protection Agency sponsored an international symposium to discuss greenhouse gas
emissions and potential mitigation technologies and practices. The 1992 Greenhouse
Gas Emissions and Mitigation Research Symposium was held in Washington, DC,
August 18-20, 1992.
The symposium proceedings provide up-to-date information on emission
sources contributing to global warming, state-of-the-art mitigation technologies and
practices, and the status of activities to refine emission estimates and develop new
technologies. The symposium addressed the following: overview of activities in EPA,
DOE, and EPRI on greenhouse gas emissions and mitigation research, and AEERL's
global emissions and technology databases; international activities of selected *
industrialized and developing countries; carbon dioxide (CO2) emissions and their
control, disposal, and reduction through conservation and energy efficiency, and
carbon sequestration including utilization of waste CO2,' methane (CH4) emissions and
mitigation technologies including such topics as coal mines, the natural gas industry,
key agricultural sources, landfills and other waste management sites, and energy
recovery by fuel cells; biomass emission sources and sinks, including cookstove
emissions and control approaches; and energy sources, solar and renewable including
renewable energy options, alternative biomass fuels, advanced energy systems, solar
energy developments, and woodstove emissions and mitigation.
The attached papers were prepared and presented by AEERL staff.
-------
-------
CONTENTS
SESSION I: OVERVIEW
Frank T. Princiotta. Chairperson
EPA-AEERL
"Greenhouse Wanning: The Mitigation Challenge,"
Frank T. Princiotta 1-1
"Methane Reductions are a Cost-effective Approach for Reducing Emissions of Greenhouse
Gases,"
Kathleen B. Hogan* and Dina W. Kniger 1-25
"Climate Change and Related Activities,"
Kenneth Freidman 1-35
"EPRTs Greenhouse Gas Emissions Assessment and Management Research Program,"
D.F, Spencer* and G.M. Hidy 1-55
"Global Emissions Database (GloED) Software,"
Lee Beck : 1-66
SESSION II: INTERNATIONAL ACTIVITIES
Jane Leggett, Chairpeison
EPA . s
"Beyond Rio,"
Hans van Zijst 2-1
SESSION ID: CO2. EMISSIONS, CONTROL, DISPOSAL AND UTILIZATION
Ken Friedman, Chairperson
U.S. Department of Energy
"Carbon Dioxide Sequestration."
Robert P. Hangebrauck*, Robert H. Borgwardt, and Christopher D. Green 3-1
"The NOAA Carbon Sequestration Program,"
Peter Schauffler 3-14
'Denotes Speaker
iii
-------
"The Role of DOE Energy Efficiency and Renewable Energy Programs in Reducing Greenhouse
Gas Emissions."
Eric Peterson 3-15
"Fuzzy Logic Control of AC Induction Motors to Reduce Energy Consumption,"
RJ. Spiegel*, P. Chappell, J.G. Cleland, and B.K. Bose ." 3-35
"Methanol Production from Waste Carbon Dioxide,"
Stefan Unnasch 3-44
SESSION IV: EMISSIONS AND MITIGATION OF METHANE AND OZONE PRECURSORS
M.J. Shearer, Chairperson
Global Change Research Center
"Global Atmospheric Methane: Trends of Sources, Sinks and Concentrations,"
M.A.K. Khalil, R.A. Rasmussen, and MJ. Shearer* 4-1
"Coal Mine Methane Emissions and Mitigation,"
David A. Kirchgessner* and Stephan D. Piccot i. 4-11
"Emissions and Mitigation of Methane from the Natural Gas Industry,"
Robert A. Lott 4-24
"Emissions and Mitigation at Landfills and Other Waste Management Facilities,"
Susan A. Thomeloe 4-46
"Fuel Cell Power Plant Fueled by Landfill Gas,"
RJ. Spiegel* and GJ. Sandelli 4-58
"Methane Emissions from Rice Agriculture,"
M.A.K. Khalil, MJ. Shearer*, and R.A. Rasmussen .. / 4-67
"Livestock Methane: Sources and Management Impacts,"
Donald E. Johnson*, T. Mark Hill, and G.M. Ward 4-81
"Ozone and Global Warming,"
Robert P. Hangebrauck* and John W. Spence 4-85
"Overview of Methane Energy and Environmental Research Programs in the United Kingdom,"
Suzanne A. Evans. Anton van Santen, Paul S. Maryan, Caroline A. Foster,
Keith M. Richards* 4-97
SESSION V: BIOMASS EMISSION SOURCES AND SINKS
Robert Dixon . .. .
EPA - * --'" -':-'--'- *A'f' ^''^'?' '-
"The Carbon Balance of Forest Systems: Assessing the Effects of Management Practices on
Carbon Pools and Flux,"
Robert K. Dixon*. Jack K. Winjum, and Paul E. Schroeder 5-1
"Denotes Speaker
iv
-------
"Soil Carbon and Agricultural Management,"
Robert B. Jackson IV*. Thomas O. Barnwffl Jr., Kevin B. Weinrich, Allen L. Rowell, Anthony
S. Donigian Jr., Avinash S. Patwardhan 5-32
"Assessment of the Biogenic Carbon Budget of the Fomier Soviet Union,"
Tatyana P. Kolchugina and Ted S. Vinson* 5-46
"Household Fuels in Developing Countries: Global Warming, Health, and Energy Implications,"
Kirk R. Smith and Susan A. Thomeloe* 5-59
"The Potential for Energy Crops to Reduce Carbon Dioxide Emissions."
R.L. Graham 5-79
SESSION VI: ENTERGY SOURCES/SOLAR/RENEWABLE
Robert Williams
Princeton University
"A Global Perspective on Biomass Energy,"
Robert H. Williams 6-1
"An Analysis^ of the Hydrocarb Process for Methanol Production from Biomass,"
Yuanji Dong*, Meyer Steinberg, and Robert H. Borgwardt ,. 6-26
"Alternative Fuels from Biomass,"
Charles E. Wyman 6-39
"Coproduction of Methanol and Power,"
William Weber*. Arden B. Walters, Samuel S. Tarn 6-55
"EPA's Cost-shared Solar Energy Program,"
Ronald J. Spiegel 6-68
"Photovoltaic Developments,"
Jack L. Stone 6-74
"Advanced Energy Systems Fueled from Biomass,"
Carol R. Purvis* and Keith J. Fritsky ... 6-92
"Programs and Policy Impacts Attributable to Regional Biomass Program Wood Stove Research
Efforts,"
Stephen Morgan 6-99
* Denotes Speaker
-------
-------
SESSION I: OVERVIEW 1A
Frank T. Princiotta, Chairperson
GREENHOUSE WARMING: THE MITIGATION CHALLENGE
FRANK T. PRINCIOTTA, DIRECTOR
AIR AND ENERGY ENGINEERING RESEARCH LABORATORY
U. S. ENVIRONMENTAL PROTECTION AGENCY
Human activity has led to an increased atmospheric concentration of certain gases, such
as carbon dioxide, methane, and chlorofluorocarbons, which resist the outward flow of infrared
radiation more effectively than they impede incoming solar radiation. This imbalance yields the
potential for global wanning as the atmospheric concentrations of these gases increase. For
example, before the industrial revolution, the concentration of carbon dioxide in the atmosphere
was about 280 ppm and it is now about 355 ppm. Similarly, methane atmospheric
concentrations have increased substantially and they are now more than twice what they were
before the industrial revolution, or about 1.72 ppm. The impact of man's activities is more
dramatic with regard to chlorofluorocarbons. These compounds do not occur naturally; they
were not found in the atmosphere until their initial discernible production several decades ago.
FACTORS INFLUENCING GREENHOUSE GAS EMISSIONS
The emissions responsible for increasing concentrations of greenhouse gases are
associated with many human activities, especially .the extraction and utilization of fuels, the
large-scale deforestation in many developing countries, and other industrial and.:-*ricultural
practices. Our goal at this conference is to discuss the state-of-the-art and the research
opportunities associated with understanding the sources and mitigating releases of these
greenhouse gases. I submit that to understand the mitigation opportunities we need to understand
the fundamental driving forces for releases of these gases.
Let us concentrate for the moment on carbon dioxide, the most important of the
greenhouse gases. The following expression relates the major factors influencing the growth of
carbon dioxide emissions over time for a given country:
(COj)f = (COa)p x (l+P+Ip+Ej+CeX
where: - , _ . ,.
projected CO2 emissions ... -
present CO2 emissions
annual population growth rate
'r .-"'" ' 'u-^...-.'.- -- 7 ,. (
\ annual growth rate: industrial production V «' annual growth rate:
Jannual growth rate: population -<;.~;J;~j / ' industrial production
'- - - per capita
>,
= I annual growth rate:!
1 energy use per unit'
Sunit of industrial
[production
(CO,),
P
V ;-
annual growth rate: energy use
annual growth rate, industrial
production
1-1
-------
| annual growth rate: carbon emissions I
\ jinual growth rate: energy use j
years into the future
I annual growth rate:
< carbon per unit of
(energy used.
The two-.major factors yielding emission growth are (1) population growth, and (2)
industrialization. These factors can be compensated for by two factors which can mitigate
growth of greenhouse emissions; these are (1) enhanced energy efficiency (i.e., reduced energy
usage per unit of industrial production) and (2) the reduction in the carbon emitted per energy
unit utilized.
It is interesting to note the relative magnitude of these factors expected to influence
emissions of carbon dioxide over the period 1990 to 2025. The Intergovernmental Panel on
Climate Change (IPCC-1992) in their most recent report included a base case of projected
emissions for the greenhouse gases all the way to the year 2100. Table 1 illustrates my
massaging of these data to extract the factors that are important for developed countries
(Organization for Economic Cooperation and Development/OECD countries) versus developing,
or poorer, countries (e.g., Asian countries). The table indicates that, for the developed
countries, the drivers are projected to be primarily economic growth, whereas population growth
is expected to be fairly modest over this time period. The mitigating factors although significant
are projected to be insufficient to counteract population and economic growth, yielding an
estimated 0.7% annual net growth of carbon dioxide emissions from the developed countries.
The situation for the developing countries is even more troublesome. Since their level of
economic activity is currently quite modest, it is projected that they will undergo rapid
industrialization, at the same time that population is growing at a relatively fast pace. Mitigating
factors, namely more efficient use of energy and less carbon intensive energy use, are projected
to be modest over this time period. This yields an expected very large growth of 3.9% annual
increase in carbon dioxide emissions over this period.
Figure 1 illustrates the expected growth in population by area (IPCC, 1992). As you can
see, the developed countries are anticipating relatively low growth rates, whereas the developing
countries, especially Asia, Africa, the Middle East, and (to a lesser extent) Latin America, are
projected to have very large growth rates over the period 1990 to 2100. As stated earlier,
population growth plus rapid industrial growth can yield large increases in carbon dioxide and
other greenhouse gas emissions.
Figure 2 illustrates, based on the expected population and industrial growth, projected
carbon dioxide emissions over the 1990-2100 time frame. The upper graphic in this figure is
the base case for the IPCC 1992 report It projects growth in emissions from about 7.3 to about
20 gigatons* of carbon over this period, with Asia, Africa, and the Middle East providing much
of the projected growth. The lower graphic is a case I developed to illustrate how important it
is for developing countries to move in a more energy efficient, less carbon intensive path than
have the developed countries during this century. This case was developed by assuming that by
the year 2100 all the developing countries would have a carbon dioxide per capita emission rate
(*) 1 gigaton = 109 metric tons
1-2
-------
-TABLE 1, -
ASSUMED ANNUAL GROWTH FACTORS INFLUENCING CO2
EMISSIONS (1990 - 2025)
(Derived from IPCC, 1992)
FACTOR
Growth of Economy Per Capita
Population Growth Rate
Growth Rate: Energy Use Per
Economic Output
Growth Rate: Carbon Emissions
Per Energy Use Unit
Annual CO2 Growth Rate
(Sum of above factors)
OECD
2.2%
0.3%
-1.1%
-0.7%
+0.7%
Asia
3.5%
1.5%
-0.8%
-0.3%
+3.9%
1-3
-------
12 .
10
0 l '
1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
1OECD minus USA 81 USA H USSR+REurope
j Latin America HAsia SaAfnca+M.East
Figure 1. Projected population by area (Source: IPCC, 1992).
1-4
-------
BASECASEIPCCI992
1990 2000
B OECD minus USA
USSR+EF
u;>j>K+EEurope
POOR AREAS EMULATE RICH BY 21 00
1990 2000 2010 2020 2030 2040 2050 2060 2070 20gO 2090 2100
OECD minus USA B USA usSR+E.Eurcpe
Bv21flJlTAmeiiCa "^ ' Aftca+M.East
By 2100 developing countries have
CO2 per capita^ 1/2 current U.S.value=2.9
Figure 2. C02 Emissions: gigatons C for two uncontrolled cases.
1-5
-------
one-half of the current U.S. value, or about 2.9 tons of carbon per capita. Stated another way,
this case assumes that over the next 110 years all developing countries will have a standard of
living and economic structure with an energy use pattern similar to contemporary Europe. In
my view, this is not a wild assumption. Assuming this, emissions are almost twice that
projected-by the IPCC. Perhaps this case represents an upper limit of what could happen in the
absence of serious changes in global energy use patterns.
PROJECTED GREENHOUSE WARMING
It is instructive to relate these projected emissions to anticipated global wanning. I have
utilized a model I developed which relates emissions of individual greenhouse gases to
equilibrium temperature increases using lifetimes, and radiative forcing functions contained in
IPCC, 1990. Realized (or actual) temperature is estimated using realized projection calculations
in IPCC, 1992, which have been correlated with the model's equilibrium calculations.
Figure 3 relates various emission grown scenarios to projected temperature rise. Note
that this is referred to as realized (or transient) temperature rise, which attempts to take into
account the thermal inertia associated with the Earth's features, especially oceans. (Equilibrium
temperatures, on the other hand, are sometimes reported which ignore the thermal inertia factor.
Typically, these temperatures are 1.5 to 2 times higher than the corresponding realized
temperature increases.) Also note that there are large uncertainties in these numbe* , probably
by at least a factor of 2 on both the high and low ends. An atmospheric sensitivity to doubling
carbon dioxide concentrations of 2.5" C was assumed in these calculations. The two
uncontrolled emission projection cases we previously discussed were analyzed: the IPCC base
case and what I call the last growth case which assumes that the developing countries will
approach the current industrialized world in terms of carbon dioxide emitted per capita. In the
IPCC base case, wanning is estimated at about 3° C by the end of the next century (consistent
with IPCC, 1992). If we were to cap emissions of all greenhouse gases during the year 2000,
it is expected that this wanning can be reduced about 30% to a little over 2° C. If the
international community would mitigate further and actually reduce emissions 1 % a year starting
in the year 2000, we can limit the rise to about 1.5° C.
This model calculates that it would require a 2%/yr emission reduction program to
stabilize wanning to about 1° C over 1980 levels. I believe that these numbers suggest the major
challenge which faces humankind, if we decide to seriously limit the projected greenhouse
warming. '- ?; - --.; $rh.._ ":'-" .*-"".-- - . ..
A LOOK AT THE IMPORTANT GREENHOUSE GASES
**"_..*"*'* ff *. t
Let us now take a look at the important greenhouse gases and their relative contributions.
Figure 4 shows the projected contributions by the major greenhouse gases assuming the IPCC
(1992) base case over the period 1980 - 2100. As you can see carbon dioxide is the most
important gas with methane and chlorofluorocarbons and their substitutes also important. Note
that the analysis assumed that all countries would reasonably implement international agreements
to phase out chlorofluorocarbons. The model, however, assumes that some of the substitutes,
like HFC-134a, which are substantial greenhouse gases in their own right, will be produced and
1-6
-------
3.5
3 -
£ 2.5
.*?
55
r-
5
a P
1980
2000
2020
2040
YEAR
2060
2080
2100
IPCC BASE CASE
CAPEMKSIONS:2000
FAST GROWTH CASE
DECLINE EMISSIONS:2%/YR
DECLINE EMISSIONS:1%/YR
Figure 3. Global warming four cases (control scarts: 2000).
1-7
-------
(60.4%)
(17,2%)
(3.7%)
(2.8%)
(u.6%) CFCs+Sub.
(4.3%)
CO2
CFCs SL Substitutes
METHANE
N2O
TROPO.OZONE
OTHERS
Figure 4. Equilibrium global warming by gas (1st year: 1980; end year:
2100),
1-8
-------
emitted in large quantities.
Note that the contribution of carbon dioxide is probably the best documented of the gases.
The methane calculations, for example, assume indirect, effects of. methane. Methane
decomposes to other greenhouse gases in the atmosphere: ozone, carbon monoxide, carbon
dioxide, and water. Note that there is a large degree of uncertainty over the magnitude, albeit
not the direction, of these indirect effects of methane. Nevertheless, I included them here since
I believe they ultimately will be important. Also note that other precursors of tropospheric
ozone have many uncertainties as well. There is the need for considerable atmospheric modeling
and measurements to relate emissions of volatile organic compounds, nitrogen oxides, and
carbon monoxide to high-level tropospheric ozone to better understand the significance of this
gas as a greenhouse wanning contributor in the upper troposphere.
Figure 5 illustrates, the relevant importance of gases over the 100-year time period of the
analysis, again using the IPCC 1992 base case. (Note that the potential cooling effect of
atmospheric aerosols has not been factored into the model calculations.) As you can note, the
short-lived gases, such as methane and ozone, are more important contributors early in this time
frame, with carbon dioxide becoming more dominant later in the time frame. This is associated
with the decay rates of the gases involved. Figure 6 illustrates an important mitigative advantage
in dealing with short-lived gases, such as methane, in that an aggressive control program can
stabilize atmospheric concentrations and mitigate wanning relatively quickly. This figure shows
that, if all gases were controlled at 1 % a year from the period 1980 to 2100, essentially all the
methane projected wanning could be mitigated whereas only about 60% of the carbon dioxide
wanning could be mitigated since the half-life in the atmosphere of carbon dioxide is so long
that emission reductions don't lead to reduction in atmospheric concentrations until many decades
later. Figure 7 illustrates this phenomenon by plotting concentration ratios relative to 1980 for
two long-lived gases (carbon dioxide and nitrous oxide) and the short-lived methane, all of which
had their emissions reduced by 1 % a year starting in the year 2000. As you can see, because
of methane's shorter half life, it responds more quickly to mitigation, yielding lower driving
forces for greenhouse wanning. Table 2 summarizes what we've discussed relative to the
important greenhouse gases. Note that this table also briefly summarizes major uncertainties
regarding each gas' wanning potential, and identifies major human sources.
It is also instructive to estimate the impact of chlorofluorocarbons (and related
compounds). Although these compounds are potent greenhouse gases based on their radiative
properties, recent data suggest that they have been responsible for ozone depletion in the lower
as well as the upper stratosphere. Since ozone in this lower region is a potent greenhouse gas,
there is a net sjojim* associated with this ozone depletion which opposes the radiative wanning
impact It appears that chlorofluorocarbons have not been the significant greenhouse wanning
contributors previously believed. Figure 8 shows the results of model calculations which do not
take into account the cooling effect. What is most interesting is that it is possible that the net
effect of replacing chlorofluorocarbons with compounds such as HFC-134a, which is a potent
greenhouse gas in its own right, could be wanning! This could occur since such replacements
which are chlorine free (and therefore non-ozone depleting) will contribute to warming without
the opposing cooling associated with ozone depletion.
1-9
-------
5,
CJ
O .4 -
UJ ;
3-
CX 1
UJ
0
fIP|BHi!|L:.V
.-;:>.£»;:; - -
tiiilil
2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
IMPACT YEAR
CO2
CFCs&Substitutes
CH4(Direct&liidirect)
N2O
OZONE
Others
Figure 5. Equilibrium warming ., °C, by gas (IPCC, 1992, base case).
1-10
-------
U
as
C02
CH4 03:xCH4
BASE WARMING
^ N20 OTHERS
3 WARM.PREVENTED
Figure 6.
iod: 1980-2100; annual
1-11
-------
0.
a.
o
oo
S
0.
a.
1980
2000 2020
(CO2 PPMX338
2040 2060 2080
(CH4 PPM)/1.605 __ N20 PPM/.303
Figure 7, Concentration ratio for three gases (relative to 1980).
1-12
-------
fr*l
CN I
211
If
III
Si I;
-"i "
Sfi Si.Si
S: fa-
&*ȣ. Si III
iiffsiai is
*'§:»: 3! :£
301
Ji
ji i
t!
-:
s
i !.ii
«£!
!i ' yi I-M s:
Ji^ci' :.f|'« S.i
iilfiJiiJii
'_i( s if, ft|~ S si
.f-iiBiiLajgi
SIJS! «, »i 51 >.-i
||-jj!^i-a!'i! = i
^1 gi-' KIWI i ,
~" B"Nif n'J
JJJ gj , £
31 -Si i|
;5i 1! :!
-^' y;; a
.101
SI
Ul
fill
.
S
! "^1 4»i
yiSiSIi
il-si&iQ!
»! ^,'H, =!
Ji & §! B;
MI^;LJ:u5i
as.
tn
I
! S|a!
1 «ai
.'llWI
»
i
j!^
.
oiT:
s
IVI
Ii
ii
ill,
81 gj i
?!.*! i
o
z
Ztf
'
= -en
en
:=!
{li
1
-
£|
tn i J"
u .
<
u
wi !
iSj '
SI
asi I
lint !
i
«*
1
.
.S ;
S
Sfl
J : CiJll
5 «!!
H Kit
J|i
±»2i
\! .A'
«icj]£
i»Ji
-------
r- sp
o d
in -T rn
o o o
I
-? fc
o
CO
j
-".
1
j _;; - .'
I
!
_!.:-.:
|
i
- .
i
:
ti
' .
.
....
.
--
'
-4- -"
< f f s >*
,.
'
r"
r
.,, <
^:
]
p^
i
,',--^
'
-
u
u
-------
WHAT PENALTIES ARE ASSOCIATED WITH DELAYING MITIGATION?
One important question relative to mitigation is: what penalties are associated with
delaying implementation of a mitigation'program? Or stated differently, .how much mitigation
opportunity would be lost if a mitigation program were started later rather than sooner? Figure
9 attempts to answer this question by plotting realized (transient) projected warming at 2100 for
three cases as a function of the year the control of all greenhouse gases would be initiated.
Included is a business-as-usual case per the IPCC 1992 emission scenario. Three control cases
are included, two where emissions are reduced by 1 % and 2% a year, respectively; and the third
where emissions are capped at the year control is-initiated at that emission level for all
greenhouse gases. Looking at the two decline cases, you can see that a 10-year delay in
initiation of a mitigation program yields a significant diminishment in the ability to mitigate
projected global warming. The graphic suggests that a 20-year delay in a 1 %/yr mitigation
program would require a 2%/yr program to achieve the same degree of wanning mitigation that
would have been achieved by the more modest program 20 years earlier.
MITIGATION CHALLENGES FOR THE U.S.
In order to understand the factors influencing greenhouse gas emissions in the U.S., I
have utilized another projection model. This model calculates emissions of carbon dioxide,
methane, and chlorofluorocarbons as a function of input factors such as population growth,
industrial growth, fuel use patterns, energy utilization efficiency, introduction of renewable
energy technology, and mobile source miles per gallon. This model incorporates the electric
utility module described in an earlier paper (Princiotta, 1990).
Figure 10 shows projected equivalent emissions for the 1980-2020 time period for a
business-as-usual case (DOE, 1987). Figure 11 shows the expected increase of carbon dioxide
emissions over the same time period for the major energy/use sectors. This projection suggests
that significant emission increases will result primarily from both increased use of coal to
generate electricity, and growth in the mobile source sector due to a larger auto, truck, and
aircraft fleet. Figure 12 shows that growth in electricity use is a critical parameter in
determining carbon dioxide emissions from the important electric utility sector. Although
introduction of renewable technologies, such as those based on solar or biomass energy sources,
would help mitigate emissions later in this time frame (Princiotta, 1990), the use of so-called
clean coal technologies such as integrated gasification combined cycle (IGCC) will have little
impact. Figure 13 shows that, even with a major introduction of IGCC technology (up to
300,000 MW), only a modest amount of carbon dioxide is mitigated. These results assumed
efficiency for IGCC is 41% vs 37% for conventional coal-fired units; yielding only an 11%
savings in coal use. * .
Figure 14 illustrates current and projected emissions from the major U.S. sources of
methane. Note the importance of landfills, emissions from cows and sheep, and coal mine and
natural gas pipeline leakage.
Last year the Administration (DOE, 1991) proposed an energy strategy that would have
a significant impact on greenhouse gas emissions. This strategy promoted a major research,
1-15
-------
m
o
«N
CO
o CiJ
gH
-------
HV3A/TOD SNOl*Ain03
1-17
-------
60+31
'SN01
-------
1980
1985
,'990 199S » 35 m537
Figure 12. C02 Emissions from electricity
annual demand growth).
2020
production (vs. electricity
1-19
-------
o
u
to
i
s
&
1980 1985 1990 1995 2000 2005 2010 2015 2020
.». 50000 MW _^_ 100000 MW_^ 200000 MW_^ 300000 MW
Figure 13. CC^ Emissions, from electric power (assuming various clean
coal scenarios).
1-20
-------
RUMINANTS PIPELINES
FUELBURN CUOTHER
2020
Figure 14. U.S. Methane emissions (for major
sources)
1-21
-------
development, and demonstration program to develop new fuel production and utilizati
technologies, with emphasis on renewable and nuclear systems. It also promoted the enhana
use of natural gas and nuclear power, and the use of renewables as electric utility ai
transportation fuels. More efficient utilization of energy, especially electricity, was ali
promoted.
The net effect of these policies, if successfully implemented, is shown in Figure 15. Th
bottom graphic compares carbon dioxide emissions for the U.S. strategy case with the previous!
discussed business-as-usual scenario (upper graphic). Such a policy can approach emission
stabilized at least over this time interval. As can be seen, the major reduction has been achieve
in the coal-electric sector. This results from a lower growth in electricity demand due t<
enhanced end use efficiency, and due to increased use of gas, nuclear, and renewable powe:
generation displacing carbon-intensive coal-fired power plants.
CONCLUSIONS
Let me summarize what I believe all these graphics and analyses seem to tell those of us
who are interested in greenhouse gas mitigation technology:
(1) Agricultural, medical, and industrial technologies have allowed for unprecedented
population and economic growth; development and application of low-emitting
technology could deal with the potential of unacceptable greenhouse wanning.
(2) Technologies and practices could be developed that provide cost effective
mitigation, not just for the developed countries that are generating the bulk of the
emissions in the short term, but also for the developing countries that will likely
be the dominant emitters in the longer term.
(3) Research could help reduce the uncertainties associated with several key gases:
For methane, emission and activity factors need improvement, and the
indirect effects of methane decomposition in the atmosphere needs
clarification. Also the apparent deceleration in the growth of methane
atmospheric concentrations cannot be easily explained -by current
source/sink information.
For tropospheric ozone, the mechanisms for formation in the upper
troposphere from the important precursor gases are not fully understood.
For chlorofluorocarbons, the balance between radiative heating and
stratospheric cooling needs to be better understood. Also, results of an
evaluation of likely substitutes for global wanning impact would be of
interest.
(4) Carbon dioxide is the key greenhouse gas which is directly linked to fossil fuel
1-22
-------
BUSINESS AS USUAL
O
U
c/3
I
1980 1985 1990 1995 2000 2005 2010 2015 2020
31 COAL-ELECTRIC COAL-INDUST. OIL-TRANSPG* T.
B OIL-OTHER 13 GAS-RES./COMM . GAS-OTHER
CO2 PROJECTIONS - U.S.ENERGY POUC Y CASE
at
yi
4?
ul
3.1
19(5 I9M
COAL-ELECTRIC
BOIL-OTHER
19U 3000 2005
COAL-INDUST.
2010 28IS
! OIL-TRANSPORT.
IGAS'OTHER
anno
Figure 15. U.S. C02 Emissions by fuel/sector.
1-23
-------
combustion, especially coal combustion. Major efforts could provide an
alternative energy path emphasizing renewable technologies such as solar, and
biomass with a focus on electric power production.
(5) Methane provides a ripe opportunity for mitigation research. By controlling the
major human sources, such as coal mines, landfills, and natural gas pipelines,
methane atmospheric levels could be stabilized within a relatively short period,
with substantial near-term mitigation impacts.
REFERENCES
Intergovernmental Panel on Climate Change (IPCC), "Climate Change 1992 - The
Supplementary Report to the IPCC Scientific Assessment," 1992
IPCC "Scientific Assessment of Climate Change," June 1990
Princiotta, F.T., "Pollution Control for Utility Power Generation, 1990 to 2020," Proceedings
of the Conference: Energy and the Environment in the 21st Century, March 26-28, 1990, MTT
Press, Cambridge, MA
U. S. Department of Energy, "Energy Security, A Report to the President of the U.S.," April
1987
U. S. Department of Energy, "National Energy Strategy," February 1991
1-24
-------
Paper 1-G
GLOBAL EMISSIONS DATABASE (GloED) SOFTWARE
by:
Lee L Beck
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
ABSTRACT
The EPA Office of Research and Development has developed a powerful software
package called the Global Emissions Database (GloED). GloED is a user-friendly, menu-
driven tool for storage and retrieval of emissions factors and activity data on a
country-specific basis. Data can be selected from databases resident within GloED
and/or inputted by the user. The data are used to construct emissions scenarios for
the countries and sources selected. References are linked to the data to 'ensure clear
data pedigree. The scenario outputs can be displayed on thematic global maps or
other graphic outputs such as bar or pie charts. In addition, data files can be
exported as Lotus 1-2-3, dBase, or ASCII files, and graphics can be saved as a .PCX
file or exported to a printer. This paper describes GloED and how it workE It also
presents future plans for software enhancements and populating the databases.
1-66
-------
BACKGROUND AND INTRODUCTION
The Global Emissions Database (GloED) was designed initially as an internal
data management tool to handle the large number of greenhouse gas data generated
as the result of international greenhouse gas research activity. Not only was there a
large amount of data but the databases.were rapidly changing based on continuing
studies and new information. Initial attempts at handling the data electronically
involved using commercial software such as Lotus 1-2-3 and dBase. These initial
efforts were frustrating partially because of the limitations of the software and
partially because of the limitations of user expertise. Initial limitations included keeping
track of which data sets were used in constructing emissions inventories and
identifying the quality of the data in each data set. It soon became apparent that a
better system needed to be developed by professionals in software development
The GloED software subsequently developed is personal computer (PC)-based
and very user-friendly with little or no computer expertise needed. Using GloED, the
storage and retrieval of data is quick and easy. This is important so that updates
can be stored as better data become available. Consequently, current best
estimates are always available. Another advantage of GloED is that units specified by
the user are automatically generated by GioED. If the user calls for units that are not
recognized by GloED, then GloED allows for the input of an algorithm that will define
the new unit.
Another important attribute of GtoED is that references are linked to the data.
Consequently, there is the ability of the user to establish the origin of every individual
piece of data in the scenario constructed. An additional advantage of GloED Is its
ability to interface with other software packages such as Lotus 1-2-3 and dBase for
those scientists and engineers familiar with these commercial software packages.
After establishing the utility of GloED as an internal data handling tool, it was
presented to an emissions workshop sponsored by the Intergovernmental Panel for
Climate Change (IPCC) in December 199H. As a result of this exposure to^the IPCC
and the Organisation for Economic Co-operation and Development (OECD), interest
grew in GioED as a standardized tool which could be used by ail researchers to
develop quality assured, country-specific, emissions inventories. The United Nations
Conference on Environment and Development (UNCED) in Rio de Janeiro just 2 months
ago underscored the need for a system to establish baseline emissions and to track
emissions reduction progress by country. GloED has the potential for providing this
system for implementing of the goals proposed at this historic Earth Summit meeting.
GLOED DESCRIPTION
GloED is a software system designed as a tool for generating estimates of
global emissions. GloED generates emission inventories by combining information
about activities with pollutant-specific emission factors for those activities. Activities
1-67
-------
are defined in terms of processes that occur in a specific pollutant source category in
a specific country or at a specific latitude and longitude. Activities are grouped into
discrete data sets within the GloED system. The user selects one or more data sets
and then has the option of narrowing the scope of the inventory by selecting a limited
number of countries, source categories, and pollutants. The final set of data selected
is called a scenario. GloED also can accept data provided by user-input.
Consequently, the emissions inventories can be updated as new data become available
to the user. GioED calculates an emissions inventory based on the scenario generated
by the user and can present summaries of the inventory graphically and in textual
form.
The contents of the emissions inventories can be reported in a variety of ways.
A text summary of the emissions inventory will print a tabular breakdown of the
results by country, source category, and/or pollutant. GloED can develop a pie chart
or bar chart showing the top pollutants or countries or source categories in a form
that allows easy comparison among them. Finally, GloED can project the results of an
emissions inventory onto a global map, using different colors to designate the type
and distribution of pollutants in the selected scenario. All of these output formats can
be viewed on the screen, saved to a file, or printed as a hard copy. The data can
also be exported to Lotus, dBase, or ASCII.
USING GLOED
Each level of the program has a menu that allows the user to select the
operations that the program should perform at that level. The user can select the
actions fn the menu either by clicking a mouse on the desired menu selection or by
typing the first letter of that selection. The GloED main menu always appears along
the top of the screen and is a set of pull-down menus, which means that the user can
'pull down' further options by selecting a menu item. When the user selects a menu
option-either with a mouse or with the cursor keys-GloED will lead the user to the
screens that apply to that menu option.
»
Most of the menu items selected will call up a screen of "scroll boxes' that
allow the user to define more specifically the way in which the selected menu function
Is performed. To move among the different scroll boxes, the mouse is used to dick in
the desired box, or the user can press the (TAB] key to move clockwise-or [SHIFT]
and [TAB] to move counterclockwise-through the scroll boxes and buttons on the
screen. The user can move within the scroll boxes either by using the cursor (or
arrow) keys on the keyboard or by clicking with the mouse on the "scroll bar" (the
vertical shaded strip with arrows at top and bottom) on the right-hand side of each
scroll box. The user can jump quickly to the very first entry in the scroll box by
pressing the [HOME] key on the keyboard and can jump to the last entry In the scroll
box by pressing the [END] key on the keyboard. The user can view the next or
previous boxful of information in the scroll box by pressing [PAGE UP] or (PAGE
1-68
-------
DOWN], respectively. When the item to be selected is highlighted, the user can select
it by pressing (ENTER] or by clicking on it with the mouse.
Any of three methods can be used to exit one of the screens in GIoED at any
time:
(1) Pressing the "escape" [ESC] key;
(2) Moving the mouse to the on-screen [CANCEL] button, and then "clicking1
the button on the mouse to select cancel; or
(3) Using the [TAB] key to tab clockwise (or [SHIFT] and fTAB] for
counterclockwise) through the on-screen scroll boxes and buttons in
sequence until the [CANCEL] button is selected, and then pressing the
[ENTER] key.
Any of these operations will take the user out of the screen in which the user is
working and return to the last active screen before arriving at this screen.
The GIoED Main Menu
Scenario Database Calculate Map Reports Help Quit
Figure 1. The GioED Main Menu
The GIoED main menu is shown in Figure 1. It is represented by a bar that will
remain at the top of the screen as long as GIoED is running, and offers choices of the .
type of function the user would tike GIoED to perform. The user can move between
menu options with the mouse or with the arrow keys. The user selects an item by
clicking on it or by typing the first letter of its name. The menu options and their
general functions are: .
Scenarlo-This menu allows the user to load a previously created
scenario, generate a new scenario, combine elements of two or more
scenarios, or delete a previously created scenario.
Database-This menu allows the user to edit data entered by the user
or to Import data sets to be combined with the system database files.
Calculated-Commands the software to create an emissions inventory
based on the currently defined scenario.
1-69
-------
Map-Allows the user to display the emissions inventory in the form of a
grldded or thematic map. (Only, thematic maps are currently
available.)
Reports-Allows the user to review the emissions inventory in a point-
by-point fashion and to search for a specific data point in the inventory.
It reports the results of the inventory calculation in text form (as
tables) or as graphics (bar charts or pie charts).
Help-Allows the user to receive on-screen assistance while operating
the software.
* Quit-Allows the user to leave the program.
The Scenario Menu Option
Scenario^! Database Calculate Map Reports Help Quit
Generate
Combine
Delete
Figure 2. Pull-Down Menu for Scenario Options
The user arrives at the step described in this section by choosing the scenario
option on the GloEO main menu. The menu that will pull down when the user selects
the scenario option is shown in Figure 2. It offers the option to load a scenario from
the user's disk, generate a new scenario, combine two or more existing scenarios, or
delete a scenario from a disk. By pressing [ESC], or the (eft-arrow key, the user
doses this pull-down menu and returns to the main menu. .
The Scenario Load Screen .
After selecting scenario in the main menu, the user can choose the load
option, which will cause a scroll box to come up on the screen. The scroll box will list
all of the scenarios that have been generated in previous sessions. To create a
different report from a previously generated scenario, the user can mouse-dick on the
name of that scenario, or use the cursor keys to select the scenario and then press
the [ENTER] key. Now the user can dick on the [OK] button and GloED will load the
selected scenario. It is important to remember that the user must now select the
calculate option on the main menu bar before using GloED to generate reports of
this scenario. .
The scenario load screen is shown in Rgure 3.
1-70
-------
Figure 3. The Scenario Load Screen
The Scenario Generate Screens
If the user selects the generate option in the Scenario menu, GloED will guide
through a series of screens in which the user names, describes, and defines the
parameters of the scenario to be generated. When this option in the scenario menu is
selected, a dialogue box entitled "Create a New Scenario" will appear. In this dialogue
box, the user enters the name of the scenario to be created. The user should enter a
valid DOS file name that is sufficiently descriptive to be remembered if it is to be used
again later. Then [TAB] to or mouse-select on the [OK] button and the Scenario
Description screen will come up to allow entering a description of the scenario to be
generated.
To replace an existing scenario with the one the user plans to generate, press
the [TAB] key or use the mouse to enter the scroll box that appears at the bottom
of the dialogue box. Then use the mouse or cursor keys to select the scenario to be
replaced. When the appropriate scenario is highlighted, press the [ENTER] key and its
name will appear on the scenario name line at the top of the box. Then mouse-dick
on the [OK] button or [TAB] to the [OK] button and press the [ENTER] key. Since
this scenario exists, a warning box will appear on the screen that says This
Scenario existsl Rebuild?' and gives on-screen buttons that say [OK], which
replaces the existing scenario with your new one, or [CANCEL], which cancels the
scenario generate routine. Select [OK] and the Scenario Description screen (Rgure 4)
will come up to allow entering a description of the new scenario to be developed.
, * ~ «». > **
. .» » ' *
The Scenario Description Screen
This screen will first ask the user to enter a long description of the scenario to
be generated.
1-71
-------
Scenario Description
Description:
This scenario generates an inventory of global NOz
emissions resulting from fossil-fuel production.
Brief Description: [SB!
Units:
OK
I Cancel
Figure 4. The Scenario Description Screen
The Unit Conversion jjtfl|ty~AfrAr entering the long and short descriptions
of the scenario, hit the [TAB] key again. The user will now be in the "Units' field of the
scenario description screen. In this field, enter the target units in which the inventory
is to be reported. The user does not need to know the units used by the individual
data sets because GloED contains a unit conversion utility that will automatically
convert the results of the scenario to the units defined as targets. This utility will be
especially useful in comparing the results of a series of scenarios. If entering the same
units for alt of the scenarios, no conversion will be required to make a comparison.
». '*.#**. - -.
The Scenario Generate Screen
Once the new scenario is described to the user's satisfaction, select [OK] and
GloED will lead to the Scenario Generate screen (Figure 5). This screen has four scroll
boxes, one each for data sets, countries, source categories, and pollutants. Move
among the scroll boxes and buttons on this screen, using the mouse to select a scroll
box item or pressing the {TAB] key to move clockwise, or [SHIFT] and [TAB] to move
counterclockwise. The other scroll boxes wilt show the source categories, countries,
and pollutants that are defined in the selected data set. When the screen first comes
up, nothing will be selected. After the user has selected the data set(s) to use, the
other boxes will fill with the countries, source categories, and pollutants in that data
set. . ;. -- - - - - ' . ...
1-72
-------
Scenario Generate
Data Sets: Source Categories:
ALL
CLEAR
GLOBAL VOC INVENTORY
BIOGENJC SOURCES
Countries:
ALL
CLEAR
ALBANIA
ALGERIA
ALL
CLEAR
AGRICULTURE (coal)
AGRICULTURE (gas)
OK
Pollutants:
ALL
CLEAR
NOx
COB
Cancel
Figure 5. The Scenario Generate Screen
When defining a new scenario, it is important to be aware of the order in which
the items are added to the scroll boxes. When a data set is selected, the other three
scroll boxes will automatically list all of the countries, source categories, and
pollutants included in that data set. GloED defines the items in the other scroif»boxes
on the basis of a hierarchy. The countries listed depend upon the data set(s) chosen.
The source categories fisted depend upon the data set(s) chosen and the countries
that have been selected within those data sets. The pollutants listed depend upon the
data set(s), countries, and source categories chosen. Thus, every time the selection
of elements in a scroll box is changed, the contents of the other boxes will change
according to this hierarchy.
The user can select a different data set or add data sets to the scenario.
Data sets can be selected or deselected by clicking them with the mouse or by using
the cursor buttons to highlight the desired data set and then pressing the [ENTER]
toy. Once the data set(s) are selected, the user can [TAB] or move the mouse
through the next boxes and select the specific source categories, countries, and
pollutants to be reflected in the emissions inventory. ALL automatically selects all of
the items in each list To select a large number of items, select ALL and then deselect
the few items not wanted in the scenario. CLEAR automatically deselects everything in
the list and can be used to clear all selections if an error has been made or to select
a few elements in a scroll box. At least one item in each scroll box must be selected
for a scenario to be generated. If the user has selected a combination of data sets,
countries, source categories, and pollutants that does not reflect an actual
combination in the system databases, an error message will request fewer
restrictions on the scenario. When all of the specific items for the desired scenario
have been defined, tab to the [OK] button on the screen, or dick on it with the mouse.
GloEO is now prepared to calculate the emissions inventory.
1-73
-------
The Calculate Menu Option
When the calculate option is selected in the GloEO main menu, the system will
generate an emissions inventory based on the current scenario. This is the inventory
that will be used in all of the following mapping and reporting menu options. Only one
emissions inventory will be generated for each scenario.
The Mao Menu Option
Scenario Database Calculate ^Reports Help 9utt
(meded
I Printer
(PCX File
Figure 6. Options in the Map Menu
Once a scenario is defined and an emissions inventory is calculated for it, the
user is ready to create a map that reflects the results of the inventory. When the
user selects the map option in the main menu, the user will be given the option to tell
GloEO to display the results either on a grldded map or on a thematic map (see
Figure 6). Gridded maps display the distribution of emissions in the form of ranges on
a latitude/longitude grid. (Gridded maps are not yet available.) Thematic maps
display ranges of emissions on a country basis. Figure 7 is an example of a thematic
map.
When the type of map to be created is selected, another menu will pull down
and request definitions of the output location for the map. If the user selects Screen,
the map will appear on the computer's video monitor. To send the map to the
printer, a dialogue box will appear on the screen and ask for selection of the available
printer output port to which the file is to be sent and for definition of the type of the
user's printer (either by selecting the name with the mouse or by pressing (ENTER] on
the appropriate output port and printer name). Finally, to save the map as a graphic
.PCX file, a dialogue box appears and requests entry of a valid DOS file name (ending
with .PCX) for the map file. At the end of each of these processes, select [OK] and
the Text Report Priorities screen will appear.
When the user has the inventory appropriately constructed and has selected
the map type, GloEO will prepare the map and then give the option of displaying the
map on the screen, sanding it to the printer/plotter, or saving it on a disk.
1-74
-------
1-75
-------
The Reports Menu Option
Data^e^atadate Map
Quit
We >
Bar >
Text >
Export >
figure 8. The Reports Menu Option
function prepares a comp.ee
Review option allows viewino
emissions inventoT
of
missions inventory, and the
'"format on individual data points in the
-suits of the em.sions
menu pulls down and e^o^nT^^0^ PM'dOWn menu- *"<*"«
ASCII format, or dBas^ fTl^^S^TT9 0° ^^ h L°tUS 1^3 format-
a GloED data file to dBase 9 S 8 Rep°rts &P°rt screen *» exporting
-
Save an Inventory to a dBase File
Filename: C:\GLOED\*.DBF
Directories Flies
» " .
OECD87.SCM
RICE^CM
>»
.~. B " - -^
C
~. n . - .
AcnvmrjDBF
ALLJCmPV.DBF
ALLDSET DBF
ALLPOLDBF
ALLSCAT.DBF
CATEGOFYJDBF
COUNnWlDBF
DATASETX>BF _ ; :; :r '
DSETBUP.DBF
EFACTOFLDBF
I OK |
I Cnncffl f
-
figure 9. The Reports Export Screen
1-76
-------
Graphical Reports
Scenario Database Calculate Map
Reports r| Help Quit
Review
Bar
Text
Export
Printer
PCX File
Figure 10. The Graphical Reports Option
In choosing to report the results of the scenario graphically, the user has the
option of presenting them in bar chart or pie chart format. Select Bar or Pie in the
Reports menu and then choose to have the graphical report appear on screen, be
sent to a printer, or saved on disk as a .PCX file. In selecting the printer option, a new
screen appears (Figure 11), asking for the printer type and output port for the file.
When the print location is designated, select [OK] and the title screen will appear. In
this screen, type the primary title for the graphical report, press [TAB] or mouse-click
on the second title field, and enter the secondary title (usually a reference for the
data, the units for the graph, or some sort of explanatory note) for the graph. Now,
select [OK] and GloED will send the report to the printer. Figures 12 and 13
examples of a bar and a pie chart, respectively.
Save a Chart to a PCX Ffle
Filename: £:\GLOED\*.PCX
JMrecteries
OECD87.SCM
RICE.SCM
m J\ *»
-B
FUes
Cancel
Rgure 11. The Reports Menu for Saving a Pie Graph to a .PCX Rle
1-77
-------
CO
"
u £
5 "
s -J
9)
1-78
-------
1-79
-------
Text Reports
Database Qdculate Map
Review
Export
Screen
Printer
File
Figure 14. Options In the Reports Menu
Then select the
Screen to
»«'on of a report. Is
Caller scroll boxes in Gloa
tars. Mouse^lickinB on the
the direction indicated by
in ft. n,^ o<
*£*. *«- to here as a text report,
enu' or *» an * to «P" The
°WSer SCreen> showinB a
TL "JnC"°ns "" much «"
? h"20"181 "l- ^^ 8ero11
^ mOV8S one " ««<*
1-80
-------
|NOx - fiance, Canada, U.S. - Fossil Fuels
COUNTRY
CANADA
CANADA
CANADA
CANADA
CANADA
CANADA
CANADA
CANADA
CANADA
CANADA
CANADA
FRANCE
FRANCE
FRANCE
FRANCE
FRANCE
^RANCE
'^KCE
Mi
ra
9
M
ra
B
H
I
ra
&
i
i
(
.-AtNCE
FRANCE
FRANCE
FRANCE
UNITED STATES OF AMERICA
UNITED STATES OF AMERICA
UNITED STATES OF AMERICA
UNITED STATES OF AMERICA
UNITED STATES OF AMERICA
UNITED STATES OF AMERICA
UNITED STATES OF AMERICA
UNITED STATES OF AMERICA
UNITED STATES OF AMERICA
SQURCP CATEGORY
AIR (oil)
AUTO OF ELEC (gas)
AUTO OF ELEC (oil)
PUB SERV ELEC (coal)
PUB SERV ELEC (gas)
PUB SERV ELEC (oil)
RESIDE.VTIAL (gas)
RESIDEiNTIAL (oil)
ROAD (oil)
TOTAL INDUSTRY (gas)
TOTAL INDUSTRY (oil)
AIR (Oil)
AUTO OF ELEC (gas)
AUTO OF ELEC (oil)
PUB SERV ELEC (coal)
PUB SERV ELEC (gas)
PUB SERV ELEC (oil)
RESIDENTIAL (gas)
RESIDENTIAL (oil)
ROAD (oil)
TOTAL INDUSTRY (gas)
TOTAL INDUSTRY (oil)
AIR (oil)
PUB SERV ELEC (coal)
PUB SERV ELEC (gas)
PUB SERV ELEC (oil)
RESIDENTIAL (gas)
RESIDENTIAL (oil)
ROAD (oil)
TOTAL INDUSTRY (gas)
TOTAL INDUSTRY (oil)
.... --
^MISSIONS
5.4O62S5E+04
1.747200E4O1
8.216900E+O2
5.6164 18E*O5
5.45748OE4O3
1.867441E4O7
1.660558E404
1.165500E+O4
5.729220E+O5
4.867460E4O5
8.1179Q9E+04
4.2064 98E+O4
7.956000E403
3.985520E+O3
7.106130E404
2.O44000E4O1
8.545440E+06
1.324I34E+O4
4.96944QE+O3
5.848920E+O5
2.864420E+05
1.024266E+05
9.154994E+05
1.05O033E+O7
6.9I2297E4O5
2.716832E+O8
1.908788E+O5
8.908956E4O4
6.855030E4O6
3.27O462E+O6
6.462365E+O5
3.249974E+O8
3.249974E+O8
Figure 15. The Text Report Browser Screen
*.>
To see a specific data point in the scenario, or to review the set of elements
chosen for the loaded scenario, use the Review option in the Reports menu. When
Review is selected in the Reports pull-down menu, the Review Emissions inventory
screen /r?gure 16) appears on the screen. When it first comes up, the screen shows
the .ssions of the first pollutant chosen for the first country and source category in
" -st database used in the currently loaded scenario.
1-81
-------
Review Emissions Inventory
Dataset
Country:
Source Category:
Pollutant:
Emissions:
Units:
Filter
Review
OK
Figure 16. The Inventory Reports Review Screen
From this point, the on-screen [UP] and [DOWN] buttons allow movement
through the scenario, one point at a time, in the indicated direction.
The [REVIEW] button opens a second window, in which appears the database
from which the data have come, the reference for the data, and the emission and
activity factors from which the emission estimate has been calculated.
If the user already knows the data point sought, or to narrow the search,
select the [FILTER] button, which activates the Search Inventory screen. The Search
Inventory Screen looks and works exactly like the Scenario Generate Screen, but the
scroll boxes on this screen contain only the database(s), countries, source categories,
and pollutants chosen for the currently loaded scenario. As a result, this screen
serves as a reminder of the complete contents of the loaded scenario and as a tool
for performing a very specific search of the emission inventory. To search for an item
in the loaded scenario, use the [TAB] key or the mouse to move through the scroll
boxes on the screen and select, with the [ENTER] key or the mouse, the parameters
of the data point When the user selects the [OK] button, GloEO returns to the Review
Emissions Inventory Screen, now containing the information about the data point of
interest : ..-,.-..' -^ i^r^- --^ .--^: ;-'*- " -' '
RELATED ACTIVITIES AND FUTURE PLANS
So far, EPA efforts have been directed toward development of the GloEO
software and development of emission factors and activity data for anthropogenic
sources of methane and nitrous oxide. With the completion of GloED software,
population of the software will begin with available information on greenhouse gas
emissions data on a country and source specific basis. This data development
1-82
-------
activity can be thought of as filling a 3*0 matrix. This 3-D matrix can be envisioned as
a matrix cube with countries along the vertical axis, greenhouse gases along the
horizontal axis, and sources or sectors filling the third dimension. EPA will begin
loading the software with information on methane emissions and will complete the
inventory data entry with available information from other sources. After published
data are loaded and quality checked, additional data estimates will enable a global
inventory of all greenhouse gases. It is recognized that some of the estimates will be
based on very weak information. However, data quality will be identified throughout
the matrix. This will be a beginning point for identifying where data quality needs to
be fortified.
Though the QloED emissions database will continue to be the primary emphasis
of database software development throughout 1993, we hope to begin development
of a companion database that will contain information on greenhouse gas mitigation
technologies. This database will be called GloTech. It will be an electronic file cabinet
that will house greenhouse gas mitigation technology and report parameters such as
emissions reduction capability, cost, and date of availability of the technology. When
populated with data, GloTech will allow scenario development and file interaction
similar to GloED. This will enable the user to perform cost effectiveness calculations
for an array of technologies that will be constructed in a scenario. Once the scenario
is constructed the user could then determine total cost, total emissions reduction
performance, and other parameters such as secondary impacts (water, solid waste,
etc.), estimated dates of availability of the technology, and limits to market
penetration. Like GloED, GloTech will allow construction of the scenarios based on
information resident within the software. It will also allow the user to input new data
or to modify data within the databases. GloTech will also have each piece of
information linked to its reference to ensure a dear data pedigree.
After GloED and GloTech are operational and fully tested, they could be linked
so that the user can perform "what if scenarios. These scenarios would provide
estimates of the effect of implementation of certain technologies on country specific
emissions, and what the cost of those technologies would be.
SUMMARY
. GioED is a powerful emissions database handling software package that is
nearing completion. It has been presented to OECD for consideration as a standard
tool for global greenhouse gas emission databases. The GloED software will continue
to be refined and updated, and enhancements are planned to enable GioED to accept
gridded information and to interface with geographical information systems. Parallel
to the development of the software, data population activities will produce a global
emissions database which may help to establish baseline emissions for an international
greenhouse gas emissions treaty such as was proposed at the U.N. Earth Summit in
Rio de Janeiro last June. GloED could also be the mechanism to track progress under
such a treaty by allowing annual updates of emissions information on a country and
sector specific basis.
1-83
-------
The development of GloTech will allow estimation of the costs associated with
greenhouse gas emissions reduction. Finally, the GloED and GloTech software working
in concert will allow the user to construct reduction scenarios, estimate impacts, and
view the results of the implemented technologies.
Reference
1. Intergovernmental Panel on Climate Change. "Workshop on National Inventories
of Greenhouse Gas Emissions and Sinks. 5-6 December, 1991. Geneva,
Switzerland. Proceedings." World Meteorological Organization/United Nations
Environment Programme. Edited by P. Schwengels.
1-84
-------
-------
SESSION III: C02, EMISSIONS, CONTROL, DISPOSAL AND UTILIZATION
Ken Friedman
Carbon Dioxide
Sequestration
by
Robert P. Hangebrauck, Robert H. .
Borgwardt, and Christopher D. Geron
Air and Energy Engineering Research Laboratory __
ABSTRACT
Mitigation of global climate change will require the stabilization
of atmospheric concentration of greenhouse gases, especially
carbon dioxide (CO,). CO, can be sequestered by flue-gas and
fuel CO, sequestration or by atmospheric CO, fixation/utiliza-
tion. Flue-gas sequestration involves separation/concentration,
transport, and either disposal or use. Disposal options are either
land or ocean based. Utilization is by either chemical or
biological utilization (recycling). Flue-gas-oriented techniques
in general have high economic and energy costs, but a few areas
show potential and warrant research and development (fl&D)
attention, especially those holding promise of combined CO/
sulfur dioxide (SO,j/nitrogen oxides (NOJ control and the inte-
grated gasification combined cycle approaches. CO, disposal
is neither a "sure thing* nor a permanent solution, with options
needing further environmental assessment. Near term, some
CO, recycling is possible, and R&D to examine, longer-term
prospects seems warranted. Atmospheric CO, fixation/utiliza-
tion involves either enhanced terrestrial or marine fixation with
utilization of the biomass in some cases. Atmospheric fixation
approaches which seem most attractive are those involving
enhanced biomass CO, sequestration combined with utilization
of the biomass for energy to displace fossil fuel. Of these the
most attractive for R&D appear to be advanced direct combus-
tion using biomass and use of biomass as a source of hydrogen
to leverage fossil fuel use for methanol production (Hydrocarb
process).
Introduction
International scientists working on the problem of
global climate have concluded that increasing concen-
340
270
jj.
1700
1000
VMT
1900
2000
Figure 1. Atmospheric CO, increase in the past 250 years
(IPCC, 1990). (Reproduced with permission.)
Paper 3-B
trations of greenhouse gases, especially C02 (Figure 1),
are leading to global warming (Figure 21 The pre-
dicted extent of such wanning is the subject of substan-
tial research with model-dependent estimates having
considerable uncertainty . The Intergovernmental
Panel on Climate Change;(IPCC) 1990 estimate was .-
0.3 ° C/decade with a range of 0.2 to 0.5 ° C/decade. The
1992 IPCC Supplement implies a wanning rate at the
lower end of this range. Wanning has also been studied
w 0.4
1170
1950
1*70
1990
1890 1910 1830
" YEAfl
Figure 2. Global-mean combined land-air and sea-
surface temperatures. 1861-1989. relative to the
average for 1951-80 (IPCC, 1990). (Reproduced
with permission.)
based on past temperature records, and has been
determined to be on the order of 0.3 to 0.6 ° C/decade
since the turn of the century. This trend is illustrated
in Figure 2. The abundance of fossil fuels and their
relatively low cost will probably ensure their use as a
principal energy source for the foreseeable future.
Once decisions are put in place to utilize fossil fuels in
the energy infrastructure, long-term commitments will
have been made to release CO, emissions to the atmo-
sphere. In the absence of constraints, the CO2 emission
rate can be expected to continue the rapid growth
indicated in Figure 3. This suggests the need for
§ :
f :
1
1 :
E
0-
. .
~-
s~
A f
t *
f
V
i
/
/
t
j
1860 1880 1900 1920 1940 1960 1980
Year
Figure 3. Annual global CO, emissions for fossil fuel
combustion, cement production, and gas flaring, 1860 to
1988(OHNL. 1990).
EPA August, 1992
3-1
Pagel
-------
prudent and timely R&D to provide a technological
base for preventing the increasing global emissions
consistent with economic growth.
Evaluating Potential and R&D Needs of
Sequestration Options
The general options for CO, sequestration, summa-
rized in Table 1, include both flue-gas C02 sequestra-
tion and atmospheric C02 fixation in the biosphere.
Flue-gas CO2 sequestration seeks to remove and dis-
pose of carbon as a part of fossil fuel use and requires
several steps to be undertaken: C02 must first be
separated and/or concentrated to obtain an economi-
cally handleable volume, which is then transported,
either to a point of disposal or to a point of utilization.
Disposal may be on the land or in the ocean. Utilization
can involve direct use of CO2 as a product (e.g., en-
hanced oil recovery) or it can be reacted to a product
such as a hydrocarbon or alcohol. Atmospheric C02
fixation/utilization on the other hand is aimed at en-
hancing or accelerating natural CO2 fixation processes
via either terrestrial or marine routes. These include
forest management and ocean fertilization, followed by
utilization for energy or other products. :
Table 1. General options for carbon dioxide seques-
tration.
Flue-gas & fuel CO2 sequestration
Separation/concentration
-Transport
' Disposal
- Land or sea
- Chemical/biological utilization
(recycling)
Atmospheric CO, fixation/utilization
-Terrestrial
- Marine
A variety of technology assessment factors are signifi-
cant for evaluation of die options for CO, sequestration
(Table 2). The status of a technology can be measured
by its level of development relative to commercial
application. Mitigation potential (possible carbon se-
questration) for a particular technology can be charac-
terized in part by measures such as breadth of appli-
cability and applicability to developing countries. Ap-
plicability to developing countries is an estimate of the
potential for use where energy infrastructures are
evolving and economic and other factors differ subi
tially from those of the industrialized world. The
effectiveness assessment factor can be measured b
engineering estimate of the cost per metric ton of
emissions prevented. .
Table 2. Technology assessment factors for C02
sequestration options.
Status of the technology
- Level of development
Mitigation potential
- Breadth of applicability
- Applicability to developing countries
Cost-effectiveness
- $/metric ton of C02 emissions prevented
Environmental and energy considerations
- Potential for adverse effects
- Product versus disposal
- Natural resource use
- Earth surface area required (land use)
- Additional fossil energy required
Probability of success and R&D requirements
Environmental and energy considerations can
gauged by the potential for adverse environmen
effects, whether or not an approach results in a use;
product or waste requiring disposal, natural resoui
use, and land use. Approaches requiring large surfc
areas of the Earth for implementation need examit
tion for social and ecological impact The additior
fossil energy requirement for implementation needs
be examined as a rough measure of overall life cyi
efficiency. Finally, the probability of success and Rfi
requirements need to be assessed.
Flue-Gas and Fuel CO2 Sequestration
A number of approaches are possible for separatio
concentration of CO, from fossil-fuel combusti<
sources. Figure 4 illustrates some of the possibility
including various flue-gas scrubbing approaches, ox
gen (0,VCO, combustion with CO2 recycle, and co
gasification combined cycles with carbon monoxu
(CO) shifted to hydrogen (H2) and CO,.
The categories of flue gas scrubbing (CO, separatio:
recovery) include absorption processes, adsorption pr
cesses, cryogenic separation, and membrane separj
tion. Of the flue-gas scrubbing approaches, amir
scrubbing is already in commercial use on a relative!
small scale, but has been demonstrated on coal-fire
power plants as large as 800 metric tons per day (TPI
(Barchas, 1992). Current application is for enhance
oil recovery (EOR) and production of soda ash, urei
EPA August, 1992
3-2
Pag
-------
Amma
Membrane sepa
Adsorption
COAL-FlflED
POWER PLANT
Compression/liquefaction
COAL-FIRED
POWER PLANT
WITH 02C02
COUBUSTtON
DISPOSAL IN
DEPLETED
GAS/OIL
RESERVES
INTEGRATED GASIFICATION
COMBINED CYCLE (K1CCI
DISSOLUTION W
SEAWATERAT
700-1000
METER DEPTH
SEA DEPOSITION
AS C02 LAKES AT
> MOO-METER
DEPTH
Figure 4. Flue-gas C02 sequestration options with CO., disposal
in the ocean or depleted gas/oil reserves
and methanol with Kerr McGee and Fluor Daniel as
major suppliers of the technology at a cost of $40/ton of
CO, recovered (not including transport, disposal, or
power replacement). Japanese research (Kansai Elec-
tric Power) is focused on this approach which they
think can operate with only 10.8% of boiler energy
output for CO, recovery and liquefaction fSuda, 1992).
Potential exists for. overcoming some of the efficiency
losses through heat recovery (Steinberg, 1991), but in
general, investigators report discouraging cost and
energy prospects. Smelser (1991) estimates that a
grass-roots pulverized coal plant with 90% CO, re-
moval can do much better than a retrofit; however, the
heat rates are 12.7 versus 15.7 MJ/kWh (12000 versus
14900 Btu/kWh). The flue-gas scrubbing approaches,
of all categories, such as sorption, cryogenic separa-
tion, and membrane separation, seem to share the
potential problems of high cost and substantial de-
creases in power plant operating efficiency, but oppor-
tunities seem to exist for research in such areas as
cryogenics and membranes (Smith, 1991). For ex-
ample, cryogenics might offer the potential for corn-
ocean injection. It has potential for
NO, and S02 removal. He-release
of CO, and other environmental
problems associated with utiliza-
tion/disposal are not quantified.
Mitigation potential includes the
utility sector (new and retrofit).
Small-scale pilot and demonstra-
tion work is underway with Japan
planning a demonstration. In
Canada Saskatchewan Power -is
planning a 150 to 350 MW demon-
stration. Canada is also research-
ing land disposal of CO,: 1) in de-
pleted hydrocarbon reservoirs start-
ing at 50,000 metric TPD, and 2}
EOR utilization at 6000 to 8000
metric TPD (Sypher: Mueller Inter-
national, 1991). Canadian EOR C02
use will be expanded substantially
in 1992. The likelihood of success is
high, but applicability depends on the ability to mini-
mize air in-leakage and environmental acceptability of
CO2 disposal. An engineering analysis of a 500-MW
power plant was discussed by British Coal at the recent
First International Conference on Carbon Dioxide Re-
moval in Amsterdam (Cross, 1992). Another analysis
was presented by Air Products (UK) (Allam, 1992).
These studies indicate that 0,-combustion/C02 recycle
will be 20% cheaper than amine (MEA) systems, can be
retrofitted to existing power plants, and would result in
lower power efficiency loss (8.6%). Capital cost for a
new plant would also be lower than for a conventional
PC plant. Cost was estimated at $49 per metric ton of
CO, removed. Distillation of liquefied C02 to recover
SO, could produce credits that would reduce cost to
$16/metric ton. **
Considering that this process or flue-gas scrubbing
processes may have the potential for inherent SO, and
NO, control, it is of interest to compare the costs of CO,
control with that for S02 plus NO, control on a cost per
metric ton of carbon basis. This is done in Figure 5 for
bined SO,, NO,, and CO, control. Accepting lower CO, low and high ranges of cost per ton of SO,, NO,, and CO,
removal efficiencies may also reduce cost
C0,/02 combustion with CO, disposal is a medium-
term option and appears to need engineering evalua-
tion along with substantial pilot-scale development
work (Wolsky, 1991). Moritsuka (1991) concludes that
such systems will be more cost-effective than flue-gas
scrubbing approaches. Herzog (1991) concluded that
removed. Costs for combined flue gas desulfurization
(FGD) and flue gas treatment (FGT) are put on a per
metric ton of carbon basis and compared with the cost
of CO, control on a per ton of carbon basis. High- and
low-range costs are assumed for FGD, FGT, and CO,
control. However, if S02 and NO, control were not
required or already in place, then this comparison
would have little meaning. Even with the offsets in cost
CO/O2 combustion would require the least incremen- for possible inherent SO, plus NO, control, a substan-
tal energy. This approach requires CO, utilization via tial cost increase for CO, control can be seen. Overall
EOR, or disposal via depleted petroleum formations or plant energy efficiency is reduced for the flue gas
3-3
EPA August, 1992
Par".?
-------
300 T
T 250 4-
u
o 200
I '
* iso 4.
D FQD (SO!) * FGT
(NO*)
FGCO2 (with
inrwf «nt SO2/NO*
control)
Low High
Range Range
Figure 5. Comparison of costs for SO, and NO, with the
costs of CO, control on a per metric ton of carbon basis.
{FGD and FGO cost ranges are derived from Emrnel, 1990).
sequestration options compared to conventional boiler
without C02 control. A similar analysis for energy use
offsets could be done by comparison to plants using SOZ
plus NOX control, but is not done in this paper.
Integrated gasification combined cycle (IGCC) with
CO, disposal-is another option shown in Figure 4. One
conceptual design is based on modification of conven-
tional gasification combined cycle plants with added
steps to shift all CO to CO3 for separation. The result-
ing hydrogen-rich fuel when burned in a turbine suf-
fers efficiency loss. An alternative design would in-
crease efficiency by applying CO/O, combustion to the
turbine combustor. No design studies are available to
determine the potential of the latter approach. Mitiga-
tion potential would include new utility applications
only. Canada is planning to investigate the approach
at pilot scale in conjunction with the design of a planned
250 MW IGCC plant (SyphenMueller International,
1991). The feasibility study was completed in 1991,
and the plant was scheduled to be in operation by 1996.
The likelihood of success seems to be reasonable, but
the process needs pilot and demonstration work.
Several innovations of the IGCC concept were reported
at a recent Amsterdam Conference on CO, removal.
They were mainly centered on the development of a
high temperature membrane for separation of CO and
Hj from the gasifier. The separated gas streams are fed
to different turbines, one of which is fed with pure O,
and recycled COr The combustion gases are then fed
to boilers and steam turbines. The exhaust from the
CO, turbine is collected (less recycle) for disposal. This
unit is expected to operate at 35% efficiency and
COj reduction with the cost of electricity increas
30%. Capital cost is $1100/kW; C02 recovery a
$16/metric ton (Hendricks, 1992). Replacement c
turbine of an IGCC plant with a molten carbonate
cell can increase the generating efficiency to m
48% while removing 90% of the C0r The fuel c
able to convert CO to COZ using air without mixir
with the oxidation products. Consequently, ni
separation plant is needed. A demonstration I
plant is under construction in Germany (RWE Ene
and will start operation in 1995 using brown
Efficiency is expected to be 45% without CO, reco
and 38.6% with recovery of 86% of C02 emissi
Investment cost without C02 recovery would be
less than for a conventional plant; with recovery,
30% more. A systems study by British Coal conch
that the IGCC system is the best choice for the
especially if operated under pressure so that the
can be absorbed in seawater for disposal without Hi
faction. Cost of pipelines for liquid C02 disposal in c
oceans was judged to be prohibitive. A thermodym
analysis of the IGCC system indicates thatMEAsc
bing after combustion is preferable to air separa
prior to combustion. Steam for CO^ stripping cai
extracted from the turbine. They conclude that
system is technically and economically feasible a
increased power cost of $0.015-0.02 per kWhr.
Fuel COj (carbon) sequestration has been evalui
extensively by Steinberg (1991). These are metl
that basically remove carbon from a fuel like
resulting in a hydrogen-enriched fuel. The remc
carbon is then sent to storage/disposal. The probl
with, carbon disposal can be minimized by usir
variation of this approach, called the Hydrocarb
cess, where biomass is used as a source of Hr Th
discussed later with the atmospheric fixation
preaches.
CO, Disposal
Ocean disposal. Any of the flue-gas CO, sequestra
approaches discussed above, implemented on a n
sive scale, require consideration of disposal of CC
the ocean as well as on land. Although ocean disp
has been and is being studied (Hoffert, 1979; Her
1991; Wilson, 1992), it may only be semi-permaE
storage and has enough environmental issues to m
the option longer term. The ocean options conside
include disposal as liquid, solid, and gaseous C0r ]
practical considerations are economic means for i
ting massive amounts of C02 down to depths whei
will stay or sink. Solid CO2 or CO2hydrates (clathra
will sink on their own. Smith (1991) suggests that
3-4
EPA August, 1992
-------
<*
jzooo
§ 1000'.
u
u
is
.
j
00 20C
^
^
b
'f\f
i '-»
j Jl
c *.~~
>0 25
^
^N
00 30
00 J5
Lonfl-wm
nnnott
00 .
-------
feasibility in terms of credits for C02 may be
limited under conditions of low oil prices.
Also the capacity seems limited, probably
amounting to only a small fraction of a GtC/
yr. After a few years of injection,CO,j is likely
to show up at the well head. We do not know
the extent of re-release of C0: used for EOR or
the potential for prevention of re-release, but
it is conceivable that measures could be taken
to prevent leaks and recycle the CO2. Other
issues arise where COZ/SO2/NOI mixtures
would be injected. These include the trans-
port of the COj/SOj/NO^ mixture over long
distances, the disposition of the 80,/NO, con-
taminants injected in oil reservoirs, and the
potential for pluggage of the reservoir (Spar-
row, 1988). We do not know the potential for
catastrophic release. Disposal in aquifers is
another option currently being discussed, but
questions on the resulting underground chem-
istry and fate remain to be resolved.
CO., Utilization
Sat*llil*-baMd
tolw photovoltaic
pantla (luium)
MAN-MADE CARBON CYCLE
(open c\dei '
COAL-FIRED POWER
WITH 02/C02
COMBUSTION
Figure 8. Long-term integrated systems for carton recycle.
In the near term, CO, utilization is the most available
and environmentally unquestioned option. A good
example is the emerging practice in the chemical in-
dustry of co-siting COj-using processes with CO,-pro-
ducers » specifically, integration of methanol produc-
tion (a COj-using process) with ammonia production (a
concentrated CO2-production process). What is the
longer-range importance of carbon recycling? Carbon
is a key transport agent for H,; e.g., methane (CH4) and
other hydrocarbons provide a "natural," convenient,
and practical means for H2 energy transport and use.
On the other hand, H2 is difficult to store, transport,
and use directly. The world's fossil fuel reserves are the
most economically available, concentrated source of
carbon, but are rapidly being consumed. Carbon re-
sources can be preserved as an energy transport me-
dium for future generations by recycling existing con-
centrated sources of COr At the same time, CO2
emissions to the atmosphere can be reduced along with
the attendant global warming. What are some of the
currently wasted CO, resources which could be man-
aged for synthesis of future hydrocarbons? They in-
clude: 1) fossil fuel combustion (e.g.t coal-fired power
plants), 2) calcination of limestone, 3) oxygen-blown
blastfurnace gas, 4) natural gas acid gas stripping, and
5) ammonia production to name a few. Carbon (as CO2)
can also be recycled from carbonate rock or from the
atmosphere to produce synthetic fuel but at much
greater cost (Steinberg, 1977).
Figure 8 illustrates some longer-term integrated sys-
tems for carbon recycling. Utilizing CO, via a "n
made" carbon cycle needs overall assessment and <
feasibility studies. It is a chemical-based recyc
approach with no CO2 disposal required. This ma
it environmentally attractive. Hj (with 0, as a
product) can be produced by using land-based so
photovoltaic produced electricity to electrolyze wa
For projected huge energy demands by the year 21
satellite- or lunar- based photovoltaic generation z
be required (Hoffert, 1991). CO, is reacted with H
form methanol. Methanol can be further reacted
dehydration to produce gasoline if desired. Steinb
(1977) investigated a similar concept using nucl
power as the energy source. The system elimins
need for an air separator and CO2 disposal. Mitigat
potential includes the utility, industrial, and trans;
tation sectors (new and-retrofit). If shown to be f
sible, the high mitigation potential and eliminatioi
the COS disposal problem argue for R&D to reduce
production cost Recent breakthroughs in product
of solar photovoltaic cells should help this somewh
but innovative means for producing H, is a fen
research area. Credits for SO2/N01 control could h
to reduce costs. Feasibility as a system needs to
determined. The Japanese are pursuing R&D
system components (Arakawa, 1992a&b). COj
combustion is in the early stages of development(smi
scale pilot, small demonstration) as discussed pre
ously. The likelihood of success depends on key comi
nents, including low-cost photovoltaic electricity gt
eration, H, production, and COj/O2 combustion. If tl
long-range concept proves to be feasible it could ha
several advantages: 1) fossil fuel can continue to
3-6
EPA August, 1992
Pa
-------
used, 2) transportation and gas tur-
bine fuels are made available, 3)
fossil carbon reserves are extended
for future use, 4) the energy storage
problem and global transportabil-
ity problem associated with solar
energy are solved, .5) the approach
appears highly applicable to devel-
oping countries, 6} solves the prob-
lem of difficult storage, transport,
and use of H2, and 7) provides a
transition to a direct-use H2 economy
by building up the necessary H2
production infrastructure. Ulti-
mately, the "C02-synfuelw being pro-
duced can be recycled itself, thus
achieving a closed cycle.
Flue-Gas Sequestration (Biologi-
cal Processes)
COAL-FIRED
POWER
PLANT
Photobioreactor
(Light-Pipe (optic fiber) Reactor
ng Super Algae)
Figure 9. Biologically based flue-gas CO, sequestration processes
(with microalgae due-gas CO2 removal).
Flue-gas sequestration via microalgae flue-gas COZ
capture is illustrated in Figure 9. The concept is
attractive, but difficult to implement because of the
large surface area required for exposing microalgae to
CO2 in the flue gas, A 100-MW power plant would
require an algae farm surrounding the plant out to a
distance of 4.3 km (Brown, 1990). Direct use of the
biomass generated appears to be difficult, but it can be
processed for lipids or to other hydrocarbon fuels such
as methane. Mitigation potential exists for new and
retrofit utility applications but is probably limited to
only certain areas with enough land, proper terrain,
nutrient capital, and adequate water and evaporation
rates. No reliable estimates of costs are available.
SERI is doing bench scale research and the Japanese
are looking at photobioreactor concepts which would
allow more efficient and concentrated
biological growth.
options in perspective, however, consider, that if the
entire U.S. current annual wood production (net growth
of 493 million dry metric tons/year) were diverted to
wood energy, it optimistically could pzvvide only 7 x
10U J (7 quads). Figure 11 is based on wood production
on U.S. timberlandsas projected by Sampson (1991). It
is likely that only a fraction of these amounts would be
used for energy. Approximately 1.5 x 10" J (1.4 quads)
is used for energy currently (High, 1990); Only a
fraction of the land area of a country is suitable, by
climate or terrain, for energy plantations. Within that
fraction, only an area within a reasonable distance of
an energy plant can be considered useful because of the
high transportation costs of biomass (one trainload of
coal is equivalent to the energy content of eight train-
loads of biomass). In any even t it appears that the most
Atmospheric CO, Fixation
enhancing Terrestrial Fixation
Figure 10 illustrates some of the op-
tions for enhancing terrestrial fixa-
tion. Options include reforestation
and forest management (0.1 -1 Gt/
yr), reducing deforestation (1 Gt/yr),
and enhanced soil C sequestration
(0.1 -0.5 Gt/yr). While there do not
appear to be any "si! ver bullets" in the
options available, it makes sense to
continue efforts underway to maxi-
mize potential benefits. To put these
ATMOSPHERIC CO, (750 GtC)
Agroforestry &
Sustainable Agri.
HMMM Opart by
flodunrawol
(RatorMiaton/Aflorwtaiion)
Low pramiM. land conflicts
Rgure 10. Means for increasing atmospheric COX fixation via enhancement of
terrestrial system uptake. Units are metric gigatons of carbon (GtC).
EPA August, 1992
3-7
Paee 7
-------
1018J (Quads)
5 10
BtanuM CM)<
15
Current
production
Potential future
production
Methanol
Synthesis
280 "C
SOi1
-------
ATMOSPHERIC CO, (750 QIC)
lilt
IBS
si--.
|lif
BJjnJ I
»f O *" I
j|o«J
!"««!
Microalgae (Phytoplankton
Carbon Stored
Poat* atctn; a
CO2 tan IB
Macroalgae
Carbon Stored
Figure 14. Illustration of means for increasing atmospheric CO2 fixation via
enhancement of aquatic system uptake. Units are metric gigatons of carbon (GtC).
Qcean fertilization. Microalgae fertilization has been
proposed as a means for enhancing carbon uptake. The
theory behind Fe, manganese, and phosphorus limita-
tions is not well understood or accepted universally.
However, considerable scientific interest in testing the
Fe limitation hypothesis exists, and it will likely be
pursued. Sarmiento (1991) and Joos (1991) have done
3-D modeling simulations of the impact of Fe fertiliza-
tion in the oceans of the Southern Hemisphere. Little
is known about potentially large, adverse effects and
therefore environmental considerations will top the
list of research priorities. Research is likely to focus on
ocean environments where key micronutrients are
thought to be limiting to productiv-
ity. This approach to C02 seques-
tration requires consideration of the
following environmental concerns:
Fertilization of large areas of
the Southern Atlantic and
Southern Pacific Oceans
The potential for deep ocean
anoxia and feedback of CH4 and
nitrous oxide (N,0)
Potential for increased release
" of CO, from ocean if fertiliza-
tion stopped -
Potential for dramatic changes
in species composition
May favor phytoplankton
species more susceptible to
ultraviolet (UV)-B radiation
Potential for altered fertility in
other ocean regions
Macroalpaefarming. Atmospheric
CO, sequestration via macroalgae
fanning is illustrated in Figure
15. These ideas are difficult to
evaluate at this point, but Lee
(1991) estimated that to generate
1 x 10" J (1 quad) of natural gas
would require 1000 kelp farms,
each 34 km long and 0.5 km wide.
Capital costs would be over $75
billion. Spencer (1991) has also
estimated mitigation potential
and cost. The following environ-
. mental issues must be considered:
Little is known about environ-
mental effects
* Ocean warming feedbacks
May require ocean dumping of
processed sludges
Potential for increased CH4 formation, especially
for sinking options
Effective use will require COZ disposal which has
several environmental consequences
Difficulty in supplying and recycling needed
nutrients
Potential for storm disturbance and loss
Potential increase in haloform production (methyl
bromide) and increased ozone depletion
* Respiration by organisms forming coral may
release C02 to the atmosphere
Observations and Conclusions.
METHANE
CO2
SEPARATOR
C02
to Dissolution
or Deposit Ion
NUTRIENT
UPWELUNQ
DEVICE
Figure 15. A proposed macroalgae ocean atmospheric CO, fixation approach.
EPA August, 1992
3-9
Page 9
-------
posaKfor limited quantities) is for product use, inc
Many cost-effective approaches will be needed to deal ing chemical feedstocks and enhanced oil recov
~ ' Many options for co-siting chemical facilities e.
where C02 can be effectively used. While this will
solve the C02 problem by itself, worthwhile contri
tions to reduce CO, can be made. One example tha
already being used is co-siting of ammonia and meti
nol production facilities. Methanol production ope
with increasing greenhouse gas emissions. Those that
are likely to be implemented before C02 sequestration
are chlorofluorocarbon/halon prevention, prevention
and control of CH4 and other tropospheric ozone pre-
cursors, accelerated conservation, and accelerated
development/use of renewables. Available data on
TECHNOLOGY
STATUS OF
TECHNOLOOr
CAHBOM
IUTKUTION
POTENTIAL
EFFECTIVENESS
CONSIOEIUTIOMS
OPPOOTVNRTES
Fiut-dA* cot SEQUEOTIUTIOM
WAflTERM
LOW
WON
C02
pmdiMH* w» COt »c*l» MMlcr tun pemr
NEAR TERM
HIOK
MEDIUM TO LOW
w«tnocn«lorNO> concwn. SHMIMMM nriyMonf
ndSQBunM NO)tfSO2 Mmwil mtefe*.
MEDIUM TEW
HOH
MEDUM
COS
atay
LONUIEHM
many CO, sequestration/fixation technology options tions can use the extra CO2 from the ammonia pla
are weak, but globally, R&D is increasing in these and the whole facility can be optimized for energy u
areas. Table 3 summarizes the Rue-gas CO, sequestration and atmospheric CO, fixation systems.
current picture of the CO: seques- " * ^ 2 H *
tration/fixation systems discussed
in this paper. Tables 4 and 5 sum-
marize the overall systems possi-
bilities on a time frame basis. The
underlying construct for these
tables was borrowed from the
Greenhouse Gas R&D Programme
(IEA, 1992) but has been modified
and expanded to cover additional
technologies and atmospheric fixa-
tion options.
Flue Gas Sequestration
Currently, there is no EPA activity
underway on flue-gas CO, seques-
tration, and there is very little U.S.
R&D in this area in general Glo-
bally, a considerable amount of
R&D work is being initiated: The
Japanese are especially active with
jouitgovernment/industry projects,
and other work is underway in Eu-
rope. The major limitations now
are high cost and increased energy
usage. Perhaps some increase in
cost-effectiveness can be achieved
from the simultaneous removal of
SO,, NO,, and particulates along
with the C0r CO/Oj combustion
appears to have research meritas a
technology for concentrating CO,,
but much development lies ahead.
IGCC options look especially at-
tractive, but apply primarily to new
powerplantcapacity. Flue-gas CO,
sequestration requires utilization
or disposal of the CO2 in conduction
with any of the above technologies.
The option which offers near-term
market solutions for utilization/dis-
MEQKMTEfW
TBW
T3BT
-555
MMM
y«r
-------
Table 4. Overall summary flue-gas sequestration systems.
Time FT* CM
Currant
Near-term
Mcdium-tarm
Long-term
Lone-term
Technology
Induilnal proniMM
(.(..ammonia 4 methane!
facilities
Conventional (oMit-fueled
utility boilare
Inteiraied ratification
combined cycle
CO2O2 combmuon (new *
retrofit) with O2 from air
M para 11 on
CO2/O2 fouil fuel combuiuon
with 02 from eolar
photovoltaic electrolytia of
. water
COBOZ-iyntheue
methenor-fueled advanced
CODIO. cyclaa (cloud carbon
cycle)
Other:
- Other adv. eytlae
-ruatcelb
Conventional utility (bnil-feel
combuition > microalfee fuel
CO2
Removal
Co-attinf for
CO2 uae -
Ami net
other flue-tii
removal lech
Abtorpuon.
cryogenic*, 4
membrane*
Not required
Not required
Not required
MiereaJfae
flue-fa*
eequMlrelion
CO2 Dlepoelllon
Chemitai feeoilocfc
Enhanced oil
recovery uie *
diipoul in depleted
oil et gaa formation*
New chamieaj
product! and
diipoul in (alt
domevaquifari
Methanol
tranaportation and
peaking fuel from
CO2*H2frem
eolar phouvottaic
electrolyui of water
- eJeo pnvidei 02
for combmuon
Ocean diipoul
Recycled u fuel or
ueed brother
praducti
Table 5. Overall summary of atmospheric fixation systems.
Tim* FniM
Ncar-tent
MedituD'tem
Long-term
TewhnQiofr
Increeaad ternstnaj biomaaa
MM toreml ttgmL, fottttation.
.. Dttvcfc ooflBintttion
* Ethanol eonv. hydrolyaia
Inereauicd terrcstria] biomau
(ihort-roUtion intensive culture)
Advanced cycle*
Hyptrofeiffe
- Ethanol adv. hydmlyaie
(ififmfjgej^ frr^hf atinn (ncnan)
Macroaigae {arming with
nerabic djge^ion (ooean)
CO2
Capture
Aim.
fixation
Atm.
Rzation
Atm.
fixation
CO2
Diipontion
None.
None
None
Ocean
diapoaaJ
without the problems of C02 disposal; re-
search needs to be done to establish feasibil-
ity and reduce potential costs, especially for
the solar hydrogen production required.
Atmospheric Fixation
EPAand others are actively working on fores-
tation, agroforestry, soil sequestration, direct
combustion/utilization, and the Hydrocarb
process development. Sequestration via mi-
croalgae and macroalgae are long term possi-
bilities requiring years to decades of ecologi-
cal effects research. There is no scientific
consensus on the duration of sequestration or
the effectiveness of biological mechanisms in
carbon fixation over long terms. As we learn
more, more potential problems seem to arise.
R&D on direct utilization of biomass offers
near term benefits. Direct biomass utiliza-
tion is currently utilized in the energy sector
and is projected to be more efficient in the
future. Competition for biomass feedstocks is
a serious limitation. For instance the wood
products industry demands increasing
amounts of fiber and can pay more for the
resource than the energy sector can afford
due to the higher value of final products. EPA
is currently examining the potential for ad-
vanced cycles using biomass. Development of
cost-effective technologies provides incentives
to grow increasing amounts of biomass. The
Hydrocarb process is a medium-term technol-
ogy and appears to provide means of mitiga-
tion of CO2 and other emissions from trans-
portation sources. Current research options
for the terrestrial biosphere research appear
justified; near-term benefits and costs appear
to be reasonable, although the sequestration
potential is limited. The serious environmen-
tal concerns associated with ocean microalgae
fertilization and macroalgae fanning make
them longer-term options requiring intensive
environmental assessment,
Disposal of CO, ia depleted oil and gas reserves is the
nearest-term straight disposal approach. Ocean CO,
disposal approaches appear to be rather costly and
have a major research need on modeling to determine
the new, longer-term atmospheric concentration levels
resulting from disposal. All of the disposal options are
longer-term options because of the environmental as-
sessment required before implementation. The man-
made carbon cycle has the potential for high mitigation
References
Allam, R. J., and Spilsbury, C., A study of the extrac-
tion of C02 from the flue gas of a 500 MW pulverized
coal-fired power station boiler, presented at the First
International Conference on Carbon Dioxide Removal,
3-11
EPA August, 1992
Par? !
-------
Amsterdam, the Netherlands (1992).
Arakawa, H., DuBois, J. L., and Sayama, K., Selective
conversion of C02 to methanol by catalytic hydrogena-
tion over promoted copper catalyst, presented at the
First International Conference on Carbon Dioxide He-
moral, Amsterdam, the Netherlands (1992a).
Arakawa, H., Methods for recycling use of emitted
carbon dioxide: catalytic hydrogenation, Now & Fu-
ture, 7,10-12 (1992b).
Barchas, R., The Kerr-McGee/ABB Lummus Crest Inc.
technology for recovery of C02 from stack gas, pre-
sented at the First International Conference on Carbon
Dioxide Removal, Amsterdam, the Netherlands (1992}.
Brown, L. M., Biological trapping of carbon dioxide,
SERI Microaigal Technology, National Renewable
Energy Laboratory (1990).
Cross, P. J. L, and Goldthorpe, S. H., System studies on
CO2 abatement from power plants, presented at the
First International Conference on Carbon Dioxide Re-
moval, Amsterdam, the Netherlands (1992.)
Emmel, T., and Maibodi, M., Retrofit Costsfor SO2 and
NO, Control Options at200 Coal-Fired Plants, Volume
I, EPA-600/7-90-021a (NTIS PB91-133322), U. S. En-
vironmental Protection Agency, Research Triangle
Park, NC (1990).
Golomb, D., Herzog, H., Tester, J., White, D., and
Zemba, S., Feasibility, modeling and economics of se-
questering power plant CO, emissions in the deep
ocean, Massachusetts Institute of Technology, Energy
Laboratory, MIT-EL 8&- -03 (1989).
Hall, D. O., Woods, J., and House J., Biological systems
for uptake of carbon dioxide, Div. of Biosphere Sci-
ences, Kings College, London, presented at the First
International Conference on Carbon Dioxide Removal,
Amsterdam (1992). -....--> . _
Hendricks, C. A., andBlok, K, Carbon dioxide recovery
using dual gas turbine IGCC plant, presented at the
First International Conference on Carbon Dioxide Re-
moval, Amsterdam, the Netherlands (1992).
Herzog, H., Golomb, D., and Zemba, S., Feasibility,
modeling and economics of sequestering power plant
C02 emissions in the deep ocean, Environmental
Progress, 10,64-74 (1991).
High, C., and Slog, K, Current and projected wood
energy consumption in the United States, Energy i
Biomass Wastes. XIII, D. L. Mass, Editor, Institu:
Gas Technology, Chicago, IL, pp 229-260 (1990).
Hoffert, M. L, Potter, S. D., Kadiramangalam, M.
and Tubiello, F., Solar power satellites: energy sot
for the greenhouse century?, Space Power, 10,131-
(1991).
Hoffert, M. I., Wey Y. CM Callerari, A. J., and Broed
W. S., Atmospheric response to deep-sea injection]
fossil-fuel carbon dioxide, Climatic Change, 2, 53
(1979).
IEA, Greenhouse Issues, IEA Greenhouse Gas R4
Programme, No. 2 (February 1992).
Intergovernmental Panel on Climate Change (IPCI
Climate Change, The IPCC Scientific Assessmei
World Meterological Organization/United Natio
Environment Programme (1990).
IPCC, 1992 IPCC Supplement, Intergovernment
Panel on Climate Change, UNEP/WMO (1992).
Joos, F., Sanniento, J. L., and Siegenthaler, U., Esl
mates of the effect of southern ocean iron fertilizatu
on atmospheric CO, concentrations, Nature, 349,77
775(1991).
Lee, K. H., Johnston, S. A., Stand], W. D., and Byrd, 1
C., Biomass state-of-the-art assessment, prepared I
Research Triangle Institute for the Electric Powi
Research Institute, EPRIGS-7471s Vols. 1 & 2 (1991
Liro, C. R., Adams, E. E., and Herzog, H. J., Modellir
the release of C02 in the deep ocean, presented at tfc
First International Conference on Carbon Dioxide Ri
moval, Amsterdam, the Netherlands (1992.)
Miller, S. L., The nature and occurrence of clathrat
hydrates, In: Natural gases in marine sediments, EC
Isaac R. Kaplan, Plenum Press, N.Y., pp. 151*17
(1974). .
Moritsuka, H., Study of power generation systems fo
carbon dioxide collection, Central Research Institute o
Electric Power Industry, Yokosuka Research Labora
tory,EW90006(1991).
Nishikawa, N., Morishita, M., Uchiyama, M.
Yamaguchi, F., Ohtsubo, K., Kimuro, H., and Hiraoka
R, CO, clathrates formation and its properties in thi
simulated deep ocean, presented at the First Intema
tional Conference on Carbon Dioxide Removal
3-12
August, 1992
Page
-------
Amsterdam (1992).
ORNL,The Carbon Dioxide Information Center, Trends
'90, A compendium of data on global change (1990).
Sampson, R. N., Biomass opportunities in the United
States, to mitigate global warming, American Forestry
Association, IGT Biomass Conference (1991).
Sarmiento, J. L., and Orr, J. C., 3-D simulations of the
impact of southern ocean nutrient depletion on atmo-
spheric COj and ocean chemistry, Limnology and Ocean*
ography (in press) (1991).
Smeiser, S. C., Stock, R. M., and McCleary, G. J.,
Engineering and economic evaluation of C02 removal
from fossil-fuel-fired power plants, Volume 1: pulver-
ized-coai-fired power plants, prepared by Flour Daniel,
Inc., EPRIIE-7365 (1991).
Smith, I. M., and Thambimuth, K V., Greenhouse
gases, abatement and control, IEA Coal Research,
IEACR/39 (1991).
Sparrow, F. T., Wolsky, A. M.t Berry, G. F., Brooks, C.,
Cobb, T. B., Lynch, E. P., Jankowski, D. J., and
Walbridge, E. W., Carbon dioxide from flue gases for
enhanced oil recovery, Argonne National Laboratory,
DE89-014017(1988).
Spencer, D. F., A preliminary assesment of CO, mitiga-
tion options, The Energy Journal (1991). :
Steinberg, M., Lee, J., and Morris, S., An assessment of
CO2 greenhouse gas mitigation technologies,
Brookhaven National Laboratory, BNL-46045, (1991).
Steinberg, M., Synthetic carbonaceous fuels and feed-
stocks from oxides of carbon and nuclear power,
Brookhaven National Laboratory, BNL-22785 (1977).
Suda, T., Development!^flue gas carbon dioxide recov-
ery technology, presented at the First International
Conference on Carbon Dioxide Removal, Amsterdam,
the Netherlands (1992).
SyphenMueller International, Development of a re-
search, development and demonstration program for
the control of greenhouse gases from coal use, prepared
for Environment Canada (1991).
Vitousek, P. M., Can planted forests counteract in-
creasing atmospheric carbon dioxide?, J. Environ. QuaL,
20,348-354(1991).
Wilson, T. R. S., The deep ocean disposal of carbon
dioxide, presented at the First International Confer-
ence on Carbon Dioxide Removal, Amsterdam, the
Netherlands (1992).
Wolsky, A M., Daniels, E. J., and Jody, B. J., Recover-
ing C02 from large- and medium-sized combustors, J.
Air Waste Manage. Assoc., 41,449-454 (1991).
3-13
EPA August, 1992
-------
-------
Paper 3-E
FUZZY LOGIC CONTROL OF AC INDUCTION MOTORS
TO REDUCE ENERGY CONSUMPTION
by: R.J. Spiegel and P.J. Chappell
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
J.G.CIeland
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC 27709
B.K. Bose
Department of Electrical Engineering
University of Tennessee
Knoxville, TN 37996
^ :IK r tt ABSTRACT
Fuzzy logic control of electric motors is being investigated under sponsorship
of the U.S. Environmental Protection Agency (EPA) to reduce energy consumption
when motors are operated at less than rated speeds and loads. Electric motors use
60% of the electrical energy generated in the U.S. An improvement of 1% in operating
efficiency of all electric motors could result in savings of 17 x 10» kWh/yr in the U.S.
New techniques are required to extract maximum performance from modern motors.
This paper describes EPA's research program, as well as early stages of work, to
implement fuzzy logic to optimize the efficiency of alternating current (AC) induction
motors. - -
This paper has been reviewed in accordance with the EPA's peer and
administrative review policies and approved for presentation and publication.
3*35
-------
INTRODUCTION .
. Electric motors use over 60% of the electrical power generated in the U.S. [1].
There'is a population of approximately 1 billion motors in the country, using over 1700
billion kWh per year. Over 140 million new motors are sold each year. A review of the
U.S. motor population reveals:
90% of the motors are less than 1 hp' (fractional motors) in size, but
use only 10% of the electricity consumed by motors;
* 95% of the electricity used by motors is consumed by approximately 2%
of the motor population (motors greater than 5 hp); and
85% of the electricity used by motors is consumed by less than 1% of
the motor population (motors greater than 20 hp).
Based on these facts, it is clear that large energy savings from improvement in motor
efficiency could be achieved with a relatively small motor population. Each 1%*
improvement in motor efficiency could result in:
17 billion kWh per year of electrical energy saved;
over $1 billion in energy costs saved per year;
an equivalent of 6 -10 million tons" per year of uncombusted coal; and
approximately 15 to 20 million tons less carbon dioxide released into the
atmosphere.
AC induction motors have high reliability and low cost and therefore perform
over 80% of the motor tasks in the U.S. Their speed of operation is determined by
the frequency of the input power, and their efficiency Is low when operating at part
load. To control the speed of an AC induction motor and thereby match motor speed
to load requirements requires the use of a device called an adjustable-speed drive
(ASD). Significant energy efficiency gains are achieved when induction motors are
controlled by ASDs [2J. ASDs use semiconductors and switching circuits to vary the
voltage or current and frequency of a motor's power supply thereby controlling the
applied torque and speed to satisfy the process or load requirements [2]. ASDs are
*lhp*746We
"1 ton = 907 kg
3-36
-------
basically power electronic devices consisting of a rectifier and a.computer-controlled
inverter. The rectifier converts the standard 60 Hz AC to direct current (DC). The
inverter then converts the OC output of the rectifier to a variable-frequency, variable-
voltage/current AC.
While ASOs can minimize power losses, they do not optimize operations for
maximum efficiency. The goaf of this program is to utilize the inherent capabilities of
fuzzy logic set theory in an integrated intelligent energy optimizer in conjunction with
an ASD to improve the energy or power efficiency of electric motors, primarily AC
induction motors, while at the same time meeting the demands of the process
equipment and load operations which are. driven by the motors. Fuzzy logic has the
proven ability to represent complex, ill-defined systems that are difficult or impractical
to model and control by conventional methods (3J. In addition, fuzzy logic is a form
of artificial intelligence that can be implemented in an integrated electronic circuit
device or microchip. This ability is especially important in the case of the modification,
or retrofitting, of existing electric motors, since microchips can .be readily added
through an add-on circuit board to existing ASD drives and require little additional
electric power for their operation.
FUZZVLOGJC ENERGY OPTIMIZER
Rgure 1 is a block diagram of the overall control approach. A fuzzy logic
energy optimizer is used to control the ASD which in turn controls the motor. A
feedback signal, usually motor speed, from the motor is shown by the dashed line in
the figure to indicate that the control scheme may be open-loop (no feedback) or
closed-loop (feedback).
RJZZT
LOGIC
ADJUSTABLE
SPEED
DRIVE
MOTOR
LOAD
J MOTOR PERFORMANCE PARAMETERS
Figure 1. Fuzzy Logic Energy Optimizer for Improved Energy Efficiency
A motor drive (ASD) may be controlled according to a number of performance
functions, such as input power, speed, torque, airgap flux, stator current, power
factor, and overall calculated motor efficiency. Normally in a drive system, the
3-37
-------
machine is operated with the flux maintained at the rated value, or with the voltage to
frequency ratio (V/Hz) held essentially constant in relation to the. value at rated
conditions; This allows speed control with the best transient response. The constant
V/Hz approach is used wherever actual shaft speed is not measured; i.e., in open-loop
speed control. The open-loop control approach is most effective when applied to
industry applications that do not require tight or accurate process control such as
pumps or blowers. The computer simulated results presented in this paper are based
on the open-loop control approach.
A fuzzy logic energy optimizer is also being developed for the closed-loop
situation which permits precision speed control. Additionally, work is proceeding on
the development of a fuzzy logic energy optimizer for the most advanced control
scheme known as closed-loop speed control with indirect vector control. This scheme
performs efficiency optimization control without sacrificing transient response. This is
very important for high performance applications such as electric vehicles.
For all the control methods the optimization approach proceeds as follows.
The input power is measured and then the control variables (input voltage, inaut
current, or input frequency) are varied from the initial setting. The input power is
measured again and compared with the previous value. Based on the sign and
magnitude of the input power signal, as well as the value of the last change in the
control variable, a new value for the control variable is computed using the fuzzy logic
energy optimizer. Sequential decrementation/Incrementation is continued until the
minimal input power level is reached. This is the operating point for best efficiency for
the particular load torque and speed condition. If a speed increase is demanded or
the load torque increases, the flux can be established to full value to get the best
transient response. When the new steady-state condition is attained, the fuzzy logic
efficiency optimization search begins again to obtain the most energy efficient
operating point.
Figure 2 is a conceptual flow diagram illustrating details of the fuzzy logic
energy optimizer which is contained within the dotted lines. Detailed explanation for
each block of the fuzzy logic energy optimizer is beyond the scope of this paper. It
suffices to say that the basic underlying principle of operation relies on the fuzzy rule
base consisting of several linguistic IF-THEN rules. A suitable rule base for the open-
loop situation is illustrate'd below. Additional information on the fuzzy logic concepts
contained in Rgure 2 can be found elsewhere [3]. The database includes the
necessary information regarding the motor parameters or other pertinent data. The
"fuzzification" stage is where the process measurements are usually represented as
fuzzy singletons, such as big, medium, and small. The "defuzzification" stage is where
fuzzy outputs are typically converted to real numbers. The most common procedure
for this conversion is the center-of-area method, much like that used for calculating
3-38
-------
r
DA
.RULE BASE
COMPUTATION
UNIT
DEFUZZIFICATION
PUZZIFICATION
PROCESS/
MOTOR
_J
Figure 2. Block Diagram of the Fuzzy Logic Energy Optimizer
r-- .
SIMULATED PERFORMANCE
The preliminary open-loop fuzzy logic controller [4] was demonstrated by
computer simulation. The control variable was the input voltage; Results show
improvement in motor efficiency using fuzzy logic control while maintaining good
performance in other areas; e.g.. maintaining desired torque and speed at steady
levels. For example, Figure 3 compares the efficiency of a motor over a broad range
of loads and operating under both conventional constant V/Hz control and fuzzy
.logic control. The load torque relation to rotor speed simulates the behavior of pump
or fan loads, where load torque is proportional to the square of the rotor speed.
Efficiency improvement by the fuzzy logic energy optimizer was achieved for all
speed/torque combinations.
3-39
-------
The following rule base was used for the calculation. Three fuzzy sets (N
standing for negative, P for positive, and 2. for zero) were chosen to relate the fuzzy
variables, along with the simple set of rules:
1. IF APin IS N AND A V0id IS N, THEN AVnew = N.
2. IF AP»n IS N AND A V0)d IS P, THEN AVnew = P.
3. IF APin IS P AND AVold IS N, THEN AVMw - P.
4. IF APin IS P AND AV0,d IS P, THEN AVnew = N.
5: IF APin IS Z AND AV0(d IS ANY, THEN AVnew = Z.
100.00
95.00
20.00 30.00 40.00 50.00 40.00 70.00 90.00 90.00 100.00
lutto
Figure 3. Fuzzy Logic Control Compared with V/Hz Control
for a 100 HP Motor with Torque Proportional to Speed Squared
Rule 5 is needed for convergence on an optimum input power; i.e., the point where any
small change in voltage results in negligible change in input power. The quantities Vdd
and Vraw represent old and new values, respectively, for the control variable (input
voltage) as the optimization approach proceeds to minimize the input power. The
change in input power level is designated by APin. To allow adjustment of step size
(for faster convergence with no overshoot), additional linguistic variables (e.g.,
3-40
-------
positive medium, PM, and negative-medium, NM) were added. A set of. 13 rules was
found to. be adequate to relate the variables for the simple control problem.
CONCLUSDN
Computer simulated results for the open-loop controller show that the fuzzy
logic energy optimizer can significantly enhance the operational efficiency of AC
motors. Future computer simulation developments will include a closed-loop controller
with a dual-variable (voltage and frequency) fuzzy logic energy optimizer. Additional
effort is taking place to provide fuzzy logic efficiency optimization for induction
motors which use indirect vector control.
Initial results further indicate that fuzzy logic energy optimizers can, in a
collective sense, consistently improve motor operational efficiency over conventional
speed control techniques (ASDs) by increments of 1 to 4%. This is highly significant in
terms of potential U.S. energy savings and pollution abatement possibilities. Figure 4
illustrates potential improvements based on conservative estimates of overage
energy savings for the motor classes indicated on the figure and typical coal-fired
power plant neat rates and emissions. The addition of a fuzzy logic energy optimizer
microchip to a 100 hp motor and ASD should result in energy savings amounting to a
cost payback within 3 to 5 months.
t *.^-~
-------
8
Tons CO, fUducad
P«r Ywr
kWh Savad par Yaar
*« Savad par Yaar
TOM SO, Raduead
Mr Yaar
All
Motors
1-20
HP
20-20,000
HP
Figure 4. Projected Collective Savings from Fuzzy Logic
Motor Control for Improved Efficiency
Once these optimizers have been thoroughly developed using computer
simulation models, prototype hardware devices will be tested in the laboratory. A
block diagram of the motor testing facility is shown in Rgure 5. The motor output
power is measured using the dynamometer, and the 3-phase input power is measured
with high precision wattmeters. This configuration allows the motor/drive efficiency to
be determined. A personal computer (PC) microprocessor will monitor the data
acquisition systems and communicate with the ASD to alter the ASD voltage and
frequency output. Various degrees of load on the motor are achieved by varying the
strength of the field in the DC brake via the dynamometer.
3-42
-------
9
Microprocessor
(v Propimatbte Micmerdp)
Efflueoey OpUuim Com)
* Ditt Acqitadae
Cunrat
Adjustable
Speed Drive
Motof Lo«l Sip.l
AdhiitedAC
Volujt.
Caimt
Siiw
Line Power
Induction
Motor
MatorSbtftSpral
T
r jtaoVAO
Dynamometer
Controller
Opfeii
Shifl
Eacoder
DCftrwtr
Dynamometer/
DC Brake
5. Block Diagram ol the Motor Testing Facility
REFERENCES
1.
2.
A
3. Kevin, S., Designing with Fuzzy Logic. IEEE Spectrum, pp.
4.
3-43
-------
-------
Paper 4-B
)NS AND MITIGATION
by: David A. Kirchgessner
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
Stephen D. Piccot
Science Applications International Corporation
3101 Petty Road
Durham, NC 27707
ABSTRACT
Estimates of methane (CH4) emissions from coal mines range from 25 to 45
Tg/yr with a recent estimate as high as 65 Tg/yr. At 46 Tg/yr, the estimate
produced by this project, coal mines contribute about 10% of anthropogenic CH4
emissions and may contribute significantly to the global change phenomenon. While
emissions from underground mines are now believed to be adequately characterized,
virtually no data are available on emissions from surface mines, and data ire totally
lacking on emissions from abandoned/inactive mines and coal handling operations.
The methodology developed to calculate emissions from underground mines is briefly
described, as is the Fourier transform infrared spectroscopy technique being
employed for measuring emissions from surface mines. A nitrogen-flooding technique
for enhancing the recovery of CH4 from cbalbeds in advance of mining is described as
a possible measure for mitigating CH4 emissions from underground mines.
- This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
4-11
-------
INTRODUCTION
Methane (CH4) is a radiativety important trace gas which accounts for about
18 percent of anthropogenic greenhouse warming. Atmospheric concentrations of
CH4 are now increasing at the rate of 1 percent per year [1]. Although the global
CH4 cycle is not fully understood, significant sources of emissions include wetlands,
ruminants, rice paddies, biomass burning, coal mines, natural gas transmission
facilities, landfills, termites, and tundra [2]. Improved emissions estimates for these
sources will allow their relative contributions to the global CH4 cycle to be better
understood, and will provide a means for focusing future emissions mitigation
research.
Attempts made to estimate global emissions from coal mining operations have
generally relied solely on global coal production data and emission factors derived
from CH4 contents of coalbeds [3,4,5]. These estimates are based on the
assumption that emissions are equal to the amount of CH4 trapped in the coal
removed from the mine. Although this trapped CH4 is liberated when coal is fractured
and removed from the mine, there are other CH4 release mechanisms in the mining
process which this assumption fails to take into account. For example, CH4 may be
released from: (1) exposed coal surfaces throughout the mine workings (e.g., the
roofs, floors, and walls); (2) gas which is trapped in the strata adjacent to the mined
seams; and (3) underlying seams close to the seam being mined. Commonly cited
global mine emissions estimates range from 25 to 45 teragrams (Tg) of CH4/year,
which corresponds to roughly 10 percent of the total annual CH4 emissions from
anthropogenic sources [5]. A recent report contains emissions estimates as high as
33 to 64 Tg CH4/year [6].
'~'^
Underground, surface, and abandoned or inactive mines comprise the three
general sources of mine related CH4 emissions. Emissions from underground mines
can be liberated from three sources: (1) ventilation shafts; (2) gob wells; and (3)
crushing operations. Ventilation air, although generally containing 1 percent or less
CH4, contributes the majority of mine emissions because of the enormous volume of
air used to ventilate mines. Gob wells are drilled into the area immediately above the
seam being mined. They provide corduits for venting CH4 which accumulates in the
rubble-filled areas formed when the mine roof subsides following longwall mining. Their
purpose is to remove CH4 which would otherwise have to be removed by larger and
more costly shaft ventilation systems. Currently, no published data for the release of
CH4 from gob wells exist. However, preliminary data obtained from the coal mining
industry indicate that gob well CH4 emissions could account for a significant fraction
of the total emissions associated with some longwall mines [7]. Emissions data for
crushing operations are also extremely limited.
4-12
-------
In surface mines, the exposed coal face and surface, and in particular areas of
coal rubble created by the blasting operation, are expected to provide the major
sources of CH4. As in underground mines, however, emissions may also be
contributed by the overburden and by, underlying strata. Emissions from abandoned
mines may come from unsealed shafts and from vents installed to prevent the buildup
of CH4 in the mines.
The main purpose of this research has been to develop an improved
methodology for estimating global CH4 emissions from underground coal mining
operations and to produce a global emissions estimate using this methodology where
country-specific estimates are not available. The -underground mine methodology
integrates data on coal production, .coal properties, coalbed CH4 contents (i.e., the
volume of CH4 per ton of coal), and coal mine ventilation air emissions from U.S.
mines. The objective was to develop a procedure which can be used to estimate mine
emissions from generally available coal analyses and production data where coalbed
CH4 data or emission estimates are not available for a country. This procedure will
be briefly described.
Since emissions data are presently not available for surface mines, the*Air and
Energy Engineering Research Laboratory (AEERL) of the U.S. Environmental Protection
Agency (EPA) has embarked upon a measurements program to quantify CH4
emissions from selected surface mines in the United States for later inclusion in this
work. The methodology employed will be discussed. Similarly, virtually no data exist
on emissions from handling operations (i.e., crushing, grinding, transport, and
storage) although their magnitude will certainly depend, to a large extent, on the
desorption characteristics of individual coals. There are also no data available on
abandoned inactive mines; therefore, AEERL is initiating assessments in both of these
categories.
Since one purpose of producing these estimates of emissions is to identify
appropriate targets for control within the coal industry, it is also necessary to
evaluate means of mitigating the emissions. Currently the most logical target for
mitigation is underground mines because they are the largest sources of emissions In
the industry and they consist of one or more point sources. The largest source of
emissions from an underground mine is ventilation air but, because of the enormous
volumes of air produced, CH4 concentrations in the air are typically less than 1
percent. No technologies are currently available to make economic use of such dilute
streams. It is believed that the most effective means of addressing the problem is to
degasify coal seams prior to mining. To make this process more economical the
efficiency can be increased by enhancing the recovery of CH4 from coal. AEERL is
studying a nitrogen-flooding technique developed by the Amoco Production Company
for the coalbed CH4 industry to accomplish this purpose.
4-13
-------
EMISSIONS
Numerous studies have examined, the physical relationships, which control the
production and release of CH4 by coal. These studies have been conducted either to.
evaluate the potential of coalbed CH4 resources or to enhance the safety of
underground mines. Generally, the studies address one of two topics: (1) factors
controlling coalbed CH4 content; or (2) factors controlling the concentration of CH4in
the mine atmosphere and mine ventilation air.
Studies in the first group have identified pressure, coal rank, and moisture
content as important determinants of coalbed CH4 content. Kim related gas content
to coal temperature and pressure, and in turn to coal depth [8J. After including coal
analyses data to represent rank, Kim produced a diagram relating gas content to
coal depth and rank. Although the validity of the rank relationship has been
questioned, it generally appears to have been accepted by recent authors
[9,10,11,12]. Independently of Kim's work, Basic and Vukic established the
relationship of CH4 content with depth in brown coals and lignite [13].
Several studies have recognized the decrease in CH4 adsorption on coal as
moisture content increases in the lowest moisture regimes [14,15,16]. Moisture
content appears to reach a critical value above which further increases produce no
significant change in CH4 content. Coals studied by Joubert et a!, showed critical
values in the range from t to 3 percent [16].
Investigations which attempt to identify correlates of CH4 content in coal mine
ventilation air include those by Irani et al. [17] and by Kissel et al. [18]. Irani et al.
developed a linear relationship between CH4 emissions and coal production depth for
mines in five seams. Kissel et al. demonstrated a linear relationship between CH4
emissions and coalbed CH4 content for six mines. Although both studies suffer from a
paucity of mines and/or seams in their analyses, Kissel et al. made the important
observation that mine emissions greatly exceed the amount expected from an
analysis of coalbed CH4 content alone. Emissions are produced not only by the mined
coal, but also by the coal left behind and by surrounding strata. For the six mines
studied, emissions per ton of coal mined exceeded coalbed CH4 per ton by factors of
from six to nine. , . .
MINE EMISSIONS ESTIMATE
Historically coal mine CH4 emission estimates have relied on coal production and
a value for coalbed CH4 content. The implicit assumption was that emissions were the
same as the CH4 content of the coal removed from the mine. A recent estimate by
4-14
-------
Boyer et al. 16] took into account the fact that emissions are six to nine times that
expected based on coal CH4 content alone. This estimate is considerably more
defensible than earlier ones for this reason. Our study refined the estimation
procedure even further by using a series of regression equations to predict coal mine
CH4 emissions from those coal characteristics known to be related to CH4 content:
depth, moisture content, and indicators of rank such as heating value and fuel ratio
(fixed carbon/volatile matter). The first step in the process uses the above coal
characteristics and one of two regression equations, depending upon heating value, to
produce estimates of in-situ coalbed CH4 content. The coalbed CH4 values along with
coal production statistics are then used in a second regression equation to predict
CH4 emissions; The equations have R* values from 0.56 to 0.71 suggesting that the
independent variables used explain 56 to 71 percent of the variability in the estimated
values. Although the equations were developed using U.S. coal data, they are believed
to be universally applicable since they employ coal characteristics which are known to
control coalbed CH4 content. The calculations are generally performed at the basin
level of disaggregation since this is the type of coal data usually available. A detailed
description of the estimation methodology is reported by Kirchgessner et al. [19J. It
produces a .global estimate of CH4 emissions from underground mines of 36.0
Tg/year. This estimate is believed to be of sufficient quality to obviate the n^ed for
further work on this category of mines.
Very little data exist on which to base estimates of emissions from surface
mines. A single emission analysis has been conducted to date by the EPA at a large
Powder River Basin surface mine in Wyoming [20]. Using open-path Fourier transform
infrared (FTIR) spectroscopy, an emission rate of about 4,814 mVday was
determined. Using a single coalbed CH4 content for the same county and coal seam,
it was estimated that, at the mine's actual coal production rate of 11.8 million tonnes
per year, potential emissions from the mined coal alone should be 1,008 m^/day. This
would suggest, as noted by Kissell et al. [18] for underground mines, that actual mine
emissions exceed, by a factor of about five in this case, the emissions which would be
expected based on coal production and coafbed CH4 content alone.
Rightmire et al. [21], in their study of coalbed CH4 resources in the United
States, report 38 analyses of shallow coals (104 m deep or less) with CH4 contents
ranging from 0.03 to 3.6 rn3/tonne coal. One analysis of 9.6 ma/tonne for the
Arkoma Basin was not included because it is known to be anomalously high for
shallow coals. Coalbed CH4 analyses for shallow coals from other countries are
lacking, so this study is temporarily making the gross assumption that the range of
0.03 to 3.6 m3/tonne coal reflects the CH4 content range for shallow coals worldwide.
Multiplying the average value for this range (1 nWtonne) by 1987 world surface coal
production of about 1.8 x 109 tonnes/year [6], and expanding the results by a factor
of five as discussed above, produces an estimate of about 6.3 Tg/year. Adjusting
4-15
-------
this value upward by 10 percent to represent 1989 coal production yields an estimate
of 6.9 Tg/year. As additional surface mine emissions are sampled under the EPA test
program, the factor by which actual surface mine emissions exceed expected
emissions may change, in which case this portion of the emissions .estimate will require
modification.
No data were found on CH4 emissions from handling operations. Boyer et al.
[6] estimate that 25 percent of the CH* contained in the mined coal is released during
post-mining operations. There is no compelling reason not to follow this precedent for
now; therefore, coal handling emissions were estimated by assuming that 25 percent
of the in-situ CH4 content for all coal produced is released in post-mining operations.
If warranted after further EPA investigation, these assumptions will be adjusted.
Country-specific results of underground mine estimates are shown in Table 1.
TABLE 1. SUMMARY OF ESTIMATED GLOBAL METHANE EMISSIONS FROM COAL MINES FOR 1989
Country
China
Former Soviet Union
Poland
United States
United Kingdom
West Germany
Australia
India
South Africa
Country Total
Rest of World
Total (Underground)
Total (Surface)
Total (Handling)
Total
1989
Undtcyiound
Mint Coil
Produaan
Cubed Mtihant Vikin
3
(ffl Aonne)
Emissions
t Disaggregaoon « 3
(to tomMi Lflyt, Average Maximum Minimum (10 m /yr) (Tg/yr)
1,053 21
418
-T - ; . -181 :'"
356
71*
73
59
95
115
2,421
567
---,.. 2,988 '
'.,...' .- - . -.-
~, ^ - ^v»
Provinces
6 Basins
3 Basins
19 Basins
12 Basins
4 Basins
3 Basins
8 States
4 Basins
-
.*
"
4.0
5.6
7.8
3.9
6.0
-
4.6
2.0
0.9
-
. <;.i ~
13.9
9.2
7.8
11.4
18.4
7.1
4.7
1.4
2.7
2.2
7.7
0.2
0.3
-
2.1
0.3
0.6
12.942
11.045
5.013
4,871
1.756
1,529
1.529
935
963
40,583
9.487
50.070
9.629
3.770
63,469
9.3
7.9
3.6
3.5
1.3
1.1
1.1
0.7
0.7
29.2
6.8
36.0
6.9
2.7
45.6
'1990-1991 Production
4-16
-------
. SURFACE MINE EMISSION MEASUREMENT
A fundamental goal of the sampling methodology is to obtain an emission rate
for total CH4 emissions from a surface.mine. The heterogeneity and size of the
source called for a creative measurement approach. Since smoke releases show that
emissions-from surface mines diffuse out of the pit in the direction of the'prevailing
winds, a near-ground-level concentration measurement downwind from the mine is.
used to estimate a total CH4 emission rate for the mine. A CH4 measurement of the
cross-wind-integrated concentration of the plume at near-ground level is made using
an open-path Fourier transform infrared (FTIR) sensor. Using this near-ground-level"
concentration measurement and a measured background or natural ambient CH4
concentration, the total mine release is estimated using an appropriate plume
dispersion model. If site-specific plume dispersion characteristics can be determined,
they can be used in the model to more accurately represent the behavioral
characteristics of the plume at a given site. Using a tracer gas. these site-specific
plume characteristics can be estimated as described below.
A tracer gas release can be assumed to be a continuously emitting point
source. Based on this assumption and on the results of the smoke release studies
conducted at strip mines, standard Gaussian dispersion equations can be applied.
When the standard Gaussian equation is integrated across the y direction (y is
assumed to tie in the direction normal to the wind direction) from - «, to + «, the
following relationship can be developed [22]:
where,
Q
u
az
H
20 exp f-1/2(H/az)2J
or
(1)
ground-level cross-wind-integrated concentration (g/m2)
emission rate (g/s)
average wind speed (m/s)
vertical dispersion coefficient (m)
effective emission height of plume centerline above ground level (m)
For a ground-level source such as a tracer release at a surface coal mine, H is
effectively equal to zero so the exponent of the expression is equal to 1. Thus,
Equation (1) can be simplified to:
4-t7
-------
2Q
(2)
az
Equation (2) can be used to obtain site-specific az values for a mine if the
values of.the "remaining unknowns can be determined. Specifically, az can be
determined for the plume given: (1) a measured tracer gas concentration (Ccvvi)
from an FTIR sensor; (2) a measured value of u from a meteorological station located
near the FTIR path; and (3) a known release rate Q from a tracer gas source, such
as a metered gas cylinder located at the mine. To use this technique to estimate
total mine emissions, a number of oz values must be determined based on tracer gas
releases conducted at several different distances upwind from the monitoring path.
These resulting values are used to construct a relationship of oz versus distance from
the path for the area source. All tracer gas releases used to determine this oz
relationship should be conducted as close in time as possible because atmospheric
stability may change, thus changing the oz relationship.
A similar and somewhat simpler technique can also be used to assess plume
dispersion characteristics using fewer tracer gas measurements. -Given measured
values for the tracer gas release rate Q, tracer release location, wind speed u.^and
wind direction, an appropriate area source plume dispersion model can be used to
predict GCWI for the tracer gas plume. The model is run to predict concentrations of
the tracer gas at various points along the FTIR monitoring path. These predicted
concentrations are integrated using the trapezoidal rule to calculate a path-
integrated concentration or Ccwi for the FTIR monitoring path. The model is run seven
times, once for each of the seven PasquiJI-Gifford (P-G) atmospheric stability classes .
[22]. These varying P-G assumptions, which incorporate the influence of oz, simulate
increasing atmospheric stability and its effect on the dispersion of the tracer gas
plume. Since several model results are produced, a range of Ccwi values are predicted
under varying degrees of atmospheric stability. The predicted Ccwi value which most
closely matches the Ccwi measured by the FTIR is used to define the P-G atmospheric
stability class which occurred during the tracer gas monitoring event. If simultaneous
CH4 measurements are also collected during this monitoring event, this stability
assumption is applied to the CH4 plume. The model is then run assuming a unity
emission rate for CH4 (i.e., a homogeneous release rate of 1 g/m2-sec) and the P-G
stability determined as described above. The model is run to predict concentrations
of CH4 at various points along the FTIR monitoring path. By again applying the
trapezoidal rule to these predicted point concentrations, a path-integrated
concentration or Ccwi for the assumed homogeneous release is predicted along the
FTIR monitoring path. Of course the FTIR is actually measuring a path-integrated
concentration due to a heterogeneous emission release pattern from the coal seam.
However, this measured value is comparable to the concentration determined from
"*4-18
-------
the model for an assumed homogeneous release because the FTJR measurements
integrated or 'averaged out1 the variable concentrations which exist in the plume from
the mine.
The-actual CH4 release rate for the mine is then calculated using the simple
relationship shown below where Q(predicted) is the unity emission rate for CH4.
'(actual)
Concentration(measured)
(predicted)
(3)
This technique is used to estimate CH4 release rates in this study. The Point Area and
Line (PAL) source model is used to predict point concentrations along the
measurements path as described above [23]. A non-reactive gas, sulfur hexafluoride
(SFe), is the tracer gas used. Use of a synthetic trace gas such as SFg is important
to the determination of plume dispersion characteristics because it is non-reactive,
does not naturally occur, and there is no background concentration to cause potential
interferences.
j»
Applying this methodology at a large, strip mine in the Powder River Basin of
Wyoming produced a CH4 emission rate of 4,814 nWday. An important observation
was, as in underground mines, that actual emissions exceed expected emissions by a
factor of about five in this instance. Details of the methodology have been discussed
previously by Piccot et al. [ 24 ] and Kirchgessner et al. [20]. A validation study of
the methodology designed to answer questions raised during the first sampling trip
has recently been completed and the data are being analyzed.
COAL MINE METHANE MITIGATION
AEERL is participating in a demonstration of the Amoco Production Company's
nitrogen-flooding process to enhance the recovery of CH4 from coal seams. Although
Amoco's interest in developing the technology is focused on CH4 as the saleable
resource, the methods involved will translate fully from the coalbed CH4 industry to
the coal mining industry. The goal of the project is to demonstrate that the 50
percent average CH4 recovery rate from coal seams using current practice can be
increased to 80 percent or more using nitrogen flooding. The final objective of the
Laboratory's involvement is to transfer the practice to the coal industry as needed.
The enhanced recovery, if achieved in a premine degasification program, will allow a
mine to reduce its costly ventilation air requirements, and to retrieve more CH4 for
utilization or sale for a given drilling cost. In this fashion a consistent program of
premine degasification may become not only less costly, but an actual economic
benefit.
4-19
-------
In conventional reservoirs CH4 is contained as a free gas. In contrast, CH4 in
coal seams is stored as a gas adsorbed on the internal micropores of the coal
matrix. The conventional practice of recovering coalbed .CH4 is to -.reduce total
reservoir pressure by pumping water out of the coal. Some CH4 desorbs from the
coal surface, migrates through the micropores to the cleat or fracture system, and
then travels to the recovery well along with the water. Although the system is simple
it is inefficient because at the lower economic limit of pumping, about 150 psi
(1000kPa), as much as 50 percent or more of the original CH4 may remain in the coal.
An additional drawback to reducing the total reservoir pressure is that the driving
force for gas expulsion is lost.
Arr alternative to reducing the total reservoir pressure is to reduce the partial
pressure of CH4 by introducing an inert, low-adsorbing gas at a constant pressure
[25]. Partial pressure of a component is equal to the total system pressure multiplied
by the component's mole concentration, in the gas phase. Therefore the injection of
nitrogen reduces the relative concentration of CH4 and hence its partial pressure while,
in some cases, increasing totai reservoir pressure. Laboratory studies have shown
CH4 recoveries of over 80 percent as well as significantly enhanced rates of recovery.
Modeling studies suggest that the cost of nitrogen is more than offset by the
improvement in production. .
The demonstration tract is located in the northern portion of the San Juan
Basin, approximately 9 miles (14.5 km) southeast of Durango, Colorado. The source
of the CH4 is in the coals of the Upper Cretaceous Fruitland formation at a depth of
about 2800 feet (853.5 m). The tract is 80 acres (32.5 hectares) in size with four
injection wells located at the four corners of the tract, and a recovery well located
approximately in the center of the tract. The objective is to demonstrate an
economic CH4 recovery rate of 80 percent or better using nitrogen flooding, with
minimaf or no effects on neighboring wells. The project began in the summer of 1992.
SUMMARY - '..
AEERL is actively involved in a program of estimating and measuring global CH4
emissions from coal mines, and of developing mitigation technology for underground
mines should control of this source be deemed prudent. The estimation of emissions
from underground mines is regarded as complete and has produced a value of 36.0
Tg/year. Emissions from surface mines were estimated to be 6.9 Tg/year using a
single measured value. A sampling campaign at selected surface mines will be
conducted using an open-path FTIR instrument and dispersion modeling. If necessary
the estimate for surface mines will be adjusted using these data. Emissions from coal
handling operations were estimated using a technique from the literature, but
4-20
-------
emissions from abandoned/inactive mines have not been addressed. Programs are
underway in both of these areas and the current global estimate of 45.6 Tg/year will
be modified as appropriate when they are complete.
A technology for enhancing the recovery of CH4 from coalbeds is being
demonstrated. It is expected that the increased efficiency will improve the economics
of premine degasification and provide a reasonable method of mitigating CH4 from
underground mines should control of these sources become advisable.
1.
2.
3.
4;
5.
6.
REFERENCES
Smith, B. and Tirpak, D. The Potential Effects of Global Climate Change on the
U.S.: Report to Congress. EPA/230-05-89-050. U.S. Environmental Protection
Agency, Office of Policy Planning and Evaluation, Washington, DC, 1989. 13 pp.
Wuebbles, DJ. and Edmonds, J. Primer on Greenhouse Gases. Lewis Publishers,
Inc., Chelsea, Ml, 1991.
Koyama, T. Gaseous Metabolism in Lake Sediment and Paddy Soils and the
Production of Atmospheric Methane and Hydrogen, journal of Geophysical
Research'. 68 (13): 3971, 1963.
Mariand, G. and Rotty, R.M. Carbon Dioxide Emissions from Fossil Fuels: A
Procedure for Estimation and Results for 1950-1982. Tellus. 36B (4): 232,
1984.
Cicerone, RJ. and Oremland, R.S. Biogeochemical Aspects of Atmospheric
Methane. Global Biogeochemieal Cycles. 2 (4): 299, 1988.
Boyer, C.M., Kelafant. J.R., Kuuskraa, VA, Manger, K.C., and Kruger, D.
M.e{hane Emissions from Cpal Mining: Iscuae *** Opportunities for Redb^umi
ED * .A ««' Agency, Office of Air and
.m,c t^mrssions frpm
EPA-400/9-90/008. U.S.
Radiation, Washington,
7.
8- Kim, A.G.
Persona, conimunjcat,,n
4-21
-------
9. Lambert, S.W., Trevits, M.A., and Steidl, P.P. vertical Borehole Design and
Completion Practices to Remove Methane Gas from Minable Coalbeds. DOE/CMTC/
80/2. U.S. Department of Energy, Washington, DC, 1980. 163 pp. .
10. Murray, D.D. Methane From Coalbeds - A Significant Undeveloped Source of
Natural Gas. Colorado School of Mines Research institute, Golden, CO, 1980.
37 pp.
11. Ameri, S., Al-Sandoon, P.T., and Byrer, C.W. Coalbed Methane Resource
Estimate of the Piceance Basin. DOE/METC/TPR/82-6. U.S. Department of
Energy, Morgantown, WV, 1981. 44 pp.
12. Schwarzer, R.R. and Byrer, C.W. Variation in the Quantity of Methane Adsorbed
bv Selected Coals as a Function of Coal Petrology and Coal Chemistry Final
Report. DE-AC21-80MC14219. U.S. Department of Energy, Morgantown, WV,
1983. pp. 1-6.
13. Basic, A. and Vukic, M. Dependence of Methane Contents in Brown Coal and
Lignite Seams on Depth of Occurrence and Natural Conditions. ia: Proceedings
of the 23rd International Conference of Safety in Mines Research institutes.
U.S. Department of the Interior, Bureau of Mines, Washington, DC, 1989. pp.
282-288.
14. Anderson, R.B. and Hofer, L.J.E. Activation Energy of Diffusion of Gases into
Porous Solids, fust 44 (4): 303, 1965.
15. Jolly, D.C., Morris, L.H., and Hinsely, F.B. An investigation into the Relationship
Between the Methane Sorption Capacity of Coal and Gas Pressure. The Mining
Engineer. 127 (94): 539, 1968.
16. Joubert, J.I., Grein, C.T., and Bienstock, D. Effect of Moisture on the Methane
Capacity of American Coals. Fuel. 53 (3): 186, 1974.
17.
18.
19.
Irani, M.C., Thimons, E.D., and Bobick, T.G. Methane Emission from U.S. Coal
Mines, a Survey 1C 8558. U.S. Department of the Interior, Bureau of Mines,
Pittsburgh, PA,: 1972. pp.- 7-15.
Kissel, F.N., McCuIloch, C.M., and Elder, C.H. The Direct Method of Determining
Methane Content of Coaibeds for Ventilation Design. Rl 7767. U.S. Department
of the Interior, Bureau of Mines, Pittsburgh, PA, 1973. pp. 1-9.
Kirchgessner, D.A., Piccot, S.D., and Winkler, J.D. Estimate of Global Methane
Emissions from Coal Mines. Chemosphere. (In Press), 1992.
4-22
-------
20. Kirchgessner, D.A., Piccot, S.D., and Chadha, A. Estimation of Methane
Emissions from a Surface Coal Mine Using Open-Path FTIR Spectroscopy and
Modeling Techniques. Chernosphere. (In Press), 1992.
21. ' Rightmire, C.T., Eddy, G.E., and Kirr, J.N. Coalbed Methane Resources of the
United States. AAPG Studies in Geology Series #17. American Association of
Petroleum Geologists, Tulsa, OK, 1984. 9 pp.
22. Turner, D.B. Workbook of Atmospheric Dispersion Estimates. EPA Report AP-
26 (NTIS PB191482). U.S. Environmental Protection Agency, Research Triangle
Park, NC, 1970.
Petersen, W.B. and Rumsey, E.D. User's Guide for PAL 2.0 - A Gaussian Plume
Algorithm for Point. Area, and Line Sources. EPA-600/8-87-009 (NTIS PB87-
168787). U.S. Environmental Protection Agency, Atmospheric Sciences
Research Laboratory, Research Triangle Park, NC, 1987.
Piccot, S.D., Chadha, A., Kirchgessner, D.A., Kagann, R.t Czerniawski, M.J., and
Minnich, T. Measurement of Methane Emissions in the Plume of a Large Surface
Coal Mine Using Open-Path FTIR Spectroscopy. in: Proceedings of the 1991 Air
and Waste Management Association Conference, Vancouver, B.C., 1991.
Pun, R. and Yee, 0. Enhanced Coalbed Methane Recovery. Presented at: 65th
Annual Technical Conference and Exhibition of the Society of Petroleum
Engineers, New Orleans, LA, September 23-26, 1990.
23.
24.
25.
4-23
-------
-------
EMISSIONS AND MITIGATION AT LANDFILLS
AND OTHER WASTE MANAGEMENT FACILITIES
Susan A. Thomeloe
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
..-.: To Be Presented at:
EPA Symposium on Greenhouse Gas
Emissions and Mitigation Research
August 18-20,1992
Omni Shozeham Hotel
Washington, D.C
-4-46
-------
EMISSIONS AND MITIGATION AT LANDFILLS
AND OTHER WASTE MANAGEMENT FACILITIES
by: Susan A. Thomeloe .
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
ABSTRACT
Landfills and other waste management sources of methane are amenable to cost-effective
control. Consequently these sources have been give a high priority for clarification of their
emission potential. The United States Environmental Protection Agency (U.S. EPA) is conducting
research to determine the emission potential and mitigation opportunities for cost-effective control
for the major sources of greenhouse gases. EPA's Air and Energy Engineering Research
Laboratory (AEERL) is responsible for developing more reliable global and country-specific
estimates for the major sources of greenhouse gases including waste management, coal mines,
natural gas production/ distribution, energy usage, cookstoves, and biomass combustion. AEERL
has gathered data which have resulted in the development of more reliable estimates for landfills.
Research has been initiated to characterize the methane potential of other waste management
facilities including wastewater treatment lagoons, septic sewage systems, and livestock waste.
AEERL is also documenting the current state of technology for utilization projects.
Currently there are 114 landfill gas to energy projects in the U.S. and about 200 worldwide.
Technology transfer/technical assistance programs have been initiated to help encourage the
utilization of waste methane and to help implement the upcoming dean Air Act (CAA) regulations
for municipal solid waste landfills. For example, AEERL is working with a consortium of local
government representatives to explore the application of EPA research on methane/energy recovery
from municipal solid waste landfills. AEERL also serves on the International Energy Agency
Expert Working Group on Landfill Gas and the Steering Committee for the Solid Waste
Association of North American AEERL is also responsible for demonstrating innovative
approaches to the control of waste methane such as the application of fuel cell technology to
recover energy from landfill gas and digester gas.
This paper describes the emission potential for waste management sources and the
mitigation opportunities. It also provides an overview of some of the barriers in the U.S. that
affect methane utilization. This research is funded through EPA's Global Climate Change
Research Program. This paper has been reviewed in accordance with EPA's peer and
administrative review policies and approved for presentation and publication.
4-47
-------
INTRODUCTION
Waste disposal results in emissions of greenhouse gases including methane (CH4), carbon
dioxide (COa), nitrous oxide (NaO), ozone precursors, and chlorofluorocarbons. The major
sources of CH* from waste management include landfills, wastewater treatment lagoons, and
livestock waste. Current estimates suggest that this source accounts for up to 125 Tg/yr or -40%
of the estimated total global anthropogenic emissions of 300 Tg/yr (1PCC, 1992). Landfills have
been estimated to contribute as much as 60 Tg/yr of CH*. Policies are being considered to reduce
greenhouse gas emissions to meet the goals of the United Nations Conference on Environment and
Development held in Rio de Janeiro in June. Emissions sources that are amenable to control -
such as landfills - have been given a high priority for clarification. (EPA, 1989)
"Waste" CHt results from the anaerobic decomposition of biodegradable waste found in
landfills, open dumps, waste piles, wastewater treatment lagoons, septic sewage systems, and
livestock waste. This waste CHt can be a source of pollution as well as a resource. There are 114
landfill gas (LFG)-to-energy projects in the U.S. (Thomeloe, 3/92) and 200 UFG-to-energy
projects worldwide (Richards, 1989). Landfill gas is utilized (1) as medium-heating-value fuel,
(2) to generate electricity using internal combustion engines, or gas and steam-fed turbines, and (3)
as high-heating-value fuel in which case the gas is upgraded and fed into a nearby natural gas
pipeline. U.S. landfills currently generate 344 MW of electricity (Thomeloe, 3/92). The gas that
is formed from anaerobic decomposition is typically 50 to 55% CHU, 45 to 50% COi, and <1%
trace constituents.
The CHU is a concern because of its global warming effects and explosive potential.
Emissions of nonmethane organic compounds (NMOQ contribute to tropospheric ozone which
aggravates urban smog and is a concern to human health and the environment Other LFG
constituents such as vinyl chloride, benzene, carbon tetrachloride, and methylene chloride are a
concern for their cancerous and noncancerous effects. The Agency has proposed CAA regulations
for emissions from municipal solid waste (MSW) landfills (FR, 1991) which will reduce five
health and welfare effects: (1) explosion hazards, (2) global warming effects from CH* emissions,
(3) human health and vegetation effects caused by ozone formed from NMOCs, (4) carcinogenkity
and other possible noncancerous health effects associated with specific landfill emissions
constituents, and (5) odor nuisance (U.S.EPA, 3/91). Estimates from the proposed regulations
indicate that 621 landfills of the 6,000 existing active landfills would be required to collect and
control MSW landfill emissions (p. 24480, FR, 1991).
The proposed dean Air Act regulations do not require utilization of the gas. Although
increased COi emissions are being traded off for reduced CRi emissions, there is a net benefit due
to the difference in the radiative forcing capacity between COi and CH* The radiative forcing
capacity of CHt to CO2 on a molecular basis is 21 tunes that of COi (p. 53, IPCC, 1990). It is
hoped that the sites affected by these regulations will consider LFG to energy as opposed to flaring
the gas. The use of energy recovery for the control of MS W landfill air emissions will result in
decreased emissions of GH* NMOCs, and toxics. Additional benefits include the conservation of
global fossil fuel resources, reduction of emissions at coal-fired power plants, reduced dependency
on imported oil, and cost savings to public entities that receive royalty payments (Thomeloe, 6/92).
However, mere are many barriers in the U.S. associated with the utilization of waste CH*
This paper provides data and infonnaribn that characterize the CH4 potential of different
waste management sources including landfills, wastewater treatment lagoons, septic sewage
systems, and livestock waste. Mitigation opportunities are identified and the different options are
described. Barriers that affect waste CH4 utilization are identified. This paper describes research
mat is being conducted through EPA's Office of Environmental Processes and Effects Research on
4-48
-------
Global Climate Change. AEERL has responsibility for characterizing the emission potential and
investigating mitigation opportunities for the major sources of greenhouse gas emissions
(Thorneloe, 1991).
"Waste" CRt potential from "Waste" Management Sources
The major sources of CEU from the anaerobic decomposition of waste include landfills,
wastewater treatment lagoons, septic sewage systems, and livestock waste. Table 1 presents an
estimate of the relative contributions of each of these sources for the U.S. and globally. These
estimates suggest that these sources on average account for 80 Tg/yr or ~30% of the total global
anthropogenic emissions of 300 Tg/yr (IPCC, 1992).
Table 1. U.S. and Global Estimates (Tg/yr) of "Waste" Methane Emissions
U.S.
Avg. Range
Reference
Global
Avg. Range
Reference
Landfills 9
Wastewater Treatment/
Sewage Treatment ?
livestock Waste 4
(6-13)« U.S. EPA, 7/92 30 20-70 IPCC, 1992
Safle, 1992
25 ?
25 (20-30)
IPCC, 1992
IPCC, 1992
Potential emissions, not corrected for the amount that is flared or utilized Approximately 12
million tonnes of CKj is being recovered from U.S. landfills (Thorneloe, 3/92).
Estimates of global CHU emissions were summarized by the IPCC and suggest that
landfills contribute -30 Tg/yr with a range from 20 to 70 Tg/yr (p. 35, IPCC, 1992, Khalil and
Rasmussen, 1990). Preliminary estimates generated using AEERL's empirical model indicate that
potential landfill CHj emissions in the U.S. range from 6.3 to 13 Tg/yr, with an average of 10
Tg/yr. Global estimates suggest a range of 20 to 40 Tg/yr of CHt emissions with an average of 30
Tg/yr. Estimates generated using Bingemer and Crutzen's approach - which is currently proposed
as the official IPCC methodology (OECD, 1991) indicate that landfill QHU emissions contribute
60 Tg/yr globally and 23 Tg/yr in the U.S. (U.S. EPA, 7/92).
The estimates generated using the empirical model are thought to more accurately reflect the
amount of CHt from landfills that is contributing to the global CK* flux (Campbell et al., 1991,
Peer et at, 3/92, Peer et aL, 1992). The estimate using the empirical model uses data from landfill
gas recovery systems and accounts for CHt oxidation and gas recovery efficiency. The data that
were used to develop the empirical model were collected from over 100 U.S. landfills. An EPA
report is being published that documents the development of the model and the estimate of CHj
emissions for U.S. landfills. Future refinements of this estimate will adjust for waste composition
using data being developed on the gas potential of different biodegradable waste streams.
Estimates for wastewater treatment are less reliable primarily due to a lack of country-
specific data needed to characterize the CHU potential of municipal and industrial wastewater
treatment There is also a lack of field data characterizing the CHj potential from lagoons
(Thomeloe, 2/92). Lagoons (or surface impoundments) are usually earthen pits used to contain
and process wastewater. AEERL is initiating a field test program in 1993 to collect lagoon
characterization data such as the biological oxygen demand (BOD) loading, flow rates, and
4-49
-------
retention time. These data will be used to develop a methodology for estimating greenhouse gas
emissions from lagoons including CH4, C02, ^0, and NMOC emissions. Initial estimates for
this source suggest that CEt emissions range from 10 to 40 Tg/yr (EPA, 7/92).
CHt emissions from wastewater treatment lagoons are not expected to be a major source in
the U.S. since many digesters flare and sometimes utilize the gas to control hydrogen sulfide
emissions. However, lagoons may be a more significant source in developing countries where
lagoons are being more frequently used and the gas is not controlled. Agencies such as the World
Bank (Bartone, 1990) recommend the use of lagoons for wastewater treatment for developing
countries since land space is readily available, operation is relatively simple, cost is low, and
energy requirements are minimal. This represents a potential opportunity to work with developing
countries to demonstrate that the CH4 can be utilized as an alternative energy source.
Individual onsite wastewater treatment systems, such as septic systems, are used
throughout the world. In China, there are an estimated 10 million biogas pits, which are designed
to produce biogas for household use. However, the majority of the world does not collect the gas
from septic systems. A portion of this CHU will be oxidized and some will be emitted to the
atmosphere. Field test work by EPA/AEERL is planned in FY94 to collect data that will result in
more reliable estimates for this source and to determine if this source is amenable to cost-effective
control.
The only published global estimate for livestock waste suggests that QLi emissions from
this source are about 28 Tg/yr with a range of about 20 to 35 Tg/yr (Safle et al., 1992). These
estimates were made by collecting information from animal waste management systems and the
quantity of animal waste managed by each system. Information was also collected from
government statistics and literature reviews. The major uncertainty regarding these estimates is due
to the assumptions and data characterizing the CHj potential from the waste of free-range animals.
AEERL is planning to conduct laboratory and field studies in FY93 that will help reduce the current
uncertainty with these estimates and will investigate opportunities for cost-effective control.
MITIGATION OPPORTUNITIES FOR "WASTE" METHANE
The recently proposed regulations for MSW landfills will result in the reduction of 5 to 7
millions tonnes of CHU. Currently U.S. landfills are recovering 12 million tonnes of CH* and
producing 344 MWe of power. In the U.S., there are 114 LFG-to-energy projects (Thorneloe,
3/92). The breakdown of these projects by energy utilization option is presented in Figure 1. The
majority of these projects (Le., 75%) generate electricity which is either used onsite or sold to a
local utility. These projects are located across the U.S. in 28 states, with 38 LFG-to-energy
projects in California and 14 in New York. Of the 24 projects using turbines, 21 projects are gas-
fed and 3 projects are steam-fed. The largest project in the world is the Puente Hills Landfill in
Whitner, California, It is operated by the Los Angeles County Sanitation Districts and generates
50 MWe (Valenri, 1992). .The majority of projects mat produce electricity (Le., -80% or 66 out
of 85) produce 1 to 5 MWe (Thomeloe, 6/92). Typically there are three to five engines or one to
two turbines per project
4-50
-------
Figure 1. Number of U.S. Landfill Gas Projects by Energy Utilization Option
(Source: Thomeloe, 3/.92)
.J2 1C Engines
I
« Mod Heating-Value Gas
!S
5
£
Turbines
g
UJ
High Heating-Value Gas
20 30 40 50
Number of LFG Projects
60
70
Globally there are about 200 LFG-to-energy projects (Richards, 1989). There are also
digesters in use for livestock waste and waste water treatment sludge that utilize the waste CHj fen-
producing steam or electricity. Data are not presently available to calculate the extent of digester
gas utilization. The Expert Working Group on Landfill Gas of the International Energy Agency is
compiling data on waste CH4 projects (Lawson, 1992). These data will be available in the future
to adjust current estimates of "waste" CH4 for the amount that is controlled. In the U.S. 1.2
millions tonnes of waste CKU is being utilized by LFG-to-energy projects (Thomeloe, 3/92).
The utilization of waste CH» can also result in a substantial cost savings to public entities
that own landfills and receive royalty payments. For example, Pacific Energy - who has
developed 25 LFG-to-energy projects - has paid out $13 million in royalties, mostly to public
entities. On average, Pacific Energy's projects are in the sixth year of operation under anticipated
20-year project lives (Wong, 1992). Other economic benefits include the purchase of goods and
services. In 1991, Pacific Energy purchased over $4 million in outside goods and services to
support its LFG projects plus a payroll of >$3 million. LFG to energy projects tend to be capital
intensive and are typically built on what is considered undevelopable acreage. Pacific Energy's
eight LFG-to-energy projects in California pay >$350,000 per year in property taxes in California
and require few public services (Wong, 1992).
There are emerging technologies for waste CHi utilization. For example, AEERL initiated
a project in 1991 to demonstrate the use of fuel cells to recover energy from landfill gas. There are
a number of advantages with the use of fuel cells including higher energy efficiency, availability to
smaller as well as larger landfills, minimal byproduct emissions, minimal labor and maintenance,
and minimal noise impact (Le., because there are no moving parts). The type of fuel cell being
demonstrated for LFG application is the commercially available 200 kWe phosphoric acid fuel cell
(PAFC) power plant The 1-year full-scale demonstration is scheduled for 1993 (Sandelli, 1992).
4-51
-------
Hie major technical issue associated with the application of fuel cell to LFG is finding a gas
cleanup system that effectively and economically cleans the gas to the fuel cell's stringent
requirements. LFG composition can be quite variable as to the type of constituents and
concentration. Chloride and sulfur compounds are quite common. "Slugs" of condensate have
also been known to cause havoc at gas turbine and internal combustion engine projects (Augenstein
and Pacey, 1992). If this project is successful, it will provide a more environmentally attractive
option for waste CFU utilization that is also more energy efficient.
A second project has been initiated by AEERL to demonstrate the use of fuel cells to
recover energy from digester gas. Digesters are frequently used at wastewater treatment (WWT)
facilities to process sludge. The digesters can utilize the waste heat from the fuel cells which
would result in an energy efficiency approaching 80%. The same type of commercially available
fuel cell will be used. The three issues to be addressed in this project are (1) feasibility of
integrating the PAFC power plant operation with the WWT plant, (2) anaerobic digester gas/waste
CH4 cleanup and processing requirements for fuel cell operation, and (3) improved fuel cell
performance on reduced heating value fuel (i.e., waste CHj versus natural gas). This project is to
begin this.fall and a 1-year demonstration is planned for FY94. It is expected that the LFG cleanup
process can also be used for digester gas. Generally digester gas trace constituents are less
variable than LFG although concentrations of sulfur compounds tend to be greater. The first phase
of die project will evaluate the requirements for the digester gas cleanup process and the application
of fuel cells to recover energy from a sludge digester at a WWT plant The design specifications for
the digester gas/fuel cell system will be provided in the Phase 1 Report
Other emerging technologies for landfill gas include the production of liquid diesel fuel
such as the process in Pueblo, Colorado, that began operation last year. A second site in the U.S.
has been proposed to produce vehicular fuel from landfill gas. The South Coast Air Quality
Management District has awarded a contract to demonstrate a process for producing methanol from
landfill gas. The site selected for this demonstration is the BKK landfill, where there was co-
disposal of hazardous and municipal waste. TeraMeth Industries is responsible for the
demonstration which is scheduled to begin in 1993.
To help promote and encourage landfill gas utilization, case studies of six sites were
conducted in FY91/91 The final report (Augenstein and Pacey, 1992) contains detailed
information on the six LFG-to-energy projects. In addition, the report provides information on a
project recently developed by Michigan Cogeneration Systems (Appendix N), 25 Waste
Management LFG-to-energy projects (Appendix M), 11 Laidlaw LFG-to-energy proiects
(Appendix L), and 4 case studies of United Kingdom LFG-to-energy projects (Appchdix K). This
report is being referenced in the upcoming CAA regulations for U.S. MSW landfills as an
"enabling" tool that provides up-to-date information on LFG utilization for landfill owners and
operators. The report has generated a great deal of interest both in the U.S. and internationally
through the International Energy Agency, The International Solid Waste Association, and the Solid
Waste Association of North America.
A follow-up technology transfer project is focusing on the technical issues associated with
LFG cleanup and energy equipment modifications (L&, application to LFG versus natural gas).
There are different philosophies pysoefaT*^ with "waste" CH4 utilization. Information provided by
industry experts in the U.S. and Europe is being collected. These projects have been in existence
since the early 1980s and much has been learned as to what constitutes a "successful** project Hie
EPA report mat is scheduled to be published in the Fall of 1993 will review the current state of
knowledge for successful waste CH4 utilization projects. This technology-transfer project is
intended to help ensure that future utilization projects are designed and operated using the must up-
to-date knowledge and information on gas cleanup and energy equipment modifications.
4-52
-------
BARRIERS TO "WASTE" Ott UTILIZATION IN THE U.S.
A major factor in helping to encourage LFG-to-energy projects is the Public Utility
Regulatory Policy Act (PURPA). It guarantees that utilities will purchase power that was
generated from landfills at a price related to the costs that utilities would experience to produce the
same amount of power. Although this guarantees a purchaser for the power, the power sale
revenues may be low if the utilities ' own generating costs are low. This is the case in pans of the
U.S. where electricity is generated using water power (i.e., hydroelectric). However, with current
low energy prices, most regions of the U.S. also have unattractive "buy-back rates" for the
electricity that is generated using waste CH*. In addition, incentives such as federal tax credits that
have helped to encourage these projects appear to being losing favor. Industry experts think that
many of the marginal projects cannot continue without these tax credits. Current trends are toward
lower energy prices, reduced tax incentives, and increasing environmental liability.
Although there are more than 6000 landfills in the U.S., there are less than 120 LFG-to
energy projects. During the oil crisis in the 1970s/1980s, when the price of oil increased from $6-
8 per barrel to S35 per barrel, there was much more interest in developing alternative sources of
energy such as LFG-to-energy projects. With the current prices of energy, it is much more
difficult to find projects that are economical. Most U.S. projects that have had to cease operation
did so primarily due to economics. Projects that are upgrading the gas to pipeline quality have
been especially hard hit due to high operating costs and low revenue. Projects of mat type are not
being planned in the U.S. However, sites in the Netherlands are rinding more favorable
economics (Scheepers, 1991).
Laidlaw Technology Inc. suggests that "successful" LFG projects need to be OVQ 1
and have an electrical price of at least $0.06-0.07/kWh including any capacity payments. Royalties
should not exceed 12.5% at this energy pricing (Jansen, 1992). Laidlaw also suggests mat, if
higher royalties are offered, the percentage should be a function of energy pricing over and above
the base energy rate as inflation occurs. The early LFG projects were based on an established firm
price for net energy which provided a substantial degree of security to developers. Contracts for
many LFG projects do not allow for fluctuations in energy rates and costs. Revenues for energy
sales are usually based on prices of the "competition" of equivalent energy sources (e.g.,
petroleum produces). Since the value of the energy base commodity can fluctuate, this can impact
profit
Administrative and development costs have increased as revenues have decreased. These
costs include legal fees, permit applications, and contract negotiations for gas lease agreements and
power purchase agreements. These costs may vary widely depending on the environmental issues,
development considerations, and regulatory requirements; John Pacey of Emcon Associates has
found that these costs can vary from $30,000 to $1,000,000 per kWh for a 1 MWe LFG-to-energy
project (Augenstein and Pacey, 1992). The costs for a 1 MWe project are summarized in Table 2.
The gas extraction/collection systems are less than 15% of the total cost of the project The major
cost component is the electricity generating equipment
Tax credits are proportional to gas energy delivery as legislated by Congress (Section 29 of
the IRS Code) in 1979 to encourage non-fossil fuel use. These credits are a direct offset to taxes
and canoe used only to offset a profit The tax credits will extend to the year 2003 and are
allowable for extraction systems installed prior to the end of the year 1992. Robert F. Hatch of
Cambrian Energy Systems - whose company has been involved in arranging financing for many
U.S. LFG to energy projects - thinks that many of the projects would not be in existence if the tax
credits were not available. Since energy prices are relatively low, some projects today can be
financed only because of the tax credits. The tax credits are intended to help promote the
development of a domestic resource as opposed to using foreign oil (Hatch, 1991). These credits
4-53
-------
have helped to encourage LFG-to-energy projects and also have helped municipalities defray the
cost of environmental regulations.
Table 2. Estimate of Capital Costs for a 1 MWe Landfill Gas-to-Energy Utilization Project
Cost*
Item
Extraction/Collection System0
Fces-Planning/Environmeru/Legald
Interconnect Cost
Generating Equipment
Contingency
Total
200
30
76
970
221
1500
Percent'
13
2
5
65
15
100
'These costs were provided by Laidlaw Technologies, Inc. (Jansen, 1992)
bAugenstein and Pacey, 1992.
(The range in cost of gas cleanup systems is $ 10,000-$500,000/kWh)
dLegal fees are approximately 50% of the total (i.e., ~$15,000-$500,000/kWh)
Range of
Value
200-1000
30-1000
20-500
500-2000
850^500
Another barrier to LFG (or waste CJit) utilization can be environmental regulations. Typically
the overall environmental benefit as well as the energy and economic benefits are not considered.
Many LFG projects axe operated such that New Source Review is avoided. This results in less
CHU being utilized. For example, a project in Phoenix, Arizona, is producing 50% of the energy
that is possible to avoid triggering New Source Review. The project was not allowed to take credit
for the offset in emissions from a coal-fired power plant The developer found out mat the New
Source Review process would take an extra 2 years to obtain a permit. Rather than delay receiving
payback on the investment, the decision was made to operate the project at half of its potential
Another dis-incenrive is that some operators of LFG-to-cnergy projects are finding that the cost
of condensate disposal is becoming a major expense. The condensate is formed when the gas is
compressed. The LFG condensate - which is being classified as a hazardous waste - requires
disposal at a Subtitle C facility. This cost [i.e., $0.18/L (~$0.70/gal)] can be significant for a site
where lean-burn engines or turbines are used as compared to the use of flares where minimal
condensate is collected [Le., 3,800 L/day (1000 gpd) for lean-bum engines or turbines versus 760
L/day (200 gpd) for Hares] (Jansen, 1992).
Industry experts are finding that air, water, and solid waste agencies have conflicting goals.
LFG-to-energy projects have been forced to shut down due to concerns for by-product emissions
of nitrogen oxides (NO*) and carbon monoxide (CO). In California last year, 48 items of state
legislation affecting solid waste were enacted (SWANA, 1992). Regulatory priorities often appear
CONCLUSIONS
Landfills and other waste management sources such as wastewater treatment lagoons, septic
sewage systems, and livestock waste are amenable to cost-effective control and.are relatively
significant sources of CH* The EPA's Global Climate Change Research Program is conducting
research to (1) reduce the uncertainty in global emission estimates for those sources amenable to
-------
control, (2) target control strategies that are cost-effective, and (3) provide data and information
that will help support regulatory activities and IPCC activities.
Currently, U.S. LFG-to-energy projects recover -1.2 million tonnes of CHi and produce 344
MWe of power. The proposed CAA regulations for MSW landfill air emissions are expected to
result in additional reductions ranging from 5 to 7 million tonnes of CHU. Utilization of LFG for
those sites affected by the proposed CAA regulations has the potential to result in increased
benefits to the national economy and global environment. The utilization of alternative energy
sources such as "waste" CHt extends our global fossil fuel resources. Not only are emissions
directly reduced when waste CHt is recovered and utilized, but emissions are also indirectly
reduced when secondary air emission impacts associated with fossil fuel use are considered.
REFERENCES
1. Augenstein, D. and J. Pacey, "Landfill Gas Energy Utilization: Technology Options and Case
Studies," EPA-600/R-92-116 (NTIS PB92-203116), June 1992.
2. Banone, C. R. 1990. Water Quality and Urbanization in Latin America. Water International,
15: 3-14.
3. Campbell, D., D. Epperson, L. Davis, R. Peer, and W. Gray, "Analysis of Factors Affecting
Methane Gas Recovery from Six Landfills," EPA-600/2-91-055 (NTIS PB92-101351),
September 1991.
4. Federal Register. Vol 56. No. 104. May 30,1991, pp. 24468 - 24528.
5. Hatch, RJF. "The Federal Tax Credit for Non-Conventional Fuels: Its Status and Role In the
Landfill Gas Industry." Proceedings from SWANA's 14th Annual International Landfill Gas
Symposium, 1991. ;'
6. Intergovernmental Panel on Climate Change. "Climate Change - The IPCC Scientific
Assessment." World Meteorological Organization/United Nations Environment Programme.
Edited by J.T. Houghton, GJ. Jenkins, and JJ. Ephraums, 1990.
7. Intergovernmental Panel on Climate Change. "Climate Change 1992 - The Supplementary
Report to the IPCC Scientific Assessment." World Meteorological Organization/United
Nations Environment Programme. Edited by J.T. Houghton, B. A. Callander, and S. K.
Varney, 1992.
8. Jansen. G.R. "The Economics of LFG Projects in the United States." Presented at the
Symposium on LFG/Applications and Opportunities in Melbourne, Australia, February 27,
1992. - - - ; ,. :.--: * .-. : - ' -(
9. Khalil, M-AJC. and RA. Rasmussen, "Constraints on the Global Sources of Methane and an
Analysis of Recent Budgets." Tellus, 42B, 229-236,1990.
20. Lawson, P. S. Landfill Gas Expert Working Group Summary Report, 1989-1991,
International Energy Agency: Biomass Conversion Agreement: MSW Conversion
Activity, Task VILAEA-EE-0305, April 1992.
4-55
-------
11. Organization of Economic Cooperation and Development (OECD). Estimation of
Greenhouse Gas Emissions and Sinks. Final Report from the OECD Experts
Meeting February 1991. Prepared for the IPCC. August 1991.
12. Peer, R.L., D.L. Epperson, D. L. Campbell, and P.V. Brook, "Development of an
.Empirical Model of Methane Emissions from Landfills," EPA-600/R-92-037 (NTIS
PB92-152875), March 1992.
13. Peer, R.L., S.A. Thorneloe, and D.L. Epperson, "A Comparison of Methods for
Estimating Global Methane Emissions from Landfills." Chemosphere, 1992 (In Press).
14. Richards, K.M. Landfill Gas: Working with Gaia, C.A.B. International, Vol. 3, No. 4,
December 1989.
15. Safle, L.M., M.E. Casada, J. W. Woodbury, and K. F. Roos, Global Methane Emissions
from Livestock and Poultry Manure, EPA/400/1-91/048, February 1992.
16. Sandelli, GJ. "Demonstration of Fuel Cells to Recover Energy from Landfill Gas
(Phase I Final Report: Conceptual Study)," EPA-600/R-92-007 (NTIS PB92-
137520), January 1992.
17. Scheepers, MJJ. "Landfill Gas in the Dutch Perspective," published in Proceedings
of the Third International Landfill Symposium, Sardinia, October 1991. *
18. SWANA. List of Solid Waste Legislation Enacted in 1991.1992.
19. Thorneloe, S A. Held Test Work for Assessing Greenhouse Gas Emissions from
Wastewater Treatment and Septic Sewage Systems. Memo to Jon Kessler, Air and
Energy Policy Division, Office of Planning and Evaluation, EPA, February 24,1992.
20. Thorneloe, S. A. "Landfill Gas Recovery/Utilization - Options and Economics,'*
presented at the Sixteenth Annual Conference by the Institute of Gas Technology
on Energy from Biomass and Wastes, Orlando, FL, March 2-6,1992.
21. Thorneloe, S.A. "Landfill Gas Utilization - Options, Benefits, and Barriers,"
presented at the Second United States Conference on Municipal Solid Waste
Management," Arlington, VA, June 3-5,1992.
22. Thorneloe, S A. "U.S. EPA*s Global Climate Change Program - Landfill Emissions
and Mitigation Researr1* '* *~J ~* *u~ '**-** ₯-»-:.-! T i*tn <;.._».,>..;.._
CagHart Italy, October
and Mitigation Research,'* presented at the Third International Landfill Symposium,
14, 1991.
23. United States Environmental Protection Agency. "Air Emissions from Municipal
Solid Waste Landfills - Background Information for Proposed Standards and
Guidelines." EPA-450/3-90-01 la (NTIS PB91-197061), March 1991.
24. United States Environmental Protection Agency, Office of Policy, Planning, and
Evaluation. International Methane Emissions. Draft Report to Congress, July 1992.
4-56
-------
25. United States Environmental Protection Agency, Office of Policy, Planning and
Evaluation. Policy Options for Stabilizing Global Climate. Draft Report to
Congress. February 1989.
26. Valenti, M. "Tapping Landfills for Energy." Mechanical Engineering, Vol. 114,
No. 1, January 1992.
27. Wong, F.P. "Alternative Energy & Regulatory Policy: Till Death Do We Part,"
presented at AWMA Conference on "Cooperative Clean Air Technology -
Advances through Government and Industrial Partnership" in Santa Barbara, CA,
March 21 - April 1,1992.
4-57
-------
Paper 4-E
by: R. J. Spiegel
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
G. J. Sandeili
International Fuel Cells Corporation
South Windsor, CT 06074
ABSTRACT
International Fuel Cells Corporation (IFC), a subsidiary of United Technologies
Corporation, is conducting a U.S. Environmental Protection Agency (U.S. EPA)
sponsored program to demonstrate methane control from landfill gas using a
commercial phosphoric acid fuel cell power plant. This is the world's first commercial-
scale demonstration to control methane emissions from landfills using a fuel cell
energy recovery system. The U.S. EPA is interested in fuel cells for this.application
because it is potentially the cleanest energy conversion technology available. This
paper discusses the project in general and describes some results to date, with
emphasis on the landfill gas pretreatment system.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
4-56
-------
IfsTTHODUCTlON
The U.S. EPA has proposed standards and guidelines [1] for the control of air
emissions from municipal solid waste (MSW). landfills. Although not directly controlled
under the proposal, the collection and disposal of waste methane, a significant
contributor to the greenhouse effect, would result from the emission regulations. This
U.S. EPA action will provide an opportunity for methane control as well as for energy
recovery from the waste methane that could further benefit the environment. Energy
produced from landfill gas could offset the use of foreign oil, and air emissions
affecting global warming, acid rain, and other health and environmental issues.
To demonstrate that methane control and subsequent energy recovery via fuel
cells are technically, economically, and environmentally feasible, two key issues must be
addressed: to define a gas pretreatment system to render the landfill gas suitable
for fuel cell uses and to design the modifications necessary to ensure that rated
power is achieved from the dilute methane fuel. Only relatively simple engineering
modifications, albeit initially costly to implement, are required to ensure that rated
power is achieved from the dilute landfill gas. The toughest and most critical problem
is the gas cleanup system. Therefore, the R&D focus of this project is on the landfill
gas contamination problem.
International Fuel Ceils Corporation (IFC), a subsidiary of United Technologies
Corporation, was awarded a contract by the U.S. EPA to demonstrate methane
destruction and energy recovery from landfill gas using a commercial phosphoric acid
fuel cell. IFC is conducting a three-phase program to show that fuel cell technology is
economically and environmentally feasible in commercial operation. Work was initiated
in January 1991. A U.S. EPA report [2] describes the results of Phase I, a conceptual
design, cost, and evaluation study, which addresses the problems associated with
landfill gas as the feedstock for fuel cell operation.
Phase II of the program includes the design, construction, and testing of the
landfill gas pretreatment module to be used in the demonstration. Its objective will be
to determine the effectiveness of the pretreatment system design to remove critical
fuel ceil catalyst poisons such as sulfur and halides. A challenge test is planned to
show the feasibility of using the pretreatment process at any landfill in conjunction
with the fuel cell energy recovery concept. The gas pretreater is described here.
Phase III of this program will be a demonstration of the fuel cell concept. The
demonstrator will operate at Penrose Station, an existing iandfill-gas-to-energy facility
owned by Pacific Energy in Sun Valley, CA. Penrose Station is an 8.9 MW internal
combustion engine facility supplied with landfill gas from four landfills. The electricity
produced by the demonstration will be sold to the electric utility grid.
4-59
-------
Phase II activities began in September 1991, and Phase III activities are
scheduled to begin in June 1993.
During Phase I, a commercial fuel cell methane destruction and energy recovery
system concept was designed. The system, shown in Figure 1, is based on
commercially available equipment adapted for operation on landfill gas. The system
was sized to be broadly applicable to a large number of landfills.
PENROSE
STATION
GAS WELLS
AND
COLLECTION
SYSTEM
(PACIFIC ENERGY)
UTILITY
POWER
LINES
GAS-GUARD
OAS PRETREAT-
MENT SYSTEM
(BIOQAS
DEVELOPMENT
INC.)
PC2S
FUEL CELL
POWER
PUNT
(ONSI CORK)
AC POWER
TO GRID
LANDFILL
COGENERATION
HEAT
NATURAL OAS
SOUTHERN CALIFORNIA GAS COMPANY
Figure 1. Project Conceptual Design
Landfill gas is collected by a series of wells in the MSW landfill and piped to a
gas pretreatment module. The pretreatment module removes contaminants such as
sulfur and halides which affect the operation of the fuel cells. The contaminants are
concentrated on an absorption bed to a predetermined level. Then during a
regeneration cycle they are stripped from the absorption media and destroyed by
incineration. Hydrocarbon condensates which form in the pretreater are also
incinerated. The resulting output is a medium heat value methane fuel suitable for use
in the fuel cell.
4-60
-------
The concept utilizes four modular 200-kW phosphoric acid fuel ceils generating
electricity to be sold tc the electric utility grid. The fuel cell power plants are
adaptations from the natural-gas-fueled PC25 fuel cell sold by ONSI Corporation, an
IFC subsidiary. Only simple modifications are required to ensure that rated power is
achieved from the dilute landfill gas.
The MSW landfills in the United States were evaluated to determine the potential
commercial market which could be derived using a 200-kW fuel cell. Each fuel cell
would consume 2832 scmd (100,000 scfd) of landfill gas to generate 200 kW,
assuming a heating value of 19 MJ/dscm (500 Btu/dscf).
The potential commerciai market available for fuel cell operation was evaluated
using a U.S. EPA estimate of methane emissions in the year 1997 [3] and an estimate
of landfill gas production rate of 0.0028 scm/yr per tonne (0.1 scf/yr per ton) of
refuse in place [4]. An estimated 4370 MW of power could be generated from the
7480 existing and closed sites that were identified [2]. The largest number of
potential sites greater than 200 kW occurs in the 400 to 1000 kW range. This
segment represents a market of 1700 sites or 1010 MW. The assessment concluded
that these sites are ideally suited to the fuel cell concept. The concept can provide a
generating capacity tailored to the site because of the modular nature of the
commercial fuel cell. Sites in this range are also less well served by competing
options, especially Rankine and Brayton Cycles, which exhibit poorer emission
characteristics at these power ratings.
ENVIRONMENTAL ASSESSMENT OF THE FUEL CBJL SYSTEM
The environmental impact from commercial application of the fuel cell concept
to the market described previously can be assessed. For the purpose of the
evaluation, a site capable of supporting four fuel cell power modules (80& kW total
capacity) was selected. The site would produce approximately 11,328 scmd
(400,000 scfd) of landfill gas per day. The gas contains approximately 50% methane
with a heating value of 19 MJ/dscm (500 Btu/dscf).
The analysis of the environmental impact shows that the fuel cell can be
designed to eliminate the methane and non-methane organic compounds (NMOCs)
from landfill gas streams. With the fuel cell system, significant amounts of CO2 and
SO2 wiH also be reduced due to the fuel cell energy generation. Using an 80% capacity
factor for the fuel cell and offsetting emissions from electric utility power generation
using a coal-fired plant meeting New Source Performance Standards, it can be shown
that for the example site the fuel cell energy conversion system provides 5.6 million
kWh of electricity per year and a reduction of emissions for CH4t NMOCs, CO2. SO2,
and CO of 1200, 35, 4200, 36, and 0.6 Mg/yr, respectively. These reductions can be
4-61
-------
used as environmental offsets, particularly in critical areas such as California or other
locations with severe environmental restrictions.
The environmental impact of the fuel cell concept to the potential U.S. market
can also be estimated. If the viable market is assumed to range in sites that have
energy capacities from 200 kW up to 1 MW, then trie fuel cell system can provide an
approximate net reduction in emissions for CH«, NMOCs, COa.SOg, and CO of 2x10«,
6x10*. 7x106, 6x104, and 1x103 Mg/yr, respectively.
FUR CELL POWER PLANT
A design of a fuel cell power plant was established to identify the design
requirements which allow optimum operation on landfill gas. Three issues specific to
landfill gas operation were identified which reflect a departure from a design
optimized for operation on natural gas. A primary issue is to protect the fuel cell
from sulfur and haiide compounds not scrubbed from the gas in the fuel pretreatment
system. An absorbent bed was incorporated into the fuel celt fuel preprocessor
design which contains both sulfur and haiide absorbent catalysts. A second issue is
to provide mechanical components in the reactant gas supply systems to *
accommodate the larger flow rates that result from use of dilute methane fuel. The
third issue is an increase in the heat rate of the power plant by approximately 10%
above that anticipated from operation on natural gas. This is a result of the
inefficiency of using the dilute methane fuel. The inefficiency results in art increase in
heat recoverable from the power plant. Because the effective fuel cost is relatively
low, this decrease in power plant efficiency will not have a significant impact on the
overall power plant economics.
The landfill gas power plant design provides a packaged, truck transportable,
self-contained fuel cell power plant with a continuous electrical rating of 200 kW. It is
designed for automatic, unattended operation, and can be remotely monitored. It
can power electrical loads either in parallel with or isolated from the utility grid.
. *
In summary, a landfill-gas-fueled power plant can be designed to provide 200
kW of electric output without need for technology developments. The design would
require selected components to increase reactant flow rates with a minimum pressure
drop. To implement the design would require non-recurring expenses for system and
component design, verification testing of the new components, and system testing to
verify the power plant performance and overall system integration.
4*62
-------
LANDFILL GAS PRETREATMENT SYSTEM
The available information on landfill gas compositions was evaluated to
determine the range of gas characteristics which a fuel ceil landfill-gas-to-energy
power plant will encounter. This information was used to set the requirements for the
gas pretreatment and fuel cell power plant designs.
The major non-methane constituent of landfill gas is C02. The CO2 ranges from
40 to 55% by volume of the gas composition with a typical value of 50%. Other
diluent gases include nitrogen and oxygen, which are indicative of air incursion into the
well (most frequently in perimeter wells). Nitrogen concentrations can range as high
as 15%, but typical values are 5% or less. Oxygen concentrations are monitored
closely and held low for safety reasons.
Landfill gas constituent compounds reported by U.S. EPA [3] indicate a typical
value for the total NMOCs of 2700 ppmv, ranging from 240 to 14,000 ppmv
(expressed as hexane). The NMOC concentration in the landfill gas is an important
measure of the total capacity required in the gas pretreatment system, while the
specific individual analyses provide a basis for gas pretreatment subcomponent sizing.
The specific contaminants in the landfill gas, of interest to the fuel cell, are sulfur and
halides (chiefly chlorides and fluorides). The sulfur level ranges from 1 to 700 ppmv,
with a typical value on the order of 21 ppmv. Sufficient data were not available to
assess the range of the halides, but a typical value of 132 ppmv was calculated for
this contaminant [3]. The range of contaminant values varies hot only from site to
site, but also at any given site with time due to seasonal weather or moisture
content. These characteristics require the pretreatment system design to be capable
of handling these gas quality variations to avoid expensive site specific engineering of
the pretreatment design which would affect the marketability and economics of the
concept
Rgure 2 illustrates the overall design strategy for the gas cleanup system. As
shown, the raw landfill gas pretreatment system is designed to reduce the primary
fuel cell contaminants (sulfur and halides) to levels between 1 and 10 ppm. A nominal
value of around 3 ppm is the design goal. An additional spool piece is added to
protect the fuel cell from sulfur and halide compounds not removed from the gas in
the pretreatment system. This is an absorbent bed incorporated into the fuel cell
preprocessor design which contains both sulfur and halide absorbent catalysts. Note
that the capacity of this bed for contaminant removal is dictated by the number of
hours the fuel celt can operate before these poisons contaminate the fuel ceil, requiring
refurbishing.
4-63
-------
I
Raw
LFG
Pretreatment
system
Spool
piece
PC25
Pretreat capability
1-10 ppm
PC25 requires
0.15 ppm for 8000 hours
0.01 ppm for 40,000 hours
LFG = Landfill Gas
Figure 2. Block Diagram of Overall Landfill Qas Contaminant Removal Process
A block diagram of the landfill gas pretreatment system is shown in Rgure 3.
The landfill gas pretreatment system, designed by Bio-Gas Development, Inc., is
optimized to handle a wide range of landfill gas contaminant levels and compositions.
This was achieved by utilizing a staged contaminant removal approach which
enhances the operation of each successive process step. This is accomplished by
removal of contaminants which adversely affect downstream processes and
controlling the temperature of each step to optimize its efficiency. Figure 3
summarizes the staged contaminant removal processes which produce clean landfill
gas to meet the fuel cell specification.
Mt "W
Rgure 3. Block Diagram of the Landfill Gas Staged Contaminant Removal System
4-64
-------
The landfill gas is first cooled to approximately 1°C to remove water and heavy
hydrocarbon contaminants. Following further dehydration in absorption media, the
landfill gas temperature is lowered to optimize the operating temperature of the
downstream activated carbpn/NMOC removal and molecular sieve/HaS removal
steps. We anticipate that low molecular weight NMOCs will be removed in the second
cooler by non-steady state adsorption in the liquid film on the heat exchanger tubes, a
concept proprietary to Bio-Gas Development, Inc. The final stage of NMOC removal is
by absorption in activated carbon bed media. Hydrogen suffide is selectively
absorbed on a molecular sieve media placed downstream of the activated carbon in
the same vessel.
The presence of CC>2 in the gas produces competition for active adsorption
sites in the H2S mole sieve and can potentially produce carbonyl sulfide (COS) by
reaction with HZS. This competition is minimized by operating at low pressure which
, favors H2S adsorption over COg adsorption. The production of COS is minimized by
maintaining low bed temperatures to slow the kinetics for this reaction. The specific
adsorption media was selected to maximize H2S removal.
Lastly a paniculate filter will remove fines which may be produced from
successive thermal regeneration cycling of the adsorption beds. The cleaned landfill
gas is delivered to the fuel cell. * <--
Approximately 15% of the cleaned gas is used to regenerate the absorption
beds. This gas. containing a high concentration of desorbed contaminants, is flared
to achieve 98% destruction of NMOCs.
CONCLUSIONS
A demonstration project design was established which addresses the key
technical issues facing commercial application of the fuel cell methane control and
energy recovery concept to the market. A site was selected (Penrose Power Station)
which represents the landfill gas market. A gas pretreatment system has been
designed, and construction of the system is underway. No technical "show stoppers"
are apparent, but the success of the project clearly will be determined by the
effectiveness of the landfill gas pretreatment system to remove critical fuel cell
catalyst poisons.. These critical tests, which will commence soon, will ultimately decide
the fate of fuel ceil application on landfill gas.
REFE3BSCES
1. US
Cf
CFR
3°' 1"1' Part IH Er°ental Protection Agency
51, 52 and 60; Standards of Performance for New Stationary^
4-65
-------
Sources and Guidelines for Control of Existing Sources; Municipal Solid Waste
Landfills; Proposed Rule, Guideline and Notice of Public Hearing.
2. Sandelli, G J., Demonstration of Fuel Cells to Recover Energy From Landfill Gas
(Phase I Final Report: Conceptual Study), EPA-600/R-92-007 (NTIS PB92-
137520), January 1992.
3. Air Emissions from Municipal Solid Waste Landfills * Background Information for
Proposed Standards and Guidelines, EPA-450/3-90-011a (NTIS PB91-197061).
March 1991, p. 3-30.
4. Maxwell, Greg, "Will Gas-To-Energy Work at Your Landfill?," Solid Waste &
Power, June 1990, p.44.
4-66
-------
-------
Paper 4-H
Ozone and Global Warming
by
Robert P. Hangebrauck, Air and Energy Engineering Research Laboratory
John W. Spence, Atmospheric Research and Exposure Assessment Laboratory .
ABSTRACT
Changes in several trace substances in the Earth's atmosphere are affecting global radiative forcing. Those
substances which seem to be in the greatest state of change now and projected into the future are carbon dioxide,
ozone (and its precursors and depleters), and aerosols. It is conceivable that countervailing changes in the
radiative forcing effects of these substances, especially ozone andaerosols, may be temporarily hiding or at least
changing the "greenhouse signal" an unfortunate circumstance, particularly if the overall impacts that will
eventually occur are unpredictable or difficult to reverse quickly. If in tact the greenhouse signal is partially being
obscured at present, there is also potential forthis effect becoming less significant in the decades ahead because
of 1) a continuation of increases in greenhouse gas emissions, 2) saturation of the troposphere aerosol effect
plus controls on sulfur emissions, and 3) increasing tropospheric ozone. The substantial complexities in factors
affecting ozone and aerosols are discussed with emphasis on ozone and its precursors, including methane,
nonmeihane hydrocarbons, carbon monoxide, and nitrogen oxides. Quantifying radiative forcing is of substantial
importance. EPA is undertaking research to enhance the ability to estimate indirect factors contributing to
forcing, including measures such Global Warming Potentials. Many of the important but difficult factors to resolve
are of the indirect type. A number of potential indirect forcing effects are identified along with an estimate of
direction (sign).
Introduction
Several trace substances affect global radiative forcing levels (changes in net downward flux of energy at the
tropopause). Some of the major ones are water vapor, carbon dioxide (CO2), ozone (O3), methane (CH4),
chlorofluorocarbons (CFCs), nitrous oxide (N2O), and aerosols. The continuing increase in the concentration of
these constituents has created considerable concern among scientists regarding the potential for climate change.
The observed increases are believed to have begun near the turn of the century with the continuing Industrial
Revolution and rapidly increasing technology. This period is marked with the beginning of an increasing energy
demand and conversion of forests to agricultural land to accommodate increasing world population. Radiative
.forcing for individual trace gases and aerosols varies greatly, and depends on their concentrations. On the other
hand these constituents are controlled by various factors in addition to emissions, including mass transfer,
chemical interactions, and atmospheric lifetime. Ozone is a major factor in global radiative forcing but is not
well quantified because its high temporal and spatial variability make the quantification difficult. Ozone levels
in the troposphere depend on both transport of ozone from the stratosphere and on local chemistry. Production
of tropospheric ozone is dependent on concentrations of precursors (substances which produce ozone), including
CH4, carbon monoxide (CO), nonmethane hydrocarbons (NMHCs), and nitrogen oxides (NO,). Ozone levels in
the stratosphere now depend on concentrations of ozone-depleting substances, including CFCs, halons, other
halogenated organics, and N2O. Temperature, amount of sunlight, global transport, amount of aerosols present,
and other factors are also of importance. On the other hand, methane and NMHCs inactivate 0,-depleting
chlorine in the stratosphere. Radiative forcing for ozone is a strong function of altitude. Because of the
importance of ozone and its precursors as radiative and photochemical trace gases, EPA has accelerated its
research on global tropospheric ozone. EPA has had ongoing research on boundary-layer ozone for many years
in support of the ambient air standards dealing with the adverse effects of ozone on human health. Ozone is also
a toxicant to young trees and leafy crops (Reich, 1987; Heck, 1982). '
Understanding of trace gas effects on global ozone and radiative forcing continues to emerge. For example, in
the stratosphere newly discovered heterogeneous reactions have been found to promote the formation of the
ozone hole in Antarctica and generally appear to promote the loss of ozone in the lower stratosphere. The loss
of ozone in the lower stratosphere has had an important impact on radiative forcing (Ramaswamy, 1992).
Similarly, initial attempts have been made recently to quantify the effects of tropospheric ozone on radiative
forcing (IPCC, 1990). This paper describes how ozone fits into the global picture and covers some of EPA's related
EPA August, 1992
4-85
Pagel
-------
1870
1890
1910 1930
YEAR
1950
1970
1990
Figure 1. Global-mean combined land-air and sea-surface
temperatures. 1961-1989, relative to the average for 1951-80.
(IPCC. 1990; Reproduced with permission.)
research.
Obscuration of the Greenhouse Effect?
Global surface temperatures have risen some-
where near 0.3 to 0.6 *C over the last century
(IPCC, 1990). Figure 1 gives the. global-mean
combined land-air and sea-surface tempera-
tures for the period 1861-1989 based on histori-
cal records and various adjustments (IPCC,
1990). In addition to these past records, model
predictions ofwarmingbased on growth in green-
house gas loadings in the atmosphere imply
substantial future warmingdPCC, 1992). How-
ever, some - e.g., Michaels (1991) - have expressed skepticism about the "popular vision" of global warmin
Example concerns include 1) adequacy of correction for the urban heat island effect, 2) less than expect*
wanning, relative to that predicted by modeling, and 3) the nature of how the warming has (or has not), take
place - such as, little increase in extreme high temperatures in the northern latitudes. However, it is likely tha
while the Earth has seen some global warming, some potential warminghas been hidden by temporary offsettin
effects of other anthropogenic global changes. Changing ozone and aerosol levels over the last several decade
may have produced some net negative radiative forcing. This negative radiative forcing, along with a sligh
decrease in solar irradiance, may have partially hidden the greenhouse signal predicted by general circulatioi
models (GCMs), especially in the Northern Hemisphere. Gl obal changes other than wanning may have resulted
The changes taking place are difficult to predict as are any future consequences.
The effects of observed trends in atmospheric ozone on climate, while difficult to quantify, have been established
directionally (Lads, 1990; Ramaswamy, 1992). Radiative forcing from loss of ozone in the lower stratosphere
implies a cooling effect on the Earth's surface and in the stratosphere. Effects from an increase in tropospheric
ozone implies higher temperatures at the Earth's surface and in the free troposphere. More indirectly, increased
ozone in the atmospheric boundary layer reduces tree growth resulting in reduced carbon sequestration (Reich,
1987). -..'.-
Estimates of the negative radiative forcing effects of aerosols range up to 2 watts(W)/m2 for tropospheric sulfate
(Charlson, 1992). Penner (1991b) estimates from modeling work that aerosols from biomass burning could
contribute a radiative forcing mask of-1.8 W/m* with aerosols from fossil fuel contributing even a larger negative
forcing. Kaufman (1991) concludes (with substantial uncertainties noted) that increased future sulfur dioxide
(SO2) emissions from fossil fuel combustion will likely produce a cooling effect from increases in cloud
condensation nuclei. On the other hand biomass burning is seen overall to have a net warming effect, because
significant ozone precursors are injected into the atmosphere in addition to aerosols (Kaufman, 1991).
Stratospheric aerosols from the eruption of Mt Pinatubo in 1991 should produce a very significant negative
forcing on the climate system (UNEP/WMO, 1991). Hansen (1992) estimates that the global mean climate
forcing due to Pinatubo will peak at about 4 W/m2 in early 1992, noting that this exceeds the accumulated forcing
due to all anthropogenic greenhouse gases added to the atmosphere since the industrial revolution. Global
climate changes that might be expected from increased tropospheric sulfate concentrations include the following
(Michaels, 1991):
* Increased cloudiness - ... .
* Enhanced brightening of low-level clouds near sulfate source regions
* A counteraction of daytime warming by the greenhouse forcing because of an increase in clouds
» Night warming from an increase in both clouds (especially stratocumulus) and greenhouse forcing
* A decrease in daily temperature range
Decrease in ultraviolet (UV)-B in affected areas
* Concentration of these effects in the industrial Northern Hemisphere
The connection between aerosols and all of the above effects is by no means proven. Other factors must be
EPA August, 1992
4-86
Page 2
-------
involved in reducing the predicted increases in the daily maximum temperature in a large portion of the Northern
Hemisphere (Karl, 1991). Some of these other factors could include changes in cloudiness and aerosol loadings,
but could also just be natural fluctuations of the climate system. These investigators were examining the nature
of and evidence for asymmetric diurnal temperature change in large areas of the Northern Hemisphere. These
are areas where, over the last several decades, most of the warming can be attributed to an increase of mean
minimum (mostly nighttime) temperatures with mean maximum (mostly daytime), temperatures displaying
little or no wanning. These are also areas where there has been increasing extreme minimum temperatures but
little change in extreme maximum temperatures, leading to a decrease in the extreme temperature range.
Reduced Obscuration of the Greenhouse Effect?
Why should all this cause concern? Many climate scientists think that serious greenhouse effects are possible,
and that lack of a strong greenhouse signal may lead to inaction in dealing with emissions, particularly C02
because of its long atmospheric lifetime. Furthermore, significant atmosphere-related global change may have
already taken place with unknown long-term consequences. In addition, if we are having substantial current
obscuration of the greenhouse signal, this obscuration may become less significant in the decades ahead because
of 1) a continuation of increases in greenhouse gas emissions, 2) saturation of the troposphenc aerosol effect, and
3) increasing troposphenc ozone.
Regarding saturation of the aerosol effect Kaufman
(1991) sees that, for a large increase in fossil fuel use,
emissions of S02 may saturate the cloud condensation
nuclei effect. For a doubling of C02, the aerosol radia-
tive forcing effect is estimated to be only 0.1 to 0.3 of
that for COj compared to 0.4 to 8 tir-es at current
emission levels.
\
A
PROFLEF
ozc
OR JJULY
ICAt
AT dO' N
Relative to troposphenc ozone, ozone is increasing on a
global basis with the rate of increase depending on
location and altitude. CH4, NMHCs, CO, and NO, are
all troposphenc ozone precursors, and their global
emissions are all increasing. In addition to surface-
based emission sources, emissions from aircraft at
cruise altitude seem to affect the ozone profile and are
similarly increasing (Wuebbles, 1990; Barrett, 1991).
How is O, Formed and Destroyed?
0 ion 2000 3000 woo sooo ana ran MOD
Volume Mixing Ratio (ppbv)
Figure 2. Ozone altitudinal profile (data derived from chart
on p. 422 of Vol. II of Atmospheric Ozone 1985 (WMO,
1985>- As the solar ultraviolet radiation enters the Earth's
atmosphere, the UV-C component is of sufficient energy to photodissotiate oxygen (O2), resulting in the formation
of ozone (O3) in the upper stratosphere and UV-C absorption. UV-B radiation, on the other hand, penetrates to
lower altitudes and has enough energy to dissociate ozone (UV-B is absorbed in this process). Figure 2 illustrates
a typical ozone concentration profile for July at 60° N. This particular profile was derived from data which did
not include boundary layer measurements, and therefore does not show a typical increase near the Earth's
surface. The profile reveals a maximum mixing ratio near 30 km for this latitude. The ozone essentially filters
out most of the short wavelength ultraviolet radiation from the Sun. Figure 3 gives a simplified overall view of
the formation/destruction of ozone in the atmosphere. The destruction of ozone proceeds until these halogen
atoms are chemically bound in stable reservoirforms such ashydrochloricaddCHCDandhydrobronucatidCHBr).
In general, atmospheric models utilizing homogeneous gas-phase chemistry provided a reasonable scientific tool
that explained the catalytic destruction of the ozone layer. With the formation of the ozone hole over the
Antarctic, atmospheric scientists began to investigate heterogeneous chemical mechanisms to explain the rapid
loss of ozone during the September-October time period in the Southern Hemisphere. The most critical aspect
of the heterogeneous mechanism is the destruction of active chlorine sinks by reactivating HC1 and chlorine
nitrate (C1ONO2) and at the same time removing NO, from the stratosphere. Lower levels of NO, limit the degree
to which chlorine can be tied up as C10N02. The sulfuric acid aerosols in the low to middle latitudes of the lower
EPA August. 1992
4-87
Page3
-------
stratosphere of the Northern and Southern Hemi-
spheres play a role similar to that of the polar
stratospheric ice crystals in contributing to ozone
depletion in the polar areas, especially Antarctica.
In the boundary layer ozone is formed photochemi-
cally from other gases (precursors). Within'urban-
suburban areas, ozone formation is driven by the
emissions of the precursors NMHCs and NOit in the
presence of sunlight. The formation of ozone in
remote areas of the boundary layer, in the free
troposphere, and in the lower stratosphere involves
the less reactive hydrocarbons such as methane.
The production of ozone is known to be complex and
is not linearly proportional to increases in precursor
emissions. Atmospheric modeling as well as photo-
chemical simulations indicate that the formation of
ozone may indeed be NOs-limited over certain por-
tions of the globe. These studies indicated that
How is O3 formed and destroyed'
Formation: """ ~ ' Daarudion: rr7!J~"
02 + UV-
02 + O-
O3
Stratosphere
CUO3>OO + O2
CIO+O->CJ*O2
(Net)0+03>202
H«»reg»n«eui
CONOZV
HO j
rw. fci
"T-*") ci
*«* 1 HI
W*i^
Troposphere
Formation (high NOx):
(Net reaction)
Oestruaion (NOx < 20 pptv):
(Nel reaction)
CH4 + 4O2
CO + 2O2-
~>CH2O + H2O+203
-> CO2 + O3
->CH20 + H2O
CO + O3 >C02+O2
. 1. A I.I _ 1_
Boundary Layer
Figure 3. Overview of ozone formation/destruction in the
atmosphere.
concentrations of NO, above a range of 10 to 25 pptv are required for ozone production, while lower concentratic
would lead to destruction (Hameed, 1979; L
1988). This implies that there are areas within t
boundary layer as well as free troposphere whi
the photochemistry is forming ozone and ott
areas where photochemistry is destroying ozoi
Scientists at Lawrence Livermore National Lai
ratory (LLNL) are developing a 3-D chemist
model to predict global atmospheric loadings a
distributions of NO, including nitric acid fr<
anthropogenic and biogenic sources (Penner, 199
Dignon, 1992). A global map of spatially allocal
NO, emissions in Figure 4 illustrates a typi
input used in the model. The model also predi
global wet and dry deposition of nitrogen sped
Figure 4. Spatially allocated global NO. emissions from Developing the capability to model NOt from
anthropogenic and natural sources in 1980.
sources is the first step in developing a 3-D Global Chemical
Model for predicting atmospheric loadings of ozone and
other greenhouse gases from emissions. Ozone is also
destroyed by surfaces, by reactions with olefmic gases, and
by photodissociation that leads to production of hydroxyl
(OH) radicals. The oxidizing power of the Earth's atmo-
sphere is controlled by the abundance of these radicals. It is
estimated that OH radicals account for about 85% of the loss
of methane (Cicerone, 1988). Some of the directional effects
of precursor emissions on ozone-related atmospheric chem-
istry are indicated in Figure 5 as derived from UNEP/WMO
(1991). For example, increasing the precursor CH4 will
decrease OH, increase O3, and increase the lifetime of
various organics such as HCFCs.
Trends in Ozone
Increased
emission
CH4
NOx
CO
NMHC
OH
-
+
-
O3
+
+
+
+
Lifetime (CH4,
HCFC, HFC)
+
-
+
+
Figure 5. Precursor influences on atmospheric
chemistry (derived from UNEP/WMO, 1991).
Figure 6 illustrates the change in total column ozone in the period 1979 - 1991 from data reported by UNE
EPA August, 1992
4*8
Page
-------
Miy-Jun Stp-Nov
Rgure 6. 1979-1991 total ozone column change
(TOMS satellite data.)
WMO (1991) for the Total Ozone Mapping Spectrometer
(TOMS) satellite. While this show substantial depletions,
more recent data reported by NASA (1992) show much greater
depletions, including in the tropics. In addition, measure-
ments show record concentrations of CIO of 1.5 ppbv over
northern New England in January 1992 as part of the Air-
borne Arctic Stratospheric Expedition (NASA, 1992). Abun-
dance of chlorine monoxide (CIO) in the lower stratosphere at
northern middle-latitudes is greater than predicted by mod-
els containing only gas-phase chemistry (UNEP/WMO, 1991).
o :
(
\
X
/
X
\
\
V
/
/
/
^
\
\
>
1
/
V
>,
-1.0
as
0.0
as
1.0
Rgure 7. Ozone attitudinal trend estimates for the
Northern'Hemisphere (1970 -1986).
Information on the vertical atmospheric distribution of ozone can be derived from ozonesonde, Umkehr,
Stratospheric Aerosol and Gas Experiment (SAGE), and Upper Atmosphere Research Satellite (UARS) data.
Unfortunately, most of the long-term measurements, particularly for the ozonesondes, are within the northern
middle-latitudes. This hinders the development of reliable global vertical distributions for the Southern
Hemisphere. There are concerns regarding calibration techniques for the long-term data from many of the
northern ozonesonde stations. Nevertheless, the average percent per year change of ozone from 1970 to 1986
as a function of altitude for nine northern stations is shown in Figure 7 (UNEP/WMO, 1989). This historical
ozonesonde record reveals that ozone appears to be increasing within the free troposphere and decreasing within
the lower stratosphere. However, variations at individual stations show widely varying profile changes for
different latitudes. Tropospheric ozone trends derived from ozonesonde data have accuracy limitations, but
could be improved by expansion of monitoring capability (Prinn, 1988). Homogeneous chemistry 1-D models
predict the ozone increases within the free troposphere but fail to predict the decreased ozone observed within
the lower stratosphere. The change in the slope of die ozone profile that is observed in the troposphere below the
tropopause may be due to-the injection of ozone precursors by aircraft (Wuebbles, 1990; Kinnison, 1991).
Tropospheric ozone trends continue to be measured at ozonesonde stations; however, measurements in the upper
troposphere are sparse, and this is where ozone has its greatest radiative forcing effect
Satellite data from the TOMS and the SAGE have been recently used to derive global maps of ozone within the
troposphere and boundary layer (Fishman, 1991). Tropospheric ozone is derived as the residual or difference
between the coincident TOMS and SAGE measurements between 50° S and 50° N. The residual, which is a
relatively small difference between two larger values, represents the ozone column in Dobson units (ozone
molecules per cm2) within the troposphere and boundary layer. Over a 10-year period the averaged seasonal
depictions show the residual ozone to begin its formation in the Northern Hemisphere during March-May,
reaching a maximum during June-August In the Southern Hemisphere, ozone forms during the September-
November season. What is so surprising is the size of the ozone plumes. The entire Northern Hemisphere is
engulfed in an ozone plume that spans the Atlantic and much of the Pacific Ocean for the June-August season.
EPA August, 1992
4-89
PageS
-------
Figure 8. Satellite-derived tropospheric ozone residuals
centered off the west coast of Africa.
In the Southern Hemisphere the ozone plume forms across lower Africa and tails across to Australia. It spans
the Atlantic Ocean to South America. The formation of this plume appears to be associated with the biomass
burning that occurs annually during the spring in the Southern Hemisphere. The formation of the observed
ozone in the Northern Hemisphere is consistent with the photochemical production of ozone from its precursors
during the spring and summer months.
Atmospheric scientists at NASA Langley Research Center are using the TOMS and SAGE data to derive daily
residual maps for global tropospheric ozone for EPA. Since SAGE provides less than 5000 measurements per
year, daily measurements are linearly interpolated from 60 day averages. The daily residual is obtained by
subtracting the interpolated SAGE daily measurements from 3-day averages of the TOMS measurements and
has a resolution of 2.5° latitude by 5.0 e longitude. Computerized images of the daily residual from 1985 to 1990
are prepared by the EPA's Scientific Visualization Laboratory at RTP. A 4-day sequence for September-October
1988 is shown in Figure 8 that is believed to represent the formation of tropospheric ozone from the biomass burn
in the springtime in the Southern Hemisphere. Scientists at NASA Langley Research Center are using the video
computerized images to assist in the planning of the TRACER A Monitoring Program of the Biomass Bum in
1992. Scientists at EPA are comparing the satellite-derived daily residual Oa data for the Northern Hemisphere
with ozone ground-based (AIRES), ozonesonde, and meteorological measurements. If the comparative analysis
shows promising results, the satellite-derived ozone residuals will provide insight into the formation, transport,
and fate of global tropospheric ozone, and may serve as a useful tool in the development of global chemical models
that predict ozone concentrations from precursoremissions. EPA and NASA scientists are planning to develop
another sequence of daily residuals using measurements from the Solar Backscatter Ultraviolet Spectrometer
(SBUV)thatflies on the Nimbus 7 satellite with the TOMS. Comparative analysis of the two daily residual data
bases should provide insight into the techniques for deriving global tropospheric ozone from satellite data.
EPA August, 1992
4-90
-------
Trends in Aerosols/Sulfates
It has been concluded that fossil fuel emissions over the past century have increased the tropospheric sulfate
aerosol concentrations (UNEP/WMO, 1991). World total sulfur emissions are estimated to be 147 TgS with about
80 TgS coming mainly from fossil fuel combustion (IPCC, 1990). Husar (1989) shows a general correlation
between sulfur emissions and extinction coefficient which is a function of visibility. Mayewski (1990) has
determined, based on ice core data, that the anthropogenic sulfate loadings in remote areas of the Northern
Hemisphere are now as high as or higher than the maximum loading from many past volcanic.eruptions.
Carbonyl sulfide (COS) is the most abundant sulfur-containing trace gas in the remote atmosphere and, with a
lifetime of 2 to 6 years, is, via photolysis, a major source of aerosol sulfate in the stratosphere along with volcanic
eruptions (UNEP/WMO, 1991). Sources of COS include anthropogenic activities, soils, biomass burning, and
oxidation of carbon disulfide (UNEP/WMO, 1991). Long-term observational records show a 40-50% increase in.,
stratospheric sulfate aerosols (UNEP/WMO, 1991).
IPCC Trace Gas Projections]
Trends in Ozone Precursors
Source emissions of CH4, NO,, CO, and NMHCs
have all been, increasing. Figure 9 shows the
IPCC predictions of future ozone precursor emis-
sions. Atmospheric concentrations/distributions
have been quantified for CH4 and CO, but are
harder to do for the more reactive species, NO,
and NMHCs.
The atmospheric concentration of CH4 has been
increasing at a rate of about 1% per year up to the
last few years. Currently this growth rate ap-
pears to have fallen sharply with no completely
satisfactory explanation. However, there is a
possibility that OH is increasing at 1.0±0.8% per
year (UNEP/WMO, 1991). The most recent data
for methane, as taken from Khalil (1990), are given in Figure 10. Unfortunately, the data after 1988 are still not
available. The data are undergoing quality assurance review, but should be available this summer in the journal
Chemosphere and a new book entitled The Global
. Methane Cycle: Its Sources, Sinks, Distributions and
Role in Global Climate Change. Figure 11 represents 5
years of methane measurements from remote marine
sites within the Geophysical Climatfe Change Sampling
Network. The averaged atmospheric concentration of
Figure 9. Projected ozone precursor emissions.
C1700
H
41680
1660
p1640~
b
V1620-
1600
1S80-
1560-
1S40
i I
1980 1982
YEAR
1984
Ii I
1986 1988
Figure 10. Atmospheric methane concentration.
Figure 11. Hemispheric methane cycles with time.
(IPCC. 1990: Reproduced with permission.)
EPA August, 1992
4-91
Page?
-------
methane is about 1.76 ppmv in the Northern Hemisphere and about 1.68 ppmv in the Southern Hemisphere. Tl
seasonality is observed to vary with latitude and is repeatable over time in the hemispheres (IPCC, 1990). Noi
the increase in methane concentration right up to the arctic polar area. The primary emission sources of methar
include: tundra/bogs/swamps, rice cultivation, biomass burning, livestock, coal mining, oil, and natural ga
systems, landfills, and a variety of industrial processes (such as coke and petrochemicals production). Methan
isotopic studies suggest that 20% of global methane is from fossil fuel and 10% is from biomass burning (UNEP
WMO, 1991).
The atmospheric concentration of carbon monoxide is increasing at about 1% per year in the Northerr
Hemisphere but is not increasing perceptibly in the Southern Hemisphere. The concentration is about 120 ppta
in the Northern Hemisphere and about 60 ppbv in the Southern Hemisphere for a hemispheric ratio of about 2
(Wuebbles, 1991). The primary sources of atmospheric CO are atmospheric oxidation of methane and
nonmethane hydrocarbons, biomass combustion-related sources such as forest clearing, and fossil-fuel combus-
tion (primarily transportation related).
Measurements of NMHCs are not adequate to establish trends
with the exception of ethane (UNEP/WMO, 1991). Ehhalt
(1991) has provided evidence of a trend for ethane over the
Northern Hemisphere of 0.9±0.3% per year. Perhaps the best
indicator of trend for NMHCs is the projected future emissions
(see Figure 9). The primary sources of atmospheric NMHCs are
biogenic in nature, including trees, oceans, and grasslands
(UNEP/WMO, 1991). Anthropogenic sources include gasoline
use, other petroleum-based solvents/chemicals, fuel wood use,
biomass burning, and waste disposal (Watson, 1991). The
reactivity and general variability of NMHCs in the atmosphere
make it difficult to establish trends. NMHCs and their reaction
by-products are found throughout the atmosphere. For ex-
ample, the work of Greenberg (1990) indicates that many
NMHCs may exist throughout the troposphere. Longer photo-
chemical lifetimes, attributed to lower temperatures at high
altitudes and latitudes, suggest to the investigators "that
NMHCs may be present throughout the troposphere in many
regions of the Northern Hemisphere." The more reactive NMHCs are generally found in smaller quantities
relative to the less reactive compounds. Transport mechanisms such as cumulus clouds, dry convection, cold
fronts, and flow of air over mountains are suggested as means for the substantial altitudinal dispersal in spite
of atmospheric lifetime considerations. Figure 12 illustrates some of the data reported by Greenbergin this
case for total NMHCs at 66° N latitude. As can be seen, substantial NMHC mixing ratios are found high in the
troposphere.
Penner (199 la) and Dignon (1992) have estimated sources and distributions globally of natural and anthropo-
genic emissions of NO,. Northern Hemisphere anthropogenic sources are seen to be mainly fossil fuel
combustion, and Southern Hemisphere anthropogenic contributions are seen to be a combination of fossil fuel
emissions and biomass burning. Fossil fuel combustion, which consists of both stationary and mobile sources,
is believed to account for about 50% of the estimated total emissions of global N0a. Ifbiomass burningis included,
the total anthropogenic contributions become more like 75%. However, there is considerably more uncertainty
in the emissions from biomass burning than from fossil fuel combustion in stationary sources. The NASA project
(TRACER A) to characterize the 1992 biomass burn in the Southern Hemisphere should provide better
quantification of the emissions. Natural sources include lightning, soil microbial activity, and input into the free
troposphere from the photodissociation of N20 in the stratosphere. Large uncertainties are also associated with
the emissions from the natural sources. Sufficient atmospheric measurements of NO, are not available to project
global concentration trends. The best indication of NO, concentration trends is from nitrate (NO,) measurements
taken from ice cores. Analyses for NO, in ice cores from Greenland and Switzerland show large increases in NO,
since the turn of the century (Neftel, 1985; Wagenbach, 1988).
O 1O 2O 3O 4O SO 6O 7O SO 9O
NHMC Mixing Ratio, pptv
figure 12. A total NMHCs profile for 66 N
latitude from Greenberg (1990).
EPA August, 1992
4-92
PageS
-------
Radiative Forcing
Determining the radiative forcing caused by various trace substances is obviously difficult, with major questions
still to be answered. Multiple effects for the same substance are possible in different parts of the atmosphere.
Many of the complex questions arise from the indirect effects of trace substances. The recent IPCC (1990) report
provided initial quantification of some of the indirect effects in terms of Global Warming Potentials (GWPs),
especially for ozone and its precursors. But in later UNEP/WMO updates of the IPCC report, reservations have
been expressed about the accuracy of the initial indirect GWPs. Also, the rather large negative-forcing indirect
effect was identified for CFC-related ozone depletion in the lower stratosphere . GWPs are evolving with
refinements taking into account such factors as sensitivity of the GWPs to the assumed background atmosphere
(Wuebbles, 1992). Many of the indirect radiative forcing effects, for which indirect GWPs would be generated,
require modeling to provide quantification, but for the most part the tropospheric models necessary to do the
evaluation have not been fully developed and validated. EPA is collaborating with the Lawrence Livermore
National Laboratory to identify, quantify, and refine potential indirect forcing effects for the various trace
substances, especially as they relate to tropospheric ozone and its precursors.
As a greenhouse gas, the radiative properties or forcing for a given quantity of ozone is a function of altitude
(Lacis, 1990). The radiative forcing of ozone has little affect within the boundary layer but maximizes within the
free troposphere at the tropopause. In the lower stratosphere, the forcing is positive but diminishes with altitude.
It is also a function of latitude. Hansen (1991) recently reported (at the 1991 AGU meeting in San Francisco)
the impacts of atmospheric ozone on radiative forcing using the GISS GCM . When ozone between 10-20 km
within the lower stratosphere was removed, the upper troposphere was found to cool by several degrees Celsius
and the Earth's surface temperature cooled by 1-2° C. This simulation provides insight into the radiative effects
of ozone removal that is believed to be now occurring by heterogeneous reactions in the lower stratosphere. When
ozone above 35 km was removed in the model simulations, the stratosphere cooled, and a warming of about 2 "
C at the surface was observed. The wanning of the surface in this case can be associated with an increase in
radiant energy transmitted through the atmosphere. Increasing ozone in the free troposphere by a factor of 10
resulted in surface warming as a function of latitude. While the GISS GCM simulations point out the radiative
importance of atmospheric ozone, the simulations also indicate a need for better ozone measurements in the
troposphere and stratosphere to establish ozone profile trends. EPA scientists are also initiating studies to
parameterize atmospheric transport processes between the boundary layer and free troposphere, such as cloud
venting, for incorporation into global chemistry models. These processes inject ozone and its precursors into the
troposphere. The atmospheric lifetime of ozone increases significantly in the troposphere where the temperature
decreases with altitude up to the tropopause. These studies are intended to enhance model determination of
global radiative influences due to changes in tropospheric ozone.
Table 1 illustrates some of the direct and indirect effects related to radiative forcing and their likely directional
effects. This table is not provided as an accurate, comprehensive compilation of all potential important effects,
but does attempt to show that there are many effects. Almost all the effects from the inorganic gases covered
in Table 1 are directionally positive. Water vapor has a positive forcing effect in both the troposphere and
stratosphere; however, there are also feedback effects related to water including clouds, snow cover, sea ice cover
and precipitation. Hydrogen is also increasing in the atmosphere (Khalil, 1990) and can contribute water vapor
to the stratosphere causing an indirect positive effect CO has positive indirect effects associated with 1)
production of ozone in the troposphere and 2) additional CO, as the final product of oxidation. Nitrous oxide is
a direct acting greenhouse gas and might also cause positive forcing via its depletion of upper stratospheric ozone.
NOS have a positive indirect forcing via production of tropospheric ozone. Radiative forcing of tropospheric ozone
may be especially sensitive to N0t emitted by aircraft at cruise altitude (Wuebbles, 1990; Kinnison, 1991;
Johnson, 1992). NO, may also have a negative forcing effect in the stratosphere via stratospheric ozone depletion.
All of the effects identified for methane in Table 1 are positive. It has a direct effect, and it produces positive
effects from ozone production in both the troposphere and the stratosphere. It has a positive effect via
deactivation of active stratospheric halogens which destroy stratospheric ozone. It produces stratospheric water
vapor, and it ultimately ends up as CO,, a greenhouse gas. All of the effects listed for NMHCs in Table 1 are also
positive, NMHCs produce positive forcing from ozone formation in the troposphere and stratosphere. Longer-
lived NMHCs that make it to the stratosphere would deactivate active CI, and Brz, thus reducing stratospheric
ozone depletion. . NMHCs, like CO and methane, will also eventually oxidize to the greenhouse gas, CO2.
EPA August, 1992
4-93
PageS
-------
Table 1. Partial listing of direct and indirect
effects related to radiative forcing.
BIO
cot
CO
N2O
NO.
mmtr* tMn*ttv
H2OTnp. (direct «)
HZCVStrai. Idiraei *)
IWStnt. H20 (indirect»>
CO2 (direct «>
COl'StrmL O3 incnaa* "a nni. cooling (direct »)
CO/Trep. 03 (indirect »>
COCO3 (indirect*)
N2O (diraet*)
NOVStrtt. O3 depletion (indirect *)
NO* aircnnVUppar Irap. O3 (indirect »)
NOz/Str»t O3 reduction (indirect -»
Orv»ajic*>
Halogenated organics would appear to cause both
positive and negative forcing. CFCs such as CFC-11
and CFC-12 are strong absorbers causing direct green-
house forcing in the troposphere. On the other hand
such long-lived halogenated organics will cause strato-
spheric ozone depletion, thereby causing negative forc-
ing. Aerosols constitute another set of forcing effects
that are not well quantified and complex, but it would
appear that their effects are mostly negative. Aero-
sols, such as sulfuric acid, can cause negative forcing
through reflective scattering of solar radiation. They
can also enhance cloud formation and reflectivity, in
the troposphere. Aerosols containing carbon black, on
the other hand, may induce positive forcing via ab-
sorption of radiant energy. Finally, sulfuric acid
aerosols in the lower stratosphere probably enhance
ozone depletion via heterogeneous catalysis and de-
struction of sink species.
Conclusions
Changes in several trace substances in the Earth's
atmosphere are affecting global radiative forcing. Those
substances which seem have the largest changes oc-
curring now and projected into the future are C02,
ozone (and its precursors and depleters), and aerosols.
It is conceivable that countervailing changes in the
radiative forcing effects of these substances, especially
ozone and aerosols, may be temporarily hiding or at
least changing the "greenhouse signal" If in fact the
greenhouse signal is being partially obscured at
present, there is also potential for this effect's becom-
ing less significant in the decades ahead because of 1)
a continuation of increases in greenhouse gas emis-
sions, 2} saturation of the tropospheric aerosol effect
plus controls on SOZ emissions, and 3) increasing
tropospheric ozone. Recent findings by international
scientists working toward more accurate assessment
of future forcing make it clear that indirect effects are playing an important role and are poorly quantified at
present. The substantial complexities in factors affecting ozone and aerosols are discussed with emphasis on
ozone and its precursors, including methane, NMHCs, CO, and NO,. A substantial number of potential forcing
effects are identified along with an estimate of direction (sign). Quantifying radiative forcing and its sources are
of substantial importance for future prevention and mitigation efforts, and to this end EPAis helpingwith efforts
to enhance the ability to estimate both the direct and indirect factors contributing to forcing. Emphasis is being
placed on better estimates of current and projected emissions and measures of radiative forcing such as Global
Wanning Potentials (GWPs). Research efforts are focused on enhanced emission/mitigation data and projec-
tions, improved data on global atmospheric trends, and enhanced capabilities for theoretical prediction via
improved models. ' " " _ ,- ^ - ^~^^:~: ";'
CH4 (UP. »)
CH4 (direct*)
CHVTnp, O3 (indirect *)
CtM/Stnt. O3 (indirect *)
CHVSlrai. O3 increaaa
CHVStral, H2O rindinct »>
CH«CO2(indinci»»
NMHC> <!)*)
NMHC»Trop. O3 (indiMtt »>
Ethan*, propane, «.
NUHC^Stnt O3 (indirect«)
Ethan*, propam, ite.
NHHCafSKal. O3 inc. m» Cla/Bn dtwtintion (inUract»)
Ethan*, prapan*. «.
NlfRCaiOOt (indirect «>
CFC-U (direct.)
CFC-U (direct*)
CFC-113(dict»)
CTC-22 (direct*)
CbaVBn/LiKfW Stret. O3 depl.tion (indinet )
CbUra/Uppar Stnt. O3 dapltlion (indinet *>
KaJeo* (direct*)
CH3Br (direct *)
A«reaoU«U«UTrop. (direct )
Aaw«oWTV»p. (indirect )
Aancato (Cwtan M«ckVTrep. (direct *)
AanmMKM VStrtt O3 d-ptotion (indirect )
EPA August, 1992
Page 20
-------
References
Barrett, M., Aircraft Pollution: Environmental Impacts and Future Solutions, WWF-World Wide Fund for
Nature Research Paper, Switzerland (1991).
Charlson, R. J., Schwrtz, S. E., Hales, J. M., Cess, J. A., Coakley, J. A., Jr., Hansen, J. E., and Hoftnan, D. J.,
Climate Forcing by Anthropogenic Aerosols, Science, 255, 423-430 (1992).
Cicerone, R. J., and Oremland, R. S., Biogeochemical Aspects of Atmospheric Methane, Global Biogeochem.
Cycles, 2, 299-327 (1988).
Dignon, J., Penner, J. E., Atherton, C. S., and Walton, J. J., Atmospheric Reactive Nitrogen: A Model Study of
Natural and Anthropogenic Sources and the Role of Microbial Soil Emissions, Proceedings of CHEMRAWN VII
World Conf. on the Chemistry of the Atmosphere: Its Impact on Global Change, Baltimore, MD (1992).
Ehhalt, D., Schmidt, U., Zander, R., Demoulin, P., and Rinsland, C., Seasonal Cycle and Secular Trend of the
Total and Tropospheric Column Abundance of Ethane Above the Jungfraujoch, J. Geophys. Res, 96,4987*
4994 (1991).
Fishman, J., Probing the Planetary Pollution from Space, ES&T, 25, 613-621 (1991).
Greenberg, J. P., Zimmerman, P. R., and Haagenson, P., Tropospheric Hydrocarbon and CO Profiles, Over the
U.S. West Coast and Alaska, Journal of Geophysical Research, 95,14015-14026 (1990).
Hameed, S., Pinto, J. P., and Stewart, R. W., Sensitivity of the Predicted CO-OH-CH4 Perturbation to
Tropospheric NO, Concentrations, JGR 84(C2): 763-768 (1979). *
Hansen, Can the Climate be Engineered?, AGU Fall Meeting Invited Talk, San Francisco (January 1991).
Hansen, J., Lacis, A., Ruedy, R~, and Sato M., Potential climate impact of Mount Pinatubo eruption, Geophysical
Research Letters, 19,215-218 (1992).
Heck, W. W., Taylor, O. C., Adams, R., Bingham, G., Miller; J., Preston, E., and Weinstein, L., Assessment of Crop
Loss from Ozone, JAPCA, 32:353-361 (1982).
Husar, R. B., and Wilson, W. E., Trends of Seasonal Haziness and Sulfur Emissions Over the Eastern U.S.,
Transactions: Visibility and Fine Particles, An AWMA/EPA Specialty Conference, Air and Waste Management
Association (1989).
Intergovernmental Panel on Climate Change (IPCC), Climate Change, The IPCC Scientific Assessment, World
Meterological Organization/United Nations Environment Programme (1990).
IPCC, 1992 IPCC Supplement, Intergovernmental Panel on Climate Change, UNEP/WMO (1992).
Johnson, C., Henshaw, J., and Mclnnes, G,, Impact of Aircraft and Surface Emissions of Nitrogen Oxides on
Tropospheric Ozone and Global Warming, Nature, 355,69-71 (1992).
Karl, T. R., Kukla, G.. Razuvayev, N., Changery, M. J., Quayle, R. G.t Heim, R. R., Easterling, D. R., and Fu, C.
B., Global Wanning: Evidence for Asymmetric Diurnal Temperature Change, Geophysical Research Letters, 18,
12,2253-2256 (December 1991).
Kaufman, Y. J., Fraser, R S., and Mahoney, R. L., Fossil Fuel and Biomass Burning Effect on Climate - Heating
or Cooling?, American Meteorological Society, 4,578-588 (1991).
Khalil, M. A. K, and Rasmussen, R A., Atmospheric Methane: Recent Global Trends, Environmental Science
EPA August, 1992
4-95
Pagell
-------
and Technology, 24,549-553 (1990).
ffinnison, D. E., and Wuebbles, D. J., Future Aircraft and Potential Effects on Stratospheric Ozone and Climate,
Proceedings of the 42nd Congress of the International Astronautical Federation, Montreal, Canada (1991).
Lacis, A. A, Wuebbles, D. J.t and Logan, J. A., Radiative Forcing of Climate Changes in the Vertical Distribution
of Ozone, Journal of Geophysical Research, 95, 9971-9981 (1990).
Lin, X., Trainer, M., and Liu, S. C., On the nonlinearity of the tropospheric ozone production, JGR, 93,15879-
15888 (1988).
Mayewski, P. A., Lyons, W. B., Spencer, M. J., Twickler, M. S., Bock, C. F., and Whitlow, S., An Ice-Core Record
of Atmospheric Response to Anthropogenic Sulphate and Nitrate," Nature, 348,22 (1990).
Michaels, P. J., Journal of Coal Quality, 10,1 (March 1991).
NASA, Summary Statement: Second Airborne Arctic Stratospheric Expedition (February 3,1992).
Neftel, A., Beer, J., Oeschger, H., Zurcher, F., and Rinkel, R. C., Sulfate and Nitrate Concentrations in Snow from
South Greenland, Nature, 314, 611-613 (1985).
Penner, J. E., Atherton, C. S., Dignon, J., Ghan, S. J., and Walton, J. J., Tropospheric Nitrogen: A Three-
Dimensional Study of Sources, Distributions, and Deposition, J. Geophys. Res., 96, 959-990 (1991a).
Penner, J. E., Ghan, S. J., and Walton, J. J., The Role of Biomass Burning in the Budget and Cycle of
Carbonaceous Soot Aerosols and Their Climate Impact, Global Biomass Burning: Atmospheric, Climatic, and
Biospheric Implications, J. S. Levine, editor, MIT Press, Cambridge, MA (I991b).
Prinn, R, G., Toward an Improved Global Network for Determination of Tropospheric Ozone Climatology and
Trends, J. Atmospheric Chemistry, 6,281-98 (1988). *
Ramaswamy, V., Schwarzkopf, M. D., and Shine, K. P., Radiative Forcing of Climate from Halocarbon-Induced
Global Stratospheric Ozone Loss, Nature, 355, 810-812 (1992).
Reich, P. B., Quantifying Plant Response to Ozone: a Unifying Theory, Tree Physiology, 3, 63-91 (1987).
UNEP/WMO, Scientific Assessment of Stratospheric Ozone Depletion: 1991 (1991).
UNEP/WMO, Scientific Assessment of Stratospheric Ozone, (1989).
Wagenbach, D., Munnich, K, Schotterer, U., and Oeschger, H., Annals of Glaciology, 10,182-187 (1988).
Watson, J. J., Probert, J. A, and Piccot, S. D., Global Inventory of Volatile Organic Compound Emissions from
Anthropogenic Sources, EPA-600/8-91-002 (NTIS PB91-161687) (January 1991).
i
WHO, Atmospheric Ozone 1985, Volume II, Report No. 16 (1985).
Wuebbles, D. J., and Patten, K. 0., Sensitivity of Global Warming Potentials to the Assumed Background
Atmosphere, Lawrence Livermore National Laboratory, UCRL-ID-109847 (1992).
Wuebbles, D. J. and Edmonds, J., Primer on Greenhouse Gases, Lewis Publishers (1991).
Wuebbles, D. J., and Kinnison, D. E., Sensitivity of Stratospheric Ozone to Present and Possible Future Aircraft
Emissions, Air Traffic and the Environment, U. Schumann, editor, Springs-Verlag (1990).
EPA August, 1992
4-96
Page 12
-------
HOUSEHOLD FUELS IN DEVELOPING COUNTRIES:
GLOBAL WARMING, HEALTH, AND ENERGY IMPLICATIONS
' - - by
Kirk R. Smith
Program on Environment
East-West Center, Honolulu, ffl 96848
and
Susan A. Thomeloe
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
Although individually small, the widespread and daily use of household stoves
with poor combustion efficiency in developing countries raises questions about possible
global wanning and other environmental implications of their airborne emissions. To
explore the possible utility of efforts to measure the emissions from representative
samples of these devices, a small pilot study of greenhouse gas emissions of biomass and
fossil-fuel stoves was undertaken in Manila 1 (Smith et al., 1992a&b). The results,
although based on only a few measurements, indicate that such stoves may have a
significant role in global greenhouse gas inventories; be subject to substantial
improvement through alternative technologies; and that policy measures should
consider energy and health implications as well. As a consequence, a larger set of
studies is being planned for India, China, Thailand, and Brazil.
This research is funded through EPA's Global Climate Change Research
Program. Research on emissions and mitigation of major sources of greenhouse gases
is being conducted by EPA's Air and Energy Engineering Research Laboratory
(AEERL). This paper has been reviewed in accordance with the EPA's peer and
administrative review policies and approved for presentation and publication.
^Organized by die East-West Center, households were selected as part of a study of urban energy
fln^j) air pollution funded by die International Development Research Center, Ottawa; physical
sampling, laboratory analyses, an
Agency; and additional data
g were supported by the U.S. Environmental
lysis and evaluation were funded by die Energy Sector
Management Assistance Program, World Bank, and the United Nations Development Program.
Collaborators included the College of Engineering of die University of die Philippines, die Oregon
Graduate Institute, and Lawrence Berkeley Laboratory of die University of California. We greatly
appreciate the contributions and suggestions of our colleagues in these institutions (Smith et al.,
1992b).
- 5-59
-------
INTRODUCTION
It has been said that wood is the fuel that heats you twice, once when you chop
it and again when you bum it. Like fossil fuels, however, biofuels also have the
potential to heat you a third time as a result of enhanced greenhouse wanning due to the
gases released by combustion. It has generally been assumed that this potential is
realized.only when the biomass being burned is harvested on a non-sustainable basis.
With sustainable harvesting, it is argued, an equivalent amount of carbon is recaptured
by the regrowing biomass as released by combustion. Thus, the net greenhouse gas
increment is zero. Even when this is true with regard to the number of carbon atoms,
however, it may not be with regard to their greenhouse equivalence. In particular,
photosynthesis captures only carbon dioxide (CCh) from the atmosphere, but actual
biomass combustion emits other carbon-containing materials including molecules other
than CX>2 with atmospheric warming impacts (Levine, 1991).
These products of incomplete combustion (PICs) are also of concern because of
their effects on human health. In many parts of the U.S., for example, smoke from
wood-fired heating stoves is the principal cause of some types of ambient pollution
during much of the year. For the nation as a whole, wood combustion is a major
emissions source for some important air pollutants, such as paniculates and polycyclic
aromatic hydrocarbons.
Most of the world's woodfuel and other forms of biofuel, such as crop residues
and animal dung, however, are burned not in metal heating stoves in developed
countries, but in simple open cookstoves in developing countries. Approximately half
the households in the world cook in this fashion. Measurements in village homes
throughout the world have shown that health-impairing concentrations of PICs are
often encountered where people use wood or other biomass for cooking or heating
under poorly ventilated conditions (Smith, 1987).
These same PICs also represent lost energy and contribute to the low engineering
efficiency with which meals are cooked in much of the developing world (Baldwin,
1987). This in turn increases pressure on biomass resources, which, along with land
clearing and other factors, has been associated with deforestation and accompanying
environmental problems in some areas. "'
The apparent opportunity for decreasing forest-stressing biofuel demand as well
as reducing health-threatening smoke exposures has lured many local, national, and
international organizations, both government and private, into programs to disseminate
improved biomass stoves in poor countries. Although there have been major successes,
such as the Chinese national improved stoves program, which has reached more than
half the nation's rural households (>100 million stoves), only in recent years has the
percentage of success been high for such programs (Barnes et al.t 1992).
Recently, rising concerns about global warming from the buildup of CO2,
methane (CHO, and other greenhouse gases in the atmosphere have focused attention
on worldwide biomass combustion. Emitting 2100-4700 Tg carbon/y compared to 5700
Tg C/y from fossil fuels, biomass burning plays important roles in the global carbon cycle
(Crutzen & Andreae, 1990). Approaching 1000 Tg C/y, household biofuel, in turn.
. 5-60
-------
accounts for a significant fraction of overall biomass combustion (Meyers & Leach,
1989). A question is thus raised: "Would alterations in household biomass combustion,
such as might be brought about by improved stoves, have significant implications for
global warming?"
This question can be divided into three parts:
1. At the global (macro) scale, what are the contributions of biomass-burning
cookstoves in less-developed countries (here abbreviated BC-LDC) to global
inventories of major greenhouse-related emissions?
2. At the project (micro) scale, what are the technical and economic potentials for
reducing greenhouse-related emissions by changing BC-LDC technologies?
3. At the policy (meso) scale, what are the health, energy, and global warming
implications of various policies affecting BC-LDC?
THE PILOT STUDY
To explore these issues, it is essential to know the BC-LDC emission factors of all
airborne species that have significant implications for energy, health, and global
warming. Some of this information is already available, for in the 1980s a number of
studies were undertaken to examine the energy efficiency and health implications of
biomass stoves, both in developed and developing country situations (Smith, 1987).
Although a significant amount of greenhouse-gas research has gone into studies of
developing-country biomass burning at large scale (forest fires, swidden agriculture,
savannah burning, etc.), however, relatively little attention has focused on the type of
small-scale combustion found in BC-LDC (Levine, 1991).
Before attempting to fill this gap by embarking on a large-scale study of BC-LDC
greenhouse-related emissions, we and our colleagues decided to conduct first a
relatively small pilot study. This could serve the double purpose of:
a. exploring whether the results were of sufficient interest to warrant
conducting a larger study; and .
b. field testing some of the sampling and analysis techniques that might be used
in the larger study.
With these goals in mind, a small study of cookstoves in Manila was undertaken.
Monitored were emissions of more than 80 greenhouse-related and health-related gases
(mostly non-methane hydrocarbons. NMHCs) from traditional cookstoves burning
wood, charcoal, kerosene, and liquefied petroleum gas (LPG), which together account
for the majority of all cooking in developing countries. Involving only a few stove/fuel
combinations in each category, it is not possible to draw statistically valid global
conclusions from this pilot study. Nevertheless, the measurements are quite suggestive,
illustrating how more detailed studies of this type could be useful in answering the three
questions posed above.
5-61
-------
Table 1 summarizes the results of the sampling and analyses in terms of emission
factors, grams of pollutant per kilogram of fuel for each of the major PICsZ. The few data
points argue that these results must be seen as tentative, until they can be verified at
larger scale. Fuels are presented in order of increasing health effects and decreasing fuel
carbon content.
TABLE 1. EMISSION FACTORS, GRAMS PER KILOGRAM DRY FUEL
LPG
Kerosene
Charcoal
Wood
n
2
7
6
9
Fuel
Carbon
Content
0.87
0.86
0.80
0.50
C02
3190
3050
2570
1620
CO
25
39
210
99
CH4
0.01
0.90
7.80
9.00
TNMOC
3
14
4
12
TSP
0.10
3.00
1.70
2.00
n » Number of data points.
TNMOC (Total Non-Methane Organic Compounds): Per carton molecular weight
taken as 18.
TSP (Total Suspended Participates): Considered 75% carbon, see footnote 3.
EVALUATION .
A useful way to evaluate both the local and global effects of BC-LDC is to detail
their impact on the carbon cycle. Shown as the framework to do so is the carbon flow
common to Figures 1-3, which is derived for the composite wood-fired cookstoves in the
Manila study. It follows the typical fate of the 500 g of carbon contained in 1.0 kg of
wood burned in such stoves. About 88% of the carbon is emitted as COi (weighing 1.6
kg) and the rest (60 g) is distributed as shown in Figures 1-3 among several kinds of
PICs, which together weigh about 126 g.3 v
Such a framework allows an examination of this flow in the context of the three
most important aspects of the emissions: energy, health, and global wanning.
field samples from the flue stream of each above into
These were detennined by taking
steel canisters that were sent back to the US.! or laboratory analysis. TTierarioofeachPICto
C0fc (net of background levels) was used to determine emission factors by taking literature values
for the carbon contents of each fuel type and constructing the carbon balance for each stove.
Details are found in Smith etaL,1992b.
«
JParticulates were not measured in the pilot study, so data from other studies were used in die
figures (Smith, 1987; Joshi et aL, 1989; Smith, 1990).
" 5-62
-------
The Cookstove Energy Cycle
89% CombiiBtlon Efficiency
2-2 kJ/C 17.8 MJ
1 kg Wood
500g Carbon
440
COj
A«h
43.S
-, .".'., .-«-""'-- r-. -*." .->''
Energy Factor (kJ/c) i_
CO
24
1.0
CH, TNMOC RSP
T4 71 4S
OS OJ 0X7
Figure 1. This shows the movement of fuel carbon through a traditional wood-fired
cookstove as measured in the Philippines (Smith et aL, 1992a&b). Sixty grains of
carbon was not combusted completely; i.e.t was released as PIOs. Based on the available
energy in each PIC, if it had all been combusted completely, another 22 MJ would have
been released as heat The stove, therefore, has a combustion efficiency of about 89%.
All the numbers refer to grains of carbon alone; e.g., the full mass of CO would be 28/12
(133) times larger. Here, total non-methane organic compounds (TNMOCs) are used
instead of NMHCs. Respirable suspended particulates (RSPs) are used instead of TSP.
Greater than 90% of TSPs are RSPs. NMHCs are about 94% of TNMOCs in this stove.
Source of energy contents: Lelieveld &. Crutzen (1992).
.. 5-63
-------
Energy To put the PICs in an energy context, each constituent needs to be weighted
by its energy content: i.e., the additional energy that could have been released if it had
been burned all the way to COi- As shown in Figure J , the result is that the PICs
contain about 1 1% of the energy originally in the wood; i.e., the combustion efficiency is
about 89%. In other words, compared to a stove with near 100% combustion efficiency,
this stove requires about 12% more fuel (1/0.89).
This inefficiency is part of the reason that traditional stoves use more fuel than it
seems they should. The other major technical reason, of course, is low heat transfer
efficiency (the fraction of heat released from the fuel that is taken into the cooking
utensil).
- As well as representing an energy loss, the 126g of PICs represent the main
health-damaging air pollutants from wood combustion. One way they can be
aggregated and compared is by use of the Relative Hazard Index (RHI). This is simply
the amount of air it would take to sufficiently dilute each pollutant until it reached the
relevant health-based concentration standard (Smith, 1987). With standards as shown,
the total RHI of the PICs is about 120,000 m3. CO: is not much of a health hazard, as
shown by the relatively small RHI, 1800 m3. (Obviously, application of different
standards (e.g., from different countries) would result in different weightings for the
pollutants.)
A practical use of these dilution volumes would be to estimate what fraction of
the needed dilution might actually be achieved in a typical cooking situation. If, to be
generous, a village kitchen is 40 m3 and its air exchange rate is equivalent to about 25
air changes per hour (Smith, 1987), the kitchen has access to about 1000 m3 of dilution
air each hour. A typical bum rate for a woodstove is about 1 kg/h, producing each hour
the amount of PICs shown in Figure 2. Even in this fairly well-ventilated situation, the
available dilution air would seem to be far below (40-70 times) what is needed to keep
total non-methane organic compounds (TNMOCs)4 and respirable suspended paniculate
(RSP) concentrations from exceeding the standards (the latter being more important for
health), a prediction consistent with many village measurements in developing countries.
In fact, it is not uncommon for indoor concentrations to reach 100 times the standard for
RSP during cooking (PandeyeiaL, 1989; Smith & Ahuja, 1990).
Two of the PIC categories shown (TNMOC and RSP) are composites containing
a vast array of mostly organic chemicals. Many of these individually are known to be
health-threatening (e.g., benzene in TNMOC and polyaromatic hydrocarbons in RSP).
Thus, if RHIs were calculated for each in turn, the total would be much larger than the
RHIs for the general categories.
^Strictly speaking, TNMOC is the appropriate category for following carbon flows and NMHC for
global warming implications. Since they are both about equal in size (NMHC are about 93% of
TNMOCs in this stove), however, we have not tried to maintain the distinction throughout the text
5-64
-------
The Cookstove Health Cycle
120.000 m»/C
1800 m»/c
1 kg Wood
SOOg Carbon
X 4.1
440
CO2
A»h
43.S
1.8
OOutlon Factor
(rafte) :
CO
. 230
tajMM
CH,
0.1
1
TNMOC
0400
ASP
30,000
4O.OOO
Figure 2. Starting with the same fuel carbon flows as Figure 1, this figure weights the
PICs not on the basis of energy, but on the basis of how many cubic meters of air would
be necessary to dilute the emission to meet U.S. air pollution standards. Where there is
only an occupational standard, an appropriate safety factor (10) has been used to
establish a public standard. The following standards were used (in mg/m3): CO2=900;
C0=10; CH4=11,000 (asphyxiation); NMHO0.16; RSP*0.05. The diluation factors
shown in the figure are on a per-carbon-atom basis. Although the NMHC standard was
actually set to prevent the formation of ozone, it represents a much less stringent
standard than would be applied if certain individual hydrocarbons were used as the
basis of the dilution factor. The benzene in NMHC from these woodstoves, for example,
would require 40 times more dilution than shown.
5-65
-------
Global Warming - Figure 3 evaluates the same PICs in terms of their greenhouse gas
potential. To do this, it is necessary to apply some index so that the impacts of the
different gases can be combined (Smith & Ahuja, 1990). This is so because the gases
have different heat-trapping abilities, lifetimes, and interactions with other gases in the
atmosphere. Here, we have used the Global Warming Potentials (GWPs) developed by
the Intergovernmental Panel on Climate Change (IPCC, 1990; Smith et a)., 1991).
These are given as a ratio to CO2 (either per molecule or per carbon atom), and thus
can be interpreted as the degree to which the total wanning of each compares to COa-
Since the gases have different atmospheric lifetimes, the relative impact (GWP) depends
on the chosen time horizon. Shown here are the results for time horizons of 20 and
JOO years. In general, shorter time horizons make the non-C02 gases look more
important relative to COi, since CO2 is the longest lived of this group.5
The result is that, depending on the time horizon chosen, the non-COi GGs (i.e.,
the PJCs) have a total GWP 20-110% as much as the COa itself. This implies that
looking only at the CO2 emissions of cookstoves may not give a good picture of their
global warming implications. It also implies that improvements in combustion
efficiency could result in much larger reductions in total GWP than would be indicated
simply by changes in CO: emissions.
Biomass-Stove PICs - Relative Weights. From all three perspectives, energy, health,
and global warming, PICs are to be avoided. As shown in Table 2, however, the three
perspectives do not weight the individual PICs in the same way relative to one another.
Note that the weights are much more skewed for the health column than the others; i.e.,
a factor of 41 between NMOC and CO, down by an additional factor of 230 to the
minor hazard of CO2, and then another factor of 10 down to the insignificant health
hazard of CH4. Both the energy and global warming viewpoints, in contrast, hold CH»
and TNMOC to have similar relative weights, and none of the differences are as large.
as for health.
GLOBAL IMPACTS
Using these preliminary data, it is instructive to note how large BC-LDCs might
appear to loom in the global picture for each perspective.
JThese GWPs are not known with certainty and changes can be expected as knowledge improves.
Indeed, in its 1992 supplement, the IPCC (IPCC, 1992) suggested that indirect effects (chemical
interactions affecting other greenhouse gases) of the non-COj gases were not well enough known
to be used in policy discussions; i.e., the values for the CH» GWP would decrease to 13,4, and
1.5 (by time horizon) and those for CO and NMHC would be 1.0 at all times. The report states,
however, that "(t)hc carbon cycle model used in these calculations probably underestimates both
the direct and indirect GWP values for all non-CCfe gases." Given this caveat and that our purpose
here is principally illustrative, we have not modified the GWPs from those recommended originally
by the IPCC (IPCC 1990). We have, however, used the updated estimates for the total effects of
as presented by Lelieveld & Crutzen (1992).
5-66
-------
The Cookstove Carbon Cycle
GWP (CO, Equivalanvc)
470 41
i
° 1 kg Wood
i
SOOg Carbon
i '
\5tpyt|/ ^
1 1
C02 ^
M
T
1 «Ut
HJttwfl ' I .'.... CO
OWP (CCfe EqutalwM/e) : 4 J
170 m-
Jtttra
OWP (CO, Equtotantfe) : 1-*
n
|*-*ih
T 1*
CH4 NMMC BSP
22 12 1
1M M \M
7.B 4.1 1
1 21 14
Figure 3. The same carbon balance for the woodstove is shown as in Figures 1 and 2.
In this case, the PICs are weighted by the Global Wanning Potentials (GWPs)
appropriate for 20-year and 100-year time horizons. Note that the PIC GWP is about
equal to that of the CC>2 for a 20-year time horizon. Sources: Smith et al. (!992a&b);
Smith et al. (1991); IPCC (1992); Lelieveld & Crutzen (1992); Joshi et al. (1989).
5-67
-------
TABLE 1 RELATIVE IMPORTANCE FOR PJCs
UNDER DIFFERENT REGIMES
Global Warming Index
Time Horizon (years)
Energy
Health
20
100
500
CX)2
CO
CH<
NMHC
RSP
1.0
0.8
1.6
1.8
1.3
I
230
0
9400
30000
1.0
4.5
22.0
12.0
1.0
1.0
1.9
7.5
4.1
1.0
1.0
1.3
3.2
2.3
1.0
All values are shown relative to CO2 on a carbon basis; i.e., NMHC
is 9,400 times worse than C02 for health.
Sources: GWPs from IPCC (1990); Smith et al. (1991) as corrected
by Lelieveld & Crutzen (1992). Energy data from Gulp (1979).
Energy
Although humans in some way utilize perhaps 40% of world net biomass
production (Vitousek et al., 1986), as shown in Table 3, that proportion used directly
for fuel accounts for only about 15% of direct human energy use. Even so, biofuel in
the form of wood, crop residues, brush, and animal dung, is today still the chief form
of energy for most humanity, just as it has been since the discovery of fire (Hall &
Rosillo-Calle, 1991). In developing countries, biofuels constitute about 35% of total
energy use, and in rural areas of developing countries, some 75%. In the poorest
developing countries, however, biomass fuels make up 80-90% of all energy use
(Smith, 1987). Based on the pilot study, therefore, the loss of energy represented by
the PICs from BC-LDC is roughly I % of total human energy use and could approach
10% for some countries.
Health
In the case of health, paniculate exposures from biomass use could be
responsible for approximately 50% of the total global human exposure. Most of this
occurs indoors in rural areas of developing countries, although there are significant
exposures in cities and outdoors as well. The vast preponderance of research,
regulation, and control of paniculate air pollution is still focused on urban outdoor
developed-country situations, which, however, account for rather a small overall
fraction of global exposures (Smith, 1988).
5-68
-------
TABLE 3. GLOBAL IMPORTANCE OF PIC FROM BIOMASS-HRED COOKSTOVES
EnerEV
Biomass makes up about 14% of all direct human energy use.
It is about 33% of energy use in developing countries.
It is about 75% of energy use in rural areas of developing countries.
It is the most important fuel for the majority of humanity.
Sources: Smith (1987); Meyers & Leach (1989); Hall & Rosillo-Calle (1991)
Health
Cause of up to 50% of total human exposure to RSP.
Second largest occupational group, after farm workers (cooks).
Known risk factor for most important killer of deveioping-country children (pneumonia).
Sources: Pandey et al. (1989); Smith (1988)
Global Warming
Human biofuel consumption: 20-40% of all biomass combustion. *
1-5% of all CH4 emissions.
6-14% of all CO emissions. .-
8-24% of all NMHC emissions. . ,
1-3% of all human-generated global warming.
Sources: Smith et al. (1992b); Ahuja (1990) .: ., -.:
These high exposure levels are due not only to high paniculate concentrations,
but also to the large populations involved. Indeed, after farmworkers, cooks represent
the largest occupational group in the world 6
It is important to note that the emissions from biomass fuels need not be high
compared, for example, to those from coal-fired industrial and power facilities in order for
the human exposures to be substantially greater. This is because a much larger
proportion of pollution released in household reaches people, compared to that from
centralized facilities. The impact per unit emissions tends to be greater for distributed
releases, and few things are more distributed than cooking, which occurs in every
household, every day. ~ - *- *
» * ~ ... ~". .*.-..--*. i_..i. .*...* -
Greenhouse Gases - Based on the few measurements taken in Manila, it would seem
possible that biomass stoves could account for fairly significant proportions of global
emissions of the three greenhouse gas categories - CO, CH4, and TNMOC (Table 3).
For
*These stoves are undoubtedly responsible for a large fraction of global exposures to a range of
other pollutants as well; e,g., CO, polycyclic aromatic hydrocarbons, formaldehyde, and benzene.
's-69
-------
CO and CRj, the percentages in Table 3 translate into contributions to overall global
wanning from biomass-fired cookstoves of 0.4-0.9% and 0.1-0.5%, respectively. These
are in same range as estimated by Ahuja (1990) for all biomass stoves, who also estimates
that the overall contribution of biomass stoves to global warming is about 2%. In
addition the contribution due to PICs is estimated to account for about 15% of net
deforestation and, thus, about 1.5% of net human C02 additions to the atmosphere
(1.1% of total warming).
CONTROL MEASURES
Assuming people will continue to need to cook and are well versed in operating
their stoves, there are basically two ways to reduce PIC emissions from biomass-fired
cookstoves: change the fuel or change the stove. With new information of the kind
made available by the Manila study, it is possible to make further judgments about these
options.
One objective of household energy policy can be to encourage people to move
up the energy ladder sooner than they otherwise might. This can be done through fuel
and stove pricing or other ways to mike new stove/fuel combinations relatively more
attractive. In most parts of the developing world, the first step beyond unprocessed
biomass is charcoal or kerosene, followed by LPG. In some areas, Thailand, for example,
little kerosene is provided and the first step after charcoal is LPG. In China, i?is often
coal, followed by LPG. Movement up the ladder generally results in substantially fewer
health-damaging PIC (RSP and CO) emissions per meal (Smith, 1990).
With a switch from biomass to fossil fuels, however, a ^ r
might at first seem inevitable because fossil rather than contemporary carbon would be
emitted. Because biomass combustion leads to a high amount of PICs with a large GWP
(Figure 3), however, the picture is substantially more complicated.
Based on the pilot study results, consider the benefits of switching from wood to
charcoal, kerosene, or LPG as summarized in Table 4.7 Prom a health standpoint, a shift
from wood to LPG reduces the overall health impact by a factor of 100. Kerosene, on
the other hand, results in a reduction by a factor of six. Use of a charcoal stove results in
an improvement by more than a factor of four.
From a global wanning standpoint, the impact depends on how the biomass is
harvested; ie., whether the COj is recycled. Table 4 shows the range between the
extremes; ie., assuming totally sustainable harvesting and complete deforestation (no
regrowth). It is perhaps not surprising that, with complete deforestation, LPG and
kerosene are better man wood, 9 and 6 times, respectively. Even with sustainable
harvesting, however, because of the significant amount of PICs released by the
woodstove, kerosene and LPG stoves release, respectively, 2.6 and 1.7 times less GWP
per meal cooked.
The values in Table 4 have been derived by taking into account die differences among the
stove/fuel combinations in overall cooking efficiency and energy per fuel carbon atom.
5-70
-------
TABLE 4. RELATIVE PIC EMISSIONS, HEALTH EFFECTS. AND GLOBAL POTENTIALS OFLPG,
KEROSENE, CHARCOAL, IMPROVED BIOMASS, AND WOOD COOKSTOVES
. Stove Biofuel
Eff. Use
LPG
Kerosene
Charcoal
me
wood
outdoor
Wood
0.70
0.50
0.30
0.25
0.15
0.00
0.00
1.60
1.60
1.00
20-YEAR WARMING
CO
0.02
0.05
0.73
0.20
0.40
1.00
CH< TNMOC
9.30E-05
0.01
0.30
-
0.20
0.40
1.00
0.02
0.13
0.10
0.20
0.40
1.00
RSP Health Deforest Regrow
0.01
0.17
0.25
0.20
0.40
1.00
0.01
0.16
0.22
020
0.04
1.00
0.11
0.17
0.91
0.20
0.40
1.00
0.39
0.61
0.29
0.20
0.40
1.00
Note: Except for the first, all the columns are normalized such that the impact of the traditional wood-fired
cookstoves is set to 1.0. Where appropriate, the impacts of the improved wood-fired cookstove with
flue are divided between those that occur indoors and outdoors. Elimination of indirect effects for CO
and TNMOC (NMHQ as suggested by IPCC (1992) changes the relative magnitudes, but does not
change the conclusions that these fossil-fuel cookstoves produce less GWP than this woodstove even
with renewable harvesting.
These values are somewhat misleading, however, because they exclude PIC
contributions from elsewhere in the fuel cycles for these fuels. The releases at oil fields
and refineries for kerosene and LPG, however, are likely to be less than 10% of those at
stoves (Smith et ah, 1975). For locally harvested wood, they should be even lower,
although there may be some wastage during harvesting, transport, and storage.
For charcoal manufacture, however, the contribution is likely to be quite large.
Although there are few emissions data from the charcoal kilns commonly used in
developing countries, even relatively modern kilns apparently emit rather large amounts
of PICs (Foley, 1986). Based on an extrapolation from emissions measurements of a U.S.
Missouri kiln (USEPA, 1986) operating at 33% efficiency (mass charcoal/dry wood).
Figure 4 shows the possible carbon flow in a kiln operating at 20% efficiency, which is
common in developing countries (Foley, 1986; Katyega and Kjeilstrom, 1991). Note that
nearly 35% of the carbon in the wood put in the kiln is released as PICs. Thus, the
20-year GWP ratio of PICs/COi for the kiln is about 7.6; i.e., the PICs produce more than
seven times as much GWP as the COz (4.8 times the 100-year GWP).
If the 64 million tonnes of wood made into charcoal each year in developing
countries (Joshi et aL, 1989) was charged to kilns like the one in Figure 4, charcoal
making might be responsible for releasing 1.5 Tg/yr of carbon as CH* This source may
be significant but direct measurements would help develop reliable estimates for the
types of the kilns typically in use.
- 5-71
-------
Wood
Wet Wood
6.0
Water
1.0
Wood
fi.O
Carbon
2.5
Kiln
1.0
0.8
Charcoal
0.9.
0.8
** PIC
0.19 - CO
0,11 -Clti
0.27 TNMOC
0.24 - TSP
Figure 4. Carbon flows for hypothetical charcoal kiln used in developing countries.
Based on measurements reported in US. EPA's Compilation of Air Pollutant Emission
Factors, 1986 (U.S. EPA, 1986) for a Missouri kiln with 33% production efficiency. This
has been modified to a 20% efficient kiln, which is more typical in developing countries,
by assuming that the difference in efficiency does not alter the ratio of PICs to COj.
TTJMOC=total non-methane organic compounds (vapor); TSP=totaJ suspended
particulates (aerosol).
5-72
-------
100-
90-
80-
#
Comtaustton
50-
40-
30-
20"
10-
X
Heat transfer
Overall
Traditional IC-1 IC-2
Stove
Figure 5. The differences among overall and internal efficiencies in three metal wood-
fired cookstoves without flues. Note that, although both improved stoves achieve
substantially more overall energy efficiency than the traditional stove, combustion
efficiencies are less. Thus, IC-2 produces 4 times more PICs per unit energy delivered
than the traditional stove (100-92)7(100-98). Thus, though its energy use per meal is 2JS
tunes less (15/40), the overall result is that about 60% more PICs are produced per meal.
The original investigators measured only CO and particulates (Joshi ct al., 1989). The
remaining PICs have been assumed to appear in the same ratios as measured in the
Manila pilot study.
-5-73
-------
Stoves
The first-level approach to improved biomass cookstoves (IBCs) is shown in
Table 4. The IBC uses 40% less fuel and releases about 35% of its PICs into the kitchen,
the rest going outdoors where it is assumed to have only 10% as much health impact per
gram as emissions released indoors. (These changes are based on results found in field
measurements of improved stove programs (Pandey et al., 1990; Ramakrishna et al.,
1989; Reid et al., 1986). Improvements measured in the laboratory can be much greater.)
Note that, due to lack of data, there is no change considered in global warming impact
other than that 40% less fuel is used; i.e., the same fraction of the carbon is released as
PICs of the same composition.
To understand how changes in stove design and operation actually affect PIC
emissions, it is important to recognize that overall stove efficiency (Et) is a function of
two internal efficiencies combustion efficiency (Ec) (i.e., the amount of chemical
energy in the fuel that is convened to heat) and heat-transfer efficiency (Eh) (i.e., the
amount of heat that reaches the food in a cooking stove or reaches the room in a heating
stove):
= (Ec)x(Eh)
0)
In general, emissions per meal of PICs and COi are an inverse function of overall
efficiency in that, all else being equal, the less fuel used for a given cooking task, the less
PICs will be released. Thus, improvements in fuel efficiency should lead to lowtr
greenhouse gas (GHG) emissions.
Changes in stove operation and design, however, often affect the two internal
efficiencies in quite different ways. In particular, thermal transfer efficiency can be
increased at the expense of combustion efficiency. Design and operation changes that
improve overall fuel utilization, therefore, sometimes actually increase one internal
efficiency at the expense of the other.
Although few data are available for biomass cookstoves, Figure 5 illustrates this
effect in a study of paniculate and CO emissions of one traditional and two improved
wood-fired metal cookstoves (Pandey et al., 1989). Overall efficiency rosesrrom 15% to
31% and 37% in the two improved stoves, greatly decreasing potential fuel demand for
cooking. In the process, however, PIC emissions per meal actually increased by 8%
because combustion efficiency dropped from 97% to 92%.
It might be thought that there is little net GHG impact from changes in
combustion efficiency. In other words* the fuel carbon that is not oxidized all the way
to COa will be released as PICs. The smaller the fraction of carbon released as CO* the
more as PICs and vice versa, .-
In rough terms, this trade-off is true for carbon mass and number of carbon atoms.
It may not be true for the net greenhouse impact, however, because these different
molecules have different greenhouse impacts. Thus it is necessary to keep track not
only of the total carbon emissions but also of their form.
. 5-74
-------
The PJCs/COa ratio can vary dramatically even at constant overall efficiency,
depending on the relative contribution of heat transfer and combustion efficiencies.
C02 emissions are in general less dependent than PICs on combustion efficiency. For
example, a shift from 90% to 80% combustion efficiency results in a near doubling of
PICs but only about 10% less O>2. (More dramatically, a change in combustion
efficiency from 99% to 98% would result in less than a 1% loss of total efficiency but a
near doubling of PICs.)
Since PICs emissions are a stronger function of combustion efficiency than they
are of total efficiency, emissions can sometimes increase along with total efficiency. A
popular means by which fuel utilization of traditional cookstoves has been raised,
regrettably, is simply to reduce airflow by enclosing the fire, thereby greatly increasing
the heat transfer efficiency to the pot, but also lowering the combustion efficiency. The
end result, therefore, can be a net increase in fuel utilization and a consequent reduction
in COa emissions, but a rise in the PICs/C(>2 ratio or even an increase in absolute PIC
emissions per cooking task, as shown in Figure 5.
The GHG implications of stove emissions depend strongly not only on the
PICs/COj ratio, of course, but also on the particular mixture of PIC molecules. Each
mixture will have a different greenhouse equivalence weighting depending on the
relative amounts of the different constituents.
The radiatively active PIC molecules, such as CHu and the molecules thtt play a
pan in their atmospheric chemistry before turning into COii such as CO and NMHC,
have total (direct and indirect) GWPs above 1.0; i.e., greater than CCh. Indeed, it would
seem that all organic molecules must have a GWP per carbon atom of at least 1.0 because
once released they would relatively soon be oxidized to CO* the rating of which, by
definition, is 1.0. Only elemental carbon particles might have a GWP less than 1.0. if they
are assumed not to be oxidized within & relevant time period.
Although there do not seem to be sufficient theory or data to predict the exact
relationship between design changes and the PICs/COz ratio, we do have rough
estimates of typical GWPs of PICs. The PIC GWP of the woodstoves in Manila varied
from 1.7 to 7.8, depending on the time horizon.
Thus, efficiency improvements to the Manila woodstove that allowed combustion
efficiency to drop in exchange for increased heat-transfer efficiency could actually lead
to significant increases in PICs with their health and greenhouse impacts.
Ion
Final decisions with regard to fuel and stove changes will of course dc
relative economics and other non-environmental issues (Smith, 1992). Ne
sless, the
carbon flow framework made possible by the monitoring data would be a valuable
grounding for these further, analyses. :: -.r-.r-r-. ~
CONCLUSION
Putting aside for the moment the few actual measurements involved, the
information available from the Manila pilot study has allowed us to make substantial
progress toward answering the three questions posed at the beginning:
5-75
-------
1. The contribution to global inventories;
2. Advantages and disadvantages of various technical options (stove/fuel
combinations); and .
3. Policy implications because of interactions among energy, health, and global
warming objectives.
As we indicate above, it would be extremely useful to have sufficient data to
resolve the carbon balances of such small combustion devices as stoves. Five
characteristics make such devices attractive for this kind of research:
a. The devices are (or are operated in a way to be) substantially different
from those in developed countries, which have been subject to much
research already;
b. They have widely varying, but generally poor, combustion efficiencies,
leading to significant amounts of PICs;
c. Although individually small, they are widely used, leading potentially to
emissions significant on the global scale;
d. There is substantial scope for technical improvement; and
e. Of interest, although not of direct concern from the standpoint of global
wanning, they are evenly dispersed with the population, making their PIC
emissions more likely to produce ill health.
Stoves fill this bill, as do other ubiquitous combustion devices such as motor vehicles.
Based on these encouraging, but preliminary findings, we are planning to embark
on a more extensive study of cooking and heating stoves. This will be undertaken
jointly with colleagues in India and China, which not only contain about 65% of the
population in developing countries, but where a wide range of stove/fuel combinations
are in use. Because of the potentially, but little studied, significance of charcoal kiln
emissions, we also plan to study the emissions of a wide range of kilns in Thailand and
.Brazil . .. . :- ; --
We also hope to examine more closely a set of greenhouse-gas sources in
developing countries that have potential for rather large contributions to global
Inventories. These are garbage dumps around large cities where not only is there
significant anaerobic decay, but also smoldering spontaneous combustion, low-tech
incineration, and intentional burning by scavenger communities to recover metals.
5-76
-------
REFERENCES
Ahuja, D.R., 1990, "Research Need for Improving Biofuel Burning Cookstove
Technologies," Natural Resources Forum 14(2): 125-134.
Baldwin, S., 1987, Biomass Stoves, Volunteers in Technical Assistance, Princeton
University, Arlington, VA.
Barnes, D.F., et al., 1992, "What Makes People Cook with Improved Stoves? A
Comparative Review," ESMAP/UNDP Paper, World Bank, Washington, D.C.
Crutzen, PJ. & MO. Andreae, 1990, "Biomass Burning in the Tropics: Impact on
Atmospheric Chemistry and Biochemical Cycles," Science 250:1169-1678.
Gulp, A.W., 1979, "Principles of Energy Conversion," McGraw Hill, New York, NY.
Foley, G., 1986, Charcoal Making in Developing Countries, Earthscan, London.
Hall, D.O. & F. Rosillo-Calle, 1991, Biomass in Developing Countries, Report to the
Office of Technology Assessment, Washington, D.C.
IPCC (Intergovernmental Panel on Climate Change), 1990, Climate Change: The IPCC
Scientific Assessment, Cambridge University Press. UK. *
IPCC, 1992, Supplement, Cambridge University Press. UK.
Joshi, V. et al., 1989, "Emissions from Burning Biofucls in Metal Cookstoves,"
Environmental Management 13(6): 763-772.
Katyega, MJ., & B. Kjeilstrom, 1991, "Assessment of Forest Biomass Technology," ATAS
Bulletin, #6,139-148.
Lelieveld, J. & PJ. Crutzen, 1992, "Indirect Chemical Effects of Methane on Climate
Warming," Nature, Vol. 355:339-341.
Levine, J.S., ed., 1991, Global Biomass Burning, MIT Press, Cambridge, MA.
Meyers S. & G. Leach, 1989, "Biomass Fuels in the Developing Countries: An
Overview," LBL-27222, Lawrence Berkeley Laboratory, Berkeley, CA.
Pandey, M.R., et al., 1989, "Indoor Air Pollution in Developing Countries and Acute
Respiratory Infections in Children," Lancet Feb.25:427-429.
Pandey, MJL, et aL, 1990, "Hie Effectiveness of Smokeless Stoves in Reducing Indoor
Air Pollution in a Rural Hill Region of Nepal." Mountain Research and Development,
10(4): 313-320.
Ramakrishna, J., et al., 1989, Cooking in India: The Impact of Improved Stoves on Indoor
Air Quality," Environment International 15(1-6): 341-352.
5-77
-------
Reid, H.F., ct al., 1986, "Indoor Smoke Exposures from Traditional and Improved
Cookstoves: Comparisons among Rural Nepali Women," Mountain Research and
Development, 6(4): 293-304.
Smith, K.R., et al., 1975, Evaluation of Conventional Power Systems, Jet Propulsion
Laboratory, Pasadena CA.
Smith, K.R., 1987, Biofuels, Air Pollution, and Health, Plenum Press, New York, NY.
Smith, K.R., 1988,4Total Exposure Assessment: Pan 2, Implications for Developing
Countries," Environment 30(8): 10-15; 33-38.
Smith, K.R., 1990, "Indoor Air Quality and the Pollution Transition," in H. Kasuga, ed.,
Indoor Air Quality, Springer-Verlag, Berlin, 448-456.
Smith, K.R., et al., 1991, "Indices for a Greenhouse Control Regime That Incorporates
Both Efficiency and Equity Goals," Working Paper 91-21, Environmental Policy and
Research Division, World Bank, Washington, D.C.
Smith, K.R., 1992, "Biomass Cookstoves in Global Perspective," in World Health
Organization, Indoor Air Pollution from Biomass Fuel, WHO/PEP/92.3B, Geneva, 164-
184.
Smith, K.R., et al., 1992a (In Press), "Greenhouse Gases from Biomass and Fossil Fuel
Stoves in Developing Countries: A Manila Pilot Study," Chemosphere.
Smith, K.R., et al., 1992b, "Greenhouse Gases from Small-Scale Combustion in
Developing Countries," EPA-600/R-92-005 (NTIS PB92-139369), Air and Energy
Engineering Research Laboratory, Research Triangle Park, NC.
Smith, K.R. & D.R. Ahuja, 1990, "Toward a Greenhouse Equivalence Index: The Total
Exposure Analogy," Climatic Change 17:1-7.
USEPA, 1986, Compilation of Air Pollutant Emission Factors, AP-42 (NTIS PB87-
150959), Washington, D.C. *
Vitousek, P., et aLv 1986, "Human Appropriation of the Products of Photosynthesis,"
Bioscience, 36(6): 368-373.
5-78
-------
AN ANALYSIS OF THE HYDROCARB PROCESS
FOR METHANOL PRODUCTION FROM B1OMASS
by: Yuanji Dong and Meyer Steinberg
Hydrocarb Corporation
232 West 40th Street
New York, NY 10018
and Robert H. Borgwardt
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
A pilot plant is being designed to evaluate the technical feasibility of producing
transportation fuel from biomass by the Hydrocarb process. The facility will be constructed in
California by Acurex Corporation with assistance from Hydrocarb Corporation under
cosponsorship of the South Coast Air Quality Management District and the EPA. As a basis for
that design, computer simulations and experimental studies have been carried out to establish
optimum process conditions for a range of feedstocks that are anticipated for pilot plant tests.
This paper discusses the results of simulations to determine the operating parameters and
performance when using urban wastes such as greenwaste and sewage sludge as feedstocks. The
simulations were used to configure the process steps for maximum fuel (methanol) production,
to determine feed rates, and to estimate thermal efficiency. The results indicate that about 77
kg of methanol can be produced from 79 kg of dry greenwaste when sludge and digester gas are
fed as co-feedstocks in a ratio of 0.2. The optimum system pressure is found to be 50 atm (5
MPa). Temperatures of 900°C for gasification and 1000°C for methane pyrolysis are
recommended on the bases of thermodynamics, kinetics, and the limitations of materials of
construction. Thermal efficiency at these conditions is estimated to be 74 percent
This paper has been reviewed in accordance with the U.S. Environmental Protection
Agency's peer and administrative review policies and approved for presentation and publication.
6-26
-------
-------
INTRODUCTION
Thermodynamic calculations suggest that the Hydrocarb process0-2' might produce carbon
black, hydrogen, and/or methanol from fossil fuels or virtually any other carbonaceous materials
and might do so with high thermal efficiency. Potential feedstocks include coals of all ranks,
residual oil, oil shaie, woody biomass, sewage sludge, and municipal wastes. This innovative
process consists of three essential reactions: (1) hydrogasification of the carbonaceous feedstocks
to produce a methane-rich gas, (2) thermal decomposition of methane to carbon and a hydrogen-
rich gas which is recycled, and (3) catalytic conversion of the carbon monoxide (CO) and
hydrogen contained in the recycled gas to produce methanol. The mix of process products is
optional: a clean solid fuel (carbon black), a liquid fuel (methanol), slurry fuel mixtures of
carbon and methanol, or gaseous fuels (hydrogen and methane) can be produced by changing the
order of the reaction steps.
Since the process operates without additional steam and oxygen--and because the oxygen
in the feedstocks is removed mainly in the form of methanol and water-the carbon dioxide (COJ
emission is significantly reduced in comparison to traditional fossil fuel conversion processes
involving steam/oxygen gasification'". .When biomass is used as a co-feedstock with fossil fuel
to produce methanol and the carbon is sequestered, the CO2 greenhouse gas emission can be
significantly reduced and even eliminated'4**'.
PRELIMINARY ANALYSES
In order to attain a better understanding of the potential of the Hydrocarb process,
particularly as a means of producing a clean transportation fuel from biomass, the U.S.
Environmental Protection Agency (EPA) developed an interagency agreement with the
Brookhaven National Laboratory to perform detailed process analyses and related experimental
studies of biomass hydrogasification and methane pyrolysis. The process was analysed using a
computer simulation model developed by Hydrocarb Corporation that performs complete mass
and energy balances for various process configurations, feedstock options, reactor types, and
operating conditions. Initial results of the simulations were published by the EPA in 1991<3); a
follow-up report is in preparation.
EPA's independent assessment of the biomass/natural-gas option of the Hydrocarb process
as a technology for production of alternative transportation fuels<6> confirmed that it can,
theoretically, produce methanol at a cost that is competitive with petroleum fuels. Most
importantly, it concluded that methanol may be produced and utilized in the transportation sector
with a 70 percent reduction in CO, emission relative to gasoline at no incremental cost and may
achieve 100 percent CO2 reduction at. marginal incremental cost Other potential advantages
derive from its higher yield of fuel energy compared to other processes for producing alternative
fuels from biomass or from natural gas by the conventional steam reforming process. The
Hydrocarb approach therefore has potential for mitigating C02 emissions from mobile sources
in a more effective manner than current alternatives. Since the transportation sector accounts for
at least 24 percent of total C02 emissions, any means of reducing that source will have to be
considered when assessing options for dealing with the global warming problem.
6-27
-------
PILOT PLANT EVALUATION
Because of methanoJ's potential role in reducing urban air pollution, the South Coast Air
Quality Management District (SCAQMD) of California is seeking new technologies for domestic
methanol production, especially from urban wastes, and has agreed to co-sponsor with the EPA
a pilot plant evaluation of the Hydrocarb process. Preliminary design of that pilot plant is now
in progress. The objectives of this joint project by SCAQMD, Hydrocarb Corporation, Acurex
Corporation, and EPA are: (1) to demonstrate the technical feasibility of producing methanol
from woody biomass and natural gas by the Hydrocarb process, and (2) to evaluate its
performance using sewage sludge, digester gas, and greenwaste as alternative feedstocks. If
successful, the results will be used to establish more comprehensive cost estimates and provide
a basis for decisions regarding further development. Prior analyses (34l6) have considered the use
of woody biomass and natural gas as feedstocks. This paper focuses on the design of a pilot
plant that will utilize 50 Ib/hr (23 kg/hr) of urban wastes as feedstock.
LABORATORY STUDIES
KINETICS OF BIOMASS HYDROGASIFICATION
The hydrogasification of biomass in the form of poplar wood having particle size less than
150 urn in diameter was investigated in a 25 mm ID and 2.5 m long tubular reactor facility
described in detail elsewhere. The tests were conducted at temperatures up to 800°C and
pressures between 30 and 50 atm. The experiments were performed in two modes depending
upon the heatup rate. In the low heatup mode, the biomass was first loaded in the reactor at
room temperature. Hydrogen was then introduced into the system until the desired initial
pressure was established. The reactor was slowly heated at a rate less than 10°C/min while the
change of pressure in the reactor and the composition of the effluent gas were monitored with
time. In the higher heatup mode, the reactor was first heated to the desired temperature and
pressurized with hydrogen before introducing the biomass. From the variation of pressure and
gas composition versus time, the rates of reaction and degree of conversion were determined.
The high-heatup-rare experiments showed that 88 to 90 wt percent of biomass could be gasified
at residence times in the order of 15 minutes'7'. The number of moles of gas formed, as
calculated from the pressure change, varies with time as shown in Figure 1.
METHANE PYROLYSIS
The rate of thermal decomposition of methane was investigated using the same reactor
facility at temperatures ranging from 700 to 900°C and pressures of 28 to 56 atmw. In these
experiments, methane was continuously fed into the reactor. On-line gas analyses were taken
upstream and downstream of the reactor to determine the reaction rate. The variation of methane
concentration with residence time at different operating conditions, shown in Figure 2, indicates
that a gas residence time of about 2 min is required for the reaction to reach equilibrium
composition at 50 atm and 900°C. The activation energy is found to be 31.3 kcai/mol. By
extrapolating it to higher temperatures, the residence time would be 41 sec at 1000°C and 12 sec
at 1140°C.
6-28
-------
o
2
0>
UJ
u.
O
cr
UJ
m
2
3
Z
0.5
10 SO 30 40
TIME (m\n.)
60
BO
Figure 1. The change in number of moles in the reactor with time at 800°C and 5.24 atm
of initial hydrogen pressure. Run No. 1152
_ 100
' Equilibrium
concentration
0 20 40 60 BO. 100 120
r RESIDENCE TIME (sec)
Figure 2. Methane concentration vs. gas residence time. (56.1 atm and 900°C)
6-29
-------
PROCESS SIMULATION STUDIES
REACTION SEQUENCE
There are two possible configurations for the Hydrocarb process, according to the
sequence of the three reactor steps involved. One of these configurations, Cycle 1, is designed
so that the process gas flows successively from the hydrogasification reactor (HPR) through the
methane pyrolysis reactor (MPR), the methanol synthesis reactor (MSR), and the methanol
condenser (COND). The gas leaving the condenser is recycled back to the HPR after
withdrawing a small purge gas stream. The other configuration, Cycle 2, directs the process gas
formed in the HPR to the MSR and condenser, then to the MPR, followed by recycle to the HPR.
Figures 3 and 4 illustrate these configurations.
Purge Ga«
Feedstocks
HPR
Carbon
MeOH
COND
HE
MPR
H2O
MSR
Figure 3. Cycle 1 of the Hydrocarb process.
In Cycle 1, the process gas from the MPR is cooled from the temperature of that reactor
down to the MSR reaction temperature, 260"C, and this recovered energy is used to heat the
process gas from the condenser, operating at SO°C, up to about the temperature of the HPR.
From an energy balance on the HPR (assumed to be a fluidized bed), the temperature of the
recycled process gas fed into it is adjusted until the reaction heat generated in the HPR is
balanced by the enthalpy difference of inlet and outlet gas streams. In Cycle 2, two gas/gas heat
exchangers are used: one cools the gas stream from the temperature of the HPR to the
temperature of the MSR by heating the gas stream from the condenser; and before entering the
MPR, the process gas is further heated in the second heat exchanger by the hot gas stream from
the MPR.
6-30
-------
Purge Gas MeOH
Carbon
MPR
Feedstocks
J HE
COND
HE
HPR
H20
MSR
Figure 4. Cycle 2 of the Hydrocarb process.
The primary interest of SCAQMD and EPA is to produce a maximum amount of
methanol from biomass. With this in. mind, computer simulations were made with the process
model to compare the above cycles and determine which is better suited to that objective. The
feedstocks assumed for this comparison are representative of urban wastes obtained from a
California wastewater treatment plant (sewage sludge and digester gas) and other sources
(greenwaste). The compositions of these materials are listed in Table 1. Included in Table 1
are the higher heating values (HHVs) of each fuel, given on a moisture-free basis (MF), and the
heat of formation, which is expressed on a moisture-and-ash-free basis (MAP).
-,
Performance Criteria ~ " -' --'.'
Several important performance criteria are used to compare the two cycles and also to
assess the performance of Hydrocarb with alternative feedstocks. They include: minimum
methane feed per unit of biomass feed, production ratio of methanol to carbon black, carbon
efficiency, thermal efficiency, and gas circulation rate. Thermal efficiency is defined here as the
total heating value of methanol, carbon black, and net purge-gas production (i.e., that which is
not used to heat the MPR) divided by the total heating value of biomass, natural gas, and any
external fuel source that is used to heat the MPR or close the energy balance. Similarly, carbon
efficiency is the ratio of total carbon in all of the products to the carbon in all of the feedstock
materials including the fuel used for MPR heating.
6-31
-------
TABLE 1. PROPERTIES OF THE FEEDSTOCKS USED
IN THE STUDY
Feedstock
Green-. Methane
waste CH«
Sewage Digester
sludge Gas
Molecular Formula
Composition (wt%)
C
H
O
H,O
Ash
S
N
CH,
CO,
High Heating Value
(Btu/lb-MF)"
(kcal/kg-MF)b
Heat of Formation
(keal/kg-MAF)
CH,
49.11
5.85
35.41
5.00
3.34
0.16
1.13
-8670
-4817
-1386
100
-23881
-13267
28.55
4.09
16.03
9.82
36.53
1.36
3.62
-5510
-3061
-1770
36.81
61 . 82
8792
4884
1 Btu/lb
b 1 kcal/kg
2 kJ/kg
» 4 JcJ/Jcg
Using the above performance criteria. Table 2 compares Cycles I and 2 under the same
operating conditions when the feedstocks are greenwaste and natural gas. The comparison is
based on a capacity of 100 kg/hr of greenwaste containing 5 wt percent moisture. The HPR is
assumed to be a fluidized bed in which 90 percent of the carbon content of the biomass is
gasified (an assumption justified by the experimental work discussed above). The material
balances calculated for the two cycles show that a minimum natural gas feed equivalent to 10
wt percent of the biomass feed is required for Cycle 1 and 6 percent for Cycle 2.
TABLE 2. COMPARISON BETWEEN CYCLE 1 AND CYCLE 2
WITH BIOMASS AND NATURAL GAS AS FEEDSTOCKS
(50 atm, HPR B 900QC, AND MPR " 1000°C)
Cycle - -- .".- 1 2
Greenwaste (5% H,O) (kg/h) ' -" '"' "'" '"''.'
CH, Rate (kg/h) : ,c~- -- -'~
Burning CH, for MPR (kg/h) ? ^^L
Methanol (kg/h) . :
Carbon Black (kg/h)
MeOH/C (kg/kg)
Carbon Efficiency (%)
Thermal Efficiency (%)
Gas Recycle (kgmol/h)
100
' 10
12.8
.- 58.0
29.7
1.95
77.8
72.0
46.4
100
10
11.6
31.6
35.6
0.89 .
75.1
64.4
43.6
6-32
-------
Table 2 compares the results of process simulations of the two cycles under the same
operating conditions using greenwaste and natural gas as co-feedstocks. It can be concluded that
Cycle 1 provides the higher methanol production and greater thermal efficiency; moreover, it
requires only one heat exchanger. Thus, Cycle 1 is the.configuration that will be used for pilot
plant evaluation.
FEEDSTOCK SIMULATIONS
Greenwaste Moisture
As shown in Table 2, 100 kg of greenwaste containing 5 percent moisture should produce
58 kg of methanol when Hydrocarb is operated in the Cycle 1 configuration. The moisture
content of the feedstock was found to have an influence on process performance as illustrated
for greenwaste in Table 3. The basis for these calculations was 95 kg/h of moisture-free
greenwaste to which varying amounts of water are added. The reaction temperature was 900°C
for the HPR and IOOO°C for the MPR. The table shows that the minimum methane feed ratio,
which is the weight ratio of methane to greenwaste fed into the HPR for hydrogasification,
increases with moisture content of the greenwaste in order to maintain material balance in the
system. Table 3 also indicates that the methanol yield and the thermal efficiency improve with
increasing moisture content of greenwaste. However, as the moisture in biomass fed into the
HPR increases, the exothermic heat generated in the reactor decreases, resulting in the
requirement of a higher temperature of the process gas recycled into the HPR in order to keep
its reaction temperature constant
TABLE 3. EFFECT OF MOISTURE IN GREENWASTE
ON THE PERFORMANCE OF THE PROCESS
(based on 95 kg/h MF greenwaste)
H,O/MF-GW (wt.%)
5.26
10.5 15.8
21.1
Min. CH«/MFH3W (kg/kg)
T of PG into HPR (°C)
Methanol (kg/h)
Carbon Black (kg/h)
MeOfl/C (kg/kg)
Carbon Efficiency (%)
Thermal Effiency (%)
Gas Recycle (kgntol/h) "
0.1
892
58.0
29.7
1.95
77.8
72.0 . _
46.4
0.12
910
65.5
28.4
2.31
77.8
72.8
47.5
0.14
927
73.1
27.0
2.71
77.7
73.4
48.6
0.16
-j 944
80.8
25.5
3.17
77.7
74.1
49.8
When the weight ratio of moisture to greenwaste reaches 21.1 percent, the temperature
of the process gas recycled to the HPR is required to be 944'C. If the moisture of greenwaste
is further increased, the overall reaction heat in the HPR will become endothermic. Therefore,
there is an optimum moisture content of the feedstock for maximum methanol yield. For
green waste, that optimum is 21 percent (Table 3), which yields about 81 kg of methanol for
every 95 kg of dry biomass. Biomass moisture contents can be adjusted to the optimum value
by water addition or by external drying using waste heat from the MPR heater or from the
methanol convener.
6*33
-------
Sewaee Sludge and Digester Gas
Additional simulations were made assuming the feedstocks to be sewage sludge and
digester gas, supplemented by green waste and natural gas. These comparisons are based on a
fixed capacity of 100 kg/h of total solid feedstocks including sludge and green waste. For these
calculations, both the greenwaste and sludge were assumed to be dried to 5 percent moisture
before entering the HPR. The process performance of Cycle 1 at 50 aim and 900°C for the HPR
and IOOO°C for the MPR is summarized in Table 4 assuming various feed ratios of sludge to
greenwaste. The methane feed was 15 percent of the greenwaste rate for all cases, while the
ratio of digester gas to sludge was 0.5, based on the gas generating capacity of the sewage plant.
A maximum allowable sludge feed ratio of 0.4 (sludge/greenwaste) was found in the calculation;
if this ratio is further increased, the HPR becomes endothermic and an additional heat source
would be needed for the HPR.
TABLE 4. THE EFFECTS OF SLUDGE/GREENWASTE FEED
RATIO ON THE PROCESS PERFORMANCE
(50 atm, HPR = 900°C, and MPR « 1000CC)
Sludge/Greenwaste
0.1
0.2
0.3
0.4
Greenwaste (kg/h) *
Sludge (kg/h)'
CH« (kg/h)
Digester Gas (kg/h)
Limestone (kg/h)b
Methanol (kg/h)
Carbon Black (kg/h) .
MeOH/C (kg/kg)
Carbon Efficiency (%)
Thermal Efficiency (%)
100.0
0
15.00
0
1.43
56.44
31.99
1.76
79.65
72.30
90.90
9.10
13.61
4.54
2.47
57.73
31.32
1.84
79.62
72.65
83.33
16.67
12.50
8.33
3.32
58.92
30.83
1.91
79.59
72.94
76.92
23.08
11.53
11.53
4.07
59.91
30.38
1.97
79.57
73.19
^71.43
"28.* 57
10.69
14.26
4.66
60.62
29.94
2.02
79.55
73.40
* 5% moisture content in greenwaste and sludge.
b Molar feed ratio of limestone to total sulfur in
feedstocks is assumed to be 2.0.
The data in Table 4 show that the thermal efficiency and the product ratio are only
slightly affected by the sludge/greenwaste ratio. This indicates that sludge can be processed with
greenwaste by the Hydrocarb process to produce considerable amounts of methanol and carbon
black. Calculations also found that there is an optimum moisture content for each
sludge/greenwaste feed ratio. . As shown in Table 5, increasing the moisture content of
sludge/greenwaste feeds increases the product ratio of methanol to carbon black and the thermal
efficiency of the process. Since the maximum allowable ratio of sludge to greenwaste is 0.4, the
pilot plant design took a sludge/greenwaste ratio of 0.2 as a representative data base. The
optimum flow rate and composition of each stream for the feed ratio (sludge/greenwaste) of 0.2
is shown in Figure 5. When feeding digester gas, it is interesting to note that part of the CO2
component of that gas reacts in the HPR with carbon in the greenwaste to form CO and methane
and the rest of that CO2 is converted to CO in the MPR. This results in an increase in the CO
6-34
-------
mole fraction of the process gas from 2.24 to 5.79 percent in the HPR and from 5.79 to 7.82
percent in the MPR. The CO is then convened to methanol in the MSR. Some waste CO2, as
well as waste biomass, is thus recycled to a clean fuel. .
TABLE 5. OPTIMUM MOISTURE CONTENT FOR
SLUDGE/GREENWASTE FEEDSTOCKS
Sludge/Greenwaste
0.2
0.4
Dry greenwaste (kg/h)
moisture (kg/h)
Sludge (kg/h)
moisture (kg/h)
CH4 (kg/h)
Digester Gas (kg/h)
Limestone (kg/h)
Methanol (kg/h)
Carbon Black (kg/h)
MeOH/C (kg/kg)
Carbon Efficiency (%)
Thermal Efficiency ( % )
79.16
14.08
15.84
1.72
12.50
8.33
3.26
76.86
25.85
2.97
77.88
74.17
67.86
9.96
27.14
2.96
10.69
14.26
4.66
73.90
26.35
2.80
78.28
74.30
OPERATING CONDITIONS
The pressure and temperature at which the HPR and MPR are operated affect the
performance of the Hydrocarb process. With regard to the HPR, the temperature is established
by the requirement for desulfurization, especially in the case of sewage sludge, which is to be
partially accomplished by limestone addition. Given the CO2 partial pressures in the HPR, the
active hydrogen sulfide sorbent, calcium oxide, can exist only at temperatures above 850°C. For
that reason, and the fact that gasification rate and carbon conversion increase with temperature,
an HPR temperature of 900°C is established.
Temperature ._ . -
The effect of temperature in the MPR was investigated at 50 atm pressure when the HPR
temperature is fixed at 900°C. The results shown in Table 6 indicate that increasing the MPR
temperature will improve the methanol production and reduce the process gas circulation rate.
However, the temperature in the MPR is limited by the thermal and structural properties of the
materials of construction. When a MPR temperature of 950°C is assumed, the HPR is no longer
energy neutral: it becomes endothermic. For these reasons, 1000°C is the recommended
temperature for the MPR.
6*35
-------
Pressure
The influence of pressure is also presented in Table 6. The methanol production rate is
found to increase with decreasing system pressure. However, as pressure decreases, the thermal
efficiency is reduced and the gas circulation rate (which affects capital cost of the plant) increases
significantly. Since catalytic methanol conveners normally require at least 50 atm pressure, some
simplification is achieved if the other two process steps that produce the synthesis gas are
conducted at the same pressure.
From this study it is concluded that the most favorable operating conditions for co-
processing biomass and methane are 50 atm system pressure, and temperatures of 900°C in the
HPR and 1000°C in the MPR.
TABLE 6. EFFECTS OF PRESSURE AND TEMPERATURE ON
THE PROCESS PERFORMANCE (100 kg/h GW,
- 0.15, HPR - 900CC)
P
atm
T/MPR
°C
MeOH/C
kg/kg
C Eff Ther Eff
Recycle Gas
kgxnol/h
50
50
50
50
1000
1050
1100
1200
1.76
2.31
2.65
2.96
79.7
78.1
77.3
76.4
72.3
72.3
72.3
71.9
47.5
34.4
27.4
20.4
60
50
40
30
1000
1000
1000
1000
1.55
1.76
2.06
2.31
80.9
79.7
77.9
74.8
72.7
72.3
71.2
67.9
44.7
47.5
53.7
68.3
PILOT PLANT DESIGN
Based on the above process analysis and kinetics experiments, design specifications are
being developed for the HPR and MPR of a 50 Ib/h Hydrocarb pilot plant utilizing biomass
feedstocks. The energy balance simulations are based on fluidized bed reactors which will be
used for both the hydrogasification of biomass and the decomposition of methane. The reaction
temperature in the HPR is controlled by adjusting the temperature of the inlet process gas from
the gas heat exchanger which recovers the heat in the hot process gas from the MPR.
The heat required for the MPR is to be provided by circulating alumina particles between
the fluidized bed MPR and a riser combustor. Other fuels such as methane or by-product carbon
are burned in the riser to heat the alumina particles. The hot panicles then enter the MPR where
their sensible heat is transferred to the entering gas. The cooled particles are then returned to
the riser combustor for reheat. Detailed design specifications for the HPR, MPR, reheater, and
other equipment for the pilot plant are being developed.
6-36
-------
Sewage
Sludge 17.6 kg/h
JJimestone 3.3 kg/h
12.5 kg/h
8.3 kg/h
I
Greenwaste
CII4
DG
Ash/Char
17 kg/h
DRYER
150 C
GW 93.2 kg/h
(15.lt moist)
Methanol Hater
76.9 kg/h 14.2 kg/h
935 C
Purge Gas
O.oi kmol/h
CH4 14.9 kg/h
52.2 kmol/h
C0 7,2%
C02 0.7
CII4 29.6
1120 7.2
H2 55.3
HB
48.9
kmol/h
CO 2.9%
C02 3.3
CH4 23.7
H20 0.1
H2 69.1
Carbon Black
25.9 kg/h
57.7 kmol/h
CO 9.8%
C02 0.4
CH4 20.1
H20 3.9
H2 65.9
52.1 kaol/h
CO 2*7%
C02 3*1
CII4 22.3
H20 1.6
H2 64.9
MeOH 5.4
MBR
260 C
6-37
-------
SUMMARY AND CONCLUSIONS
The process design and analysis as well as experimental work on the kinetics of
hydrogasification of biomass and methane decomposition for a pilot plant using the Hydrocarb
process to produce methanbJ from biomass are reviewed. The hydrogasification kinetic
experiments showed an overall conversion of 88 to 90 wt percent of biomass in the HPR. The
gas residence time required for the reaction in the MPR to reach an equilibrium composition was
found to be 2 min at 50 atm and 900°C. Process simulations show that a process having the flow
sequence from the HPR to the MPR to the MSR to the condenser, produces more methanol and
provides a higher thermal efficiency than the alternative cycle. A minimum methane co-feed is
required for processing greenwaste, and depends on the moisture content of the greenwaste.
Sewage sludge and digester gas from a municipal sewage treatment plant can be co-processed
with greenwaste by the Hydrocarb process. The CO, in digester gas is convened to CO in the
HPR and MPR so that more methanol can be produced in the methanol synthesis reactor. The
optimum process conditions were determined to be a pressure of 50 atm in the total system and
temperatures of 900 and 1000°C in the HPR and MPR, respectively.
REFERENCES
(1) Steinberg, M. The Direct Use of Natural Gas (Methane) for Conversion of Cafrmaceous
Raw Materials to Fuels and Chemical Feedstocks. In: Hydrogen Systems,Volume 2,
Proceedings of the International Symposium, Beijing, China. Pergamon Press, Oxford,
1986. pp. 217-228.
(2) Steinberg, M. and E.W. Grohse. HYDROCARB-M"" Process for Conversion of Coals to
a Carbon-Methanol Liquid Fuel. BNL-43569, Brookhaven National Laboratory, Upton,
NY, 1989.
(3) Steinberg, M., E.W. Grohse, and Y. Dong. A Feasibility Study for the Coprocessing of
Fossil Fuels with Biomass by the Hydrocarb Process. EPA-600/7-91-007 (NTIS DE91-
011971), U.S. Environmental Protection Agency, Research Triangle Park, NC, 1991.
(4) Steinberg, M. Biomass and Hydrocarb Technology for Removal of Atmospheric CO*
BNL-44410, Brookhaven National Laboratory, Upton, NY, 1991.
(5) Borgwardt, R.H., M. Steinberg, E.W. Grohse, and Y. Dong. Biomass and Fossil Fuel to
Methanol and Carbon via the Hydrocarb Process. Energy Biomass Wastes, 15: in press,
199L
(6) Borgwardt, R.H. A Technology for Reduction of CO, Emissions from the Transportation
Sector. Energy Conversion & Management. 33* No. 5-8, pp. 443-449, 1992.
(7) Kobayashi, A. and M. Steinberg. Hydropyroiysis of Biomass. BNL-47158, Brookhaven
National Laboratory, Upton, NY, 1992.
(8) Kobayashi, A. and M. Steinberg. The Thermal Decomposition of Methane in a Tubular
Reactor. BNL-47159, Brookhaven .National Laboratory, Upton, NY, 1992.
6-38
-------
Paper 6-F
EPA'S COST-SHARED SOLAR ENERGY PROGRAM
by: Ronald J. Spiegel
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
ABSTRACT
The objective of this program is to establish and demonstrate solar energy
cost-shared commercialization projects to demonstrate how they can be used to
displace fossil fuels. The program will also have a major impetus to validate the
ability of solar energy to be used as a pollution mitigation technology. Further, the
demonstrations will assist in removing obstacles to the marketplace for solar
technologies by assisting in quantifying environmental concerns. This paper discusses
a project which has just commenced in the area of photovoltaic demand-side power
supplies. Additional discussion is provided relative to future projects which are being
contemplated. .
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative policies and approved for presentation
and publication.
6-68
-------
INTRODUCTION
Growing use of fossil fuels has generally contributed to environmental .
.. degradation including such problems as acid rain and the greenhouse effect (the
buildup of global warming gases such as carbon dioxide, carbon monoxide, and
methane). It is, therefore, imperative to investigate technologies that have potential
to prevent or mitigate these environmental problems and also make sense from
energy, engineering, and economic viewpoints. Solar energy technologies meet these
criteria. Because of solar energy's huge potential and the clear environmental benefits
from its usage, EPA intends to establish and demonstrate the practical potential of
solar energy appfications that are close to being commercialized and thereby provide
added impetus to their commercialization and, at the same time, utilize the
technology as a pollution reduction strategy or to comply with potential environmental
concerns.
The general goal of the program is to.demonstrate the technical and economic
feasibility of solar technologies as a pollution mitigating energy replacement of fossil
fuels. The general approach is to install and monitor solar energy systems at
different geographic locations in the United States for users of retail electricity who
are interested in having such systems on their premises. These systems should supply
electrical power directly to the user, where they have to compete only with the retail
cost of electricity. These systems would be monitored for performance and reliability
using on-line remote monitoring all tied into a central facility at EPA's Air and Energy
Engineering Research Laboratory (AEERL) at Research Triangle Park, NC. Participants
in the demonstration projects will be required to have considerable experience in the
selected technology, to select an appropriate solar energy system, to determine the
requirements for the technical and economic aspects of the technology to be a
pollution mitigating energy strategy, to select appropriate host sites, to conduct the
required demonstration activities, and to cost-share at least 50% of the project (at
least 30% for companies qualifying as small businesses).
PHOTOVOLTAICS (AN EPA PROSPECTUS)
Within the solar area, perhaps the most promising and potentially ubiquitous
energy option is based upon photovoltaic (PV) conversion, the transformation of
solar radiation directly into electric power. Tremendous opportunities exists for solar
PV technologies to assist in meeting the energy needs of the 1990's and beyond. The
costs have come down dramatically (about 68% per decade) since solar cells were
first used in space. The current levelized cost of energy for PV is around 30 to
35c/kWh, with costs expected drop to the 4 to 7e/kWh range by the year 2010 [1].
Thus, in the domestic bulk power markets, head-to-head competition with gas- and
coal-fired plants before 2010 will be tough. However, even at today's costs there is
an untapped remote (off-grid power supplies) world market of perhaps 200-300 MW.
6*69
-------
For grid-connected applications, nearer term opportunities appear to be on the
customer (retail) side of the meter, where load growth could be met with PV demand-
side and distributed power supplies. For example, summer peaking rates can range
from 11 to 440/kWh which would imply that for demand-side power supplies to shave
peaking loads on buildings could be cost effective today in some locations. The utility
companies or customers could own the on-site PV power generators.
Another near term possibility for grid-connected PV is distributed power
sources for standby generation and dispatchable power for the utility grid. For load
centers where demand growth has strained local transmission and distribution
capabilities, PV distributed power supplies could be an option to upgrading a
substation or constructing a new substation. In fact, it has been reported [2] that
current PV systems could be marginally competitive today in certain locations. For
these sites, avoided utility costs would accrue due to savings in equipment capacity
additions/upgrades (generation, transmission, and distribution) and in increased
reliability of customer service. That study also found that, while costs somewhat
exceeded benefits today, by 1995 the benefits provided by PV distributed systems
would exceed costs. This projection, of course, assumes that the levelized cost of
energy for PV will decline. "!" . .
With the potential increase of electric vehicles (EVs) in the near future, solar PV
carports are expected to provide daytime recharging for EVs. The primary
recharging stations for EVs will be the users' homes, with batteries charged at night.
These daytime recharging stations could also be grid-connected to provide solar-
generated electricity to buildings when the vehicles are not plugged-in. There is a
good match between the peak electricity provided by the recharging station and
daytime recharging needs of the EVs. Of course, the number of recharging stations
will depend on how many EVs are ultimately on the road and how inexpensive PV units
become.
The future environmental impact of PVs is difficult to estimate at this time. The
U.S. Department of Energy (DOE) [1] has projected that PV market penetration in the
U.S. by the year 2010 could range from approximately 6 to 27 GW. Using a 27.5%
capacity factor for PV and offsetting emissions from electric power generation using
a coal-fired plant, it can be shown that for each gigawatt of PV generating capacity,
approximately 2.2 million tonnes less of carbon dioxide (CO2) is emitted to the
atmosphere per year. Thus, if these market projections are correct, by the year
2010 a yearly reduction of emissions of CO2 could be achieved by PV on the order of
13 to 59 million tonnes.
Thus, EPA's R&D efforts will tend to be in the PV systems application
development arena, where DOE traditionally has not allocated much funding. EPA's
future activities are likely to be in: demand side power supplies, distributed power
6-70
-------
supplies, integrated building applications, daytime recharging stations for EVs, and
advanced control systems for PV/hybrid systems.
CURRENT PROJECT (DEMAND-SIDE MANAGEMENT)
Demand-side management (DSM) activities by utilities have been around for a
while. Most utilities conduct their own DSM programs, but some have turned to
energy service companies to acquire load management capabilities. Usually, peak
demand reduction is their primary goal, but energy reduction has also been targeted.
However, most utilities have been somewhat reluctant to commit to demand side
power supplies, especially new and emerging technologies, as a means for load control
(DSM). To fill that void, EPA has embarked on a project to install and monitor PV
energy systems in several geographical locations in the U.S. for retail-end users of
electricity. The goal of the project is to investigate how PV technology may be used
as a pollution mitigating energy replacement of fossil fuels by reducing electrical
demand by commercial and residential buildings. The project requires that the PV
systems:
be located on the user's premises;
supply electric power directly to the user;
reduce the electrical demand by commercial and/or residential buildings;
compete only with the retail cost of electricity;
be modular and replicable;
be capable of generating 5 to 15 kW at each site; and
obtain performance data over a complete heating and cooling season
and transmit the data to EPA.
In August, 1992, a contract is expected to be awarded to Ascension
Technology, Lincoln Center, MA, to conduct the study. Additionally, Ascension
Technology has 10 utility partners that will provide significant co-funding, along with in-
kind support. Seventeen PV systems (eight commercial buildings and nine residences)
will be installed and monitored at various sites in the U.S. The actual site selections
have not been finalized yet. The total PV power involved is 116 kW. Applying the
previously stated yearly emissions reduction potential of PV (1 GW reduces CO2
emissions by 2.2 million tonnes), this project will produce a reduction of 255 tonnes of
COa per year.
The total environmental benefits of DSM using PV power supplies could be
significant. For example,' if 50 million residential users in the U.S. installed 4 kW rooftop
PV generators to produce daytime electricity, the electricity produced could displace
200 GW of central station generated electricity. This calculation assumes that this
offset electricity is produced from coal to estimate an upper value for the pollution
reduction potential of PV power supplies in a DSM role. Obviously, the offset
6-71
-------
electricity could be produced by a complicated mix of gas, oil, coal, and nuclear
plants; With this assumed, over 400 million tonnes less C02 would be emitted yearly.
. FUTURE PROJECT (PHOTOVOLTAIC - ELECTRIC VEHICLE RECHARGING STATIONS)
Electric vehicles (EVs) are rapidly becoming a focus of public attention [3].
Legislation has been passed in California and Massachusetts, and is pending in 10
other states, to require a specified percentage of automobile sales in these states to
be EVs by 1998. The impetus behind this legislation is that EVs offer the promise of
cleaner air to cities. In the United States today there are 185 million gasoline fueled
vehicles on the road with an additional 7 million added annually. Total vehicle
kilometers traveled doubled in the last two decades from approximately 1.6 trillion in
1970 to 3.2 trillion in 1990, and current estimates predict another doubling by the
year 2000. The number of cars on American roads and the kilometers they travel are
increasing much faster than the human population. Although the United States has
the most stringent emission regulations in the world, motor vehicles account for
approximately 35% of total volatile organic compound (VOC) emissions, 40% of total
nitrogen oxides (NOX) emissions, and over 65% of total carbon monoxide (CO)
emissions. *
Adoption of EVs by the public would necessitate changing production (car
manufacturers) and buying (consumer) patterns. Several recent legislative policies
may provide catalysts for these changes. For example, the California plan does
mandate EVs: 2% of all cars sold in California in 1998, rising to 10% by 2003. If the
eight Northeast States for Coordinated Air Use Management adopt the California
plan, between 250,000 and 300,000 EVs could be on U.S. roads by 2003. Also,
several large automobile manufacturers (e.g., GM, Ford, Chrysler, Nissan, and BMW)
have major EV projects ongoing. The GM Impact has changed the vision of EV
performance to that of high performance sports cars [3]. It is anticipated that a
significant increase in EV sales could begin as early as 1995.
The environmental benefits from EVs are difficult to quantify at this time. A
study conducted by the South Coast Air Quality Management District [4] showed that
by replacing 5% of conventional gasoline fueled vehicles with EVs in the South Coast
Air Basin would reduce yearly VOC and NOX emissions by approximately 8950 and
6600 tonnes, respectively. However, California would probably benefit more than
other states because their electricity is produced primarily by natural-gas fired steam
plants. In other areas, there could be an increase in emissions associated with a
growth in electricity consumption and production. However, many of the power
plants may be located in less environmentally sensitive geographical locations, and if
EV recharging occurs at night, emissions will be produced at off-peak ozone forming
hours.
6-72
-------
It is generally assumed that most EV batteries will be charged at night when
demand for electric power is low. However, it is almost a certainty that some
infrastructure must exist for daytime recharging, especially for commuter vehicles.. If
EVs are bought in large numbers, driving patterns are likely to be similar to internal
combustion vehicles, which-will require daytime "refueling." EVs charged with power
from PVs would truly be zero emission vehicles.
EPA is currently contemplating a PV recharging scenario for EVs. While no
decision has currently been reached, the likely approach will be to install two or three
PV recharging stations in non-attainment ozone metropolitan areas. One year's data
would be gathered to demonstrate that EVs and a PV recharging facility can be a
transportation option to reduce greenhouse gases.
CONCLUSION
EPA has embarked on a study to determine the environmental benefits potential
of DSM using PV. The R&D focus will be in the systems application development area
as a pollution reduction strategy.
1. The Potential of Renewable Energy: An InterJaboratory White Paper, U.S.
Department of Energy, SERI/TP-260-3674, DE 90000322, March 1990.
2. Benefits of Distributed Generation in PG & E's T&D System: A Case Study of
Photovoltaics Serving Kerman Substation, Pacific Gas and Electric Company, GM
663024-8, August 1991.
3. Fischetti, M., Here Comes the Electric Car - It's Sporty, Aggressive, and Clean,
Smithsonian, Vol. 23, pp. 34-43, April 1992.
4. Long Range Strategies for Improving Air Quality, South Coast Air Quality
Management District and the Southern California Association of Governments,
September 1985.
6-73
-------
Paper 6-H
ADVANCED ENERGY SYSTEMS FUELED FROM BIOMASS
by: Carol R. Purvis and Keith J. Fritsky
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
ABSTRACT
The concentration of carbon dioxide (CO2), a greenhouse gas, is increasing by
an estimated 0.5 percent per year. CO2 emissions from fossil fuel combustion
quadrupled between 1950 and 1980. Conversion of renewable biomass to energy is
CO2 neutral and produces lower sulfur dioxide (SO2) and nitrogen oxides (NOJ
emissions than fossil fuel combustion.
- The U.S. Environmental Protection Agency/Air and Energy Engineering
Research Laboratory is studying two biomass conversion technologies: conventional
combustion in a boiler coupled with'a steam turbine system and gasification in a -
gasifier coupled with an aeroderivattve turbine system. In-house research is
addressing the problems encountered in conventional systems with regard to
emissions, tui>e fouling, bed agglomeration, and low thermal efficiency. Extramural
research is addressing the problems of advanced systems with regard to
fixed/fluidized-bed gasifiers, alkali/particulate cleanup, gas compatibility with turbines,
and system efficiency. The results will provide data for owner/operators to improve
system performance and for designer/developers to demonstrate advanced systems.
This research will help promote biomass-for-energy as a global warming mitigation
strategy by focusing on a need to maximize biomass resources through increased
utilization efficiency. '.,-
*; - » * " ', '
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
6-92 -
-------
INTRODUCTION
The eading culprit in the greenhouse gas arena is carbon dioxide (CO2), and its
concentration is increasing by an estimated 0.5 percent per year [1J. Anthropogenic
activities are responsible for the unwanted buildup of C02 in the atmosphere. CO2
emissions from fossil fuel combustion quadrupled between 1950 and 1980 alone;
therefore, reducing fossil fuel C02 emissions is an obvious place to start mitigation
activities [1]. The efficient use of biomass is CO2 neutral and can reduce pollution by
lowering sulfur dioxide (S02) and nitrogen oxides (NOJ emissions. Figure 1 is a
simple representation of the biomass closed carbon cycle and the fossil fuel (oil, coal,
and gas) open carbon cycle.
CLOSED
-CARBON
CYCLE
OPEN
CARBON
CYCLE
Figure 1 - Carbon Cycles
Biomass is the principal single source of energy for 75 percent of the world's
population [2]. It provides about 14 percent of the world's energy; 35 percent of the
total energy supply in developing countries, and 3 percent of the total energy supply in
developed countries (see Figure 2) (3J. This energy contribution is equivalent to 35
million barrels (5.6 x 10s L) of oil per day globally [2]. Some analyses suggest that the
amount of energy contributed by biomass could be increased by a factor of 13 [2]. By
the second quarter of the 21st century, biomass could provide 25-35 percent of the
total global power generation capacity [4J. In 1990, the United States had 9,000 MW
of biomass-fueled electric capacity, 36 times the capacity in 1980, due to the Public
Utilities Regulatory Policies Act of 1978 (PURPA) {4).
6-93
-------
FOSSIL FUELS 75%
HYDRO 6%
NUCLEAR 5%
BIOMASS 14%
WORLD
FOSSIL FUELS 86%
FOSSIL FUELS 58%
HYDRO 6%
NUCLEAR 5%
BIOMASS 3%
HYDRO 6%
NUCLEAR 1%
DEVELOPED COUNTRIES
BIOMASS 35%
DEVELOPING COUNTRIES
... ., Figure 2 - Energy Use Distribution
Increasing biomass availability on a sustainable basis may not be viewed as a
high priority in developing countries and what is appropriate in one country might not
work as well in another. But energy demands will increase in all countries and all
countries have forestry end agricultural wastes and the potential for energy crops.
Therefore, greater use or indu trial wood wastes, agro-processing wastes, and
agricultural residues will have tne greatest near term impact. The potential for the use
of c"!sting biomass resources in the world is 65,381 x 106 GJ; 41,068 x 106 GJ in
d: Ding countries and 24,313 x 106 GJ in developed countries [3]. Tfcs
di vpment and commercialization of small, residue-fueled facilities should fill the
needs of rural communities and biomass industries as well as developing countries.
Promotion of carbon sequestration and biomass utilization requires research on
various issues including land use, forest management, production, harvesting,
reforestation, urban reforestation (carbon sequestration and mitigation of the urban
heat island effect), displacement of fossil fuels with biomass-based t.ectricity and
liquid fuels, and increased use of wood products. The U.S. Department of Agriculture
(USDA)/Soil Conservation Service, USDA/Forest Service, U.S. Department of Energy
(DOE)XOffice of Conservation and Renewable Energy, U.S. Environmental Protection
Agency/Air and Energy Engineering Research Laboratory (EPA/AEERL), National
Renewable Energy Laboratory, and Oak Ridge National Laboratory are some of the
domestic organizations studying these issues.
6*94
-------
CONVERSION TECHNOLOGIES
In ethanol production from woody energy, crops, 75 percent of the energy
ultimately produced can actually count toward fossil fuel displacement. When woody
crops are directly convened to electricity, this figure can increase to 90 percent [1J.
This 15 percent increase indicates that the conversion technologies with the greatest
near term potential will focus on direct biomass-to-energy production. This near term
production should occur through utilization of residues in small stand-alone energy
production facilities.
EPA/AEERL's goal is to promote activities to achieve renewable production,
provide modern energy carriers, develop high-efficiency conversion processes,
develop high efficiencies for end use, and develop technologies that have favorable
economics on a small scale.
EPA/AEERL's program is studying two conversion technologies: conventional
combustion in a boiler coupled with a steam turptne system and gasification in a
gasifier coupled with an aeroderivative turbine system (see Figure 3}.
EXHAUST
EXHAUST
.CONVENTIONAL
ADVANCED
Figure 3 - Cycle Flow Diagrams
Combustion
Throughout the world, process heat/steam and electricity generated from
biomass fuels are almost entirely a result of direct combustion. Direct combustion
technologies fueled with biomass include dutch ovens, spreader-stokers, total or
suspension fired combustors, and fluidized or circulating bed combustors. Since these
technologies are firmly established, they are readily transferable to biomass energy
applications throughout the developed and developing world.
In applying these technologies to biomass fuels, problems are often
encountered with regard to unacceptable, uncontrolled emissions, boiler tube fouling,
bed agglomeration, and less than optimum thermal efficiency. These problems arise
as a result of the combustor's inability to respond to rapid variations in certain
properties of the fuel feed (moisture content, particle size, and inorganic content).
6-95
-------
In-house research at EPA/AEERL is addressing the above problem, and by
doing so, will provide data to owner/operators of biomass combustion systems for the
purpose of improving system performance through modifications to the combustion
process. This research will help promote biorhass-for-energy as a global warming
mitigation strategy by focusing on a need to maximize biomass resources through
increased utilization efficiency.
The approach of the research is to burn different biomass fuels (sawdust
briquettes, bagasse pellets, and switch grass briquettes) in a pilot-scale stationary
grate combustor. A first series of tests is to use fuels with relatively uniform
properties. Standard thermal and chemical analyses will be performed to quantify fuel
properties. Data obtained from this series of tests will establish a baseline and
characterize relationships between fuel properties and system performance. In a
second series of tests, fuel properties will be measurably altered (by increasing
moisture content or inorganic fraction, for instance) and the system will be monitored
and adjusted to op: mize combustor performance under adverse operating conditions.
The purpose of performing IMS series of tests will be to minimize emissions of criteria
and non-criteria pollutants -- carbon monoxide (CO), NO,, paniculate matter (PM), and
volatile organic compounds (VOCs) - and fouling of heat transfer surfaces while
maximizing thermal efficiency by controlling air flows, combustion air temperature, etc.
Techniques for controlling the combustion process will be developed so thaHhey can
be applied by operators of biomass combustion systems. Testing is anticipated to
begin in June 1993 when construction of the combustor facility is complete.
Gasification
Gas turbines fueled with biomass gas offer a great advantage in the 10 ^ 50
MW range. Net plant conversion efficiencies could exceed 50 percent, especially if
the turbine exhaust is used in a heat recovery steam generator or an air bottoming
cycle [4], Aeroderivative gas turbines (i.e., gas turbines derived from aircraft jet
engines) offer high efficiency, low unit capital costs at modest scales, and low
maintenance due to their compact modular nature.
Fixed-bed Gasifiers
Temperatures of the gas exiting the gasifier are expected to be 500-600°C for
the fixed-bed updraft design. Within this temperature range, the alkali compounds
(formed primarily from potassium and sodium in the feedstock) appear to condense on
PM and can be controlled by paniculate collection devices. The temperature is high
enough that the tars formed will remain in the vapor phase and actually boost the
heating value of the gas. It would be desirable to have the gasifier and gas turbine
dc^e-coupled to eliminate any condensation problems [4].
The State of Vermont was funded by EPA/AEERL, DOE, and U.S. Agency for
International Development to evaluate the compatibility of gasified biomass feedstocks
with an aeroderivative gas turbine power generation system. The EPA/AEERL is also
6-96
-------
funding Vermont to perform a feasibility study on demonstrating this technology. The
compatibility evaluation was performed at the General Electric Corporate Research
and Development (GE CR&D) coal gasification pilot plant. The objectives of the.study
were: ..
- to determine the composition of the product gas and gasification rate,.
- .to determine the nature of the contaminants in the gas that might have a
detrimental effect on the gas turbine,
- to measure the contaminants in the gas that might have a negative
environmental effect, and
- to determine the effectiveness of cyclones for paniculate removal.
The-gasification plant consisted of a feed system, a fixed-bed updraft gasifier,
and a cyclone (see Figure 4). The State of Vermont provided 83.8 tons (76 tonnes) of
dried wood chips and Winrock International provided 42.5 tons (39 tonnes) of bagasse
pellets.1' These highly reactive biomass fuels were gasified at 20 aim (2.03 MPa).
The bibgas product had a higher heating value than coal gas and was compatible with
gas turbine combustors. The paniculate carry over and the alkali metal contained in
the particles indicated that a single cyclone was not sufficient for cleanup. The sulfur
emissions were lower than for coal combustion facilities equipped with flue gas
desulfurization systems, and fuel bound nitrogen levels were lower than for coat [5J.
Ftera
Fuel
ni
Ash
Fuel
ftrdcutate
Oust
Figure 4 - GE CR&D's Equipment Row Diagram
'Pelletized switch grass was the third fuel considered but the pellets crumbled In shipping. Briquetting
Marketing and Services, Inc. was able to produce a switch grass briquette per the GE size specification
on a demonstration briquettor. This demonstration size briqueitor had a small throughput that would not
produce the quantity required for the test in the time allowed prior to testing.
6-97
-------
Fluidized-bed Gasifiers
The most promising long-term prospect for biomass gasification is the use of
fluidized-bed gasifiers that can accommodate a wide variety of feedstocks. Fluidized-
bed gasifiers have higher throughput capabilities and greater fuel flexibility than fixed-
beds, "including the ability to handle low-density feedstocks like undensified crop
residues or sawdust. Gas quality control with fluidized-bed gasifiers may require
different treatment for two reasons. First, the exit gas temperatures are higher, 800- .
900°C, which may result in the vaporization of alkalis. Second, there is more
paniculate carry-over, which cyclones alone will not handle, that will require ceramic or
sintered-metal barrier filters [4],
EPA/AEERL plans to cooperate with a researcher in the development of a
system consisting of a pressurized feed system, a fluidized-bed gasifier, and a filter for
gas cleanup. All components of the system have been designed to operate on
biomass fuel with minimal preparation and produce a gas suitable to fuel an
aeroderivative turbine. Testing of the system will begin in late '92.
REFERENCES
1. Trexler, M.C. Minding the Carbon Store: Weighing U.S. Forestry Stratec.es to
Stow Global Warming. World Resources Institute, Washington, DC, 1991.
2. Hall, D.O.. and Woods, J. Biomass: Past, Present and Future. Jn: Proceedings
of the Conference on Technologies for a Greenhouse-Constrained Society. Lewis
Publishers. Chelsea, Michigan, 1992.-
3. U.S. Congress, Office of Technology Assessment, Fueling Development: Energy
Technologies for Developing Countries. OTA-E-516 (Washington, DC: U.S.
Government Printing Office, April 1992).
4. Williams, R.H., and Larson, E.G. Advanced Biomass Power Generation -The
Biomass-Integrated Gasifier/Gas Turbine and Beyond. jn: Proceedings of the
Conference on Technologies for a Greenhouse-Constrained Society. Lewis
Publishers, Chelsea, Michigan, 1992.
5. Furman, A.H., Kimura, S.G., Ayala. R.E., and Joyce, J.F. Biomass Gasification
Pilot Plant Study. Draft report prepared for the State of Vermont by GE Corporate
Research and Development, Schenectady, New York, 1992.
6-98
-------
-------
APPENDIX 2
Additional AEERL Journal Articles and Book Chapters
Relating to
EPA's Global Climate Change Research Program
-------
-------
Technological Considerations
for Planning the Global Car-
bon Future
by
Robert P. Hangebrauck
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park. NC
ABSTRACT
The atmospheric level of carbon dioxide (CO,) is the dominant
variable in the anthropogenic influence of future global climate
change. Thus, it is critical to understand the long-term factors
affecting its level, especially the longer-range technological
considerations. Most recent analyses of the problem focus on
the next 20 to 100 years. While great uncertainties exist in
longer-term projections of CO2, it is of concern that the longer-
range buildup (over many centuries) may be of substantial
magnitude and might be so despite major efforts to reduce use
of fossil reserves for energy, save world forests, and/or collect
and dispose of CO2. This paper summarizes some of the recent
literature relating to the longer-term CO, problem and discusses
some of the technological considerations for known prevention
and mitigation approaches in the context of this longer-term
problem. These approaches include: renewables (solar photo-
voltaics, wind, and biomass), conservation, flue-gas and fuel
CO, sequestration via disposal on land or in the ocean, carbon
recycling (chemical/biological utilization), and atmospheric CO2
fixation/utilization via terrestrial and marine approaches. These
are discussed along with other strategies to identify'those that
1) could be major factors in preventing long-term CO2 buildup,
2) would be environmentally sound but likely to have more
limited long-range CO2 impact, 3) would be environmentally
uncertain or uncertain for other reasons, and 4) would be
environmentally questionable or unlikely. solutions for other
reasons.
KEYWORDS
Carbon dioxide; carbon: technology; global warming; preven-.
tion; mitigation
Introduction
The potential for a long-range CO2 buildup has recently
been reported (Hasting and Walker, 1992; Walker and
Kasting, 1992; Kheshgi, 1989; Hoffert et al., 1979).
"Long-range" in the context of these studies is several
centuries into the future. However, most recent analy-
ses of prevention of CO2 impacts on global climate have
focused on a relatively near-term time frame - gener-
ally, out to the years 2030,2050, or perhaps 2100. The
To be published in "Energy
Conversion and Management."
problem is that shorter-term evaluations may miss the
resulting long-term C02 buildup noted by the previ-
ously cited investigators. While great uncertainties
exist in these projections, it is of concern that this
longer-range buildup could be of substantial magni-
tude ~ on the order of 500 to 2800 ppmv. Furthermore,
these concentrations might be achieved despite major
efforts to reduce use of fossil reserves for energy (Kast-
ing and Walker, 1992) or collect and dispose of C02
(Hoffert et al., 1979). In addition, the recovery time for
atmospheric CO2 concentrations to return to
preindustrial or even current levels could be in the
range of thousands to hundreds of thousands of years,
if at all. While accepting the uncertainties of making
such projections, it is of substantial interest to explore
the technological options that might work under the
requirement for drastic reductions in release of fossil
carbon. This paper is aimed at exploring this problem
and potential solutions.
Long-Range CO2 Buildup
In 1990 the atmospheric concentration of C02 was 354
ppmv (ORNL, 1991). It is projected to increase rapidly
in the future with increasing emissions (IPCC, 1992).
In the absence of constraints, the CO2 emission rate can
be expected to continue the rapid growth indicated in
Figure 1. However, Figure 1 includes only fossil
energy emissions. Contributions to the atmosphere
also are made by deforestation and volcanic activity.
Figure 2 adds estimates for the historic emissions from
deforestation (net biota flux). Note that the estimates
associated with deforestation are highly uncertain.
Figure 3 illustrates the magnitude of current loadings
6000
.2000
1000
~^r
I
1860 1880 1900 1920 1940 1960 1980 2000
Year
Figure 1. Annual global CO, emissions for fossil fuel ;
combustion, cement production, and gas flaring, 1860 to
1989 (ORNL, 1991).
-------
9000
7 T
1860 1870 1880 1890 1000 1(10 1920 1830 1940 1860 1060 1970 1880 I860
VMT
Figure 2. CO2 emissions (or both fossil energy and
deforestation. Data derived from ORNL (1991} for fossil
energy; Kasting and Walker (1992) for deforestation.
from fossil fuel, deforestation, and volcanic outgassing,
and the values include both low and high estimates,
reflecting the upper and lower bounds of uncertainty.
Figure 4 charts the increase in atmospheric concentra-
tion of CO, against cumulative emissions from all
sources starting with the year I860, and shows an
almost linear relationship for the most recent emission
accumulations. However, as a function of time, both
cumulative emissions and atmospheric concentrations
are increasing more rapidly. For the 100 years from
1860 to 1960 about 217 GtC was added to the atmo-
sphere and the atmospheric concentration of C02 in-
Deforestation
Fossil fuel
*
g
s
B
o
I
360-
350-
340-
a on
320-
31O-
30O-
290-
1C
*
1 ' t
1 j>
f
9 years
mission!
tm. Cora
J>
+217 C
. +29 1
f
...
1C
xnv
I**
*. -
I
*"
i
f .
U^:
JO y»an
>Emtesi<
Aim. C<
yt
ns +21
nc. +3
^X
.
'GtC
ppmv>
K
i
X"
Volcanic
outgassing
Figure 3. High and low estimates for current annual
emissions of CO2 reflecting the upper and lower bounds of
uncertainty. Data derived from ORNL (1991); IPCC (1992);
Houghton (1991); and Kasting and Walker (1992).
creased by 29 ppmv. For the 30 years from I960 to
1990, another 217 GtC was added to the atmosphere,
and the atmospheric concentration increased by 37
ppmv. Note that the scatter in the points plotted in
Figure 4 is not a true reflection of the variability in the
data; it results from plotting measured atmospheric
concentrations (ORNL, 1991} versus the total emis-
sions. The total annual emissions used to derive
cumulative emissions in Figure 4 include some reason-
ably good estimates of the fossil-fuel
component (ORNL, 1991). However,
the deforestation (land use change)
component values, as mentioned pre-
viously, are highly uncertain esti-
mates. Note the range of annual val-
ues in Figure 3 (IPCC,
Houghton, 1991).
1992;
50 100 150 200 250 300 350 400 450
Cummulative CO2 Emissions, GtC
Figure 4. Atmospheric concentration of CO2 as a function of cumulative CO2
emissions (fossil energy and deforestation). Total emissions for each year are
from Figure 2.
Moving to the longer-term picture,
future atmospheric CO2 loadings de-
pend substantially on the rate at which
the fossil hydrocarbon reserve is ex-
pended for energy use. However, as
we will see later, even with drastic
cutbacks in the use of fossil energy,
CO, emissions are still high enough to
produce major elevations in atmo-
spheric CO2 levels. Figure 5 illus-
trates some scenarios for consump-
tion of the fossil energy reserve. Three
of the curves are based on a logistic
function similar to that used by
-------
sooo
4500
4000
3500
3000
I
2500
2000
1500
1000
500
Z
igictic rite,
'. InHiil a ewth
LogMte, 1
1962 c Hnumptloi
LogMI:, 0.5%
0.24 < tO/yaw «ft r 2050
1900
.2000
2100
2200
2300
YEAR
2400
2500
2600
2700
Figure 5. Scenarios for consumption of the loss!! energy reserve.
Siegenthaler and Oeschger (1978) and Kheshgi (1989)
with three different initial rates. A fourth curve as-
sumes an immediate cutback to the 1982 rate of use
and holding at this rate until the. entire fossil energy
reserve is expended. A fifth curve assumes a drastic
cutback to 0.24 GtC/year after 2050. Examining these
scenarios provides some perspective on the time peri-
ods (centuries) over which the fossil energy reserve
would be expended including some cases where fossil
energy would be used at much lower rates than are
characteristic of present trends. The straight line
representing an immediate cutback of fossil fuel emis-
sions to 1982 levels (and a hold at that level) is roughly
consistent with the most stringent mitigation scenario
included in the IPCC Climate Change 1992 report
(IPCC,1992), but the IPCC scenario covers only the
period to the year 2100.
3000
2500
"2000
1500
1000-
500
1800 2200
Figure 6 includes various long-range
projections of atmospheric CO, con-
centration. Estimates modeled by
three investigators are plotted to show
examples of the possibilities. Bach of
the investigators used different sce-
narios for the consumption rates for
the fossil energy reserve. Hoffert et
al. (1979) assumed a higher initial
growth rate, 5%, than any shown in
Figure 5. Many factors can affect
such projections. Kasting and Walker
{1992) and Walker and Kasting (1992)
showed a dramatic effect from saving
the forests versus burning the forests,
which is not incorporated in Figure 6.
Long-term ocean uptake is a function
of several factors, for example: ocean
upwelling velocity, phosphate in run-
off, carbon to phosphate ratio, and alkalinity of runoff
(Kheshgi, 1989).
This carbon imbalance problem suggests the need for
R&D to provide enhanced long-range modeling capa-
bility for the biogeochemical system and a technologi-
cal base for preventing the increasing global emissions
consistent with economic growth. The issue remains as
to what is acceptable change, but that is not the focus
of this paper. As noted earlier, this paper is aimed at
exploring the nature of the problem and potential
solutions. The cause of the carbon imbalance is of
course the steadily increasing loading of anthropogenic
C02 emissions, which now far outweighs the capacity
for timely biogeochemical uptake. Kasting and Walker
(1992) and Walker and Kasting (1992) point out that,
in times prior to substantial human influence, the
atmosphere-ocean system had to deal with only the
relatively low emissions from volcanic outgassing of
CO4. Thus, current annual CO2 inputs are 50 to 100
timeshigher than the long-term input rate of CO2 to the
atmosphere-ocean system.
The long-range mismatch between sinks of CO, and
sources of CO2 is illustrated in Figure 7 using sink data
derived from Kasting and Walker (1992) and source
data from Figure 5 (2% logistic rate curve). A rather
optimistic net photosynthesis uptake of 3 GtC was
assumed at year 50.
Given the potential for long-term carbon imbalance,
what levels of reduction in carbon emissions may be
4600 sooo necessary to minimize future'atmospheric C02 eleva-
tions and how can they be achieved?
Figure 6. Long-range projections of atmospheric C02
concentrations modeled by three investigators.
-------
N»1 nmr uptik* capicHy, QIC
1000000
10000
1000
100 -
10 <
1 .
Carbon released from ^ *<**"""
TO&SII Tuei reserve
\
\ "
I-- 1 if
L--- hi
./ a
y
V
n
ff
j
5
i
I
I
t.,
.
1 10 1 100 1000 10000 100000 1000*
Figure 7. Illustration of the mismatch between sinks of CO2
and sources of CO2.
Restoring the Carbon Balance
Hasting and Walker (1992) note "because the removal
processes for atmospheric CO2 are slow, the maximum
CO2 level reached is relatively insensitive to die fossil
fuel burning rate unless the burning rate is many times
smaller than its present value," and estimate that
major reductions in carbon emissions would be re-
quired to avoid substantially elevated future atmo-
spheric CO, concentrations. Figure 8 derived from
Hasting and Walker (1992) illustrates the rather
modest difference in CO2 concentration/time profiles
for a high rate of fossil fuel consumption versus a low
one, which holds fossil energy releases constant at the
currentemission rate. To return to preindustrial levels
1200
1100
200
1800 2200 2600 3000 3400 3800 4200 4600- 5000
Y«ar.
Figure 8. Illustration of the insensftivtty of long-range
projections of atmospheric CO2 concentrations to the rate of
consumption of fossil energy. Both calculations assume that
the forests will be a substantial sink for CO..
would mean cutting emissions by a factor of 100. Even
where optimistic assumptions are made about forests
(they are saved and enhanced growth results from C02
fertilization) and where fossil combustion emissions
would immediately be cut by a factor of 25, the C02
concentration level would increase for approximately
the next 17,000 years, eventually reaching 500 ppmv.
Cuttingback currentfossil fuel emissions by a factor of
25 means cutting the current emission rate of approxi-
mately 6 GtC/year to approximately 0.24 GtC/year.
This would still be 2.4 times greater than preindustrial
volcanic outgassing, estimated at about 0.1 GtC/year.
Complicating factors are involved here beyond the
need to stem the carbon buildup in the atmosphere.
First, increasing energy needs over the globe will have
to be satisfied at the same time carbon releases are
being decreased. Second, there is a need to preserve
concentrated carbon resources for essential uses. Es-
sential uses might include 1) special energy needs with
limited alternative options such as certain transporta-
tion fuels; 2) essential chemicals and chemical feed-
stocks for a variety of petrochemicals, plastics, rubber,
and Pharmaceuticals; and 3) energy-related products/
by-products used for a variety of purposes such as coke,
roofing, and paving. If the current usage of fossil
hydrocarbons is assumed to be 8.3% for such purposes
(UNEP, 1990; JCAE, 1973), then about 0.55 GtC^ear
of the world fossil reserve would be required currently.
Assumingthat the intensity of use of such products will.
increase with population and increased standard of
living (e.g., in developing countries), then we might
assume a growth in this type of use of 1 to 2% per year.
Further, assuming that such growth might continue
until theyear 2100 and then level off at that rate, about
1.6 to 4.6 GtC/year would be used for such purposes.
Nearly all of this carbon could eventually be released to
the atmosphere. Considering the 25-fold reduction
target of 0.24 GtC/year, much of this carbon would need
to be recycled or permanently sequestered.
The Prevention and Mitigation Options
Before discussing the mitigation options, the following
paragraphs cover the considerations which are taken
to be of major importance in this assessment:
Long-term mitigation potential
Environment and energy/resource use
Costreffectiveness
Level of development
Probability of success
Perhaps the primary consideration in assessing the
technology options is the potential for major reduction
-------
or elimination of CO, emissions. This assessment
assumes that drastic reductions in C02 emissions are
required now, carrying on into the future. The levels of
reduction required are assumed to be at least equiva-
lent to the 25-fold reduction in current emission levels
by the year 2050 consistent with one of the reduced
impact scenarios evaluated by Kasting and Walker
(1992). This would require extraordinary levels of
control applications to sources, many of which would
not be very amenable to control. Figure 9 illustrates
that currently only about 40% of fossil fuel C02 emis-
Cement
Gas flaring
Figure 9. CO., emissions resulting from the combustion of
various portions of the fossil hydrocarbon reserve (ORNL.
1991).
sions result from combustion of solid fuels. Globally,
only a portion of the solid fuels are burned in large
power plant boilers. Coal-fired power plants contrib-
ute only about 30% of the U.S. emissions of C02 (Johnson
et al., 1992). Much of the fossil fuel combustion occurs
in the transportation, industrial, and residential/com-
mercial sectors. These sources are generally smaller
and not likely to be ready candidates for C02 recovery
and disposal, even if such technology were deemed to be
acceptable. Thus, it is assumed that it is more likely
that such sources will be mitigated via.reduced direct
use of fossil fuels.
The technology options for consideration are:
* Renewables
- Solar photovoltaics
- Wind . ' .'.. .'""'"
- Biomass
Conservation . . . - -
- Increased end use efficiency
Increased generation efficiency
Fuel substitution
Flue-gas and fuel C02 sequestration
- Separation/concentration
- Transport
- Disposal on land or in the ocean
- Carbon recycling (chemical/biological
utilization)
* Atmospheric C02 fixation/utilization
- Terrestrial ,
- Marine
Hangebrauck et al. (1992) includes a more extensive
discussion of the specific C02 sequestration approaches
covered in the last two "bullets." Nuclear (fission and
fusion) options may also play a role long term but are
not assessed in this paper.
To help structure a discussion of the global carbon
future, the various technology options and related
strategies are categorized and discussed as follows: 1)
those that could be major factors in preventing long-
term CO2 buildup, 2) those that would be environmen-
tally sound but likely to have a more limited long-
range C02 impact, 3) environmentally uncertain or
uncertain for other reasons, and 4) environmentally
questionable or unlikely solutions for other reasons.
Could Be Major Factors in Preventing Long-
term CO2 Buildup
As discussed previously, saving the existing forests,
versus burning them, can have a major impact on
reducing the long-term peak atmospheric CO2 concen-
tration. For example, Kasting and Walker (1992) show
a reduction in the peak future concentration from 2000
to 1100 ppmv. However, despite current efforts, the
deforestation trend is still not reversed. Therefore, it
is remains important to
Slow net deforestation
Next, the technologies which seem to have the most
new potential here are renewables and recycling in-
cluding:
Solar photovoltaics (terrestrial-based and perhaps
space-based)
Solar (passive and thermal)
Wind
Increased terrestrial biomass in conjunction with
use for energy production
Recycling or utilization of CO,
Reaching a sustainable use of fossil carbon will obvi-
ously require major restructuring of how we use the
fossil hydrocarbon reserves in conjunction with a grow-
ing renewable energy supply sector. Figure 10 pro-
vides an illustrative example of how this might be done.
-------
Wind
Solar
Biomass ^v
*
Fossil Hydrocarbons -.
(gas, oil, coal) * °
» He
/T"
H2/
K2
H2
1 '
at, Power, and
insportation Energy
Chem
Convi
Hydro
i
II and Coal Refining - . -
I
Hydrocarbons for »
Petrochemicals
leal
srslon of
6
carbons
,
CO2
3etrochemlcal ,
Production
i
Petr
»>Feec
ProC
^chemical
(stocks &
fucts
t
Post-use Recycling
Figure 10. Example major strategic restructuring to reduce global carbon releases
and waste by emphasizing renewables and use of the fossil hydrocarbon reserve for
products versus energy.
reduce tropospheric ozone and air
toxics.
While the potential for greatly accel-
erated renewables exists, there are
now severe .production limitations
for renewable capacity (NREL,
1990). These limitations would need
to be eliminated as rapidly as pos-
sible for technologies, such as solar
photovoltaics, by accelerated KD&D
and other means (Stone, 1992). The
National Energy Strategy currently
projects only 23 exajoules (EJ) (22
quads) for the U.S. from renewables
by 2030 (DOE, 1991). Another esti-
mate suggests 50% of U.S needs
within 40 years (Brower, 1992).
The major objective would be to make renewable en-
ergy the backbone of the system while constraining and
preserving the fossil hydrocarbon reserves. Renew-
ables would be used not only for direct production of
energy but also to generate hydrogen needs for 1)
certain energy applications and 2) coal and oil refining
operations used to generate "petrochemical" products.
Even though it would be renewable-energy-use inten-
sive, wind- or solar-derived hydrogen would also need
to be used to recycle some of this CO2 to hydrocarbons.
A small, but increasing, amount of industrial C02 is
already being recycled where concentrated sources are
matched with process requirements; e.g., co-sited
methanoVammonia production. Considerable Japa-
nese research is underway to develop chemical fixation
of C02 as recycling technology (Arakawa, 1992; Arakawa
et al., 1992). Methods of fixation are being explored in
depth, and selected methods have been identified for
greater emphasis. The following list covers most of the
methods under investigation:
Photocatalytic fixation
* Electrochemical fixation
Catalytic hydrogenation
* Decomposition to carbon
* Reaction of CO2 with methane
Reaction of CO2 with hydrocarbons . .,, , .
Organic synthesis , ',',-','
Polymer synthesis , ... .....
The phase-in of renewables will substantially reduce
global, regional, and local loadings of conventional air
pollutants, including sulfur dioxide (SO2), nitrogen
oxides (NO,), carbon monoxide (CO), volatile hydrocar-
bons, and particulate matter This in turn will help to
Considerable variation exists in es-
timates of the potential for increased biomass in con-
junction with energy production; however, Lee et al.
(1991) estimate that short-term rotation energy crops
could provide an additional to 13 EJ (6 to 12 quads).
The most promising options seem to be 1) direct use via
combustion or advanced cycles and 2) variations of the
Hydrocarb process to produce methanol along with
carbon to be stored or recycled. The near-term version
BkxniM Mjrtharw
t ^f
Methanol
Carbon
Figure 11. The Hydrocarb process involves three basic
process steps which allow conversion of biomass to methanol
and carbon.
of the Hydrocarb process is illustrated in Figure 11
(Borgwardt, 1992; Steinberg, 1993). The carbon se-
questered can run from 55 to 100% for the configura-
tion shown at retail selling prices equivalent to gaso-
line prices ranging between $0.26 and $0.37/liter ($1.00
and $1.40/gallon). Longer range versions of this pro-
cess could use renewable hydrogen, allowing imple-
mentation of a fully renewable carbon cycle as illus-
trated in Figure 12. Wind-derived hydrogen would be
more cost-effective initially.
1
Methanol
Synthesis
260 °C
50x10* Pa (SO atro.
t
Hydro-
gasification
Reactor
MWC
50x10* Pa (50 aim.]
jf
Pyroiysis
Reactor
1100 "C
SOx10*Pa(SOatm.)
-------
Und-bMWJ Wind
Farms or Solar
Photovoltaic*
By-product Oj
Electrolytic
H2&02
Production
+ 1.28rV> CH..OH
Fnt-1
Figure 12. Illustration of a renewable carbon cycle.
Other options for using biomass include hydrogen
production via indirectly heated gasification (Will-
iams, 1992), gasification combined with methanol
synthesis, and advanced enzymatichydrolysis of biom-
ass to produce ethanol. As previously mentioned, the
biggest question on biomass for energy seems to be the
projected amount available. Two additional issues
seem to be of importance also. . One is increasing
environmental concerns about project- or site-specific
use of forests, such as short-term rotation energy crops
on energy plantations, that might result in soil conser-
vation problems and increased runoff. Another issue is
the question raised by Vitousek (1991) , i.e., in an
energy-starved world, would increased use of biomass
for energy really displace fossil fuel or just make more
energy available?
Other renewables, such as geothermal, will undoubtly
play a role also.
Environmentally Sound But Likely To Have More
Limited Long-range CO, Impact
Options in this category include: . .
Conservation .... .:
Forest enhancements such as reforestation and
agroforestry . . ;. . , . s : .
Conservation options for fossil energy, as used here,
include: ....- - . . .--; . . -
* Increased end-use efficiency _. .., ...
Cogeneration
Increased electric generation efficiency
Conservation, especially end use management, is per-
haps the most useful and cost-effective step which can
be taken in the near term to reduce carbon loadings.
The National Academy of Engineering concluded that
about 1 GtC/year of U.S. emissions could be mitigated
for their "net benefit" and "low cost" options (NAE,
1991). NAE felt that this represented an optimistic
upper bound. Conservation can play an important role
to 1) help delay fossil C02 buildup and 2) help make use
of solar/renewables more cost-effective by reducing the
renewable capacity required.
Increased power generation efficiency also makes sense
where heat rates of existing plants can be improved via
such measures as improved plant operation and main-
tenance. On the other hand, modest improvements in
efficiency via replacement of existing fossil-fuel-fired
plants with capital-intensive advanced systems, such
as integrated gasification combined cycles (IGCC),
may need further consideration. Such systems repre-
sent large investments in fossil-energy capacity with a
substantial commitment to future carbon emissions
and at fairly high cost(ICF, 1990). Adding C02 control
to such systems will greatly increase the cost and
reduce efficiencies. While IGCC and coal-gasification
fuel cell systems incorporating integrated CO2 control
may cost less than application of CO, controls to con-
ventional power plants, the costs and energy penalties
are still high. Furthermore, unless there were eco-
nomically effective ways of recycling the carbon, the
CO2 released from disposal could be too great; there are
major uncertainties in the sequestration effectiveness
of C02 disposal in the ocean or on land. None of these
options may really be permanent. As shown for ocean
disposal in Figure 13, some part of the C02 will return
12000'
1500
1000
800
.1500
2000
2500
3000
YMT
t*f\
I>UM
3500
4000
4500
a100% into the atmosphere
* - b 50% at 1500 m depth into the ocean
- c 50% at 4000 m depth into the ocean
d 100% at 1500 m depth, into the ocean
e 100% at 4000 m depth into the ocean
Figure 13. Modeled atmospheric CO, resulting from use of
all fossil fuel over time with CO2 disposed of at various
depths.
-------
to the atmosphere as a function of disposal depth and
time (Hoffert et al., 1979). Aside from this re-release
issue, there other major uncertainties in C02 disposal
including technical feasibility, safety, environmental
acceptability, and cost.
It is also not clear when taking a broader, and more
e
ui
£
"3
E
UI
Trajectory Step
Figure 14. Simplified energy trajectory for coal to delivered
power via a conventional power plant with air pollution
controls for paniculate, flue gas desulfurization (FGO) for
sulfur dioxide, and flue gas treatment (FGT) for nitrogen
oxides, plus COZ control. -
systems-oriented view of the efficiency of fossil hydro-
carbon use, that it makes sense to use these valuable
resources so inefficiently. The energy use trajectory of
coal to power is already inefficient overall, but with the
added energy-intensive step of CO, control, it becomes
considerably less efficient. Figure 14 depicts the en-
ergy trajectory for coal to delivered electricity for con-
ventional coal combustion with C02 control. This step
would include CO, recovery, compression, pipelining,
and disposal. A base case coal-fired power plant is
assumed to have an efficiency of 34.5%; the plant
incorporating CO2 control is assumed to have an effi-
ciency of 22.8%. This represents an additional energy
loss of about 35% (Smelseretal., 1991). The combined
energy losses for all other air pollution control systems
are minor by comparison (Emmel and Maibodi, 1990).
Figure 15 does the same for a more energy efficient
trajectory including IGCC with integrated CO, control
(including CO2 pipelining, and disposal). A base case
IGCC plant is assumed to have 35.4% efficiency; the
plant incorporating CO2 control is assumed to have an
efficiency of 28.5% (Smelser et al., 1991). This repre-
sents an additional energy loss of about 20%. The
X
a
. e
"c
I
c
K
Trajectory Step
Figure 15. Simplified energy trajectory for coal to delivered
power via a IGCC system with integrated cleanup for
contention air pollutants plus CO, control.
additional loss of the original energy and carbon in
either case gives rise to the question... in planning for
the future, is this a wise or even acceptable use of the
Earth's valuable fossil hydrocarbon resources? The
extraction, preparation, and transportation step was
assumed to be 15% in both cases and would include
mining, reclamation, coal handling, coal beneficiation,
drying, and transportation. Also a 10% loss was as-
sumed in both cases for power line transmission losses.
Environmentally Uncertain or Uncertain for
Other Reasons
Options in this category include:
Options relying on CO8 disposal
Microalgae flue-gas capture
Aside from any questions on cost and energy efficiency
losses, options relying on CO2 disposal have major
uncertainties related to 1) technical feasibility, 2) the
actual COj sequestration effectiveness achieved, 3)
sustainability with time, 4) other environmental ef-
fects, 5) safety/liability, and 6) acceptance at all levels
by the concerned public. C02 sequestration effective-
ness is perhaps the major issue of uncertainties 2*6.
The rate of re-release CO, to the atmosphere is going to
be very difficult to predict with confidence. If CO2
disposal is undertaken, it will be undertaken at great
cost. If the end result is nothing but a modest delay in
arriving at unacceptable atmospheric CO2 levels, a
costly mistake may have been made. It may also be a
-------
mistake for which there is no technically feasible or
economically acceptable remedial action a new an-
thropogenic source of CO2 which could emit even if all
direct fossil energy emissions were brought to a halt.
Considering the consequences, can the uncertainty
associated with such options be reduced to the point
where the options can be utilized? The technology
options of 1} conventional combustion with C02 recov-
ery and disposal, and 2) higher-efficiency, fossil-fuel
power generation systems, such as IGCC with inte-
grated CO2 recovery/disposal, seem to be costly, tech-
nologically risky choices for the future.
Ocean C02 disposal environmental issues which
need to be resolved include:
C02 return to the atmosphere by long-term circula-
tion.
Potential for re-release of C02 via up-welling of
CO2 (e.g., large amounts of C02 pumped into a
limited region of the deep ocean might become
unstable to convective overturn).
Conversion of high density hydrates to low density.
Catastrophic releases in transport or storage.
Potential for local/regional acidification of ocean
with biological impacts.
Chances for reduction of biological diversity in the
column and ocean floor.
Potential for accumulation of clathrates on sea-
floor that would impair normal biological pro-
cesses in and on the sediment and disturb the
bottom sediment. . ...
Potential for local extinction of animals beneath
the deposited material.
Similar concerns exist relating to disposal on land. The
options for land disposal include disposal in depleted
gas formations, oil formations, salt domes, aquifers,
and other geological voids. Major, complex questions
exist relating to the capacity, rates, and suitability of
these options. We do not know for sure if land-based
disposal is more secure than ocean disposal and the
extent of probable re-release. Leakage would seem to
depend on many factors including the current and
future deterioration/condition of the mechanical and
geological structures associated with potential reser-
voirs (Johnson et al., 1992). We do not know the
potential for catastrophic release.
Smith and Thambimuthu (1991) estimate a CO, capac-
ity for depleted gas fields worldwide at 1.3 GtC, with
new capacity becoming available at 0.7 GtC/year. On-
land disposal would appear to be less costly than ocean
disposal, especially inland. However, the majority of
U.S. known capacity for storage is located in the gen-
eral area of the Gulf of Mexico, far from the largest
population centers. The potential exists for use in
enhanced oil recovery (EOR), but economic feasibility
in terms of credits for C02 may be limited under
conditions of low oil prices. Also the capacity seems
limited, probably amounting to only a small fraction of
a GtC/yr. After a few years of injection, C02 is likely
to show up at the well head. We do not know the extent
of re-release of C02 used for EOR or the potential for
prevention of re-release, but it is conceivable that
measures could be taken to prevent leaks and recycle
the C02. Other issues arise where CO/SOj/NO^ mix-
tures would be injected. These include die transport of
the CO/SO./NO, mixture over long distances, the dis-
position of the SOZ/NOI contaminants injected in oil
reservoirs, and the potential for pluggage of the reser-
voirs (Sparrow et al., 1988). Use of salt domes might
provide substantial capacity, but appears to be a very
high cost option (Johnson et al., 1992). Disposal in
aquifers and other porous structures is another option
currently being discussed, but questions on the result-
ing underground chemistry and fate remain to be
resolved. All cases of disposal would be site specific and
require careful study and piloting to avoid problems
like 1) encroachment of salt water or contaminated
water into potable water, 2) escape of CO2 into the
atmosphere, and 3) underestimated capacity (Johnson
etal.,1992).
The general limitations of these approaches would
need to be explored in depth before disposal decisions
could be made. The time frame for resolution may be
long, but if these approaches are to be considered, then
much research lies ahead before adequate models and
measurement are available for decision making.
Other approaches fall into this category; i.e., environ-
mentally uncertain or uncertain forother reasons. One
of these is microalgae flue gas capture of CO2. If used
as a control means for fossil-fuel C02 emissions, one of
the major concerns would be the effectiveness of CO2
capture. Also, while the concept is attractive, it ap-
pears to be difficult to implement because of the large
surface area required for exposing microalgae to CO2 in
the flue gas. A100-MW power plant would require an
algae farm surrounding the plant out to a distance of
4.3 km (Brown, 1990). This area would have to be
available in the immediate vicinity of the power plant,
which wouldbe a problem for alarge part of the existing
power plant capacity. Direct use of the biomass gener-
ated appears to be difficult, but it can be processed for
lipids or to other hydrocarbon fuels such as methane.
Mitigation potential exists for new and retrofit utility
applications but is probably limited to only certain
areas with enough land, proper terrain, nutrient capi-
-------
tal, and adequate water and evaporation rates. No
reliable estimates of costs are available. NRELis doing
bench scale research and the Japanese are heavily
researching the area including photobioreactor con-
cepts, which would allow more efficient and concen-
trated biological growth. If the CO2 being recovered by
the 'microalgae-capture approach ..was the result of
combusting the microalgae biomass itself or its conver-
sion products, then this approach could fall into a more
positive category of long-range solutions. Information
is not available to assess this option at present.
Environmentally Questionable or Unlikely Solu-
tion for Other Reasons
Two of the most discus sed atmospheric fixation/utiliza-
tion geoengineering approaches are:
Microalgae ocean fertilization.
Macroalgae farming with anaerobic digestion and
ocean disposal.
Microalgae ocean fertilization has been proposed as a
means for enhancing carbon uptake. The theory be-
hind iron (Fe), manganese, andphosphorus limitations
is not well understood or accepted universally. How-
ever, considerable scientific interest in testing the Fe
limitation hypothesis exists, and it will likely be pur-
sued. Sarmiento and Orr (1991) and Joos et al. (1991)
have done three-dimensional modeling simulations of
the impact of Fe fertilization in the oceans of the
Southern Hemisphere. Little is known about poten-
tially large, adverse effects; therefore environmental
considerations would top the list of research priorities.
To the extent that research is undertaken it is likely to
focus on ocean environments where key micronutri-
ents are thought to be limiting to productivity. This
approach to C02 sequestration requires consideration
of the following environmental concerns:
Fertilization of large areas of the Southern Atlan-
tic and Southern Pacific Oceans.
The potential for deep ocean anoxia and feedback
of CH4 and nitrous oxide (N2O).
* Potential for increased release of CO, the from the
ocean if fertilization is stopped.
Potential for dramatic changes in species composi-
tion. . - =
Potential for favoring phytoplankton species more
susceptible to ultraviolet [(UV)-B] radiation.
Potential for altered fertility in other ocean re-
gions.
Atmospheric CO2 sequestration via macroalgae farm-
ing is difficult to evaluate at this point, but Lee et al.
(1991) estimated that to generate 1 x 1018 J (1 quad) of
natural gas would require 1000 kelp farms, each 34 km
long and 0.5 km wide. Capital costs would be over $75
billion. Spencer (1991) has also estimated mitigation
potential and cost. The following environmental issues
must be considered:
Little-known environmental effects.
Ocean warming feedbacks.
Possibility of required ocean dumping of processed
sludges.
Potential for increased CH4 formation, especially
for sinking options.
* Possibility of required C02 disposal, which has
several environmental consequences.
Difficulty in supplying and recycling needed
nutrients.
Potential for storm disturbance and loss.
Potential increase in haloform production (methyl
bromide) and increased ozone depletion.
CO2 release to the atmosphere from respiration by
organisms forming coral.
A related option is that of increased marine biomass
(coastal kelp) with direct combustion or conversion use
of the kelp. An additional limitation in this case is the
limited coastal areal capacity available for CO2 fixa-
tion.
Conclusions
Present and future inhabitants of this planet are and
will be confronted with a long-range imbalance of
sources and sinks of CO2 as a result of anthropogenic
use and release of carbon into the atmosphere. Many
recent analyses of the C02 problem have focused on
near term aspects - out to the year 2030, 2050, or
perhaps 2100. Longer-term modeling of the atmo-
spheric buildup of COa (over several centuries) by
several investigators reveals the prospects of greatly
increased atmospheric concentrations even with dras-
tic reductions in current use of the fossil hydrocarbon
reserves for energy. While there are large uncertain-
ties associated with this longer-term modeling, longer-
range strategic planning of the global carbon future is
warranted. A rather arbitrary target was used to help
shape the consideration of the technical options. This
target was based on a recent analysis by Kasting et al.
(1992), which showed that by limiting CO2 emissions to
about one twenty fifth of current level, the peak atmo-
spheric concentration of COv could optimistically be
limited to 500 ppmv for several thousand years into the
future. The options have been examined in the light of
this emissions target and the technical options' current
status and prospects. These options are then catego-
10
-------
CmUgory
Could be major factors In
preventing long-term COZ
buildup
Would be environmentally
sound but likely to hire
more limited long-range
CO2 Impact
Would be environmentally
uncertain or uncertain Tor
other reasons
Would be environmentally
questionable or unlikely
solution for other reasons
Technology Options
Halt deforestation
Solar photovoltaic;
* Solar (passive and thermal)
Wiad
Recycling or utilization of CO2
Increased terrestrial biomass in conjunction with use for energy
Conservation
- Increased end-use efficiency
- CogeDeration
- Increased electric generation efficiency
Forest enhancements such as reforestation and agroforestry
Options relying on CO2 disposal
Microalgae flue-gas capture
Microalgac ocean fertilization
Macroalgae farming with anerobic digestion and ocean disposal
Table 1. Summary of applicability ol the technical options to the longer-
term elimination of global CO2 releases. (Note that nuclear (fission and
fusion) options may also be important, but are not included in this
assessment).
against incorrect, uncertain international
decisions to use potentially high-risk tech-
nology. Enhancing longer-term carbon
cycle model capabilities, and factoring in
mitigation, will be essential to doing fu-
ture longer-term strategic analyses with
confidence.
References
Arakawa, H.. Methods lor recycling use of emitted carbon dioxide:
catalytic hydrogenation. Now & Future, 7,10-12 (1992).
Arakawa, H.. DuBois. J. L., and Sayama, K., Selective conversion ol
COZ lo methanol by catalytic hyorogenation over promoted copper
catalyst, presented at the First Internationa! Conference on Carbon
Dioxide Removal, Amsterdam, the Netherlands (1892).
Borwardt. R. H., A technology tor reduction of CO2 emissions from the
transportation sector, Energy Convert. Management, 33, 443-449,
(1992).
Brewer, M.. Cool energy: renewable solutionato environmental prob-
lems. Revised Edition. MIT Press (1992).
Brawn. L M.. Biological trapping of carbon dioxide. SERI Microalgal
Technology, National Renewable Energy Laboratory (1990).
DOE, National Energy Strategy, Department ol Energy (1991).
rized accordingly in Table 1.
Perhaps the more important conclusion is that major
strategic changes will be necessary to have any hope of
cutting back CO2 emissions to near such a target. It is
likely that by the end of the next century even non-
energy use of the fossil hydrocarbon reserve may pro-
duce emissions of a few GtC/year. Therefore, the
relevance of international mitigation R&D would prob-
ably be enhanced if it is helpful in implementing a
strategy which would include:
* Halting deforestation
* Maximizing the use of renewables for energy and
non-energy uses
Preserving the use of fossil hydrocarbon reserves
for petrochemical and other industrial production
along with recycling of the products of this indus-
try and recycling or utilization of the waste C02
generated in production
Using renewables to produce the hydrogen input
requirements of the petrochemical industry
Maximizing near-term end-use energy conserva-
tion to both delay the C04 buildup and make the
use of energy from renewables more cost-effective
Continuing R&D on the feasibility of CO, disposal,
with emphasis on quantifying sequestration effective-
ness and environmental acceptability, would help guard
Emmel. T.. and Malbodl, M.. Retrofit Cods lor 8O2 and NOx Control Options at 200 Coal-Fired
Plants, VoL 1. Introduction and Methodology, EPA-600/7-90-021a (NTIS PB91-133322). U. S.
Environmental Protection Agency, Research Triangle Park. NC (1990).
Hangebrauck, R. P., Borgwardt, R. H., and Geren, C. 0., Carbon Dioxide S»
*tn
presented at EPA Symposium on Greenhouse Gas Emissions and Mitigation Research,
Washington. D.C., B/1B-2QV92 (1992).
Hotfwt, M. I, Wey Y. C., Catemrl. A. J, and Smacker. W. S., Atmospheric response to deep-
sea bijecuon* ol lossMusI carbon dioxide. Climatic Change, 2, S3-8 <1979).
HougMon, R. S., Tropical deforestation and atmospheric carbon dioxide, Climatic Change, In
press (1981).
ICF Resource*, Preliminary technology cost estimates ol measures available to reduce U.S.
greenhouse gaa emissions by 2010 (IBM).
IPCC, 1902 IPCC Supplement, Intergovernmental Panel on Clmete Change, UNEP/WMO
(1902).
JCAE (Joint Commute* on Atomic Energy), Understanding the national energy dilemma, 03rd
Congress, U. S. Government Priming OHioa stock number 52704194? (1973).
Johnson. H. E., Veflasa. 8. A., PeBine. J. E,. Biases. F . E., Sknbeck, D. R.. and Dlceraon. R. L.
Screening analysis ol CO2 utilization and fixation: Final Report, DOE/FE/S1680-H2 (1992).
Joos. F., Sarmiento. J. U and Siegenthaler, U., Estimates^ the ettect ol southern ocean Iron
lertllzation on atmospheric COZ concentrations. Mature, 349, 772-77$ (1991).
Kastlng, J. F., and Warner, J. C. G., The geochemteal carbon cyd» and the uptake of fossil fuel
CO2, AIP Conference PmoMdriga 247, Global Warming: Physics and Facia, B. Q. Levl, D.
HafemeWer, and R. Scrtbnei, ads., American Institute ol Physics, New York, pp. 175-200
0992).
Kheatigi. H S., The sansMvtty of CO2 projections to ocean processes, presented at the Third
International Conference on Analysis and Evaluation ot Atmospheric COZ Data - Present and
Past, HMerzarten (1988).
Lae,K.H.,Jc4maon,S.A.,Standl,W.D.,ar^Byrt,D.C.,eiomissEtate-ol-me-ertesieismen1,
Vols. 1 & 2, prepared by Research Triangle Intitule lor the Electric Power Research Institute,
EPRIGS-7471s{1B9l).
NAE, National Academy of Engineering. Poky Implications of greenhouse warming (1991 ).
MREL (SEHI), Th*pMerrUaldi
3674 (DE90000322) (IBM).
njy.AnlMerlalxxatory whit* paper. SEHI/TP-260-
ORNUThe Carbon DtoxWe Information Center, Trends "81 , A compendium of data on global
change (1991).
Sermiento, J. L, and On. J. C, 3-D sJmutaUons at the Impact of southern ocean nutrient
depletion on atmospheric CO2 and ocean chemistry, limnology and Oceanography On press)
Siegenthaler, U., and Oeschger. H.. Predicting future atmospheric carbon dioxide levels,
Soieno*, 199. pp. 368-384 (1B7B).
11
-------
Smalm, S. C.. Stock. R. M.. and McCtaary, G. J., Engineering and economic evaluation at CO2
removal bom fottiMuel-tlrad power plants. Vol. 1: pufcertzed-conMired power plants, prepared
by Fluor Daniel, Inc., EPRIIE-7365 (1991).
Smith, I. M., and Thambimuthu, K. V., Greenhouse gases, abatement and control, IEA Coal
Research, IEACR/39 (1991).
Sparrow, F. T.. Wobky, A. M.. Berry, Q. F., Brooks, C., Cobb, T. S., Lynch, E. P., JankowsW, 0.
J.. and Walbridge. E. W., Carbon dioxide from Hue gases lor enhanced oil recovery, Argonne
National Laboratory. DE89-014017(t9S6).
Spencer, O. F., A preliminary assessment of COZ mitigation options, The Energy Journal (1991).
Sione, J. (.., Photovoltaic developments, presented a) EPA Symposium on Greenhouse Gas
Emissions and Mitigalion Research Washington, D.C., &1B-2O/92 (1992).
Steinberg, M., Kobayashi, A., and Dong, Y., Rales of reaction and process design data lor the
Hvdrocarb process, EPA-600VR-93-020. U. S. Environmental Protection Agency, Research
Triangle Park. NC (1993).
UNEP.Thegreenhouse effect and energy efficiency: sometacts and fig ures.UMEP Industry and
Environment (1990).
VJiousek, P. M., Can planted forests counteract increasing atmospheric carbon dioxide? J.
Environ. Qual., 20,346-354 (1B9f).
Walker. J. C. G., and Kastlng, J. F., Effects of fuel and forest conservation on future levels ol
atmospheric carbon dioxide, Paleooeooraphy, PileocBmatology, Paleoeoology (Global and
Planetary Change Section), 97, 161-169 (1992).
Williams, R. H., A global perspective on biomass energy, presented al EPA Symposium on
Greenhouse Gas Emissions and Mitigation Research, Washington, D.C.. B/ia-20/92 (1992).
12
-------
Energy Onfotn. Mgmt Vol. 33. No. 3-8, pp. W-449. 1992
Mined in Cmu Britain
0196-8904/92 S5.00-fO.00
Pergamon Press IM
A TECHNOUX3Y FOR
REDUCTION OF CO, EMISSIONS FROM THE TRANSPORTATION SECTOR
Robert H. Borgwardt
Global Wanning Control Branch
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC
ABSTRACT
By sequestering byproduct carbon and replacing petroleum fuels with biomass-derived methanol, the
Hydrocarb process can nullify the net effect of CO, emissions from motor vehicles. This paper gives
a preliminary assessment of the process which indicates that substantially more fuel energy could be
produced~and at lower cost-than other current options for mitigating CO2 from mobile sources. The
incremental cost of eliminating net CO2 emissions is estimated at $0.05 per gallon (3.78 liters) of
gasoline displaced by methanol. About 80 percent reduction should be achievable at no incremental
cost.
KEYWORDS
Carbon dioxide; methanol; biomass; global warming; transportation; alternative fuel
INTRODUCTION -''
Annual consumption of petroleum fuel in the U.S. is approximately 4.2 x 10" liters and increasing.
COj emissions from the transportation sector account for 24 percent of the U.S. total. Other emissions
from auto exhaust are largely responsible for non-compliance with ozone standards in more than 100
U.S. cities. These facts have prompted the Air and Energy Engineering Research Laboratory of the
U.S. EPA to examine processes that might produce clean alternative fuels which could reduce both the
human health risks associated with petroleum fuels and the ecological risks anticipated from climate
changes that may ensue from the global warming effect of accumulating atmospheric CO2. Also
important is the fact that U.S. domestic petroleum supplies are being depleted with an exacerbating
effect on national security and balance of payments. The most desirable alternative fuel should
therefore be derived from domestic resources to the greatest possible extent and must be producible
in amounts that are very large if petroleum is to be displaced sufficiently to affect any of the above
problems. This paper summarizes the results of an evaluation of the Hydrocarb process, conceived by
ichers at the Brookhaven National Laboratory (Steinberg, 1990; Steinberg et al.. 1991) as a
promising means to the desired objectives.
In the rear term, there are three alternatives to the fuels in current use: compressed natural gas, ethanol,
and methanol. The substitution of natural gas for gasoline could reduce CO, emissions by 16 percent,
mainly because of its higher H/C ratio.. .Ethanol derived from grain by the fermentation-distillation
process is already used, as a 10-percent blend with gasoline, in 9 percent of the fuel sold in the U.S.
Prospects for significantly increasing that percentage are limited by the available grain supply and
production cost of the alcohol which is currently sustained by subsidy. Newer ethanol processes, such
as acid hydrolysis and enzymatic hydrolysis of woody biomass, will expand the availability of fuel
ethanol by utilizing lower grades of feedstock. Unlike compressed natural gas, the new ethanol
443
-------
444
BORGWARDT; REOl/CING CO2 EMISSIONS FROM TRANSPORTATION
processes can theoretically reduce GQ, emissions to tern since all of the carbon used to fuel the
process, as well as the carbon content of the alcohol, is biogenic and reabsorbed from the atmosphere
by the subsequent photosynthesis cycle.
Methanol, the third alternative, is currently produced from natural gas at a yield of 0.78 mol per mol
of CH,. As a liquid fuel, methanol has the advantage of greater compatibility with the existing auto
refueling infrastructure than natural gas and would avoid the powerful greenhouse effect of CH,
emission from natural gas fueled vehicles, estimated at 0.05-1.6 g/km. Even if produced from fossil
CH«, methanol can reduce CO2 emissions from vehicle exhaust by about 19 percent relative to gasoline
as a result of its higher H/C ratio and greater energy efficiency in internal combustion engines. When
the 002 emissions from methanol distribution and production by rhe conventional technology are
accounted for, the net CO2 reduction is estimated to be 7 percent (DeLuchi et al., 1987). By
supplementing natural gas with biomass as cofcedstock in methanol production, the Hydrocarb process
can reduce net CO2 emissions to zero while increasing the yield of methanol to 2.15 mol per mol of
CH<, or a 175 percent increase of fuel obtainable from a given supply of natural gas. Furthermore,
where biomass supply is the limiting factor, this analysis indicates that the maximum yield of alcohol
fuel energy can be derived from that biomass if it is convened to methanol via the Hydrocarb route.
The following discussion outlines how these results might be achieved
PROCESS ANALYSIS
Figure 1 is a flow diagram for the Hydrocarb process as configured for zero net CO2 emission. Woody
biomass, dried to 18.8 percent moisture, is fed with natural gas to a hydrogasification reactor (HGR)
at 800 °C and 50 atm. Gasification products of the indicated equilibrium composition are fed to a
methane pyroJysis reactor (MPR) where methane is thermally cracked to carbon black and hydrogen.
In the third and final step, methanol is produced by reaction of CO and hydrogen in a stands: catalytic
converter. Equilibrium conversion for this step is based on the data of Strclzoff (1970); other equilibria
are based on the data of Baron et al. (1975).
BJOMASS MOISTURE
13.4kg
BIOMASS 100KQ I CH, 23.5kg.
Ill
I 4 4
PURGE
0.309 kg mol
*
CO 7.63%
CO, .3.65
-CH, 16.04
Hf> 0.02
H, 71.63
CH.OH 1.03 :,.
23.81 k^rool
CHiOH
9.42kg
HO
3.11kg
HVDRO-
OASIFICATION
REACTOR
800 «
S0«tm
I
CO 6.40%
CO, 2.90
CH> 35.60
H|O 16.00
Hi 38.90
25,06 kg mol
PVROtYSB
REACTOR
11001C
50 Mm
CARBON RESIDUE
1247kg :
CO 6.70%
CO, 331
CHt 14.10 '
HjO 0.68
H* 62.97
CONDENSER
SOT ~
SO *m
26 J7 kg mol
CARSON BLACK
19.82kg
CO 17.60%
CO, OJO
CH, 11.30
Hfl 2.80
H, 68.10
2186 kg mol
METHANOL
CONVERTER
260 «C
SO «tm
Fig. 1. Hydrocarb flow sheet and equilibrium stream compositions.
-------
BORGWARDT: REDUCING CO, EMISSIONS FROM TRANSPORTATION
445
Part of the product carbon is burned in a reheater to supply energy for the pyrolysis reactor; the
remaining carbon is sequestered to offset CO, emissions. Zero net emission of CO2 occurs when the
mols of sequestered carbon (S) is:
S = mols methane! produced - mols biomass fed + mols carbon burned in the reheater +
mols purge x total mol fraction- carbon in purge - total mols carbon produced -
mols carbon used for reheat
The relative simplicity of sequestering carbon, rather than"CO2, for the purpose of mitigating
atmospheric emissions would favor this process in any situation where methanol can be substituted for
direct use of fossil fuels-including stationary sources such as turbine generated electric power.
Material and energy balance data are summarized-in Table 1. Heat transfer between the exothermic
hydrogasification reactions and endothermic pyrolysis reactions is accomplished by alumina recycle
which also acts as a cracking catalyst. This energy is supplemented by the carbon burned in the
reheater. A second alumina recycle transfers heat from the converter feed gases to the gasifier feed
stream. The indicated drier heat load assumes an initial biomass moisture content of 50 wt percent
which is reduced to the three values shown in Table 1 using waste heat from the converter and
reheater.
Table 1. Material and energy balance summary
Biomass moisture content, wt %
5.5 11.8 18.5
Biomass feed, kg (moisture free basis)
CH, feed, kg
HGR recycle feed, kg-mol
HGR gas out, kg-mol
HGR carbon out, kg
MPR carbon black, kg
Reheater heat load, 10* kcal
MPR gas out, kg-mol .
MeOH convener gas out, kg-mol
MeOH converter heat load, 10* kcal
MeOH product, kg
Water out, kg
Condenser gas out, kg-mol
Purge gas, kg-mol
Purge gas heating value, 10* kcal
Carbon requirement for reheat, kg
Biomass drier heat load, 10* kcal
100
20.2
20.4
21.9
19.3
14.9
17.3
29.4
23.5
-7.55
86.8
3.2
20.7
0.293
-2.4
19.1
63
100
23.5
23.4
25.1
12.1
19.8
20.6
33.7
27.0
-8.65
99.4
3.1
23.7
0.309
-2.5
23.1
5.8
100 *
27.3
26.6
28.4
13.5
15.5
24.6
38.3
30.6
-10.1
115
3.2
26.9
0.272
-2.2
28.5
5.3
Because the amount of carbon available for sequester depends on the reheater heat load, the COj
emissions are a function of the biomass moisture content: when that moisture exceeds 11.8 percent, a
net COj emission occurs. Higher biomass moisture, however, also increases the methanol yield per
unit of biomass fed, improving energy conversion efficiency and process economics. At 18.5 percent
biomass moisture, all product carbon is required for the reheater and none is available for sequester.
This configuration therefore yields maxhmmi methanol at minimum cost, but also a net CO, emission.
Process Economics
The absence of an oxygen plant, a steam plant, and major desulfurization and ash handling equipment,
is expected to reduce the capital cost and simplify the process compared to other gasification
technologies, especially those involving coal. The capital cost is estimated at $4.5 x 10' for a
-------
446
BORGWARDT. REDUCING CO2 EMISSIONS FROM TRANSPORTATION
Hydrocarb plant processing 4500 metric tons per day of biomass. The plant size is constrained by the
cost of biomass production and transportation from dedicated energy plantations. The delivered cost
of biomass from larger plantations cancels the economy of scale for a larger Hydrocarb plant The
delivered biomass cost used here is based on the data of Ismail and Quick (1991) for short-rotation
willow crops and is the largest single cost factor. The following data were used to evaluate methanol
production costs for this plant operating at the conditions defined in Table I:
Biomass, per dry metric ton $53.7
Natural gas, per 100 Nnr1 $14.1
Carbon sequester, per metric ton $23.1
Ash disposal, per metric ton $9.0
O & M, other, per metric ton biomass $5.0
Return on investment 15%
Annual utilization factor 90%
Total capital charge 20.9%
The results, shown in Table 2, range from $0.51 per gallon (3.78 liter) for zero CO2 emission, to $0.46
per gallon at a maximum CO2 emission of 27.5 kg per 10' J of alcohol energy produced.
Table 2. Performance estimates
Biomass moisture content, wt %
5.5 11.8 18.5
Biomass fed, kg (MF)
Biomass moisture, kg
Carbon sequestered, kg
Carbon available for sale, kg
Thermal efficiency, %
Mol MeOH/mol CH< fed
MeOH production cost. $/gaI
Incremental cost, S/gaT
Net COj emission, kg/109 J
COj reduction*, kg 00/10' J of MeOH
Reduction of gasoline COj emission, %
Effective cost of CO, emission
reduction, $/1000 kg CO,
$71000 kg carbon
100
5.8
7.6
7.6
62
2.15
0.55
0.11
0
96
100
11.9
43.6
100
13.4
8.8
0
67.5
2.12
OL51
0.05
29.0
96
100
5.6
20.5
100
22.7
0
0
73.6
2.11
0.46
-0.03
27,5
68.5
; 71.4
-4.7
5*7.2
"Relative to 1989 gasoline price.
Cost of CO, Emission Reduction
Included in Table 2 are other performance data derived from the methanol production costs. Of
principal interest is the cost of achieving zero CO, emission by displacing gasoline with hydrocarb
methanol which is most meaningfully expressed in terms of its incremental cost above mat of its
gasoline equivalent For a methanol production cost of $0.51/gal, the equivalent gasoline selling price
is obtained by adding die costs of distribution ($0.06), markup ($0.07), and taxes ($0.12) and
multiplying by the ratio of energy/volume of the two fuels (1.54). This ratio is variously estimated
according to different assumed magnitudes of the increase in energy efficiency of internal combustion
engines using methanol; the value used here is based on measurements by EPA's Office of Mobile
Sources (1989) which show a 30 percent increase of efficiency. The resulting equivalent price of
gasoline is $1.17/gal. Given the average price of $1.12/gal for gasoline sold in the U.S. during the year
1989, the incremental cost of reducing CO, emissions to zero is $0.05 per gallon of gasoline displaced.
-------
BORGWARDT: REDUCING CO, EMISSIONS FROM TRANSPORTATION
447
The CO, emission from gasoline combustion is 9.0 kg per 3.78 liters (1 gallon). It follows that the cost
of eliminating the effect of that emission by substitution of Hydrocarb methanol for gasoline would be
$0.05(1000)/9 = $5.6/metric ton of C02
An additional 10 percent CO2 emission occurs during the petroleum refining process which is assumed
here to be balanced by Hydnxarb emissions from natural gas compression and delivery plus the
emissions from biomass production and its transport--reliable estimates of which are unavailable.
Referring to Table 2, the minimum methanol production cos: of $0.46/gal is obtained with a net CO2
emission of 27.5 kg/iff J of useable energy. Assuming 30 percent greater energy requirement for
gasoline engines, the same 10* J of delivered energy would result in the generation of 96 kg of CO2.
The percentage reduction of CO2 emission for this case is therefore 68.5 kg, or 71.4 percent, and the
effective cost is negative; i.e., at any methanol production cost less than $0.48/gal, an economic benefit,
as well as substantial CO2 reduction, results from replacing gasoline (at 1989 prices) with biogenic
methanol. The crossover point at which the effective cost is zero occurs at a CO2 reduction os about
80 percent.
Comparison with Other Alcohol Processes
Although the expected cost of methanol production bt Hydrocarb is based on data from a much less
advanced stage of development than other alcohol processes, it is nevertheless instructive to compare
the above cost estimates with those options. This is done in Table 3 which compares the production
cost of Hydrocarb methanol at zero net GQj emission ($0.51/gal) with other alcohol processes
normalized on the basis of fuel energy content.
Table 3. Production costs of alcohol fuels derived from biomass
Process
Alcohol production cost, $/10* J
Ethanol by enzymatic hydrolysis
Methanol by steam-oxygen gasification
Ethanol by acid hydrolysis
Ethanol by fermentation of com
Methanol by Hydrocarb
22.3
17.6
17.0
15.1
8.5
In addition to production cost, the availability of alcohol feedstocks-cither natural gas or biomass-will
determine the extent to which petroleum can be displaced as a future source of motor fuel.
Production cost is, in fact, of little significance in terms of CO, mitigation if the available feedstocks
are insufficient to displace a major portion of the petroleum fuels. Table 4 compares the quantities of
alcohol that could be produced from given domestic supplies of natural resources and the degree to
which those supplies might be leveraged by available process options.
By die year 2020, projections based on current trends suggest that the U.S. annual consumption of
petroleum fuels will increase by about 49 percent. If CO, emissions from the transportation sector
could be stabilized at 1990 levels solely by partial displacement of petroleum with biogenic fuels, a
total fuel production equivalent to 33 x 10" liters of gasoline would be required by 2020. Clearly,
a substantial increase in vehicle fuel economy as well as maximum utilization of fuels derived from
biomass would be necessary to achieve stabilization. In the case of light duty vehicles, this would
require an average fuel economy of about 35 miles/gal (14.9 km/liter) for gasoline and an annual
production of 1.3 x 10" liters of methanol with zero net CO, emission-a formidable commitment of
industry and resources.
-------
448
BORGWARDT: REDUCING C02 EMISSIONS FROM TRANSPORTATION
Table 4. Fuel alcohol yields from limited resources
Process
Conventional methanol (steam reforming)
Hydrocarb methanol
Conventional ethanol (fermentarion/dist.)
New ethanol (enzyme hydrol.)
Hydrocarb methanol
CO* reduction,*
.%
19
100
14
100
100
71
Limiting
resource
Natural gas
Natural gas
Biomass (com)
Biomass (wood)
Biomass (wood)
Biomass (wood)
Leverage
factor*
Base case
2.75
Base case
2.0
5.9
7.8
'Relative to emission from gasoline fueled vehicles
"Of fuel energy production
Our analysis indicates that the maximum methanol yield from biomass would be obtained by use of
coal as cofeedstock instead of natural gas (Borgwardt ct al., 1991) This version of Hydrocarb, still in
preliminary study, may double the methanol yield from the zero-GO2 natural gas option. Additionally,
it is expected to be capable of achieving a negative CO2 emission that could further leverage the
stabilizing effect by balancing pan of the gasoline emissions. Although the engineering problems
involved with the use of coal as cofeedstock-especially with respect to desulfurization, ash handling,
and nitrogen purge-are more difficult, that option is expected to provide the greatest potential for
displacing petroleum with a domestically produced alternative fuel. "*
SUMMARY AND CONCLUSIONS
At current crude oil prices, methanol produced from natural gas by the conventional route could be an
economically viable alternative to gasoline as a motor fuel while reducing CO, emissions from vehicles
by at least 7 percent Assuming the existence of a distribution-network, methanol produced in the U.S.
or delivered to U.S. ports at a cost less than $0.48/gal would achieve this CO, reduction at no
incremental cost relative to the gasoline it displaced.
If derived from biomass and natural gas by the Hydrocarb process, methanol may be capable of
reducing CO, emissions by about 80 percent at no incremental cost A reduction of 100 percent might
be achievable at an incremental cost of about $O.OS/gallon of gasoline displaced, based on 1989 fuel
prices in the U.S. For the latter scenario, the effective cost of CO, mitigation in the transportation
sector is estimated at $5.67ton or $20.5/ton of carbon emissions avoided.
The extent to which methanol or other alcohol fuels might reduce CO, emissions will depend largely
on die size of the biomass supply, and the amount of fuel energy that can be derived from it
Hydrocarb is expected to leverage energy production by a factor of 3-6 times that of other alcohol
process options and at about 50 percent lower cost When used a feedstock for Hydrocarb, natural gas
can be converted to a liquid fuel that contains over twice the energy of the original gas and permits
this fossil fuel to be utilized without CQj emission. This assessment concludes that evaluation of the
process should proceed to the pilot plant stage. A study of that type will be undertaken with EPA
support in 1992. : '.=.-',---'.-. ^ , ,. -.. - -.:. ".--.,
-------
BORGWARDT: REDUCING CO2 EMISSIONS FROM TRANSPORTATION
449
REFERENCES
Baron, R.E., Porter, J.H. and O.H. Hammond (1975). Chemical Equilibrium in Carbon-Hydrogen-
Oxygen Systems. M.I.T. Press, Cambridge, MA. ...
DeLuchi, M.A., Sperling, D. and R.A. Johnston (1987). A'Comparative Analysis of'Future
Transportation Fuels. Research report UCB-ITS-RR-87-13, Institute of Transportation Studies,
University of California, Berkeley.
Borgwardt, R.H., Steinberg, M., Grohse, E.W. and Y. Tung (1991). Biomass and fossil fuel to
methanol and carbon via the Hydrocarb process. Energy Biomass Wastes. 15. in press.
Ismail, A. and R. Quick (1991). Advances in biomass fuel preparation, combustion and pollution
abatement technologies. Energy Biomass Wastes. 15. in press.
Steinberg, M. (1990). Biomass and Hydrocarb Technology for Removal of Atmospheric CO2. Research
report BN:L-44410, Brookhaven National Laboratory, Upton, NY.
Steinberg, M., Grohse, E.W., and Y. Tung (1991). A Feasibility Study for the Coprocessing of Fossil
Fuels with Biomass by the Hydrocarb Process. Research report EPA-600/7-91-007 (NTIS DE91-
011971), U.S. EPA, Air and Energy Engineering Research Laboratory, Research Triangle Park, NC.
Strelzoff, S. (1970). Methanol: its technology and economics. Chem. Ene. Prog. Svmp. Ser. No. 98.
661 54-68.
U.S. Environmental Protection Agency, Office of Mobil Sources (1989). Analysis of the Economic and
Environmental Effects of Methanol as an Automotive Fuel," Research report 0730 (NTIS PB90-
225806).
-------
-------
Journal of Power Sources, 37 (1992) 255-264
255
Fuel cell energy recovery from landfill gas"
G. J. Sandelli . . ' :
International Fuel Cells Corporation, 195 Governors Highway, South Windsor, CT 06074 (USA)
R. J. Spiegel**
U.S. Environmental Protection Agency, Air and Energy Engineering Research Laboratory,
Research Triangle Park, NC 27711 (USA)
Abstract
International Fuel Cells Corporation is conducting a US Environmental Protection Agency
(EPA) sponsored program to demonstrate energy recovery from landfill gas using a commercial
phosphoric acid fuel cell power plant. The US EPA is interested in fuel cells for this
application because it is the cleanest energy conversion technology available. This paper
discusses the results of Phase I, a conceptual design, cost, and evaluation study. The
conceptual design of the fuel cell energy recovery concept is described and its economic
and environmental feasibility is projected. Phase II will include construction and testing
of a landfill gas pretreatment system which will render landfill gas suitable for use in the
fuel cell. Phase III will be a demonstration of the energy recovery concept.
Introduction
The US Environmental Protection Agency (EPA) has proposed standards and
guidelines [1] for the control of air emissions from municipal solid waste (MSW)
landfills. Although not directly controlled under the proposal, the collection and disposal
of waste methane, a significant contributor to the greenhouse effect, would result from
.the emission regulations. This EPA action will provide an opportunity for energy
recovery from the waste methane that could further benefit the environment. Energy
produced from landfill gas could offset the use of foreign oil, and air emissions affecting
global wanning, acid rain, and other health and environmental issues.
International Fuel Cells Corporation (IFC) was awarded a contract by the US
EPA to demonstrate energy recovery from landfill gas using a commercial phosphoric
acid fuel cell. IFC is conducting a three-phase program to show that fuel cell energy
recovery is economically and environmentally feasible in commercial operation. Work
was initiated in Jan. 1991. This paper discusses the results of Phase I, a conceptual
design, cost, and evaluation study, which addressed the problems associated with landfill
gas as the feedstock for fuel cell operation.
The research described in this article has been reviewed by the Air and Energy Engineering
Research Laboratory of the US EPA and approved for publication. Approval does not necessarily
reflect the view and policy of the Agency nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
"Author to whom correspondence should be addressed.
0378-7753y92/$5.00
01992-Elsevier Sequoia. All rights reserved
-------
256
Phase II of the program includes construction and testing of the landfill gas
pretreatment module to be used in the demonstration. Its objective will be to determine
the effectiveness of the pretreatment system design to remove critical fuel cell catalyst
poisons such as sulfur and halides. A challenge test is planed to show the feasibility
of using the pretreatment process at any landfill in conjunction with the fuel cell
energy recovery concept. A preliminary description of the gas pretreater is presented
here.
Phase III of this program will be a demonstration of the fuel cell energy recovery
concept. The demonstrator will operate at Penrose station, an existing landfill gas-to-
energy facility owned by Pacific Energy in Sun Valley, CA. Penrose Station is an 8.9
MW internal combustion engine facility supplied with landfill gas from four landfills.
The electricity produced by the demonstration will be sold to the electric utility grid.
Phase II activities began in Sept. 1991 and Phase III activities are scheduled to
begin in Jan. 1993.
Landfill gas
Availability
The MSW landfills in the US were evaluated to determine the potential power
output which could be derived using a commercial 200 kW fuel ceil. Each fuel cell
would consume 100 000 SCFD of landfill gas to generate 200 kW, assuming a heating
value of 500 Btu per cubic foot.
The potential power generation market available for fuel cell energy recovery was
evaluated using an EPA estimate of methane emissions in the year 1992 [2aj. An
estimated 4370 MW of power could be .generated from the 7480 existing and closed
sites identified. The largest number of potential sites greater than 200 kW occurs in
the 400 to 1000 kW range. This segment represents a market of 1700 sites or 1010
MW.
The assessment concluded that these sites are ideally suited to the fuel cell
concept The concept can provide a generating capacity tailored to the site because
of the nodular nature of the commercial fuel cell. Sites in this range are also less
well served by competing options, especially Rankine and Brayton cycles which exhibit
poorer emission characteristics at these power ratings. '
As a result of the assessment, the conceptual design of the commercial concept
was required to be modular and sized to have the broadest impact on the market.
Characteristics
The available information on landfill gas compositions was evaluated to determine
the range of gas characteristics which a fuel cell landfill gas-to-energy power plant
will encounter. This information was used to set the requirements for the gas pretreatment
and fuel cell power plant designs.
A summary of landfill gas characteristics is shown in Table 1. The heating value
of the landfill gas varies from 350 to 600 Btu per cubic foot, with a typical value of
500 Btu per cubic foot. The major non-methane constituent of landfill gas is carbon
dioxide. The carbon dioxide ranges from 40 to 55% by volume of the gas composition
with a typical value of 50%. Other diluent gases include nitrogen and oxygen, which
are indicative of air incursion into the well (most frequently in perimeter wells).
Nitrogen concentrations can range as high as 15% but typical values are 5% or less.
Oxygen concentrations are monitored closely and held low for safety reasons.
-------
257
TABLE 1
Landfil! gas characteristics
Characteristic
Heating value ' '
(HHV).(Btu/ft3)
CH4 (%)
CO, (%)
N2 (%)
02 (%)
Sulfur as H3S (ppmv)
Halides (ppmv)
Non-methane organic compounds
(NMOCs) (ppmv)
Range
350-600
35-58
40-55
0-15*
0-2.5'
1-700
N/A
237-14000
Typica)
500
50
45
5
<1% (for safety)
21
132
2700
'Highest values occur in perimeter wells.
Landfill gas constituent compounds reported by EPA [2bJ indicate a typical value
for the total non-methane organic compounds (NMOCs) of 2700 ppmv (expressed as
hexane). The NMOC concentration in the landfill gas is an important measure of the
total capacity required in the gas pretreatment system, while the specific individual
analyses provide a basis for gas pretreatment subcomponent sizing. The specific
contaminants in the landfill gas, of interest to the fuel cell, are sulfur and halides
(chiefly chlorides and fluorides). The sulfur level ranges from 1 to 700 ppmv, with a
typical value in the order of 21 ppmv. Sufficient data were not available to assess the
range of the halides, but a typical value of 132 ppmv was calculated for this contaminant
PC]. The range of contaminant values varies not only from site to site, but also at
any given site with time due to seasonal weather or moisture content. These characteristics
require the pretreatment system design to be capable of handling these gas quality
variations to avoid expensive site specific engineering of the pretreatment design which
would affect the marketability and economics of the concept.
Emissions requirements
Existing US regulations do not address methane emissions from landfills directly.
Proposed new EPA regulations [1J would control NMOCs from large landfills (150
Mg per year and up) and hence would indirectly control methane emissions. "'
Landfill gas emission requirements are primarily determined at the state and local
level. State requirements are generally limited to controlling explosion hazards, typically
limiting methane concentrations to below 25% of the lower explosion limit. An evaluation
of state regulations revealed that collection and control requirements generally necessitate
venting, or the use of a flare. However, Federal dean Air Act requirements are
driving the state and local air quality rules, especially in areas identified as non-
attainment regions. For instance, non-attainment regions for ozone may lead to strict
requirements for secondary emissions including NO,, carbon monoxide and NMOCs.
The best known example of strict local emission requirements is the South Coast Air
Quality Management District (SCAQMD) in southern California.
-------
258
Commercial fuel cell landfill gas to energy system conceptual design
This section describes the commercial fuel eel) landfill gas to energy system
conceptual design. The design is based on providing a modular, packaged, energy
conversion system which can operate on landfill gases with a wide range of compositions
as typically found in the US. The complete system incorporates the landfill gas collection
system, a fuel gas pretreaiment system and a fuel cell energy conversion system. In
the fuel gas pretreatment section, the raw landfill gas is treated to remove contaminants
to a level suitable for the fuel cell energy conversion system. The fuel cell energy
conversion system converts the treated gas to electricity and useful heat.
Landfill gas collection systems are presently in use in over 100 MSW landfills in
the US. These systems have been proven effective for the collection of landfill gas.
Therefore these design and evaluation studies were focused on the energy conversion
concept.
Overall system description
The commercial landfill gas to energy conversion system is illustrated in Fig. 1.
The fuel pretreatment system has provisions for handling a wide range of gas con-
taminants. Multiple pretreatment modules can be used to accommodate a wide range
of landfill sizes. The wells and collection system collect the raw landfill gas and deliver
it at approximately ambient pressure to the gas pretreatment system. In the gas
pretreatment system the gas is treated to remove NMOCs including trace constituents
which contain halogen and sulfur compounds.
The commercial energy conversion system shown in Fig. 1 consists of four fuel
cell power plants. These power plants are designed to provide 200 kW output when
operating on landfill gas with a heating value of 500 Btu per standard cubic foot and
for accommodating higher contaminant concentrations. The output from the fuel cell
is utility grade a.c. electric power. It can be transformed and put into the electric
grid, used directly at nearby facilities, or used at the landfill itself. The power plants
are capable of recovering co-generation heat for nearby use or rejecting it to air.
800 KW FUEL CELL POWER PLANT
OPERATING ON LANDFILL GAS
Fig. 1. Fuel cell energy recovery commercial concept.
-------
259
. As configured in Fig. 1, the commercial system can process approximately 18 000
standard cubic feet per hour of landfill gas (mitigate 9050 SCFH of methane) with
minimum environmental impact in terms of liquids, solids or air pollution. Details of
the individual sub-elements in the energy conversion system follow this discussion.
'Fuel pretreatment system ~ ' ' \
A. block diagram of the landfill gas pretreatment system is shown in Fig. 2. The
fuel pretreatment system incorporates two stages of refrigeration combined with three
regenerable adsorbent steps. The use of staged refrigeration provides tolerance to
varying landfill gas constituents. The first stage significantly reduces the water content
and removes the bulk of the heavier hydrocarbons from the landfill gas. This step
provides flexibility to accommodate varying landfill characteristics by delivering' a
relatively narrow cut of hydrocarbons for the downstream beds in the pretreatment
system. The second refrigeration step removes additional hydrocarbons by a proprietary
process and enhances the effectiveness of the activated carbon and molecular sieve
beds, which remove the remaining volatile organic compounds and hydrogen sulfide
in the landfill gas. This approach is more flexible than utilizing dry bed adsorbents
alone and has built-in flexibility for the wide range of contaminant concentrations
which can exist from site to site and even within a single site varying with time.
The three adsorbents are regenerated by using heated gas from the process stream.
Each step consists of two beds in parallel. In operation, one bed is adsorbing while
the parallel bed is being regenerated. The regeneration path and sequence are shown
as dashed lines in Fig. 2. A small portion of the treated landfill gas (approximately
8%) is heated by regeneration with the incinerator gases and then passes through the
beds in the sequence shown. After exiting the final bed, the regeneration gas is fed
into the low NO, incinerator where it is combined with the vaporized condensates
from the refrigeration processes and the mixture is combusted to provide 98%. destruction
of the NMOCs from the raw landfill gas. The exhaust from the incinerator is essentially
COj and water. The pretreatment system design provides treated gas to the fuel cell
power plant in an efficient, economic, and environmentally acceptable manner.
The pretreatment system design provides flexibility for operation on a wide range
of landfill gas compositions, it has minimal solid wastes, high thermal efficiency, and
low parasite power requirements. The pretreatment system is based upon modification
of an existing system and utilizes commercially available components. The process
train and operating characteristics need to be validated by demonstration. Key dem-
JK 77
Fig. 2. Simplified block diagram of commercial landfill gas pretreatment system.
-------
260
onstrations in Phase II will include: the achievement of low total haiide contaminant
levels in the treated gas; effectiveness of the regeneration cycle as affected by regeneration
time and temperature; durability of the regenerable beds; and low environmental
emissions.
The fuel pretreatment system described above was analyzed to estimate the overall
thermal efficiency, internal electric power requirements and maintenance characteristics.
The estimated thermal efficiency is 92% with the balance of thermal energy used for
regeneration, vaporization of the condensate and incineration of regeneration gases.
Electric power is used for pumping the gases and the refrigeration stages and is
accounted for as a parasite power characteristic of the system. Maintenance requirements
consist of maintaining and adjusting controls and valves in the regeneration system
and replacement of fully regenerated spent bed materials on an annual basis.
The pretreatment system was evaluated to define anticipated air emissions, and
liquid and solid effluents. The incinerator is designed for 98% destruction of all NMOCs
and NO, emissions of less than 0.06 pounds per million Btu of fuel consumed. There
is no liquid effluent from the system since all condensates are vaporized and subsequently
incinerated. Solid disposal involves removing spent regenerable bed materials at the
factory and treatment by an approved reclamation processor.
Fuel cell power plant
The commercial landfill gas energy conversion conceptual design incorporates four
200 kW fuel cell power units. Since each of the four units in the concept is identical,
this discussion will focus on the design issues for a single 200 kW power unit.
A preliminary design of a fuel cell power plant was established to identify the
design requirements which allow optimum operation on landfill gas. Three issues
specific to landfill gas operation were identified which reflect a departure from a
design optimized for operation on natural gas. A primary issue is to protect the fuel
cell from sulfur and haiide compounds not scrubbed from the gas in the fuel pretreatment
system. An absorbent bed was incorporated into the fuel celt fuel preprocessor design
which contains both sulfur and haiide absorbent catalysts. A second issue is to provide
mechanical components in the reactant gas supply systems to accommodate the larger
flow rates that result from use of dilute methane fuel. The third issue is an increase
in the heat rate of the power plant by approximately 10% above that anticipated from
operation on natural gas. This is a result of the inefficiency of using the dilute methane
fuel. The inefficiency results in an increase in beat recoverable from the power plant.
Because the effective fuel cost is relatively low, this decrease in power plant efficiency
will not have a significant impact on the overall power plant economics.
The landfill gas power plant design provides a packaged, truck transportable, self-
contained fuel cell power plant with a continuous electrical rating of 200 kW. It is
designed for automatic, unattended operation, and can be remotely monitored. It can
power electrical loads either in parallel with the utility grid or isolated from the grid.
In summary, a landfill gas fueled power plant can be designed to provide 200
kW of electric output without need for technology developments. The design would
require selected components to increase reactant flow rates with a minimum pressure
drop. To implement the design would require non-recurring expenses for system and
component design, verification testing of the new components, and system testing to
verify the power plant performance and overall system integration.
Environmental and economic assessment of the fuel ceO energy conversion system
The commercial application of the concept to the market described previously
was assessed. For the purpose of the evaluation, a site capable of supporting four
-------
261
fuel cell power modules was selected. Hie site characteristics assumed are the typical
values discussed earlier. The site would produce approximately 434 000 standard cubic
feet of landfill gas per day. The gas contains approximately 50% methane with a
heating value of 500 Btu per standard cubic foot.
The analysis of the environmental impact shows that both the fuel cell and the
flare system can be designed to eliminate the methane and the NMOCs from the
landfill gas system. For the example site considered, the methane elimination is
essentially complete for both systems and 98% of the NMOCs are destroyed. Trace
amounts of SO, and NO, will be emitted in each case. With the fuel cell system,
however, significant reductions of NO, and SO, will be achieved due to the fuel cell
energy generation. This analysis assumes an 80% capacity factor for the fuel cell and
offsetting emissions from electric utility power generation using a coal-fired plant
meeting New Source Performance Standards. For the example site, the fuel cell energy
conversi9n system provides 5.6 million kW h of electricity per year, with a net reduction
of NO, of 35.2 tons per year and a reduction of SO, of 16.8 tons per year. These
reductions can be used as environmental offsets, particularly in critical areas such as
California or other locations with severe environmental restrictions.
The environmental impact of application of the fuel cell concept to the potential
. -market is shown in Table 2. The data show that both the flare and the fuel cell
mitigate methane and NMOCs under the proposed standards and guidelines [2].
However, the flare merely converts these emissions to CO2, and rain and other unhealthy
pollutants. The fuel cell can provide a net reduction in global pollution by offsetting
energy production from coal.
Economically the fuel cell energy system has the potential for deriving revenues
from electric sales, thermal sales and emission offsets credits. These revenues can be
used to offset the investment cost associated with gas collection, gas pretreatment and
fuel cell power units. The level of these revenues depends upon the value of the
electricity, the amount and value of the heat used, and the value of the emissions
offsets. . - ' ~ - - "
Economics
Electric rates vary considerably with geographic location and the purchaser of
the electric energy. Commercial rates are applicable where the electricity can be used
at the landfill or in nearby commercial facilities. Commercial rates vary from a high
of 13.68 cents per kW h to a low of 2.71 cents per kW h. The median rate in the
US is approximately 7 cents per kW h. The rates charged to industry are generally
TABLE 2
Emissions impact of fuel cell energy recovery from landfill gas
Abatement .
technology
Venting [2]
only
Flare
Fuel cell
Global wanning
Methane
(Mg/yr.)
1.8X10*
0 .
0
NMOC
-------
362
lower and are closer to the fully burdened avoided cost lor the utility. These rates
range from 10.0 cents per kW h to a low of 1.64 cents per kW h with the mean value
of approximately 5 cents per kW h. In general, both the commercial and industrial
rates are higher in locations with high population density and/or with air emissions
problems. These locations are ideal for the use of the fuel cell energy conversion
. system with its favorable environmental impact. Since' the rates vary considerably, the
analysis in this section is done on a parametric basis for a wide range of electric rates.
The fuel' cell energy conversion system was studied to establish the net revenues
or costs for processing landfill gas to mitigate methane emissions. For the purposes
of the analysis it was assumed that the fuel ceil energy conversion system and the
flare system would have an overall annual capacity factor of 80%. For this analysis,
two levels of fuel cell installed costs were considered. The lower level represents a
fully mature cost when the power plant has been accepted into the marketplace, and
is routinely produced in large quantities. The upper level represents a price level when
the power plant is being introduced into the marketplace, and is produced on a
moderate and continuous basis.
Figure 3 shows the fuel cell revenues for the most stringent application situation
(no emission credits or thermal energy utilization). In this case, the fuel eel] receives
revenues only from the sale of electricity. Although the emissions are lower from the
fuel cell, no specific credit or value is attached to them for this example. Under these
conditions the fuel cell is still the economic choice for most locations at the mature
product installed cost. At the entry cost the fuel cell is economical in those areas
where the value of electricity is 9 cents per kW h or higher. This would primarily be
areas such as California, New York, and other parts of New England. With the potential
for revenue from thermal energy or emission offset credits, the economics become
more competitive. Thus the applicability of the concept would become attractive to
a broader market.
Other energy conversion systems could also produce electric and/or thermal energy.
Both the internal combustion engine and the gas turbine engine have been suggested
as options for methane mitigation at landfill sites. For the landfill size selected for
FUEL CELL INSTALLED COST
MATURE PRODUCT
MARKET
COLLECTION
MO FLAME
6.0
fcO
10.0
110 14.0
VALUE RECEIVED FOR FOB. CELL ELECTRICITY ~ «/KWh
Fig. 3. Comparison of fuel cell .to flare for methane mitigation
electric revenues only.
-------
263
this analysis, the internal combustion engine is more effective than the gas turbine
options for cleanup. This is used as the basis for the comparisons provided here. The
internal combustion engine can provide both heat and electric energy while consuming
the methane at the landfill gas site. With the present state-of-the-art technology,
however, a lean-burn internal combustion engine has higher levels of NOr emissions
than a fuel cell unless special precautions are taken to clean the exhaust. For our
analysis two cases were considered. The first case assumes no cleanup of the-internal
combustion engine exhaust, and the second assumes that the exhaust is cleaned with
selective catalytic reduction (SCR). Since the SCR employs a catalyst in the cleanup
system, the landfill gas will have to be pretreated in a manner similar to the fuel cell
system. For those cases with a SCR cleanup system, a pretreatment system has also
been included as part of the total system cost.
Figure 4 shows the results of the economic analysis for the fuel cell system and
the internal combustion engine system. Since both systems can provide electricity, the
comparison between the systems is based on the cost of electricity generated from
the energy conversion system with appropriate credit for thermal sales and/or emission
offsets. The fuel cell is competitive at the full mature price when no exhaust cleanup
is required with the internal combustion engines. However, the operation of the internal
combustion engine at the landfill site would be quite dirty, and significant amounts
of NO, would be added to the ambient air. For many locations where the fuel cell
would be considered, such as California or other high emissions areas, the exhaust
cleanup option is required. Consequently, the fuel cell option would be fully competitive
with the internal combustion engine option for most cases where on-site cleanup of
the internal combustion engine is required. In areas where a SCR would be employed
to clean up an internal combustion engine exhaust, the fuel cell concept is competitive
at entry level cost.
Based on the analysis of both the flare option and other energy conversion options,
the fuel cell power plant is fully competitive in all situations in the mature production
situation. For initial power plant applications with limited lot production, the fuel cell
power plant is competitive in areas with high electric rates and/or severe emissions
restrictions at the local landfill site.
ELECTRICITY SALES
THERMAL RECOVERY
EMISSIONS OFFSETS
MATURE
PRODUCT
COST
WTTM
sen
EXHAUST
CLEANUP
HO
EXHAUST
CLEANUP
FUEL CELL
ENERGY CONY
SYSTEM
LCf.
ENERGY COWL
SYSTEM
Fig. 4. Comparison of fuel cell to internal combustion engine energy conversion system.
-------
264
Conclusions
Based on the environmental and economic evaluation of the commercial fuel cell
energy system, the following conclusions can be made.
The fuel cell landfill gas to energy conversion system provides a net reduction in
total emissions while simultaneously mitigating the methane from the landfill gas.
Fuei cells will be competitive at initial product prices on landfill sites located in
high electric cost areas or where the thermal energy can be utilized. The fuel cell
will also be attractive where there is a credit for the environmental impact of fuel
cell energy conversion.
When the projected mature product price is achieved, fuel cells will be competitive
for most application scenarios. In many situations, fuel cells will provide net revenues
to the landfill owners. This could, in the long term, result in methane mitigation
without additional cost to the ultimate consumer.
References
1 US Ftderal Register, May 30, 1991, Pan HI Environmental Protection Agency. 40 CFR Parts
51, 52 and 60; Standards of Performance for New Stationary Sources and Guidelines for
Control of Existing Sources; Municipal Solid Waste Landfills; Proposed Rule, Guideline and
Notice of Public Hearing.
2 Air emissions from municipal solid waste landfills background information for proposed
standards and guidelines, EPA-450l3~90-011a (NT1S PB91-197061), Mar. 1991, (a) p. 3-30;
(b) p. 3-23; (c) Tables 3-6, pp. 3-25 to 3-28.
-------
Editor-in-Chief: Professor O. Hutzinger
ATMOSPHERIC METHANE:
SOURCES, SINKS AND ROLE IN
GLOBAL CHANGE
GUEST EDITORS
M. A. K. Khalil and M. J. Shearer
0045-6535(1993)26:1/A;1-8
(INDEXED/ABSTRACTED IN: AnalAbstr, ASCA, Aqua Abstr, Biosis Data, CAB Inter, Cam SciAbstr,
7u/T Cont/AgriBio EnvSci, CABS, Chemosphere, Environ PerBibl, Excerp Med, PASCAL-CNRS Data,
SciCitlnd, SCI SEARCH Data
ISSN 004&-6535
CMSHAF 26 (1-4) 1-814 H993)
Pergamon Press
Oxford New York Seoul Tokyo
-------
-------
<£
A Global Methane Emissions Program for
Landfills, Coal Mines, and Natural .Gas Systems
L.L. Beck
The U.S.~Environmental Protection Agency
Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711 USA
(Received in USA 17 October 1991; accepted 21 May 1992)
ABSTRACT
The Air and Energy Engineering Research Laboratory (AEERL) of EPA's Office of Research and Development has
"chosen anthropogenic methane emissions as a principal focus in its global climate research program. Three of the major
.sources are municipal solid waste landfills, coal mines, and natural gas systems. This paper presents the scope and
methodology of the AEERL methane emission studies and discloses data accumulated thus far in the program. A major
'emphasis in the landfill program is measurement of emissions from operating landfills and calculation of country-specific
Remissions. Landfill methane emissions are not estimated, but factors affecting emissions are discussed and estimates
^developed by others are provided. For coal mines, existing data collected by other researchers on underground mines are
^combined with EPA data on emissions from surface mines to provide an estimate of global emissions of 43 Tg/yr.
"iMtthane from natural gas production, transmission, and distribution systems is estimated to be 4.4 Tg/yr for the United
Stties.
&£.:<
1. INTRODUCTION
The U.S. Environmental Protection Agency's (EPA's) Office of Research and Development has been providing
i since 1988 on the magnitude and sources of radiatively important trace gases (RITGs), with special emphasis on
spheric methane (CH4>. CH4 was chosen as an area of primary concern because of its short atmospheric lifetime (10
pyears) relative to some of the other RITGs such as carbon dioxide (50-200 years); chlorofluorocarbons (65 years for CFC-
['11,130 years for CFC-12); and nitrous oxide (150 years) (IPCC, 1990). Because of its relatively short lifetime and the
ha that the abundance of CH4 increases is due to human activity, the EPA believes that efforts to stabilize atmospheric
'concentrations of this important RTTG have the potential to produce positive results in a relatively short time frame (EPA,
*1990a).
The Air and Energy Engineering Research Laboratory (AEERL) of EPA's Office of Research and Development is
I compiling data on CHj emissions from all sources. However, the AEERL has decided to focus initially on three major
(sources of Cfy: municipal solid waste (MSW) landfills, coal mines, and natural gas systems. These three sources
comprise an estimated 20 percent of anthropogenic CH4 globally (EPA, 1990a), and may be more amenable to cost
I effective mitigation measures than CH4 emissions from other human-related activities such as rice cultivation, enteric
[fermentation, and biomass burning. Also, there is a strong possibility that mitigation measures can utilize the recovered
{4 from these sources as fuel. This would reduce RITG emissions (e.g., carbon dioxide) which would otherwise be
eleased from the burning of fossil fuels. Mitigation of these sources could therefore be justified for reasons apart from
global wanning concerns.
447
-------
There is uncertainty in the magnitude of emissions from landfills, coal mines, and natural gas systems. First, there
ait different emission characteristics among specific categories of sources and individual sources within the categories.
These differences are quantified by developing "emission factors" (quantity of emissions per source). Similarly, there is
uncertainty in the numbers and kinds of sources within each country, or "activity data." The AEERL has chosen first to
reduce the uncertainty in national and global emission inventories by developing more accurate emission factors and activity
data (emission factor x activity data = inventory). An additional programmatic emphasis is to develop first inventories of
U.S. sources, then global inventories. This is because of the accessibility of sources and the need in some cases to develop
or refine methodology for measuring or estimating emissions.
Information on R1TG emissions is being developed in cooperation with the global climate research community, and
is being coordinated with international agencies such as the Organisation for Economic Co-operation and Development
(OECD). The information which follows is accurate from a technical perspective; however, nothing in this paper should be
construed to represent EPA policy.
2. EMISSIONS FROM LANDFILLS, COAL MINES, AND NATURAL GAS SYSTEMS
A. Municipal Solid Waste (MSW) Landfills
CH4 is generated from MSW landfills by the anaerobic decomposition of organic material. Current estimates of
CH« emissions from MSW landfills range from 25 to 40 teragrams per year (Tg/yr) (EPA, 199Qa) (a teragram Is 10«
grams or 1 million metric tonnes), and emissions can continue for 100 years or longer (Thomeloe, 1994). The major
variables being evaluated which are expected to influence the volume and rate of emissions are waste composition,
temperature, moisture, disposal method, and time. The AEERL is developing emission factors for several types of landfills
in a Held testing program (Thomeloe and Peer, 1991). To date, research is in progress at AEERL to collect data at more
than 30 landfills to develop more reliable emission estimates. Research is also being initiated to correlate CH* emissions
with biodegradable components in landfills (i.e., food vs paper vs yard waste) and to validate a procedure which uses
shallow probes to determine gas emission potential directly from specific landfills.
Another important factor in determining global Jandfill-CH4 emissions is the total amount of refuse in place. This
total for any given country in a given year is the sum of the MSW landfilled that year plus the landfilled MSW remaining
from previous years. Country-specific data being collected by AEERL on the amount and composition of landfill waste will
provide data to correlate CH4 generation rate with key variables.
Climate for a given country is also an important variable since Cffj generation rates are dependant on moisture,
temperature, and pH (Thomeloe and Peer, 1990). Because of this, the AEERL program will consider dividing large
countries such as the U.S. and Canada into climatic regions.
Waste composition is another important consideration when estimating the CH» generation potential for a given
landfill, and an AEERL report on (he effect of waste composition on landfill air emissions is planned for June 1992
(Thomeloe, 1991). As might be expected, the greater the proportion of paper, wood, textiles, and other biodegradable
materials, the higher the CH< emission potential. Data are available on the waste composition for industrialized countries
such as the U.S., the U.K., and Canada. Also waste composition is available for some of the major developing countries
such as India. The AEERL will develop estimates where data are not available.
-------
Finally, Die use or systems to collect and utilize or dispose or (e.g., flare) landfill gas will affect the potential for a
landfill to emit CH4 to the atmosphere. The U.S. has the greatest number of existing and planned landfill collection projects,
followed by Germany, (he U.K., and Sweden (Thomeloe and Peer, 1991).
As this brief discussion indicates, landfill emission rates depend on the interrelationship of many complex variables,
most of which have a non-linear relationship to CH4 generation. The AEERL has developed a landfill CH4 emissions model to
assist in processing these data to determine CH4 emission rates from MSW landfills (Peer and Epperson, 1992). The AEERL
is also .continuing to refine the input data and algorithms. In addition to the physical parameters which appear to affect chemical
reaction kinetics, CH4 emissions from MSW landfills appear to correlate well with some demographic data such as population
and gross national product (GNP) (Peer, et al., 1991).
The AEERL is also conducting a demonstration for operating a 200 kW, commercially available, phosphoric acid fuel
cell on landfill gas. Using this technique, the first of its kind, landfill CH4 is reduced and electricity is produced as a
byproduct. Fuel cells are a potentially superior technology to boilers, internal combustion engines, and gas turbines for
mitigation of landfill CH4 because they are highly efficient, quiet, and appear to be environmentally clean. This 4-year project
will Include a I-year demonstration of the fuel cell and landfill gas cleanup equipment.
B. Coal Mines
CH4 is trapped in coal scams and is released to the atmosphere when the coal seam is exposed to the atmosphere. This
happens during mining operations and also during pre-mine degassing operations performed at some mines. The amount of
CH< in the scam is a function of coal age, moisture, and depth. The amount and rate of release to the atmosphere depends on
the physical and chemical characteristics of the coal, mining techniques, and degassing (if performed). CH4 is afco released
from abandoned mines and to a small extent by natural processes such as erosion and diffusion from the coal seam to Ihe
Earth's surface.
Research into coalbed CH4 generation has historically been performed primarily for safety reasons. This has been to
protect miners because of the explosive nature of CH4 when mixed with air in concentrations ranging from 5 to 15 percent.
Because of this, the practice has been to evacuate the CH4 from the mine or to dilute it to less than I percent prior to entry by the
miners. This has been accomplished through mine ventilation and more recently by gob gas wells and pre-mine degassing.
Mines are most commonly ventilated by using continuously operating fans which circulate fresh air across the actively
mined coal face. Because of the large volumes of air needed to keep CH4 concentrations below the lower explosive limit,
ventilation air contains less than 1 percent CH* and therefore has little or no economic value. It is vented directly to the
atmosphere.
CH< is also released from gob wells. Underground mines can release large amounts of CH4 when the fractured area
behind the working longwall face collapses. This collapsed stratum, called "gob," releases CH4. To prevent tWs CH4 from
entering the mining area, gob wells are drilled 2 to 15 m above the area being mined prior to the mining of the longwall panel
[EPA, 1990a). The well operates at a negative pressure and thus removes mine air. Consequently, most gob gas Is usually
vented to the atmosphere. There are exceptions, however. In Alabama over 849.000 mV day from 80 gob wells is captured
md sold as natural gas. Since the program was initiated, over 1.1 x 10" m3 of CH4 has been used as fuel rather than vented to
he atmosphere (Boyer, et al., 1990).
-------
Pic-mine degassing can produce fuel-quality CH« and is an excellent technique for reducing Cfl* emissions to UK
atmosphere, This is practiced for purely economic reasons at some locations with gassy coal, such as in the Black Warrior and
San Juan Basins in Alabama and Colorado, respectively (The Energy Daily, 1989; GRI, 1990).
The AEERL evaluated existing estimates of CH< from coal mining and found that most estimates neglected one or more
Important variables. Most notably, most estimates do not consider emissions from abandoned mines or from surface mines,
even though surface mines are responsible for approximately 40 percent of global coal production. Emissions from surface
mines arc Jess than from underground mines because these coal reserves are closer to the surface (may have "leaked" more), are
younger, and have been exposed to less heat and pressure than the deeper reserves. During the geologic past, the deeper coals
were covered by 10 km or more of overlying strata, resulting in high pressures and temperatures in excess of 300°C (Boyer, et
aJ., 1990). CH< is a product of geochemical coalification, and heat and pressure are the two primary agents which drive the
chemical reactions that comprise the coalification process.
Because of the paucity of data on CHj emissions from surface mines, the AEERL is undertaking a program to measure
emissions from active surface mines. The methodology chosen is to make in-situ, open-path infrared speclroscopy
measurements using a Fourier Transform Infrared Remote (FOR) sensor. This technique involves directing an infrared beam
across an open area to a reflector. The reflected beam is then subjected to absorption analysis to quantify the gases of interest
along the path of the beam. The behavioral characteristics of a plume at a given site are approximated using a tracer gas (sulfur
hexafluoridc) and application of standard Gaussian dispersion equations (Kirchgessneret ai., 1992a).
The AEERL measured emissions from the Caballo mine in Campbell County, Wyoming, which is the Wyoming
Powder River region The Powder River region is recognized for its coalbed CH« resources. Even so, the estimated CH4
emission rate (based on measurements) of 62 million ft 3/yr was five times higher than would have been expected if only
coalbed CH4 content and coal production rale had been used.
Using emission measurements from surface mines, combined with data from underground mines collected by EPA and
others, The AEERL has revised our global emissions inventory for CH< emissions from coal mines. The initial estimate based
on published data and emission measurements is 36 Tg/yr for underground mines and 6.9 Tg/yr for surface mines for a total
global estimate of about 43 Tg/yr (Kirchgessner et al., I992b).
C. Natural Gas Systems
For the purpose of the AEERL study, EPA defines "natural gas systems" as natural gas production, transmission, and
distribution systems. This breakdown of the Industry has been used by other researchers and covers the industry from the gas
production wellhead to the exit of the end-use meter. Oil wells which produce natural gas are also included in the study.
Quantifying leakage (emissions) from natural gas systems is important for several reasons. The leaks are significant from their
direct injection into the atmosphere because they are predominantly CHLj. It is also important to know the leakage rate in order
to evaluate the effectiveness of fuel-switching strategies. Several researchers have suggested switching from coal and other
fossil fuels to natural gas as a sh'ort-term mitigation strategy until cleaner energy sources can be developed and implemented,
since natural gas has 55-60 percent of the carbon per urril of energy as coal (EPA, 1990a;EPA, I990b). In order for such
strategies to be effective, however, the natural gas leakage rate must be known and subtracted from the combustion gains.
Since the relative radiative forcing of CH< is 21 times that of carbon dioxide on a molecule for molecule basis and 58 times
more gram for gram (1PCC, 1990), leakage in the systems carrying natural gas to (lie fuel-switched combustor could offset the
combustion gains.
-------
"Ihe AEERL has eniered into a multi-year program with UK Gas Research Institute (OKI) to quantify leakage rates for
natural gas systems in the United States. As pan of this program a spreadsheet has been developed to assist in tabulating what
is known or estimated regarding leak rates. By using this methodology and attendant statistical evaluation, the initial estimate
for natural gas system leakage is 4.4 Tg/yr for U.S. sources.
The AEERL/GRI study of CH4 emissions from natural gas systems began in 1990 and is expected to continue at least
through 1993. The study is currently evaluating data needs and methodology for filling data gaps. Some of the methodology
being considered is:
Release of tracer gas coupled with upwind/downwind sampl ing,
- Meter balance techniques,
Use of "sniffing" devices like organic vapor analyzers to determine rales of leakage from valves and fittings, and
- Direct measurement of emissions by bagging the component and sampling the collected emissions.
Once emissions data are collected, statistical methodology will be used to apply the data to estimate U.S. emissions.
3. SUMMARY AND COMMENT
The AEERL has concurrent programs directed at reducing the uncertainty of estimates of CfU emissions from landfills,
coal mines, and natural gas systems. These programs were begun in 1990 and are in various stages of completion. The near-
term focus of the MSW landfill program is to develop emission factors from U.S. landfills in a field test program. The coal
mine program is applying state-of-the-art, in-situ measurement techniques to establish CH< emissions from surface mines, and
will combine these data with knowledge about underground mine emissions to estimate global emissions. The natural gas study
is concentrating first on reducing the uncertainty of emissions estimates from U.S. sources, and will apply this knowledge to
establish Cfy emission estimates for other countries which have natural gas systems. Table 1 presents current estimates of
emissions from MSW landfills, coal mines, and natural gas systems.
Table 1. Estimates of CH4 Emissions from MSW Landfills,
Coal Mines, and Natural Gas Systems (Tg/Yr) -
MSW Landfills (Global)
Coal Mines (Global)
Natural Gas Systems (U.S. Only)
N/A-Not Available
b Range given in IPCC (1990)
Estimate
N/A«
43
4.4
Range
20 - 70<»
19 - 50b
0.51 - 8.4
15'I
-------
Ilie AEERL is also researching CH< emissions from cookstoves and biomass burning, though these
programs are not as mature as the areas highlighted by this paper. The AEERL is also planning to begin
research on other anthropogenic sources of CH* emissions. Some of the areas being considered are industrial
and hazardous waste land/Ills, waste water treatment lagoons, septic systems, disposal of agricultural wastes,
open burning of municipal solid waste, and sludge disposal.
The information being collected on CH< and other RITCs will be entered into a computerized database
management system being developed by AEERL. The soflwarc, called Global Emissions Data (GloED)
system, will serve as a repository for data and will be updated as new or better data become available. The
software will be available early in 1992 for peer review and appraisal by other researchers.
REFERENCES
Boyer. CM., J.R. Kclafanl, V.A. Kuuskraa, et al. (1990). Methane Emissjons from Coal Mining; Issues
and Opportunities for Reduction. Office of Air and Radiation. EPA/40U/9-90/U08. September.
EPA (U.S. Environmental Protection Agency) (I990a), Methane Emissions and Opportunities for Coptrol.
Workshop Results of Intergovernmental Panel on Climate Change. EPA/400/9-90A)07.
EPA (U.S. Environmental Protection Agency) H990b). Policy Options for Stabilizing Global Climate. Office
of Policy, Planning, and Evaluation. EPA Document No. 21P-2003.
GRI (1990), Quarterly Review of Methane fromJToal gcaros Technology. March, (Gas Research Institute,
Chicago, IL). Vol. 7, No. 3, ISSN 8756-9655.
IPCC(1990), (Intergovernmental Panel on Climate Change), CJjmate Change The IPCC Scientific
Assessment. Haughton, J.T., G.J. Jenkins, and J.J. Ephraums, Editors, (University Press, Cambridge).
ISBN 0-521-40720-6.
KJrchgcssner, D.A., S.D. Piccot, and A. Chadha (I992a), Estimation of Methane Emissions from a Surface
Coal Mine Using Ooen-Palh FT1R Spectroscopv and Modeling Techniques. Accepted for Publication in
Chcmosphcre.
KJrchgcssner, D.A., S.D. Piccot, and J.D. Winkler (1992b), Estimate of Global Meihane Emissions, from
Coal Mines. Accepted for Publication in Chcmosphere.
Peer, R.L., A.E. Leininger, B.B. Emmel, et al. (U.S. Environmental Protection Agency) (1991). Approach
for Estimating Global Landfill Methane Emissir " """" "~' ':** »--« *»*
600/7-91-002 (NT1S PB9M49534), January.
for Estimating Global Landfill Mcjhane Emissions. Air and Energy Engineering Research Laboratory. EPA-
Peer, R.L., and D.L. Epperson (1992), Development of an Empirical. Model fpr Methane Emissions from
Landfills. Air and Energy Engineering Research Laboratory. EPA-600/R-92-037 (NTIS PB92-I52875),
March.
The Energy Daily (1989), How "Moonbeam Gas" Made Its jVav from Dream (o Market. July 18.
Thomeloe. S.A. (1991), U.S. EPA's Global Climate Change Program - Landfill Emission^ and. Mitigation
, Paper presented at Sardina 91 Third International Landfill Symposium in Cagliari, Italy.
Thomeloe, S.A., and R.L. Peer (1991), EPA's Global Climate Change Program Global Landfill Methane.
AWMA Annual Meeting, Vancouver, B.C.
Thomeloe, S.A., and R.L. Peer, (1990), L.andfHI Gas and the (Greenhouse Effect. Presented at ihe
International Conference on Landfill Gas: Energy and Environment, Bournemouth, England.
-------
To be included in NATO Book
The Global Methane Cyde:
Its Sources, Sinks, Distributions and Role in Global Change
1993
Methane Emissions from Industrial Sources
" L. LEE BECK,1 STEPHEN D. Piccor,2 AND DAVID A. KiRCHGEssNER1
'United States Environmental Protection Agency, Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
1 Southern Research Institute, Environmental Studies Division,
Research Triangle Park, North Carolina 27709
Introduction - -
This chapter identifies and describes major industrial sources of methane (CH4) emissions. For
each source type examined, CH* release points are identified and a detailed discussion of the factors
affecting emissions is provided. A summary and discussion of available global and country-specific
CH4 emissions estimates are also presented.
The major emission sources examined include coal mining operations and natural gas
production and distribution systems. However, a variety of minor industrial sources are also
examined because their collective contributions to the global CH4 budget may be significant.
Although the treatment of these minor sources may not be comprehensive, the limited available data
are presented for several different sources. Among the minor industrial sources examined here are:
coke production facilities, chemical manufacturing operations, peat mining operations, light water
nuclear reactors, fossil fuel combustion equipment (boilers and automobiles), geothermal electricity
generation facilities, salt mining operations, residential refuse burning, and shale oil mining
operations. .-'- '
Coal mining operations
Historically, coal use in the United States suffered a nearly catastrophic decline after World
War II. It was not until the mid 1970s that bituminous coal production once again equalled the levels
seen in the mid 1940s. Production in 1970 was about 545 million tonnes and had grown
more or less consistently to 830 million tonnes by 1987. This growth will undoubtedly be mirrored
-------
in countries that have significant coal reserves and are undergoing industrialization. The trend is
perhaps epitomized in China where coal production rose from 620 to 1050 million tonnes between
1980 and 1989.
The global environmental effects of coal production and use will increase as coal production
increases. The underground mining of coal is accompanied by the emission of CH4 and the
production of large quantities of waste water and coal wastes. The use of longwall mining often
results in subsidence at the surface, which can produce significant property damage and can be costly
to prevent. Surface mining can produce significant scarring of the land, but, if proper land
reclamation techniques are applied, surface mine sites can be restored to near original conditions.
Because of the high cost of these techniques, this restoration may not be done hi all countries.
Combustion of coal results in the release of large quantities of "conventional" pollutants including
sulfur dioxide, nitrogen oxides, and paniculate matter. It is also one of the largest sources of
anthropogenic carbon dioxide emissions. Although several of these pollutants can be controlled with
today's pollution control technologies, the large capital investments required may not always be
available in developing countries.
Sources of methane emissions in the coal mining industry. Three categories of mines within
the coal mining industry are known to emit CH4 to the atmosphere: underground mines, surface
mines, and abandoned or inactive mines. Much more is known about the emissions from
underground mines than from any other mining category. Based on currently available data,
underground mines are generally believed to be the most significant source of CH4 within the coal
muiing industry.
In general, CH4 can be released into a coal mine from all of the seams disturbed during the
mining process. Seams other than the one being mined may be disturbed by the mechanical or
blasting operations which occur at underground and surface coal mines. In an underground longwall
mine, some studies suggest that this zone of disturbance may extend up to 200 meters (m) into the
roof rock and 100 m below the worked seam (Greedy, 1983). For safety reasons, underground mines
use various methods to remove CH4 from the mine; however, CH4 is not a safety concern in surface
mines. In some countries, a portion of the CH4 removed from underground mine workings is burned
for energy, but in most cases the CH4 is released to the atmosphere.
Methane from underground mines can be released from three sources: ventilation shafts; gas
drainage systems; and coal crushing and handling operations (Boyer et al., 1990; Piccot et al., 1990).
Figure 1 illustrates these sources. Ventilation air, although generally containing 1 percent or less
CH4, is known to contribute the majority of underground mine emissions because of the enormous
volume of air used to ventilate mines. Gas drainage wells are drilled into the area immediately above
the seam being mined. They provide conduits for venting CH*, which accumulates in the rubble-
-------
CH4
Advance!
Degas /
Well
UNDERGROUND MINE
Shaft/Ventilation System
Caved-in
Area
CH,
SURFACE MINE
Drilling and Blasting
Overburden
Coal Loading
Figure 1. Sources of methane emissions at underground and surface coal mines.
-------
filled areas formed when the mine roof subsides following longwall mining. Other types of drainage
systems are used that extract CH4 from coal seams well in advance of mining operations. The
purpose of all gas drainage systems is to remove CH4 that would otherwise have to be removed by
larger and more costly shaft ventilation systems. Currently, few published data exist for the release
of CH4 from gas drainage systems. However, unpublished data obtained from industry
representatives indicate that drainage well CH4 emissions may account for a significant fraction of
the total emissions associated with longwall mines (Boyer et al., 1990; Kirchgessner et al., 1993a).
In some parts of the world, such as western and eastern Europe, CH4 from gas drainage systems is
utilized and is not released to the atmosphere.
Very few measurements have been taken at surface mines. Data from six surface mines in the
United States suggest that the primary sources of emissions are exposed coal surfaces, in particular
the areas fractured by coal blasting (Piccot et al., 1991; Kirchgessner et al., 1993b). In general, the
strata overlying the coal do not appear to be a significant source of emissions, but, as in underground
mines, emissions may be contributed by underlying seams or faults.
Emissions from abandoned mines may come from unsealed mine shafts or from vents installed
to prevent the buildup of CH* in the mines. There has been little research to characterize this
potential source. However, in a continuing study by the U.S. Environmental Protection Agency
(USEPA), CH« emissions have been measured as they escape from vents installed at abandoned
underground mines. Although only a small number of mines have been examined, preliminary results
indicate that emissions vary significantly and that the rate of emissions from an abandoned mine may
be strongly influenced by barometric pressure changes. For several mines, emission rates were
negligible while at one mine emissions were almost half of those produced when the mine was active.
The rate of release of CH4 from mined coal varies depending on coal type, local geology,
geologic history, and other factors. Methane can be released quickly, in a matter of hours, or slowly,
over a period of several months. Since mined coal is typically removed within a day, there is a
potential for emissions to occur in the post-mining operations. Post-mining operations can include
coal breaking, crushing, drying, storage, and transportation. Although measurement data are very
limited, Boyer et al. (1990) estimate that on average 25 percent of the CH4 contained hi the mined
coal could be emitted after the coal has left the underground mine. Researchers at British Coal
suggest that coal handling and transport operations in Great Britain may produce about 2 nrVtonne
of coal when a coal with a CH4 content of 5 mVtonne is mined (i.e., 40 percent of the CH4 contained
in the mined coal is released after it leaves the mine; Watt Committee on Energy, 1991).
Factors affecting methane emissions from coal mines.
Numerous studies have examined the physical factors which control the production and release
of CH4 by coal. These studies have been conducted to evaluate the potential of coalbed CH4
-------
resources, enhance the safety of underground mines, or estimate global CH4 emissions. Generally,
the studies address one of two topics: controlling the CH* content of coals, or controlling the
concentration of CH4 in the mine atmosphere .and mine ventilation air. Most of these studies have
focused on underground mining operations.
Studies of coalbed CH4 contents have identified pressure, coal rank, and moisture content as
important determinants of coalbed CH4 content. Kim (1977) related gas content to coal temperature
and pressure, and in turn to coal depth. After including coal analysis data to represent rank, Kim
produced a diagram relating gas content to coal depth and rank. Although the validity of the rank
relationship has been questioned, it generally appears to have been accepted by recent authors
(Schwarzer and Byrer, 1983; Lambert et al.. 1980; Murray, 1980; Ameri et al., 1981).
Independently of Kim's work, Basic and Vukic (1989) established the relationship of CH< content
with depth in brown coals and lignite.
Several studies have recognized the decrease in CH4 adsorption on coal as moisture content
increases in the lowest moisture regimes (Anderson and Hofer, 1965; Jolly et al., 1968; Joubert et
al., 1974). Moisture content appears to reach a critical value above which further increases produce
no significant change in CH4 content. Coals studied by Joubert et al. (1974) showed critical values
ranging from 1 to 3 percent. . .
In a recent study by Kirchgessner et al. (1993a), a set procedure was developed which
integrates the influences of several factors known to affect CH4 content in coalbeds. Two equations
were produced for estimating coalbed CH4 contents in cubic meters per tonne at the basin or seam
level: one for coals with heating values less than 34,860 joules/gram (J/g) (equation 1 below), and
one for coals with heating values equal to or greater than 34,860 J/g (equation 2 below). The
equations for estimating coal bed CH4 content (IS) were developed by performing multivariate
regression analyses on a database of 137 U.S. coal samples. The R1 for equation 1 is 0.56, and the
R2 for equation 2 is 0.71. Coal properties in the equations have sound geological bases for inclusion
and the parameters included follow patterns predicted in the literature. Coal depth (D) in meters
appears in both equations. Moisture content (M) in percent and parameters closely tied to coal rank
such as heating value (HV) hi joules/gram, and the ratio of carbon content to volatile matter (FR) are
also included. As a final step hi the procedure, the IS values obtained from the equations are
multiplied by a factor to yield an estimate of total CH* content (i.e., anthracite 1.11; low volatile
bituminous 1.10; medium volatile bituminous 1.20; high volatile B and C bituminous 1.12).
IS = 0.0159D + 2.781 ± - 2.228
(1)
Although more difficult to quantify, the amount of CH4 contained in a given quantity of coal
may be influenced by the burial and erosional history of a coal seam (Watt Committee on Energy,
-------
IS = 0.0136D + 0.0015HV + 2.6809FR - 56.490
(2)
1991). In some geological configurations, the CH4 content of coal increases with coal rank but
decreases on approach to the Permo-carboniferous erosion surface.
Early investigations in the United States which attempt to identify correlates of CH, emissions
from coal mine ventilation air include those by Irani et al. (1972) and Kissel et al. (1973). Irani et
al. developed a linear relationship between CH4 emissions and coal seam depth for mines located in
five seams. Kissel et al. demonstrated a linear relationship between CH4 emissions and coalbed CH4
content for six mines. Although both studies suffer from a paucity of mines and seams in their
analyses, Kissel et al. made the important observation that mine emissions greatly exceed the amount
of CH4 associated with the mined coal seam alone. Emissions are produced not only by the mined
coal, but also by the coal left behind, overlying and underlying seams, and nearby gas deposits. For
the six mines studied, emissions per tonne mined exceeded coalbed CH4 per tonne mined by factors
of six to nine.
In studies conducted by Boyer et al. (1990) and Kirchgessner, et al. (1993a), regression
equations were developed relating coal production rate and coalbed CH4 content to the emissions from
underground mines. These equations were developed to estimate global emissions from underground
mines, using similar data and techniques. To develop these equations, multivariate regression
analyses were performed using mine-specific data for the United States. In the analysis performed
by Kirchgessner et al., a database of 269 mine-specific emission measurements was used to produce
an equation with an R* value of 0.59. This means that about 60 percent of the variation in CH4
emissions from the mines in the database can be explained by the independent variables included in
'-/'.
the equation. In the analyses performed by Boyer et al., a database of about 60 mine-specific
observations was used to produce an equation with an R2 value of 0.35. Although not fully
understood, one likely reason for the difference hi R2 values between the two studies is that the
equation developed by Kirchgessner et al. was estimated using a database which contained over 4
times more individual mine measurements than the Boyer equation.
Summary of global emissions estimates. Over the past 30 years there have been several
attempts to estimate the global emissions of CI^ from coal mining operations. Although we do not
address them all, the estimates presented represent the approximate range of published emission rates.
This section summarizes these estimates, describes and compares the basic assumptions used hi their
development, and identifies key relationships that exist among them.
Table 1 presents a summary of global CH« emissions estimates developed by various
researchers for coal mining operations. Estimates range from 7.9 to 64 teragrams (Tg) per year.
The lower estimate of 7.9 Tg/yr is unrealistic. Although the specific assumptions used in developing
-------
this estimate are not clear, it appears to be based on the implicit assumption that emissions from coal
mines are equal to the amount of CH4 trapped in the coal removed from the mine (Hitchcock and
Wechsler, 1972; Bates and Witherspoon, 1952). Although this trapped CH4 is released when coal
is fractured and removed from the mine, this assumption fails to account for other CH4 release
mechanisms that occur. These release mechanisms, which were described earlier, can significantly
contribute to the total emissions from mining operations. A second low estimate of 13.1 Tg/year
reported by Darmstadter et al. (1984) for 1980 appears to be based on an unrealistically low emission
factor.
A review of the global estimates presented in Table 1 reveals that many are closely related; that
is, the basis of several estimates can be traced back to key assumptions made by some of the earliest
researchers. Figure 2 illustrates the relationships which exist between various global estimates of CH4
DATA ON COAL PROPERTIES, ETC.
i i
V
J_
KOYAMA
I
BATES AND
WITHERSPOON
SEILER
_1_
HITCHCOCK fi
WECHSLER
I
EHHALT,
EHHALT AND
SCHMIDT
CICERONE AND
OREMLAND
CRUTZEN
BOYER et al.
KIRCHGESSNER, et al.
Figure 2. Relationships among various global estimates of
methane emissions from coal mines.
emissions from coal mines. As the figure shows, many published estimates have been based
-------
primarily on methodologies developed by Koyama (1963, 1964) and Bates and Witherspoon (1952).
In general, estimates developed by Seiler (1984), Hitchcock and Wechsler (1972), Ehhalt (1974),
Ehhalt and Schmidt (1978), Crutzen (1987), and Cicerone and Oremland (1988) were developed
based in large part on CH4 emission factors developed by Bates and Witherspoon (1952) and Koyama
(1963, 1964). Koyama's emission factors have been used by Hitchcock and Wechsler, and Ehhalt
and Schmidt to estimate an upper end of the range of coal mine emissions.
Estimates by Boyer et al. (1990) and Kirchgessner et al. (1993a) are on a country-specific basis
and were not developed based on other researchers' emission factors. Instead, new emissions
relationships were developed based on measurement data contained in databases on coal properties
and mine emissions rates.
The coal mine estimate of Fung et al. (1991) presented in Table 1 was developed differently
from the other estimates discussed here. Fung et al. used a combination of global CH4 mass balances
and atmospheric modeling techniques to infer a budget for all CH« sources including coal mines. In
general, several CH4 budget scenarios were constructed and tested to determine which was best able
to reproduce the meridional gradient and seasonal variations of CH« concentrations observed i$ the
atmosphere. One budget scenario for all CH4 sources, including coal mines, was selected by Fung
et al. because it was judged to reproduce the atmospheric record best. The coal mine estimate
associated with this budget scenario is shown in Table 1. There are methodological and other
differences among the estimates that contribute to the variations observed in the results hi Table 1.
First, the estimates are developed for different years: estimates from 1960 through 1989 are
presented. Since coal production increased significantly during this period, an increase in emissions
is expected and a direct comparison of these estimates cannot be made. Second, Table 1 shows that
several estimates fail to account for all the coal produced globally. Estimates developed by Koyama
(1963, 1964), Crutzen (1987), and Cicerone and Oremland (1988) are known to account for the
emissions associated with hard coal production only and do not include the emissions associated with
brown or lignite coals. Although these types of coals typically do not contain much CH«, their
contribution to global emissions cannot be neglected. Estimates developed by Hitchcock and
Wechsler (1972). Ehhalt and Schmidt (1978), Boyer et al. (1990), and Kirchgessner et al. (1993a)
include emissions associated with all coal types. Another difference among these estimates is that
many do not appear to include the global emissions associated with post-mining operations (i.e.,
crushing, grinding, handling, and transport). The estimates developed by Boyer et al. and
Kirchgessner et al. are the only ones that specifically include an estimate of the emissions from post-
mining operations. None of the estimates includes the emissions associated with abandoned mines.
A simple evaluation of the global estimates presented in Table 1 could lead to the potentially
erroneous conclusion that the two-fold increase in CH4 emissions since 1960 can be explained
-------
primarily by the two-fold increase in global coal production that occurred during the same period.
However, the actual change in global coal mine emissions cannot be determined from the estimates
hi Table 1 because significant differences exist hi the emissions rates used to develop those estimates,
and many estimates do not include the emissions from lignite coals. Table 2 compares CH4 emission
Table 2. Summary of methane emission rates associated with various global emissions
estimates.
Source
Koyama (1963, 1964)
Hitchcock and Wechster
(1972)
Seller (1984)
Crutzen (1987)
Boyer, et al. (1990)
Kirchgessner, et al. (1993 a)
Year
1960
1967
1975
not specified
(assume middle
1980s)
1987
1989
Emission Rate
(m3 methane/tonne
coal mined)
19.5
. 5.0 to 17.5
19.5
18 to 19»
14.2
13.8
As cited hi Boyer et al. (1990).
rates associated with a variety of studies. These data indicate that the emission rates used to estimate
emissions in the most recent studies tend to be lower than those used in early studies. . "''
In two recent studies, emissions from coal mines were estimated on a country-specific basis
(Boyer et al., 1990; Kirchgessner et al., 1991a). Both studies represent relatively comprehensive
attempts to characterize key country-specific factors which may significantly influence emissions by
(1) estimating country-level or basin-level coalbed CH4 contents, (2) estimating the emissions
associated with different mining techniques (i.e., surface mining and underground mining), and (3)
subtracting CH4 recovered and used at coal mines from the global estimate. Summaries of the
country-specific estimates from both studies are presented hi Table 3. A comparison of the two
studies shows that total emissions are relatively similar but significant differences exist hi the emission
estimates for key countries and regions. For China, the United States, South Africa, and India,
Boyer et al. produce estimates that are about two tunes higher than Kirchgessner et al. Conversely,
estimates for the "Rest of the World" and "Surface Mining" are higher in Kirchgessner et al. by a
factor of about two. The reasons for these differences have not been determined.
-------
Table 3. Country-specific emission estimates developed by Boyer, et al. (1990)
and Kirchgessner et al. (1993a).
Country
Underground
Mining
China
Former Soviet
Union
United States
Poland
South Africa
India
United Kingdom
West Germany
Australia
Czechoslovakia
Rest of the World
Subtotal
Surface Mining
Subtotal
Post-Mining
. Emissions
Total Mining
Boyer et
1987 Coal
Production
( I O6 tonnes)
891
429
337
193
111
85
86
79
_47
26
Not Reported
-
Not Reported
-
4,630°
A range of emissions from 30.3 to
underground mines.
* A range of emissions from 2.6 to
surface mines. . ,
c This value cannot
.
be obtained by
al. (1990)
1987 CH4
Kirchgessner et al. (1993a)
1989 Coal
Emissions Production
(teragrams)
16.0
7.7
6.1
3.3
2.0
1.5
1.5
1.4
0.9
0.4
2.9
43.71
3.7"
. . Included
above
47.4
59. 1 million tonnes
5.0 million tonnes
" ; '
summing th« nmHi
(10* tonnes)
1,053
418
356
181
115
95
71
73
59
-
567
2,988
2,154
-
5,142
was established
was established
;* ' *
1989 CH4
Emissions
(teragrams)
9.3
7.9
3.5
3.6
0.7
0.7
1.3
1.1
1.1
-
6.8
36.0
6.9
2.7
45.6
in this study for
in this study for
iction rates listed above because
values for Surface Mining and the Rest of the World were not reported by Boyer et al.;
however, total production was reported.
-------
Other researchers have also estimated emissions for individual countries. Although these
estimates are not global in nature, they are reported here because they generally represent the results
of a focused and detailed assessment of country-specific coal and mine emission characteristics. As
a result, they can be used to independently examine the representativeness of the country-specific
estimates presented in Table 3. The results are summarized in Table 4. The reader should be
cautioned that Table 4 is not intended to be a comprehensive summary of individual country
estimates. Although the table includes mainly those results obtained from participants in the NATO
Advanced Research Workshop on the Global Methane Cycle (this volume and in Khalil and Shearer,
1993), other independent assessments are also included.
Table 4. Summary of other country-specific mine emission estimates.
Country
Emissions
(tg/yr)
Source
Comments
Australia 0.4S
D.J. Williams,
CSIRO, Minerals
Research
Laboratories
Australia
Preliminary 1989 estimate for underground
mines only.
Former
Soviet
Union
Poland
Turkey
United
Kingdom
United
States
3.5 to
11.2
(average
7-4>
3.3
0.22
0.75 ±
0.1
3.0
Andronova and Karol
(1993)
Pilcher et al. (1991)
Personal communi-
cation with H. Kose
and T. Onargan*
Greedy (1993)
Piccot and Saeger - , -
(1990)
Estimate is for 1988. The maximum
potential emissions are 17.3 Tg.
Estimate is for 1988.
Estimate is for 1990.
This measurements-based estimate includes
emissions from all coal types and coal
handing and transport losses (1990/1991).
Estimate includes emissions for under-
ground coal mines only (estimate for 1985)
Both are from the University of The 9 Eylul, Izmir, Turkey.
The country-specific estimates in Table 4 generally agree with the estimates in Table 3. The
estimates of Kirchgessner et al. and Boyer et al. generally agree with estimates developed for Poland
and the former Soviet Union by Pilcher et al. (1991) and Andronova and Karol (1993). However,
estimates for Australia and the United Kingdom developed by Kirchgessner et al. and Boyer et al.
are about two times higher than the estimates presented hi Table 4. Estimates for the United
Kingdom developed by Kirchgessner et al. are based on coalbed CH4 content measurements for seams
mined in the United Kingdom, and on mine emissions relationships developed from 260 U.S. coal
-------
mine emissions measurements. The Greedy estimate is based on relationships developed from United
Kingdom measurements data supplied by British Coal for both coalbed CH4 content and mine
emissions rates. The discrepancy between the two estimates lies in the assumptions that relate
coalbed CH4 content by depth to mine production to calculate total emissions from underground
mines.
Natural gas production and distribution
Natural gas has long been recognized as the environmentally preferred fossil fuel. It produces
virtually no sulfur dioxide or paniculate emissions, and far fewer nitrogen oxide and carbon
monoxide emissions than other fossil fuels. For this reason, and because it is widely available,
relatively easily recovered, and readily usable, the global consumption of natural gas has
approximately doubled since 1970. Among the industrialized nations, the United States is an
exception to this pattern in that it consumes about 10 percent less natural gas today than in 1970.
This has variously been attributed to an excessively restrictive regulatory structure (DOE, 1991) and
to a misconception of the future natural gas price structure stemming from an underestimate of
available U.S. reserves in the 1970s (Hay et al., 1988). The U.S. National Energy Strategy produced
in 1991 has recommended removing or revising excessive regulation inhibiting natural gas
transactions (DOE, 1991), and the Gas Research Institute has stated that domestic natural gas reserves
are sufficient for the next several decades (Hay et al., 1988). Both of these factors should accelerate
the slow rate of increase in U.S. domestic natural gas utilization which is already occurring.
Natural gas emits about half as much carbon dioxide per unit of energy output as coal, and
about two-thirds as much as oil. Recognizing this, the Intergovernmental Panel on Climate Change
-'.;
has formalized the recommendation to switch to natural gas as fuel where possible to achieve short
term mitigation of the global climate change problem (Environment Agency of Japan, 1990). It must
first be demonstrated, however, that CH4 leakage from the increased production and utilization of
natural gas would not nullify the benefit of decreased carbon dioxide production.
Sources of methane emissions in the natural gas industry. The natural gas industry can be
broadly divided into the production, transmission, and distribution sectors diagramed in Figure 3.
Each of these sectors can contribute steady or fugitive CH4 emissions and intermittent emissions.
Fugitive emissions result from normal operations and result primarily from leaking components such
as valves, flanges, and seals. Intermittent emissions result from routine maintenance procedures,
system upsets, and occasional large scale accidents.
Methane emissions from the production sector usually include those from well drilling, gas
extraction, and field separation facilities. In this discussion gas processing plants are also included.
Emissions from well drilling result primarily from occasional venting and flaring employed to prevent
-------
Compressor
Stations
PRODUCING
WELLS
TRANSMISSION
LINES
Underground
Storage
DISTRIBUTION
SYSTEM
Industrial Consumer
Residential
Consumers
Figure 3. Natural gas pipeline system.
IS
-------
blowouts. During extraction, CH4 may be emitted by natural-gas-fired engines used for power
generation, various wellhead components collectively referred to as the "Christmas tree," and
occasional venting and flaring when gas volumes do not warrant recovery. Field separation may
involve gas heating, gas or liquid separation, and gas dehydration. Principal sources of emissions
are fugitive leaks, venting and flaring, natural-gas-powered pneumatic devices, and combustion losses
from heaters and dehydrators. Gas processing plants are usually located close to the production area
and may be regarded as part of the production process. Gas plants are used to separate natural gas
liquids from the gas stream and to fractionate the liquids into their components. The processes which
are currently most commonly used in these plants are cryogenic expansion, refrigeration, and
refrigerated absorption. Primary emissions sources from gas processing plants are fugitive losses,
compressor exhaust, and venting and flaring.
Methane emissions associated with the transmission sector are produced by the pipelines,
compressor stations, and metering and pressure regulation stations. Leaks from the pipelines are
caused by corrosion, material and construction defects, miscellaneous leaks at valves, flanges, and
finings, and earth movement which can cause strains and cracks. Venting can occur at points in the
pipeline where residual liquids collect and must be drained. Pneumatic devices powered by natural
gas are found throughout the transmission sector and are typically vented to the atmosphere.
Maintenance procedures such as pipe scraping result in emissions during launching and retrieving of
the scraper. Dehydrators must receive periodic blowdowns and purges which are vented, and
pipelines must occasionally be purged during installations, abandonments, replacements, repairs, and
emergency shutdowns. Compressor stations produce fugitive emissions from the usual sources (e.g.,
flanges, seals), occasional unfiared venting from system overpressure, and gas turbine start-up and
operating emissions.
The primary sources of emissions from the distribution system, which delivers natural gas to
the end users, are pipeline leaks. These leaks result, in varying degrees, from all of the same causes
as leaks in transmission pipelines. Gas is intentionally vented after isolating segments of lines for
repair, and is used to purge ah- from the pipeline after repair. Blow and purge operations on meters
and regulators are typically vented to the atmosphere. The distribution system, because of its size,
is generally regarded as the most significant source of CH4 emissions in the natural gas network.
Injection facilities can be located at various points in the system, depending upon the facilities*
function. Gas is frequently reinjected at the production site to maintain oil or gas reservoir pressure.
Gas is also injected into underground reservoirs for storage. Normal operations at these facilities
produce the usual fugitive emissions, releases during routine maintenance, and venting for
overpressure protection of compressors, scrubber vessels, and wellhead injection stations,,
The final category of emission sources (not discussed under the three-pan industry breakdown
-------
above) is liquefied natural gas (LNG) facilities. Functions performed at an LNG facility include
receiving, storage, and regasification. Equipment consists of unloading piping, pumps, insulated
storage tanks for LNG, and heaters and compressors for regasification. During normal operation,
fugitive releases occur but, because of the nature of these facilities, maintenance can be scheduled
well in advance and the necessary controlled venting can be directed to the flare system. Pressure
relief system releases are typically flared as well.
Summary of Emission Estimates. Numerous estimates of emissions for the natural gas industry
are available. Global emissions estimates from as long as 20 years ago have been produced,
primarily for the purpose of determining global balances of atmospheric trace gases, but more
recently for assessing global climate change issues. In the past few years, as a result of the
awareness of methane's role in global climate change, country-specific and sector-specific estimates
of emissions have also become available. Estimates of global emissions from the natural gas industry
are summarized in Table 5. Estimates produced during the 1980s typically range from 25 to SO
Tg/yr and assume leakage rates of from 1 to 4 percent. Ehhalt and Schmidt's (1978) estimate of 7
to 21 Tg/yr is a notable exception but is explained by their acceptance of Hitchcock and Wechsler's
(1972) estimate. This estimate would correspond with the others if expanded to 1985 gas production
values. Seller's (1984) estimate of 19 to 29 Tg/yr is also low but is explained by his use of 1975 gas
production data. Estimates at the higher end of the typical range by Sheppard et al. (1982), Blake
(1984), and Cicerone and Oremland (1988) are derived by adding assumed values for vented gas to
the calculated values for gas leakage. Keeling (1973) assumed a leakage rate of 6-10 percent and
estimated emissions of 40-70 Tg/yr.
Table 5. Estimates of global emissions for the natural gas industry.
Source
Hitchcock and Wechsler
(1972)
Keeling (1973)
Ehhalt and Schmidt (1978)
Sheppard et al. (1982)
Reported Year
1968
1968
1968
1975
Estimate (Tg/yr)
7-21
40-70
7-21
50
Assumed Loss
1-3
6-10
1-3
2 (leakage) +
Rates %
25% for
Blake (1984)
Seiler (1984)
Bolle et al. (1986)
1975
1075
Not specified
50-60
19-29
35
vented and flared
2-3 (leakage) + 30 Tg
for vented
2-3
-------
Crutzen (1987)
Darmstadter et al. (1984)
Cicerone and Oremland
(1988)
Fung et al. (1991)
Not specified
1980
Early 1980s
1986
33
10
25-50
40
4
1
. 2.5 (leakage)
for vented &
Unknown
+ 14 Tg
flared
Sheppard et al. (1982) and subsequently Blake (1984) estimate emissions from venting and
flaring at wellheads to be about 30 Tg/yr. Cicerone and Oremland (1988) provide a later,
independent estimate of 14 Tg/yr. It can be inferred from the discussions that these estimates are for
both gas and oil fields and it is assumed that oil fields produce the majority of emissions. The
estimates are not separated by industry, however, and such matters as flaring efficiencies and venting
versus flaring practices by individual countries are not discussed. There are currently no reliable data
on global venting and flaring emissions from oil and gas fields.
While it is not always explicitly stated in the literature, the assumed leakage rates fall into the
estimate that the gas industry refers to as "unaccounted for" gas (UAG). UAG is the difference
between the volume of gas that a utility reports as purchased versus the volume sold, less any
company use or interchange. It is a statistical figure attributable to numerous diverse components,
including meter inaccuracies, gas theft, variations hi temperature and pressure, billing cycle
differences, and gas leakage or other actual losses. In the United States roughly 2 percent of gas
marketed annually is classified as UAG (American Gas Association, 1986). Because UAG reflects
nuances of gas companies' accounting systems rather than actual gas losses, it is generally recognized
that emissions estimates should not be based on these statistics.
Table 6. Summary of studies showing gas losses by industry sector.
Source
Location & Year
Summary of Findings
Description
Emissions
Tilkicioglu and
Winters (1989)
Cottengham et al.
(1989)
USA-1988
PG&E"
Gas field and field
separation facilities
Gas transmission
Gas distribution
Gas process plants
Total
Distribution
Transmission
Dig-ins
Total
U2Tg/yr
0.97 Tg/yr
0.43 Tg/yr
0.31 Tg/yr
2.83 Tg/yr
498,000 Mcf/yr"
42,000 Mcf/yr
106,000 Mcf/yr
646,000 Mcf/yr or
0.08% of adjusted
operating receipts
-------
Chem Systems
International Ltd.
(1989)
USA-1988
W. Germany
1989
Netherlands 1989
UK 1989
Production
Transmission
Distribution
Overall
Production
Transmission
Overall
Overall
Overall
%UAGC
0.13
0.54
2.2
% Loss
0.13
0.5-1.0
0.16
0.01
Negative*
Former USSR
1989
Transmission
Distribution
-2.0
>2.0
1 Pacific Gas and Electric (PG&E) is a gas distributor in the United States. k 1 ft3 = 28.3L
(0.0283 m3). e Unaccounted for gas. d An apparent gain was indicated.
Table 6 summarizes those studies that break down gas losses by industry sector. Although
the number of countries covered is limited, the estimates in these studies suggest that, except for the
former Soviet Union, gas actually leaked to the atmosphere consists of less than 1 percent of
throughput. The most detailed study is specific to the Pacific Gas and Electric Company in the
United States which suggests that actual leakage represents only 0.08 percent of gas received. These
estimates could vary considerably if a larger sample of countries is considered.
Table 7 is reproduced from a report by Piccot et al. (1990) and is derived from International
Energy Agency energy balances for selected countries. The average loss weighted by each country's
throughput is estimated to be 2.3 percent. .This table reports a percentage loss for the United
Kingdom of 3.6 percent which is considerably higher than the value reported in Table 6.
Unfortunately, the source of this discrepancy cannot be determined from available information. It
is possible that for some of countries included in Table 7 the estimates may suffer from the same
shortcoming as UAG statistics in that they may not provide sufficient detail to allow differentiation
of actual leakage from other accounting losses. In those cases, the estimates could be high.
Table 8 contains country-specific emission estimates from various other sources. Again, a
relatively high estimate of 5.3 to 10.8 percent of supply for the United Kingdom emerges, as does
another low estimate of 0.3 Tg/yr for West Germany.
Clearly, the wide range of sometimes conflicting estimates reflects the high degree of
uncertainty associated with current global estimates of CH4 emissions from natural gas extraction,
processing, transmission, and distribution systems. Despite the plethora of emissions estimates
-------
available for the gas industry, it is clear that they are based on very little hard data. It is likely that
the loss rates for natural gas system- will vary significantly among countries because the types of
systems, system operating characteristics, and system ages likely vary significantly from country to
country.
Table 8. Country-specific gas emission estimates.
Source
Mitchell et al. (1990)
Year
Estimates
Not United Kingdom
specified Low: 1.9% of supply
Medium: 5.3% of supply
High: 10.8% of supply
(preferred estimate is medium to high)
Selzer (1990)
Tilkicioglu (1990)
Dixon (1990)
Not
specified
1988
1987-
1988
West Germany: 0.3 Tg/year
USA: 3.1 Tg/year
Australia: 2% of production
Minor industrial sources '""..
Industrial sources of CIL. that have been given the greatest attention hi the literature include
coal mining operations and the production and distribution of natural gas. Logically, attention has
been focused here because both sources are responsible for producing significant quantities of CH4.
However, other industrial sources also release CH4 into the atmosphere. Individually these sources
emit minor quantities of CH4 but collectively their contribution to the global budget may be
significant. Very little research has been done to characterize the CH4 emissions from these minor
industrial sources, and they are rarely included in assessments of the global CH4 cycle. Although
current estimates of CH4 emissions for minor industrial sources require more study to reduce
uncertainties, estimates developed so far suggest that the combined emissions from all sources may
be as significant as the more traditional sources (i.e., coal mines, natural gas production and
distribution systems, and solid waste disposal landfills). These estimates are briefly discussed here.
An early attempt to estimate CH4 emissions from industrial and other sources located in urban
areas was conducted by Blake (1984). In Blake's analysis, a limited number of ambient air samples
were collected in several cities, and it was observed that these samples routinely exhibited elevated
levels of CH4. Based on these measurements, an emission flux rate for the world's cities was first
calculated (0.06/nrVday) and then multiplied by an estimate of the land surface area covered by the
world's cities. Based on these rather crude calculations, Blake estimated that non-automobile-related
-------
emissions from the world's urban areas are about 10 Tg/yr. Urban CH4 sources can include a variety
of industrial and other activities, including natural gas distribution system leaks.
At a recent international workshop on. methane and .nitrous oxide emission sources, new
emissions estimates for minor sources were reported by Piccot and Beck (1993), and by Berdowski
et al. (1993). Sources examined in both studies included industrial activities such as coke production
facilities, petroleum refineries, printing operations, gasoline storage and marketing facilities, fossil
fuel combustion, organic chemical manufacturing operations, and others. Non-industrial sources such
as the combustion of solid waste in the residential sector were also examined.
The approach used to estimate global emissions in both studies was based on the use of source-
specific emission factors (i.e., CH4 emissions per unit of source-specific activity). For example,
emissions from petroleum refineries were estimated by Piccot and Beck by multiplying an individual
country's refinery crude oil throughput by an emission factor which quantifies the amount of CH4
emitted from all refinery processes per tonne of crude oil throughput. In both studies, many of the
CH4 emissions factors were based on the emissions characteristics of industrial and other operations
in the United States. In some cases an attempt was made to represent specific factors influencing
emission rates in individual countries (e.g., Piccot and Beck estimated the emissions reductions due
to the use of 3-way catalytic emission controls on light duty motor vehicles in different countries).
Although differences in the types of industrial processes and pollution control equipment used among
countries may affect the magnitude of the emission factors used, most studies have not attempt to
rigorously characterize these differences. At this point, there can only be speculation about how these
differences may affect the emission estimates.
Estimates of the CH4 emissions from selected minor sources examined by Piccot and Beck and
Berdowski et al. are compared in Table 9. The base year for both sets of estimates is 1990. The
most significant source categories identified include residential on-site waste burning (estimated by
Piccot and Beck only), mobile sources, fossil fuel combustion, coke production and iron and steel
processes, and petroleum refining. Although emission estimates for these sources differ somewhat
between the two studies, it appears that their overall contribution to the global budget is close to 10
Tg/yr. Emissions from coke production facilities were also estimated by Darmstadter et al. (1984)
to be 4 Tg for 1980-higher than that estimated by both Piccot and Beck and Berdowski et al.
Table 9. Emissions estimated by Piccot and Beck (1993) and Berdowski et al. (1993) for minor
industrial and other sources of methane.
Source Description
Berdowski et al.
(Tg/year)
Piccot and Beck
(Tg/year)
Residential on-site waste
burning
not estimated
3.3
-------
Mobile sources
Fossil fuel combustion
Petroleum refining
Selected industrial sources
Organic chemical MFG.
Coke production and iron and
steel processes
Other miscellaneous sources (industrial
waste, forest fires and managed burning,
sewage treatment plants, kraft paper
manufacturing, printing and publishing,
petroleum storage and distribution)
3.0 (±2)
4.5 (±1.5)
0.5
0.2
<2.0
1.8
1.6
0.6
0.1
0.3
1.4
Global estimates of CH4 emissions from several minor industrial sources have also been
estimated by Lacroix (1993). Table 10 summarizes the emissions estimated by Lacroix for five
industrial sources of CH4. Total emissions for 1990 from the five sources are 13.3 Tg/yr. The
approach used by Lacroix is based on emission factors (JEA-EPA, 1990) and commercial fuel use
for 1987 in compilations of the International Energy Agency adjusted for a world average energy
growth rate of 2 percent per year to the present.
The estimates of Lacroix include several sources addressed by Piccot and Beck and Berdowski
et al. For most sources examined, agreement in the emissions estimates between the three studies
is variable. For example, the combined emissions from both petrochemical plants and refining
operations estimated by Lacroix range from 4.0 to 7.0 Tg/yr. Both Piccot and Beck, and Berdowski
et al. estimate that emissions from these two sources are less than 1 Tg/year.
Table 10. Emissions estimated by Lacroix (1993) for various industrial sources of methane.
Source Description
Emissions (Tg/yr)
Combustion of Fossil Fuels
Petroleum Refining
Petrochemicals
Peat Mining
Geothermal Electricity Production
Total Emissions
4.6 ± 0.4
3.5 ± 0.5
2.0 ± 1.0
2.0 ± 1.0
1.2 ± 0.7
13.3 ± 3.6
Agreement on the emissions from fossil fuel combustion related activities are only moderate among
all of the studies. Piccot and Beck estimate that emissions from all fossil fuel combustion related
activities (i.e., stationary and mobile sources) are 3.5 Tg/yr while Berdowski et al. estimates the
value to be much higher-7.5 Tg/yr. Lacroix estimates combustion related emissions to be 4.6
-------
Tg/year-closer to the value reported by Piccot and Beck. However, there is significant disagreement
in the distribution of combustion related emissions between stationary and mobile sources. Estimates
by Lacroix and Berdowski et al. can be used to place a range on mobile source emissions of 3.0 to
4.3 Tg/yr-much higher than estimates developed by Hitchcock and Wechsler (1972) (0.5 Tg/yr), and
by Piccot and Beck (1.8 Tg/yr).
Lacroix includes estimates for other industrial sources not addressed in most studies including
peat mining and geothermal electricity production. The total emissions from these two sources are
3.2 Tg/yr.
There is some evidence that mining oil shale and salt may release CH4 into the atmosphere.
Although we are not aware of attempts to estimate the global emissions from these mining operations,
measurements collected by the U.S. Bureau of Mines (USBOM) suggest that some salt and oil shale
deposits contain CH4 that can be released during mining. Data provided by the USBOM show that
normal production-grade salt adjacent to anomalous salt zones (i.e., zones with brine seeps, gas
seeps, and other factors that differ from normal salt) can contain CH4 in quantities of 0.1 m3 of CH4
per tonne of salt mined. Salt within the anomalous zones can contain between 0.4 and 1.8 m3 of CH4
per tonne mined. With oil shale, USBOM data show that oil shale samples at two mines in the
United States have CH4 contents of from 0.195 to 1.3 m3 per tonne muled. In general, these CH4
contents are much lower than the CH« contents typically encountered hi coal produced at underground
mines.
Radiocarbon emissions estimates for industrial sources.
In a recent, Cicerone and Oremland (1988) recognized that emissions from sources of
radiocarbon-free CH4 may be understated. Radiocarbon-free QK4 "sources include CH4 hydrate
deposits and many of the industrial sources included here (i.e., coal mines, natural gas processing
and transmission systems, and a tew of the minor industrial sources). Cicerone and Oremland cite
studies suggesting that sources of radiocarbon-free CH4 may release twice the emissions than earlier
studies by Koyama (1963), Ehhalt (1974), Sheppard et al. (1982), Seiler (1984), Crutzen (1987) and
others have estimated. It was suggested that collectively these sources contribute up to 50 Tg/yr
more than has previously been estimated (i.e., total emissions from fossil fuel sources, CH4 hydrates.
and others are 135 Tg/yr). ...... .
Several other studies also suggest that the CH* budgets for sources of radiocarbon-free CH4
may have been underestimated in the early studies cited above. In a recent study conducted by the
National Aeronautics and Space Administration, model calculations based on carbon-14 data,
atmospheric CH4 concentrations, and other information indicate that 123 Tg/yr of atmospheric CH4
was emitted from fossil carbon sources during 1987 (Whalen et al., 1989). This is about 50 percent
-------
higher than Cicerone and Oremland estimate for coal mines and natural gas drilling, venting, and
transmission systems. If the more recent coal mine estimates by Boyer et al. (1990) are considered,
and if the emissions from minor industrial sources are taken into account, the gap begins to close
between Whalen's estimate of 123 Tg/yr and the "bottom-up" estimates for sources of radiocarbon-
free CH4.
Summary
Over the past 30 years there have been several attempts to estimate the global emissions of CH,
from coal mining operations. The estimates presented in this section are representative of the range
of emission estimates published. Emissions estimates range from 7.9 to 47.2 Tg/yr for various years
between 1960 and 1989. The estimates vary because of differences in assumptions made for emission
factors (i.e., emissions/tonne of coal mined) and coal production rates. Although it is not possible
to identify the most "correct" estimate, it is likely that for the late 1980s emission estimates which
are less than 25 to 30 Tg/yr are unrealistically low. We believe a more realistic range for this period
would be 35 to 48 Tg/yr. Several recent studies have attempted to characterize country-specific
emissions and have included mine emission sources not previously addressed (i.e., coal handling
operations). Emission estimates from these studies, which are at the upper end of the range cited
above, may be the most representative available in spite of the significant uncertainties which still
remain. None of the estimates presented here include emissions from abandoned underground coal
mines, so the range of 35 to 48 Tg/yr may still be low.
Numerous estimates of emissions for the natural gas industry are available. Global emissions
estimates from as long as 20 years ago have been produced, primarily for the purpose of determining
emission inventories and global balances of atmospheric trace gases. Methane emissions estimates
produced during the 1980s typically range from 25 to 50 Tg/yr and assume leakage rates of from 1
to 4 percent. Clearly, the wide range of sometimes conflicting estimates reflects the high degree of
uncertainty associated with current global estimates of CH4 emissions from natural gas extraction,
processing, transmission, and distribution systems. Despite the plethora of emissions estimates
available for the gas industry, it is clear that they are based on very little hard data. More detailed
country-specific assessments will be needed before a "best estimate* or narrow range of emissions
can be established. It is likely that the loss rates for natural gas systems will vary significantly among
countries because the types of systems, system operating characteristics, and system ages likely vary
significantly from country to country.
Several minor industrial sources also release CH4 into the atmosphere. Individually these
sources emit small quantities of CH4 but collectively their contribution to the global budget may be
significant. Although current estimates of CH4 emissions for minor industrial sources are limited and
-------
highly uncertain, estimates suggest that the combined emissions from all minor industrial sources
identified so far may be between 10 and 15 Tg/yr, nearly as significant as some of the more
traditional CHt sources. It is likely that emissions from these sources account for some of the
difference observed between the total fossil fuel contribution estimated by carbon-14 analysis, and
the bottom-up estimates of major industrial sources. Based on the literature cited here the most
significant minor sources include residential on-site waste burning, peat mining and geothermal power
production, mobile sources, stationary source fossil fuel combustion, coke production and iron and
steel processes, and petroleum refining. Estimates of emissions from specific minor sources vary
widely between different studies indicating that further work is needed to develop more representative
estimates.
-------
References
Ameri, S., F.T. Al-Sandoon, C.W. Byrer. 1981. Coalbed methane resource estimate of the
Piceance Basin (Report No. DOE/METC/TPRJ82-6). U.S. Department of Energy, Morgantown,
West Virginia, 44 p.
American Gas Association. 1986. Lost and unaccounted for gas. In: Issue Brief 1986-28:
American Gas Association, Arlington, Virginia, 4 p.
Anderson, R.B., L.J.E. Hofer. 1965, Activation energy of diffusion of gases into porous solids.
Fuel, 44:303.
Andronova, G.A., I.L. Karol. 1993. The contribution of USSR sources to global methane emission.
Chemosphere, 26(l-4):\ 11-126.
Basic, A., M. Vukic. 1989. Dependence of methane contents in brown coal and lignite seams on
depth of occurrence and natural conditions. In: Proceedings of the 23rd International
Conference of Safety in Mines Research Institutes. U.S. Department of the Interior, Bureau of
Mines, Washington, D.C.
Bates, D.R., A.E. Witherspoon. 1952. The photo-chemistry of some minor constituents of the
Earth's atmosphere (COj, CO, CH4, N20). Roy. Astronom. Soc. Monthly Not., 112.
Berdowski, J.J.M., Olivier, J.G.J. and Veldt, C. 1993. Methane emissions from fuel combustion
and industrial processes. In: Proceedings of The International Vbrkshop on Methane and
Nitrous Oxide: Methods in National Inventories and Options For Control, February 1993,
Amersfoort, The Netherlands.
Blake, D.R. 1984. Increasing Concentrations of Atmospheric Methane, 1979-1983 (Ph.D.
Dissertation). University of California, Irvine, California.
Bolle, H.J., W. Seller, B. Bolin. 1986. Other greenhouse gases and aerosols. In: B. Bolin (ed.)
The Greenhouse Effect, Climatic Change, and Ecosystems. John Wiley and Sons, New York,
157-203.
Boyer, CM., J.R. Kelafant, V.A. Kuuskraa, K.C. Manger, D. Kruger. 1990. Methane emissions
from coal mining: issues and opportunities for reduction (EPA Report No. EPA-400/9-90/008).
U.S. Environmental Protection Agency, Office of Air and Radiation, Washington, D.C.
Chem Systems International, Ltd. 1989. Methane Losses from Natural Gas Utilization, prepared by
Chem Systems International, Ltd., for the National Energy Administration, Sweden, 42 p.
Cicerone, R.J., R. Oremland. 1988. Biogeochemical aspects of atmospheric methane. Global
Kogeochem. Cycles, 2:299-327.
Cottengnam, T.L., R.M. Cowgill, J.B. Godkin, J.R. Grinstead, D.J. Luttrell, F.A. Nelson, R.H.
Noistering, J.D. Peterson, R. Quintanilla, J.L. Robertson, E.R. Walden, R.L. Waller, R.E.
Wlasenko, R.M. Wong. 1989. Unaccounted-for Gas Project, Pacific Gas and Electric
Company. San Francisco, California.
Creedy, D.P. 1983. Seam gas-content data-base aids in firedamp prediction. The Mining Engineer,
August 1983:79-82.
Creedy, D.P. 1993. Methane emissions from coal-related sources in Britain: Development of a
methodology. Chemosphere, 26(1-4):419-44Q.
Crutzen, P.J. 1987. Role of the tropics hi atmospheric chemistry. In: R. Dickinson (ed.)
Geophysiology of Amazonia. John Wiley and Sons, New York.
Dannstadter, I, L. Ayres, R.U. Ayres, W.C. Clark, P. Crosson, P. Crutzen, T.E. Graedel, R.
McGill, R.F. Richards, J.A. Tarr. 1984. Impacts of world development on selected
characteristics of the atmosphere: an integrative approach, volume 2-appendices (Report
No.ORNL/Sub/86-22033/l/V2). Oak Ridge National Laboratory, Oak Ridge, Tennessee.
Department of Energy. 1991. National energy strategy (Report No. DOE/S-0082P). U.S.
Department of Energy, Washington, D.C., 217 p.
Dixon, B.F. 1990. Methane losses from theAustralian natural gas industry. In: International
Vtbrkshop on Methane Emissions from Natural Gas Systems, Coal Mining, and Vbste
Management Systems. Environment Agency of Japan, U.S. Agency for International
-------
Development, and U.S. Environmental Protection Agency, Washington, D.C., 709 p.
Ehhalt, D.H. 1974. The atmospheric cycle of methane. Tellus, 26:58-70.
Ehhalt, D.H., U. Schmidt. 1978. Sources and sinks of atmospheric methane. Pageoph, 116:452-
463.
Environment Agency of Japan. 1990. International workshop on methane emissions from natural
gas systems, coal mining, and waste management systems. Environment Agency of Japan,
U.S. Agency for International Development, and U.S. Environmental Protection Agency,
Washington, D.C., 709 p.
Fung, I., J. John, J. Lerner, E. Matthews, M. Prather, L.P. Steele, PJ. Fraser. 1991. Three-
dimensional model synthesis of the global methane cycle. /. Geophys. Res., 96:13,033-13,065.
Hay, N.E., P.L. Wilkinson, W.M. James. 1988. Global climate change and emerging energy
technologies for electric utilities: the role of natural gas. American Gas Association,
Arlington, Virginia, 30 p.
Hitchcock, D.R., A.E. Wechsler. 1972. Biological cycling of atmospheric trace gases (Report No.
NASW-2128). National Aeronautic and Space Administration, Washington, DC, 415 p.
Irani, M.C., E.D. Thimons, T.G. Bobick. 1972. Methane emission from U.S. coal mines, a survey
(Report No. 1C 8558). U.S. Department of the Interior, Bureau of Mines, Pittsburgh,
Pennsylvania.
JEA-EPA. 1990. Methane emissions and opportunities for control: workshop results of
intergovernmental panel on climate change, response strategies working group. Japan
Environment Agency/United States Environmental Protection Agency, Washington, D.C.
Jolly, DC, L.H. Morris, F.B. Hinsely. 1968. An investigation into the relationship between the
methane sorption capacity of coal and gas pressure. The Mining Engineer 127:539. *
Joubert, J.I., C.T. Grein, B. Bienstock. 1974. Effect of moisture on the methane capacity of
American coals, fuel, 53:186.
Keeling, CD. 1973. Industrial production of carbon dioxide from fossil fuels and limestone. Tellus,
25:174-198.
Khalil, M.A.K., MJ. Shearer, Eds. 1993. Chemosphere, 26(1-4).
Kim, A.G. 1977. Estimating methane content of bituminous coalbeds from adsorption data (Report
No. RI8245). U.S. Department of the Interior, Bureau of Mines, Pittsburgh, Pennsylvania.
Kirchgessner, D.A-, S:D. Piccot, ID. Winkler. 1993a. Estimate of global methane emissions from
coal mines. Chemosphere, 26(1-4)A53-412.
Kirchgessner, D.A., S.D. Piccot, A. Chadha, T. Minnich. 1993b. Estimation of methane emissions
from a surface coal mine using open-path FTIR spectroscopy and modeling techniques.
Chemosphere, 26(l-4):23-44.
Kissel, F.N., C.M. McCulIoch, C.H. Elder. 1973. The direct method of determining methane
content of coalbeds for ventilation design (Report No. RI 7767). U.S. Department of the
Interior, Bureau of Mines, Pittsburgh, Pennsylvania.
Koyama, T. 1963. Gaseous metabolism in lake sediments and paddy soils and the production of
atmospheric methane and hydrogen. /. Geophys. Res., 58:3971-3973.
Koyama, T. 1964. Biogeochemical studies on lake sediments and paddy soils and the production
of hydrogen and methane. Recent Researches in the Fields of Hydrosphere, Atmosphere, and
Geochemistry, 143-177. "..-..
Lacroix, A.V. 1993. Unaccounted-for sources of fossil fuel and isotopically-enriched methane and
their contribution to the emissions inventory: A review and synthesis. Chemosphere 26 (1-
4*507-558.
Lambert, S.W., M.A. Trevits, P.F. Steidl. 1980. Vertical borehole design and completion practices
to remove methane gas from minable coalbeds (Report No. DOE/CAfFC/TR-80/2). U.S.
Department of Energy, Washington, DC, 163 p.
Mitchell, C, J. Sweet, T. Jackson. 1990. A study of leakage from the UK natural gas distribution
system. Energy Policy, November 1990:809-818.
Murray, D.D. 1980. Methane from coalbeds-a significant undeveloped source of natural gas.
-------
\
\
Colorado School of Mines Research Institute, Golden, Colorado, 37 p.
i, P.A., T. Kram. 1989. CH4/COj-emission from fossil fuels global wanning ±
Proceedings of the International Energy Agency Wrkshop on Greenhouse Gases, Ji.
Paris.
S.D., A. Chadha, D.A. Kirchgessner, R. Kagenn, M.J. Czerniawski, T. Minnich. h
Measurement of methane emissions in the plume of a large surface coal mine using open-pau
FT2R spectroscopy. In: Proceedings of 84th Annual Meeting of the Mr and Vbste
Management Association, June 1991, Vancouver, British Columbia, Canada.
Hecot, S.D., M. Saeger. 1990. National- and state-level emissions estimates of radiatively
important trace gases (RTTG) from anthropogenic sources (Report No. EPA-600/8-90-073;
NTIS PB91-103572). U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina.
Mccot, S.D., A. Chadha, J. DeWaters, T. Lynch, P. Marsosudiro, W. lax, S. Walata, J. D.
Winkler. 1990. Evaluation of significant anthropogenic sources of radiatively important trace
gases (Report No. EPA-600/8-90-079; NTIS PB91-127753). U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina.
Piecot, S.D., and L. Beck. 1993. Estimation of Methane Emissions From Minor Anthropogenic
Sources. In: Proceedings of The International Vbrkshop on Methane and Nitrous Oxide:
Methods in National Inventories and Options For Control, February 1993, Amersfoort, The
Netherlands.
Richer, R.C., C. Bibler, R. Glickert, L. Machesky, J. Williams. 1991. Assessment of the potential
for economic development and utilization of coalbed methane in Poland (Report No. EPA-
400/1-91-032). U.S. Environmental Protection Agency, Office of Air and Radiation,
Washington, D.C.
Schwarzer, R.R., C.W. Byrer. 1983. \fcriation in the quantity of methane adsorbed by selected
coals as a function of coal petrology and coal chemistry (Report No. DE-AC21-80MC14219).
U.S. Department of Energy, Morgantown, West Virginia.
W. 1984. Contribution of biological processes to the global budget of CH< in the
atmosphere. In: M.J. Klug and C.A. Reddy (eds.) Current Perspectives in Mcrobial Ecology.
American Society for Microbiology, Washington, D.C.
H. 1990. Anthropogen methane emission. In: International Vbrkshop on Methane
Emissions from Natural Gas Systems, Coal Mining, and Wtste Management Systems,
Environment Agency of Japan, U.S. Agency for International Development, and U.S.
Environmental Protection Agency, Washington, D.C., 709 p.
, J.C., H. Westberg, J.F. Hopper, K. Ganesan. 1982. Inventory of global methane sources
and their production rates. J. Geophys. Res., 57:1305-1312.
Tllklcioglu, B.H. 1990. Annual methane emission estimate of the natural gas systems in the United
States. In: International Wbrkshop on Methane Emissions from Natural Gas Systems, Coal
Mining, and Vbste Management Systems; Environment Agency of Japan, U.S. Agency for
International Development, and U.S. Environmental Protection Agency, Washington, DC.,
709 p.
, M., N. Tanaka, R. Henry, B. Deck, J. Zeglen, J.S. M>gel, J. Southon, A. Shemesh, R.
Fairbanks, W. Broecker. 1989. Carbon-14 in methane sources and in atmospheric methane:
the contribution from fossil carbon. Science, 245:286-290.
Committee on Energy. 1991. Quantification of methane emissions from British coal mine
sources (Draft Report).
-------
o
O
To be included in NATO Book
The Global Methane Cyde:
Its Sources, Sinks, Distributions and Role in Global Change
1993
GLOBAL METHANE EMISSIONS FROM WASTE MANAGEMENT
Siisan A. 'Hiunicloe1, Morton A, Ilarlaz2, Rebecca !V-r3, L.C. IluflJ, Lit- Davis^. Joe Manghio3
United States Hnvironmental Protection Agency, Office of Research and DeveloptnenI
Air and Energy Engineering Kese:arcli Laboratory, Kesearch Triangle Park, North Carolina, U.S.A.
2Norlh Carolina Slate University, Civil Engineering Department
Kaleigh, North Carolina, U.S.A.
3Radian Corporation, Research Triangle Park, North Carolina, U.S.A.
Introduction
Landfills, wastewaier treatment lagoons, and livestock waste management are all wasie
management operations representing major sources of methane (CII^) Initial estimates of 014
emissions from these sources suggest approximately 72 teragrams per year (Tg/yr) globally with a
95% confidence Interval of 54 to 95 Tg/yr. 'lliis represents approximately H% of total global CII^
emissions of 500 Tg/yr (IPCC, 1992). Iliis chapter begins with a brief overview of how CH4 is
generated from the anaerobic decomposition of waste and then discusses in detail landfills,
wastewater treatment lagoons, and livestock waste management. Current techniques for
estimating CH* emissions from waste are summarized, and sources of uncertainty and areas where
f UN her study is needed are Identified.
lite potential control of CILj emissions from waste management has been targeted by the
United Slates (U.S.) and oilier countries as part of greenhouse gas reduction programs designed to
meet the goals of treaties signed at the United Nations Conference on Environment and
Development (UNCCD) held in 1992. Consequently, reducing the uncertainty associated with CH<
emission estimates is a high priority.
Methane Production During the Anaerobic Decomposition of Waste
'Hie anaerobic decomposition of organic matter is a complex process which requires ihat
several groups of microorganisms act synergisticly under favorable environmental conditions (see
Boone, this volume). 'IIic pathway described below has been demonstrated to apply to anaerobic
decomposition in sludge digesters and in livestock waste management systems, 'llils anaerobic:
pathway Is also expected to occur in landfills and anaerobic wastewater lagoons. (See also Harlaz.
et al., 1989a.)
-------
Three trophic groups of anaerobic bacteria must be present to produce CH^ from biological
polymers such as, cellulose, hemicellulose, and protein: (1) hydrolytic and fermentative
microorganisms, (2) obligate proton-reducing acetogens, and (3) memanogens (Wolfe, 1979;
Zchnder et al., 1982). The hydrolytic and fermentative group is responsible for the hydrolysis of
biological polymers. The initial products of polymer hydrolysis are soluble sugars, aniino acids,
long-chain carboxylic acids, and glycerol. Following polymer hydrolysis, the hydrolyiic and
fermentative microorganisms ferment the initial products of decomposition into short-chain
carboxylic acids, alcohols, carbon dioxide (CC>2), and hydrogen. Acetate, a direct precursor of
CH
-------
solid waste (MSW) production, as compared 10 developing countries. However, this scenario will
not be realized if per capita income decreases in developing countries. Recently, declining
economic conditions have resulted in reduced MSW generation in Caracas, Venezuela, Mexico
City, Mexico, and Buenos Aires, Argentina (Bartone et al., 1991 >
Methods of managing MSW vary widely, ranging from open dumps and open burning 10
sanitary landfills wiih leachate collection systems and landfill gas control. The majority of the
world's MSW is managed using either sanitary landfills or open dumps. In the U.S., recent
estimates indicate that 72% of MSW is buried in landfills (Kaldjian, 1990). Anaerobic decomposition
prevails in landfills. both anaerobic and aerobic processes occur ai open dumps. The CH4 poienlial
of other waste management processes such as incineration, recycling, and composting is
considered insignificant in comparison to landfills and open dumps.
Landfilled waste conlains numerous constituents that have the potential to biodegrade under
anaerobic conditions. However, optimal conditions for anaerobic decomposition within a landfill
may not exist and may thus result in overestimated emissions. Many emission methodologies
assume that optimal conditions exist. In a recent study (Peer et al., 1993) field data were gathered
to develop an empirical model that is intended to reflect actual emissions to the atmosphere. "Iliis
model adjusts for gas recovery efficiency and CH* oxidation. Estimates are presented later in this
chapter, along with updated estimates using the approach developed by Bingemer and Crutzen
(1987).
Factors Affecting CHt Potential of Buried Waste. The traditional method of classifying MSW
according tosortable categories (e.g., paper, plastic, food waste, yard waste, glass, metals, rubber,
wood, textiles, dirt, and miscellaneous (Kaldjian, 1990]) is appropriate for recycling studies and
overall solid waste management planning. However, data specific to the chemical composition of
refuse are more applicable to analyses of refuse decomposition. Studies of refuse in Madison,
Wisconsin, found cellulose plus hemicellulose to be about 60% of landfill waste and to account for
91% of the CH4 potential of refuse (Barlaz, 1985, 1988, 1989b).
The components of MSW thai contain significant biodegradable fractions are food waste, yard
waste, and paper, which have a combined cellulose and hemicellulose content of 50 to 100%.
Lignin is the other major organic component of refuse; 'however, Itgnin does not undergo
significant decomposition under anaerobic conditions (Young and Frazer, 1987).
Methane formation does not occur immediately after refuse is placed in a landfill or dump.
It can take months or years for the proper environmental conditions and the required
microbiological populations to become established. .Numerous factors control decomposition,
including moisture content, nutrient concentrations, presence and distribution of microorganisms,
particle size, water flux, pH level, and temperature. For a review of the effect of each of these
-------
factors on CH+ production see Barlaz et al., 1990; Pohland and flarper, 1987; Halvadakis et al.,
1983.
The two factors that appear to have the most impact on CH4 production are moisture conteni
and pH. The effect of refuse moisture content lias been summarized by Halvadakis ei al. (1983),
although some of the data in the summary relate lo manure and noi to municipal waste. The
broadest data sets are those of Emberton (1986) and Jenkins and Fettus (1985). Emberton
measured CH^ production rates in excavated landfill samples under laboratory conditions. Jenkins
and Pettus sampled refuse from landfills and tested how CH.j production was affected by the
moisture content of refuse. In both studies, the CH4 production rate exhibited an upward trend
with increasing moisture content, despite differences in refuse density, age, and composition. H is
difficult to translate the results of these laboratory studies to actual landfills. An attempt by the U.S.
EPA's Air and Energy Engineering Research Laboratory (AEERL) 10 identify u statistically significant
correlation between landfill gas recovery and precipitation (which affects refuse moisture content)
found no such correlation (Peer et al., 1992).
A second key factor influencing the rate and onset of CH^ production is pH. The optimum
pH level for activity by methanogenic bacteria is between 6.8 and 7.4, CH+ production rates
decrease sharply with pH values below about 6.5 (Zehnder et al., 1982). When refuse is buried in
landfills, there is often a rapid accumulation of carboxylic acids; this results in a pH decrease and a
long lime lapse between refuse burial and the onset of CH4 production.
Neutralizing leachate and recycling it back through refuse lias been shown to enhance the
onset and rate of CH4 production in laboratory studies (Pohland, 1975; Btiivjd et al., 1981; Barlaz et
al., 1987, International Energy Agency, 1992). Given that moisture and pH have been reported as
the two most significant factors limiting CH^ production, the stimulatory effect of leachate
neutralization and recycling is logical. Neutralization of leachate provides a means of externally
raising the pH of the refuse ecosystem. Recycling neutralized leachate back through a landfill
increases and stabilizes refuse moisture conteni and substrate availability and provides mixing in
what would otherwise be an immobilized batch reactor.
Notably, field experience with leachate recycling systems is limited and more infonnation is
needed to fully document their value. In addition, the lapsed time preceding the onset of CH*
production in landfills is an important aspect when considering the management of individual
landfills for biogas recovery or emissions mitigation. The age at which landfills and uncontrolled
dumps begin to produce CH4 is of lesser importance when evaluating global CH4 emissions from
MSW management systems. In this case, the total CH4 production potential is more critical.
-------
Determination of the CH4 Potential of Landfills and Dumps. Knowledge of the chemical
composition of refuse buried in a landfill makes it possible to estimate the volume of CHj that may
be produced. The mass of CH* that would be produced if all of a given constituent were convened
'to CH4, COi, and ammonia may be calculated from Equation 1 (Parkin and Owen, 1986):
CnHaObNc + [n - l/4a l/2b +3/4c]H2O -»
il/2n - l/8a + l/4b + 3/8c]CO2 + [l/2n + l/8a - l/4b - 3/8clCH4 + cNH3
(1)
using this stoichiomeiry, the CH* potential of cellulose (C6Hi0O5) and heinicellulose (C5H8O4) is
415 and 424 liters G9 CH4 at standard temperature and pressure (0°C, 1 atmosphere) per dry
kilogram (kg), respectively (18.5 and 18.9 grams Igl CH4/dry kg).
These methane potentials represent maximum CH4 production if 100% of the cellulose and
heinicellulose were converted to CM*. However, decomposition of these constituents in landfills is
well below 100% for several reasons but mainly because (1) some cellulose and heinicellulose is
surrounded by lignin or other recalcitrant materials (such as plastic) and, therefore, is not
biologically available; and (2) without active intervention, buried refuse is not evenly closed to
moisture, microorganisms, and nutrients. Barlaz etal. (1989b) applied mass balances to shredded
refuse incubated in laboratory-scale lysimeters with leachate recycle. Carbon recoveries of 87 to
111% were obtained, where a perfect mass balance would give a carbon recovery of 100%.
Mineralization of 71% of the cellulose and 77% of the hemicellulose was measured in a container
"sampled after 111 days. Mass balances were useful for documenting the decomposition of specific
chemical constituents and demonstrating the relationship between cellulose and hemicellulose
decomposition and CH« production.
Mass balances may be used to estimate the CH« potential remaining in a landfill by sampling
the refuse, performing the appropriate chemical analyses, and calculating the CH* potential.
Ideally, the initial chemical composition and CH4 potential of the refuse would be known, in which
case comparing that initial CH< potential with the potential at the time of sampling would provide
information on the fraction of the refuse that has been degraded. Indisputably, representative
sampling of a full-scale sanitary landfill is not realistic. However, it is possible to obtain multiple
samples at presumably representative locations within a landfill to get an estimate of the range and
extent of decomposition. Samples should be as large as can reasonably be handled and reduced by
proven techniques.
Another technique for assessing the CH* potential of refuse is the biochemical CH* potential
(BMP) test (Shelton and Tiedje, 1984; Bogner, 1990). In the BMP test, the anaerobic
biodegradabiliry of a small sample of refuse (5 to 10 g) is measured in a small batch reactor (100 to
-------
200 in
-------
1990). However, there are no data on the quantitative significance of CH^ oxidation above
landfills. Cfy escaping through cracks in a landfill cover likely will not reside in the cover for a
period sufficient to undergo significant biodegradation. .
Emissions Estimate Methodology for Landfills and Open Dumps. Two techniques for
estimating emissions from landfills are reviewed here: the Organization of Economic Cooperation
and Development (OECD) and CPA/AEEKL methods. These methods were developed to estimate
CH4 emissions on a global basis. Models that estimate CH^ production from individual landfills are
reviewed by Augenstein and Pacey (1990) and Peer et al., (1993) While global estimates focus
on ultimate CH4 release, models of individual landfills emphasize the rate and duration of CF-L$
production as these factors affect the economics of landfill gas recovery projects.
Organization of Economic Cooperation and Development Methodology. OECD (1991)
used the mass balance approach developed by Bingemer and Crutzen (1987), where an
instantaneous release of CH4 is assumed to enter the atmosphere during the same year that refuse is
placed in a landfill. This method also assumes that (1) all of the CH« that is produced escapes to the
atmosphere (i.e., none is oxidized en route to the atmosphere) and (2) all developing nations
generate and dispose of MSW at die same per capita rate.
To calculate the annual emission from MSW, OECD used the following equation from
Bingemer and Crutzen:
CH4 Emission = Total MSW Generaied (kg/yr) x MSW Landfilled (%) x DOC in
MSW (%) x Fraction Dissimilated DOC (%) x 0.5 g CH4/g Biogas
x Conversion Factor (16 g CH4/12 g C) - Recovered CM, (fcg/yr)
where: DOC is degradable organic carbon; Fraction Dissimulated DOC is (he portion of carbon in
substrates that is converted to landfill gas; and Recovered CH4 is the amount of CH4 that is
recovered through gas recovery systems and never emitted to die atmosphere.
The uncertainties of this approach are attributed to assumptions regarding anaerobic decomposition.
Many factors inhibit this process, and this approach tends to overstate potential emissions. For
example, this methodology does not adjust for CH< oxidation, which is known to occur.
BPA/AEERL's Regression Model Methodology. The EPA/AEERL methodology uses an
empirical model derived using landfill gas recovery data. The quantity of CH4 estimated by this
model is much less than that predicted by sioichiometric analyses or by laboratory studies (Peer el
-------
al., 1992; Barlaz et al., 1990; Barlaz et a!., 1989b; EMCON, 1982). The data gathered from U.S.
landfills that were used to develop this model represent a broad range of climate zones and waste
composition (Campbell et al., 1991; Peer et al., 1992) as described below. This model is intended
to reflect the amount of gas that is ultimately released to the atmosphere by adjusting for gas
recovery efficiency and CH4 oxidation. For the estimates presented in this chapter, it was assumed
that the recovery efficiency is 80% and that 10% of non-recovered CH4 is oxidized. Refinements of
this methodology include adjustments for recovery efficiency and CM., oxidation based on factors
derived through an uncertainty analysis. Comparison of the refined estimates with the earlier
estimates indicates a slight increase.
Data from 21 landfills were used to determine if there is a correlation between CH^
recovery and landfill characteristics such as waste quantity, age, depth, and climate. The SAS
(Slalislical Analysis System) regression procedure was used to generate regression statistics for
various models. Selection of variables for the regression models was based on the results of die
correlation and scatter plots of selected variables (e.g., climate, landfill depth, refuse mass, refuse
age). The main conclusion of the study was that the annual CH^ recovery rate was linearly
correlated with the mass of refuse in the landfill, and with landfill depth (Peer et al., 1992). The
data, regression line, and the 95% confidence limits of the regression coefficient are shown in
Figure 1. The regression was significant (P < 0.01), but much of the variability in the data is
unexplained (adjusted RZ = 0.50). The intercept was not significant, so the final model was forced
through the zero. Peer et al. (1992) provide information on the characteristics of each site that may
contribute to data variability. This report also provides an overview of development of the
empirical model. "
No statistically significant relationships were identified between annual CH4 recovery and
climate variables such as precipitation, temperature, and dew point. The effect of refuse age on gas
production was also analyzed. Gas recovery correlated most strongly with refuse between 10 and
20 years old. Although these results were not conclusive, they suggest that the generation time for
gas production is 20 to 30 years (Peer et al., 1992). This generation time 5s within the range of
generation times assumed in many landfill gas recovery models (EMCON, 1982; Augenstein and
Pacey, 1990).
One advantage of using the EPA/AEERL model is that refuse mass is the only variable
required to estimate CH4 emissions. Furthermore, it will be relatively easy to update the
relationship between CH+ recovery and refuse mass as more data become available. The
confidence limits of the regression coefficient can be used to bound emission estimates:
-------
'100
I
§ 80 -
"- 60 -
£
§?
o
o
01
or
c 40
at
§> 20
I
10
IS
Amount of Refuse (millions of lonnes)
Figure 1. Methane Recovery Data as a Function of
Buried Refuse (Peer et al., 1992)
-------
TABLE 1. WASTE TOTALS (Tg) BY GEOGRAPHIC
REGION USED TO DEVELOP EMISSION ESTIMATES
Geographic Region
Africa
Asia and the Middle East
Europe
North America
Oceania
South and Central America
Waste
80
363
224
301
14
64
Waste
Landfilled
38
175
165
216
12
47
References*
1-14
15-28
29-46
47-49
50-51
52-58
'Reference Kev
1.
2.
3-
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19-
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
El-Halwagi et al., 1988. 30.
El-Halwagi et a!., 1986. 31.
Kaltwasser, 1986. 32.
UNDP et al., 1987. 33.
Holmes, 1984. 34.
Monney, 1986. 35.
Cointreau, 1984. 36.
Cointreau, 1987. 37
The World Bank, 1985. 38.
Mwiraria et al., 1991. 39.
United Republic of Tanzania, 1989. 40
Verrier, 1990. 41.
World Resources Institute, 1990. 42.
Rettenberger and Weiner, 1986. 43]
Bhide and Sundaresan, 1990. 44.
United Nations, 1989- - 45.
Manialis et al., 1987. 46.
'Lohani and Thanh, 1980. 47.
Ahmed, 1986. 43.
Pairoj-Boriboon, 1986. 49.
Gadi, 1986. 50.
Mei-Chan, 1986. 51
Kaldjian, 1990. 52.
Diaz and Goulueke, 1987. 53.
Xianwen and Yanhua, 1991. 54.
Cossu, 1990a. 55.
Hayakawa, 1990. 56.
Swartz, 1989. 57.
Lechner, 1990. 58.
World Resources Institute, 1990.
Carra and Cossu, 1990.
Ettala, 1990.
Stegmann, 1990.
Ernst, 1990.
Cossu and Urbini, 1990.
Beker, 1990.
Carra and Cossu, 1990.
Gandolla, 1990.
Cossu, 1990b.
Swartz, 1989-
Richards, 1989-
Kaldjian, 1990.
Scheepera, 1990.
Bartone and Haley, 1990.
Bartone, 1990.
Bingemer and Cruizen, 1987.
U.S. EPA, 1988.
Kaldjian, 1990.
El Rayes and Edwards, 1991.
Bateman, 1988.
Richards, 1989.
Kessler, 1990.
Kaldjian, 1990.
World Resources Institute, 1990.
Diaz and Golueke, 1987.
Bartone et al., 1991.
Yepes and Campbell, 1990.
Bartone, 1990b.
10
-------
The upper and lower 95% confidence limits are 6.5 and 2.5 m3 CH^/min/lO6
Mg wet refuse (4.6 to 1.8 g CH^/min/kg of wei refuse), respectively.
Assuming an average generation time of 25 years gives an average CH4
recovery of 59.4 m3 CHy'Mg wet refuse (42 g CH4/kg of wet refuse).
A range of 33 10 86 m3 CII^/IO*5 Mg wet refuse (24 10 6l g CH^/kg wet refuse)
results. These values were derived using data collected at U.S. MSW landfills.
The use of this factor assumes that other countries have a waste composition
similar 10 U.S. landfills. However, U.S. MSW generally has a higher organic
content than most other countries (Bingemer and Crutzen, 1987). Therefore,
the use of ihis factor may overestimate landfill emissions for other countries.
Future refinements of the model will adjust for waste composition using gas
potential daia for different biodegradable wasie streams.
Assumptions and Data Used lo Estimate Waste Generation Rates. To estimate global
CM., emissions it was necessary to make several assumptions regarding waste generation and
disposal. Emission factors in all industrialized countries were assumed to be equal to those in the
U.S. The CH-j emission factors for die less-developed countries (LDCs), where adequate data were
not available, were assumed to be 25 to 75% of the average U.S. estimate. Several factors were
taken into account to develop this range. The composition of waste is different in third world
countries. For example, much of the garbage is scavenged before it is placed in the landfill,
especially paper, textiles, and metal products. More putrescibles end up in dumps and probably do
not generate as much CH4 because more oxidation (aerobic process) takes place. In addition,
garbage is often burned, which decreases the amount of material available for anaerobic
decomposition, but may increase CH4 emissions from inefficient combustion. FinallylHmost landfills
in the LDCs are open dumps which are scavenged by humans and wild and domestic animals.
While anaerobic conditions may form in some of these dumps,, the potential for. aerobic
decomposition is much greater than in sanitary landfills. Based on these assumptions, the 25 to 75%
range was chosen as die default value for LDCs, since no data on actual CH^ emissions are
available.
Estimates of waste generation and burial are presented in Table 1. This table was developed
using the data presented in the references shown. A recent review of global waste management
trends (Davis et al., 1992) found that information on MSW generation in developing countries is
difficult to obtain; in many cases, it is anecdotal. As shown in the reference list for Table 1, much
country-specific data were available to determine MSW generation and land disposal values,
11
-------
especially for developed countries. Where no data were available, daia from similar countries
were used.
Most of the available data for developing countries are provided on a per capita basis for
only the larger cities; this information was combined with population (United Nations, 1990) and
percent of the country lhat is considered urban (Population Reference Bureau, Inc., 1989), to
determine the amount of waste generated in urban ceniers of these countries. An estimate for rural
per capita refuse generation rate (Kessler, 1990) was then combined with the rural population
value to determine rural waste generation. Once the amount of waste generated for the entire
country was estimated, a default value of 50% disposed of on land (whether in landfills or open
dumps) was used to estimate the amount of waste lhat may degrade anaerobically. The 50%
default was chosen to represent waste that is actually collected in some manner and disposed of in
landfills or large enough open dumps for anaerobic conditions to occur. The remaining 50% of
waste generated is assumed to be (1) incinerated or combusted, (2) dumped in rivers or other
bodies of water, or 0) scattered or buried in small piles lhat degrade anaerobically. These disposal
methods do not produce large amounts of CH4.
The amount of MSW landfilled in the U.S. is approximately 189 million tonnes (U.S. EPA,
1988). Paper was die largest single component of die DOC fraction in both the U.S. and Canada.
Per capita MSW generation was in die range of 1.7 to 1.8 kg/person/day for both the U.S. and
Canada (Kaldjian, 1990; El Rayes and Edwards, 1991), and the average DOC content of paper or
MSW is 20%. The average MSW generation rate in other OECD countries is 1.1 kg/person/day.
MSW in diese countries has a DQC content of approximately 15-3%. The value used for the U.S. is
for MSW only; an additional 15 Tg/yr of biodegradable industrial solid waste is also landfilled (U.S.
EPA, 1987), which is unaccounted for in die present estimates of landfill CH4. In most cases,
country-specific information does not state specifically whedier industrial waste is codisposed of
with MSW.
Information on the amount of MSW generated and landfilled in the European countries lhat
are not OECD members and in die former Soviet Union is limited. Nozhevnikova et al. (1993)
used 0.8 kg/person/day for MSW generation in die former USSR. Average MSW generation for
Greece, the former Soviet Union, and Eastern Europe is approximately 0.6 kg/person/day
(Frantzis, 1988; Papachristou, 1988; Peterson and Perlmutter, 1989; Bingemer and Crutzen, 1987),
and the available data indicate that putrescibles make up a large portion of the MSW (estimates
-<'' ---,
range from 32 to 60%). This MSW contains approximately 15% DOC (Franizis, 1988; Papachristou,
1988; Peterson and Perlmutter, 1989; Bingemer and Crutzen, 1987; Zsuusa, 1990).
For most Asian countries, estimates of MSW generation were identified for one or two major
cities, but'not for the entire country. National per capita MSW generation estimates were identified
for Indonesia, Sri Lanka, the Philippines, Singapore, Taiwan, and Pakistan. These estimates range
12
-------
from 0.4 kg/person/day for the Philippines to 1.0 kg/person/day for Singapore. The average per
capita MSW generaiion lor these countries is estimated to be 0.6 kg/person/day (Davis el al.,
1992). . .
Few, data are available on MSW production and management in Central America, South
America, the Caribbean Islands, and Mexico. Most of the available information is only for the
larger cities. The average per capita MSW generaiion rate in seven South and Central American
countries (Brazil, Colombia, Chile, Paraguay, Peru, Venezuela, Costa Kica), and Mexico is estimated
to be 0.8 kg/person/day. The components are mainly vegetable and putrescible waste paper and
cardboard. The average DOC for the seven South and Central American countries and Mexico is
17% (Davis el al., 1992).
Information on MSW generation and disposal for African and Middle Eastern countries is also
very limited. In Africa, it appears ihat toxic and hazardous industrial and commercial wastes are
purposely or inadvertently disposed of with the MSW stream. Some information pertaining to
generation rales for African countries was located; but information for only two Middle Eastern
countries, Israel and Yemen, was obtained. Based on the very limited information for these two
continents, it is estimated that per capita generaiion rates range from 0.3 10 1.1 kg/person/day, and
the DOC content ranges from 3 to 20%.
Global and Country-Specific Estimates of CHf Emissions from Landfills and Dumps-.
ORCD Proposed Methodology. The MSW generation and landfill disposal data in Table 1
were used to calculate CH^ emissions for each country using die OECD methodology discussed
earlier.
This methodology is identical to that of Bingemer and Crutzen (1987), but makes use of
more recent waste generaiion data and the results of an exhaustive study by the EPA/AEERL to
gather country-specific waste generation and disposal data. As shown in Table 2, the global
estimaie of landfill CH^ emissions using this methodology is 57 Tg/yr. The estimate for die U.S.
(i.e., 21 Tg/yr) has been adjusted for the amount of CH4 that is recovered for energy utilization (1.2
Tg/yr, Thorneloe, 1992).
EPA's Regression Model Methodology. Based on the EPA's regression model methodology,
global landfill CH* emissions are estimated to range from 11 to 32 Tg/yr, with a midpoint of
21 Tg/yr. These are the initial estimates using the empirical model and database. Table 2 presents
country-specific global estimates using the methodology described in Peer et al. (1992) and
provides the lower, most probable, and upper-bound estimates of CH< emissions by continent.
This estimate for the US. (i.e., 4 to 12 Tg/yr) has also been adjusted for the amount of CH-j that is
recovered for utilization based on a recently completed survey (i.e., 1.2 Tg/yr, Thorneloe, 1992).
13
-------
1 TABLE 2. METHANE EMISSION ESTIMATES (Tg) FROM BURIED REFUSE USING THE OECD AND
REGRESSION MODEL METHODS
.-v.1' % k l " ' \ '? " '^ ' /'" rtv .-,0'-- '. >'-1 ' r '-' '; ' ' ' >!'"' * '* *' 't ' - ' ' " '*" ' '-'v " *
Country
OECD Method
v > ,\ . t " , ' . ' "" "-!" '' "
Africa
Congo
Egypt
Gambia
Ghana
Kenya
Liberia
Morocco
Nigeria
South Africa (Customs Union)
Sudan
Tanzania, United Republic of
Uganda
Zimbabwe
Other Africa
TOTAL - AFRICA
Asia
Bangladesh
China (Mainland, NMP)
India
Iran, Islamic Republic of
Iraq . ; -
Israel
Japan
Korea, People's Democratic Rep.
Korea, Republic of
Kuwait
"Malaysia
Mongolia
Myanmar
Pakistan
Philippines
Saudia Arabia
Sri Lanka
Thailand
Turkey
United Arab Emirates
Vietnam v --,:... ~ «> - - s"
Oilier Asia
TOTAL - ASIA
0.01
0.32
0.01
0.05
0.09
0.01
0.12
0.48
- 0.43
0.09
0.09
0.06
0.08
1.27
....
0.42
3.87
0.80
0.31
0.12
0.05
1.04
0.14
0.64
0.02
0.08
0.01
0.16
0.75
0.42
0.10
0.10
0.28
0.20 '
0.01
0.22
2.04
11.8
Regression Model Method
Lower Bound
' " '
<0.01
0.05
0.00
0.01
0.01
0.00
0.02
0.07
0.07
0.01
0.01
0.01
0.01
0.19
.,...
0.06
0.59 .
0.12
0.05
0.02
0.0 1
0.28
0.02
0.10
0.00
0.01
0.00
0.02
0.11
0.06
0.01
0.01
0.04
0.07
0.00
0.03
0.32
1.93
Midpoint of CH4
Emissions
<0.01
0.10
0.00
0.02
0.03
0.00
0.04
0.15
0.13
0.03
0.03
0.02
0.02
0.39
"0.96"
0.13
1.18
0.24
0.09
0.04
0.01
0.50
0.04
0.20
0.00
0.02
0.00
0.05
0.23
0.13
0.03
0.03
0.09
0.12
0.00
0.07
0.62
3.82
Upr>er Bound
<0.01
0.15
0.00
0.02
0.04
0.01
0.06
0.22
0.20
0.04
0.04
0.03
0.03
0.58
................
0^19
1.77
0.37
0.14
0.05
0.02
0.72
0.06
0.29
-'! 0.01
0.04
0.00
0.07
0.34
0.19
0.04
0.04
0.13
0.18
0.01
0.10
0.94
5.70
14
(Continued)
-------
I TABLE 2. METHANE EMISSION ESTIMATES (Tg) FROM BURIED REFUSE USING THE OECD AND
REGRESSION MODEL METHODS (CONT.)
Country
Europe
Albania
Austria
Belgium
Bulgaria
Czechoslovakia
Denmark
Finland
France
German Democratic Republic
Germany, Federal Republic of
Greece
1 Hungary
1 Ireland
1 Italy
1 Netherlands
1 Norway
1 Poland
1 Romania
1 Spain
1 Sweden
1 Switzerland & Liechtenstein
1 Union of Soviet Socialist Republics
1 United Kingdom
1 Yugoslavia
I Other Europe
[TOTAL -EUROPE
1 North America
I Canada
1 United Stales of America
1 Other North America
1 TOTAL - NORTH AMERICA
1 Oceania
Australia
New Zealand*
Other Oceania
TOTAL - OCEANIA
| Latin America
Argentina
Brazil \ -...-
Colombia
Costa Rica
Mexico . . 1 .-"->"-
Venezuela
Oilier South America^
TOTAL LATIN AMERICA
OECD Method
0.01
0.17
0.14
0.06
0.15
0.07
0.24
0.88
0.17
1.80
0.48
0,24
0.11
1.45
0.44
0.11
0.37
0.13
0.85
0.08
0.12
2.49
2.85
0.17
0.08
13.7
2.02
21 .04
0.36
23.4
1.16
0.15 '
0.03
1.34
0.28
" 2.23' "
0.44
0.02
: 0.65
0.05
0.45
4.12
Regression Model Method
Lower Bound
0.00
0.04
0.04
0.01
0.02
0.02
0.06
0.23
0.02
0.48
0.13
0.03
0.03
0.39
0.12
0.03
0.05
0.02
0.23
0.02
0.03
0.32
0.76
., 0.02
0.02
3.12
0.47
3.86
0.06
4.39
0.27
0.04
0.00
0.31
0.04
0.34
0.07
0.00
0.10
0.01
0.07
0.63
Midpoint of CH4
Emissions
0.00
0.08
0.07
0.02
0.04
0.04
0.12
0.42
0.04
0.87
0.23
0.06
0.05
0.70
0.21
0.05
0.10
0.03
0.41
0.04
0.06
0.65
1.37
0.04
0.04
5.74
0.84
8.00
0.11
8.95
0.48
0.07
0.01
0.56
0.09
0.68
0.13
0.01
0.20
0.02
0.13
1.26
tipper Bound
0.00
0.12
0.10
0.02
0.06
0.05
0.17
0.61
0.06
1.25
0.33
0.09
0.08
1.01
0.31
0.07
0.14
005
0.59
0.06
0.08
0.97
1.98
0.07
0.05
8.32
1.21
12.14
0.17
13.5
0.70
0.11
0.01
0.82
0.13
1.02
0.20
0.01
0.30
0.02
0.21
1.89
GLOBAL TOTAL 57 11 21 32
Tins source has been estimated to emit 0 2 Tg/yr by Lassey et al (1992)
-------
Emission estimates for countries other than die U.S. have not been adjusted for the amount of CH4
that is recovered for utilization. Richards (1989) and Tliorneloe (1992) estimate that worldwide
there are 269 sites in 20 countries where landfill gas is recovered, including 114 sites in the U.S.
No estimates are available, however, on the amount of CH4 recovered in other countries.
Trends in Waste Management and Their Impact on CH4 Emissions. As methods of MSW
disposal change, there will be changes in CH4 emissions. Trends in global wasie management and
iheir impact on CH4 release are discussed here.
US. and Canada. Landfilling is the predominant MSW management method in both the U.S.
and Canada. However, there is a trend in both countries toward more recycling, more incineration
(especially in the U.S.), and less landfilling (Kaljian, 1990; U.S. EPA, 1988; Swartz, 1989; El Rayes
and Hdwards, 1991; Alter, 1991). Both countries also have a growing number of landfill gas
recovery sites. The U.S. generates 12 times as much MSW as Canada.
Although the percentage of MSW to be placed in landfills is predicted to decline, increases
in ilie amount of MSW generated and in the percent DOC will cause the fraction of degradable
MSW in landfills to remain close to current levels. For example, if reliance on landfill disposal in
the U.S. were to decline to 50% in the year 2010, 114 of the predicted 227 million Mg of MSW
generated would be placed in landfills. Assuming an increase in carbon content to 22.2%, it is
reasonable to assume that the U.S. will place 25 million Mg of DOC in landfills in 2010. Therefore,
it can also be assumed that the current rate of landfill gas recovery will continue for several more
decades (Boyner et al., 1988; Willumsen, 1990).
Although landfill gas production in the U.S. and Canada is expected to remain relatively
steady over the next two decades, ihe startup of new landfill gas recovery systems is expected to
shift the balance of landfill gas emissions. The amounts of CH4 emitted to the atmosphere will
decrease as more is controlled through flaring or utilization. Concurrently, the amount of CO2
released will increase (Thorneloe, 1992; Bonomo and Higginson, 1988; El Rayes and Edwards,
1991; Boyner et al., 1988; Willumsen, 1990; Rathje, 1991). U.S. landfills are currently recovering
about 1.2 Tg/yr of CH« and producing 344 MWe of power (Tliorneloe, 1992). Clean Air Act
regulations proposed in May 1991 are expected to have a major impact on reducing landfill CH«
emissions from both new and existing MSW landfills in the U.S. The proposed air emission
regulations are expected to result in an additional emission reduction ranging from 5 to 7 Tg/yr of
CRj (Federal Register. 1991; U.S. EPA, 1991).
ORCD Countries. In the future, most OECD countries are considering policies that would
increase the amount of MSW handled by recycling and incineration and to decrease the amount
placed in landfills. Landfills will be large, regional sites that also will be used to dispose of
incinerator ash. The decreases in the amount of MSW landfilled, coupled with increases in landfill
16'
-------
gas recovery and incineraiion, are anticipated to lead to reduced CH4 emissions and increased
emissions of CO2 and other combustion gases.
European Countries and the Former Soviet Union. Sanitary landfills or open dumps (e.g.,
placing refuse in a scattered fashion near residences or along roadsides) are used almost
exclusively for MSW management in Greece, Hungary, Portugal,'Poland, Romania, Bulgaria,
Yugoslavia,'and die former Soviet Union (Bartone, 1990c; Bartone and Haley, 1990; Curi, 1988;
Mnatsaknian, 1991). In the future, Poland plans to close open dumps and dispose of MSW in larger,
regional sanitary landfills. The former Soviet Union hopes to establish an effective recycling
program. No landfill gas recovery sites were identified in these countries.
Asian Countries. Some Asian countries are upgrading their collection methods by
introducing compactor trucks and covered containers. These changes could serve to decrease the
amount of scavenging and increase the amount of MSW that is dumped in an uncontrolled fashion.
Economic constraints, in tandem with a history of slow MSW management development, indicate
that the use of sanitary landfills will net-increase markedly in the near future for most of Asia.
However, increases in population and total MSW will likely lead to increased CH4 emissions.
South America. Central America. Mexico, and the Caribbean Islands. In the future, Brazil
hopes to build recycling and composting plants, and sanitary landfills. Mexico also hopes to
increase its number of sanitary landfills. Few landfill gas recovery sites are currently operating in
South America and Mexico; there seems to some interest in increasing the number of landfill gas
recovery sites in Brazil (Kessler, 1990; Richards, 1988).
Africa and the Middle East. In Africa the only MSW management methods reported to hold
promise for expanded and successful application are recycling, composting, or possibly biogas
recovery if markets and appropriate technologies can be developed to support these systems
(Conner, 1978; Cointreau, 1982; Belts, 1984; Oluwande, 1984; Vogler, 1984; El-Halwagi et al.,
1988). Much of the recyclable material in the MSW stream is currently being recovered
(Cointreau, 1982; Wright et al., 1988), at least when comparing the amount of material recycled in
low-income countries with that of middle-income and developed nations. However, recyclable
materials are still available in the waste streams and, because wages are low in the developing
countries, further recycling may be a viable waste management option (Cointreau, 1982).
Uncertainty Associated with Estimating Landfill CHj Emissions. There are several sources of
uncertainty in estimating emissions of CH
-------
Two issues contribute to the difficulty of estimating CH-j potential from open dumps: (1) ilie
physical characteristics (size, configuration, temperature, moisture, compaction) of open dumps are
unknown, and (2) die quantity and composition of open-dumped waste are also unknown.
However, CH^ is generated from open dumps. Blikle et al, (1990) reported biogas recovery from
two uncontrolled landfills in Nagpur, India. Each of these sites was atom 8 hectares in surface area
and about 3 to 5 m deep. Neither site contained any cover material, and the older of the two
landfills accepted waste from 1971 to 1984. Most of the organic matter had decomposed by the
time me tests were performed, but biogas was obtained from wells 50 mm in diameter at a rate of
0.240 m3/hr (CH4 content not identified). Waste had been deposited in the second site "only
recently" and the rate of biogas recovery was from 5 to 9 nWlir. The CH4 content of the biogas
from the second site was 30 to 40%. The work of Bhide et al. (1990) suggests that open dumps are
a source of CH^. Therefore, they have been included in die emission estimates.
The CH4 potential of other types of landfills, such as those containing industrial and
hazardous wastes, is not well understood. Industrial waste contains waste streams those will
decompose under anaerobic conditions. Certain industrial waste streams, such as those of die food
industry, may have high organic content and are, therefore, potentially significant sources of CH,j.
However, landfills containing hazardous waste will have a low CH4 potential because of the low
-*
moisture content and the requirement that only solid materials are accepted. In addition, die
chemicals in die waste stream may be toxic to the microbes. Therefore, global emissions are
negligible compared to those from MSW land/ills or for industrial landfills.
The disposal of industrial and hazardous waste with MSW was common in the U.S. dtrough
1975. Many closed landfill sites in die U.S. and worldwide contain biodegradable mixtures of
these waste streams. Some industrialized countries such as the United Kingdom consider landfills
as an acceptable treatment option for hazardous and industrial wastes. However, regulations being
considered by die European Commission may prevent this practice. Waste streams in developing
countries are less controlled and mingling of MSW, industrial wastes, and raw sewage in landfills is
common (Cointreau, 1982). Co-disposal sites will generate CH< and may have emission potentials
similar to MSW landfills having no history of co-disposal. Currently, published estimates of CH^
emissions specific to open dumps and industrial and hazardous waste landfills are not available.
Estimates presented in Table 2 include open dump emissions, but do not specifically consider die
potential of land/tiled industrial and hazardous waste.
Wastewater Treatment and Septic Sewage Systems
Wastewater from domestic, commercial, and industrial facilities is also a source
Wastewater treatment lagoons, particularly those diat treat wastewater with high biochemical
oxygen demand (BOD), are suspected of emitting significant amounts of CH4. Global CH«
18
-------
emissions from wastewater treatment lagoons are estimated to be about 25 Tg/yr (Orlich, 1990).
This estimate is based on data from a wastewater lagoon study in Thailand and the assumption that
300 /O3.4 g at standard temperature and pressure) of CH4 is produced per kg of BOD. This
estimate is considered uncertain due to a lack of field data on the ChLj production potential for
different types of lagoons. In addition, uncertainty results from limiiations in available data on
quantities of wastewater being treated and the characteristics of lagoons worldwide that affect CH4
production.
Lagoons are commonly used for wastewater treatment and disposal in developing countries,
primarily because land is available, operations are relatively simple, minimal energy is needed,
and capital and operating expenses are low. Anaerobic condiiions are favored because of ihe
limited maintenance and control of these lagoons and limited use of expensive aeration devices.
Moreover, the average temperature in many tropical and subtropical developing countries is close
10 the optimum biological temperature of methanogenic bacteria (i.e., 35°C). This results in
greater biological activily and a higher CH^ production potential, as compared to lagoons found at
cooler latitudes (Gloyna, 1971).
Mean daily gas production rates were estimated by Toprak (1993) to be 51,000 mVday for
wasiewater treatment lagoons for a treatment plant being constructed in Izmir. This city is one of
the fastest growing metropolitan areas in Turkey. Plans are being considered to utilize thc^ltj as
an alternative source of energy. The wastewater treatment plan in Izmir is to be completed in the
year 2000 (Toprak, 1993).
The World Bank predicts that, because lagoons are relatively inexpensive and easy to
operate, they will continue to be a preferred wastewater treatment method for developing
countries (Bartone, 1990d). Furthermore, with increasing population growth and urbanization in
developing countries, the number and variety of sources discharging into lagoons will increase,
resulting in higher BOD loading rates. Increased CH4 emissions from lagoons will result as these
changes occur.
The U.S. EPA estimates that 130 billion /"(34 billion gal.) of domestic, commercial, and
industrial wastewater is treated in the U.S. each day (U.S. EPA, 1987). Approximately 433
domestic and industrial lagoon systems receive various types and quantities of industrial
wastewater, in addition to domestic wastewater. Furthermore, another estimated 5,000 municipal
lagoons contain domestic wastes from residential, commercial, and institutional sources (U.S. EPA,
1987). Insufficient data are available to characterize the CH4 emission potential of wastewater
treatment lagoons both in the U.S. and globally. An empirical model for wastewater treatment
lagoons using field data would be of value in relating BOD and any other significant factors to CH^
emissions.
The CH^ potential is expected to be minimal for the lagoons that receive pretreated
wastewater. However, when no pretreaiment occurs or pretreatment is minimal, lagoons may be
19
-------
a significant source of CH^. QHLj formation varies depending on temperature, retention time, BOD
loading, lagoon depth, oxygen content, and the frequency at which the lagoon is dredged. The
majority of the U.S. lagoons are facultative; that is, aerobic decomposition occurs in the upper strata
and anaerobic degradation occurs in the lower strata. It is likely that facultative lagoons proceed to
a more anaerobic state as BOD loading increases and surface aeration diminishes (e.g., due to low
wind speed). In particular, information on lagoons used for wastewater from food-processing
industries and other industries that have high BOD wastewater streams would be of value.
Most of the world's population, including about 25% of the U.S. population, relies on
individual septic systems. This is the mosl common treatment and disposal method for domestic
wastewater. Septic tanks are anaerobic digesters of the simplest form and release CH4 to the
atmosphere at a rate thai is dependent on temperature, retention time, and system configuration.
However, a certain portion of the CII4 will be oxidized as the gas diffuses through the soil. Iliese
findings were confirmed in work by Khali! et al. (1990), who showed that very lillle CH4 escapes
to die atmosphere from underground biogas pits in China.
The process of CH
-------
than 1% of the potential CII4 resulting from anaerobic lagoons. Another recent study (Williams,
1993) found that CH4 emissions from dairy cow patties contribute less than 1% of the potential. In
ihe U.S., livestock accounis for about 6 Tg/yr of Cfy, most of which originates in the rumen. CH4
emissions from livestock manure are estimated lo be 0.6 Tg/yr in the U.S. (Johnson et al., this
volume).
In another siucly, Safley et al. (1992) esiimatecl lhat CH,j emissions from livestock manure
contribute 21 to 35 Tg/yr, wilh an average of 28 Tg/yr. The estimate for ihe U.S. is 4 Tg/yr, as
compared tojolmson et al.'s estimate of 0.6 Tg/yr. The major difference between lliese estimaies
is the use of an emission factor representing free-range livestock waste. Ihe factor used in the
Safley study is at least 10 times higher than that derived by Lodman el al, (1993) and Williams
(1993). Using Safley et al.'s methodology and reducing the emission faciors for pasture/range,
clrylot, and daily spread animal waste disposal by a factor of 10, leads to a global estimate of
approximately 15 Tg/yr from animal waste.
The principal factors controlling-potential CH, production from manure include the quantity
and characteristics of the animal waste, the type of wasie management sysiem, and the temperature
and moisture content of llie waste (Safley et al., 1992). Cattle in the U.S. produce larger quantities
of organic waste than any other type of livestock. The average head of cattle produces 24 kg of
wet feces per day (Overcash et al., 1983), including 2.8 kg of organic matter. A portion of this
organic matter can be decomposed by methanogenic bacteria. Oilier animals such as sheep, goats,
horses, and fowl, have a larger fraction of volatile solids in their feces, but because cattle produce a
larger quantity of manure per individual and are more populous than other livestock, tiiey
contribute the largest portion of CH4 from livestock manure. Volatile solids are lhat portion of
organic matter that can be decomposed by microorganisms (Safley et al., 1992). Safley et al.
estimate that global cattle populations contribute 53% of the CH., emissions from livestock waste.
In addition to ihe quantity and characteristics of livestock manure, animal waste
--.;
management systems largely determine the potential for CH4 generation. Livestock are managed
in conditions ranging from open pasture and range to complete confinement. Concurrently,
manure may simply be left on the pasture or range where it is deposited, or it may be prepared
using dry storage methods or liquid treatments. The CH^ production potential of all waste
management systems is highly dependent on tempera lure and moisture (Safley and Westemnan,
1987; Jolinson et al., this volume). Generally, warm lemperatures and high moisture content
provide for maximum CILj production.
Deep pit stacking and daily spreading of solid and semi-solid manure have die least CH4-
producing potential of all livestock waste management systems (Safley et al., 1992). Under these
two management systems, manure generally has a very low moisture content. By contrast,
anaerobic lagoons have the greatest CrVprodudng potential of all livestock waste management
21
-------
systems. Due to die high moisture content of the waste, almost all of the CH^-producing potential
of waste can be realized with proper design and operation of the anaerobic lagoons. Safley et al.
(1992) estimate that the CH^-producing potential of anaerobic lagoons is 70% higher than any other
form of livestock waste management system. In warm and tropical latitudes, anaerobic lagoons
have their greatest CH.(-producing poieniial. Although liquid/slurry management systems have
70% less CH^-producing poiential than anaerobic lagoons (Safley et al., 1992), they rank second
among livestock waste management systems for CH^-producing poteniial, because the moisture
content of the waste is high. Unfavorable temperatures and short residence time of the waste in
storage are probably the factors that limit the CRj production poteniial of liquid/slurry systems. By
virtue of their relatively widespread use in western and eastern Europe, Asia, and North America,
liquid/slurry management systems are esiimated to contribute 26% of the CH4 from livestock waste
management systems. Together, liquid/slurry systems and anaerobic lagoons have been estimated
10 account for 10 Tg/year, or almosl 36% of the total CH^ emissions from livesiock waste
management (Safley et al., 1992).
Because more livesiock waste is deposited in pasture and range systems than in any other
management system globally, assumptions regarding the CH^-producing poieniial of this waste
have significant effects on the total emission estimate. Research being conducted at Colorado State
University indicates that manure CH< production varies under feedlot conditions and under
simulated grazing conditions. Temperature, moisture, and animal diet were the variables that had
the greatest influence on CItj production. More accurate estimates of how manure-handling
systems affect fermentation would help obtain a better estimate of CH^ production.
Uncertainty in the estimates of global CH4 emissions from livestock waste results primarily
from limitations in available data. Data are particularly limited for developing countries and for
free-range livestock waste management. Even in developed countries uncertainly is associated
with animal population estimates, animal sizes and diet, and the types and numbers of animal waste
management systems. In addition, refinements to the current estimates using field test data on the
CH^-production potential of livestock waste both under free range conditions and in livestock
management facilities would be of value. Currently, the only published global estimate for this
source is by Safley et al. initial revised estimates by EPA/AEEIIL suggest that this source contributes
2 to 5 Tg/yr of CH^ globally. Data from field and laboratory studies could help reduce the current
uncertainty of this estimate.
Summary
Global and U.S. estimates of CH4 emissions from landfills, wastewaier treatment, and
livestock waste are presented in Table 3. The recent estimate by EPA/AEERL for landfills (i.e., 11
to 33 Tg/yr) is thought to more accurately reflect CR* emissions from landfilled waste lhan
-------
previous estimates. However, there is siiJI some uncertainty with this estimate, and future
refinements are planned. The global estimates for wastewater treatment and livestock waste are
very uncertain due to limitations in available data, particularly the lack of data on the emission
potential of wastewater treatment lagoons and free-range livestock waste. There is also a lack of
country-specific data for this source category. Uncertainties in ihese estimates result from
optimistic assumptions regarding the extent of anaerobic decomposition and limitations in available
data characterizing (1) waste quantities and composition, and (2) treatmeni or disposal practices,.
Using the ranges presented in Table 3, two alternative estimates of the contribution of
waste sources to global CH4 emissions can be derived by calculating the joint probability
distributions of the estimates as described by Khali! (1992), Assuming that any value within the
range shown for landfills, wastewater and sewage treatment, and livestock waste are equally
probable, a Monte Carlo model was used to generate random values for each source. Repealing
this process for 200 iterations gives a distribution of estimates. Using (1) Hingemer and Cruizen's
(1987) estimate for landfills, (2) Orlich's (1990) estimate for wastewater treatment, and (3) Safley
ei al.'s (1992) estimate for livestock-Waste, a global estimate of 103 Tg/yr with a 95% confidence
interval of 75 to 103 Tg/yr results. Using the EPA/AEERL estimate for landfills (i.e., 11 to 33 Tg/yr)
and Orlich's and Safley's estimates, results in a global estimate for waste management of 72 Tg/yr
with a 95% confidence interval of 54 to 95 Tg/yr Using Lodman et al.'s (1993) and Williams'
(1993) emission factors with Safley et al.'s methodology, plus EPA/AEERL's landfill estimate and
Orlich's wastewater estimate, gives a global estimate of 60 Tg/yr.
TABLE 3. GLOBAL AND US. ESTIMATES OF CBU EMISSIONS FROM WASTE MANAGEMENT
landfills a
Wasiewaier Treatment a
Livestock Waste "
>>:.K-S:'K'::::K^M':*!'S?:;SvR':*:'?>>X'S:::;:?':-:'X-:':^:-:
Global (Tg/yr)
Avg.
50
15
60 c
..22 '
25
25
15'
SS-x-xfcM'S":.!?
Rang?
30-70
10-20
n-33
12-38 d
20-35
^X'K'MPS'ftStfv:*
*:SfS8:f:i&"3:^:.:W:£^
Reference
Bingemer & Crutzen, 1967
Richards, 1989
Bingemer & Crutzen, 1987
TiiomeJoe, 06/92
Oriid), 1990
Safley etal., 1992
S:S»x«i'ttS<'R«S^:->R'>K::-:w:':'S₯x'>;-:;:;::>X':-":^:-:'>?;
US. (Tg/yr)
Avg.
6
23 C
9a
4
0.7
<::*HW>Kv*.v
Range
3>
4-H
S-^'X-^^w!:-:^-
w:i:-:^:-:Sx^:::':!A-::W':<%*:'w::^-S:w:ftx^>::{*^>:WX'>:
Reference
Augensiein & Pacey, 1990
Bingemer & Crutzen, 1987 c
Thorndoe, 08/92
Safley et al, 1992
Lodman a al.. 1992
«<-K:rty:':>K:?>:'>:-ft:;:>:^S₯»x::;:^::^>7^:«iw?:'r':':';':':'!--
1 Potential emissions, not collected for the amount that is dared or utilized. Approximately 1.2 million tonnes of CH4 is being recovered
. from U.S. landfills Olicmeloe, 1992).
c Uses estimated annual placement rates from 1950 to 1990.
d Uses Bingemer & Crurzen's (1967) methodology and updated country-specific data on MSW generation rates.
Assumes 50% uncertainty of Oriich's (1990) estimate of 25 Tg/yr.
e Uses Lodman et al.'s pasture/range emission factor, and Safley et al.'s methodology
-------
literature Cited
Ahmed, MF. 1986. Recycling of solid wastes in Dhaka In: Waste Management in Developing
Countries, 1, KJ. Thome-Kozmiensky, ed. EF-Verlag fur Energie und Urm'elttechnik
GmbH, Berlin, pp. 169-173.
Alter, H. 1991. The future course of solid waste management in the U.S., Waste. Mgmt. & Res,,
#3-20,
Augenstein, D., J. Pacey. 1990. Modeling landfill CH^ generation, International conference on
landfill gas: energy and etwironment, 10/17/90, Bournemouth, England.
Barlaz, M.A. 1985. Factors affecting refuse decomposition in sanitary landfills, M.S. Thesis, Dept.
of Civil and Environmental Engineering, Univ. of Wisconsin - Madison.
Barlaz, M.A. 1988. Microbiological and chemical dynamics during refuse decomposition in a
simulated sanitary landfill, Ph.D. Dissertation, Department of Civil and Environmental
Engineering, University of Wisconsin - Madison.
Barlaz, M.A., D.M. Schaefer, R.K. Ham. 1989a. Bacterial population development and chemical
characteristics of refuse decomposition in a simulated sanitary landfill. Appl. Env. Microbiol.
55O):55-65.
Barlaz, M.A., R.K. Ham, D.M. Schaefer. 1989b. Mass balance analysis of decomposed refuse in
laboratory scale lysimeters. ASCEJ. of Environ. Engineering, JJ5(6):l 088-1 102.
Barlaz, M.A, R.K, Ham, D.M. Schaefer. 1990. CH4 production from municipal refuse: a review of
enhancement techniques and microbial dynamics. CKC Critical Revieivs in Environmental
Control, 19, Issue 6.
Barlaz, M.A., M.W. Milke, R.K. Ham. 1987. Gas production parameiers in sanitary landfill
simulators. Waste Mgmt. & Res., 5:27.
Barione, C.R, 1990a. Economic and Policy Issues in Resource Recovery from Municipal Solid
Wastes, Resources, Conservation and Recycling, 4:7-23.
Bartone, C.R. 1990b. Investing in Environmental Improvements Through Municipal Solid Waste
Management. Paper presented at the WHO/PEPAS Regional Workshop on National Solid
Waste Action Planning, Kuala Lumpur, 02/26/90-03/02/90.
Bartone, C.R. 1990c. Economic and policy issues in resource recovery from municipal solid
wastes. Resour. Conserva. Recylc., 4:7-23:
Bartone, C.R. 1990d. Urban wastewater disposal and pollution control: Emerging issues for sub-
Saharan Africa. Proceedings of the African Infrastructure Symposium, The World Bank,
Baltimore, Maryland, 01/08-09/90, p. 6.
Bartone, C.R., C. Haley. 1990. The Bled Symposium: Introduction. Resour. Conserva. Recycl.,
4:1-6.
Bartone, C.R., L. Leite, T. Triche, R. Schertenleib. 1991. Private sector participation in municipal
solid waste service: Experiences in Latin America. Waste Mgmt. &Res., .9:495-509.
Bateman, C.S. 1988. Landfill gas development in Australia. Y.R. Alston and G.E. Richards, eds. In:
Proceedings of the International Conference on Land/Ill Gas and Anaerobic Digestion of
Solid Waste, October 4-7, Harwell Laboratory, Oxfordshire, UK. pp. 156-161.
Beker, D. 1990. Sanitary landfilling in the Netherlands. In: International Perspectives on
Municipal Solid Wastes and Sanitary Landfilling, J.S. Carra and R. Cossu, eds. Academic
Press, New York, NY. pp. 139-155.
Belts, M.P. 1984. Trend in solid waste management in developing countries. In: Managing Solid
Wastes in Developing Countries, J.R. Holmes, ed. John Wiley & Sons, Ltd. Chicbester,
England, pp. 291-302. ....
Bhide, A.D., S.A. Gaikwad, B.Z. Alone. 1990. CH4 from land disposal sites in India. Proceedings
of the International Workshop on CHf Emissions from Natural Gas Systems, Coal Mining
and Waste Management Systems. Environment Agency of Japan, the U.S. Agency for
International Development, and the U.S. Environmental Protection Agency,. Washington,
DC, 04/09-13/90.
Bhide, A.D., B.B. Sundaresan. 1983. Solid waste management in developing countries. Indian
National Scientific Documentation Centre, New Delhi, India.
-------
Bingemer, H.G., PJ. Crutzen. 1987. The production of CH4 from solid wastes. / Geopbys. Res.,
#2(D2):2181-2187.
Bogner, J.E. 1990. Controlled study of landfill biodegradatiori rates using modified BMP assays.
Waste Mgmt. & Research, 8:329-552.
Bonomo, L, A.E. Higginson. 1988. International Oferi/ieu> on Solid Waste Management.
Harcoun Brace Jovanovich, New York, NY. p. 268.
Boyner; J., M. Vogt, R. Piorkowski, C. Rose, M. Hon. 1988. U.S. Landfill Gas Research. Y.R.
Alston, and G.E. Richards, eds. In: Proceedings of the International Conference on Landfill
Gas and Anaerobic Digestion of Solid Waste, 10/04-07/88, Harwell Laboratory,
Oxfordshire, UK. pp, 313-338.
Buivid, M.G., et al. 1981. Fuel gas enhancement by controlled landfilling of municipal solid
waste, Resource Recovery and Conservation 6:3,
Campbell, D.( D. Epperson, L Davis, R. Peer, W. Gray. 1991 Analysis of Factors Affecting
Methane Gas Recovery from Six Landfills. Prepared for Air and Energy Engineering
Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC.
EPA-600/2-91-055 (NTIS PB92-10135D.
Carra, J.S., R. Cossu. 1990. International Perspectives on Municipal Solid Wastes ami Sanitary
Landfilling, J.S. Carra, and R. Cossu, eds. Academic Press, New York, NY. pp. 1-14.
Coinireauj'SJ. 1982. Environmental Management of Urban Solid Wastes in Developing Countries.
A Project Guide, Urban Development Tecluiical Paper Number 5, World Bank, Washington,
DC. pp. 1-17.
Cointreau, S.J. 1984. Solid waste collection practice and planning in developing countries. In:
Managing Solid Wastes in Developing Countries, J.R. Holmes, ed. John Wiley & Sons, Ltd.
Chichester, England, pp. 151-182.
Cointreau, S.J. 1987. Solid Waste Management Study for the Greater Banjul Area, The Gambia.
Ministry of Economic Planning and Industrial Development, Banjul, The Gambia. 09/87.
Conner, M. A. 1978. Modern Technology for Recovering Energy and Materials from Urban Wastes
- Its Applicability in Developing Countries. Resour. Conserv. Recycl., 2:85-92.
Cossu, R. 1990a. Sanitary landfilling in Japan. In: International Perspectives on Municipal Solid
Wastes and Sanitary Landjilling, J.S. Carra, and R. Cossu, eds. Academic Press, New York,
NY. pp. 110-138.
Cossu, R. 1990b. Sanitary landfilling in the United Kingdom. In: International Perspectives on
Municipal Solid Wastes and Sanitary Land/tiling, J.S. Carra and R. Cossu, eds. Academic
Press, New York, NY. pp. 199-220.
Cossu, R., G. Urbini. 1990. Sanitary landfilling in Italy. In: International Perspectives on
Municipal Solid Wastes and Sanitary Land/tiling, J.S. Carra and R. Cossu, eds. Academic
Press, New York, NY. pp. 94-109.
Curi, K. 1988. Comparison of solid waste management in tourisitic areas of developed and
developing areas. In: Proceedings of the 5 tb International Solid Waste Conference, Internal
Solid Waste and Public Cleansing Association, Copenhagen, Denmark, 09/88.
Diaz, L.F., C.G. Golueke. 1987. Solid waste management in developing countries. Biocycle,
28:50-55.
El-Halwagi, M.M., S.R. Tewfik, M.H. Sorour, A.G. Abulnour. 1986. Municipal solid waste
management in Egypt: practices and trends. In: Waste Management in Developing
Countries, /, K.J. Thome-Kozmiensky, ed EF-Verlag fur Energie und Emvelitechnik GmbH,
Berlin, pp. 283-288.
EI-Halwagi, M.M., S.R. Tewfik, M.H. Sorour, A.G. Abulnour. 1988. Municipal solid waste
management in Egypt. In: Proceedings of the 5 tb International Solid Waste Conference,
International Solid Waste and Public Cleansing Association, Copenhagen, Denmark,
September, pp. 415-424.
El Rayes, H., W.C. Edwards (B.H. Levelton & Associates, Ltd.) . 1991. Inventory of CH^ Emissions
from Landfills in Canada. Prepared for Environment Canada, Hull, Quebec, pp. 25-69.
Emberton, J.R. 1986. The biological and chemical characterization of landfills. In: Proceedings of
Energy from Landfill Gas, Solihull, West Midlands, UK, 10/30-31/86.
25
-------
EMCON Associates. 1982. CH^ Generation and Recovery from Landfills, Ann Arbor Science
Publishers, Inc., Ann Arbor, MI.
Ernst, A. 1990. A review of solid waste management by composting in Europe. Resour. Conserva.
Recycl., 4:135-149.
Ettala, M.O. 1990. Sanitary landfilling in Finland. In: International Perspectives on Municipal
Solid Wastes and Sanitary Landfilling, Jf.S. Carra and R. Cossu, eds. Academic Press, New
York, NY. pp. 67-77.
Federal Register, Vol. 56, No. 104. May 30, 1991, pp. 24468 - 24528.
Framzis, I. 1988. Recycling in Greece. BioCycle.30-31.
Gadi, M.T. 1986. In: Waste Management in Developing Countries, 1, KJ. Thome-Kozmiensky,.
ed. EF-Verlag fur Energie and Umvelnechnik GmbH, Berlin, pp. 188-194.
Gandolla, M. 1990. Saniiary landfilling in Switzerland. In: International Perspectives on
Municipal Solid Wastes and Sanitary Landfilling, J.S. Carra and R. Cossu, eds. Academic
Press, New York, NY. pp. 190-198."
Gloyna, E.F. 1971. Waste stabilization ponds. World Health Organization, Geneva p. 76.
Halvadakis, C.P., et al, 1983. Landfill Melhanogenesis; Literature Review and Critique, Technical
Report No. 271, Department of Civil Engineering, Stanford University.
Hayakawa, T. 1990. The status report on waste management in Japan - special focus on methane
emission prevention. In: Proceedings of the International Workshop on Methane Emissions
from Natural Gas Systems, Coal Mining, and Waste Managemetu Systems, Washington, DC.
April 9-13. pp. 509-523.
Holmes, J.R. 1984. Solid waste management decisions in developing countries. In: Managing
Solid Wastes in Developing Countries, J.R. Holmes, ed. John Wiley & Sons, Ltd. Chichesier,
England, pp. 1-17.
IPCC. 1992. Climate Change 1992. The Supplementary Report to the IPCC Scientific Assessment.
Published for The Iniergovernmental Panel on Climate Change (IPCC), World
Meteorological Organization/United Nations Environment Programme. Cambridge
University Press. Edited byJ.T. Houghton, G.J. Jenkins, and J.J. Ephraums.
International Energy Agency. 1992. Landfill Gas Enhancement Test Cell Data Exchange-Final
Report of die Landfill Gas Expert Working Group. Editor: Pat Lawson, AEA-EE-0286.
Jenkins, R.L., J.A. Pettus. 1985. In: Biotecbnological Advances in Processing Municipal Wastes for
Fuels and Chemicals, A.A. Antonopoulos, ed, Argonne Natl. Lab. Report ANL/CNSV - TM -
167, p. 419.
Kaldjtan, P. 1990. Characterization of Municipal Solid Waste in the United States: 1990 Update,
EPA-530/SW-90-042, PB90-215112. Office of Solid Waste, Washington, DC.
Kaltwasser, BJ. 1986. Solid waste management in medium sized towns in the Sahel area. In:
Waste Management in Developing Countries, 1, KJ. Thome-Kozmiensky, ed. EF-Verlagfur
Energie und Umvelueclmik GmbH, Berlin, pp. 299-307.
Kessler, T. 1990. Brazilian trends in landfill gas exploitation. ETATEC Consultores s/c Ltda, S&o
Paulo, Brazil.
Khalil, M.A.K. 1992. A statistical method for estimating uncertainties in the total global budgets of
atmospheric trace gases. J. Environ. Sci. Health, A27:777-770.
Khalil, M.A.K., R.A. Rasmussen, M.-X. Wang. 1990. Emissions of trace gases from Chinese rice
fields and biogas generators: CH«, tyO, CO, CO2, chlorocarbons, and hydrocarbons.
Cbemospbere, 20:207-226.
Kinman, R.N., et al. 1987 Waste Management and Research. 5, p. 13-
Lassey, K.R., D.C. Lowe, M.R.Manning. 1992. A Source Inventory for Atmospheric CH4 in New
Zealand and Its Global Perspective, Journal of Geophysical Research, _97fZ>f>.3751-3765.
Lawson, P. 1992. Landfill gas expert working group summary report, 1989-1991. International
Energy Agency, AEA-EE-0305.
Lodman, D.W., M.E. Branine, B.R. Carmean, P. Zimmerman, G.M. Ward, D.E. Johnson. 1992.
Estimates of CH4 Emissions from Manure of U.S. Cattle. Chemosphere (In Press).
Lodman, D.W., M.E. Branine, B.R. Carmean, P. Zimmerman, G.M. Ward, D.E, Johnson. 1993
Estimates ofCHj Emissions from Manure ofU.S. Cattle. Chemosphere, 26(1-4)
26
-------
Lohani, B.N., N.C. Tlianh. 1980. Problems and practices of solid waste management In Asia. /.
Envion. Set. 06/80, pp. 29-33.
Maniatis, K., S. Vanhille, A. Hartawijaya, A. Buekens, W. Verstraete. 1987. Solid waste
management in Indonesia: Status and Potential Resources Conservation Recycle, 15 (87):
277-290.
Mwiraria, M, J. Broome, R. Semb, and W.P. Meyer. 1991. Municipal Solid Waste Management in
Uganda and Zimbabwe. Draft report of the United Nations Development Program and The World
Bank. 05/18/91.
Mei-Chan, L 1986. Waste management in the Taiwan area. In: Waste Managemeni in
Developing Countries, 1, K.J. Thome-Kozmiensky, ed. EF-Verlag fur Hnergie und
Umvelttechnik GmbH, Berlin, pp. 247-250.
Mnaisaknian, R.A. 1991. Legislation and public control of waste sites in the U.S.S.R. In:
Proceeding? of the Third International Landfill Symposium, Sardinia, Italy. 10/91. pp.
1747-1753.
Monney.J.G. 1986. Municipal solid waste management-Ghana's experience. In: Waste
Management in Developing Countries, /, K.J. Thome-Koi:miensky, ed. EF-Verlag fur
Energie und Umveltiechnik GmbH, Berlin, pp. 32-1327.
Mwiraria, M.,J. Broome, R. Semb, W.P. Meyer. 1991. Municipal Solid Waste Managemeni in
Uganda and Zimbabwe. Draft report of the Uniled Nations Development Program and lite
World Bank. 05/18/91.
Nozhevnikova, A.N., A.B. Lifshitz, V.S. Lebedev, G.A. Zavarzin. 1993 Emission of methane into
the atmosphere from landfills in ihe former USSR. Chemo$pbere26(l-4)-.(\n press).
Oluwande, P.A. 1984. Assessment of solid waste management problems in China and Africa. In:
Managing Solid Wastes in Developing Countries, J.R. Holmes, ed. John Wiley & Sons, Ltd.,
Chichester, England, pp. 71-89.
OECD. 1991- Estimation of Greenhouse Gas Emissions and Sinks. Final Report from the OECD
Experts Meeting, 02/18-21/91. Prepared for the IPCC. Revised 08/91.
Orlich, J. 1990. CH4 Emissions from Landfill Sites and Waste Water Lagoons. Proceedings of the
Internationa] Workshop on Cftj Emissions from Natural Gas Systems, Coal Mining, and Waste
Management Systems, 04/09-13/90. Environment Agency of Japan, U.S. Agency for
International Development, and the U.S. Environmental Protection Agency, Washington,
DC.
Overcash, M.R., F.J. Humenik, J.R. Miner. 1983- Livestock Waste Managemeni. Vol.11. CRC
Press, Boca Raton, FL.
Pacey.J. 1989. Enhancement of Degradation: Large-Scale Experiments. In: Sanitary Landfitting:
Process Technology and Environmental Impact, T. Christensen, R. Cossu, R. Stegmann, eds.,
Academic Press, London, pp. 103-119.
Pairoj-Boriboon, S. 1986. State-of-the-art of waste management in Thailand In: Waste
Management in Developing Countries, 1, K.J. Thome-Kozmiensky, ed. EF-Verlag fur
Energie und Umvelttechnik GmbH, Berlin, pp. 208-219.
Papachristou, E. 1988. Solid wastes management in Rhodos. In: Proceedings of the 5 tb
International Solid Wastes Conference. International Solid Waste and Public Cleansing
Association, Copenhagen, Denmark, September.
Parkin, G.F., W. F. Owen. 1986. Journal of the Environmental Engineering Division, ASCE, 112,
5,p,867. - ->.--
Peer, R.L., D.L. Epperson, D.L. Campbell, P. Von Brook. 1992. Development of an Empirical
Model of Methane Emissions from Landfills. EPA-600/R-92-037, PB92-152875- Prepared
for the U.S. Environmental Protection Agency, Air and Energy Engineering Research
Laboratory, Research Triangle Park, NC. '.1 -.-".-,
Peer, R.L., S.A. Thorneloe, D.L. Epperson. 1993. A comparison of methods for estimating global
methane emissions from landfills. Chemosphere, 26(1-4).
Peterson, C, A. Perlmutter. 1989. Composting in the Soviet Union. BioCycle, July, 74-75.
Pohiand, F.G. 1975. Sanitary landfill stabilization with leachate recycle and residual treatment.
EPA-600/2-75-043, PB248524.
27
-------
Pohland, F.G., S.R. Harper. 1987. Critical review and summary of leachate and gas produciion
from landfills. EPA/600/2-86/073, PB86-24018L
Population Reference Bureau, Inc. 1989. World Population Data Sheet of the Population
Reference Bureau, Jnc. Demographic data and estimates for the countries and regions of the
world. Washington, DC.
Rathje, W.L 1991. Once and future landfills, National Geographic, May, pp. 117-134.
Richards, K.M. 1988. Landfill gas - a global review. In: Biodeterioraiion 7, D.R. Houghton,
R.N. Smith, H.O.W. Eggins, eds. Elsevier Applied Science, London, pp. 774-790.
Richards, K.'M. 1989. Landfill gas: working with Gaia. Biodelurioration Abstracts, 3:317-331.
Safley, L.M., P.W. Wesiennan. 1987. Biogas production from anaerobic lagoons. Biological
Wastes, 23=181.
Safley, L.M., M.E. Casada, J.W. Woodbury, K.F. Roos. 1992. Global CH4 emissions for livestock
and poultry manure. EPA/400/1-91/048, U.S. Environmental Protection Agency, Air and
Radialion, Washington, DC, 02/92.
Scheepera, M.J.J. 1990. Landfill gas in the Dutch perspective. International Conference on
Landfill Gas: Energy and Environment '90, Session 3.2.
Shelton, D.R., J.M. Tiedje. 1984. General method for determining anaerobic biodegradaiion
potential. Appl. Env. Microbiol., 47:850-857.
Stegmann, R. 1990. Sanitary landfilling in the Federal Republic of Germany. In: International
Perspectives on Municipal Solid Wastes and Sanitary Landfilling, J.S. Carra, and R. Cossu,
eds. Academic Press, New York', NY. pp. 51-66.
Swartz, A. 1989- Overview of International Solid Waste Management Methods. Stale Government
Technical Brief 98-89-MI-2. The American Society of Mechanical Engineers, Washington,
DC.
Thorneloe, S. June 1992. Landfill Gas Recovery/Utilization - Options and Economics, Published in
Proceedings for the l6th Annual Conference by the Institute of Gas Teclinology on Energy
from Biomass and Waste, 03/92, Orlando, FL.
Thorneloe, S.A. August 1992. Emissions and Mitigation at Landfills and Other Waste Management
Facilities. Presented at the EPA Symposium on Greenhouse Gas Emissions and Mitigation
Research, Washington, DC. To be published in conference proceedings.
Toprak, H. 1993. CH« emissions originating from die anaerobic waste stabilization ponds.
Cbemospbere,26(14).
United Nations. 1989. City Profiles. Prepared by the United Nations Centre for Regional
Development and the Kitakyushu City Government. 64 pp.
United Nations. 1990. 1988 Demographic Yearbook. Fortieth Issue. Department of International
Economic and Social Affairs, Statistical Office. New York, NY.
United Nations Development Programme (UNDP), The World Bank, and the Canadian
International Development Agency. 1987. Master Plan for Resource Recovery and Waste
Disposal, City of Abidjan. Final Report. Prepared by Roche Ltd. Consulting Group, Sainte-
Foy, Quebec, Canada. February.
United Republic of Tanzania. 1989. Masterplan of Solid Waste Management for Dar es Salaam.
Volume II: Annexes. Ministry of Water, Department of Sewerage and Sanitation. Prepared
by HASKONING, Royal Dutch Consulting Engineers and Architects, Nijmegen, The
Netherlands, and M-Konsult Ltd., Consulting Engineers, Dar es Salaam, Tanzania, pp. 1-39,
47-49, 74-94, 109-124, 142-145. -
U.S. Environmental Protection Agency. 1987. Report to Congress, Municipal Wasietvater Lagoon
Study, Office of Municipal Pollution Control.
U.S. Environmental Protection Agency. 1988. Report to Congress, Solid Waste Disposal in the
United States, Volume 1, EPA/530-SW-88-011, PB89-110381, Office of Solid Waste,
Washington, DC. -
U.S. Environmental Protection Agency. 1991. Air Emissions from Municipal Solid Waste Landfills
- Background Information for Proposed Standards and Guidelines, EPA-450/3-90-01 la,
PB91-197061. Office of Air Quality Planning and Standards, Research Triangle Park, NC.
28
-------
Verrier, SJ. 1990. Urban Waste Generation, Composition and Disposal in South Africa. In:
International Perspectives on Municipal Solid Wastes and Sanitary Landfilling. J.S. Carra, and
R. Roccu, eds. Academic Press, Harcourt Brace Jovanovich, London, England, pp. 161-176.
Vogler, J.A. 1984. Waste recycling in developing countries: A review of the social, technological,
and market forces. In: Managing Solid Wastes in Developing Countries, J,R. Holmes, ed.
John Wiley and Sons, Ltd., Chichester, England, pp. 241-266.
Whalen, S.C., W.S. Reeburgh, K.S; Sanbeck. '1990. Rapid methane oxidation in a landfill cover.
Applied and Environmental Microbiology 56:3405-3411.
Williams, DJ. 1993. Methane emissions from the manure of free range dairy cows. Cbemosphere,
2tfC/-O(in press).
Willumsen, H. 1990. Landfill gas. Resour. Conserva. Recycl. 4(90}. 121-133.
Wolfe, R.S. 1979. Meihanogenesis, In: Microbial Biochemistry, International Review of
Biochemistry, Vol. 21, Quayle, J.R., Ed.k University Park Press, Baltimore, MD.
Wolin, M.J. and T.L. Miller. 1985. In Biology of Industrial Microorganisms, ed. A.L Dernsin and
N.A. Solomon, Benjamin/Cumings Publishing Company, Inc., Menlo Park, CA. pp. 189-221.
World Bank, The. 1985. Metropolitan Area of Douala. Study of Waste Management and Resource
Recovery. Part A. Phase I. Prepared by Motor Columbus, Consulting Engineers, Inc., CH-
5401 Baden, Switzerland.
World Resources Institute. 1990. World Resources 1990-91. Oxford University Press, New York,
NY.
Wright, F., C. Bartone, S. Arlosoroff. 1988. Integrated resource recovery: optimizing waste
management. In: Proceedings of the 5 tb International Solid Wastes Conference,
International Solid Wastes and Public Cleansing Association, Copenhagen, Denmark, pp.
619-624, 09/88.
Xianwen, C., Z. Yanhua. 1991. Landfill gas utilization in China. In: Proceedings of the Third
International Landjill Symposium, Sardinia, Italy, October, pp. 1747-1753.
Yepes, G., T. Campbell. 1990 (draft). Assessment of Municipal Solid Waste Services in Latin
America. Report in progress prepared for The World Bank, Technical Department,
Infrastructure and Energy Division, Urban Water Unit, Latin America and the Caribbean
Region, pp. 1-6, 20-26. -
Young, LY,, A.C. Frazer. 1987. GeomicrobiologyJ.5;26l.
Zehnder, A.J.B., et al. 1982. Microbiology of CH< bacteria. In: Anaerobic Digestion, D.E.
Hughes, ed., Elsevier Biomedical Press B.V., Amsterdam, p. 45.
Zsuzsa, K.P. 1990. Possibilities for utilization of the energy content of the solid wastes of
settlements. Resour, Conserva. Recycl., 4:173-180.
29
-------
-------
Chemosphere, 26 (1993) 453-472 '
ESTIMATE OF GLOBAL METHANE EMISSIONS FROM
COALMINES
David A. Klrchgessner1, Stephen D. Piccot1*, J. David Winkler3
'U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park. North Carolina 27711, USA
'Science Applications International Corporation
Air Pollution Research Programs Branch
3101 Petty Road, Durham, North Carolina 27707, USA
'Alliance Technologies Corporation
100 Europa Drive, Chapel Hill, North Carolina 27514, USA
ABSTRACT
Country-specific emissions of methane (CH«} from underground coal mines, surface coal mines, and coal
crushing and transport operations are estimated for 1989. Emissions for individual countries are estimated by
using two sets of regression equations (R5 values range from O.S6 to 0.71). The first set is used to estimate the
CHi content of coals in selected countries based on country-specific coal depth and other relevant paran&ers.
The second equation relates this CH4 content and the country's coal production rate to the emissions'from coal
mining operations. The regression equations developed in this study rely on documented relationships which exist
between mine emissions, coalbed CH« content, coal production rate, and other coal properties. Only those
independent variables which could be included at 95 percent confidence or greater were retained in the regression
equations. Estimated global CH, emissions from coal mining are estimated to be 45.6 Tg for 1989.
1. INTRODUCTION
Methane (CHJ is a radiative!)' important trace gas which accounts for about 18 percent of anthropogenic
greenhouse wanning. Atmospheric concentrations of CH, are now increasing at the rate of 1 percent per year
(Smith and Tirpak, 1989). Although the global CH, cycle is not fully understood, significant sources of emissions
(in order of decreasing emissions) include wetlands, ruminants, rice paddies, biomass burning, coal mines, natural
gas transmission facilities, landfills, termites", and tundra (Wuebbles and Edmonds, 1991). Improved emissions
estimates for these sources will allow their relative contributions to the global CH, cycle to be better understood,
and will provide a means for focusing future emissions mitigation research.
Attempts made to estimate global emissions from coal mining operations have generally relied solely on
-------
global coal production data and emission factors derived from CH« contents of coalbeds (Koyama, 1963; Marland
and Rotty, 1984; Cicerone and Oremland, 1988). These estimates are based on the assumption that emissions
are equal 10 the amount of CH« trapped in the coal removed from the mine. Although this trapped CH4 is
liberated when coal is fractured and removed from the mine, there are other CH, release mechanisms in the
mining process which this assumption fails to take into account. For example, CH< may be released from: (1)
exposed coa! surfaces throughout the mine workings (i.e., the roofs, floors, and walls); (2) gas which is trapped
in the strata adjacent to the mined seams; and (3) underlying seams close to the seam being mined. Commonly
cited global mine emissions estimates range from 25 to 45 teragrams (Tg) of CJVyear, which corresponds to
roughly 10 percent of total annual CH, emissions from anthropogenic sources (Cicerone and Oremland, 1988).
A recent report contains emissions estimates as high as 33 to 64 Tg CH«/year (Boyer et al., 1990).
Underground, surface, and abandoned or inactive mines comprise the three general sources of mine related
CH4 emissions. Emissions from underground mines can be liberated from three sources: (1) ventilation shafts:
(2) gob wells; and (3) crushing operations. Ventilation air, although generally containing 1 percent or less CH4,
contributes the majority of mine emissions because of the enormous volume of air used to ventilate mines. Gob
wells are drilled into the area immediately above the seam being mined. They provide conduits for venting CH,
which accumulates in the rubble-filled areas formed when the mine roof subsides following longwall mining.
Their purpose is to remove CH, which would otherwise have to be removed by larger and more costly shaft
ventilation systems. Currently, no published data for the release of CH4 from gob wells exist. However,
preliminary data obtained from the coal mining industry indicate that gob well CH, emissions could account for
a significant fraction of the total emissions associated with some longwall mines (Sddt, 1991). Emission data
for crushing operations are also extremely limited. ....
In surface mines, the exposed coal face and surface, and in particular areas of coal nibble created by the
blasting operation, are expected to provide the major sources of CH,. As in underground mines, however,
emissions may also be contributed by the overburden and by underlying strata. Emissions from abandoned mines
may come from unsealed shafts and from vents installed to prevent the buildup of CH« in the mines.
The main purpose of this research is to develop an improved methodology for estimating global CH*
emissions from underground coal mining operations and to produce a global emissions estimate using this
methodology where country-specific estimates are not available. The underground mine methodology integrates
data on coal production, coa! properties, coalbed CH« contents (i.e., the volume of CH« per ton of coal), and coal
mine ventilation air emissions from U.S. mines. The objective is to develop a procedure which can be used to
estimate mine emissions from generally available coal analyses and production data where coalbed CH» data or
emission estimates are not available for a country. . .
Since emissions data are presently not available for surface mines, this methodology is currently restricted
to underground mines. The Air and Energy Engineering Research Laboratory (AEERL) of the U.S.
Environmental Protection Agency (EPA) has embarked upon a measurements program to .quantify CH4 emissions
from selected surface mines in the United States for later inclusion in this work. Until that time, surface mine
emissions estimates are included here using simplified assumptions since the lack of data prevents more precise
-------
inclusion of surface mines at this time.
Similarly, virtually no data exist on emissions from handling operations (i.e., crushing, grinding, transport.
and storage) although their magnitude will certainly depend, to a large extent, on the desorption characteristics
of individual coals. These emissions will be estimated for now following precedents from the literature.
2. BACKGROUND
Numerous studies have examined the physical relationships which control the production and release of CH<
by coal. These studies have been conducted either to evaluate the potential of coalbed CH« resources or to
enhance the safety of underground mines. Generally, the studies address one of two topics: (1) factors
controlling coalbed CH< content; or (2) factors controlling the concentration of CH, in the mine atmosphere and
mine ventilation air.
Studies in the first group have identified pressure, coal rank, and moisture content as important determinants
of coalbed CH4 content. Kim related gas content to coal temperature and pressure, and in turn to coal depth
(Kim, 1977). After including coal analyses data to represent rank, Kim produced a diagram relating gas content
to coal depth and rank. Although the validity of the rank relationship has been questioned, it generally appears
to have been accepted by recent authors (Lambert et el.. 1980; Murray, 1980; Ameri et a!.. 1981; Schwarzer
and Byrer, 1983). Independently of Kim's work, Basic and Vukk established the relationship of CH, content
with depth in brown coals and lignite (Basic and Vukic, 1989).
Several studies have recognized the decrease in CH, adsorption on coal as moisture content increases in the
lowest moisture regimes (Anderson and Nofer, 1965; Jolly a al., 1968; Joubert et al., 1974). Moisture content
appears to reach a critical value above which further increases produce no significant change in CH« content.
Coals studied by Joubert et al. showed critical values in the range from 1 to 3 percent (Joubert et al., 1974).
Investigations which attempt to identify correlates of CH, content in coal mine ventilation air include those
by Irani et al. (1972) and by Kissel et al. (1973). Irani et al. developed a linear relationship between CH*
emissions and coal production depth for mines in five seams. Kissel et al. demonstrated a linear relationship
between CH, emissions and coalbed CH« content for six mines. Although both studies suffer from a paucity of
mines and/or seams in their analyses, Kissel et al. made the important observation that mine emissions greatly
exceed the amount expected from an analysis of coalbed CH* content alone. Emissions are produced not only
by the mined coal, but also by the coal left behind and by surrounding strata. For the six mines studied,
emissions per ton mined exceeded coalbed CH« per ton by factors of from six to nine.
3. METHODS
As shown in Figure I, the development of the emissions estimation procedure for underground mines focuses
on two areas: (1) evaluations of characteristics which affect coalbed CH< content; and (2) evaluations of mine
shaft emissions characteristics. The first area examines the independent variables controlling coalbed in-situ CH4
-------
DATABASE DKVEUOPHEVT
Shaft Ventilation Bmli*ieni
Cob Hill Emiaiiona
Coal Production Rate
DEVZLOP TIE NIKt
MISSIONS
DATABASE
Coalbed Caa Co'htent
Coal Heating Value-
Coal Carbon Content
Coal Moicture Content-
Coal Volatile Matter-
Coal Depth
DEVELOP THE
DATABASE
DATA BOtZEXINO SO
UMOVE INCOMPLETE C
INCONSISTENT MSA
STXf 31 DATA
BZANIHZ BAZA AND IDEMTIPT BZI
VAXIA8LES/TKENDS NBICB MAY APFECT
COALBED METHANE AMD MINZ EMISSIONS
CONDUCT NULTI-VARIATE U6RESSION
ANALYSIS TO EXAMINE THE PAXAKETERS
IDENTIFIED IN IKE ABOVE ANALYSES
DEVELOP A SET OP EQUATIONS PO* BSTIMAZINC
EMISSIONS ntON tmDCROROUND MINIS OIVEN COAL
QUALITY, HANK, AND PRODUCTION RATE DATA
FIGURE 1. Overview of the technical approach.
-------
content, and applies regression analyses to estimate key relationships. The dependent variable, coalbed CH«, and
numerous combinations of independent variables were subjected to regression analyses. Hie equations having
the highest RJ and the lowest Mallows CP statistic were selected. In addition, only equations with terms having
students t-statistics of two and above were used (i.e., there is at least 95 percent confidence that the term is
statistically significant). Hie second path follows the same course in attempting to quantify factors controlling
CH4 emissions in mine ventilation air. In both cases, analyses include data collection, data screening,
identification and evaluation of independent variables, and the estimation of regression equations with supporting
statistics. The remainder of this section describes these procedures in detail.
3.1 Coalbed Data Analysis
The purpose of this portion of the analysis is to produce a method for estimating coalbed CH, values when
actual CH* values are not available for a country. ' The method relies on the use of generally available coal
properties such as depth, heating value, and proximate analysis data. Four sources of data were used to estimate
regression equations for coal bed CH« content. Data on U.S. coalbed CH, contents have been compiled by
Diamond et a!, for the U.S. Bureau of Mines and by Tremain for the Colorado Geological Survey (Tremain,
1980; Diamond er a/,, 1986). The Tremain data include coal depth and coal analyses. The Diamond database,
in addition to CH, values, includes only coal depth, rank, and ash content. The coal analyses needed to
supplement the Diamond data were found in Schwarzer and Byrer and in Bureau of Mines' files (Schwarzer and
Byrcr, 1983). A total of 148 data sets were identified which were both complete and for which the CH. analyses
were internally consistent. Since the deepest active U.S. coal mines are approximately 670 meters (m) deep, the
11 data sets which exceed this depth were eliminated. A total of 137 data sets were finally included In the
'coalbed database. " " '"*"" "":- '''' ''"'*-' ->-- ---'*' -''
Table 1 describes the final coalbed database used to estimate coalbed CH4 regression equations. As the table
shows, the database represents most major coal producing regions in the United States; there are coal samples
from eight states and 14 coal seams. A wide range of coal ranks is included, as evidenced by the range of heating
values covered: 22,078 to 38,555 J/g (moisture- and ash-free basis). Although no overriding bias is apparent in
the database, it is noted that the Illinois Basin coals are not well represented.
Coalbed CH4 contents in both databases were measured using the Direct Method adopted by the U.S. Bureau
of Mines and were reported in three components of the total (Diamond and Levine, 1981):
]) desorbed gas - the amount of gas released from a coal sample placed in a sealed canister at
atmospheric pressure. This gas may take from a few hours to several months to desorb.
2) lost gas - the amount determined by extrapolation of the desorption curve back to time zero,
which accounts for gas lost to the atmosphere after the coal is sampled and before it is placed
in the canister.... . ,._..,, ..
3) .residual gas - the amount of gas released when the coal is ground after the desorption process
is completed. . .....
-------
TABLE I. CHARACTERISTICS OF THE COALBED DATABASE
Number of Coal Samples = 137
Number of Seams = 14
* Range of Coal Characteristics
Moisture: 0.3 3.5 %
Heating Value: 22,078 - 38,555 J/g
Seam Depth: 23 - 546 m
Fuel Ratio:* 0.67 5.46
Geographic Coverage
State
Alabama
Colorado
Illinois
Ohio
Pennsylvania
Utah
Washington
West Virginia
Total
Number of Data Sets
26
29
1
3
33
1
2
J2
137
fixed carbon/volatile miner.
At this point, decisions were made regarding the form of the dependent variable and the terms to include as
independent variables in the regression equations. Lost gas, because of the varying amounts of time which may
elapse before the sample is placed in a container, and because different methods of extrapolation are used, may
introduce uncertainty into the total gas analysis (Kim, 1977; Rightmire a al., 1984). Since it would be
undesirable to introduce this additional uncertainty into the regression analyses, it was concluded that values for
lost gas would be removed prior to regression analyses and would be factored back in after the analyses were
complete. Based on the literature cited earlier, the principal coat characteristics affecting CH« content appeared
to be measures of pressure, rank, and moisture content. Therefore, depth, as a surrogate for pressure, and
moisture content were included as independent variables. Heating value was chosen as an independent variable
since it is known to increase with coal rank. Fuel ratio, the ratio of fixed carbon to volatile matter? was chosen
as the fourth independent variable. It is recognized that fixed carbon increases and volatile matter decreases as
rank increases (Such et a!., 1975).- Therefore, the fuel ratio was considered to be a reasonable surrogate for coal
rank'. _ . .,,. . ... , . ...
The next step was to produce the plots shown in Figures 2.3,4, and 5 of desorbed plus residual CH. versus
each independent variable. The purpose was to verify visually the relationships reported in the literature and to
confirm the validity of the surrogate variables chosen for pressure and rank. The plots were also used to identify
both the nature of the relationships and breaks in the data which might suggest dividing one or more of the
independent variables into separate regimes. The plot in Figure 3 suggests that the moisture relationship is not
-------
_ It
1
u
ft ,
I
I"
A * *
I
A
*
Figure 2. Depth verms CH4 content.
c-J0
* A A
A I T * I
>' it 13 J«
.Figure 3. Moisture versus CH, cootent.
-------
.-. 70
P :
X
u
5
I*
o
o
8
V
O
A
A
*
212X0 «40C 0 MtPC C
Heotinq Vmue (
XUDO MMOO
«j2»oc
figure 4. Heating value versus CH4 content.
c
r
E
I
'T
I
C> 10 1» JO It 9C it «0 «.» t« tt (0
FU«. Katie (rixa Cer&on/veic MoMtr)
Figure 5. Fuel ratio versus CH4 content.
-------
linear. The heating values in Figure 4 suggest thai the data might appropriately be divided into two regimes at
about 34,860 J/g. Regression analyses were performed using breakpoints from 32,536 to 36,022 J/g to determine
the best fit, and a heating value of 34,860 J/g was chosen as the most appropriate breakpoint. Dividing trie data
into two regimes produced better results than treating all of the data together. Multivariate regressions were
performed on all combinations of independent variables for both regimes. The two equations finally selected are
shown in Table 2, along with the supporting R} and Mallows C, statistics. The regression t-statistic for each
equation term is noted in parentheses immediately below the term.
TABLE 2. RESULTS OF REGRESSION ANALYSIS
Regime
HV < 34,680 J/g
HV > 34,680 J/g
where: . HV -
-"is -
M «
FR =
Equation
IS - 0.0159 D + 2.781 (1/MJ) - 2.228
(4.682) (4.332) (-2.659)
IS » 0.0136 D + 0.0015 HV -I- 2.6809 FR - 56.4901
(4.318) (3.587) (6.526) (-3.808)
Heating value (J/g coal [moisture- and ash-free basis])
In-situ residual + desorbed gas (m9 CH
-------
TABLE 3. RATIOS OF LOST TO DESORBED PLUS RESIDUAL CH« BY COAL RANK
ASTM Coal Rank
Lost CH4/{Desorbed + Residual
Anthracite
Low-volatile Bituminous
Medium-volatile Bituminous
High-volatile Bituminous
High-volatile B Bituminous
High-volatile C Bituminous
All Other Types
0.11
0.10
0.20
0.05
0.13
0.10
0.11
' In-situ CH4 content from the coalbed database
Coal production
Gob well emission factors
Table 4 summarizes the characteristics of the database. The database includes 269 mine-specific observations
of actual CH4 emissions from mine shafts. The year for each of these observations is also noted (data from 1970
to 1985 were used). Annual coal production rates for each mine where emissions information exists are included
in the database. In-situ CH, contents were assigned to each mine using the data contained in the coal bed CH,
database described earlier. In most cases, an in-situ value for the same county and seam was assigned to a mine.
For longwall mining methods using gob wells, the use of shaft emissions data alone would understate actual mine
emissions by the amount of gas withdrawn from the gob wells. To correct for this understatement, gob well
emissions data were obtained from Sodt (1991). Using these data, it was assumed that, if longwall mining began
in the year for which shaft emissions data were available, the emissions for that mine were increased by 30
perfrnt. If longwall mining began .in a year prior to the year for which shaft emissions data were available, the
emissions were increased by 60 percent. No factor was applied if longwall mining was not used.
A multivariate regression analysis was performed on the database to examine the effects of coalbed CH*
content and coal production rate on mine emissions. The relationship that best predicts mine CHi emissions is
shown in Table 5 (Rs 0.59). T-statistics are parenthetically noted below each term. Figure 6 is a plot showing
the actual emissions data and the resulting 'best fit' equation for this analysis (see "predicted" line in figure).
The form of the equation used represents the standard slope-intercept form for a linear relationship. This
form was chosen for several reasons. First, examination of the emissions database and results from earlier
research efforts suggest that a linear relationship may exist. In addition, the emissions database and observations
made earlier by Irani ei al. (1972) indicate that emissions can occur at underground mines even ai very low or
zero production rates. Thus, the equation form used should have a non-zero intercept, allowing a value for
emissions to be determined when the production term is zero.
-------
TABLE 4. CHARACTERISTICS OF THE MINE EMISSIONS DATABASE
Number of Observations 269
Years Included in (he Database: 1970-1985
Number of Seams = 7
Range of Mine Characteristics
Methane Emissions:
Coal Production:
0.0028 0.4871 million m3 per day
323 -13,641 tonnes per day
Geographic Coverage
State
Alabama
Illinois
Kentucky
Ohio
Pennsylvania
Virginia
West Virginia
. Total
Number of Data Sets
27
33
5
9
47
29
112
269
TABLE 5. RESULTS OF MINE EMISSIONS REGRESSION ANALYSIS
Mine Emissions - 1.08 x la1 (Cot! Production x In-situ CH4 Content) + 31.44 - 26.76 (Dummy Variable)
- ''< (8.23) .. - («.67) (-4.8S)
where:
Mine Emissions ' total emissions of CH, from the mine sheft and gob wells
IK present) (10* cubic meters per year)
Coal Production ' * - annual production of coel (tonnes of coal per ye*r)
In-situ CH.' - - total CH4 content of the unmined coal (cubic meters CH. per tonne coel)
Dummy Variable - 1 if (Co»l Production x In-situ CH. content) < 7.6 x 10*
- 0 if (Co»l Production « In-situ CH. content! ^ 7.6 x 10*
4. GLOBAL EMISSIONS ESTIMATE
4.1 Underground Mines
A summary of the country-specific data assembled on coal characteristics and coal depth is shown in Table
6. These are representative of the data which were used to calculate coalbed CH, contents using the regression
equations from Table 2. Maximum, minimum, and average values are presented only to offer the reader an idea
of the range of values found and to show where in the range the majority of values reside. These average values
-------
.
Cool Production Oonn«*/y«O' J In Situ (f *
Figure 6. CH, emissions from underground mines.
cannot S>e used to adequately predict coalbed CH, contents on a country-by-counuy basis. Gas content
measurements data were used instead of the regression equation to represent in-situ CH, content for the United
Kingdom (Greedy, 1991).
Table 7 shows the country-level 1989 coal production, the range of calculated coalbed CH, values, and
calculated CH, emissions. The coalbed CH4 values are calculated using the types of data presented in Table 6
and the appropriate regression equation from Table 2. CH, emissions estimates are derived using the coal
production and coalbed CH, contents in the regression equation from Table 5. For West Germany, emissions
data developed by German coal industry analysts were used (Treskow and Fitzner, 1987). The U.K. emissions
may be considered an upper-end estimate since it does not account for the small fraction of CH« that is collected
and used rather than released. When information on coal mine CH, recovery and use was available (i.e.. United
States, Poland, and West Germany), the amount of CH, recovered and used was subtracted from a country's total
emissions.' For example, in Poland it is assumed that 19 percent of the potential CH, emissions is utilized
(Pilcher«o/., 1991).
The countries listed in Table 7 produce BI percent of the world's coal from underground raincs. Thus,
emissions for the remaining 19 percent of underground production were assumed to be equal to 19 percent of total
underground mine emissions. Total global emissions from underground mines are estimated to be 36.0 Tg/year.
4.2 Surface Mines
Very little data exist on which to base estimates of emissions from surface mines. A single emission analysis
has been conducted to date by the EPA at a large Powder River Basin surface mine in Wyoming (Kirchgessner
a of., 1992). Using open-path Fourier transform infrared (FUR) spectroscopy, an emission rate of about 4,814
mVday was determined. Using a single coalbed CH, content for the same county and coal seam, it was estimated
-------
a
i
K^
P
<
PC
S
u .
1
g
D
§5
>
r or COUNT
K
(A
^
s
1
I
J
1
*
Is
i!
2|
B
u
B
I
«
g
ft
«
1
g
i
~
I
1
I
1.
r
i
X
6
r
i
K
g
t
'. '*
*~ j * i
w *
^ i i
b 5 !
!
< f*i i
O . cj oe
* Irt
r*' ri O
IM f») S
i - .-
IE § "5 jS _
«e «
_J _J '
^ ^) i
.
^
« I
\C CM
^& 1*** c*4
H R S
8 S K
1 1 1
... '.
*! " ^
K!^ (fNI tf\
S 1 " 1
«* "i j
~ ! '
1
5 1 .a
1 1 2
*}
ci
\£
S
2
M
S
eo
a
s
-------
ee
£
w
I
I
S
S
c
I
o
ee;
V,
u
at
g
i
8
I
I
Meth
£
I
[
i *
>
O
II
8
CNOO
00 Ov O
CM O
t-. oc
g^oe
w
a.
i
-------
that, at the mine's actual coal production rate of 11.8 million tonnes per year, potential emissions from the mined
coal alone should be 1,008 mVday. This would suggest, as noted by Kissel! et al. (1973) for underground
mines, that actual mine emissions exceed, by a factor of about five in this case, the emissions which would be
expected based on coal production and coalbed CH, content alone.
Rightmire ft al. (1984), in their study of coalbed CH, resources in the United States, report 38 analyses of
shallow coals (104 m deep or less) with CH, contents ranging from 0.03 to 3.6 mVtonne coal. One analysis of
9.6 mVtonne for the Arkoma Basin was not included because it is known to be anomalously high for shallow
coals. Coalbed CH, analyses for shallow coals from other countries are lacking, so this study is temporarily
making the gross assumption that the range of 0.03 to 3.6 mVtonne coal reflects the CH, content range for
shallow coals worldwide. Multiplying the average value for this range (1 mVtonne) by 1987 world surface coaJ
production of about 1.8 X 10* tonnes/year (Boyer et a/., 1990), and expanding the results by a factor of five as
discussed above, produces an estimate of about 6.3 Tg/year. Adjusting this value upward by 10 percent to
represent 1989 coal production yields an estimate of 6.9 Tg/year. As additional surface mine emissions are
sampled under the EPA test program, the factor by which actual surface mine emissions exceed expected
emissions may change, in which case this portion of the emissions estimate will require modification.
4.3 Handling Operations
No data were found on CH, emissions from handling operations. Boyer et a!. (1990) estimate that 25 percent
of the CH, contained in the mined coal is released during post-mining operations. There is no compelling reason
not to follow this precedent for now, therefore coal handling emissions presented in Table 7 were estimated by
assuming that 25 percent of the in-situ CH, content for all coal produced is released in post-mining operations.
5. RESULTS AND DISCUSSION
5.1 Estimates of Coalbed Methane and Mine Emissions
The regression equations developed in this study are generally believed to be satisfactory for predicting CH,
emissions from underground coal mining. Two equations were produced for estimating coalbed CH, contents.
one for coals greater than 34,680 J/g heating value and one for coals equal to or less than 34,680 J/g. The
equations are felt to be of high quality in that the coal characteristics which could be retained in .tb.e equations
at a 95 percent confidence level or better also have a sound geological base for inclusion and produce satisfactory
R7 values. The coal characteristics included follow patterns predicted in the literature cited. Depth or pressure
is a dominant factor in coalbed CH, content because it affects not only the generation of CH, but also its retention
in the reservoir. Depth appears in both equations. Moisture also plays an important role in CH, retention, which
may explain its strong effect on the equation for lower heating value and, presumably, shallower coals. At
shallow depths, where geologic factors may have compromised a reservoir's integrity and the coals are generally
of lower rank, moisture levels are known to strongly affect sorption capacity. The link between heating value
and fuel ratio (fixed carbon/volatile matter) in the equation for higher heating value coals is predictable because
-------
the parameters are process-related. With increasing depth-related pressures, the process of coalification eliminates
volatile matter, which in turn increases the heating value and fuelTatio. These two characteristics define coal
rank, and CH« content is strongly tied to coal rank.
It should be noted that the equations for predicting coalbed CH, are extremely sensitive to coal depth and do
not produce credible results for shallow coals at depths of less than about 90 m. This is not surprising because
very few coalbed CH, analyses for shallow coals were available during the initial development of the regression
equations.
The single regression equation which predicts mine emissions suggests that CH, from an underground coal
mine will be emitted even when coal production is not occurring, as originally noted by Irani et ai. (1972). The
equation predicts that, at zero production, CH, emissions will be approximately 12,80! m'/day. The actual
emissions at zero production would depend on factors such as the size of the mine and the in-situ CH, content
of the coal. This equation also predicts that CH, emissions will increase from this zero production level in
proportion to both the in-situ CH, content and the production rate. The equation has an R' value of 0.59 which
means that almost 60 percent of the variation in CH, emissions from the mines can be explained by the
independent variables. This R' value is satisfactory, given the small number of independent variables and, more
importantly, the difficulty of predicting a complex physical phenomenon with so few variables.
A dummy variable is included in the equation to help explain CH, emissions from coal mines. An
understanding of mining processes suggests that there are technological and geological factors which are not
adequately characterized in the database and which may affect CH, emissions from mines. These factors may
include the extensiveness of the mine workings (i.e., a surrogate for the area of exposed coal surfaces) and the
presence of underlying or adjacent coal seams. In this case, a dummy variable is helpful in determining if there
are differences in CH4 emissions from mines that cannot be adequately explained by the existing data. The
statistical significance of this variable strongly suggests that there are factors which differ between mines with
low and high values of the product of coal production and in-situ CH, content.
The mine emissions regression results were tested to determine if serial correlation of the error terms is
present. This analysis shows thai serial correlation of the error term is present and that there is a relationship
between the year in which the emissions data were gathered and the level of emissions from mines in the data
set used. This trend may be due to effects associated with technological change, such as the increased use of
longwall mining, which has taken place since 1970. It could also be due to a recently observed trend in the
United States toward mining increasingly deeper and gassier coal seams (Grau, 1987). The final regression
equation used was selected from a series of equations which were estimated with the expressed intent of excluding
this U.S.-specific trend. This was done to avoid biasing the global estimate.
The data were also organized by level of coal production in a mine. The regression results show that there
is no serial correlation when data are organized by production level. This suggests that the equation used to
predict CH, emissions does not systematically over- or under-predict emissions from mines with high or low
production levels.
-------
5.2 Global Emissions Estimates .
When country-specific data are entered, the mine emissions equation produces a global estimate of CHi
emissions from underground mines of 36.0 Tg/year for 1989. Gross estimates for surface mine emissions and
coal handling emissions based on minimal data produce average values for 1989 of about 6.9 and 2,7 Tg/year,
respectively. Together, these values produce an estimate of global CH4 emissions from coal mining of 45.6
Tg/year. This represents the upper end of the range cited by Cicerone and Oremland (1988), and about the
middle of the range estimated by Boyera al. (1990). Comparing the results of these studies could be misleading
since this study is unique in basing its estimates on those country-specific properties of coal which are known to
affect coal CH4 contents. To examine the accuracy of this study, it is probably more appropriate to compare the
U.S. estimate from this study to an estimate of U.S. mine emissions for 1985 prepared by Grau (1987), based
on actual mine emission measurements. Grau estimates total emissions from underground bituminous coal mines
to be about 2.3 Tg/year. Expanding this number by 10 percent to simulate a 1989 production level and adding
8 percent to allow for unaccounted for gob well emissions, the estimate becomes 2.7 Tg/yr. Total U.S.
underground mine emissions from this study are 3.5 Tg/yr, making the Grau estimate about 75 percent of the
estimate in this study. Even with these simple comparisons, the estimates are strikingly similar and suggest the
validity of the procedure used in this study.. It should also be noted that the estimated emissions of 3.6 Tg/year
from Polish mines agree well with an independent estimate of 3.3 Tg/year developed from Polish mine
measurements (Pilcher, 1991).
Finally, the sensitivity of the estimates to changes in various coal parameters must be addressed. While
equations of the type used in this study are believed to be capable of producing representative results, their ability
to do so depends on the quality of the country-specific data. To determine the sensitivity of the global emissions
estimates to the parameters of concern, analyses were performed on the three largest foreign coal producers. As
expected, the country-specific estimates are quite sensitive to the depth values. Changing depths by 10 percent
changes the emission estimate by 10 to IS percent. Changing the depth values by 30 percent changes the estimate
by about 40 percent. This emphasizes the need for having representative data on depth. This is especially critical
for China, since only minimal depth data were available for that country and it is estimated to be the largest
source of mine related CH. emissions globally (see Tables 6 and 7).
The sensitivity of the estimates to heating value was a matter of concern because many coals have heating
values near the 34,860 J/g breakpoint, which determines the regression equation to be used for coalbed CH«
determinations (see Table 2). Where all coals within 7 percent of this breakpoint were moved to the other side
of the breakpoint and reanalyzed, emissions estimates changed by only 10 percent, suggesting that errors in the
heating value estimates for countries would not cause large errors in the global estimate. Moisture does not
appear to be of concern, since those countries having a significant number of values near the knee of the curve
in Figure 3 are relatively small producers.
Little can be said at this time about the estimation procedure for surface mines. Shallow coals are
infrequently analyzed for CH< because they are not regarded as an economic CHi resource and because CH, is
" *+
not a safety problem in surface mines. With this general lack of coalbed CH« data, and only a single
-------
\
measurement of surface mine CH« emissions, an admittedly unsophisticated estimation procedure has been
temporarily adopted. The estimate of 6.9 Tg/yr is not strongly supportable but a better estimation procedure
cannot be developed until better data are available.
Data are also lacking for CH« emissions from coal handling operations. However, the CH« desorption rates
for many coals are sufficiently slow that post-mine emissions must certainly be significant and cannot be ignored
here. It has been assumed that 25 percent of the CHi content of mined coals is released after the coal leaves the
° mine. This produces an estimate of global emissions from handling operations of 2.7 Tg/year.
Data for abandoned mines appear to be totally lacking at this time. It is known that some mines may
continue to emit CH, for a period of time after abandonment, and in 1966 the U.S. Bureau of Mines estimated
mat there were over 60,000 inactive and abandoned coal mines in the United States (Scott and Hays, I97S).
While it is clear that this category of mines must make a contribution to coal mine CHt emissions on a global
basis, an estimate of the emissions must be deferred until data become available. The Air and Energy
Engineering Research Laboratory is currently engaged in collecting measurements data for a group of abandoned
coal mines.
6. CONCLUSIONS
The methodology employed in this study relies on documented relationships which exist between mine
emissions, coalbed CH4 content, coal production rate, and other coal properties. A new set of mathematical
equations was developed which quantifies these relationships. In the process of developing these equations,
analyses suggested that the relationships between emissions and coal properties are even more comp^c than
expected and that all causative factors in the relationships have not been fully defined. Several geological and
, technological factors, not currently included in the databases, could potentially influence mine emission rates.
The relationships are of sufficient quality, however, that the emissions estimation procedure for underground
mines is regarded as representative and reliable.* Additional surface and abandoned mine data, and handling
operation data could lead to a revised global estimate at a later date, since the full impact of these categories is
not known. Improved country-specific data could also have a significant impact on the estimate.
The global estimate of 45.6 Tg/yr for CH4 emissions from coal mines is thought to be reasonable. It is the
first attempt to produce this estimate using country-specific coal properties which actually influence production
and retention of CH, by the coal reservoir. It also appears that the methodology developed here is applicable at
the province, basin, and possibly seam levels if a sufficiently detailed database on coal properties and production
is available.
ACKNOWLEDGMENTS
This study was conducted under EPA Contract Number 68-D9-0173 for the Air and Energy Engineering
Research Laboratory, Research Triangle Park, North Carolina. The authors wish to acknowledge the technical
-------
assistance of Sushma Masemore of Science Applications International'Corporation, and Walid Ramadan, Teresa
Lynch, Terri Young, and Terry Wilson of Alliance Technologies Corporation. They also wish to thank Ross
Leadbctter of the University of North Carolina at Chapel Hill, who helped with the statistical analyses. The
authors also gratefully acknowledge the assistance provided by Peet S66t of Northwest Fuel Development, Inc.
REFERENCES
S. Ameri, F.T. Al-Sandoon, and C.W. Byrer (1981), Coalbed Methane Resource Estimate of the Piceance Basin.
DOE/METC/TPR/82-6, U.S. Department of Energy, Morgantown, WV, 44 pp.
R.B. Anderson and L.J.E. Hofer (1965), Activation Energy of Diffusion of Gases into Porous Solids, fuel. 44
(4): 303.
A. Basic and M. Vukic (1989), Dependence of Methane Contents in Brown Coal and Lignite Seams on Depth
of Occurrence and Natural Conditions, in Proceedings of the 23rd International Conference of Safety in Mines
Research Institutes. U.S. Department of the Interior, Bureau of Mines, Washington, DC, pp 282-288.
C.M. Boyer, J.R. Kelafant, V.A. Kuuskraa, K.C.Manger, and D. Kniger (1990), Methane Emissions from Coal
Mining: Issues and Opportunities for Reduction. EPA-400/9-90/008, (U.S. Environmental Protection Agency,
Office of Air and Radiation, Washington, DC), p.3.
R.J. Cicerone and R.S. Oremland (1988), Biogeochemical Aspects of Atmospheric Methane, dfifeaj
Biogeocheniicaj C.yCJC* 2 (4): 299.
D.P. Creedy (1991), An Introduction to Geological Aspects of Methane Occurrence and Control in British Deep
Coal Mines, Quarterly Journal of Engineering Geolopy. Volume 24.
W.P. Diamond and J.R. Levine (1981), Direct Method Determination of the Gas Content of Coal: Procedures
and Results. R] 8S1S, U.S. Department of the Interior, Bureau of Mines, Pittsburgh, PA, 36 pp.
W.P. Diamond, J.C. LaScola, and D.M. Hyman (1986), Results of Direct-Method Determination of the Gas
Content of U.S. Coalbeds. 1C 9067, U.S. Department of the Interior, Bureau of Mines, Pittsburgh, PA, 95 pp.
G.E. Eddy, C.T. Rightmire, and C.W. Byrer (1982), Relationship of Methane Content to Coal Rank and Depth:
Theoretical vs. Observed. SPE/DOE 10800 (Society of Petroleum Engineers, Dallas. TX) pp 117-119.
R.H. Gnu HI (1987), An Overview of Methane Liberations from U.S. Coal Mines in the Last IS Years.
Proceedings from The Third U.S. Mine Ventilation Symposium. University Park. PA. October 12-14.
"'''
M.C. Irani, E.D. Thimons, and T.G. Bobick (1972), Methane Emission from U.S. Coal Mines, a Survey. 1C
8558. U.S. Department of the Interior, Bureau of Mines, Pittsburgh, PA, pp 7-15.
M.C. Irani, E.D. Thimons, and T.G. Bobick (1974), Methane Emission from U.S. Coal Mines, a Survey. 1C
8659, U.S. Department of the Interior, Bureau of Mines, Pittsburgh, PA, 47 pp.
M.C. Irani, E.D. Thimons, and T.G. Bobick (1977), Methane Emission from U.S. Coal Mines, a Survey. 1C
8733, U.S. Department of the Interior, Bureau of Mines, Pittsburgh, PA, 55 pp.
D.C. Jolly, L.H. Morris, and F.B. Hinsely (1968), An Investigation into the Relationship Between the Methane
Sorption Capacity of Coal and Gas Pressure, The Mining Engineer. 127 (94): 539.
-
-------
J.l. Joubcrt, C.T. Griein, and D. Bienstock (1974), Effect of Moisture on the Methane Capacity of American
Coals, Bid- 53 (3): 186.
A.G. Kim (1977), Estimating Methane Content of Bituminous Coalbeds from Adsorption Data, RI 8245, U.S.
Department of the Interior, Bureau of Mines, Pittsburgh, PA, pp 1-22.
D.A. Kirchgessner, S.D. Piccot, and A. Chadha (1992), Estimation of Methane Emissions From a Surface Coal
Mine Using Open-Path FTIK Spectroscopy and Modeling Techniques, Chettjosphere (In press).
F.N. Kissel, C.M. McCuIIoch, and C.H. Elder (1973), The Direct Method of Determining Methane Content of
Coalbeds for Ventilation Design. RI 7767, U.S. Department of the Interior, Bureau of Mines, Pittsburgh, PA,
PP 1-9.
T. Koyama (1963), Gaseous Metabolism in lake Sediment and Paddy Soils and the Production of Atmospheric
Methane and Hydrogen,* Journal of Geophysical Research. 68 (13): 3971.
S.W. Lambert, M.A. Treviti, and P.F. SteidI (1980), Vertical Borehole Desipi and Completion Practices to
Remove Methane Gas from Minable Coalbeds. DOE/CMTC/TR-80/2, U.S. Department of Energy, Washington,
DC, 163 pp.
G. Marland and R.M. Rotty (1984), Carbon Dioxide Emissions from Fossil Fuels: A Procedure for Estimation
and Results for 1950-1982, IcUus. 36B (4): 232.
D.D. Murray (1980), Methane From Coa]beds-A Significant Undeveloped Source of Natural Gas. Colorado
School of Mines Research Institute, Golden, CO, 37 pp.
G.F. Nielson (1977) ed., Keystone Coal Industry Manual. McGraw-Hill, Inc. Annual publication.
R.C. Pitcher, C. Bibler, R. Glicken, L. Machesky, and J. Williams (1991), Assessment of the Potential for
Economic Development and Utilization of Coalbed Methane in Poland. EPA-400/1-91-032, U.S. Environmental
Protection Agency, Office of Air and Radiation, Washington, DC.
C.T. Rightmire, G.E. Eddy, and J.N. Kirr (1984), Coalbed Methane Resources of the United States, AAEP
Studies in Geology Series *\7. (American Association of Petroleum Geologists, Tulsa, OK), 9 pp..,
R.R. Schwarzer and C.W. Byrer (1983), Variation in the Quantity of Methane Adsorbed by Selected Coals §s
a Function of CoaJ Petrology and Coal Chemistry FinaJ Report. DE-AC21-80MC14219, U.S. Department of
Energy, Morgantown, WV, pp 1-6.
_ . ."'....-'".." *r-.
R.L Scott and R.M. Hays (1975), Inactive and Abandoned Underground Mines. EPA-440/9-75-007 (NT1S
PB258263) (U.S. Environmental Protection Agency, Washington, DC), p. 7.
B. Smith and D. Tirpak (1989), The Potential Effects of Global Climate Change on the U.S.: Report to
Congress. EPA/230-05-89-050, (U.S. Environmental Protection Agency, Office of Policy Planning and
Evaluation, Washington, DC), 13 pp.
P. Soot, (1991), Northwest Fuels Development, Inc., Portland, OR, personal communication.
Soviet Geography (1989), (Winston, V.H. and Sons, Silver Springs, MD, and WeJwyn, England), Volume 30.
E. Such. M.-Th. Mackowsky, and M. Teichmuller .(1975). Stach's Textbook of Coal Petrology, 2nd ed
(Gebruder Borntraeger, Berlin, Stuttgart), 38 pp.
C-M. Tremain (1980) The Coalbed Methane Potential of the Raton Mesa Coal Region. Raton Basin. Colorado.
Colorado Geological Survey Open-File Report 80-4, Colorado Geological Survey, Denver, CO, 1980. 48 pp.
-------
A. Treskow and G. Fitzner (1987), Gas Drainage Result in the German Coal Mining Industry in 1985/86,
gBckauf (English translation), Vol 123, Nr 19, pp. 540-542.
D.J. Wuebblcs and I. Edmonds (1991), Primer on Greenhouse Gases. (Lewis Publishers, Inc. Chelsea, MI).
-------
-------
Chemosphere, 26 (1993) 23-44
ESTIMATION OF METHANE EMISSIONS FROM A SURFACE
COAL MINE USING OPEN-PATH FTIR
SPECTROSCOPY AND MODELING TECHNIQUES
David A. Kirchgessner1, Stephen D. Piccot**, AJay Cbadha*
'U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711, USA
'Science Applications International Corporation
Air Pollution Research Programs Branch
3101 Petty Road, Durham, North Carolina 27707, USA
'Alliance Technologies Corporation
100 Europa Drive, Chapel Hill, North Carolina 27514, USA
ABSTRACT
A new measurements methodology has been developed which allows the rapid and efficient measurement of
methane (CH«) emissions from surface coal mines. An initial field trial of this methodology has been completed,
and results from the field trial revealed that emissions from one surface coal mine in the U.S. are estimated to
be 1,735,000 mVyear. The results provide some evidence that CH4 concentrations determined by the FTIR may
be low by 20 to 75 percent but the overall effect of this potential bias on the mine emissions estimate cannot be
adequately quantified. The initial trial demonstrated that the methodology is an applicable and feasible approach
for measuring CH< emissions from very large surface coal mines. It also highlighted several uncertainties and
methodology .questions which if resolved could further improve the performance and reliability of the
methodology.
I. INTRODUCTION
Recent global methane (CH<). emission estimates indicate that coal mining operations may contribute as low
as 25 or as high as 64 teragrams (Tg) of CH, per year (Cicerone and Oremtand, 1988; Boyer et ah, 1990). Thus,
coal mines may account for 7 to 18 percent of the total global anthropogenic emissions burden. For underground
coal mines, CH< emissions measurements data are readily available, thus facilitating the development of global
emissions estimates for this important source. For surface coal mines, emissions measurements data are not
available, and measurements techniques have not been developed and tested.
A measurements methodology has been developed and tested by the authors which involves the use of an
open-path Fourier Transform Infrared (FTIR) spectrometer and Gaussian based plume dispersion modeling
-------
techniques (Piccot et al., 1991). The open-path FT1R spectrometer used here is a remote sensing device which
directs a collimated beam of infrared radiation through the ambient air along a path of several hundred meters.
The beam is reflected back to the FT7R with mirrors and a spectral analysis is performed to determine the
infrared absorbance of individual compounds passing through the beam. Use of more conventional methods, such
as emission isolation flux chambers, was considered for directly measuring the flux of CH, from small sections
(about 0.5 m1) of exposed coal surfaces (Klenbusch, 1986). However, it was concluded that these in-situ devices
could not be safely or effectively used in a surface mine environment. In addition, because of the heterogeneity
of the emissions patterns observed, a prohibitively large and expensive number of measurements would be needed
to obtain a statistically representative set of flux measurements for a large mine. The use of conventional ambient
point sampling techniques to determine mine plume CH, concentrations was also considered and dismissed for
a similar reason. The plumes from surface coal mines can be over 1000 meters across requiring that a substantial
number of point sampling systems be used. In addition, ambient measurements collected remotely do not contain
potential errors which can result from sample line leaks and the loss or production of compounds in sampling
containers.
Initial site surveys were conducted at surface mine sites in the U.S. to identify an initial mine site suitable
for conducting full scale measurements. Six sites were visited, three in the Powder River region of Wyoming
and Montana and three in the Illinois coal basin. Preliminary ambient measurements data were collected during
these visits using a portable flame ionization detector (Fbxboro OVA). Although these measurements provide
only rough approximations, they indicate that surface mines are a very heterogeneous source of CH< emissions
(ambient CH4 concentrations ranged from 3 to over 1000 parts per million) and that disturbed coal areas, such
as the coal rubble produced from blasting operations, may be the most potentially significant CH, source at the
mines. As pan of these initial site surveys, wind flow patterns within the mine area were studied to examine
plume behavior and to assess the viability of measuring plume concentrations using remote sensing techniques.
Several smoke releases were initiated, observed, and filmed at several mines. In most cases, these films revealed
that wind flow patterns within the mine pits can be complex (i.e., air sometimes swirled in a helical pattern down
the mine pit at 90° angles to the direction of the prevailing wind) but that the overall plume had Gaussian
dispersion properties and dispersed in the direction of the prevailing wind once it reached the top of the pit. The
Caballo mine, located in the Powder River region of Wyoming, was selected for conducting full scale
measurements because the site has a straightforward rectangular configuration, the plume demonstrated consistent
Gaussian dispersion characteristics under the meteorological conditions observed at the site, and the concentrations
of CH, measured were moderate. - - .
2. DESCRIPTION OF THE SAMPLING AND ANALYSIS APPROACH
2.1 Introduction and Mathematical Basis
A fundamental goal of the sampling methodology is to obtain an emission rate for total CH« emissions from
a surface mine. The heterogeneity and size of the source called for a creative measurement approach. Since
-------
smoke releases show that emissions from surface mines diffuse out of the pit in the direction of the prevailing
winds, a near-ground-level concentration measurement downwind from the mine is used to estimate a total CH,
emission rate for the mine. A CH* measurement of the cross-wind-integrated concentration of the plume at near-
ground level is made using an open-path FUR sensor. Using this near-ground-level concentration measurement
and a measured background or natural ambient CH* concentration, the total mine release is estimated using an
appropriate plume dispersion model. If site-specific plume dispersion characteristics can be determined, they can
be used in the model to more accurately represent the behavioral characteristics of the plume at a given site.
Using a tracer gas, these site-specific plume characteristics can be estimated as described below.
A tracer gas release can be assumed to be a continuously emitting point source. Based on this assumption
and on the results of the smoke release studies conducted at the CabaJlo and Rawhide mines, standard Gaussian
dispersion equations can be applied. When the standard Gaussian equation is integrated across the y direction
(y is assumed to be in the direction normal to the wind direction) from -» to + «, the following relationship
can be developed (Turner, 1970):
20 e»p f -
(1)
where,
QWI = ground-level cross-wind-integrated concentration (g/mj)
Q = emission rate (g/s)
u » average wind speed (m/s)
a, «= vertical dispersion coefficient (m)
H o effective emission height of plume centerline above ground level (m)
For a ground-level source such as a tracer release at a surface coal mine, H is effectively equal to zero so
the exponent of the expression is equal to 1. Thus, Equation (1) can be simplified to:
Qv*
(2)
Equation (2) can be used to obtain site-specific ol values for a mine if the values of (he remaining unknowns
can be determined. Specifically, 9, can be determined for the plume given (1) a measured tracer gas
concentration (Ccw,) from an FTIR sensor; (2) a measured value of u from a meteorological station located near
the FTIR path; and (3) a known release rate Q from a tracer gas source, such as a metered gas cylinder located
at the mine. To use mis technique to estimate total mine emissions, a number of a, values must be determined
based on tracer gas releases conducted at several different distances upwind from the monitoring path. These
resulting values are used to construct a relationship of at versus distance from the path for the area source. AH
tracer gas releases used to determine this 9, relationship should be conducted as close in time as possible because
atmospheric stability may change,-thus changing the at relationship.
-------
A similar and somewhat simpler technique can also be used to assess plume dispersion characteristics using
fewer tracer gas measurements. Given measured values for the tracer gas release rate Q, tracer release location,
wind speed u, and wind direction, an appropriate area source plume dispersion model can be used to predict Qw,
for the tracer gas plume. The model is run to predict concentrations of the tracer gas at various points along the
FTIR monitoring path. These predicted concentrations are integrated using the trapezoidal rule to calculate a
path-integrated concentration or Ccw, for the FTIR monitoring path. The model is run seven times, once for each
of the seven Pasquill-Gifford (P-G) atmospheric stability classes (Turner, 1970). These varying P-G assumptions,
which incorporate the influence of a,, simulate increasing atmospheric stability and its effect on the dispersion
of the tracer gas plume. Since several model results are produced, a range of Qw, values are predicted under
varying degrees of atmospheric stability. The predicted Qwi value which most closely matches the C^wi measured
by the FTIR is used to define the P-G atmospheric stability class which occurred during the tracer gas monitoring
event. If simultaneous CH, measurements are also collected during this monitoring event, this stability
assumption is applied to the CH« plume. The model is then run assuming a unity emission rate for CHj (i.e.,
a homogeneous release rate of 1 g/m'-sec) and the P-G stability determined as described above. The model is
run to predict concentrations of CH« at various points along the FTIR monitoring path. By again applying the
trapezoidal rule to these predicted point concentrations, a path-integrated concentration or Ccwi for the assumed
homogeneous release is predicted along the FTIR monitoring path. Of course the FTIR is actually measuring
a path-integrated concentration due to a heterogeneous emission release pattern from the coal seam. However,
this measured value is comparable to the concentration determined from the model for an assumed homogeneous
release because the FTIR measurements integrated or 'averaged out' the variable concentrations which exist in
the plume from the mine. . .., .',.-.
The actual CH4 release rate for the mine is then calculated using the simple relationship shown below where
, is the unity emission rate for CH«.
'Concentration -------- 1
This technique is used to estimate CH4 release rates in this study: The Point Area and Line (PAL) source model
is used to predict point concentrations along the measurements path as described above (Petersen and Rumsey,
1987). A non-reactive gas, sulfur hexafiuoride (SFJ, is the tracer gas used. Use of a synthetic .trace gas such
as SF« is important to the determination of plume dispersion characteristics because it is non-reactive, does not
naturally occur, and there is no background concentration to cause potential interferences.
2.2 Experimental Setup and Caballo Mine Site Description .....-
Sampling was conducted using an FTIR sensor to measure concentrations of CH« and SF» in the plume along
a path located 1 meter above the ground. The FTIR sensor, shown in Figure 1, was assembled by MDA
Scientific, Inc. of Norcross, Georgia. In general, the FTIR is a unistatic open-path device which contains a
Bomem interferometer and a liquid nitrogen cooled mercury-cadmium-telluride detector. A modulated beam of
-------
OPTICAL TELESCOPE
FT IB SPECTBOMETEB
/
FIGURE 1. Simplified illustration of the FTTR sensor system.
infrared radiation is generated by a glowbar within the FTIR system and is directed into the atmosphere through
an optical telescope which has a 30.5 cm (12 inch) primary mirror. Triangular apodization is used. The beam
passes through the atmosphere and is reflected back to the detector via a corner cube retroreflector where it is
subjected to absorption analysis to identify and quantify the gases present along the path of the beam. The
divergence of the beam over the path is 2.2 mili-radians. Based on the absorption analysis, a path-integrated
concentration is determined for the gases of interest. The analysis is carried out on a portable personal computer
(PC). In addition to the analytical functions, the PC is used to operate the instrument and is an integral pan of
the FTIR sensor. The spectral resolution of the FTIR is I wave number at standard atmospheric conditions.
Pressure broadening is 1/10 to 2/10 of a wave number for each atmosphere above atmospheric pressure. At the
spectral resolution associated with this instrument, pressure broadening effects are not significant.
The absorbance of CH4 was determined by performing a spectral analysis of the CH, peak located at 2915.8
cm0. The absorbance of SF» was determined by performing a spectral analysis of the SF» peak located at 948.7
cm'1. Both peaks absorb in a region where interferences due to water vapor are minimized. Discrete
measurements were collected by taking 32 scans over a 2.S minute interval. Path-integrated concentrations for
both gases were determined by comparing and analyzing the collected spectra against a 'reference" spectra
determined for both gases using known concentrations at standard atmospheric conditions.
The Caballo mine is owned by Exxon Corporation and operated by The Carter Mining Company. It is
located in Campbell County, Wyoming, approximately 15 kilometers southeast of the town of Gillette. It is in
an area of the northwestern Wyoming Powder River region which has been recognized for containing coalbed
CH4 resources (Rightmire et al., 1984). This mine is about 12 years old and has been operating since the fourth
quarter of 1978. The annual coal production in 1989 was about 14.3 million metric tons of high moisture sub-
bituminous coal. The mine operates primarily on the Smith coal seam and has one active pit that operates 24
hours a day, 7 days a week.
-------
No coal bed CH« content measurements data have been collected at the Caballo mine. Only one sample is
available for the Smith seam in Campbell County and it shows that CH, contents are low: less than 0.0013
cmVmetric ton of coal. Despite this low value, at least one mine in this area may have contributed to a sustained
release of significant quantities of CH* in a nearby residential subdivision. This subdivision has since been
abandoned by most of the residents due to CH, hazards. .
Figure 2 shows a top view of the Caballo mine. The shaded area shows exposed coal-surface areas at. the
mine. The series of five areas to the north and west of the coal seam are referred to as overburden benches.
These benches are where strata overlying the coal seam are placed after removal from the working area just north
of the coal seam. Structurally, they are constructed like plateaus and increase in elevation from bench 1 north
to bench 5. The lowest point on the pit floor is 1,338 meters above mean sea level on the west end of the pit.
The coa! seam is in a line which runs along the southern edge of the pit. The coal seam, on average, rises 21.3
meters above the floor of the pit. The overburden on top of the coal rises an additional 20 to 25 meters above
the pit floor (overall it is about 45 meters from ground level just north of the coal seam to the pit floor).
The initial site survey indicated that the coal seam itself was a primary source of emissions. With winds from
the north the plume from the coal seam was sampled on the southern edge of the mine parallel to the southern
edge of the coal seam. With winds from the south sampling was conducted north of the seam on overburden
bench 1. Because the width of the mine plume is greater than the maximum path length the FTIR sensor can
measure (i.e., the maximum path length claimed in. the manufacturer's literature is 650 meters), the plume was
divided into east and west segments.
Figures 3 and 4 show the sampling configuration on the east and west plume segments when the wind is out
of the north. Examination of data coDected by the OVA revealed that on the west side of the mine the primary
source of emissions was a slow seepage of-gas from all exposed coal surfaces while on the east side of the mine
the primary source was the rapid seepage of gas from the blast area and surrounding coal surfaces that were
fractured by the blasting operations. FTIR path lengths for the east and west segments ranged from about 375
to 525 meters. The segments were measured in close succession. That is, a series of several 2.5 minute
measurements, or interferograms, were obtained on the east segment followed by a series of several 2.5 minute
interferograms on the west segment. Background or ambient CH, concentrations were determined using shorter
path lengths at various locations upwind from the mine. Site-specific wind speed, wind direction, and other data
needed to estimate the total emissions from the mine were collected from a meteorological station erected 3 meters
off the ground on bench 5 as shown in Figures 3 and 4. Two monitoring paths were used for the east plume in
this configuration as Figure 4 shows. The reasons for using two paths are discussed later in the results section.
Calibration cell measurements were conducted at the site in an effort to assess FTIR drift and performance.
This was done by passing the FTIR infrared beam in a closed-path configuration through a 15 centimeter
calibration cell containing a CH« calibration gas certified to be 19,500 ppm (about 2 percent CH,). This
approximates the total optical density anticipated to occur in the long path measurements taken at the site. For
these calibration measurements, the calculated CH, concentration from the FTIR sensor was compared with the
expected 19,500 ppm.
-------
ill
o
1
I
t
I
u
-------
LEGEND
sauwo wicc
OF «ul«
©WIMX.OGI'
SUTION
-MOHlTOBlfC MTM
FIGURE 3. Sampling configuration for the west plume segment with a north wind.
FIGURE 4. Sampling configuration for the east plume segment with a north wind.
On several days, a series of equally spaced point measurements were taken with the OVA along the FT1R
path. The OVA was calibrated using certified calibration gases at two concentrations: 5 and 10 ppm. The
accuracy of the OVA cited by the manufacturer at full range (1 to 1000 ppm) is ± 20 percent. The point
measurements were collected to develop "a second data set for comparison with the FTIR data. Another purpose
-------
3f the OVA measurements was to obtain a series of point samples that could be used to identify and examine the
CH, concentration profile along the path of the plume.
3. RESULTS FROM THE FIELD TRIAL
3.1 Overview .
Measurements activities at the mine began on December 10 and concluded on December IS, 1990. Severe
weather conditions (i.e., heavy snow and high winds) suspended sampling activities on December 14. The
ixtremdy cold temperatures and consistently high winds at the mine complicated measurements activities.
Table 1 presents an overview of the number and. types of measurements data collected at the Caballo mine.
As the table shows, several types of data were collected, including 48 mine plume CHi measurements, 44 mine
plume SF, measurements, 16 background or ambient CH, measurements, and 7 CH» calibration measurements.
Nine CH« and five SF« measurements were eliminated from the data set as a result of FTIR operational problems
or because the FTIR signal was too weak to perform a credible analysis.
i.-
TABLE1. TYPES OF MEASUREMENTS DATA COLLECTED AT THE CABALLO MINE
Type of Measurement
FTIR CH4 Measurements
Mine Plume Concentration
Ambient Background Concentration '
Calibration Concentration ...-.-»,
FTIR SF« Measurements
Plume Concentration
Meteorological Measurements -'*'< "- ' ' -
Wind Speed - "
Wind Direction
Standard Deviation of the Wind Direction
Organic Vapor Analyzer Measurements
Mine Plume Concentration Profiles
Number of Measurements
48
16
7
44
measured daily*
measured daily*
measured daily*
y
*Stmpl*d eoniinuoudr *nd raeordw! S-nvnul* (vtngt v»lu»«
'A Mritt of 10 to 20 point m*Mur*nwnti wtrt m*dt lor «*ch ptofM.
After the sampling trip was completed a preliminary review of the mine plume concentration measurements
revealed that most were unrealistically low.- The FTIR vendor (MDA Scientific) proposed that, for the instrument
used at the mine, interferences from light scattering within the instrument itself were significant due to the long
paths and accompanying weak return signals experienced. For example, scattered light accounted for about 50
-------
XHints of the total 175 counts typically associated with the longest open-path measurements nude at the Caballo
nine. A correction factor formula was developed based on the premise that the internal scattering was diluting
he overall signal and, thus, the detection of ambient CH, and SF, concentrations in the path. This formula was
ised to estimate a concentration correction factor which, when multiplied by the measured concentration from
he FTTR sensor, provided an adjusted CH» or SF* concentration. The correction factor equation used for CH,
vas: . . ' . .' .
Correction Factor
A'
(4)
A «= corrected absorbance of CH. at the 2915.8 cm'1 peak
A1 = measured absorbance of CH* at the 2915.8 cm'1 peak associated with the collected sample
Die corrected absorbance (A) is estimated based on a C05 spectrum observed by the detector. With an ideal
measurement (i.e., the instrument is completely purged with nitrogen), no CO, spectrum will be observed.
However, with very long path measurements such as those taken at the Caballo mine, an absorbance associated
with the residua] CO, within the instrument can be detected. The residua! CO] absorbance observed in the spectra
collected is assumed to be directly proportional to the loss of signal associated with the CH, peak at 2915.8 cm'1.
The following describes the plume concentration measurements data collected at the Caballo mine. Following
this, emissions estimates are presented and discussed.
3.2 Summary and Analysis of CH, Concentration Measurements
The results associated with the four types of CH« measurements collected (FUR ambient or background
measurements, FTTR mine plume measurements. FTTR calibration measurements, and OVA mine plume
measurements) are discussed below.
FTIR Background or Ambient CH1 Concentration Measurements. Sixteen background or ambient
measurements were taken. Three are not considered valid because inspection of the meteorological data after the
field program indicated that emissions from the mine may have entered the FTIR path while these samples were
being taken. This may explain why the concentrations associated with these samples are generally higher than
the other background samples collected.
Figure 5 shows the background concentrations associated with the 13 valid samples collected. The average
concentration from these samples is 1.64 ppm, slightly lower than the global average of about 1.7 ppm. As the
figure shows, ambient concentration measurements appear to be within the same range with the exception of the
measurements taken on day 6. Measurements taken on days 2, 3. and 4 range from about 1.8 to 2.3 ppm, while
measurements taken on day 6 are consistently below 1.6 ppm. On day 6, the measurements were taken in the
afternoon, when winds were much stronger and more variable than at any other time when background
measurements were collected. In addition, by the time these measurements were collected some etching of the
instrument's main hygroscopic optical window had occurred due to moisture condensation. The effect of this
-------
r\
T-
z
c
JCFNT'HAT
-o
U
c
. u-
d
c\
--.
-i
"
7 ..V.
<
t£
r\
u
L
U
I.
c
H
u
C
C
-..
' - V
»J
C\
'
;".
...
-v ..
f\
a. 0
..
1 f\
l
CUK
-'
a
r
303
A
l
t/J
*
g
1-
1
V
r
+. ..
) ir
r
NO 11
,,
3 ^
r
veUi'
r o
.T
JEDN
. -";
4 *
DO 3
- .
a
c
3 U
3 _ C
.3W
i»"
3 ^!
3' C
r r
3 C
* c
3
03
ID
1C
OJ
(D
*
V V
O
a
i/)
n
cv
rv
j
.
,1
!
T-
!
t
Id
g
I
-------
etching on the measurements is not known but it likely reduced the incoming and outgoing signals.
Fluctuations in the background measurements can be significant. Although concentrations on day 3 are very
consistent, die maximum variation in the concentrations for days 2 and 4 is on the order of 0.5 ppm. This
variation may have occurred because measurements were taken close to the coal seam on day 4 (i.e., mine
emissions could have potentially entered the path on occasion). Even so, variations seen in the ambient
concentrations are almost as great as the apparent CH, contribution from the mine for at least some of the
measurements. Because the emission flux from the mine is estimated by -subtracting ambient CH, from the
concentrations measured in the mine plume, it is mandatory that a representative ambient value be used to reduce
uncertainty in the emissions estimates.
I^UK Mine Plume CH> Concentration MHSIirclTHlllS Figure 6 shows the CH4 plume concentrations
measured on December 12, 1990, while Figure 7 presents the data collected on December 15,1990. For these
two days, hut monitoring path lengths (i.e., the distance between the FTIR sensor and the retroreflector) ranged
from 375 to 525 meters.
As Figure 7 shows, only one east leg measurement is reported for December 15. This is because there were
difficulties in analyzing the spectral data collected for the east leg. For most east leg measurements, signal
strengths were low, and the interferograms exhibited low signal-to-noise ratios. This is likely a result of the long
path length associated with the east leg measurements (about 460 meters) and high wind velocities vibrating the
retroreflector and FTIR on December 15. All west leg measurements were collected on December 15 with a path
length of 375 meters and none contained unacceptably low signal-to-noise ratios. In addition, the west teg
retroreflector was more sheltered from the strong winds than the east leg retroreflector.
The data in Figure 6 show that on December 12 measured path-integrated concentrations in the plume varied
from a low of just under 2.0 ppm to a high of about 4.3 ppm. Figure 7 shows that on December 15
concentrations varied from a low of 1.6 ppm to a high of about 2.8 ppm. Three general trends have been
observed in the December 12 and 15 data as outlined below,
* bverall Mine Flume. Ground-level CH, concentrations may increase by 20 to 50 percent as ambient
air passes over the Caballo mine. Mine emission rate calculations discussed in the next section provide
a more representative measure of the emissions contribution from the mine.
* Effects of Measurements Configuration. Figure 6 shows that measured concentrations significantly
increased when the east path retroreflector was relocated in an effort to more fully encompass the plume
(see the sampling configuration in Figure 4). When the retroreflector was moved, the path-integrated
concentrations increased dramatically by a factor of 1.8 on average.
Effects of Coal Blasting. Concentrations in the plume from the east side of the mine were higher than
from the west side. The east side of the mine contained the blast area which, based on initial OVA
measurements, was identified as a significant source of emissions at the mine.
Calibration MP**"**1"6"15- Although the FTIR cannot be 'tuned' in the field based on the results of
a calibration measurement, these measurements can be used to examine instrument performance and drift relative
to a known standard: The results show that the percent difference between the measured concentration of the
-------
3 -
-
11
7*
ta
s
IIP
*>T
FIBE
>
it
»
V
>
1.
«C
*-«ET«O
V
1
2P -
9:i
S
BP
r»
4P
t->
IP
1-
lOVft
71
»
3'
>-»
Jf
Til*
6»ST LEG CUT* B%a VEST LEE 0»T»
FIGURE 6. Summary of mine plume CH4 measurements collected on 12/12/90.
s <
1
IB
I '*
* * *
I . a. a
I
5 "
a , .
s
ID . 9
| ...
5
0 4
0 9
.
f
F
t
-
"
11 S
5 -
Ai.
\
12
n
DP
i
3
,j
,.
"
-
SP
1 0
n
IP
1 ft
W>
1 i
I
I
I
I
1
1
1
10
1
1
|
1-4
2P
.,
7P
\
i o
EP
'
tine
LEG 0»T»
FIGURE 7. Summary of mine plume CR4 measurements collected on 12/15/90.
-------
calibration gas in the calibration cell and the known concentration generally ranges between -5 and -25 percent.
In addition, this percent difference appeared to progressively increase between the first and final sampling day.
On average, the FTIR measurements appear to consistently underestimate actual CH, calibration concentrations
by about -20 percent. This suggests that some systematic error in the instrument's performance was occurring.
As with me OVA measurements described below, these results provide strong evidence that concentrations
determined from the particular FTIR used at. the Caballo mine may be low.
OVA Poiqt MegiiMffl|« Point samples were collected with an OVA on several different days after
calibrating the OVA as described earlier. On December 12. a series of point samples was collected at 30-meter
(100-foot) intervals along the FUR path. These measurements were conducted simultaneously with a series of
five FTIR measurements collected during the period from 3:10 to 3:40 pm. As expected, the point samples
collected show that plume concentrations significantly increase at locations immediately downwind from the coal
blast area. Concentrations outside the blast area plume ranged from 3.0 to 7.0 ppm while concentrations within
the blast plume were generally between 10 and 20 ppm. On occasion, the concentration within the blast plume
was as high as 30 to 50 ppm.
The OVA data collected between 3:10 and 3:40 pm were used to simulate a path-integrated measurement for
the east leg plume. The trapezoidal rule was applied to the point measurements to determine a path-integrated
#.
concentration. The value estimated was 7.0 ppm. This is higher than the average of the east leg FTIR path-
integrated measurements taken during the same time: 4.0 ppm. This finding is generally consistent with the
earlier finding that the FTIR underestimated calibration gas concentrations. However, it should be noted that the
OVA measures total hydrocarbons and it is possible that some emissions from a diesel truck operating near the
coal loading area contributed somewhat to the OVA measurement. The degree of the potential contributiorfcom
a diesel truck has not been quantified but it is doubtful that, if truck emissions did cause an interference, their
contribution accounts for all of the discrepancy seen between the OVA and FTIR measurements. An evaluation
was conducted to identify other potential sources of interference (e.g., natural background hydrocarbon
emissions): none were found to be significant.
3.3 Summary and Analysts of CH. Emission Rate Estimates
Tracer Gas Results and Atmospheric Stability Determination. As discussed earlier, CH< emission rates are
estimated based on the plume dispersion characteristics determined from.the SF» plume measurements.
Specifically, the SF« measurements are used to identify the Pasquill-Gifford atmospheric stability class"which most
likely occurred during a specific tracer gas monitoring event. If simultaneous CH« measurements are also
collected during this tracer gas monitoring event, the same stability class assumption is applied to the CH« plume.
For several of the CH. concentration measurements collected, a simultaneous or near simultaneous' SF,
measurement was not available. In many of these cases, SF« was released during the monitoring event but the
f
resulting measurement could not be used because either low signaJ-to-noise ratios corrupted the measurement,
N*w tfcnuiumau* n d««iMd M *fl SF,
» 4 to 6 frinutM of CH, RWMUWMRI.
-------
or an inadequate portion of the SF* plume was captured in the monitoring path (he., less than 70 percent was
captured). The data developed in this study suggest that atmospheric stability at the mine could change
significantly over a period of several hours. For this reason, CH, concentration measurements where
simultaneous or near simultaneous SFt measurements were unavailable for determining stability were not used
to estimate CH« emission rates.
Table 2 presents measured path-integrated SF, concentrations for the 10 SF, monitoring events which were
not adversely affected by low signal-to-noise ratios or partial plume capture. It also presents a series of path-
integrated concentrations predicted by the PAL model under different assumptions for the Pasquill-Gifford
atmospheric stability class. The percent difference between the measured concentration and the predicted
concentration is shown. Percent difference was calculated as shown below.
Percent difference = 100"[(measured predicted)/predicted]
(5)
The predicted concentration which most closely matches the measured concentration is used to define the
atmospheric stability for each monitoring event in Table 2. In some cases, the measured value is about half way
between two predicted values. In these cases it is assumed that a stability which is half way between the predicted
values actually occurred (see measurements WCOAL 20 and WCOAL 24). Otherwise, the closest stability is
.used.
Prior to conducting field measurements at the mine it was anticipated that atmospheric stability class C or
D would be encountered during the cool months in Wyoming. This was determined using conventional techniques
to select a stability class based on such factors as the time of year, the solar angle, the anticipated cloud cover,
and the anticipated wind speed (Turner, 1970). The data in Table 2 show that for 6 of the 10 SF» measurements
stability A or B is indicated. For the remaining measurements stability C is the most prevalent. These results
indicate that the atmospheric layer containing the mine plume was more unstable than expected (i.e., plume
dispersion occurred more rapidly than anticipated). Although this was not an expected outcome of the study, it
is possible that the dispersion properties of the near-ground atmospheric layer are greater than those of the overall
atmosphere. Surface features at ground level can introduce turbulence and mixing in the air stream which do not
occur in the overall atmosphere (i.e., overburden high walls, coal high walls, and din piles were located between
the source and the FTIR sensor). In addition, if the surface temperature is significantly higher than the ambient
air temperature, some convective mixing may occur near ground level. Since most of the surface over which the
plume passed was black coal (an efficient absorber of solar radiation) it is conceivable that some thermal mixing
occurred - ...-,' , _._... ...... '.
The SF« results also.suggest that atmospheric stability changed throughout the day over a wide range.
Because of this potential for .significant variation, the use of monitoring event-specific stability classes for
estimating CH< emissions from the mine appears to be warranted.
-------
TABLE 2. COMPARISON OF PREDICTED AND MEASURED SF. CONCENTRATIONS
Assumed
Stability
ECOAL15
Stability A
Stability B
Stability C
Stability D
ECOAL17
Stability A
Stability B
Stability C
Stability D
WCOAL20
Stability A
Stability B
Stability C
Stability D
WCOAL21 ^
Stability A
G»«hili*«* D ' i
aiaoiHiy 0
Stability C -*-..
Stability D ~
WCOAL22
Stability A
Stability B
Stability C
Stability D
WCOAL23
Stability A
Stability B
Stability C
Stability D
Predicted
Path-average
Concentration (ppm)
0.0046
0.0096
0.0151
0.0261
0.0035
0.0080
0.0138
0.0254
0.0069
0.0103
0.0149
0.0241
-. . " '-
0.0064
i- » - * A JY)O£,
'' U.UWD
0.0138
- i- 0.0223
- . - -
0.0061
0.0096
0.0140
0.0229
0.0061
0.0096
0.0140
0.0228
Measured
Path-average
Concentration (ppm)
0.0036
0.0042
0.0181
0.0157
,,
0.0103
0.0100
, _
Percent
Difference
-22
-63
-76
-86
20
-48
-70
-83
162
76
21
-25
145
j*j
64
14
-30
69
7
-26
-55
64
4
-29
-56
(Continued)
-------
\
TABLE 2. COMPARISON OF PREDICTED AND MEASURED SF« CONCENTRATIONS
.-,-,.,-: ..... ....... .--.,-;...;:,;,....,_ (Continued). . ^
Assumed
Stability
WCOAL24
Stability A
Stability B
Stability C
Stability D
WCOAL25
Stability A
Stability B
Stability C
Stability D
ECOAL26
Stability A
Stability B
Stability C
Stability D
ECOAL3Q '""
Stability A '
Stability B
Stability C
'Stability D ~"
Predicted
Path-average
ConcentratioB (ppm)
0.0064
0.0096
0.0143
0.0232
0.0067
0.0104
0.0152
0.0247
0.0037
0.0053
. ,
0.0079
6.0131
. i - - «« ... -i ; * j . \
., ,*^;:.. '- - - ' -, .
.-.: ...,,." ..'-
Percent
Difference
19
-21
-47
-67
27
-18
-44
-66
186
100
34
-19
97
40
' "' '"' - -8
-44
-^'^J*j»::&^
-------
Methane Results and Mine Emission Estimates. As stated in the previous section, CH« concentration
measurements where simultaneous or near simultaneous SF« measurements were not available are not used to
estimate CH» emission rates. In addition, CH< concentration measurements which were collected on December
15 were not used because the background CH, measured on that day was suspect (see earlier CH, concentration
results discussion). Without a representative background CH, concentration, it is not possible to distinguish
between the contribution from background CH, and. the mine's contribution to the plume. Finally, five CH,
measurements were not included because low signal-to-noise ratios corrupted the measurement.
The data in Table 2 were used to estimate CH4 emission rates for the mine as discussed earlier. The results
are shown in Table 3. Monitoring event-specific stability classes determined from Table 2 were used to assign
a stability class to the following CH, measurements: ECOAL15, ECOAL17, ECOAL26, ECOAL30, and
WCOAL20 through WCOAL25. For ECOAL16, stability A was assumed because measurements taken
immediately before (ECOAL1S) and after (ECOAL17) indicated stability A was likely occurring. For ECOAL18
and ECOAL19, stability B was used because measurements taken immediately before (ECOAL15 and ECOAL17)
and soon after (WCOAL20 and WCOAL21) indicated the atmospheric stability was changing from A to C. A
similar approach was used to select stability C for ECOAL27 to ECOAL29.
Estimated emission rates for die mine range from 0.70 to 6.31 mVmin for the east side, and 6.77 to 6.24
mVmin for the west side. As expected, emissions from the east side are somewhat higher than those from the
west side because the coal blast area was producing emissions on the east side of the mine at the time these
measurements were taken! The average east side emission rate was 1.85 mVmin and the average west side
emission rate was 1.45 m'/min.
Based on these two average values the total annual emissions from the Caballo mine were estimated to be
about 1,735,000 mVyear (61.3 million ftVyr). This emission rate is greater than the emission rates associated
with 62 percent of the underground mines in the United States (Gnu, 1987). However, it is lower than the
emission rates associated with the most significant emitting underground mines. In addition, the specific
emissions (emissions per ton of coal mined) are about a factor of 10 lower than many underground mines due
to the low relative methane contents associated with surface minable coals. Great caution should be exercised
in making these types of comparisons. Measurements at only one surface mine have been collected, and there
is a great potential for significant variation to occur between individual mines (as is the case with underground
coal mines).
-.;-
4.0 SUMMARY AND CONCLUSIONS
The results presented here suggest that potentially significant quantities of CHi are emitted from the Caballo
mine, and that the coal blasting operation is a significant contributor to total mine emissions. Based on the SF«
and CH, concentration measurements collected, it is estimated that the mine emits about 1,735,000 m3 of CH,
per year. The results also provide evidence that CH« concentrations determined by the FTIR are low (particularly
those associated with the long-path mine plume measurements). Although the exact magnitude of this bias is not
-------
TABLE 3. ESTIMATED METHANE EMISSIONS FOR SPECIFIC SAMPLING EVENTS
AT THE CABALLO MINE .
Measurement
ECOAL15
ECOAL16
ECOAL17
ECOAL18
ECOAL19
ECOAL26
ECOAL27
ECQAL28
ECOAL29
ECOAL30
Average Values
. . : .
WCOAL20 - -' * !--:T"--~-
WCOAL21 - *:.-
WCOAL22
WCOAL23r-! ?:£>*=}&* ''0.00'
" .5.87
0.00
17.31
Emissions
(mVmln)
1.59
0.64
0.99
5.72
1.00
1.06
1.76
1.74
1.84
2.21
1.85
--.-. . -
'-';*-* -<- -- 5.66'-
0.70
., , ,. - i.gg
?.*,'.. v«~-'.v 0.00
:,. , Q49
0.00
1.45
Stability Class
A
A
A
B
B
D
C
C
C
C
-
.
> ;-% C-D "
'..... c
- - - B '
D *
D
A-B
B
-
.""?-
v*~&^^
^s^ivi
-------
mown, it could be as low as -20 percent (based on the calibration tube measurements) or as high as -75 percent
$ased on the OVA measurements). Unfortunately, these data cannot be used alone to assess the accuracy of the
overall mine emissions estimate. This is because the accuracy associated with the SF, measurements made by
he FT1R is not known so the overall effect on the estimation of stability classes cannot be determined.
Key objectives of this field trial were to: (1) demonstrate that the methodology developed here can be applied
it large scale surface coal mines,- (2) identify and assess uncertainties in the methodology, (3) identify
.tiethodology validation needs, and (4) develop necessary methodology modifications. The results of this field
rial revealed that the methodology is an applicable and feasible approach for measuring CH, emissions from very
large surface coal mines. It also highlighted several uncertainties, methodology questions, and areas where
.mprovements could be made. These are outlined below.
How accurately does the FHR spectrometer measure path-integrated CHi and SF«
concentrations and what is the effect of monitoring path length on the accuracy of the
measured concentrations? Field validation studies are needed to determine detection limits
and to identify the maximum FTIR path length which can be used and still produce
acceptable results.
What is the effect of meteorological station location and data averaging times on estimated
CH< emissions rates. It may be more appropriate to locate the meteorological station at the
FTIR path and to use averaging periods which are much shorter (e.g., 1 minute) than those
used here.
What is the reliability of the measurements methodology under partial plume capture
conditions and under source conditions where the 'emissions release is strongly
heterogeneous? '
Variability in the background or ambient CR, concentration can be on die same order as the
CH« contribution from the mine. In future field investigations, more background
measurements are needed to reduce this uncertainty. Alternatively, simultaneous mine plume
and background sampling could be conducted.
Changes in the wind direction occurred throughout the day which complicated measurements
activities and which invalidated some of the measurements taken (i.e., only a small portion
of the plume passed through the path on occasion). Several procedural changes would
significantly reduce the number of invalid samples occurring as a result of meteorological
variation. These include improved FTIR and retroreflcctor mobility, on-site/real time plume
mapping, and the meteorological modifications described above.
The reader is cautioned not to use the data presented here as being representative of all surface mining
operations. Since these measurements were completed, initial site surveys were conducted at other mines located
in the vicinity of the Caballo mine (i.e., within 3 to 8 kilometers). Methane concentrations measured with the
OVA at one small mine were an order of magnitude below the concentrations measured during the initial survey
conducted at Caballo. Conversely, a large surface mine which also operates near Caballo has been recognized
as likely producing a sustained release of significant quantities of CH,. Interviews with former employees of this
mine suggest that CHi liberations from within the site are significantly greater that those produced at Caballo.
-------
These initial site surveys and interviews indicate that mines within the same general vicinity may have
dramatically different emission rates/ This phenomenon is consistent with the known emission characteristics of
underground mines (Gnu, 1987). As a result, it has been concluded that additional measurements will be needed
before a representative set of surface mine emission factors can be developed.
ACKNOWLEDGMENTS
This study was performed in partial fulfillment of Contract No. 68-D9-0173, Work Assignment No. 1/102
by Alliance Technologies Corporation, 100 Europa Drive, Chapel Hill, NC 27514, tinder sponsorship of the U.S.
Environmental Protection Agency. The authors wish to acknowledge the guidance and assistance of Julian Jones
of the U.S. EPA/AEERL and David Mobley, formerly of the U.S. EPA/AEERL, and now in the Office of Air
Quality Planning and Standards. Also acknowledged is the technical assistance and overall guidance of Peet Soot
of Northwest Fuel Development Corporation. The authors also wish to thank Bill Morrell and the staff at the
Caballo Mine. Without their assistance, guidance, and dedication this research would not have been possible.
Finally, the authors wish to thank the staff of MDA Scientific for their help in examining field results and in
coping with the instrument related difficulties encountered.
REFERENCES
C.M. Boyer, J. Kelafant, V. Kuuskraa and K. Manger (1990), Methane Emissions From Coal Mining: Issues
and Opportunities for Reduction. EPA-400-9-90-008 (Office of Air and Radiation, U.S. Environmental Protection
Agency). _ .; .. i; ':__ ' .. -. .. . %
R.J. Cicerone and R.S. Orcmland (1988), Biogeochemical Aspects of Atmospheric Methane, Qlobjl
i 2 (4):299...-.
CoaJbed Methane Resources of the United Slates, (1984) AAPG Studies in Geolopv Series No. 17. C.T.
Riehtmire, O.E. Eddy and J.N. Kirr, eds., (American Association of Petroleum Geologists, Tulsa, OK).
R. H. Grau ID (1987), An Overview of Methane Liberations From U.S. Coal Mines in The Last IS Years,
Proceedings of the Third U.S Mine Ventilation SymposiumUniversity Park. PA. October 12-14.
M.R. Klenbusch (1986), MC85UrCmen^ °f Gaseous Emission Rates from I rand Surfaces Using an Emission
Isolation Flux Chamber - User's Guide. EPA-600/8-86-008 (NTIS PB86-22316I), (Environmental Monitoring
Systems Laboratory, U.S. Environmental Protection Agency, Las Vegas, NV).
T. Koyama (1963), Gaseous Metabolism in Lake Sediment and Paddy Soils and the Production of Atmospheric
Methane and Hydrogen," Journal of Geophysical Research. 68(13):3971.
W.B. Petersen and E.D. Rumsey (1987), User's Guide for PAL 2.0 - A Gaussian Plume Algorithm for Point.
Area, and Line Sources. EPA-600/8-87-009 (NTIS PB87-168787), (Atmospheric Sciences Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park, NC). .
S.D. Piccot, A. Chadha. D.A. Kirchgessner. R. Kagann, M. Czemiawski and T. Minnich (1991), Measurement
of Methane Emissions in the Plume of a Large Surface Coal Mine Using Open-Path FTIR Spectroscopy.
Proceedings of the 1991 Air and Waste Management Association Conference in Vancouver B C
--",*." '-T;-~ J
-------
D. Bruce TUmer (1970), Workbook of Atmospheric Dispersion Estimates. EPA Report AP-26 (NTIS PB191482)
(U.S. Environmental Protection Agency, Research Triangle Park, NC).
------- |