United States T'P A-fifiO/R-Q9-n*}7
Environmental Protection *** t>UU/K Ut UJ /
Aflencv March 1992
Research and
Development
DEVELOPMENT OF AN
EMPIRICAL MODEL OF
METHANE EMISSIONS
FROM LANDFILLS
Prepared for
Office of Air and Radiation and
Office of Policy, Planning and Evaluation
Prepared by
Air and Energy Engineering Research
Laboratory
Research Triangle Park NC 27711
-------
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.
-------
EPA-600/R-92-037
March 1992
DEVELOPMENT OF AN EMPIRICAL MODEL
OF METHANE EMISSIONS FROM LANDFILLS
Final Report
By:
Rebecca L Peer, David L Epperson,
Darcy L Campbell, and Patricia von Brook
Radian Corporation
3200 E. Chapel HBI Road/Nelson Highway
Post Office Box 13000
Research Triangle Park, North Carolina 27709
EPA Contract No. 68-D9-0054
Work Assignment No. 31
EPA Project Officer Susan A. Thomeloe
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory (MD-63)
Research Triangle Park, North Carolina 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C. 20460
-------
ABSTRACT
The U.S. Environmental Protection Agency's (EPA's) Air and Energy Engineering Research
Laboratory (AEERL) began a research program in 1990 with the goal of improving global landfill
methane (CH4) emissions estimates. Part of this program is a field study to gather information that can
be used to develop an empirical model of CH4 emissions. The field study is the subject of this report.
Twenty-one U.S. landfills with gas recovery systems were included in the study. Site-specific
information Includes average CH4 recovery rate, landfill size, tons of refuse (refuse mass), average age
of the refuse, and climate. A correlation analysis showed that refuse mass was positively linearly
correlated with landfill depth, volume, area, and well depth. Regression of the CH4 recovery rate on
depth, refuse mass, and volume was significant, but depth was the best predictive variable (R2 - 0.53).
Refuse mass was nearly as good (R2 « 0.50). None of the climate variables-precipitation, average
temperature, dewpoint-were correlated with the 0(4 recovery rate or with CH4 recovery per metric ton
(Mg) of refuse. A large amount of the variabPlty In CH4 recovery remains unexplained, and Is likely due
to between-slte differences In landfill construction, operation, and refuse composition. A model for
global landfill emissions estimation is proposed.
-------
CONTENTS
Abstract ii
Rgures Iv
Tables v
1. Introduction 1
Background 1
Objectives 3
2. Landfill Characteristics 5
3. Data Reduction and Model Development 20
Data Quality 20
Data Reduction 21
Statistical Analyses 30
Model Development 32
4. Results and Discussion 43
Regression Models 43
Effect of Data Quality on Analysis 48
Comparison to Landfill Model 49
5. Summary and Conclusions 52
Evaluation of the Methodology 52
A Global Model 54
References 56
Appendices
A. Landfill Site Visit Reports 58
B. Metric/U.S. Equivalent Conversion Chart 90
C. Data Summary Tables in U.S. Equivalents 92
D. Landfill Data Sheet 96
Hi
-------
FIGURES
Number
1 Methane Recovery versus Average Landfill Depth 34
2 Methane Recovery versus Refuse Mass 35
3 Methane Recovery versus Landfill Volume 36
4 Methane Recovery Rate per Unit Mass versus Refuse Age 37
5 Methane Recovery per Unit Mass versus Normal Annual Dewpoint Temperature 38
6 Methane Recovery Rate per Unit Mass versus Annual Average Rainfall 39
7 Climatic Normals for the Landfills 40
8 Methane Recovery Regression with 95% Confidence Interval
of Regression Coefficient 46
)v
-------
TABLES
Number
1 Summary of Gas Row Data Obtained from the Landfills .......................... 22
2 Summary Statistics for Methane Recovery Rates
Grouped by Measurement Type ......................................... 23
3 Summary of Data Used to Estimate Landfill Parameters .......................... 25
4 Summary of Landfill Parameters Used in the Statistical Analyses ............... .... 26
5 Summary of Climatic Data for the Landfills ................................... 29
6 Correlation Coefficients of Methane Row Variables wtth Landfill Parameters
and Summarized Weather Data ......................................... 31
7 Correlation Coefficients Between Landfill Parameters and Landfill Refuse in Place ....... 33
8 Correlation Coefficients Between Weather Variables ............................. 33
9 Landfill Regression Summary ............................................. 44
10 Significance of Refuse Age ............................................... 47
11 Actual and Predicted Methane Values ....................................... 50
12 Comparison of Model Performances ..... ................................... 51
C-1 Comparison of Summary Statistics for Methane Row Rates at the Landfills ............ 93
O2 Summary of Landfill Parameters Used in the Statistical Analyses (U.S. Equivalents) ...... 94
C-3 Summary of Climatic Data for the Landfills (U.S. Equivalents) ...................... 95
-------
SECTION 1
INTRODUCTION
BACKGROUND
The U.S. Environmental Protection Agency's (EPA's) Air and Energy Engineering Research
Laboratory (AEERL) began a research program in 1990 with the goal of improving global landfill
methane (CH4) emissions estimates (Thomeloe and Peer, 1990). Methane is a greenhouse gas of
particular concern because its radiative-forcing potential is thought to be much greater than that of
carbon dioxide (C02) (Shine et al., 1990). Considerable uncertainty remains about the quantitative
emissions of CH4 from each of its known sources (KhalB and Rasmussen. 1990).
Municipal solid waste (MSW) includes a considerable amount of degradable organic carbon,
moisture, and a variety of bacterial species. These bacteria are the primary agents of the decomposition
of organic wastes, producing COa, CH4, and trace amounts of other gases. In the presence of oxygen,
aerobic bacteria are active and the primary product is CO£. However, in sanitary landfills, the oxygen is
quickly used up and oxygen infiltration is usually very low, in an oxygen-poor environment, anaerobic
bacteria are the primary decomposers and methane as well as CO2 are produced. The MSW deposited
in both sanitary landfills and open dumps is a large potential source of methane.
Recent CH4 budgets attribute approximately 11 percent (Cicerone and Oremland. 1988) and
17 percent (Bingemer and Crutzen, 1987) of global anthropogenic Cfy to landfill emissions. However.
these estimates are based on limited data and assume near optimal conditions for anaerobiosis. The
optimal moisture, temperature, pH, and substrate conditions are not likely to be found in sanitary landfills
in the United States and Europe, let alone in the open dumps typical of the less-developed countries.
One objective of AEERL's research program is to develop a database and a methodology that can be
used to reduce the uncertainty of CH4 emission estimates from landfills on a global basis (Thomeloe
and Peer, 1991).
The research program began with a review of currently avafiable models and data (Peer et al.,
1991). Several theoretical models and laboratory experiments used to estimate CH4 production in
individual landfills were identified. However, adapting these methodologies for global estimates posed
several problems, the worst being that site-specific data would be needed for every country. The few
1
-------
global emissions methodologies that were found were reasonable, but were hampered by a paucity of
data. In particular, reliable refuse generation rates and waste composition data were not available for
many countries. In addition, many landfill experts believe that climate (partiojlarly as It affects moisture
input to the landfill) has an effect on CH4 generation rates. No currently available model incorporates
climate as a controlling variable.
In order to accurately estimate CH4 generation in landNIs on a global basis, a model is needed
that is responsive to a wide range of climates, types of waste, and landfiHing practices. Understanding
the effects of climate on CH4 production Is especially important to dimate modelers who are studying
feedback effects of global dimate change. Therefore, AEERL initiated a field testing program to gather
data to:
Identify key variables that affect Cfy generation; and
Develop an empirical model of Cfy generation based on those variables.
From the literature review, several variables were identified as potentially important: refuse
moisture content, refuse composition, refuse age. pH. and a variety of other variables related to landffll
characteristics and waste-handling practices. The scope of global landffll emissions estimation, however.
limits the number and type of variables that can be considered.
LandfiUs with gas recovery systems, where landfiO gas is collected and measured by the gas
recovery operators, offer unique opportunities for studying Cfy production. If data from these landfills
can be verified to be reasonably accurate, and V sufficient information is avalaUe about the landfill itself,
the landfiN gas measurements collected at the site may be used to estimate total 014 generation. If
sufficient sites are available to provide a representative sample of current U.S. landfills, then an empirical
model may be developed. Eventually, this model can be expanded to estimate methane emissions from
landfills globally. Gas recovery systems are widely used in Europe, and are becoming more common in
other parts of the world. They represent an important source of data for estimating global landfill 014
emissions, and may be used to calibrate a model developed from U.S. data.
The first step in developing the field testing program was a plot study of six U.S. landfills that have
CH4 recovery systems (Campbell et ai., 1991). The primary objectives of the plot study were to
determine the types and quality of landfill data avalable and to assess the feasiblity of expanding the
study to indude other sites. The study Included gas sampling and testing. For the six sites visited, the
sampling program demonstrated that landffll gas composition, as measured by landfill monitoring
equipment, was reasonably accurate.
-------
The results of the plot study were sufficiently encouraging so that a full-scale field program was
begua However, the field program was limited to acquisition of CH4 data gathered by on-stte monitors.
Furthermore, no other sampling or testing was planned. Data acquisition was confined to historical
records kept at individual sites.
The limitations of this approach were recognized at the outset:
Collection efficiencies of the recovery systems are unknown;
Waste composition data are usually not avaBaWe for a particular site; and
It is Impossible to fully account for differences in the structure and operating characteristics
of landfills.
Given these caveats, a certain amount of unexplainable variability was expected in the final results.
Nevertheless, by choosing sites from a wide range of climates, any relationship between climate and
CH4 production might be detected if the effects of site-specific factors (such as refuse age) could be
identified and removed.
OBJECTIVES
The objectives of the study described in this report were to:
Develop a statistical model of annual landfill CH4 emissions as a function of climate, refuse
mass and age, and other physical characteristics (if warranted);
Compare the performance of the statistical model to a deterministic kinetics-based model of
landfill CH4 production; and,
Develop a simple model that can be used to estimate global CH4 emissions from landfills.
It is important to note that CH4 recovery is being used as a surrogate for CH4 emissions in this study,
thus affording the potential to either underestimate and overestimate emissions. The method may
underestimate if gas recovery is not 100 percent efficient; some CH4 may still be lost through the cap or
by lateral gas migration out of the landfill. On the other hand, the method may overestimate if gas
recovery circumvents the reoxidation of CH4 by methanotrophs, methanogens, and suKate-reducing
bacteria. Given that strong arguments can be made for both cases and no quantitative data exist for
either, the approach used in this study is to assume that both cases are true but the net effect is zero. If
data that refute this assumption become available, adjustments to the model wOl be made.
-------
The landfills Included in this study are described in Section 2. Data acquisition and development,
and analytical methods and models are described in Section 3. The results of the analyses are
discussed In Section 4. Section 5 presents conclusions.
-------
SECTION 2
LANDFILL CHARACTERISTICS
Data from 25 landfills were obtained for the field study. Brief descriptions of the landfills are
presented below. Landfills 1 through 6 were part of a pilot study (Campbell et al., 1991). Only
Landfills 23, 24, and 25 were not visited by Radian personnel; data for these three landfills were provided
by on-sfte operators. Data from Landfills 14,15,18, and 19 were not used in this report because they
either did not maximize gas recovery or they lacked historical refuse/gas recovery data. Therefore,
these landfills are not described below. The landfill site visit reports written during this study are
provided in Appendix A (excluding Landfills 14,15,18, and 19). The site visit reports for Landfills 1
through 6 are provided in the pilot study report (Campbell, et al.. 1991) and are, therefore, not
reproduced here. Landfill data are given in metric units. A metric/U.S. equivalent conversion chart is
provided in Appendix B.
Landfill 1
Landfill 1 is located in Wisconsin and was visited August 6,1990. The landfill covers about
35 hectares. Refuse was accepted at this site from sometime in the 1950s untfi June 1989, when the
landfill was closed. The site accepted hazardous waste until the early 1980s; the hazardous waste was
placed in separate cells from other refuse. The landfill is reportedly covered by a 1.5-meter-thick cap.
Gas recovery began at this site on December 31.1985. Forty-five wells are in place, 25 along the
perimeter of the site (installed in 1985) and 20 on the interior portion of the site (added in 1987). Six
wells that are placed over older sections average 12 to 15 meters in depth. The wells installed most
recently are 24 to 27 meters deep.
Landfill personnel do not routinely monitor for surface or perimeter gas migration. Problem areas
are usually identified by visual inspection of the surface for vegetative stress. Once a problem area is
identified, the decision is made by landfill personnel whether or not to install a new well. Except for
roadbeds, the entire landfill surface was seeded with grass. The only fissures noted in the landfill
surface appeared to be due to water erosion. It is not known, however, if gas is escaping through these
fissures.
-------
LandfBI2
LandfiD 2 is located in Illinois. It was visited on August 7.1990. Two closed sections cover about
54 hectares. Refuse was first placed in the older closed section of the landfll In 196a Refuse
acceptance at the newer closed section began in November 1982. The original landfll owners were very
Inconsistent In cap placement and cap thickness. Cap thickness in the older section varies from 0.15 to
2.4 meters. The day cap on the newer section has an average thickness of 0.9 meters. The current
owner uses visual vegetation inspection and routine surface monitoring to identify areas that need a new
or thicker day cap.
The older section of the landfll produces from 19.000 to 30.000 liters of leachate per week. The
leachate and the landfll gas condensate are transported by truck to a local wastewater treatment plant
The newer section of the landfll produces a much smaller quantity of leachate (not quantified, however).
Gas is being recovered from the two closed sections. The current owners installed a flare system
In 1988, and converted to turbines in January of 1989. Of a total of 72 wells. 65 are currently on-line. Of
the on-line wells. 44 are considered very active and were used in the data analysis. The other on-line
wefls are very low flow and were installed primarly to control odors. The older section of the landftt has
40 wefls in place, and the remainder are on the newer section. Routine gas monitoring reports on
permanent probe testing for pressure, percent CH* and water levels are prepared by landfll personnel.
Wefls are added as needed to improve gas recovery.
The landfll operators place a great emphasis on controlling any gas migration problems so as to
prevent odor complaints and vegetative stress. A visual inspection of the vegetation growing on the
landfll surface revealed only one area with vegetative stress; a wefl had already been installed to correct
the problem, but it was not yet under a vacuum. Gas was bubbling through the water that had collected
in the bottom of the wen. No odors were detected In any other part of the collection area. The only
location at which measurable CH4 concentrations could be found with a handheld organic vapor
analyzer (OVA) was within a new wen enclosure. No other significant leaks were found.
Landfill 3
Landfll 3. located in Pennsylvania, was visited on August 9.1990. The landfll comprises a closed
portion, which covers about 51 hectares, and an active portion, which covers about 24 hectares. Refuse
acceptance began in 1970 and essentially ceased in 1988 for the portion of the landfll with gas recovery.
Hazardous wastes were accepted until 1981. and make up about 1 percent of the total refuse. Refuse is
-------
stBI being added in small amounts to the closed portion as settling occurs. Average day cap thickness
is 0.6 meters.
Gas was originally vented to the atmosphere to control lateral off-site migration. Gas recovery
began in January 1988 on the closed portion of the landfll. Gas wil be recovered from the active
portion at some future date. Two turbines are currently operating full time. A total of 31 wells are on the
site, with an average well depth of 30 meters. Landfll personnel report that they are encountering
problems on the eastern slope of the recovery area, with organic vapor surface probe readings of 25 to
48 percent as Ofy. One suspected reason for this problem is the fact that this slope has several
leachate manholes that are not tied into the gas collection system.
A few areas of the landfill had sparse vegetation, but it could not be concluded that these areas
had migration problems because the topsoD applied to the site was of very poor quality, and prior to the
site visit there had been a 6-week period with very little rain. A very strong landfill gas odor was
detected in at least five separate areas on the eastern slope (even with a brisk wind), but there were no
signs of .vegetative stress.
Landfill 4
Landfill 4 is located in Florida. The site was visited on August 20,1990. One area of the landfill is
closed; another area is currently accepting refuse. The closed portion covers about 57 hectares.
Refuse acceptance in the closed area began in 1971 and ceased in Aprl 1989. The current area was
then opened. Portions of the landfill also accepted sludge from a nearby wastewater treatment plant in
the past and continue to do so. Construction and demolition debris make up approximately 5.5 to
6 percent of total refuse volume. Final cover on the dosed area is 45.7 cm of topsoB, 45.7 cm of day
(rock tailings), and 45.7 cm of sand. This cover is very permeable to rainfall and the permeability also
limits the amount of vacuum that can be applied.
The Plant Manager estimated leachate collection up to 5.3 x 106 liters/month, depending on
rainfall. Because the cap is so permeable, the amount of leachate produced is greatly affected by
rainfall amounts. Leachate from the area currently accepting waste is shipped off site to a wastewater
treatment facility along with condensate from the gas collection system. The dosed portion of the
landfill with gas recovery does not have a true leachate collection system.
Gas is recovered from the dosed portion of Landfill 4. One hundred and eleven wells are in place.
Average well depth is 21 meters, with depths ranging from 18 to 46 meters. Five turbines were installed
and brought on line during March and April of 1989. Official start-up began July 1989. Prior to this time,
-------
recovered gas was processed in a gas plant and/or flared. Currently, the fadfty continuously operates
four turbines at 95 percent capacity. The Plant Manager hopes to increase recovery by installing eight
new deep wells.
Permanent bore holes have not been installed for routine perimeter gas migration monitoring.
Buldings near the perimeter of the site (up to 305 meters from the landfiU) are routinely tested for gas
levels. On-ske personnel indicated that when vegetation stress is Identified, they first try to adjust the
vacuum on nearby wells. If required, a decision is made as to whether a new well should be added to
alleviate the vegetative stress. Sol dehydration due to lateral gas may also result in vegetative stress.
LandffllS
Landfill 5. located in southern California, was visited on August 23,1990. Since refuse acceptance
began at this site in 1952. refuse has been placed in the pit left from a gravel mining operation. During
the 1950s and 1960s, the site primarily received inert waste, but at that time the waste also contained a
high proportion of orange trees. The very center portion of the landfill reportedly contains a high
proportion of construction and demolition debris. Landfill personnel have no knowledge of hazardous
waste being disposed of at this site.
The dosed portion of the tandfll covers about 32 hectares. The closed portion did not have a
final cover in place at the time of the site visit although a cover was to be Installed In the Fall of 1990.
The area is currently covered with a permeable slty sand and is not vegetated. Landffll personnel
estimate the moisture content of the refuse in place to be 12 percent Rejection of condensate to
boost moisture in the refuse was permitted by the local authorities until 1985. Landfill personnel note,
however, that since this practice ceased, there has not been any appreciable drop in either gas or
condensate production. There is no active leachate collection system In place where gas is being
collected.
Gas is being recovered from the closed portions of the bndffl. Gas collection first began at this
site in 1976; the previous owners periodically used an internal combustion engine to produce energy or
flared the gas. The closed portion of the landfin has a total of 102 weds, 42 interior and 60 perimeter.
Orifice plates are used on each wen to measure and control gas flow. The lines connecting the well
systems are kept separate and lead to flares on the perimeter of the site. There are three flares at the
site, one for each well system and one for backup use only. The interior weds are better producers than
the perimeter weds, with high gas flow and higher gas Cf^content The depths of the interior wells
range from 46 to 76 meters, whle the perimeter wens, designed primarily for migration control, are much
shallower.
8
-------
Local regulations often limit customer use of landfill gas. Therefore, there is currently only a
sporadic market for gas sales at this faclity, and the majority of recovered gas is flared. Operators of
this facility are optimistic that gas sales will increase in the future, and predict that as waste acceptance
rates Increase, gas recovery rates wfll also increase.
Perimeter gas migration is controlled by the GO shallow perimeter wells that encircle the closed
portion of the landfill. Monthly surface test data typically indicate organic vapor readings below 50 parts
per million. No measurable organic vapors levels were detected during the site visit Vegetation has not
been established on any portion of the landfill.
Landfills
Landfill 6, located in northern California, was visited on August 24,1990. Three areas of the landfill
(Areas 1, 2, and 3) are now closed. Refuse is currently being placed in another area. The closed areas
comprise approximately 40 hectares.
Refuse was first accepted at this site in 1975. Paper waste accounts for approximately 46 percent
of the total refuse at the site; garden wastes account for 13 percent, and glass/ceramics and food
waste, 10 percent each. The average refuse moisture content is reported to be 23 percent The final
cover on Areas 1 and 2 and part of Area 3 consists of a 1.2-meter-thick day cap and 0.3 meters of soH.
with vegetation established. Parts of Area 3 have not yet been seeded with vegetation.
Gas recovery from the closed portions of the landfill began in August 1988. The current system
consists of 68 wells, three internal combustion engines, and a backup flare that is used if one of the
engines fails. This flare is constantly burning, and normally runs on propane (with only a small stream of
recovered CH4). On-site personnel indicate that the volume of Cfy burned in the flare has been steadily
decreasing over time.
Condensate collected at the well heads is fed back into the fill area. Condensate collected at the
gas recovery plant is combined with the leachate collected from the landfill and transferred to one of two
surface collection ponds. The liquid is then allowed to stand unto it reaches a solids-to-liquid ratio of
50:50. After testing the mixture's toxicity, it is placed in the landfill if permitted.
Information received from the County Environmental Health group, the party responsible for gas
migration testing, showed that there are no areas with any significant gas migration problems. At the
time of year testing was performed for this study, all vegetation was dry and in a generally dormant
condition. Visual inspection indicated one small area of possible distress during the previous growing
9
-------
season. The cap Is fuH of targe cracks, caused by excessive dryness of the sol; however, no leakage
from the cracks was detected.
Landfill?
Landfll 7 is located In North Carolina and was visited on March 14, 1991. The landfiD, which was
opened in 1972. covers 51 hectares and is divided into three areas based on past and current refuse
placement patterns. Closed areas are covered with a 0.61 -meteMhtek day cap. This landfill has no
active leachate collection system; however, groundwater monitoring weds are located around the
perimeter of the property.
The current refuse acceptance rate is approximately 1270 metric tons per day, with at least
36 metric tons being construction and demolition debris. Landfll operators have no knowledge of any
commercial hazardous wastes being placed at this site, but asbestos-containing bulding tiles were
accepted unti 1985. The bidding ties are placed with the construction and demolition debris in areas
that are separate from the areas that receive residential garbage.
Gas recovery began at this site in December 1989. The average depth of the 48 existing weds is
12 meters, but as new weds are instaBed in higher fn areas, they wfl be drffed 26 to 28 meters deep.
The landfll gas is used as boier fuel and is the main fuel source for a nearby Industrial facility.
Condensate from the gas stream is sent to the d»y sewer system. One flare is located by the blower.
which is computer-operated to automatically run when demand for steam from the boier is low. Gas
composition measurements are taken daly at a point along the pipe leading to the flare. Percent CH4
typically runs from 40 to 53 percent, and averages 51 percent At this point, the gas contains an
estimated 2 to 3 percent water by volume.
The dty has installed 14 perimeter wells to monitor off-site gas migration. Monitor levels
reportedly never exceed 0.05 percent CH* Prior to installation of the gas recovery system, there were
frequent odor complaints, which have now ceased. LandflR personnel believe that only 65 percent of the
avalabie 0(4 is being recovered (based on amount of refuse in place), with the majority of the gas
escaping along the steep slopes without being detected at the perimeter migration welts.
LandflU 8 is located in Georgia. It was visited on March 19. 1991. Refuse was accepted at this
22-hectare site between 1973 and 1987. Waste composition is estimated to be 60 percent commercial.
25 percent residential, and 15 percent construction and demolition debris. Paper reportedly accounts
10
-------
for SO percent of the landfll refuse. No liquids or sludges were accepted during the life of the landfill.
The landfll is covered with a 0.61- to 1 ^-meter-thick day cap and has an established vegetative cover,
mostly grasses with some wid flowers. There is no active leachate collection system at this landfill.
Laachate is reportedly well contained by the natural day liner under the landfill.
Gas recovery began at this site in 1986. One hundred and seven wells are in place with plans to
reduce the number of wells by approximately half to reduce competition between wells and increase
individual well efficiency. Average well depth is 17 meters. Replacement wells (approximately 20 per
year) wDI be drilled deeper than original wells in most cases. There is one flare located by the gas
recovery facility. The flare can bum 28 cubic meters per minute, which is not enough to prevent all
migration when the system is down. The flare is not used when the plant is operating.
The gas treatment process to produce pipeline quality, high-Btu gas is considered proprietary.
Before treatment, the CH4 content of the landfBI gas ranges from 55 to 59 percent After processing, the
gas contains at least 95 percent CH4. The gas is sold to a local gas company.
Perimeter wells have been installed at 305-meter intervals around the landfBI to monitor off-site gas
migration. The wells average 9 to 14 meters in depth. They are monitored monthly when the gas
recovery plant is in operation, and weekly when it is not operating. When the recovery plant is not
operating, the vacuum on the perimeter wells is increased and the vacuum on the interior wells is
decreased to prevent off-site migration.
Landfill 9
Landfill 9 is located in Mississippi and was visited on March 19,1991. The landfill has been
operational since 1979 and has 73 hectares available for refuse disposal. The landfill is divided into
three cells, two of which (Cell 1 and Cell 2) have been fflied. The two filled cells cover approximately
12 hectares and are covered with a 0.61- to 0.91-meter-thick day cap and have established vegetative
cover, mostly grasses and some wOd flowers. Cells 1 and 2 do not have active leachate collection
systems. Cell 3 does have a system and, ultimately, the gas condensate sumps will be tied into this
system.
Landfill personnel estimate that approximately 20 percent of incoming refuse is construction and
demolition debris, 60 percent is commercial, and 25 percent residential. LandfBI personnel have no
knowledge of accepting commercial hazardous waste, gas tanks, or car batteries. Tires are accepted if
they have been shredded or cut in half.
11
-------
A gas recovery system was installed at this landfiH in September 1990 in response to migration
problems. There are a total of 32 wells. 20 of which are in Cell 1. and 12 in CeH 2. Interior wells are
18 to 21 meters deep and perimeter wells are about 9 meters deep. An additional 15 deep interior wells
should be operational by summer 1991. Recovered gas averages 40 percent CH4; gas from perimeter
weds has a lower Cfy content There is one flare located on site, which is burning about 17 cubic
meters per minute.
Perimeter probes have been installed at 305-meter intervals to monitor off-site gas migration.
These probes are tested monthly. The site is flaring the minimum amount of gas necessary to keep the
perimeter probes at 0 percent CH*
LandfflMO
LandfD110 was visited on March 21.1991. It is located In Alabama on approximately 40 hectares,
which are divided into two 20-hectare areas based on past and current refuse placement patterns. The
landfiU has been operational since 1958; however, accepted refuse was burned In place in the early
1960s. Waste acceptance is expected to cease in 2006. Waste is covered daly with sol excavated
from an area behind the landfffl. A 0.91-meter sandy day sol cap was placed on the newer area of the
landfiD in 1985. This landfiO does not have a leachate collection system, but has had no off-site leaching
problems. However, leachate monitoring weds are maintained at the landfffl perimeter and checked
twice each year.
Incoming refuse is not segregated. LandfiH personnel have no knowledge of liquid, hazardous, or
infectious wastes being accepted. Periodic spot Inspections of incoming refuse is conducted and actual
dumping monitored two or three times per week to ensure that no prohibited waste is entering the
landfill. Asbestos was accepted in the past, but is now banned. Tires are stiff being accepted. In a
waste characterization study done for EPA in 1986. approximately 50 percent of total waste was
determined to be from household sources. 20 percent from construction and demolition, and 5 percent
from industrial processes.
Landfill gas recovery began in 1988. The 96-weD system draws gas from approximately
32 hectares of the landfffl. The weds are approximately 21 meters In depth, with the bottom 6 meters
filed with rock. Gas recovery personnel expect 50 new wells to be driNed in the near future. The
recovered gas is treated and sold to a local gas company. The volume of pipeline-quality gas sold is
about half the volume of total tandfll gas extracted. The other 50% of the landfill gas is unusable water,
hydrocarbons, carbon dioxide (CO;?), oxygen (02), and nitrogen (N2).
12
-------
In 1979, a polyethylene curtain was placed In a 7.6-meter ditch on the perimeter of the landfill to
stop CH4 gas migration. Gas migration is checked once each year at the landfill perimeter. No
vegetative stress was observed.
Landfill 11
Landfill 11, located in Georgia, was visited on March 22,1991. The landfill covers 49 hectares and
is divided into two approximately equal areas (Parcel A and Parcel B), based on gas rights; there are no
obvious distinctions between the two areas. The landfill began accepting waste in 1962 and is expected
to be active until at least 1993.
The landfill has accepted nearly all of the residential and commercial waste from the surrounding
county (population 170,000) since the landfill opened. Landfill personnel have no knowledge of
hazardous waste placed at the site; however, waste is not routinely inspected for the presence of
hazardous waste. Sewage sludge from the local wastewater treatment plant is accepted. Prior to
landfBling, the sludge moisture must be reduced such that the sludge will not pass through a paint filter.
Construction and demolition debris was accepted until 1986. Medical waste is not accepted. This
landfill has no active leachate collection system. However, there are groundwater monitoring wells
around the perimeter of the property.
The company that owns the gas rights to Parcel B has not recovered gas from that area since
1989. A brick manufacturing company has been collecting gas from Parcel A since 1985. Parcel A
contains 39 wells with average depths of 14 meters. Landfill gas is treated by mechanical refrigeration to
remove water. All of the gas collected is used as fuel in a brick kfln. None of the gas is vented or
flared.
Gas migration monitoring is not required at this site, and there is no monitoring program in place.
Landfill personnel state that vegetative stress on the landfill surface is not a problem.
Landfill 12
Landfill 12 is located in Oregon. It was visited on April 4,1991. The landfill covers approximately
34 hectares and is divided into two segments: north and south. Filing began in the south area in 1969
and progressed to the northern boundary untl 1980. Between 1980 and 1983 waste was placed in a
second lift in the south area. The site was closed in 1983.
13
-------
The tandffl is butt on land that was formerly a marshy wetland. Since 1963, the final landfll cover
has been upgraded to increase slope and water runoff. Currently, the landfll cover consists of an
average of 1.8 meters of low-permeablity sol. The slope of the landfill averages just over 2 percent, and
grass covers the entire surface. No active leachate collection or treatment system exist at this site.
However, groundwater monitoring wells are located around the perimeter of the landfll.
The landfill accepted primarily residential and commercial wastes. The tendril's operating permit
specifically excluded medical wastes, chemicals, ote, liquids, septic tank pumpings, and nondigested
sewage sludge; however, household hazardous wastes were not excluded. Some construction and
demolition debris was accepted and placed separately from the general refuse.
Gas recovery began in 1984. Of a total of 78 active wells, 50 are horizontal wells placed In
trenches about 1.8 meters below the landfll surface and ranging in depth from 61 to 305 meters. The
remaining wells are vertical, with an average depth of 11 meters. A private company operated the gas
recovery system from 1984 until 1989, when gas production at the site decreased to the point where gas
purification was no longer profitable. Currently, some gas is extracted from the existing well system and
used to fuel holers for the county; however, the system is no longer operating In an optimal fashion.
Perimeter monitoring for gas migration was not practiced at this site.
Landffl 13
Landfffl 13 is located in Washington state and was visited on Aprl 5,1991. The landfll is 50 years
old and is scheduled to dose in December 1991. The landfH was bull on an old gravel mine. The
underlying sol is sandy gravel. The bottom of the tandfiU is about 1.5 meters above the groundwater
table and is lined with a layer of demolition refuse. The landfll does not have a leachate collection
system. Groundwater contamination from landfll leachate has been documented. The landfll owners
are currently bulding a water treatment plant that wU commence operation at the end of 1991.
Contaminated water from the underlying aquifer wH be extracted, treated, and returned to the ground.
The landfll is divided into two areas. The north area comprises approximately 20 hectares and
has been closed since 1988. A60-ml polyethylene liner was placed on 14 hectares of this section tn
1989. The south area comprises approximately 12.1 hectares and refuse is stJI being accepted in
portions of the area, A 0.15-meter cover of sandy dirt is placed over the active area daly. Open
burning of refuse was practiced from the time the landfll was opened until the mid-1960s, after which
refuse was either shredded or placed directly in the landfll and compacted.
14
-------
LandfiH gas recovery began in the late 1970s. Separate CH4 gas recovery systems were operated
in the north and south areas until March 1991. The landfill operates a total of 71 vertical gas extraction
wells, 20 in the south area and 51 in the north area. Eight of the 20 in the south area are located In the
active dumping area Each well is 9 to 15 meters deep. Twenty-five new wells wfll be Installed In the
currently active area at landfill closure. Initially, recovered landfill gas was burned by separate flares in
each section. The south area flare is currently inoperable and the south area wells are tied Into the
north area collection system.
Gas migration is measured with 20 gas probes located on the perimeter of the landfill. Very little
gas migration has been noted since the flare system was installed. However, substantial amounts of
landfill gas were observed bubbling in puddles along the road between the north and south areas.
LandfilMS
Landfill 16 is located in New York and covers 30 hectares. It was visited on Apr! 10,1991. Waste
acceptance at this site began in 1976. All white goods, wood, tires, and construction debris are
separated out of incoming refuse. Sewage sludge is accepted and is placed on the south slope of the
fill area. Asbestos is also accepted and placed in a separate area of the landfill. There is no final cap
on any part of the landfill, but a cap is planned in the near future. Leachate is collected in several ponds
around the site.
Gas recovery began at this site in December of 1988. The gas is treated and then used to operate
a 3MW turbine generating electricity. Thirty-six wells, spaced 30 to 61 meters apart, are connected to
one header. Average well depth Is 18 meters; however, wells located on the south side of the fill area,
where the county is actively dumping waste, average only 11 meters.
The county is responsible for monitoring off-she gas migration and has installed several perimeter
wells. Migration is thought to be minimal.
Landfill 17
Landfill 17, located in Massachusetts, was visited on Apr! 23.1991. The landfill covers
16 hectares and is divided into two sections, one active and one inactive. A 60-mH synthetic liner is In
place under the active section; it extends over the inactive half to also act as a cap. A 0.3-meter-thick
day cap also covers the inactive section. Leachate is collected from the approximately 8 hectares of the
landfill that are lined.
15
-------
The landfill started accepting refuse in 1952, but much of the refuse currently in place was
received within the past 10 years. Commercial waste accounts for 85 percent of the refuse, with the
remainder being residential waste. The site contains tires, which were accepted by a former owner.
Hazardous waste was accepted between 1968 and 1979. Batteries are banned from landfills by State
regulation.
A candlestick flare (small, unenclosed flame) was introduced at this site in 1987; however, it was
not large enough to handle the amount of landfill gas collected. A new flare system went on-line in
February 1990. The current gas control system consists of 41 wells which feed one enclosed flare.
Perimeter wells from the 1987 gas collection system are monitored periodically for Cfy, but show
no gas migration from the site. Interior wells appear to be pulling gas toward the center of the site,
thereby controlling off-site migration.
Landfill 20
Landffl 20 is located In Michigan. The site visit took place on May 1,1991. The landfill covers
approximately 100 hectares, which are divided into two sections based on past and current refuse
placement patterns. The entire landfffl is surrounded by a day dike and a trench to control surface
water run-off. Refuse was accepted in the northern 65-hectare section from 1967 to 1985. Refuse was
then placed in the southern 31 hectares, which are stil being filled. Ten hectares of the northern section
have a final cap (a 0.61-meter-thick day cap and 0.15 meters of topsol seeded with grass). The
remaining 55 hectares-apart from the valley now being fflled-have a 0.7-meter-thick day cap. A final
cap wB be placed over this area in the near future.
The site contains no active leachate collection system under the oldest refuse in the northern
section of the landfll. The southern section and the upper fiH areas of the northern section do have a
system. Condensate is combined with leachate and sent to leachate collection ponds and aerated.
During dry periods, some of the liquid is redrculated in active fill areas and haul roads to reduce dust
The remaining liquid is tanked and trucked off site.
Refuse entering the tandfiH is classified as general refuse (municipal solid waste), compact general
refuse (municipal solid waste), construction and demolition debris (usually indudes wood), bulky wastes
(white goods) and process residue from a local resource recovery fadity, ferrous metal (now recyded),
contaminated soB, non-contaminated sol. sand and concrete, asbestos, industrial and wastewater
sludge, and ash from a local incinerator. Only incinerator ash is segregated; it is placed in a lined,
2.8-hectare monoffll.
16
-------
Landfill gas collection began in January 1990. Initially, 69 wells collected gas from 73 hectares of
refuse. In December 1990. an additional 11 wells were brought on line. Eighty wells are currently
distributed over 85 hectares, with the majority being located in the northern section, primarily where
refuse is deepest Average well depth is 17 meters; maximum depth is 20 meters. The gas is currently
flared, although there are plans to sell electricity to the local utility company. In order to prevent off-site
migration, the amount of gas flared is maximized.
Gas migration is monitored at 31 perimeter probes, which are tested monthly. Landfill personnel
indicate that two or three of the probes along the southern boundary of the landfill (near the newly-
placed refuse) show slightly elevated CH4 levels. As new wells are installed along the southern slope,
this migration will be controlled. In addition, four leachate collection wells in this area had CH4 levels so
high that flares were attached.
Landfill 21
Landfill 21 is located in Michigan and was visited on May 2,1991. It began operation In 1971 and
is currently divided in three areas based on temporal waste acceptance. The oldest area encompasses
18.6 hectares and accepted waste between September 1971 and the fall of 1978. The second area
encompasses 8 hectares and accepted waste between 1982 and 1986. Both of these areas are now
dosed and are graded to the original contour of the land. The oldest area of the landfBI is capped with
0.61 meters of day and covered with approximately 9 meters of dirt The second area is capped with
0.61 meters of day and is scheduled to have a low-density polyethylene liner.
The third area, the only active portion of the landfll, encompasses 16 hectares. Waste is placed in
excavated cells to an approximate depth of 20 to 30 meters. The area is lined with 41 mil polyethylene.
Asbestos and contaminated soil are the only wastes separated from incoming refuse. Asbestos is kept
in a controlled area, separate from all other landfill waste acceptance areas. Contaminated soil is co-
disposed with other refuse. Municipal sludge is not accepted. The landfill has an active leachate
collection system in all areas except the oldest
Landfill gas recovery began in February 1990. The landfill gas collection system consists of
56 wells with an average depth of 18 meters. Forty-three wells are in the oldest area and 13 are in the
second area.
A flare was installed to control off-site migration, and no CH4 has since been detected during the
monthly inspection of the 20 perimeter probes. No vegetative stress was detected in the dosed areas
on the perimeter of the landfill.
17
-------
LandfH22
LandflU 22 is located in Minnesota and covers 40 hectares. The site was visited on May 3,1991.
Refuse acceptance began in 1970 and ended in 1988. The landfill accepted primarly municipal solid
waste. No known commercial hazardous wastes, little construction and demolition debris, and little
sludge were accepted. The landfill does not have final closure, but the entire surface is covered with at
least 0.61 meters of sflty day. which b 3 to 4.5 meters deep In some areas. The refuse is settling 1.5 to
3 meters per year, and day is added to control surface water drainage. No active leachate collection
system exists at this landfiH. The IB area is surrounded by 54 groundwater test wells.
LandfiU gas collection began in February 1989. The gas is currently flared, as gas recovery
personnel feel that electricity or pipeline-quality gas cannot be produced profitably at this site. The well
field has a total of 54 weds; 14 of them are in the east area to prevent off-site migration and are not
placed in refuse. The 40 high-flow wells range in depth from 11 to 24 meters.
Off-site gas migration is monitored with 74 perimeter probes that are tested weekly. Because of
concerns over the proximity of houses on the east side of the tendfU, wens were installed to contain
potential migration. The amount of gas recovered from these weOs is negligible. Gas migration through
the cap is monftored with an organic vapor analyzer. There Is reportedly no significant amount of gas
escaping to the atmosphere.
LandHI23
LandfiB 23 is located in Colorado and covers 40.5 hectares. Refuse was accepted at this site from
the late 1950's untl 1986. The majority of refuse was placed between 1981 and 1985. The surface of
the tendfffl has a day cap 0.6 to 1.5 meters thick.
Gas recovery for electric power generation began in December 1986. A reciprocating engine is
used to generate the electricity. The wefl field consists of 19 high-flow weds and 12 low-flow wens. The
average depth of the wells is 19.8 meters, with the minimum and maximum depths corresponding to the
refuse fiH depths. There is one flare located at the site which is used to bum off landfill gas not
consumed for power generation.
Landfill 24
LandfiH 24 is located in Texas and covers a total of 109 hectares. Waste acceptance began in
1972 and continues to date. A large portion of the refuse was received between 1983 and 1988. giving
18
-------
an average refuse age of 5.6 years. Refuse composition is estimated to be 55 percent commercial,
10 percent residential, and 35 percent demolition debris. Closed portions are covered with a day cap
1.5 meters thick.
Gas recovery began in May of 1988. The landfill gas is burned to generate electric power using a
turbine. The well field has 23 high-flow wells and 5 low-flow wells.. Well depth averages 15 meters.
Landfill 25
Landfill 25 is in Illinois and covers 75 hectares. The landfill began accepting waste in 1970.
Average refuse age is 12 years, with most waste accepted between 1973 and 1982. Refuse composition
is 60 percent residential, 20 percent commercial, and 20 percent demolition debris. Closed portions of
the landfill are covered with a day cap 0.6 to 3.6 meters thick.
Gas recovery began in July 1988. The landfill gas is burned to generate electric power using two
turbines. The system consists of two turbines, 43 high-flow wells, and 30 low-flow wells. There is also
one flare located on site used to bum off the landfill gas not used for power generation.
19
-------
SECTION 3
DATA REDUCTION AND MODEL DEVELOPMENT
All the data in this report, except for the climate data, came from the records kept at individual
sites. Therefore, considerable effort was expended in evaluating each set of data, standardizing units.
and preparing the data for analysis, as described below. (Note: All data and results tables in the main
body of this report are presented In metric units; Appendix C contains select tables in U.S. equivalents.)
DATA QUALITY
Since the data used in this analysis were gathered by site operators to suit their own needs, the
quality of the data could not be controlled by the measures that are normally used for actual emissions
sampling. The acceptable error for this study is therefore greater than it would be for a controlled
testing program. Gas composition estimates calculated by on-site operators were verified by
independent sampling for the first six sites (Campbell et al.. 1991). The relative accuracy for the percent
of CH4 by volume ranged from 2 to 19% and averaged 10.5% for LandfiNs 1 through & Furthermore, no
biases in the on-site data were found. Based on this analysis, the on-site CH4 data were found to be
acceptable for the purposes of this study.
The accuracy of the flow rate data could not be verified directly. Instead, the calibration
procedures used by the operators were evaluated. Based on the calibration procedures and instrument
types, the flow rate data were judged by Radian personnel to be reasonably accurate. The expected
accuracy of the flow monitors was assessed at 10% for the first six sites (Campbell et al.. 1991). Due to
greater variety of equipment and calibration procedures for the 22 sites, the average error is likely to be
somewhat higher. Therefore, the expected accuracy of flow rate data is estimated to be about 15%.
The greatest source of error is the refuse data The composition is not well documented for most
of the sites. The mass of refuse is calculated by several different methods. Refuse is weighed at the
gate at a few sites, but not at most The accuracy of these data cannot be quantified, but the error is
certainly greater than for the methane flow rate. Other landfill characteristics, such as depth and
volume, are equally uncertain.
20
-------
Overall, the data quality is good. Certain variables, such as refuse mass, are as good as they can
be given that the study required data from existing landfills with gas recovery. The accuracy of other
variables might be improved by use of independent gas testing, but the small increase in accuracy is not
worth the increased cost, especially given the large (and uncontrollable) uncertainty in refuse mass and
composition. The data reduction discussion below elaborates further on the data quality for specific
variables.
DATA REDUCTION
Methane Data
The data used to calculate the average CH4 recovery rate at each landfill were provided in
printouts or computer files listing total gas flow, percent CH4 composition of the total gas flow, and
other information applicable to an individual landfill. Data were reported for each gas recovery system,
usually in the form of dairy averages of hourly recovery rates or of monthly totals of gas flow. Table 1
summarizes the type and units of the gas flow data, the type of percent CH4 composition, and the time
period over which the gas flow data were available for each landfill.
The gas flow data were put into Statistical Analysis System (SAS) datasets and summarized to
obtain an average CH4 recovery rate for each landfill. For landfills with multiple turbines or flares, a total
gas flow value was calculated by summing over all operating systems. The total gas recovery rates
were then converted to cubic meters per minute, and the CH4 recovery rates were calculated by
multiplying the gas recovery rate by the appropriate CH4 composition percentage.
The resulting CH4 recovery rate distributions were examined for obvious outliers; however, no
statistical outlier screening was performed. Methane data points that were separated by large gaps from
the main body of data-sometimes by two or more orders of magnitude-were considered obvious
outliers. These data points often occurred at shut down or start up and were considered
unrepresentative of the optimum gas recovery operation; therefore, they were excluded from the CH4
averaging calculations. In addition, some portions of data were excluded for landfBIs that changed the
refuse acceptance rate, weds, acres, or other important variables during the period of record for which
gas flow data were received. Thus, only periods of record where the landfill parameter variables
remained constant were used to calculate GH4 averages.
Table 2 shows the average CH4 recovery rate for each landfill, as well as other summary statistics.
The number of measurements available varied a great deal between sites. For example, Landfills 13
and 21 had only 11 data points from a 12-month and 14-month period, respectively. On the other hand,
21
-------
TABLE 1. SUMMARY OF GAS FLOW DATA OBTAINED FROM THE LANDFILLS
Landfill
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
20
21
22
23
24
25
Gas How Data
Type
daDy average
daily average
daily average
daly average
daily average
daly average
monthly total
daily total
daly average
monthly total
monthly total
daly total
dally average
monthly total
point total
daBy total
daly average
daly average
daly total
daly total
daly total
Units'
CFH
CFH
CFH
CFH
CFM
CFM
CF
CF
CFM
CF
CF
CF
CFM
BTU
CFM
CF
CFM
CFM
CF
CF
CF
Methane Composition
(datatype)
daly percent
daly percent
daly percent
daly percent
daly percent
daly percent
average percent
daly percent
average percent
average percent
average percent
average percent
daly percent
average percent
daly percent
daly percent
daly percent
average percent
daly percent
daly percent
daly percent
Time Period
Covered
5/89 to 5/90
1/89 to 7/90
3/89 to 7/90
7/89 to 9/90
1/90 to 8/90
8/88 to 8/90
1/90 to 12/90
2/89 to 2/91
9/90 to 3/91
1/90 to 2/91
1/86 to 12/90
2/87 to 12/87
1/90 to 12/90
1/89 to 3/91
2/90 to 3/91
1/90 to 4/91
3/90 to 1/91
2/89 to 5/90
4/90 to 1/91
11/89 to 5/91
7/90 to 5/91
'Units abbreviations:
CFH = cubic feet per hour
CFM - cubic feet per minute
CF -cubtefeet
Btu - British thermal units
22
-------
TABLE 2. SUMMARY STATISTICS FOR METHANE RECOVERY RATES GROUPED BY MEASUREMENT TYPE
Measurements
Landfill
1
2
3
4
5
6
7
B
9
1.0
11
12
13
16
17
20
21
22
23
24
25
Type
daily
daily
daily
daily
dally
dally
monthly
dally
dally
monthly
monthly
dally
dally
monthly
minute
monthly
dally
dally
dally
dally
dally
Number
194
302
314
85
209
37
12
626
15
6
12
232
11
15
13
12
11
51
202
333
331
Average
55.3
18.0
40.0
98.4
24.8
16.7
9.7
11.7
7.7
29.3
11.3
8.0
10.4
16.0
13.8
35.0
27.4
33.2
2.2
17.7
20.2
Methane Recovery (m3/min)
Median
55.3
18.2
40.3
98.7
24.9
16.8
10.2
12.4
7.0
30.5
11.7
7.7
11.0
16.6
13.9
35.1
26.5
31.8
2.3
17.9
20.8
Standard
Deviation
2.12
1.19
2.32
1.33
1.70
2.07
2.01
2.46
1.42
3.34
1.22
1.02
1.50
4.13
1.50
4.75
2.94
7.84
0.51
2.18
2.80
Minimum
Value
48.0
12.3
30.2
93.3
20.5
12.6
4.0
0.5
5.7
23.4
9.1
5.4
7.8
7.4
10.1
26.5
24.5
21.6
0.3
3.1
1.4
Maximum
Value
61.4
20.5
44.6
101.5
27.9
22.6
12.0
17.1
10.5
32.4
12.7
10.4
11.7
21.6
16.5
41.3
32.9
60.0
2.9
22.1
24.4
Range
13.4
8.2
14.4
8.2
7.4
10.0
8.0
16.6
4.8
9.0
3.6
5.0
3.9
14.2
6.4
14.8
8.4
38.4
2.6
19.0
23.0
Coefficient
of Variation
(CV)*
3.8
6.6
6.3
1.4
6.8
12.4
20.9
21.0
21.2
11.4
10.8
12.9
14.5
25.7
10.8
25.5
10.8
23.6
22.9
12.4
13.9
*CV » 100 * (Standard Deviation/Mean); a unltless measure of relative variability allowing comparison of variance
between samples of various sizes or with very different means.
-------
landfll 8 had 626 data points from a 25-month period and Landfill 24 had 333 data points from a 1-year
period. More confidence can be placed in the reliability of the estimated means for the latter two
landfills; however, the gas flow variability is fairly low (as shown by coefficients of variation), so the small
number of measurements for some sites is not likely to be as problematic as It would be if flow were
highly erratic.
Amount of Refuse In Place
The amount of refuse in place was analyzed by weight for each landfill. As shown in Table 3,
which summarizes the methods used by different landfils to estimate landfill parameters, refuse data
from some landfills were provided by weight, whie data from others were provided by volume. It was
preferred that refuse data be provided by weight, particularly if the amount of refuse accepted each year
was measured with a scale, as In the case of Landfill 16. For the most part, landfill personnel did not
indicate how they arrived at their estimate of refuse weight In some cases, scales were installed after
some refuse had already been placed in the landfll.
The landfils that provided refuse data by volume did so because volume is the parameter used for
recordkeeping or permitting purposes, and is often the basis for the landfill's tipping fee. LandfOI
personnel were able to recommend a conversion factor for volume to weight in most cases. The
conversion factor used by personnel at LandfiOs 21 and 22 was 337 rr^/Mg.* For Landfill 21, refuse
weight was calculated with this conversion factor using estimates of the daly volume placed in the
landfiU. For Landfll 13, a conversion factor of 2.81 nft/Mg was recommended, and weight was again
calculated based on an estimate of the daly volume.
For Landfills 3.10,11, and 17, landfll personnel were able to provide refuse weight only for recent
years; for earlier years, only rough estimates were avaBaWe. The amount of refuse in place calculated
by the methods discussed above is included in Table 4, which shows the various landfill parameters that
were used in the statistical analyses.
Refuse Age
As shown in Table 3, the average age of the refuse at each landfll was determined by several
methods. The most accurate method was where the quantity of refuse placed in the landfill each year
was provided by landfll personnel. If this type of detal was not avalable, landfill personnel were asked
'Conversion factors supplied by site operators were usually in the form of yd3/lb. Metric equivalents
are reported here for consistency.
24
-------
TABLE 3. SUMMARY OF DATA USED TO ESTIMATE LANDFILL PARAMETERS
Variable
1 234 56789 10 11 12 13 16 17 20 21 22 23 24 25
Refuse Weight:
Provided as:
Weight
Volume
Other
XX XXXXXX X X
X XX
X XX XX
XXX
Refuse Age:
Weighted Average
Estimated
Open/dose Dates Only
XX XXXXXXX
XXX XXXXXX X X
X
Number of Hectares:
Entire Landfill
Distinct Section
Other
Not Known
XX X XXX
XXX XX
X X X X X X X
XXX
Landfill Depth:
Uniform
Benched
Pyramid-Shaped
Not Known
X XX X XX
X X
XXX X X XX
XXX
XXX
Well Depth:
Related to Rtl Depth
Not Related to Fill Depth
X XX XX XXXXXXXXXX
XXX XX X
Number of Wells:
Total Used
High Row
Other
X XXX XX XXXX XXXX
x x xxxx
X
n|a.076\report
-------
TABLE 4. SUMMARY OF LANDFILL PARAMETERS USED IN THE STATISTICAL ANALYSES
Landfill
Identification
Code
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
20
21
22
23
24
25
Refuse Mass
L«tt« (Id8 Mg)
A
B
C
D
E
F
G
H
1
J
K
L
M
P
0
T
U
V
w
X
Y
6.35
6.12
7.35
13.79
0.89
2.40
2.95
2.72
1.63
5.26
1.61
2.78
0.96
3.38
5.17
9.71
2.60
3.97
2.87
6.21
10.65
Average
Refuse
Age
6.0
10.0
10.0
9.5
15.0
7.0
10.0
10.0
7.0
12.0
10.0
8.5
7.0
5.5
10.0
11.0
13.0
12.0
10.7
5.6
12.0
Landfill
Area
(hectares)
34.80
54.64
50.99
56.66
32.38
40.31
50.59
22.26
12.14
32.38
27.92
34.40
14.17
30.35
16.19
72.85
26.71
40.47
40.47
109.27
74.87
Average
Landfill
Depth
(meters)
67.06
25.91
66.14
56.39
45.72
9.83
18.29
16.76
15.24
21.34
24.38
10.67
12.19
27.43
18.29
18.29
27.43
24.38
21.34
30.48
22.86
Average
Well
Depth
(meters)
13.72
14.33
23.47
21.34
34.14
9.83
12.19
16.76
15.24
13.72
13.72
10.67
12.19
18.29
18.29
17.37
18.29
24.38
19.81
15.24
21.34
Number
of Wells
45
44
31
111
102
68
48
107
32
98
39
78
51
36
41
69
56
40
19
23
43
Landfill
Average
Methane Average Methane
Recovery Recovery Rate
Volume Rate Per Unit Mass
(lO63) (m3/mln) (mfymln/IO6 Mg) Gas End Use
23.34
14.15
33.73
31.95
14.80
3.96
9.25
3.73
1.85
6.91
6.81
3.67
1.73
8.33
2.98
13.32
7.33
9.87
8.63
33.30
17.11
55.3
18.0
40.0
98.4
24.8
16.7
9.7
11.7
7.7
29.3
11.3
8.0
10.4
16.0
13.8
35.0
27.4
33.2
2.2
17.7
20.2
8.71
2.94
5.45
7.14
2.28
6.97
3.28
4.32
4.71
5.57
6.22
2.87
10.87
4.74
2.67
3.61
10.52
8.35
0.78
2.85
1.90
ELEC.-TURBINE
ELEC.-TURBINE
ELEC.-TURBINE
ELEC.-TURBINE
FLARE
ELEC.-IC ENGINE
BOILER FUEL
HIGH BTU
FLARE
HIGH BTU
BRICK KILN FUEL
HIGH BTU
FLARE
ELEC.-TURBINE
FLARE
FLARE
FLARE
FLARE
ELEC.-IC ENGINE
ELEC.-TURBINE
ELEC.-TURBINE
-------
to provide a rough estimate of the percent of refuse placed in the landfill during certain time periods.
For example, the majority of refuse in Landfill 23 was placed between 1981 and 1985, even though the
landfill accepted refuse from 1960 through 1986. Rather than basing average age only on landfill open
and closure dates, a weighted refuse age of 10.7 years was used to better reflect average refuse age.
Landfill 13 is the only landfill for which only open and closure dates were available; the average age for
refuse at this site was set at the median between the two dates. The average age for each landfill
resulting from the above calculations is shown in Table 4.
Number of Hectares
Table 3 shows that the parameter for hectares for each landfill can be viewed in three different
ways. Some landfills (1, 3, 8,12.17, and 22) had gas recovery systems that covered the entire fill area.
Others (2,5, 6,16. and 21) had gas recovery systems that did not cover the entire fill area, but covered
a distinct section of the landfill. For these landfills, the hectares shown in Table 4 are only for the
hectares covered by the gas recovery system.
It was more difficult to estimate the hectares for the remaining landfills because the gas recovery
system covered a portion of the landfill that was not distinctly separate from other sections of the landfill.
For example, Landfill 11 had a second gas recovery system that is no longer operating. The hectares
presented in Table 4 for Landfill 11 are only those covered by the currently-operating gas recovery
system, for which gas flow data were gathered. It is impossible to determine, however, whether or not
gas was migrating from the area where the second system was in place. Another example where the
hectares in the gas recovery area were difficult to delineate is Landffll 7, which has some steep slopes
that do not have wells. It is possible that some gas is escaping to the atmosphere from these slopes.
Landfill Depth
Landfill depth is included in Table 3 because It may be useful when evaluating hectares and well
depth as parameters. Several of the landfills had relatively steep side slopes and were pyramid shaped.
Landfills 8 and 9 are tall, but were buBt in benches so there are no large, steep slopes that may limit well
placement The remaining landfills were uniform in depth so that no well placement problems were
caused by landfill design. Again, Table 4 snows the value of this parameter for each landfill.
Well Depth
For most of the landfills visited, well depth is directly related to fill depth. For Landfills 1.3,4,7, 8,
and 11, however, this may not be true. Except for Landfill 8, these are the landfills that are pyramid
27
-------
shaped. LandfBI 8 is unique because the weBs were not drilled by the current operator of the gas
recovery system. The current operator plans to drll replacement wells deeper than the existing wells. In
tact, in most cases, landfill personnel indicated that future wells wil be drilled deeper than existing wells
in cider to maximize gas recovery. At Landfll 11. some of the older wells were covered with refuse.
Landfffl 12 is the only landfill visited that had horizontal trench wells in addition to vertical wells-ell of the
other landfUs had only vertical wells. Table 3 shows whether well depth is related to fill depth and
Table 4 lists the actual well depths for each landfiU.
Number of Wells
The parameter for the number of wells for most landfills is the total number of active wells
(excluding abandoned or capped wells). For Landfills 2,13, 22, 23, 24, and 25, however, only wells
considered "high flow* were included in the analysis, as shown in Table 3. The remaining low-flow wells
were not Included because they were installed to eliminate gas migration, and they produce very little
gas. At LandfBI 22. for example, wells excluded from the analysis were not even placed In refuse.
Landfill 8 is once again unique because the previous operator of the gas recovery system installed more
weds than the current operator feels is necessary. Table 4 lists the number of weds for each landffll.
Climate Data
Monthly average temperature and total rainfall data were obtained from the Southeast Regional
Climate Center for a cooperative National Weather Service (NWS) station nearest each landfH. The
monthly average temperature and total rainfall values were summed and converted to average annual
temperature and total annual rainfall values for each year. The annual temperature and rainfall values for
the years of refuse acceptance were then averaged for comparison to landffll data for each landffll.
In addition to the daily weather data, the 30-year averages of annual mean temperature (NOAA,
1985). mean dewpoint temperature (NOAA. 1979), and annual rainfall (NOAA. 1985) were obtained for
the NWS stations. These 30-year averages of temperature and rainfall are referred to as the 'normal'
values.
Table 5 shows the temperature and rainfall averages and the time periods from which they were
calculated, as well as the 30-year normals for each landfill. The average annual temperatures calculated
from the refuse acceptance time period were usually fairly dose to the normals; however, the rainfall
normals sometimes varied considerably from the refuse acceptance period averages.
28
-------
TABLE 5. SUMMARY OF CLIMATIC DATA FOR THE LANDFILLS
Annual Climate Averages
(duriny refuM aootptanoo poriod)
Landfill
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
20
21
22
23
24
25
Weather
Average
* * i
renoo
1955 to 1989
1968 to 1982
1970 to 1988
1971 to 1988
1951 to 1989
1975 to 1989
1972 to 1989
1973 to 1987
1979 to 1989
1962 to 1989
1962 to 1989
1969 to 1983
1980 to 1988
1976 to 1989
1952 to 1989
1967 to 1989
1971 to 1986
1970 to 1988
1958 to 1986
1972 to 1989
1970 to 1989
Average
Temp.
ro
7.31
9.28
12.59
24.63
17.29
16.14
15.10
16.43
17.99
16.59
18.06
1221
10.67
9.31
8.31
9.23
9.13
724
10.06
19.01
9.53
Maximum
Temp.
rc)
13.13
14.93
17.58
29.68
25.33
23.94
21.23
21.98
24.42
22.74
24.61
17.68
16.49
15.26
13.14
14.47
14.41
12.46
17.79
24.72
14.77
Minimum
Temp.
CC)
1.49
3.53
7.61
19.58
9.26
834
&97
10.89
11.56
10.43
11.52
6.73
4.86
335
3.47
359
334
2.03
232
1329
4.29
Annual
Rainfall
(cm)
75.34
94.34
109.19
140.87
42.16
47.40
107.77
12629
148.31
13729
114.02
12537
103.78
97.64
122X0
8237
8324
73.18
39.01
88.90
94.62
30-Year Annual Averages
Average
Temp.
rc)
7.50
9.28
12.39
23.94
17.11
16.17
15.00
16.22
18.11
16.78
18.17
12.17
11.06
9.50
8.22
9.22
922
7.06
10.17
18.89
9.56
Average
Dewpoint
ro
333
3.89
6.11
18.33
10.00
833
8.89
10.00
1222
10.56
11.11
6.67
6.67
^78
1.67
3.89
339
1.11
-222
11.11
339
Annual
Rainfall
(cm)
73.05
90.47
10521
155.91
4323
45.44
106.07
123.47
134.16
138.48
11354
12254
10432
102.01
12050
78.66
78.66
6655
38.89
7430
84.68
29
-------
Dewpoint temperature was included In this analysis because It is a readiy available variable that
provides a better measure of moisture avaiabifty than either tempetature or precipitation alone. Better
composite variables could be chosen (such as actual evapotranspiration). but calculating these values
was beyond the scope of this project
STATISTICAL ANALYSES
Data Reduction
In order to reduce the number of predictive variables tested, several summary descriptive analyses
were performed. The first analysis examined correlations between possible CH* recovery rate predictors
and CH4 recovery rates and also between the many landfill parameters. Correlation is a measure of the
degree to which variables vary together or a measure of the intensity of association. The resulting
correlation coefficient, r, is bounded below by -1.0 and bounded above by +1.0. A correlation
coefficient of +1.0 or -1.0 indicates that one variable can be expressed exactly as a linear function of
another variable, depending on whether the two variables are directly or inversely related, respectively.
When the correlation is small, r is near zero. Correlation is calculated as follows:
x - xf * E (K - y )*
where: x and y are the sample means of x and y.
Table 6 shows the Pearson correlation coefficients between average Cfy recovery rates and
recovery rates per unit mass with average and normal weather data, as well as other landfill parameters
for the 21 landfills. The significance level of the correlation value Is determined by calculating the
probabllty that the true value is equal to zero.
Three Cfy recovery rate correlations-mass of refuse, landfin volume, and landtn depth-were
significant at the 95-percent confidence level. This means that a 5% chance exists that a correlation this
high could have happened by chance. A fourth 014 recovery rate correlation-number of wells-was
significant at the 90-percent confidence level. None of the OUreccvery rate per unit mass correlations
were significant The correlation coefficients for the significant CH4 recxD^ry rate correlations show a
moderate amount of correlation and reflect expected relationships between the CH4 recovery rate and
some of the landfill parameters.
30
-------
TABLE 6. CORRELATION COEFFICIENTS OF METHANE FLOW VARIABLES WITH LANDFILL
PARAMETERS AND SUMMARIZED WEATHER DATA
Dependent Variables
Independent Variables
Average Annual Temperature"
Normal Annual Temperature
Average Annual Rainfall*
Normal Annual Rainfall
Mean Age of Refuse"
Number of Wells
Refuse Mass
Mean Depth of Landfill
Volume of Landfill
Mean Well Depth
Average Methane Average
Recovery Rate
0.27
0.24
0.14
0.23
0.13
0.40C
0.71 d
0.73d
0.66"
0.29
Methane Recovery Rate
per Unit Mass
-0.07
0.06
0.07
0.10
-0.12
0.09
0.25
0.18
-0.06
-0.19
Averaoe calculated from neriod of mfusa *rx»at»ne» at Meh lanrifill
b Mean age of refuse based on recovery data date.
6 Correlation coefficient significant at 90 percent confidence level.
d Correlation coefficient significant at 95 percent confidence level.
31
-------
It is likely that many of the landfill parameters are correlated. For example, the amount of refuse
placed In a landfill (refuse mass) should correspond to the depth of the landfiN. In fact, refuse mass was
found to be the only landffll parameter that was significantly correlated with an the other landfill
parameters. Table 7 shows the Pearson correlation coefficients between refuse mass and the other
landfill parameters. Table 8 shows the Pearson correlation coefficients between the 30-year averages
and refuse acceptance period averages for the weather variables.
Scatter Plots of Selected Variables
visual inspection of data plots is a useful method of identifying functional relationships. Scatter
plots* for two correlated variables show the general relationship or trend between the two variables and
how well defined that relationship is. Scatter plots for CH4 recovery rate versus landfill depth, refuse
mass, and landfill volume are shown in Rgures 1,2. and 3, respectively.
Although no linear correlations between CH4 per ton and the independent variables were found,
nonlinear relationships are possible. Methane flow per ton was plotted against refuse age (Figure 4),
average annual rainfall (Figure 5). and normal annual dewpoint temperature (Figure 6). No functional
relationships could be identified from these plots, although the dewpoint temperature plot shows a slight
linear trend. Figure 7 shows the location of each site on a grid of temperature and precipitation
normals. The sample population is fairly representative of the range of climates within the continental
United States; however, the more extreme climates are underrepresented.
MODEL DEVELOPMENT
The final step in this study was to develop the simplest model that could be used to predict CH4
emissions from landfills. Because information on refuse mass is readBy avaBable on a global scale, the
preferred method for predicting Cfy was based on this variable, although other variables were also
considered.
Based on the preliminary data analyses (see Table 6), a linear model appeared to be sufficient to
model CH4 recovery rate. The scatter plots of CH4 per ton did not reveal any obvious functional
relationships. However, a linear trend seemed to be present in the climate variable data. The
relationship between Cfy generation and climate is a strongly held belief among gas recovery modelers.
Therefore, some regression analysis were also attempted for these variables.
*For clarity, letters are used to identify landfills in plots. See Table 4 to match letters to number
identification.
32
-------
TABLE 7. CORRELATION COEFFICIENTS BETWEEN LANDFILL PARAMETERS
AND LANDFILL REFUSE IN PLACE
Dependent Variable
Independent Variables Refuse Mass
Refuse Age 0.38*
Number of Hectares 0.54b
Landfill Depth 0.58b
Well Depth 0.58"
Number of Wells 0.37s
Landfill Volume 0.69s5
a Correlation coefficient significant at 90 percent confidence level.
b Correlation coefficient significant at 95 percent confidence level.
TABLE 8. CORRELATION COEFFICIENTS BETWEEN WEATHER VARIABLES
Normal Rainfall
Average Rainfall
Normal Temperature
Average Temperature
Normal Dewpoint
Normal
Rainfall
1.00"
0.98*
0.39"
0.40b
0.54'
Average
Rainfall
1.00-
0.36
0.36
0.52"
Normal
Temperature
1.00-
0.99"
0.94'
Average Normal
Temperature Dewpoint
1.00*
0.94' 1.00*
* Correlation coefficient significant at 95 percent confidence level
b Correlation coefficient significant at 90 percent confidence level
33
-------
100
80 -
I
60 -
40 -
U
20 -
B X
W
~
10
20 30 40 SO
Depth (meters)
60 70
Figure 1. Methane Recovery versus Average Landfill Depth
-------
100
80 -
c
E
E
00 -i
40 -
o
O)
20 -
U
B(
M
W
S 10
Tons of Refuse (millions of metric tons)
15
Figure 2. Methane Recovery versus Refuse Mass
-------
too
80 -
n -
40 -
I
20 -
M
1
H
L
U
W
I
5
I
10
1
15
20
1
25
Volume (million m3)
30
30
Figure 3. Methane Recovery versus Landfill Volume
a
-------
9.8
8.4 -
7.0 -
1
^42"
o
0) 2.8 -
1.4 -
u
K
C
H
Q
Q
W
10
Average Refuse Age (years)
1
15
20
Figure 4.
Recovery Rate per Unit Mass versus Refuse Age
-------
9.6
8.4 -
S.
u -
1.4 -
-5
W
U
T
M
K
H
Q
L X
E
,
0
-1 T-
S 10
Dewpoint Temperature (°C)
T
is
Figure 5. Methane Recovery per Unit Mass versus
Normal Annual Dewpoint Temperature
20
-------
10
7.5 -
D
(5
a.
5 -
o
9
I
2.5 -
T
10
"T
20
W
U
M
K
H
"T
30
1
40
1
50
1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 T
60 70 80 90 100 110 120 130 140 150
Annual Average Rainfall (cm)
160
Figure 6. Methane Recovery Rate per Unit Mass versus Annual Average Rainfall
to
-------
25
20 -
£1.
K
H
10 -
W
C
M
IP Y B
5 -
1 I T I I I I I I I I
10 20 30 40 50 60 70 80 90 100 110
Precipitation (centimeters)
Figure 7. Climatic Normals for the Landfills
120 130 140 150 160
-------
Regression Analysis
The SAS* (Statistical Analysis System) regression procedure (PROC REG) was used to generate
regression statistics for various models. Two general models were used-one to predict CH4 recovery
rate, the other to predict CH4 recovery rate per unit mass. Selection of variables for the regression
models was based on the results of the correlation and scatter plots summaries discussed above. In
addition, the data distribution of potential regression variables was examined tor normality. Although
most variables were not normally distributed, the distributions were not so far off as to warrant data
transformations.
Landfill Air Emissions Estimation Model
In order to validate the statistical model, its performance was compared to that of the U.S. EPA's
Landfill Air Emissions Estimation Model (Pelt et al., 1990), which is a deterministic computer model that
was developed for regulatory purposes. Assuming that the refuse has been accepted at the same
annual rate over time (i.e., all submasses are of the same size), the model equation is as follows:
QCH4 - LO R {«P(-te) - exp(-kt)}
where:
QCH4 * methane generation rate at time t, ftS/yr*
LO - potential methane generation capacity of the refuse, ftS/Mg refuse
R = average annual refuse acceptance rate during active life, Mg/yr
k = methane generation rate constant, 1/yr
c = time since landfill closure, year (c = 0 for an active landfill)
t = time since the initial refuse placement, year
The Landfill Model methodology is based on the Scholl Canyon model (EMCON, 1982) which is a
first order decay equation. Because site-specific characteristics are required as model input, the Landfill
Model is impractical for use on a global scale.
Two critical input parameters, k and LQ. are not known for individual sites. The CH4 generation
rate constant, k. is assumed to be a function of moisture content, nutrient availability, pH. and
temperature. The default value of 0.02 yr1 was used for k. The CH4 generation potential, LQ, is an
estimate of the total amount of CH4 that will be produced by a given mass of refuse as the refuse
'Conversion chart for metric/U.S. equivalent is provided on page 91.
41
-------
decomposes. Lg is partly a function of refuse composition, but It is mitigated by landfill characteristics.
The model default is 298 m3 CH4/Mg refuse* Published landfill waste CH4 recovery potentials for
individual sites in the U.S. range from 50 to 162 rrP/Mg (Ham and Barlaz. 1987; Augenstein and Pacey.
1990). The published ranges were included In this study since the model default waste CH4 potential
was designed specifically for regulatory national impacts assessment Therefore, three sets of model
runs were performed using LQ values of 50,162, and 298 rrr^/Mg in order to determine the sensitivity of
the model to LQ values.
*The model actually uses units of ft3/Mg, so default LQ specified in the model is 8120 ft3 CH4/Mg
refuse. Metric units are used in text for consistency.
42
-------
SECTION 4
RESULTS AND DISCUSSION
REGRESSION MODELS
Table 9 shows the results of several linear regression models. For most of the models that use a
single landfill parameter, the intercept term was found to be insignificant This was expected because
when the mass of refuse is zero, no CH4 is produced. Therefore, a second regression model was
calculated which forced the line through the origin (e.g., a no-Intercept model). It should be noted that
the R2 term in a no-intercept model cannot be interpreted in the same manner as in a model with an
intercept (Rawlings, 1988); therefore, both intercept and no-intercept models for the same variable are
shown for several cases in Table 9.
and Waste Variables
From the regression model results shown in Table 9, landfill depth appears to be the best
predictor of Cfy emissions (P = 0.0002, R2 = 0.53). However, refuse mass is very nearly as good
(P = 0.0003. R2 = 0.50). Depth and tons were shown to be linearly correlated (Table 7). This is not
surprising, as larger landfills tend to be deeper and hold more mass. A bivariate regression of Cfy
emissions on both refuse mass and landfill depth gave a better R2 than any of the unh/ariate models as
show in Table 9. None of the regressions on climate variables were significant and the R2 values were
very low (near zero). Therefore, the best empirical model based on these data for predicting CH4
recovery is:
CH4 = 2.06 + 0.15
where:
CH4 = methane flow rate (nP/min);
W = mass of refuse (106 Mg); and,
D = landfill depth (m).
However, the purpose of this analysis is to develop a model for predicting CH4 emission for a large
population of landfills. Landfill depth is not known for most landfills, and in many cases, is difficult to
estimate. Because waste production data are much more widely available than landfill depth data on a
43
-------
TABLE 9. LANDFILL REGRESSION SUMMARY
Regression Model*
methane depth
methane depth
methane 10* Mg
methane 10° Mg
methane volume
methane volume
methane wells
methane wells
methane - depth + 108 Mg
methane depth + 10* Mg
methane 10* Mg + mean rain
methane 10* Mg + mean temp
methane - 10* Mg + dewpolnt 30
methane/Mg mean rah)
methane/Mg - mean temp
methane/Mg - dewpolnt 30
Prob >F
0.0002
0.0001
0.0003
0.0001
0.0011
0.0001
0.0701
0.0001
0.0001
0.0001
0.0011
0.0015
0.0009
0.7688
0.7607
0.6127
R2
0.53
-
0.50
-
0.44
-
0.16
-
0.65
-
0.53
0.52
0.54
0.00
0.01
0.01
bO
1.09
-
1.89
-
7.38
-
6.87
-
-5.95
-
10.31
4.67
2.98
4.48
6.64
4.61
b1
9.13E-1
8.84E-1
4.27
4.52
1.37E-6
1.73E-6
3.06E-1
4.07E-1
2.36
2.056
4.32
4.11
3.97
6.19E-3
4.21E-2
7.03E-2
b2 Comments
Intercept not significant
no Intercept In model
Intercept not significant
no Intercept In model
Intercept not significant
no intercept In model
model fit & wells borderline; Intercept not significant
no Intercept In model
0.18 Intercept not significant
0.15 no Intercept In model
1 .22E-1 intercept & mean rain not significant
5.61E-1 intercept & mean temp not significant
9.49E-1 intercept & dewpolnt 30 not significant
poor model fit; mean rain not significant
poor model ftt; mean temp not significant
poor model ftt; dewpolnt 30 not significant
Methane - bO + bl * variable 1 * b2 * variable 2.
-------
global basis, the no-intercept regression of CH4 on refuse mass is the better choice of model. This
model is:
CH4 - 4.52W
where: CH4 and W are the same as defined above.
Figure 8 shows the regression line for CH4 recovery rate as a function of refuse mass. The
95-percent confidence interval of the regression coefficient is shown by the dashed lines.
Average refuse age was not shown to have a significant effect on the CH4 recovery rate. To
evaluate the effect of refuse age in more detail, the mass of refuse for each landfill was assigned to three
age groups: 1 to 10 years, 11 to 20 years, and greater than 20 years. A multivariate regression of CH4
recovery rate on refuse mass in the three age groups was calculated. No single age group was
significant, although the F value of 4.194 for the overall model was significant (Prob > F = 0.0164). The
significance level is higher for the wastes between 11 and 20 years of age. as is shown in Table 10.
Although not significant, the somewhat better correlation of CH4 recovery rates with wastes older than
10 years may be indicative of the generation time. Generation times of 20 to 30 years are generally
assumed for landfill gas recovery (Augenstein and Pacey, 1990). This analysis suggests that generation
times are at least 20 years if not longer.
Climate Variables
No other variables were found to have any effect on CH4 production. In particular, no functional
model linking CH4 production to climate variables was found. This does not mean that climate is not
important. Given the unexplained variability in the regression of CH4 recovery on refuse and depth,
some aspect of climate may actually play a controlling role. However, as shown in this study, site-
specific factors and difficulties in accurately quantifying key parameters confound the problem.
The effects of climate-particularly precipitation-are thought to be significant on gas generation
and recovery by many landfill experts (e.g., EMCON, 1982; Augenstein and Pacey, 1990). Although the
cover on sanitary landfills is designed to be impermeable, cracks in the cover are common and may
allow some infiltration of water. Also, during the period when waste is being deposited in landfills, no
barriers to precipitation exist. When the cap is put in place, that moisture will be trapped below the
cover.
For these reasons, it was thought that landfills in very wet climates should show greater methane
generation than those in dry climates. However, no statistically significant correlations were found in this
45
-------
100
80 -
»^M
I
60 -
40 -
..-e
20 -
U
/>^
F ,.--^^.... *
.-;>^..- Q
W
5 10
Tons of Refuse (millions of metric tons)
Figure 8. Methane Recovery Regression with 95% Confidence Interval
of Regression Coefficient
15
§
-------
TABLE 10. SIGNIFICANCE OF REFUSE AGE
Refuse Age Category
(years)
Average Mass of
Refuse in Category
(106 Mg)
Total Mass of Refuse
in Category
(106 Mg)
Prob > |TI'
1-10
11-20
>20
2.78
2.01
3.74
58.28
42.25
78.52
0.281
1.598
1.328
0.782
0.129
0.202
* Student's T value is calculated by dividing the parameter estimate (or slope) for the variable by the
standard error. The null hypothesis that the parameter is zero is tested.
**The probability of a T statistic larger than the one actually obtained occuring if the parameter estimate
(slope) is actually zero. The smaller the probability, the greater the likelihood that the variable exerts a
strong effect on the dependent variable.
47
-------
study. The Ngh degree of variabflity in these data makes It impossible to detect any but very strong
correlations (such as the effect of refuse mass).
One method for removing any confounding effects In a muhh/ariate analysis is the use of partial
regression leverage plots (Rawlings, 1988). The residuals of a regression model measure the
discrepancy between the model predictions and the data. The residual for each landftt, i, Is calculated
by:
*, - Y, - Y,
where:
YJ - the actual value of landfH I, and
Y, - the predicted value for landfill L
The residuals from the regression of CH4 recovery rate on refuse mass and landfill depth were
plotted against the residuals from a regression of mean rainfall on refuse mass and depth. The
regression of these residuals is not significant (F * 2.496. Prob > F « 0.13); however, this is much
closer to sionfficance than the regression of Cfy recovery rate per ton on mean rain (Prob > F - 0.77,
see Table 9). Furthermore, the trend is positive (slope of regression is 0.35) which suggests a positive
correlation between methane recovery and annual precipitation. Unless an additional confounding
variable can be Identified and quantified, no statistically significant relationship between CH4 recovery
and dimate can be established with these data. However, a larger data set may allow detection of
climate effects even if no other independent variables are included.
EFFECT OF DATA QUALITY ON ANALYSIS
The greatest data uncertainty concerned the refuse Itself. As shown in Table 3, refuse mass In
place was sometimes difficult to quantify accurately. The best data came from landfills that weighed
incoming refuse at the gate (and had done so from the day they opened). However, not all landfills had
scales; these sites generally had volume estimates, which had to be converted to mass. Moreover, the
quality of recordkeeping varied widely from site to site. Despite these concerns, refuse mass was the
most reliable of the site capacity variables.
LandfU depths and acreage (both of which were used to calculate volume) were less reliable.
Depth at some sites was quite variable; acreage was unreliable because the topography of sites ranged
from very flat to pyramid shaped. Therefore, it is not surprising that landffll volume is a poorer predictive
variable (as measured by significance level and R2 value) than tonnage.
48
-------
COMPARISON TO LANDFILL MODEL
The results of the Landfill Model runs are shown in Table 11. Also shown are the CH4 emissions
predicted by the regression model and the actual mean CH4 recovered annually. The year for which the
CH4 values apply is shown in the second column.
The ratio of the predicted value to the true mean for each model run was used to compare model
performance. A ratio of 1 shows perfect agreement between the predicted value and the actual value.
The ratios are shown in Table 12. The LandfBI Model with LQ of 50 irP/Mg (Run 1) tends to underpredict
(ratio less than 1). When Lg is set to 162 mfyMg, the model gives the best overall result (the mean ratio
of 1.07 approximates 1). As this analysis shows, the Landfill Model is very sensitive to LQ. No one value
of LO is necessarily the best; for some landfills, the closest agreement was given by the default LQ
(sites 1, 4.11,13, 21, and 22). The lowest LO was best only for two sites (23 and 25). For the
remainder of the landfills, the closest agreement is given by an LQ of 162 rr^/Mg.
The regression model performs reasonably well compared to the Landfill Model. The regression
model's mean ratio of 1.39 falls between Landfill Model runs 2 and 3. It also produces slightly more
variable results, as shown by comparing CVs (78.8 percent for the Landfill Model, 89.3 percent for the
regression model). Thus, the regression model represents the considerable variability of the sample
population and should provide a good estimate of the expected CH4 recovery rate from U.S. landfills as
a group.
One particular advantage of using this statistical model is that only one variable is required.
Furthermore, it is relatively easy to add new observations and further refine the model, as only average
CH4 recovery and refuse mass are required. The confidence limits of the regression coefficient can be
used to bound estimated CH4 emissions. The upper and lower 95-percent confidence limits are 6.52
and 2.52 m3 CH4/min/Mg refuse, respectively.
49
-------
TABLE 11. ACTUAL AND PREDICTED METHANE VALUES
Site
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
20
21
22
23
24
25
Run 1 at LO
Run 2 at LQ
Run 3 at LO
Year
1989
1990
1990
1990
1990
1989
1990
1989
1991
1990
1990
1987
1990
1990
1990
1990
1990
1989
1990
1990
1991
, = 50 m3/Mg
= 162 m3/Mg
= 298 m3/Mg
Actual
Methane
(m3/min)
55.3
18.0
40.0
98.4
24.8
16.7
9.7
11.7
7.7
29.3
11.3
8.0
10.4
16.0
13.8
35.0
27.4
33.2
22
17.7
20.2
(model default)
Landfll Air Emissions
Estimation Model*
(m3/min)
Run 1
8.83
8.73
11.37
21.54
14.34
4.01
4.50
4.37
2.78
7.46
2.57
4.29
1.64
5.28
6.77
14.30
3.98
6.27
3.89
9.47
16.68
Run 2
22.08
21.83
28.44
53.90
35.85
10.02
11.25
10.94
6.96
18.66
6.43
10.72
4.10
13.21
16.93
35.77
9.95
15.69
9.72
23.68
41.72
Run 3
40.65
40.19
52.35
99.18
66.01
18.45
20.71
20.14
12.82
34.35
11.85
19.73
7.54
24.32
31.16
65.85
18.32
28.88
17.89
43.59
76.78
Regression
Model
(m3/min)
28.67
27.71
32.88
60.06
47.81
11.98
14.31
13.36
8.76
24.08
9.53
13.59
5.91
16.15
23.69
42.84
12.85
18.64
13.97
28.06
46.82
50
-------
TABLE 12. COMPARISON OF MODEL PERFORMANCES
Landfill Air Emissions Estimation Model
Site Number
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
20
21
22
23
24
25
Mean
Standard Deviation
Run 1
Pred./Actual
0.16
0.48
0.28
0.22
0.58
0.24
0.46
0.37
0.36
0.25
0.23
0.54
0.16
0.33
0.49
0.41
0.15
0.19
1.74
0.54
0.82
0.43
0.34
Run 2
Pred./Actual
0.40
1.21
0.71
0.55
1.44
0.60
1.16
0.93
0.90
0.64
0.57
1.34
0.39
0.82
1.23
1.02
0.36
0.47
4.35
1.34
2.06
1.07
0.85
Run3
Pred./Actual
0.73
2.23
1.31
1.01
2.66
1.10
2.14
1.71
1.67
1.17
1.05
2.47
0.72
1.52
2.26
1.88
0.67
0.87
8.00
2.46
3.79
1.97
1.56
Regression Model
Pred./Actual
0.52
1.55
0.83
0.62
1.95
0.73
1.50
1.15
1.15
0.83
0.85
1.72
0.57
1.02
1.73
1.24
0.47
0.57
6.32
1.60
2.34
1.39
1.24
51
-------
SECTION 5
SUMMARY AND CONCLUSIONS
The research program described In this report and hi the piot study report (Campbell et al., 1991)
had as its goal the development of an empirical model of CH4 emissions from landHls. It was
successful in meeting its major objectives, but much remains to be learned. Some of the successes and
limitations of the research methodology are discussed below.
EVALUATION OF THE METHODOLOGY
The piot study included some gas sampling and analysis procedures that were not used in the
rest of the program. The cost of gas sampling at each landfill would have limited the number of sites
that could have been included in.the study. In order to get as large a sample size as possible, only data
avalable from landfll personnel were used. This approach Is admittedly subject to error, but the plot
study showed that 0(4 measurements from on-site monitors were reasonably accurate (generally within
10 percent of the reference method). Moreover, no methods could be found to measure the variables
that were most subject to error (landfll size, refuse age, and refuse mass in place). Early in the
program, an intensive sampling effort involving analysis of core samples from the landfills was
considered; however, the expense of this procedure would have severely limited the sites that could be
included. Furthermore, the usefulness of the results to a global emissions model was doubtful.
The field program required the cooperation of both landfll operators and gas recovery operators.
At some sites, local government officials were also Involved. The logistics of arranging and completing
site visits consumed a large portion of the project resources. Sites were screened initially by telephone.
If the site operator was willing to participate in the program, as much information as possible was
obtained prior to visiting the site (the set of information needs is Included in Appendix D). In many
cases, this information could be used to delete sites from the proposed list
Initially, 43 landfills were considered for inclusion in the program. These sites were chosen to
represent a wide range of climates. Some sites were quickly eliminated because operators did not
choose to participate. Others were eliminated for a variety of reasons, such as lack of critical data
records, or the fact that the landfill's gas recovery system was either very new, gas recovery was not
being optimized, or some gas was not being measured.
52
-------
Despite this Initial screening, four of the 22 landfills visited had to be dropped from the final
analysis because of data Inadequacies. Three other landfills were added but not visited. For these three
sites, the landfill operations were known to the project team, and the operators were also fairly
knowledgeable about the program and Its objectives; therefore, these data were considered reliable and
included in the study.
Given the above limitations, this study was successful and has added to the general knowledge
base for landfills. Moreover, the main objective-developing a model for global emissions-was achieved.
The strengths and successes of this program Include:
A model was developed that accurately reflects real world variability of landfill CH4 recovery;
The model Is very simple and easily adapted to global emissions estimation;
The uncertainty associated with Cfy recovery was quantified; and,
The program was cost-effective, allowing maximization of sample size.
The weaknesses of this approach are:
The model Is not mechanistic, and Is therefore limited In its usefulness.
Between-site variability is high, and much of the variability remains unexplained by the
model.
Recovery is used as a surrogate for emissions. The validity of this assumption is unknown.
From here, the research program could take several parallel paths. It would be useful to continue
to add sites and refine the model. However, this approach has limited usefulness and should not take
priority over other research areas. The limitations on the available data, particularly refuse mass and
composition, make it unlikely that this method wOl ever be successful at quantifying the effects of all the
controlling factors. The exception may be climate, but a large number of observations are likely to be
required In order to detect any effect. In particular, more observations from the climatic extremes (cool
dry and warm wet sites) are needed.
One area that needs further study is the use of Cfy recovery as a surrogate for emissions. This
Issue can be addressed by independently measuring total 0)4 gas production and by measuring landfill
gas emissions and analyzing gas constituents. A method of measuring total gas production by pressure
53
-------
probes is currently being developed/ An intensive program of measuring landfill gas
emissions is needed; multiple samples from multiple sites will be needed in order to fully
characterize U.S. landfills.
A GLOBAL MODEL
A factor for estimating global landfill CI-U emissions can be proposed based on the ChU
recovery rate per refuse mass regression model. The intercept was not significant
(Table 9), so the simpler model (with the line forced through the origin) can be used. The
slope for this line is 4.52 m3 CI-U per min/106 Mg of refuse. If all of the waste is assumed
to decay in 25 years, the following emission factor can be calculated:
4.5 m3 ChU/min/106 Mg of refuse * 60 min/hr * 8760 hr/yr * 25 yr
- 5.9 x 107 m3 ChU/106 Mg of refuse
On a global basis, this factor may overestimate ChU production for many countries. The
composition of wastes from less-developed countries in particular is lower in paper and
therefore less likely to produce CHU. Also, global landfilling practices vary much more than
those of the sample population of U.S. sites. On the other hand, if waste decays more slowly
than assumed in this study (25 years), then this factor underestimates CH4 per ton of
refuse.
Despite these concerns, the ChU potential factor developed in this study should yield
more reasonable estimates of global landfill methane emissions than are currently available
because the factor is based on actual landfill data rather than theoretical models. By careful
consideration of all the mitigating effects, some of which have been discussed in this report,
'Memorandum from Stan Zison, Pacific Energy, to Susan Thomeloe, U.S. Environmental
Protection Agency. April 26, 1991.
54
-------
is based on actual landfill data rather than theoretical models. By careful consideration of all the
mitigating effects, some of which have been discussed in this report, this simple model can be used to
quantify and reduce some of the uncertainty in global estimates.
55
-------
REFERENCES
Augenstein, D. and J. Pacey. 1990. Modeling Landfll Methane Generation. Presented at the
International Conference on Landfill Gas: Energy and Environment. October 17.1990. Bournemouth,
England.
Bingemer. H.G. and P.G. Crutzen. 1987. The Production of Methane from Solid Wastes. J. Geophys.
Res. 92: 2181-2187.
Campbell. 0.. D. Epperson. L Davis, R. Peer, and 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-
101351).
Cicerone, R.J. and R.S. Oremland. 1988. Biogeochemical Aspects of Atmospheric Methane. Global
Biogeochemical Cycles 2: 299-327.
Emcon Associates. 1982. Methane Generation and Recovery from Landfills. Ann Arbor Science
Publishers. Inc., Ann Arbor, Ml. 135 pp.
Ham. R.K. and MA Barlaz. 1987. Measurement and Prediction of Landfill Gas Quality and Quantity.
Presented at ISWA Symposium "Process, Technology, and Environmental Impact of Sanitary Landfill,'
Cagliari. Sardinia. Italy. October 20-23.
Khali, MAK. and RA Rasmussen. 1990. Constraints on the Global Sources of Methane and an
Analysis of Recent Budgets. TeHus.42B: 229-236.
National Oceanic and Atmospheric Administration. 1979. Climatic Atlas of the United States. National
Climatic Data Center, Ashevfle. NC.
National Oceanic and Atmospheric Administration, 1985. Climates of the States. Gale Research Co..
Detroit, Ml.
Peer, R.L. A.E Leininger. B.B. Emmel. and S.K. Lynch. 1991. Approach for Estimating Global Landfill
Methane Emissions. Prepared for Air and Energy Engineering Research Laboratory. U.S. Environmental
Protection Agency, Research Triangle Park, NC. EPA-600/7-91-002 (NTIS PB91-149534).
Pelt. W.R.. R.L Bass. I.R. Kuo. and AL Blackard. 1990. Landfill Air Emissions Estimation Model User's
Manual Prepared for Air and Energy Engineering Research Laboratory, U.S. Environmental Protection
Agency. Research Triangle Park. NC. EPA-600/8-90-085a (NTIS PB91 -167718).
Rawlings, J.O. 1988. Applied Regression Analysis: A Research Tod. Wadsworth and Brooks/Cole:
BelmontCA. p. 265.
Shine. K.P.. R.G Derwent. D.J. Wuebbtes, and J.J. Morcrette. 1990. Radiative Forcing of Climate, pp.
41-68 In J.T. Hougnton. GJ. Jenkins, and JJ. Ephraums (eds.). Climate Change: The IPCC Scientific
Assessment Cambridge University Press, Cambridge. 365 pages.
Thomeloe. SA and R.L Peer. 1990. Landfll Gas and the Greenhouse Effect Presented at the
International Conference on Landfill Gas: Energy and the Environment, October 17,1990, Bournemouth,
England.
56
-------
Thomeloe, S.A. and R.L Peer. 1991. EPA's Global Climate Change Program: Global Landfill Methane.
Proceedings, Air and Waste Management Association Annual Meeting, Vancouver, BC.
Kaldjlan, P. 1990. Characterization of Municipal Solid Waste in the United States: 1990 Update.
Prepared for Office of Solid Waste, U.S. Environmental Protection Agency, Washington, D.C.
EPA-530/SW-9Q-042 (NTIS PB90-215112).
57
-------
APPENDIX A
LANDFILL SITE VISIT REPORTS
Data from 25 landfills were obtained for this report Site visit reports for 12 landfills are included
in this appendix. Site visit reports for Landfills 1 through 6 are contained in a plot study (Campbell
et al.. 1991). Data from Landfills 14.15,18, and 19 were not used in this report because the landfills
either did not maximize gas recovery or they lacked historical refuse/gas recovery data; therefore, these
landfills are not described in the report or in this appendix. LandfiUs 23,24, and 25 were not visited.
The purpose of the site visits was to evaluate the landfil gas collection system and methane
(CH4) gas recovery systems at the landfills, and to obtain historical data on gas flow rates and gas
composition, amount of refuse in place, refuse composition and age, as well as information on other
characteristics of interest Whle on site, the accuracy of instrumentation used to measure gas flow
rates, gas composition, and refuse weights or volumes was also evaluated.
The data gathered on 0(4 recovery rates and factors that may influence these rates wfll be used
as inputs for a global model to predict 014 emissions from landfills.
58
-------
LANDFILL 7
Place and Date
State: North Carolina
Date: March 14,1991
Background Information
The landfill covers 125 acres, and is divided into three areas based on past and current refuse
placement patterns. The estimated total amount of refuse in place at this site is 3.25 mBlion tons, with a
total capacity of 6 mBlion tons. Area 1 was the first to be filled, from 1972 when the site opened until
1984. There are an estimated 1.8 million tons of refuse in Area 1. Area 2 was filled from 1984 to 1986,
and from 1989 to present. Prior to 1988, there was approximately 800,000 tons of refuse in Area 2.
Refuse was placed (400,000 tons) in Area 3 from 1987 to 1988.
When the landfill first opened in 1972, approximately 400 tons of refuse were accepted each day.
From 1981 until 1986, refuse acceptance increased to 900 tons per day, and since 1986,1400 tons are
brought in each day. Landfill personnel maintain records on the refuse composition, and copies of these
records wDI be sent to Radian personnel. Construction and demolition debris is included In the total
tonnages reported here, and of the current acceptance rate of 1400 tons/day, at least 40 tons/day are
construction and demolition debris.
Incoming refuse is weighed via a tare system, by which trucks are weighed and assigned a code
number. This code number and the corresponding truck weight can then be called up each time the
truck passes over the scales at the weigh station, and the weight of the refuse is automatically
calculated. The scales are calibrated annually by the North Carolina Department of Agriculture.
The landfill height varies greatly, though eventually all fill areas will be brought up near the
maximum height of 110 feet in Area 2. The maximum refuse height in Area 3 is 90 feet, and Area 1
maximum fill height is 50 feet Some refuse was placed below grade at this site, at depths down to
20 to 30 feet below grade.
There is no knowledge of any commercial hazardous wastes being placed at this site, but
asbestos-containing bunding tPes were accepted until 1985. This material, along with construction and
demolition debris, is placed in areas of the landfill that are separate from the areas that receive
residential garbage. To assure that no hazardous materials are disposed illegally in the landfill, landfill
personnel randomly select one truck each day and inspect the load to make certain that no hazardous
wastes were brought to the landfill.
Closed areas are covered with a 2-foot thick day cap. Only Area 3 has an established
vegetative cover. Because more refuse wil be placed on Areas 1 and 2 in the future, these areas have
59
-------
not been intentionally vegetated. Some vegetation has invaded the exposed sofl. however, such as wfld
grasses and conifers.
Landfill personnel indicated that the dty has plans to iiwlall a oxnposting facility at the landfffl In
the near future. Initially they plan to accept only leaves and grass clippings, but they may eventually
install a chipper for tree limbs.
Landfill Gas Collection System
Gas recovery began at this site in December, 1989. There are 48 wells on the landflU, with
24 wells in Area 1,17 wells in Area 2, and 7 weds in Area 3. The average depth of existing weds is 40
feet but as new weds are installed in higher fill areas they wB be drilled 85 to 90 feet deep. Lateral
polyethylene collection tines move the gas from the wells to a blower station, which then sends the gas
to a boiler located at a nearby nutrition supplement production facility. The landfiU gas is used as boiler
fuel and is the main fuel source for this fadity. These lines are buried below the surface (primarily to
prevent vandalism) and throughout the landfill there are 12 condensate traps on these lines. These traps
are placed In low areas along the lines where condensate collects, and the gravel hi the bottom of the
trap allows the condensate to be fltered back into the landfiN.
At the blower station, more condensate is removed from the gas stream (estimated by gas
recovery personnel to be 75 gallons/day). This condensate is sent to the dty sewer system. The gas
stream is not conditioned in any way before being sent to the bolen this condensate probably forms as
the gas cools. Overall, the boiler is estimated to be 81.5% efficient by gas recovery personnel and the
boler manufacturer. Gas recovery personnel indicate that records are maintained on total gas flow as
wed as steam produced.
There is one flare located by the blower which ts computer-operated to automatically run when
demand for steam from the boSer is low. For example, when demand is greater than 17,500 Ibs/hr, the
flare never runs. As demand falls, however, the computer wU adjust the amount of landfill gas to be
flared proportionally. This system is designed so that the vacuum applied to the wells wtUnotbe
adversely affected by fluctuations in the steam demand.
Gas Sampling Points
Ad gas flow and composition measurements are taken at the blower station after lateral lines
from tandtU Areas 1,2, and 3 combine to form one gas stream. No measurements are taken on flow
from the individual areas, but gas recovery personnel felt that 80% of the 1.8 mllion cubic feet of gas
recovered each day comes from Area 1. Total flow is measured via a differential pressure transducer
connected to a pitot tube In a 12-inch pipe where the gas enters the blower. The amount of gas flared
issimilarly measured in a 4-inch pipe leading to the flare. Trie pressure transducers send flow rate data
to a computer which is located in a separate office (not at the landfill). Gas recovery personnel
indicated that there is no strict procedure followed regarding calibration of the pressure transducers, but
60
-------
they estimated that the equipment is calibrated about twice a year. Gas composition measurements are
taken daily with a Gas Tech combustibles analyzer at a point along the pipe leading to the flare. Percent
methane typically runs from 49 to 53%, and averages 51%. At this point, the gas contains an estimated
2 to 3% water by volume. No details were provided on the frequency of calibration of the gas analyzer.
Leachate Collection
There is no active leachate collection system at this landfill. Around the perimeter of the
property there are ground water monitoring wells, however.
Gas Migration
To monitor off-site gas migration, the city has installed 14 perimeter migration wells which are
tested monthly with a methane gas monitor (Industrial Scientific Corporation, model number CD212).
Landfill personnel indicated that monitored levels never exceed 0.05% methane. Prior to installation of
the gas recovery system, there were frequent odor complaints. Gas recovery personnel indicated that
the complaints ceased when the recovery system was installed. Discussions with gas recovery
personnel indicate, however, that they think they are recovering only 65% of the available methane
(based on amount of refuse in place), and that the majority of gas loss occurs along the steep slopes (of
Area 3, for example) and escapes to the atmosphere without being detected at the perimeter migration
wells.
Information Gathered
No records were gathered the day of the site visit, but discussions were held with landfill
personnel and gas recovery personnel about the type of information that is of interest
Information to be Sent
Landfill personnel were asked to provide Information on the refuse tonnage and composition,
broken down annually. In addition, a contour map of the landffll showing gas recovery wells was also
requested.
61
-------
LANDFILLS
Place and Date
State: Georgia
Date: March 19, 1991
Background Information
The landfffl is inactive and covers 55 acres. Refuse was accepted from 1973 to 1987. There are
an estimated 3 mfllion tons of refuse in the landffll. The waste acceptance rate was approximately the
same for each of the 14 years of operation of the landfU. This Is equivalent to approximately 215,000
tons of refuse per year or 590 tons daly.
Waste composition is estimated to be 60% commercial and 25% residential. The remaining 15%
is construction and demolition debris. A large amount of the commercial waste came from office
buddings and includes white office paper. It te estimated that paper accounts for 50% of the landfill
refuse. No liquids or sludge were accepted during the life of the tandfin.
The landfill was butt in benches; the refuse height varies from 30 feet to 75 feet with an average
depth of 55 feet No refuse was placed below grade at this site. The landfU is covered with a day cap
that is 2 to 4 feet thick The permeablity of the day used In the cap is 10 -6. The landfffl has an
established vegetative cover, mostly grasses with some wld flowers.
Gas Collection System
Gas recovery began at this site in 1986. The end-use of gas from this site Is high Btu. pipeline
quality gas. There are 107 wells connected by 15 lateral lines that lead into the header line. Wells
installed by the original gas recovery system owner are being replaced at a rate of about 20 wells per
year. There are plans to reduce the number of weds by approximately half so that each well is not in
competition with neighboring wells and wll perform more efficiently. The average depth of the wells is
55 feet Replacement weds wil be driUed deeper than original wells in most cases. The lateral lines
move the gas to the header which transports gas to the recovery plant These lines are buried below
the surface.
The gas treatment process is currently considered proprietary, and wll not be discussed in detail
here. Before treatment the CH4 content of the landffl gas ranges from 55 to 59%. Methanol is used to
remove the carbon dioxide (CO2) from the landfU gas; after processing the remaining gas contains at
least 95% CH. During processing, there is an estimated 2% CH4 loss, as some CH4 is removed along
wtththeCO2bythemethanol. There is less than 1/2% C02 In the final sale gas product Processed
gas must meet the specifications of the gas company that purchases the gas. They require a 950 Btu
gas with less than 4% nitrogen. The recovery fadity routinely provides gas with a 960 to 970 Btu
62
-------
content Based on operating records, gas recovery personnel estimate that the recovery plant runs 91%
of the time and processes approximately 1.4 million cubic feet of gas per day, although the system has
a 3.2 million cubic foot capacity. Gas recovery personnel estimate that they are recovering
approximately 75% of the landfill gas produced at this site. This estimate is based on the conditions of
the existing well field and the cap.
At the recovery plant, condensate is removed from the gas and collected. The condensate is
sent through two oil/water separators before being tanked and trucked off-site. The landfill is currently
trying to get a permit to discharge this treated condensate into the city sewer system.
There is one flare located by the gas recovery facBity. The flare can bum 1000 CFM, which is
not big enough to prevent all migration when the system is down. The flare is not used when the plant
is operating.
Gas Sampling Points
Sampling points are located on each lateral line for a total of 15 sampling ports. An Air Data
Multimeter (ADM-870) manufactured by Shortridge Instruments Inc. is used to measure flow rate,
temperature and pressure. A gas chromatograph located on-site is used to evaluate gas composition.
Nitrogen content is routinely checked. Perimeter wells are monitored using an Mine Safety Appliance
(MSA 62-S) which indicates percent lower explosive limit (LEL) and percent methane.
Leachate Collection
There is no active leachate collection system at this landfil. Leachate loss from the landfill is
reportedly well-contained by the natural day liner under the landfil.
Gas Migration
To monitor off-site gas migration, perimeter migration wells have been installed at 1000-foot
intervals around the landfill and are monitored monthly. When the gas recovery plant is not operating,
these perimeter wells are monitored on a weekly basis. The perimeter wells average 30 to 45 feet in
depth. If the plant is not running, the vacuum on the perimeter wells is increased to prevent off-site
migration, and the vacuum on the interior wells is decreased.
Information Gathered
DaBy reports of the amount of gas sold from February 1989 to March 1990 were obtained from
gas recovery personnel. Also included in this data is the average carbon dioxide, nitrogen, methane and
Btu content of the gas.
63
-------
LANDFILL 9
Place and Date
State: Mississippi
Date: March 19. 1991
Background Information
The landfill is active and opened in 1979. It is divided into 3 cells, two of which have been fined
and are capped. These two c^s cover approximately 30 acres of the 180 acres avalable for refuse
disposal There Is an estimated 1.8 mllion tons of refuse in the landfill. Cell 1 was Hied from about
1979 to 1984; cell 2 was filed from 1984 to 1989. Refuse is currently placed in cell 3.
When the landfill first opened in 1979, approximately 1,300 yards/day of refuse was accepted
each day. Over the last 3 years that acceptance rate has increased to at least 3,000 yards/day.
Incoming trucks are charged a fee based on the size of the truck bed, regardless of whether or not it is
Ml. Gas recovery personnel estimate mat there Is a total of 1.8 mllion tons of refuse in place, and
indicated that they would provide a conversion factor between yards and tons of refuse. Construction
and demolition debris is included in the estimates of amount of waste accepted each day, and accounts
for approximately 20% of incoming refuse. The remaining waste is 75% commercial and 25% residential.
The landftt personnel have no knowledge of accepting commercial hazardous waste, gas tanks, or car
batteries. The landfill wBI accept tires if they have been shredded or cut In half.
The landfiK height varies greatly, as SI was placed in benches, with a maximum interior height
of 68 feet Eventually all interior fin areas wB be brought up to this height Refuse was placed 25 to
30 feet below grade.
Closed areas are covered with a 2 to 3-foot thick day cap. Cells 1 and 2 have established
vegetative cover, mostly grasses and some wM flowers.
Gas Collection System
Gas recovery began in September 1990. The system was installed in response to migration
problems. There are a total of 32 wells; 20 of the weds are in cell 1; and 12 weds are in ced 2. Interior
wells are 60 to 70 feet deep and perimeter wells are about 30 feet deep. AH weds are tied to lines that
lead to a blower, with lines from cells 1 and 2 brought in separately. Fifteen deep interior weds are
currently under construction and should be operating by the end of the summer. Nine or ten are
located in cell 2.
There is one flare located on-site. It is a 5 burner IT-McGiH unit with louvres to adjust the
temperature and flame. It also has a fai safe valve that activates If the system shuts down. Vibration
isolation valves can control flow rate to the flare. Two 40 hp blowers are alternately used on a weekly
64
-------
basis. The flare is currently burning about 600 CFM. The gas averages 40% CH4; gas from perimeter
wells has a lower CH4 content The site is flaring the minimum amount of gas necessary to keep the
perimeter wells at 0% methane. The site will maximize gas recovery when untreated, medium Btu gas is
sold to a nearby concrete manufacturer at the end of the summer. At that time landfill personnel
estimate recovery to be 1000 to 2000 CFM.
Gas Sampling Points
Sampling points to monitor flow are located before and after the gas enters the blower bunding.
Only total flow is measured, the first sampling port is located after the lines from cells 1 and 2 are
combined. Gas flow rate, temperature, and pressure are measured with an Air Data Multimeter
(ADM-870) manufactured by Shortridge Instruments Inc. Gas composition is not routinely measured.
Perimeter gas wells are tested with a Mine Safety Appliance (MSA 62-S) methane meter which measures
percent lower explosive limit (LEL) and percent methane.
Leachate Collection
Cells 1 and 2 do not have active leachate collection systems. A leachate collection system is
present in ceil 3. Ultimately, the gas condensate sumps wOl be tied into this collection system.
Gas Migration
To monitor off-site gas migration, perimeter migration wells have been installed at 1000-foot
intervals. These perimeter wells are 35 to 40 feet deep and are tested monthly.
Information Gathered
Landfill and gas recovery personnel provided weekly flow rate testing data for September 1990
through the present In addition, a summary of the test results of canister samples analyzed for gas
composition was also provided. Maps of the site plan, methane gas pipeline plan and profile, and layout
of the gas production lines were also provided.
65
-------
LANDFILL 10
Place and Date
State: Alabama
Date: March 21.1991
Background Information
The landfifl covers approximately 100 acres and is divided Wo two areas based on past and
current refuse placement patterns. Each area encompasses 50 acres. The two areas are divided by a
dirt road leading to the active dumping area. The landffl has been open since 1958. However, the
garbage was burned throughout the 1960's. Unbumed waste has been placed in the ok! and new areas
since 1971. with each rise overlapping previously filed rises. Waste acceptance in the old area ceased
in October 1989. however, the currently active rise may be extended over the old area during 1991.
Waste acceptance tn the new area ceased in early 1990. The new area currently has a grass cover.
Waste acceptance in all areas is expected to cease in the year 2006.
The landfill is permitted to accept approximately 2000 tons of waste per day and there is no total
capacity limitation. During 1990. the landffl received about 1200 to 1500 tons per day. The tipping fee
was recently raised, and they are currently receiving about 600 tons per day. Landfll personnel estimate
the average waste density to be 0.63 tons per cubic yard of air space. An on-site, open-ended, cross
sectional survey was performed over a 30-day period to determine the average density of the incoming
refuse. The average depth of the tendfiB is 110 feet A 2% grade is maintained with the use of lasers
attached to the bulldozers.
Waste has been covered daiy with sol excavated from an area behind the landfill. A three-foot
cap. consisting of sandy day sol, was placed on the new area in 1985.
The landfill began weighing incoming refuse In 1980 with a computerized scale. Trucks were
initially weighed and assigned a code number. This code number and the corresponding truck weight is
called up each time the truck passes over the scales. The weight of the incoming refuse is automatically
calculated. The scales are calibrated quarterly. In addition, the State Division of Weights and
Measurements checks the scale calibration twice a year.
The landfU does not practice waste segregation. There is no knowledge of liquid, hazardous.
and infectious wastes being accepted at the landfiU. Small quantity generator waste and household
hazardous waste are accepted at Subtitle D fadties. however. Tires are stll being accepted. Landfill
personnel conduct spot inspections of incoming refuse periodically and monitor actual dumping two or
three times a week to ensure no prohibited waste is entering the landfU. Asbestos is currently banned
from the landfffl, however. It was accepted in the past The tendfiB maintains a receivables assessment
management system (RAMS) to determine waste composition on a monthly basis. Additionally, a waste
66
-------
characterization study was conducted for the U.S. Environmental Protection Agency in 1986. The results
of this study indicate approximately 50% of total waste originates from household sources, 25% from
commercial sources. 20% from construction and demolition, and 5% from industrial processes. Landfill
personnel suggested using these proportions rather than those generated by the RAMS. The primary
difference is that in the RAMS household wastes are separated into two categories: refuse (23%) and
yard waste (17%).
Landfill Gas Collection System
Landfill gas recovery began in 1988. The methane gas recovery system consists of 96 wells.
The wells draw gas from approximately 80 acres of the landfill. Waste continues to be dumped on top
of 60 of the 96 wells in operation. The gas recovery personnel have had occasional problems with
broken pipe lines and buried wells. The wells are arranged in parallel rows, with 125 feet between each
row. The wells are approximately 70 feet in depth, with the bottom 20 feet filled with rock. The
collection pipe is perforated, starting 7 feet below the ground surface. A concrete cap covers the pipe
end. Gas recovery personnel expect 50 new wells to be drilled in the near future. The average vacuum
draw on the wells is 5 inches of H2O.
The recovered gas stream is treated by a metnanol cooling process to remove H2O,
hydrocarbons, CO2, O2. and Nj. The refined gas stream is pipeline quality gas, containing less than
3 percent CO2. The refined gas is sold to the local gas company via a nearby pipeline. The average
Btu content is 950. Approximately 30 mDlion cubic feet per month of refined gas is sold to the gas
company. Monthly gas records have been maintained since January 1990. The volume of pipeline
quality gas sold is about half the volume of total landfill gas extracted.
Gas Sampling Points
A standard gas meter is used to measure methane gas flow to the pipeline. Raw landfill gas
recovered before treatment is not measured. Gas recovery personnel dalm that the treated gas sales
are 50% of raw gas recovered based on periodic monitoring of raw gas methane composition. Gas
composition is measured every 15 minutes with a Bendix 7000 gas chromatograph after the gas is
treated and before discharge to the pipeline. The gas chromatograph is calibrated dafly with a sample
gas of known composition. The gas recovery personnel sample raw gas composition in each of the
recovery wells on a monthly basis. The raw gas methane content ranges from 50 to 58%. Wells which
draw gas with a methane content below 45% are temporary dosed. Excess nitrogen in the raw landfill
gas occurs periodically due to the breakage of collection pipes and wells from the expansion of the
active dumping area.
67
-------
Leachate Collection
The landfill does not have a leachate collection system and has had no off-site leaching
problems. However, leachate monitoring wells are maintained at the landfill perimeter and checked
twice a year.
Gas Migration
A polyethylene curtain was placed in a 25-foot ditch on the perimeter of the landffil in 1979 to
stop methane gas migration.
Methane gas migration is checked once a year at the landfill perimeter. An Exploslmeter
model 214 is used to check for the presence of methane. The Exploslmeter is calibrated with 95 percent
methane. No vegetative stress was observed.
information Gathered
The landfill personnel provided daly waste acceptance records from 1988 through 1990. A solid
waste landfill survey prepared for the U.S. EPA in November 1986 was also obtained. An aerial
photograph of the landfill and surrounding areas was provided for copying back at the office. A seven-
month waste composition record from July 1990 to January 1991 was also obtained.
The gas recovery personnel provided a schematic diagram of the gas extraction weOs. Methane
gas sales records (in cubic feet) for 1990 were also obtained.
-------
LANDFILL 11
Place and Date
State: Georgia
Date: March 22,1991
Background Information
The landfill covers 120 acres and is divided into two areas based on gas rights. Parcel A covers
69 acres, and Parcel B covers 51 acres. There are no obvious distinctions between the two parcels.
Waste Is placed in rises which are bult progressively over the landfill surface. Each rise overlaps
previously filled rises containing older waste. The landfill is active, and it is expected that waste will be
accepted for at least two more years. Currently, about 10,000 to 13,000 tons of waste are accepted per
month.
The average depth of the landfill is about 80 feet Except for some excavation to obtain landfill
cover, most of the waste has been placed above grade. Less than 10 feet of the waste is placed below
grade. The top of the landfill is slightly rounded and the slopes are steep.
The landfill began accepting waste in 1962. The landfill has accepted nearly all of the residential
and commercial waste from the surrounding county (population 170,000) since the landfill opened. Prior
to 1986. the landfill accepted construction and demolition debris. In 1986, a separate landfill opened
and began accepting nearly all of the construction and demolition debris in the county. Therefore, waste
placed in Landfill 11 after 1986 does not include construction and demolition debris.
Incoming refuse is weighed via a tare system, by which trucks are weighed and assigned a code
number. This code number and the corresponding truck weight can then be called up each time the
truck passes over the scales at the weigh station, and the weight of the refuse is automatically
calculated. The scales are calibrated monthly by a scale company service technician who uses standard
weights to calibrate the scales. The State Department of Agriculture also inspects the scales annually.
The weigh station became operational in 1981.
There is no knowledge of hazardous waste placed at the site (however, household wastes and
small quantity generator wastes are accepted at Subtitle D landfills). Waste is not routinely inspected for
the presence of hazardous wastes.
Medical waste is not accepted at the landfill, but each truck carrying general refuse from the
local hospital is scanned with a geiger counter for the presence of radioactivity.
Sewage sludge from the local waste treatment plant is accepted at the landfill. All sewage
sludge is tested for moisture content using a paint filter test Prior to landfBling. the sludge moisture
must be reduced such that the sludge wOl not pass through a paint filter.
-------
As each rise is completed, it is covered with a 2-foot cap of day and grass is planted. These
areas wffl eventually be covered over with new waste until the landffll reaches capacity. It is estimated
that an additional 20 feet of waste wfll be placed on the landfill before It is permanently closed.
Currently, there are no large-scale recycling or composting programs in the county; however,
there have been some wood chipping demonstration projects.
Gas Collection System
A brick manufacturing company has been collecting gas from Parcel A (69 acres) since 1985.
Currently no gas is collected from Parcel B. A separate company has gas rights to Parcel B. but has
not recovered gas from the landfill since 1989.
On Parcel A, 39 wells placed about 180 feet apart collect landfill gas. The average depth of
each wen is about 45 feet As additional waste is placed on the landfBI, pipe extensions are added to the
wells so that they wBI not be covered.
Lateral polyethylene collection lines move the gas from the wells to a blower station which is
located at the brick factory. The average vacuum at the main landfill gas header is about 5 inches of
water. Landfill gas from the main header is treated by mechanical refrigeration to remove water. The
dehydrated gas is then used as fuel in a brick kin.
About 500 gallons of condensate are collected from the gas each day by the refrigeration
system. About half of this condensate is Injected back into the landfill, and the rest is evaporated by a
boiler.
AH of the gas collected is burned in the brick kin. None of the collected gas Is vented or flared.
Records of the amount of gas collected have been kept continuously since the gas collection began.
Gas Sampling Pohfls
Total gas flow measurements are taken at the brick factory downstream of the condensate
removal system. The gas flow data, therefore, reflects total landfiR gas collected minus water. Gas flow
is measured with a turbine gas flow meter.
The landfill gas is also monitored for methane at a point downstream of the condensate removal
system. Percent methane is measured daly using a hand-held gas monitor. An automatic gas
chromatograph is also installed in the gas line downstream of the condensate removal system, but It has
not operated reliably. Records on percent methane are not maintained. According to the representative
of the brick company, the percent methane is measured daly just to ensure that the gas contains
between 50 and 55 percent methane. Lower methane readings indicate breaks in the gas lines which
signal to the gas collection personnel that repairs are needed. The brick company representative stated
that the landfill gas has averaged 53 percent methane since gas collection began. Other gas
constituents, such as CO2, N2, and O2, are not monitored regularly.
70
-------
Laachate Collection
There is no active leachate coBection system at this landfill. Around the perimeter of the
property, there are ground-water monitoring wells, however.
Gas Migration
Gas migration monitoring is not required at the site, and there is no monitoring program.
Landfill personnel stated that vegetative stress on the landfill surface is not a problem.
information Gathered
Maps of the landfill as well as a map of the gas collection system were obtained. Monthly
records of total gas flow for the period 1985 through 1990 were provided by the brick company.
Information was also provided regarding periods when gas collection equipment was not operating
continuously.
Landfill personnel provided monthly waste acceptance tonnages for the period 1981 though
1990. No records of waste acceptance were kept prior to 1981. However, landfill personnel suggested
that since the landfill served the entire county from 1962 to 1980 (waste was not exported or imported
across the county line), waste acceptance prior to 1981 should follow the trend of the county population.
Information to be Sent
A private consulting firm conducted a study of waste composition accepted at the landfill during
the past 2 to 3 years. Landfill personnel stated that they would obtain and send a copy of that report
71
-------
LANDFILL 12
Place and Date
State: Oregon
Date: Apr! 4. 1991
Background Information
The landfn covers approximately 85 acres and is divided into two segments: a north area and a
south area. Filing began in the south area and progressed northward to the northern boundary of the
landfill between 1969 and 1980. From 1980 until site closure in 1983. waste was placed in a second lift
on the south area. The two areas were initially separated by an east-west drainage ditch, but the ditch Is
no longer present The estimated total amount of waste at the site is 3.1 million tons (1.9 mllkxi tons in
the south area and 1.2 million tons in the north area).
The landfill is buBt on land that was formerly a marshy wetland with an elevation of 20 to 22 feet
The surface elevations after landfill closure ranged from 45 to 58 feet Since 1983. the final landfill cover
has been upgraded to increase slope and water runoff. Currently the landfiU cover consists of an
average of 5.8 feet of low permeablity sol. The slope of the landfn averages just over 2 percent and
grass covers the entire surface.
A 1986 hydrogeological study of Landfn 12 conducted for the EPA estimated that the net
amount of water from precipitation that percolates through the landfiH is about 24 inches per year.
Boring tests have shown that much of the waste near the bottom of the landfiU is water saturated. This
is mainly due to the high water table at the site.
The landfll accepted primarly residential and commercial wastes. The landfn operating permit
specifically excluded medical wastes, chemicals, ois. liquids, septic tank Dumpings, and nondigested
sewage sludge (however, household hazardous wastes were not excluded). Some construction and
demolition debris was accepted at the landfffl, and it was placed in the southern end of the landffll
separate from the general refuse. The amount of construction and demolition debris accepted at
Landfll 12 is not known; however, a representative of the State environmental agency said that the
amount was probably not large since most area construction and demolition debris was taken to a
separate landfill.
Gas Collection Sstem
LandfiH gas recovery began at this site In 1984. There are a total of 78 active wells. About 50 of
these wells are horizontal wells placed in trenches about 6 feet below the landfll surface. These
horizontal trench wells range in length from about 200 to 1.000 feet The remaining 28 wells are vertical
wells with an average depth of 35 feet The perforation in the vertical wells begins about 3 feet below
72
-------
the landfll surface. The wells and the connecting piping are constructed of PVC. Gas wells cover nearly
all of the landfill surface. However, a representative of the landfill said that wells were not located over
the section where construction and demolition debris was placed.
From 1984 to 1989. a private company operated the gas recovery system. The landfill gas was
conveyed by an internal combustion compressor which was powered by landfill gas. The vacuum
pressure at the main landfill gas header was typically 2 to 4 inches of water. whQe the vacuum at
individual wells was about 0.5 inches of water. The total flow of raw landfill gas varied between 300.000
to 800.000 standard cubic feet per day. The gas constituents had the following average composition:
CH4-53%
CO2-42%
N2 - 4%
O2 - 1%
H2O - <1%
Btu's - 530
The raw landfill gas was treated by removing water with a triethylene glycol dehydration system.
The dehydrated gas was then treated with selectively permeable membranes to separate the methane
from the carbon dioxide. The resulting methane portion of the gas stream (95% methane) was sold as
pipeline quality gas via a natural gas pipeline near the landfill. Whenever the gas refining system was
not in operation, landfill gas was diverted to a flare.
By 1989, landfill gas production at the landfill decreased to a point where gas purification was no
longer profitable. Therefore, the company that had been operating the system ceased operation.
Currently, some gas is extracted from the existing well system and used to fuel boilers for the county,
but the system is no longer operating in an optimal fashion.
Gas Row Measuring Points
An 8-inch annubar was installed at the terminus of the landfill gas collection system at a point
where it enters the gas purification system. This measured the inlet flow of raw landfill gas in inches of
water, and was recorded on a 7-day circle chart
Grab samples of raw landfill gas were taken once per week; these samples were analyzed by
GC for methane, oxygen, and nitrogen. These grab samples were also tested for moisture. An
automatic GC was used to continually measure methane, oxygen, and nitrogen In the processed gas
stream. Processed gas samples were measured every 9 minutes. 24 hours a day.
The equipment that was used to measure gas flow and composition is no longer present at the
site; it was removed when the gas refining operation ceased in 1989.
73
-------
Leachate Collection
There is no active leachate collection or treatment system at this landffll. However, ground-water
monitoring wells are located around the perimeter of the landffll.
Gas Migration
Monitoring at the perimeter of the landffll for gas migration was not practiced at this site.
Information Gathered
The operator of the landffll provided monthly tonnages of waste received at the landffll for the
period 1971 through 1983. A study of ground-water impacts at Landffll 12 was also obtained from the
State environmental agency. This study contained annual waste acceptance tonnages for the entire life
of the landffll (1969 though 1963) as well as a map showing the location of waste placement by year. A
study published by the local solid waste authority was obtained which provides a breakdown of typical
waste composition in the area.
The company that operated the gas recovery system from 1984 to 1989 provided records of
total landffll gas recovered for the period April 1987 to October 1987. They also supplied a schematic
diagram showing the placement of wells at the landfflL
Two additional reports were also obtained One is a 1983 report that projects the amount of
recoverable gas from Landffll 12. The other is a 1988 report presenting a risk assessment of the use of a
mixture of landffll gas and natural gas in customer homes.
74
-------
LANDFILL 13
Place and Date
State: Washington
Date: April 5.1991
Background information
The landfill is 50 years old and is scheduled to dose in December 1991. Open burning of refuse
was practiced from the time the landfill opened in the late 1930s until the mid 1960s. After open burning
of refuse ceased, the refuse was either shredded or placed directly in the landfill and compacted. A
layer of demolition refuse lines the bottom of the landfill.
The landfill was bunt on an old gravel mine. The underlying soil is sandy gravel. The refuse is
approximately 50 feet in height from the prior excavation level. The bottom of the landfill is about 5 feet
above the ground water table.
The landfill is divided into two areas. The north section is approximately 50 acres and has been
closed since 1988. Refuse acceptance in the north area began in 1980. The cap for the closed portion
consists of a synthetic membrane (60-mO polyethylene) which was placed on 35 acres of the north
section in 1989. On top of the synthetic membrane is a 1-foot thick earthen cap. The south area is
approximately 30 acres and portions of it are still actively accepting refuse. The south area contains the
oldest refuse. A 6-inch cover of sandy dirt is placed over the active area daily.
The landfill receives approximately 440,000 to 480,000 cubic yards of refuse annually. The
density of the in-place compacted waste is approximately 1200 pounds per cubic yard. The density of
incoming uncompacted waste is approximately 600 pounds per cubic yard. Last year, approximately
134,000 tons of refuse were accepted in the landfU. A weigh station was installed at the landfill in 1988.
Vehicles are weighed once and assigned a code which is entered hi the computer. Each time the
vehicle enters the landfill, the code is punched in the computer. Incoming refuse weight is determined
automatically. Several studies of estimated waste acceptance and in-place volume have been performed
by the landfill's consulting firm and local health district officials. The landfill serves the entire county in
which it is located. The county population is approximately 230,000.
Landfill Gas Collection System
The landfill operated separate methane gas recovery systems hi the north and south sections of
the landfill up until March 1991. Landfill gas recovery began in the late 1970s in the south section and in
1988 in the north section. In the south section, landfill gas was flared using an open candle flare. The
south area flare ceased operating in March 1991 and 20 perimeter wells from the south section were tied
into the north area collection system. The north area flares landfill gas using an enclosed flare. From
75
-------
1988 to February 1991. the north flare combusted only landfffl gas from the north area of the landfill.
When the south area perimeter weds were tied into the north area collection system, the north flare
began combusting perimeter gas from the south area.
The landfill operates a total of 81 vertical gas extraction wells. Twenty wells are located in the
south area and fifty-one are located in the north area. Eight of the twenty wells in the south area are
located in the active dumping area. Each wen is approximately 30 to 50 feet in depth. Twenty-five new
wens wil be installed in the currently active area at landfill closure. The average vacuum pressure on all
wens is 34 inches of water column measured at the main header before the blower and 5 inches at each
individual weft. Condensate is collected from the gas recovery system and pumped via nine pneumatic
pumps to a treatment tank. The condensate is filtered with carbon and pH adjusted before discharge to
the local sewage treatment plant
Gas Sampling Points
Gas composition is monitored in each of the gas extraction wells every two weeks using a hand-
held gas analyzer. Raw landfifl gas methane content averages between 40 and 50% by volume. Gas
flow is measured in the main header to the blower on a daly basis. Based on the measurements, landfill
personnel reported that the average gas flow rate is 835 cubic feet per minute. Daly flare temperature
readings are also recorded. The average flame temperature of the flare was reported to be 1875* F.
Leachate Collection
The landfn .does not have a leachate collection system. Ground-water contamination from
landfill leachate has been documented. The bndflB owners are currently building a water treatment plant
which will commence operation at the end of the year. The contaminated water from the underlying
aquifer wfll be extracted, treated, and returned to the ground.
Gas Migration
Gas migration is measured by 20 gas probes located on the perimeter of the landfill. The gas
probes were installed in 1979 when methane gas recovery began. Very little gas migration has been
detected since the recovery system was installed. However, substantial amounts of landfill gas was
observed bubbling In puddles along the road to the active dumping area. The road is located between
the north and south areas.
Information Gathered
Landfill personnel provided a topographic diagram of the landfiN and a schematic of the gas
recovery system. Also provided were monthly records of gas mkjratkxi probe readings and gas
composition measurements taken at Individual wells during 1990 and 1991. The landfill's consulting firm
76
-------
sent gas flow rate records for 1990, and the local district health office personnel provided data from
several landfill studies on waste composition.
Information to be Sent
Waste acceptance records maintained by the landfill were requested.
77
-------
LANDFILL 16
Place and Date
State: New York
Date: Apr! 10,1991
Background Information
The landfll covers 75 acres and refuse averages 80-100 feet deep and 120 feet deep at the
highest point Refuse was placed 20-30 feet below grade. Waste acceptance began in 1976 and the
landfD currently has approximately 3,700,000 tons of refuse in place. This tonnage estimate is for
municipal solid waste only, and does not include tires, construction and demolition debris, or ash.
Approximately 800 tons of refuse are accepted each day, although the amount varies with season.
Refuse acceptance may be as high as 1500 tons per day in the spring. Trucks hauling refuse are
weighed as they enter and leave the landfll. All white goods, wood, tires, and construction debris are
separated out of incoming refuse. Sewage sludge is accepted and is placed on the south slope of the
fiP area. Asbestos is also accepted and placed in a separate area of the landfill. There is no final cap
on any part of the landfll now. but there are plans to add a cap in the near future.
Landfill Gas Collection System
Gas recovery began at this site in December. 1988. There are 36 weds, spaced 100-200 feet
apart and connected to one header. The average depth of the wells is 60 feet: however, wells located
on the south side only average 37 feet deep where the county is actively dumping waste.
One Solar Centaur T4500 turbine (3 MW) is currently operating. Landffll gas enters the gas plant
at 120* F and flows through two inlet scrubbers. The first stage filters the gas and the second stage
provides mechanical separation of contaminants in the gas. In this process, the gas is compressed.
The gas then flows to a separator and ol knockout which catches impurities and water. The gas then
passes through a cooler and water knockout before ft enters the turbine.
The single turbine runs on full load continuously with an annual range of 1300 to 1800 scfrn.
This value changes with temperature as the gas becomes more or less dense. The turbine operates
more efficiently when the gas is cooler. The turbine burns approximately 2.1 mflion cubic feet of gas
per day (at 100% capacity). Methane content of the gas is typically 52% by volume.
The recovery system generates 4,000 gallons of condensate every three days. This condensate
is tanked and trucked off-site.
78
-------
Gas Sampling Points
An on-site gas chromatograph (GC) is used to measure gas composition twice a week.
Standards are run on the GC every morning. Gas flow is measured with a Kurz Air Velocity Meter
(model 4418) flow meter once a week just before the gas enters the turbine (at an orifice plate).
Leachate Collection
Leachate is collected in several leachate ponds around the site. Approximately 35,000 gallons of
leachate (and rainwater runoff) are collected daily and tanked and trucked off-site.
Gas Migration
The county is responsible for monitoring gas migration off-she and has installed several
perimeter wells. Migration off-site is thought to be minimal.
Information Gathered
A record of the tons of refuse placed by year in the landfill was provided by the County
Department of Public Works for 1976 to present These values are for municipal solid waste only, and
do not include tires, construction and demolition debris, or ash. Gas recovery personnel provided a
map of the landfill showing well locations, an example of a gas monitoring
record for each well, and an example GC printout Brochures were also provided discussing the gas
turbine.
Information to be Sent
Monthly gas flow records were requested from the gas recovery personnel.
79
-------
LANDFILL 17
Place and Date
State: Massachusetts
Date: Apr! 23.1991
Background Information
The tandfiU started accepting waste in 1952 and covers 40 acres. The landfill is rounded in
shape with a maximum height of about 80 feet There are approximately 5.7 mlllon tons of refuse in
place. The refuse acceptance rate is about 15,000 tons/month; the landfill operators indicated that
much of the refuse In place has been accepted in the past 10 years. The site is divided into two
sections, an active and an inactive section. A 60-mB synthetic liner is placed under the active half of the
landfill and extends over the inactive naff to also act as a cap for the inactive naff of the landfill. The
liner is under 20 acres and covers the remaining 20 acres. A one-foot thick day cap also covers the
inactive section.
Refuse composition is mostly commercial restaurant/office type waste which accounts for 85%
of the refuse. Construction debris accounts for another 5 to 7% of the waste, and the remaining refuse
is residential. The site contains tires, which were accepted by the former owner. Most of these are
found in the northeast section of the landfiR. The site accepted hazardous waste from 1968 to 1979 but
does not allow hazardous waste at this (andffl any longer. Batteries are banned in landfiP by the State
and are separated out of incoming trash.
Incoming refuse is weighed via a tare system, by which trucks are weighed and assigned a code
number. This code number and the corresponding truck weight can then be called up each time the
truck passes over the scales at the weigh station, and the weight of the refuse is automatically
calculated. The scales are calbrated annually by a scale company service technician who uses
standard weights to calibrate the scales. The State Department of Weights and Measures also inspects
the scales annually. Also, hauling trucks pass by a radiation detector as they enter the site.
Once in place, trash is pushed into place with compactors. The operators estimate that the
density of waste compacted in place averages 1440 Ibs/cubte yard.
Flare System
Flaring began at this site in 1987. This candlestick flare was not large enough for the landfiU, so
a new system went on-line in February 1990. The current gas control system consists of 41 wells and
one enclosed flare manufactured by McGII. The system flares an average 1200 cubic feet per minute
(cfm) of landfill gas from the well field. Gas passes from the main header Into a knockout tank, through
80
-------
a blower to a second knockout tank, and then to the flare. The blower is manufactured by Aerovent and
has a maximum capacity of 2 million cfm.
The gas is very dry and there is generally little to no condensate. When condensate is collected,
It is generally after rainfall when moisture has been added to the landfill. The site has never collected
any condensate in the second knockout tank.
Gas Sampling Points
All gas flow measurements are taken in the pipe between the blower and the flare. Sampling for
flow is taken with an Air Data Multimeter ADM-870. This same device measures vacuum in the system.
The sampling point for measuring vacuum is directly before the gas enters the blower. Vacuum
averages about 42" of water. Gas composition (percent methane) is measured weekly using a hand-held
gas analyzer. Gas composition measurements are taken at the pipe between the blower and the flare.
Leachate Collection
Leachate is collected from that portion of the landfill that is lined - approximately 20 acres.
About one and a half million gallons of leachate are collected annually. This leachate is sent to a public
wastewater treatment plant
Gas Migration
Perimeter wells from the 1987 gas collection system are monitored periodically for methane but
show that no gas is migrating from the site. Interior wells appear to be pulling gas toward the center of
the site, controlling off-site migration.
Information Gathered
Landfill personnel provided monthly tonnages of waste received for the period February 1990
through April 1991. They also provided an estimate of the amount of total waste in place in the landfill
prior to February 1990. Landfill personnel also provided weekly gas extraction reports containing total
landfill gas flow rate and methane concentrations for the period February 1990 through April 1991. A
landfill map and gas recovery schematic was also provided.
81
-------
LANDFILL 20
Place and Date
State: Michigan
Date: May 1.1991
Background Information
The entire landfill covers approximately 247 acres and is divided into two sections based on past
and current refuse placement patterns. The entire landfll is surrounded by a day dike and a trench for
surface water run-off. The landfll began accepting refuse in 1967 in the northern 160 acre section. This
section was fflled untl 1985. Refuse was then placed in the southern 77 acre section, which is stBI
actively being fflled. One small area in the northern section is being filed to bring it up to the height of
the entire section. Refuse is placed in cells; new ceils in the northern section were added above older
fill areas with a 2-foot thick day cap and leachate collection system under the new cells. The refuse age
in the northern section ranges from 6 to 24 years, with an average age of 15 years (assuming refuse
placement was relatively consistent from 1967 to 1985). The oldest refuse in the southern section is
approximately 6 years old.
The total refuse height on the northern section is irregular and ranges from 40 to 80 feet The
refuse in the southern section was placed 40 feet below grade and wll be 125 feet in height
Twenty five acres of the northern 160-acre section have a final cap (2-foot thick day cap, 6
inches of topsoB, and seeded with grass); the remaining 135 acres (apart from the valley now being
filed) have a 1-foot thick day cap. A final cap wll be placed over this area in the near future.
Annual waste acceptance figures were estimated by the previous landfill owner for 1967 to 1982
(as tons). A total of 5.62 MM tons of refuse were placed during this period. Records of gate yards
accepted from 1982 to 1989 were provided by landfll personnel, and refuse density values wll be
provided. A total of 16.96 mllion gate yards were placed from 1982 through 1989. Incoming gate yards
were .estimated based on the truck bed size, regardless of whether or not It was full.
Scales were installed in 1989, and a tare system is used by which vehides, their refuse types,
and their vehide weights are coded into a computer. Scales are calibrated annually by the State and
quarterly by the scale manufacturer. The refuse entering the landfll is classified as follows:
1. General refuse (municipal solid waste);
2. Construction and demolition debris (usually indudes wood);
3. Compact general refuse (MSW);
4. Bulky wastes from local resource recovery facitty (i.e.. white goods);
5. Process residue from local resource recovery facility;
6. Ferrous metal (now recyded);
82
-------
7. Non-contaminated dirt;
a Sand and concrete;
9. Contaminated soD;
10. Asbestos;
11. Industrial and wastewater sludge; and
12. Ash from local incinerator.
The only refuse that is segregated is the incinerator ash, which is placed in a lined, 7-acre
monoffll.
Landfill Gas Collection System
Landfill gas collection began in January of 1990. The gas is currently flared and landfill
personnel have plans to sell electricity to the local utility. Of the 247 acres, wells are distributed over
210 acres. The majority of wells are located in the northern 160 acre section; there are 16 wells in the
southern section. Gas was recovered from about 180 acres of refuse (with 69 wells) untl December of
1990. when 30 additional acres (with 11 wells) were brought on-line in the southern section.
Wells are primarily located where refuse is deepest There are currently no wells on the shallow
side slopes of the landfill. Well depth varies with refuse depth,, with a maximum well depth of 67 feet
and an average depth of 57 feet The 80 wells are connected to 14 lateral lines which lead to a 16 inch
header. Gas first passes through a separator where free liquids fall out. then onto two Hoffman
centrifugal blowers, each with a 100 horsepower engine and a capacity of 4 million cubic feet/day
(2778 scfm). From the blowers the gas is send to a McGll 5-bumer flare with a 99.75% destruction
efficiency of total hydrocarbons. The flare has a capacity of 168.6 MM Btu/hr, and currently bums about
4 mPlion cubic feet/day. The flare temperature is maintained at 1850*F. In order to prevent off-she gas
migration, the amount of gas flared is maximized.
Gas Sampling Points
Gas flow is read daOy from a Daniels Paymeter (Model 2272) at the inlet to the flare. The flow
meter is calibrated on an annual basis. Gas composition is also tested daOy, and the gas is drawn from
the header line before entering the blower. A standard sample gas of known composition is used daily
to calibrate the Hach Carte (Series 400 AGC) gas chromatograph. Individual wells and lateral lines are
tested twice each week for flow and gas composition. Landfill personnel try to adjust the vacuum on
each well so that a nitrogen content of 2% by volume is maintained.
Leachate Collection
There is no active leachate collection system under the oldest refuse in the 160 acre northern
section of the landfill. The southern section and the upper fill areas of the northern section have
83
-------
leachate collection systems. Condensate is combined with leachate from these areas and sent to
leachate collection ponds and aerated. During dry periods, some of the liquid Is redrculated In active fill
areas and haul roads to reduce dust The remaining liquid (38 million gallons per year) is tanked and
trucked off-site.
Gas Migration
Gas migration off-site is monitored at 31 perimeter wells that are tested monthly with a Mine
Safety Appliances Exploslmeter. LandfiU personnel indicated that only 2 or 3 of the wells alongthe
southern boundary of the landfill (near the newly-placed refuse) show slightly elevated CH4 levels. As
new wells are Installed along the southern slope, this migration wBI be controlled. In addition, leachate
collection wells in this area have high enough CH4 levels that 4 had to have flares attached untl they can
be brought into the header line to the blowers.
Information Gathered
Gas recovery personnel provided well boring logs for each well which show total well depth,
bottom condition, maximum temperature, date, and sol zone. Refuse information (in tons or gate yards)
was provided for 1967 through 1989. Gas recovery personnel also provided monthly gas flow and
composition records for January 1990 through Apr! 1991 and a topographic map of the site (1984).
Information to be Sent
Gas recovery personnel were asked to provide refuse tonnages for 1990 and 1991 (and a
recommended conversion factor for gate yards) and an up-to-date site map.
84
-------
LANDFILL 21
Place and Date
State: Michigan
Date: May 2.1991
Background Information
The landfill began operation in 1971 and is currently divided into three distinct areas based on
temporal waste acceptance. The oldest area encompasses 46 acres and accepted waste from
September 1971 until the fall of 1978. Area A encompasses 20 acres and accepted waste from 1982
untO 1986. Area B is the only active portion of the landfBI and encompasses 40 acres. Waste is placed
in excavated cells to an approximate depth of 80 to 100 feet The closed areas are graded to the
original contour of the land. The underlying son is day with an average depth of 35 feet The active
area is lined with a 41 mB polyethylene liner. The area was covered at the end of each day with
six inches of random soil. For the last three months, the active area has been covered at the end of
each day with a temporary synthetic fabric which is removed at the beginning of the next business day.
The oldest area of the landfill is capped with 2 feet of day and covered with approximately 30 feet of
dirt Area A is capped with 2 feet of day and is scheduled to have a low density polyethylene liner.
Daily soD covering practices in these areas averaged 10% of the daily refuse accepted.
The oldest area has a thick grass cover. The area wil be used as a local park in the near future,
with baseball, soccer, and tennis areas planned.
The landfill does not operate a scale to weigh incoming refuse. Waste acceptance rates are
measured by "gate* yards (cubic yards). Landfill personnel estimate the landfill will contain
approximately 4 million tons of refuse upon dosure. Approximately 1.1 million tons of refuse were
placed in the oldest area over its active life Approximately 800,000 tons of refuse were placed in Area A
over its active life.
The landffll separates asbestos waste and contaminated son from other incoming refuse.
Asbestos is kept in a controlled area separate from all other landfill waste acceptance areas. No further
segregation of wastes is practiced at the landfill. Contaminated sofls are co-disposed with other refuse
in the active area B only. Approximately 750 yards of contaminated sol are accepted daily. The landfill
does not accept municipal sludge.
Landfill Gas Collection System
LandfOi gas recovery began in February 1990. An LFG Industries flare is used to destroy the
extracted landfill gas. The flare was installed to control off-stte gas migration. Two Lamson centrifugal
blowers are used to draw the gas from the landfill. Each 100 horsepower blower has a capacity to pull
85
-------
3200 cubic feet of gas per minute (dm) and they are operated alternately. Currently the system is
operating at 2200 dm.
The landfill gas recovery system consists of 56 wells. The average depth of the wells is 60 feet
Thirteen of the gas extraction wells are in Area A. The remaining 43 wells are in the oldest area. The
average vacuum draw on the entire system is 90 inches of water column. The average methane gas
composition is 45%.
Gas condensate is collected and hauled off-site for treatment and disposal. A condensate
knock-out tank is located prior to the blower station. An average of less than 100 gallons per day of
condensate is collected. The condensate is sampled and TCLP analyzed on a routine basis.
Gas Sampling Points
Temperature, vacuum pressure, methane gas composition, and flow rate are measured once a
month at the flare inlet and at each extraction well (well flow rates are not measured). An Air Data Row
Meter is used to measure the gas flow rate at the inlet to the flare (after the blowers). Landffll personnel
do not calibrate the gas flow meter.
A Gas Scope is used to measure methane gas content at the inlet to the flare and at each
recovery wen. The gas scope is calibrated once a month.
Leachate Collection
The older section, which consists of approximately 46 acres, does not have an active leachate
collection system. The rest of the landfiR does have an active leachate collection system.
Gas Migration
The flare was installed in February 1990 to control off-site gas migration. Since installation of the
flare, gas migration off-site is controlled, as evidenced by no detectable CH4 at the 20 perimeter probes.
The gas probes are inspected on a monthly basis. No vegetative stress was detected in the dosed
areas or on the perimeter of the landfiH.
Information Gathered
Two maps of the landfill were received during the site visit (attached). Monthly measurements of
vacuum, percent methane, and temperature at the flare inlet and extraction wells, examples of the
routine gas condensate TCLP tests, and methane probe monitoring records were provided. Gas flow
records from the flare inlet were also provided. Finally, average waste acceptance rates and tonnage in
place for areas with gas extraction wells were received from tendfiU personnel.
86
-------
LANDFILL 22
Place and Date
State: Minnesota
Date: May 3,1991
Background Information
The landfill covers 100 acres and refuse acceptance began in 1970. Refuse acceptance ceased
In 1988, with the majority of refuse accepted until 1985. The total tons of refuse In place is
4,380,003 tons, with only 260,421 tons accepted from 1986 to closing in 1988. These tonnage estimates
are based on gate records of cubic yards brought in and assuming a conversion of 4 yards per ton.
The refuse age ranges from 20 years to 3 years, with an average of 13 years.
The landfill accepted primarily municipal solid waste, no known commercial hazardous wastes,
little construction and demolition debris, and little sludge. The depth of the fill area ranges from a
minimum of 5 to 6 feet at the edges to an average (and maximum) of 80 feet The landfill does not have
final closure, but the entire surface is covered with least 2 feet of sOty-day. which is 10 to 15 feet deep in
some areas. The refuse is settling 5 to 10 feet/year, and day is added to control surface water
drainage.
Landfill Gas Collection System
Landfill gas collection began in February of 1989. The gas is currently flared, as gas recovery
personnel feel they cannot produce electricity or pipeline quality gas profitably. The well field has a total
of 54 wells, but 14 of these wells are in the east area to prevent off-site migration and are not placed in
refuse. The 12-inch header line from these wells is brought into a separate blower house on the east
side of the property. The two 15 hp, 1100 scfm capacity Lamson blowers run only when nearby
perimeter probes detect methane gas migration. The gas from the east side header was originally
brought to a separate flare, but the methane content (<10%) and flow of gas from these wells
(120 scfm) were too low to maintain the flare. This header is now tied to the header from the high-flow
wells before the two lines enter the blower house.
The 40 high-flow wells range in depth from 35 to 80 feet and draw the majority of gas from this
landfill. These wells are connected to an 18-inch header which brings the gas first to the blower house
and the condensate knockout tank. Gas recovery personnel feel that this site produces a low amount of
condensate; they tank and truck approximately 500 gallons/day. The gas then flows to one of two
Lamson blowers (each 50 hp with a capacity of 3600 scfm). The blowers always run alternately. The
system is currently operating at 1500 scfm with the methane content of the gas averaging 40%. Gas is
87
-------
flared in a John Ztok open candle flare with an NAO exterior. There is an ultraviolet sensor to detect the
flame and keep the automatic system from cutting the flare off when winds shift the flame.
Gas Sampling Points
Total gas flow from both header lines is measured daly with an Air Data Multimeter model 870 at
the Wet and outlet of the main blower house. The Multimeter is calibrated onee each year. Gas
composition is monitored once each year by an independent laboratory. Gas flow, and composition
from the east side header is also measured separately. Weds are tested weekly for temperature, percent
methane (wtth Mine Safety Appliances MSA 625 Exploslmeter). Row is tested less often at the wells.
Leachate Collection
There is no active leachate collection system at this landfill; the fill area is surrounded by 54
ground-water test wells.
Gas Migration
Gas migration off-site is monitored wtth 74 perimeter probes that are tested on a weekly basis
with an MSA 62S Explosimeter Historical test data on these probes must be viewed with caution
because the probes along the northwest edge of the HI area (probes G-5. G-6, and G-7) are actually
placed in fill and are not representative of gas migration off-site. Due to concern over the proximity of
houses on the east side of the landfiU. the east side wells were installed (not on fifl) to contain any
migration. Gas recovered from these weDs is negligible. Gas migration through the cap is monitored
weekly with an organic vapor analyzer. Gas recovery personnel indicate that there is no significant
amount of gas escaping to the atmosphere.
Information Gathered
Gas recovery personnel provided maps showing wen and perimeter probe locations, perimeter
probe test results for 1990, well boring logs, and condensate test analyses. Records of refuse fin tons)
deposited each year were provided, along with three graphs of estimated methane gas generation rates
over time assuming three different production (decomposition) rates. Gas recovery personnel provided
test results of flare inlet gas analyses from two independent consultants, and total CH4 flow data taken at
the flare inlet, as well as flow data for wells on the east side of the landfiO.
Information to be Sent
Gas recovery personnel were asked to provide the most recent gas flow data.
88
-------
REFERENCE
Campbell, D., D. Epperson, L Davis. R. Peer, and 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-
101351).
89
-------
APPENDIX B
Metric/US. Equivalent Chart
Data in the main body of this report are
presented in metric units. This conversion
table is provided for easy reference.
90
-------
Metric/U.S. Equivalent Conversion Chart
Multiply
feet
meters
acres
hectares
cubic feet
cubic meters
inches
centimeters
tons (2000 Ib)
metric tons (Mg)
tons per acre
Mg per hectare
cubic feet per acre
cubic meters per hectare
cubic yards
cubic meters
acres
acres
cubic yards per ton
cubic meters per Mg
cubic feet per ton
cubic meters per Mg
cubic meters of CH4 (at STP)
by
0.3048
3.2808
0.4047
2.4710
0.0283
35.3145
2.5400
0.3937
0.9072
1.1023
2.2417
0.4461
0.0700
14.2913
0.7646
1.3079
4.046.9
43,560
0.8428
1.1865
0.0312
32.0563
713.8
To obtain
meters
feet
hectares
acres
cubic meters
cubic feet
centimeters
inches
metric tons (Mg)
tons (2000 Ib)
Mg per hectare
tons per acre
cubic meters per hectare
cubic feet per acre
cubic meters
cubic yards
square meters
square feet
cubic meters per Mg
cubic yards per ton
cubic meters per Mg
cubic feet per ton
grams CH4 (at STP)
Temperature
Fahrenheit = (1.8) (Celsius) + 32
Celsius = 0.556 [(Fahrenheit) - 32]
91
-------
APPENDIX C
Data Summary Tables hi U.S. Equivalents
This appendix contains data summary
tables in U.S. Equivalents. The
corresponding metric table in the main
body of the report is referenced.
92
-------
TABLE C-1. COMPARISON OF SUMMARY STATISTICS FOR METHANE FLOW RATES AT THE LANDFILLS'
Measurements
Landfill
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
20
21
22
23
24
25
Type
dally
dally
daily
dally
dally
dally
monthly
dally
dally
monthly
monthly
dally
daHy
monthly
minute
monthly
daHy
dally
dally
dally
dally
Number
194
302
314
85
209
37
12
626
15
6
12
232
11
15
13
12
11
51
202
333
331
Average
1955
637
1415
3477
878
590
342
415
272
1036
399
282
368
567
488
1237
967
1172
79
625
715
Methane Recovery (ft3/min)
Median
1954
644
1424
3489
880
592
359
437
247
1077
415
273
390
585
491
1240
937
1125
81
631
734
Standard
Deviation
75
42
82
47
60
73
71
87
50
18
43
36
53
46
53
68
04
77
18
77
99
Minimum
Value
1696
435
1068
3296
725
445
143
18
202
827
322
191
276
260
358
935
865
764
11
110
51
Maximum
Value
2169
723
1576
3585
987
800
423
604
372
1145
449
366
414
765
583
1458
1164
2121
104
781
862
Range
474
289
508
289
262
355
260
587
170
318
127
176
138
505
225
522
299
357
93
677
811
Coefficient
of Variation
(CV)
3.8
6.6
6.3
1.4
6.8
12.4
20.9
21.0
21.2
11.4
10.8
12.9
14.5
25.7
10.8
25.5
10.8
23.6
22.9
12.4
13.9
* Corresponds to Table 2.
-------
TABLE C-2. SUMMARY OF LANDFILL PARAMETERS USED IN THE STATISTICAL ANALYSES (U.S. Equivalents)*
Land
Wentifle
Cod
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
20
21
22
23
24
25
«- ^^^=
VIII
atlon
>e
Latter
A
B
C
D
E .
F
0
H
1
J
K
L
M
P
Q
T
U
V
w
y
Y
sssBssar
Refuse Man
(106 ton«)
7.00
6.75
8.10
15.20
12.00
2.64
3.25
3.00
1.80
5.80
2.00
3.06
1.06
3.73
5.70
10.70
2.87
4.38
3.16
6.84
11.74
Average
Refuse
Age(yrs)
8.0
10.0
10.0
9.5
15.0
7.0
10.0
10.0
7.0
12.0
10.0
8.5
7.0
5.5
10.0
11.0
13.0
12.0
10.7
5.6
12.0
Landfill
Area
(acres)
86.0
135,0
126.0
140.0
80.0
99.6
125.0
55.0
30.0
80.0
69.0
85.0
35.0
75.0
40.0
180.0
66.0
100.0
100.0
270.0
185.0
Average
landfill
Depth
(feet)
220.00
85.00
217.00
185.00
150.00
32.25
60.00
55.00
50.00
70.00
80.00
35.00
40.00
90.00
60.00
60.00
90.00
80.00
70.00
100.00
75.00
Average
w«n
Depth
(feel)
45.00
47.00
77.00
70.00
112.00
32.25
40.00
55.00
50.00
45.00
45.00
35.00
40.00
60.00
60.00
57.00
60.00
80.00
65.00
50.00
70.00
^KB^^on
Number
of Welle
45
44
31
111
102
68
48
107
32
96
39
78
51
36
41
69
56
40
19
23
43
Landfill
Volume
(106 feet3)
824.9
500.3
1192.1
1129.2
523.2
140.0
327.0
131.9
65.4
244.2
240.7
129.7
61.0
294.3
104.6
470.9
259.0
348.8
305.2
1177.2
605.0
Average
Methane
Flow Rate
1955
637
1415
3477
878
590
342
415
272
1036
399
282
368
567
486
1237
967
1172
79
625
715
Average
Methane Rate
Per Unit Mast
(ftfymln/
10* tons)
279.29
94.37
174.69
228.75
73.17
223.49
105.23
138.33
151.11
178.62
199.50
92.16
348.49
152.01
85.61
115.61
337.29
267.58
25.00
91.37
60.90
Gat End Uee
ELEC.-TURBINE
ELEC.-TURBINE
ELEC.-TUR8INE
ELEC.-TURBINE
FLARE
ELEC.-IC ENGINE
BOILER FUEL
HIGH BTU
FLARE
HIGH BTU
BRICK KILN FUEL
HIGH BTU
FLARE
ELEC.-TURBINE
FLARE
FLARE
FLARE
FLARE
ELEC.-IC ENGINE
ELEC.-TURBINE
ELEC.-TURBINE
Corresponds to Table 4.
-------
TABLE C-3. SUMMARY OF CLIMATIC DATA FOR THE LANDFILLS
(U.S. Equivalents)'
Annual Climate Averages
(from refuse acceptance period)
Landfill
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
20
21
22
23
24
25
Weather
Average
Period
1955 to 1989
1968 to 1982
1970 to 1988
1971 to 1988
1951 to 1989
1975 to 1989
1972 to 1989
1973 to 1987
1979 to 1989
1962 to 1989
1962 to 1989
1969 to 1983
1980 to 1988
1976 to 1989
1952 to 1989
1967 to 1989
1971 to 1986
1970 to 1988
1958 to 1986
1972 to 1989
1970 to 1989
Average
Temp.
PF)
452
48.7
54.7
76.3
63.1
61.1
592
61.6
64.4
61.9
64.5
544)
512
484
47.0
48.6
48.4
45.0
50.1
662
492
Maximum
Temp.
fF)
55.6
58.9
63.7
85.4
77.6
75.1
702
71.6
76.0
72.9
76.3
634
61.7
SOS
55.7
58.1
57.9
54.4
64.0
76.5
58.6
Minimum
Temp.
fF)
34.7
38.4
45.7
672
48.7
47.0
48.1
51.6
52.8
504
52.7
44.1
404
38.0
384
392
384
35.7
362
554
39.7
Total
Rainfall
Cinches)
29.66
37.14
4Z99
55.46
16.60
18.66
4243
49.72
58.39
54.05
4449
4946
4046
38.44
48.03
32.43
32.77
28.81
15.36
35.00
3725
30-year Annual Averages
Average Average Total
Temp. Dewpoint Rainfall
fF) fF) finches)
45.5
48.7
54.3
75.1
62.8
61.1
59.0
612
64.6
622
64.7
534
514
49.1
464
48.6
48.6
44.7
504
664
492
38
39
43
65
50
47
48
50
54
51
52
44
44
37
35
39
39
34
28
52
39
28.76
35.62
41.42
6148
17.02
1749
41.76
48.61
52.82
5442
4446
48.40
4147
40.16
47.60
3047
3047
2646
1541
29.45
3344
* Corresponds to Table 5
95
-------
APPENDIX D
Landfill Data Sheet
This data sheet was used as a guide by
project team members in order to obtain
relevant and complete data prior to and
during the site visit
96
-------
LANDFILL DATA SHEET
Call Made By Date Call Hade
Landfill Facility Name:
Address:
Contact at Landfill: Phone Number:
Please provide the following Information for only that portion of your landfill where methane Is being
recovered. Please provide this information for the period of time that data has been collected. For Items
such as the number of wells that may have changed over time, please provide the current information.
PRIORITY DATA
Active Landfill?
Date Waste Acceptance Began ^
Date Waste Acceptance Ceased _____________________________^^__^^______^__
Date Methane Recovery Began ____^___^_^_____^_^^______^_^^_^^_____^_____
Gas End Use
Annual Methane Production Rate
Tons of Refuse In Place
Age of the Refuse ___^^__
Number of Acres
ADDITIONAL INFORMATION (provide as necessary)
Number of Active Wells (Regular- or High-How Wells)
Number of Low-How Wells
Depth of Active Wells: Minimum Average Maximum
Depth of Low-How Wells: Minimum Average Maximum
Depth of Landfill: Minimum Average Maximu
Methane Recovery System (I.e.. turbine, 1C engine, other):
Landfill Design (I.e., cell, canyon, trench, or other)
Cap Composition Cap Thickness
Cap Permeability
No. of Hares (if applicable)
Acceptance Rate of Waste by Year
Total Capacity (by weight):
[If capacity Is provided by volume, what is the average refuse density?]
Daily Soil Cover Information (does volume number Include ALL refuse or soil and refuse?)
Results of Routine Testing for Surface or Perimeter Leaks .
Any other data available on:
Refuse Composition? ___^^____^_^________^^^^_^^^^_________
Gas Composition?
Moisture Content of Refuse?
Compliance Testing of Power Generation or Control Equipment Exhaust?
97
-------
TECHNICAL REPORT DATA
(Ptease read Imonictiont on the revene before completing)
1. REPORT NO.
EPA-600/R-92-037
12.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Development of an Empirical Model of Methane
Emissions from Landfills
S. REPORT DATE
March 1992
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
R. L. Peer, D. L. Epperson, D. L. Campbell, and
P. von Brook
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
P.O. Box 13000
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
11. CONTRACT/G RANT NO.
68-D9-0054, Task 31
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 3-12/91
14. SPONSORING AGENCY CODE
EPA 7600/13
is.SUPPLEMENTARY NOTES AEERL project officer is Susan A.
541-2709.
Thorneloe, Mail Drop 63, 919 /
16. ABSTRACT
The report gives results of a field study of 21 U. S. landfills with gas re-
covery systems, to gather information that can be used to develop an empirical mo-
del of methane (CH4) emissions. Site-specific information includes average CH4 re-
covery rate, landfill size, tons of refuse (refuse mass), average age of the refuse,
and climate. A correlation analysis showed that refuse mass was positively linearly
correlated with landfill depth, volume, area, and well depth. Regression of the CH4
recovery rate on depth, refuse mass, and volume was significant, but depth was the
best predictive variable (R2 = 0. 53). Refuse mass was nearly as good (R2 = 0.50).
None of the climate variables (precipitation, average temperature, dewpoint) were
correlated with the CH4 recovery rate or with CH4 recovery per metric ton of re-
fuse. Much of the variability in CH4 recovery remains unexplained, and is likely due
to between-site differences in landfill construction, operation, and refuse composi-
tion. A model for global landfill emissions estimation is proposed.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Earth Fills
Methane
Emission
Mathematical Models
Refuse
Pollution Control
Stationary Sources
Gas Recovery
13 B
13C
Q7C
14G
12A
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report}
Unclassified
21. NO. OF PAGES
103
20. SECURITY CLASS (This pageJ
Unclassified
22. PRICE
5PA Form 2220-1 (9-73)
98
------- |