MEMORANDUM
To:	Docket ID. No. EPA-HQ-OAR-2019-0424
From:	Liz Goodiel, U.S. EPA/CCD
Date:	November 21, 2022
--------------------------------------------------------------------------------
Subject:	Technical Support for Supplemental Revisions to Subpart HH: Municipal Solid Waste Landfills
1.0	Introduction
The U.S. Environmental Protection Agency (EPA) is issuing a supplemental proposal to amend specific provisions in the Greenhouse Gas Reporting Rule (40 CFR part 98) to improve the quality and consistency of the data collected under the rule. As part of this proposal, the EPA is proposing revisions to the Municipal Solid Waste (MSW) Landfills source category (40 CFR part 98, subpart HH; hereafter referred to as "subpart HH") to account for anomalous or large-release events of methane emissions from municipal solid waste landfills, including but not limited to a poorly operating recovery system of the gas control and collection system (GCCS), a poorly operating control device, or leaks in the cover system due to cracks, fissures, etc. Existing subpart HH emission calculation equations include methodologies to quantify and account for emissions released due to poor operations in recovery systems and control devices. However, the EPA is proposing slight revisions and clarifications to these terms to improve accuracy of reporting. Additionally, there is not currently a methodology in place under subpart HH to account for methane emissions from cover system leaks. The EPA is proposing the following revisions for subpart HH Equations HH-6 through HH-8 to improve clarity for calculating emissions due to cover system leaks:
 A new term in Equations HH-6, HH-7, and HH-8 to quantify emissions from cover system leaks from landfills with gas collection systems for which surface methane concentration measurements across the landfill are conducted. 
 A lower default collection efficiency in Equations HH-7 and HH-8 for landfills with gas collection systems for which surface methane concentration measurements are not conducted.
This memorandum provides technical support for these proposed revisions.
2.0 	Background on Subpart HH Calculation Methodologies  
Methane (CH4), carbon dioxide (CO2), nitrogen (N2), hydrogen (H2), and small quantities of other gases are generated by the anaerobic decomposition of organic waste in a MSW landfill. These gases, collectively referred to as landfill gas (LFG), are produced throughout the operating life of a facility (i.e., while the facility receives waste) as well as many years, often several decades, after the facility has closed. 
Subpart HH of the Greenhouse Gas Reporting Program (GHGRP) requires landfill operators to estimate annual CH4 generation based on the amount of decomposable material in the waste coupled with a decay rate. This approach is generally referred to as a "first-order decay" (FOD) model and is the top down national GHG inventory approach recommended in the 2006 Intergovernmental Panel on Climate Change (IPCC) Guidelines. For landfills without gas collection systems, the FOD model is the only method available under subpart HH for estimating annual CH4 generation and emissions. For landfills with gas collection systems, subpart HH requires landfill operators to measure the annual quantity of CH4 recovered and use this measured CH4 to calculate annual CH4 emissions using two methods. Method 1, represented by Equation HH-6, calculates CH4 emissions by subtracting the amount of CH4 recovered from the amount of CH4 generated based on the FOD model. Method 2, represented by Equation HH-8, calculates CH4 emissions by subtracting the amount of CH4 recovered from the amount of CH4 generated based on Equation HH-7. (Equation HH-7 calculates CH4 generation based on the amount of CH4 recovered divided by the collection efficiency and the fraction of annual operating hours of the collection system.) Method 2 is commonly referred to as the "back-calculation method" since CH4 generation is based on measured or recovered CH4 rather than modeled CH4. Landfill operators with gas collection systems must report annual CH4 emissions using both methods but have the option to select the method they consider the best representation of their annual emissions reported under subpart HH. 
3.0 	California Study Indicating Large Emissions from Landfills  
Recent CH4 measurement data collected from five aerial survey campaigns over a variety of different CH4 emission sources in California suggest that CH4 emissions from landfills may be considerably higher than CH4 emissions quantified or reported under subpart HH. Based on the study measurements, Duren, et al. (2019) concluded: 
   "Solid-waste management is the largest methane point-source emission sector in California, with persistent plumes observed at only 32 of 436 surveyed landfills and composting facilities. Our imaging of landfills identified methane plumes associated with construction, gaps in intermediate cover and leaking gas-capture wells -- indicating a subpopulation of anomalous emitters. The fact that we did not detect a larger population of smaller methane point sources across the landfill sector suggests that most of those facilities emit methane as area sources that cannot be detected with this method."
As noted by Duren, et al., (2019), the aerial method used in this study is best suited to detect high emitting "point source" releases of CH4 and could not quantify emissions for landfills with lower emissions or with emissions spread out more evenly across the landfill. It is important to note that only landfills with anomalous emissions could be quantified by the aerial methods used by Duren, et al., (2019) and that these emissions only occurred at 7 percent of the surveyed landfills. However, when these anomalous emissions occur, the CH4 emissions reported to the EPA under subpart HH are consistently lower than the measured emission rates extrapolated to annual estimates.
      Because the California aerial study of Duren, et al., (2019) could not quantify the emissions from 93 percent of the landfills that did not have anomalous emissions, this study does not provide evidence that the subpart HH methodologies are inaccurate or biased under typical conditions that exist for most landfills. Therefore, we conclude that this study indicates a potential bias in the subpart HH methodologies for the fraction of landfills that have anomalous emissions. 
4.0 	Assessment of Subpart HH Methods to Account for Large Emissions Events  
      We identified the following potential causes of high emissions from landfills:
 Poor operations in the recovery system of the GCCS;
 Poor operation of the control device; and/or
 Leaks in the cover system due to cracks, fissures, gaps around protruding wells, etc.

4.1 	Revisions Considered for Poorly Operating Recovery Systems
Typically, recovery system periods of poor operation are noticed quickly by landfill operators. If the recovery system goes down, the amount of CH4 recovered (term "R" in Equations HH-6, HH-7, and HH-8) will be lower. In this situation, Equation HH-6 will automatically yield higher CH4 emissions. While it may appear that a lower R in Equation HH-8 would result in lower projected emissions, the fRec,n term is included in the equation specifically to account for times when the recovery system is not operating. Together, these terms (R and fRec,n) will result in the same quantity of CH4 generation (Equation HH-7 and the first term in Equation HH-8), and the reduced CH4 recovery will yield higher emissions in Equation HH-8. 
In reviewing the fRec,n term, issues were raised during the verification of submitted subpart HH emission reports in how the electronic Greenhouse Gas Reporting (e-GGRT) system calculates emissions at times when one recovery system may have two measurement locations or multiple control devices are associated with one measurement location. For example, the landfill may have turbines used to generate electricity from the landfill gas and a backup flare for when the turbines are down for maintenance or repair and separate measurement locations for the turbines versus the flare. The gas collection system may operate a total of 8,760 hours per year and the landfill gas may only be directed to the flare for 100 hours. In this case, the fRec,n term should be 1; however, the e-GGRT internal calculations will estimate the fRec,n term associated with the flare to be 0.011 (100/8,760). To help clarify the correct calculation procedure, we are separating out calculations that should be performed at the gas collection system level rather than the measurement location. Thus, we are replacing the fRec,n term evaluated at the measurement location with an fRec,c term evaluated at the collection system level. 
4.2 	Revisions Considered for a Poorly Operating Control Device
Like recovery systems, control device periods of poor operation are typically noticed quickly by landfill operators. For example, some flares may have automatic shut-offs to prevent sending landfill gas to a flare when the pilot light is not lit. Both Equation HH-6 and HH-8 include an fDest,n term (fraction of operating time for the control device) to account for times the recovered gas is vented rather than destroyed. There may be times when the control device may be operating, but not at the temperature where the device is expected to efficiently destroy CH4. These periods of poor performance can yield significant unintended CH4 emissions. Therefore, to ensure the fDest,n term accurately accounts for these issues, the EPA is proposing additional language to clarify that, "The annual operating hours for the destruction device should include only those periods when flow was sent to the destruction device and the destruction device was operating at its intended temperature or other parameter indicative of effective operation. For flares, times when there is no pilot flame present must be excluded from the annual operating hours for the destruction device." With this clarification, the performance of the fDest,n term should account for periods when the destruction devices are non-operating or operating poorly. 
4.3 	Revisions Considered for Leaks in the Cover System
Unlike recovery system or control device periods of poor operation, leaks in the cover system may go unnoticed for quite some time. 
For landfills without gas collection systems, subpart HH assumes that nearly all the CH4 generated is emitted, save a small fraction of CH4 (typically 10 percent) that is oxidized by aerobic microbes near the soil surface as the landfill gas percolates through the upper soil layer. If there is a large crack or fissure in the landfill surface for gas to escape, the CH4 oxidation fraction for the gas escaping via the fissure is likely near zero. Thus, the subpart HH methodology for landfills without gas collection systems may underestimate emissions in this case, but only by about 10 percent.
The majority of MSW landfills reporting under subpart HH have landfill GCCS and these "controlled" landfills account for 76 percent of subpart HH emissions in 2021. Presently, Equations HH-6 and HH-8, which are used to quantify CH4 emissions from controlled landfills, do not account for fugitive CH4 emissions from cover system leaks. Cover system leaks can lead to large release events of CH4 emissions and may explain the discrepancy between the anomalous emitters from the California aerial surveys (Duren, et al., 2019) and those reported under subpart HH. Controlled landfills have fugitive CH4 emissions from areas of the landfill where the GCCS is not in place; these areas must use a collection efficiency (CE) of 0 percent. Controlled landfills also have fugitive CH4 emissions from controlled areas where the GCCS is in place. Fugitive emissions from controlled areas are dependent on the CE of the GCCS, which can vary across a landfill depending on soil cover depth and compaction. For example, capped areas with 3 feet of compacted clay are assumed to have higher collection efficiencies (95 percent) and lower emissions than areas with a daily or intermediate cover (60 percent and 75 percent, respectively). 
Because a significant portion of the CH4 generated may be collected via the GCCS, the rate at which landfill gas percolates through the soil surface layer is reduced. This reduced rate tends to increase the fraction of CH4 oxidized, so controlled landfills may be able to use higher soil oxidation factors (0.25 or 0.35) instead of the basic 0.10 default oxidation factor. As such, fissures or cracks in the cover material can significantly impact the projected emissions from controlled landfills using Equation HH-6 and have even more of a significant impact on landfill operators using Equation HH-8 because these leaks can affect both the overall CE and the surface oxidation factor.
To ensure GCCS are efficiently capturing the landfill gas, landfills subject to New Source Performance Standards (NSPS) or Emission Guidelines (EG) for MSW landfills under part 60 or part 62 of section 111 of the Clean Air Act that meet certain criteria are required to conduct periodic surface methane concentration measurements (SMCM). We note that MSW landfills are subject to the NSPS or EG based on size and date of operation and are required to install and operate a GCCS once annual emissions of non-methane organic compounds (NMOC), which are present in landfill gas along with CH4, reach a certain threshold. Both the NSPS and EG include certain maintenance requirements for a mandatory GCCS, including periodic SMCM monitoring across the cover system. Specifically, SMCM involves measuring the CH4 concentration measurements within 5 to 10 centimeters (cm) of the ground every 30 meters across the entire coverage area of the system. For any location where a measurement equals or exceeds 500 parts per million (ppm), the landfill is required to take corrective action at the location, which can include cover maintenance or adjusting the vacuum of the collection wells in the area, and re-monitor the location within 10 days. If the latter shows that the concentration is still at or above 500 ppm, additional corrective action and re-monitoring must take place within 10 days of the first re-monitoring. If the second re-monitoring concentration equals or exceeds 500 ppm, corrective action and re-monitoring must take place within 10 days of the second re-monitoring. If the third re-monitoring is still out of compliance, a new well or other collection device must be installed within 120 days of the initial exceedance. For most landfills, SMCM is required quarterly, with the measurement data sent to the permitting authority as well as to the regional EPA. Closed landfills that have four consecutive quarterly monitoring events with no "exceedances" (i.e., no measurements equal to or greater than 500 ppm) may conduct annual monitoring; however, if an exceedance occurs in the future, the landfill must resume quarterly monitoring.
It makes sense that landfill operators that conduct SMCM monitoring to ensure their GCCS is effectively capturing CH4 would have better overall CEs than landfills that do not conduct SMCM monitoring, all other things being equal. It also follows that landfills that have more SMCM exceedances than another landfill (all other things being equal) would have higher emissions than the equivalent landfill with fewer SMCM exceedances. Therefore, we sought methods to identify means to quantify these differences.
4.3.1	Quantifying Differences in Collection Efficiencies Between Landfills Conducting and Not Conducting Surface Methane Concentration Measurements
The Environmental Integrity Project (EIP) evaluated CH4 emissions from Maryland landfills. The EIP study compiled data for 40 Maryland landfills, primarily from data reported to the Maryland Department of the Environment (MDE) in their 2017 state GHG inventory. Twenty-one of the 40 landfills for which data were compiled have GCCS required by either state or federal requirements. Four of the 21 mandated landfill GCCS are required under federal rules: three active landfills and one closed landfill. According to the EIP study authors, 17 landfills have voluntarily installed GCCS and adequate data were available to estimate GCCS CEs for 16 of these landfills. The EIP study authors primarily assessed landfill gas CE by comparing measured quantities of CH4 collected by the GCCS and modeled CH4 generation rates using LandGEM, which uses the FOD model. Thus, this study did not directly measure the emissions from the Maryland landfills.
The EIP study authors compared the landfill gas CE for the four landfills mandated by federal rules, which are required to conduct SMCM, to the CE for the 16 landfills with voluntary GCCS. EIP found that the four landfills with federally mandated GCCS and SMCM had an average CE of 76 percent, while the 16 landfills with voluntary GCCS had an average CE of 55 percent. Thus, the CEs for landfills with voluntary GCCS (no SMCM) were 21-percentage points lower than the CEs for landfills with federally mandated GCCS (with SMCM).
However, some of this difference could be caused by differences in the relative coverage of the GCCS across the landfill as well as differences in the types of soil covers used in areas where landfill gas is collected. Table HH-3 of subpart HH contains default CEs for areas under active gas collection that differ by the type of soil cover. Areas of the landfill with a final soil cover of 3 feet or thicker of clay or a geomembrane cover system have a default CE = 95%; areas with an intermediate soil cover have a default CE = 75%; and areas with only daily soil cover have a default CE = 60%. Thus, some of the differences in the Maryland landfill CEs may be due to differences in the relative areas of the cover system. We reviewed subpart HH reported data from Maryland landfills. There are 19 Maryland landfills reporting under subpart HH for RY 2017, 14 of which have GCCS.  The 14 Maryland landfills with GCCS that report under subpart HH include the three active landfills with federally required GCCS and 11 landfills identified by EIP as having voluntary GCCS. Relevant data for these landfills are summarized in Table 1.
Table 1. Comparison for Maryland Landfill Gas Collection Efficiencies Reported in EIP (2021) and Subpart HH. 
                           Subpart HH Facility Name
                                EIP Reported CE
                            Subpart HH Reported CE
Landfills with Federally Required GCCS
Millersville Landfill
                                                                           0.77
                                                                           0.89
Prince George's County Brown Station Road Sanitary Landfill
                                                                           0.67
                                                                          0.805
Eastern Sanitary Landfill Solid Waste Management Facility
                                                                           0.64
                                                                           0.73
Landfills with Voluntary GCCS
Midshore I Regional Solid Waste Facility
                                                                           0.45
                                                                           0.65
Harford Waste Disposal Center
                                                                           0.38
                                                                           0.59
Norris Farms Landfill
                                                                           0.95
                                                                           0.95
Beulah Sanitary Landfill
                                                                            0.3
                                                                            0.6
Quarantine Road Landfill
                                                                           0.38
                                                                          0.569
Mountainview Sanitary Landfill
                                                                           0.62
                                                                          0.787
Alpha Ridge Landfill
                                                                           0.41
                                                                           0.82
Reichs Ford Municipal Landfill & Recycling Center
                                                                           0.53
                                                                            0.8
Newland Park Landfill
                                                                           0.55
                                                                           0.68
Cecil County Central Landfill
                                                                           0.63
                                                                            0.7
Central Sanitary Landfill
                                                                         0.0077
                                                                           0.38
                                                                               


Average for Federally Mandated GCCS
                                                                            69%
                                                                            81%
Average for Voluntary GCCS
                                                                            47%
                                                                            68%

As seen in Table 1, considering the subset of Maryland landfills reporting to subpart HH, we see a 22-percentage point difference between the CEs for landfills with voluntary GCCS (no SMCM) and those with federally mandated GCCS (with SMCM) (69% - 47%). However, we see from the subpart HH reported CEs that there is a 13-percentage point difference (81% - 68%) that can be attributed to differences in the coverage area of the GCCS and/or the relative areas of the landfill having different cover types (final, intermediate, or daily covers). 
The subpart HH calculated collection efficiencies are consistently higher than those determined from the Maryland inventory data. This may indicate that the subpart HH defaults are too high, but it may also indicate that the "potential methane generation capacity" input term for LandGEM (which is analogous to the DOC term used in Equation HH-1) is higher than the actual potential CH4 generation capacity of the waste being disposed. That is, because the CEs determined from the Maryland data are based on modeled CH4 generation rather than actual measured CH4 emissions from the landfills, it is just as likely that the Maryland CEs are biased low as the subpart HH CEs are biased high. However, it is clear that landfills that have federally required GCCS appear to have higher CEs than voluntary systems and the difference in the CEs is larger (by about 10-percentage points) than can be explained by differences in the relative areas of the different cover types. Therefore, we are proposing additional default CEs in Table 3 of subpart HH (Table HH-3 to Subpart HH of Part 98 - Landfill Gas Collection Efficiencies) to account for the reduced CEs observed for landfills for which SMCM monitoring is not performed. The proposed new CEs for landfills for which SMCM monitoring is not performed are 10-percentage points lower than the current defaults in Table HH-3. The current default CEs in Table HH-3 will remain applicable for landfills for which SMCM monitoring is conducted.
4.3.2	Quantifying Methane Emissions from Exceedances in the Surface Methane Concentration Measurements
To account for differences in performance between similar landfills with different numbers of SMCM exceedances, we sought to quantify the emissions from an individual exceedance. We reviewed several studies[,] and identified a correlation from Heroux, et al., (2010) between the CH4 flux rate in micrograms per square meter per second (ug/m[2]-s) and the CH4 concentration in ppm by volume (ppmv) at 6 cm of the ground surface as follows.
Methane Flux μgm2­s =0.3647xMethane Concentration (ppmv)
                                   R2=0.9188
                                       
Because the SMCM must be conducted within 5 to 10 cm of the ground, the value of the SMCM can be used directly for this correlation. The SMCM are conducted every 30 meters, so each measurement location represents a 900 m[2] area (30 m x 30 m). When the SMCM are conducted daily since the last monitoring event, each measurement represents 1 day or 86,400 seconds (24 hrs/day x 3,600 sec/hr).  Noting there are 10[12] ug/Mg (Mg = megagram = metric ton), the emissions associated with a single daily SMCM exceedance can be calculated as:
Emissions Mg/day= 0.3647xSx900x86,4001012=0.0000284xS 
Where, 
    S	=	Surface CH4 concentration measurement that exceeds 500 ppm above background (ppmv).
For landfills that conduct SMCM at different frequencies, the above equation can be multiplied by the number of days between monitoring events to estimate the emissions for that given monitoring frequency. For example, if quarterly monitoring is conducted, the emissions associated with an exceedance would be calculated using 91 days in one quarter. One quarterly SMCM with a value of 1,000 ppmv would represent 2.6 Mg of "extra" CH4 emissions from the landfill compared to a facility with no exceedances. A landfill can have more than one exceedance per monitoring period and have additional exceedances in subsequent monitoring events during the calendar year. 
4.3.2.1	Added Term for Equation HH-6
When using the FOD model and Equation HH-6, all CH4 generated that is not recovered percolates through the cover material. As noted previously, some fraction of the CH4 that percolates through the cover material is expected to be oxidized and the rest emitted. When there are fissures or cracks in the surface, this does not alter the quantity of CH4 generated or the measured quantity of CH4 recovered, but it does cause the landfill gas to "bypass" the aerobic microbes near the soil surface. So, when using Equation HH-6, we need to account for the amount of CH4 that the model assumes is oxidized but is not due to bypassing. To accomplish this, we are proposing to add the following term to account for SMCM monitoring exceedances to Equation HH-6:
Added Term for Equation HH­6 in Mg=OXxm=1M0.0000284xdmxSm
Where:
   OX	=	Oxidation fraction. Use appropriate oxidation fraction default value from Table HH-4 of subpart HH (Table HH-4 to Subpart HH of Part 98 - Landfill Methane Oxidation Fractions).
   M	=	Number of individual surface measurements that exceed 500 parts per million (ppm) above background in the reporting year. If surface monitoring is not performed or no measurement exceeded 500 ppm above background in the reporting year, assume M = 0.
   0.0000284	= 	Correlation factor (metric tons CH4 per ppm surface concentration per day).
   dm	=	Leak duration (days), estimated as the number of days since the last monitoring event at the specified location from company records. Alternatively, you may use the following defaults for d: 10 days for 10-day monitoring events; 30 days for monthly monitoring, 91 days for quarterly monitoring, and 365 days for annual monitoring.
   Sm	=	Surface measurement methane concentration for the m[th] measurement that exceeds 500 parts per million above background (parts per million by volume). 
4.3.2.2	Added Term for Equations HH-7 and HH-8
When using the back-calculation method to determine CH4 generation in Equation HH-7 and CH4 emissions in Equation HH-8, the quantity of CH4 generated is estimated by the quantity of CH4 recovered and the expected CE of the GCCS. When there are exceedances during SMCM monitoring, the CE of the GCCS is lower than expected by the quantity of excess CH4 released from the soil surface in the area associated with the exceedance. Thus, when using Equations HH-7 or HH-8, the following term is proposed for the additional CH4 generation or emissions, respectively, associated with the SMCM exceedance. 
Added Term for Equations HH­7 and HH­8 in Mg=m=1M0.0000284xdmxSm
4.3.3	Estimating the Number of Subpart HH Reporters Impacted by Revisions to Account for Leaks in Cover Systems 
A total of 834 landfills with GCCS reported CH4 emissions to subpart HH in RY 2021. All but 42 of these landfills are believed to be subject to the federal NSPS or EG for MSW landfills based on their design capacity and year of waste acceptance. However, as noted previously, some landfills may be subject to NSPS or EG for MSW landfills but not required to install a GCCS and conduct SMCM monitoring until they exceed certain NMOC emission thresholds (typically 34 Mg/year, or 50 Mg/year for certain closed landfills). To estimate the number of landfills for which operators are required to conduct SMCM monitoring, estimates of NMOC emission rates were needed. NMOC emission estimates for 2021 were found for 701 of the 792 controlled landfills believed to be subject to NSPS or EG for MSW landfills. Among the 701 landfills for which data were available, 592 (84%) had NMOC emissions requiring controls (based on exceeding the 34 Mg/year emission threshold). Based on this information, we estimate that at least 70 percent (592/834) of landfills with GCCS reporting to subpart HH are conducting required SMCM monitoring according to the provisions in the federal NSPS or EG for MSW landfills. These facilities should be able to provide the additional data needed to calculate emissions from cover leaks with minimal burden.
By difference, we estimate that 15 to 30 percent of landfills with GCCS reporting under subpart HH are not conducting SMCM monitoring. These facilities may elect to conduct SMCM monitoring at least annually or use the new, lower CEs proposed for landfills with GCCS that are not conducting SMCM monitoring. Some of these landfills may be required by state regulations to conduct SMCM monitoring so these facilities should be able to provide the additional data needed to calculate emissions from cover leaks with minimal burden. We expect nearly all of the other landfill operators to elect to use the new, lower CEs and this option will not change the overall reporting burden for these facilities. 
