                                                                               
MEMORANDUM   

TO:		Jodi Howard, EPA/OAQPS/SPPD

FROM:	Bradley Nelson, EC/R, Inc.

DATE:	April 20, 2016

SUBJECT:	Comparison of State Leak Detection and Repair Programs 
--------------------------------------------------------------------------------


      The purpose of this memorandum is to provide a summary of state leak detection and repair programs and compare, if available, the potential methane reduction with the methane reductions in the proposed rule. This memorandum only includes information from State programs that were publically available at the time of production of this memorandum, and may not include State programs that are currently being drafted or proposed. In the September 18, 2015 proposed Oil and Natural Gas Sector NSPS (80 FR 56593), the EPA determined that the best system of emission reductions (BSER) for oil and natural gas production well sites and compressor stations was the implementation of a semiannual optical gas imaging (OGI, also referred to as infrared or IR camera) monitoring and repair program. The proposed rule estimated 2.72 tons per year (TPY) methane reductions from natural gas production well sites, 21.1 TPY methane from gathering and boosting stations, 37.4 TPY methane at transmission stations, and 98.7 TPY methane at storage facilities.
      For this memorandum, we reviewed leak detection and repair programs from Colorado, Wyoming, Utah, Ohio, Pennsylvania and West Virginia. A summary of each of the programs are provided in the following sections.
Colorado Regulation 7
      The Colorado Department of Public Health & Environment (CDPHE) finalized regulations in April 2014 to address hydrocarbon emissions from oil and gas facilities, including well production facilities and natural gas compressor stations. For well production facilities and compressor stations, the monitoring frequency is determined by the estimated uncontrolled actual VOC emissions leak from the highest emitting tank or if no tanks are present, the controlled actual emissions from all permanent equipment. The monitoring frequency for fugitives at well production facilities vary depending on emissions. There is a one-time inspection (0-6 TPY VOC), annual inspections (6-12 TPY VOC), quarterly inspections (12-20 TPY VOC w/o tanks, 12-50 w/ tanks), or monthly inspections (> 20 TPY VOC w/o tanks, > 50 TPY VOC w/ tanks). Monthly audio-visual-olfactory (AVO) inspections are also required for well production facilities that do one-time, annual, and quarterly monitoring. For compressor stations the monitoring frequency is annual (0-12 TPY VOC), quarterly (12-50 TPY VOC), or monthly (> 50 TPY VOC). A leak is defined as hydrocarbon concentration greater than 500 ppm. The rule does not cover methane emissions from well sites or compressor stations. Monitoring is required using IR camera, Method 21 device or a state approved monitoring instrument. The first attempt to repair leaks found during monitoring must be made no later than five working days after discovery, unless parts are unavailable or the equipment requires shutdown to complete repair.  If parts are unavailable, they must be ordered promptly and the repair must be made within fifteen working days of receipt of the parts. If a shutdown is required, the leak must be repaired during the next scheduled shutdown.
      The CDPHE estimated emission reductions for annual, quarterly and monthly monitoring, whereas the NSPS emission estimate is for semiannual monitoring. The CDPHE published a summary of the 2014 leak detection and repair (LDAR) results for facilities having to meet the reporting requirements in Regulation 7. Regulation 7 requires well production facilities constructed on or after October 15, 2014 to submit annual reports for LDAR activities in 2014. A total of twenty one companies reported a total of 4,869 leak inspections (1,697 approved instrument monitoring method inspections at well sites, 2,818 AVO inspections, and 41 AIMM inspections at compressor stations) which were conducted at 1,803 well production facilities and natural gas compressor stations. 1,706 leaks were found during the inspections (745 valves, 688 connectors, 86 flanges, 16 pump seals, and 171 pressure relief devices). Of these leaks found, 1,544 (90.5%) were repaired and 80 (4.7%) were placed on the delay of repair list. See Table 1 of the Appendix for emissions reduction comparison. 
Wyoming Chapter 8
      The Wyoming Department of Environmental Quality issued regulations in June 2015 for existing (as of January 1, 2014) PAD facility (location where more than one well and/or associated production equipment are located, where some or all production equipment is shared by more than one well or where well streams from more than one well are routed through individual production trains at the same location) and single-well oil and gas production facilities or sources, and all compressor stations that are located in the Upper Green River Basin (UGRB) ozone nonattainment area. The rule requires operators with fugitive emissions greater than or equal to 4 tons per year of VOC to develop and implement an LDAR protocol by January 1, 2017. Operators must monitor components (flanges, connectors (other than flanges), open-ended lines, pumps, valves, and "other" components listed in Table 2-4 of the EPA's Protocol for Equipment Leak Emissions Estimates) quarterly using a combination of Method 21, IR camera, other instrument based technologies, or AVO inspections. However, an LDAR protocol consisting of only AVO inspections does not meet the requirements of the rule.  No specific repair timeframes are included in the regulation.
       A review of Wyoming's permitting data shows that the average number of wells for the natural gas production permitted facilities was 16 with an average number of components of 5,042, which is eight times the number of wells and components that were used for EPA's model plant (2 wells, 671 components). For oil wells, the average number of wells per well site from the permit data was 2, with a total of 1,003 components per well site. The EPA model plant estimated 2 oil wells per well sites and a total of 127 components. See Table 3 of the Appendix for emissions reduction comparison. 

Utah General Approval Order
      The Utah Department of Environmental Quality approved a "General Approval Order for a Crude Oil and Natural Gas Well Site and/or Tank Battery" on June 5, 2014. This General Approval Order (GAO) requires LDAR for equipment (e.g., valve, flange or other connection, pump, compressor, pressure relief device or other vent, process drain, open-ended valve, pump seal, compressor seal, and access door seal or other seal that contains or contacts a process stream with hydrocarbons) based on annual throughput of crude oil and condensate. Annual inspections are required for sources that have a projected annual throughput of crude oil and condensate combined that is greater than or equal to 10,000 barrels or for sources that do not have a crude oil or condensate storage tank on site, and quarterly inspections are required for sources that have a projected annual throughput of crude oil and condensate combined that is greater than or equal to 25,000 barrels. For sources performing quarterly monitoring, provisions are provided for less frequent monitoring if no leaks are found during a year of monitoring. Repairs must be made within 15 days of finding a leak. A delay of repair is allowed if replacement parts are unavailable (must order parts within 5 days of detection and repair leak within 15 days after receipt of the parts) or technically infeasible to repair without a shutdown (shutdown must occur within 6 months of finding leak or operators must demonstrate emissions from shutdown would be greater than the uncontrolled leaking component.
      The monitoring can be performed using Method 21, a tunable diode laser absorption spectroscopy (TDLAS) or an infrared (IR) camera. A leak is defined as a reading of 500 ppm with Method 21 analyzer or TDLAS, or visible leak with IR camera.
      The methane reductions estimated by Utah are much higher than the methane reductions estimated for the NSPS. This difference is the result of the model plant used by Utah to estimate emissions. The model plant used by Utah assumed 50 valves, pump seals, other, connectors, and flanges for the component counts and the uncontrolled VOC emissions were calculated using Tables 2-4 & 2-8 of the 1995 EPA Protocol for Equipment Leak Emission Estimates. See Table 2 of the Appendix for emissions reduction comparison. 
Ohio General Permit
      The Ohio EPA approved two types of general permits in May 2014 for oil and gas well-site production operations (small flares and large flares) and high volume horizontal hydraulic fracturing for facilities that emit less than 1 ton per year of any toxic air contaminant (not including HAP emitting sources that are subject to MACT subpart HH). Each permittee is required to develop and implement an LDAR program for ancillary equipment (pumps, compressors, pressure relief devices, connectors, valves, flanges, vents, covers, any bypass in a closed vent system, and each storage vessel) that requires monitoring using a forward looking infrared (FLIR) camera or Method 21. Leak definitions vary depending on component (most are 500 or 10,000 ppm). Quarterly monitoring is required for the first year and varies after that depending on performance. Repairs must be made within 30 days of finding a leak but if leaks cannot be repaired within that time frame, the general permit references the delay of repair provisions allowed under NSPS subpart VVa.
      Ohio has also proposed a general permit for natural gas compressor stations that has the potential to leak greater than 10 tons per year of VOC. The general permit requirements for compressor stations are similar to the LDAR requirements for oil and gas well-site production operations.  No emissions data were available for this LDAR program.
Pennsylvania General Permit 5 and Exemption Category No. 38
      General Permit 5 is a General Plan Approval and/or General Operating Permit for mid-stream natural gas gathering, compression and/or processing facilities that are minor air contamination facilities. Exemption Category No. 38 of the Air Quality Permit Exemption List applies to sources located at a well pad. The general permit requires operators to conduct leak detection and repair programs monthly using AVO methods. Equipment to be monitored include: valves, flanges, connectors, storage vessels/storage tanks, and compressor seals. In addition, the general permit requires annual monitoring at wells and quarterly monitoring for compression and processing facilities. Operators must use a FLIR camera or approved device to detect gaseous hydrocarbons leaks. All leaks at production sites, compressor stations or processing facilities must be repaired within 15 days of finding the leak. The general permit also requires operators of certain traditional oil and natural gas sources to report their emissions annually. A search of the Pennsylvania Department of Environmental Protection (PA DEP) website found no emissions information. However, in their comments on the proposed rule, the PA DEP stated that they performed independent cost-effectiveness analyses for LDAR and leak quantification surveys for sources at natural gas compressor station, processing plant and transmission station facilities. Using cost information received from two vendors for the LDAR surveys, the PA DEP estimated the cost-effectiveness for 5 percent leaking components at $41.96 per ton of methane reduced and $2.10 per ton of methane reduced for 100 percent leaking components.
West Virginia Class II General Permit G70-B
      General Permit G70-B is for natural gas production facilities. The permit requires quarterly monitoring using AVO, Method 21 analyzers, IR cameras, or some combination. The AVO inspection shall include, but not limited to, defects as visible cracks, holes, or gaps in piping; loose connections; liquid leaks; or broken or missing caps or other closure devices. If a Method 21 analyzer is used a leak (fugitive emissions of regulated air pollutants) is defined as no detectable emissions (less than 500 ppm). If an IR camera is used no detectable emissions is defined as no visible leaks detected in accordance with US EPA alternative IR camera work practices (40 CFR 60, subpart A). The first attempt at repair must be made within 5 calendar days of discovering the leak, and the final repair must be made within 15 calendar days of discovering the leak. No emissions data are available for this LDAR program
 San Joaquin Valley Air Pollution Control District Rule 4409 
      The San Joaquin Valley Air Pollution Control District requires the development of an operator management plan that establishes inspection, replacement, re-inspection requirements, maintenance, repair periods and replacement retrofit requirements for components at light crude oil production facilities, natural gas production facilities and natural gas processing plants. 
      For manned facilities, the District requires owners and operators to audio-visually inspect for leaks daily, and for unmanned sites the District requires owners and operators to audio-visually inspect for leaks weekly. Additionally, the District requires owners and operators to conduct inspections for leaks quarterly using Method 21. Leaks discovered are required to be repaired within two to seven days of discovery, depending on the magnitude of the leak. An extension of up to seven days is allowed if the leak is minor. Owners and operators are also allowed to apply for written approval to change the Method 21 monitoring inspection frequency from quarterly to annually if they meet specified criteria. Components at oil production facilities and gas production facilities that exclusively handle gas/vapor or liquid with a VOC content of ten percent by weight or less are exempt from requirements. 
      Component leaks that are discovered during monitoring must be repaired or replaced, vented to a closed vent system, or removed from operation as soon as practicable but no later than 7 days for minor leaks greater than or equal to 2,000 parts per million by volume (ppmv) but equal to or less than 10,000 ppmv, 3 days for major leaks greater than 10,000 ppmv but equal to or less than 50,000 ppmv, or 2 days for major leaks greater than 50,000 ppmv. The operator may be allowed to extend the repair period for 7 days for minor leaks greater than or equal to 2,000 ppmv but equal to or less than 10,000 ppmv or 2 days for major leaks greater than 10,000 ppmv but equal to or less than 50,000 ppmv. The repair time for major leaks greater than 50,000 ppmv cannot be extended.
Equivalency of State Regulations
      The final fugitive monitoring requirements in 40 CFR Part 60, subpart OOOOa include semiannual monitoring of fugitive components for oil and natural gas production well sites and quarterly monitoring of fugitive components at compressor stations. Because of the variability of the state rules or general permit requirements and the applicability thresholds, we are unable to determine equivalency with any of these state regulations or general permit requirements. Therefore, we are finalizing the standards with revisions, where appropriate, to expand the source category, promote gas capture and beneficial use, and provide opportunity for flexibility and expanded transparency in order to yield a consistent and accountable national program that provides a clear path for states and other federal agencies to further align their programs.

      
      
      
      
      
      
      
      
      Appendix A
      Summary of State Leak Detection and Repair Requirements and 
      Methane Emission Reductions
      
Table 1 - Colorado Regulation 7

[1] Methane reduction data obtained from Initial Economic Impact Analysis for Proposed Revisions to AQCC Regulations No. 7, November 15, 2013. Colorado methane-ethane emissions estimated to be 7.3 TPY for well production facilities and 15.5 and 44.3 for compressor stations from the CO impact analysis. The emissions were converted to methane using the average percent methane (86.1% for production, 93.1% for transmission) calculated from data from Table 5 of the gas composition memorandum.

Table 2 - Utah General Approval Order

[2] Methane emissions obtained from GAOEmissionCalcs.xls spreadsheet obtained from http://www.deq.utah.gov/Permits/GAOs/oilgas/oilgasinfo.htm. Fugitive emissions were calculated using VOC emission factors from Table 2-4 (uncontrolled) and 2-8 (controlled) of the EPA Equipment Leaks Protocol document. The analysis assumed 50 valves, pump seals, other, connectors, and open-ended lines for the component counts and selected the maximum emission factor from each service. Methane emissions were calculated using the maximum methane/VOC ratio from Utah well site speciation data.
Table 3 - Wyoming Chapter 8

[3] Methane emissions were estimated using WT permit application data for 4 oil production sites well and 5 natural gas well sites. A summary of the permit application information is provided in the spreadsheet WY Fugitive Permit Data.xls. The 95% emission reduction for quarterly monitoring is based on information from WY NSR.
Table 4 - Ohio General Permit Requirements


Table 5 - Pennsylvania General Permit and Exemption Category No. 38


Table 6 - West Virginia Class II General Permit G70-B


