Summary of Comments and Responses Memo

Date: September 21, 2016

To: Oil and Gas ICR Project File

Subject: Summary of Comments and Responses

Oil and Gas Information Collection Request

Docket Id. No. EPA-HQ-OAR-0204

Background

On June 3, 2016 (81 FR 35763), EPA solicited comment on the first draft Information Collection Request (ICR) for the oil and natural gas industry, and we received 66 comment letters during the comment period. We reviewed and considered each of these comments and, as a result, revised the ICR as noted below. This document addresses some of the general comments and concerns, regardless of whether they have been incorporated into the revised ICR. The comments not specifically addressed in this document included more detailed remarks on particular sections of the ICR. A table identifying these comments and how they were addressed is included as an Attachment to this memorandum. A redline version of the ICR supporting statement showing changes made from the proposed ICR (including Attachments showing redline changes to specific questions) has been placed in the docket (see Docket ID No. EPA-HQ-OAR-2016-0204). Any changes made to the ICR as a result of these comments are also reflected in the revised version of the ICR spreadsheet questionnaires.

Definition of Facility

Some commenters mentioned that using the source determination rule as a way to define a facility does not make sense across the industry for existing sources, particularly the production segment. The draft ICR referenced the source determination rule’s clarification of the term “adjacent” to determine the scope of a “facility.” However, commenters stated that there are non-contiguous areas that are adjacent, and flow into the same centralized tank battery but do not all reside within a quarter of a mile of each other. Commenters pointed out that the first draft of the ICR would not require the owners or operators to include information on the centralized production area that would include centralized tank batteries and other processing equipment such as separators, acid gas removal units (AGRUs), and glycol dehydrators under the original approach, unless these operations were within a quarter mile of a well site. Similarly for Part 2, some commenters indicated that owners or operators of facilities that have large numbers of operations within a quarter mile of each other would likely incur a large burden of responding to Part 2 if they were required to include all these operations in the detailed survey.

In order to address these concerns, we revised Part 1 to request limited information on all centralized production surface sites as well as all well surface sites. For Part 2, we request detailed information on the individual well sites that contain the randomly chosen wells, as well as detailed information on the centralized production surface sites that the selected well site feeds.

Statistical Sampling Approach

We received many comments suggesting changes to our statistical sampling approach for Part 2 respondents and we have modified our strategy as a result. In the first FR notice, we proposed two options for stratifying the production segment for Part 2 sampling: by geographical region and by gas-to-oil ratio (GOR). Many commenters suggested that the GOR approach for stratifying the production segment was the more preferable method. Several commenters also suggested adding a new category for stripper wells with low production, which are defined as wells that produce 15 barrels of oil equivalent (BOE) or less per day on average over a 12-month period. We decided to use the GOR approach for stratifying the production segment; however, we have also included another set of categories for stripper wells, stratifying both the non-stripper well and the stripper well populations by GOR using the proposed GOR groupings. For the stripper well populations, we doubled the acceptable error so that the sampling size for the stripper populations is about 96 wells. Having separate categories for non-stripper wells with the GOR groups will help us collect better information on higher-producing wells and cost-effective emissions control opportunities. More information on this approach is detailed in the supporting statement.

Publish Facility Lists

In the first draft of the ICR, we assumed that response rates would be approximately 75 percent due to inaccurate contact information and facility closures. One way to increase this response rate is to have better facility contact information. Several commenters suggested that we publish our entire list of facilities and have individual companies adjust their addresses as appropriate. We agree that this approach should provide us with better information and have therefore changed our response rate assumption to 90%. Facility lists are now available at: https://oilandgasicr.rti.org/ where owners and operators may adjust and correct their appropriate address information prior to the mail out of the survey forms.

Gathering and Boosting Facilities

In the first draft of the ICR, we recognized that facility information on gathering and boosting facilities was incomplete. Commenters provided several suggestions for acquiring necessary gathering and boosting facility information. One approach was to wait until 2017, when EPA will receive a more complete list of gathering and boosting stations from the Greenhouse Gas Reporting Program (GHGRP), since these facilities will be reporting for the first time for the 2016 calendar year. Another approach was for EPA to publish a Federal Register notice that all gathering compressor stations must complete Part 1 of the ICR. After EPA receives the list, commenters suggest that EPA would then randomly select facilities on the list to receive a Part 2 questionnaire.

Another suggestion by commenters would be for EPA to require natural gas processing plants at the highest level parent company to select a certain percentage of its gathering compressor facilities using a random number generator imbedded in an EPA-provided spreadsheet. Parent companies of natural gas processing plants are likely to own most gathering facilities, so that the majority of gathering and boosting facilities should be captured. In addition, commenters suggested that EPA should require natural gas well production facilities to randomly select gathering facilities that they may also own. Because of our need to acquire information in a timely manner, we have decided to follow the third approach suggested by commenters to require highest level parent companies of production and natural gas processing facilities to randomly select a percentage of gathering and boosting facilities to respond to Part 2.

Extend Deadline to Complete Part 1 and Part 2

Many commenters requested more time to respond to both Part 1 and Part 2. The original time to complete Part 1 and Part 2 were 30 days and 120 days, respectively. Several commenters requested 60 days for Part 1 and from 180 days to 240 days for Part 2. Reasons to extend the Part 1 response period included the need for more time to access data spanning across several databases and hard copies for responders of many facilities. Numerous commenters also indicated that some elements that may have appeared to not be burdensome on the surface such as providing distances from facilities to field offices and providing information on availability of electricity would require GPS mapping and detailed records searches, which would take a lot more time and effort than EPA accounted for. Reasons to extend the Part 2 response period included: time needed to conduct field data collection and sampling and analysis, especially time to familiarize industry and laboratories with the CARB sampling and analysis methodologies; the possibility of inclement weather and delays in conducting field work; avoiding overlap with GHGRP and permit compliance reports, which are heavily weighted in the first quart of 2017; avoiding winter heating season where personnel are already stretched thin with holiday seasons and personal vacations; and time needed to hire contractors to complete the ICR.

For Part 2, while we acknowledge commenters’ concerns with completing the ICR within the required timeframe, we are still keeping the response deadline to 120 days from receipt of the ICR. As discussed earlier, we expect that limiting the scope of Part 2 by revising the definition of “facility” to a “well site facility” will reduce the burden for some operators. For Part 1, we note that we have eliminated reporting of some elements, such as electricity, distances, and the occurrence of manned facilities that commenters indicated would require significantly greater burden than expected. Because we are simplifying Part 1, we see no reason why commenters could not meet the original 30 day response deadline. Therefore, we are also not extending the timeframe for responses to either Part 1 or Part 2 at this time.

Moving Data Elements from Part 1 to Part 2

Multiple commenters mentioned there would be added burden in collecting information on manned facilities, electrification, and access to gas infrastructure. In order to decrease this burden, we have determined that it would be best to move these data elements into Part 2. Part 2 will be collected completely by the e-GGRT system which is capable of storing and managing confidential business information. Further, some commenters requested that we ask more detailed questions about manned facilities, electrification, and access to gas infrastructure, indicating that more information is necessary to understand these issues. We eliminated these questions from Part 1 and added more detailed and directed questions for these elements in Part 2.

CARB Method

Many commenters requested that EPA provide flexibility in sampling requirements for separators to determine potential flash emissions from atmospheric tanks. Flashing emissions occur when pressurized liquids (crude oil, condensate, and produced water) are routed into atmospheric storage tanks. To assess flashing emissions, we are requesting that selected facilities in Part 2 use the California Air Resources Board (CARB) method for sampling and analyzing pressurized liquids from any separators prior to atmospheric storage tanks. Comments on the CARB method include: the current amendment to the CARB method is still in the draft stage and has not had adequate peer review; it has been applied primarily for produced heavy oil sampled at low pressure separators; only four laboratories are familiar with the CARB method; some laboratories have mentioned that the method will not be appropriate for areas in the Uintah and the Denver-Julesburg Basin; and finally, the high cost of the CARB method.

While we understand the issues that have been raised, we intend to require a consistent set of procedures for sampling and analysis. The CARB method simply clarifies an existing industry sampling protocol (GPA method 2174) for the purposes of sampling pressurized liquids from separators, and the CARB method utilizes existing industry methods for analysis. One commenter requested simply using GPA method 2174, as opposed to the CARB method. GPA method 2174, however, does not specify a volumetric flow rate requirement into the sampling containers. We believe it is important to limit the sampling rate because samples taken at a lower rate are likely to yield more reliable results than those taken at faster rates because of the minimized risk of flashing within the canister. Consequently, we are requesting that selected facilities use the latest version of the CARB method, but use the volumetric flowrate from the December 2010 version that requires a 60 mL/min sampling rate as opposed to later versions that require a 120 – 180 mL/min rate. In addition to concerns about the sampling rate, some commenters stated that the CARB method would not be appropriate for the Denver-Julesburg basin nor for non-heavy oil wells.. However, preliminary information suggests that the CARB method can and does produce reliable results for a range of GORs. While a higher GOR may make extracting the sample more difficult, this method, with volumetric sampling rate limitations and clear instructions, should provide a consistent sampling approach.

Several commenters also requested that the overall burden for pressurized liquid sampling be reduced, particularly for downstream segments. The commenters suggested that the natural gas composition in transmission pipelines is uniform and that only one representative sample should be required for any company. They also noted that separators in this industry segment do not have significant liquids accumulation and do not typically have ports for liquids sampling. In response to these comments, we have added a provision that if the separator’s liquid output (oil, condensate, and water) is less than 10 bbl/day, then sampling is not required. However, we are concerned that this may effectively exempt all separators in the transmission compression industry segment. Therefore, we are requiring parent companies of selected transmission compression facilities that do not have any separators with liquid output exceeding 10 bbl/day to select any separator at a selected facility to perform a single flash gas analysis using the CARB method. We expect that allowing parent companies the opportunity to select the separator to sample will eliminate the need to install sampling ports for separators that do not have them.

Some commenters asked if separators are in series, which separator should be sampled via the CARB method. Our intent is that, for separators in series, only the separator that feeds into an atmospheric storage tank should be sampled. We have clarified this question in the instructions to the Part 2 survey.

GHGRP Duplication

Many commenters stated that there are many data fields within the ICR that will also be reported to the EPA during the Greenhouse Gas Reporting Program (GHGRP). We agree that reporting the same information on the ICR and the GHGRP would be duplicative. On Part 2 of the ICR, we ask for the company’s GHGRP ID number. If this information is provided, the spreadsheet blacks out the cells containing information that is identical to the data elements already reported in the GHGRP.

Additional HAP Data

Some commenters mentioned that the focus of the ICR was skewed towards only collecting data on methane emissions, and these commenters requested that more information be collected on HAPs. While we acknowledge that the primary goal of this ICR is to collect data on emissions of greenhouse gases, information that EPA plans to collect will also support the Agency’s ongoing efforts to evaluate hazardous air pollutants (HAP) emitted from the oil and gas industry. EPA is currently in the process of evaluating issues raised in several petitions seeking administrative reconsideration of the 2012 revisions to the National Emission Standard for Hazardous Air Pollutants (NESHAP), 40 CFR Part 63, Subparts HH and HHH. The ICR includes the collection of information that would be relevant to this effort.

Storage wells and Injection Wells

One commenter indicated that EPA only asked for information on gas wells, and therefore it was unclear on how operators of storage wells would report information. We agree with this commenter and have added a tab for storage and injection wells. We also request information on injection wells. For storage wells, we included questions regarding the storage formation type, the total and working gas capacity, and the deliverability and injection capacity. For injection wells, we included questions regarding the type of injection well, the formation type and capacity, and the maximum injection rate.

Well Economics

One commenter suggested that we ask for operational costs associated with marginal wells so that we can understand the impact of potential control requirements. In response, we have added questions to the ICR that will help us better understand the economics of oil and natural gas wells.

Confidential Business Information

Several commenters asserted that the information sought by the surveys that accompanied the first draft of the ICR could potentially be confidential business information (CBI), and that CBI ought to be protected from public disclosure by EPA. To support the position that some of the data elements in the surveys should be withheld from public disclosure, comments alleged that making these data elements available to the public could have an unfair impact on competition, compromise trade secrets, breach contracts, implicate national security, and/or violate the law (such as CIPSEA). Some commenters asserted that certain data elements in the surveys were not “emissions data” and thus cannot be excluded from confidentiality protection under section 114(c) of the Clean Air Act. These comments were submitted in response to the discussions related to the treatment of confidential information in the Federal Register Notice (81 FR 35763), the Support Statement for Public Comment, and a memo included in the docket titled “Data Category Assignments and Draft Confidentiality Determinations for Data Elements in the Draft Information Collection Effort for Oil and Gas Facilities” (CBI Memo).

EPA is taking this opportunity to clarify that no confidentiality determination is being made via the ICR process. The charts that were included in the CBI Memo were intended for informative purposes only and were meant to assist with the solicitation of comments from possible industry respondents. EPA appreciates the comments and emphasizes that all information submitted to the Agency in response to the ICR surveys will be managed in accordance with applicable laws and EPA’s regulations governing treatment of confidential business information at 40 CFR Part 2, Subpart B. Any information determined to constitute a trade secret will be protected under 18 U.S.C. § 1905.

Legal Authority

Several commenters proclaim that the stated purpose of the ICR, to gather information to support future regulatory actions, does not comport with the Clean Air Act (CAA) and therefore cannot justify the ICR under the Paperwork Reduction Act. Put another way, commenters allege that EPA does not have the legal authority to utilize the information that the Agency plans to collect via the ICR to regulate the oil and gas industry. To support this argument, commenters state that EPA cannot promulgate regulations in accordance with CAA § 111(d) for the oil and gas industry because this industry is already regulated by various entities, including by states and the EPA, under different statutory authorities, including CAA § 112. Moreover, some commenters argue that EPA must make an additional endangerment finding, must promulgate additional CAA § 111(b) standards, and/or revise the source category listing before promulgating a 111(d) standard.

EPA disagrees with these comments. The Agency does have the authority to utilize the data that it plans to collect via the ICR for numerous purposes, including, but not necessarily limited to, exercising EPA’s authority under CAA § 111(d). EPA does not agree that it lacks the legal authority to regulate the oil and gas industry for greenhouse gases pursuant to 111(d) due to the fact that that same industry is subject to regulations for hazardous air pollutants under CAA § 112. Commenters have not cited any compelling authority to substantiate their claims that EPA lacks the legal authority to effectuate the purpose of the ICR.

CAA § 114 grants the EPA authority to collect the information sought via the ICR. Per the CAA, the Agency has legal authority to collect this information from the oil and gas industry. Moreover, the purpose of the ICR serves a goal that is within the bounds of the Clean Air Act. EPA acknowledges that some issues raised by commenters, including the interaction of CAA sections 112 and 111(d), are contemplated in active litigation. EPA has stated its view on these issues in its filings in West Virginia v. EPA, No. 15-1363 (and consolidated cases) (D.C. Cir.) and in the Clean Power Plan final rule (at 80 FR 64662, 64710-64715 (Oct. 23, 2015)). As such, EPA does not find it necessary to restate the Agency’s arguments here.

To be clear, the data collected via the ICR will only be utilized in accordance with applicable laws, regulations, and agency guidance. This data has utility, is not duplicative with currently available information, and has value that outweighs the ICR’s burden. The analysis required by the Paperwork Reduction Act is included in the docket for the second draft of the ICR, as it was for the first draft. Please see the Support Statement and other docket materials for further discussion on this matter.

2