National Oil and Gas CommitteeMonthly Meeting– Minutes

June 12, 2014

Attendees

Tom Moore, WESTAR

Theresa Pella, CenSARA

Julie McDill, MARAMA

Roy Huntley, OAQPS EIAG

Mike Pring, ERG

Regi Oommen, ERG

R3: WV (Bob Betterton), PA (Sherry Bogart), MD (Tim Wallace), Allegheny County, PA (Tony Sadar), VA (Thomas Foster)

R4: AL

R5: MI (Rick Dalebout, Dave Mason), OH

R6: OK (Mark Gibbs), TX, NM (Mark Jones, Mike Snyder)

R7: KS (Lynn Deahl)

R8: CO (Dale Wells), ND

R9: CA (Chris Nguyen)

Moderator: Mark Gibbs of OK

Agenda

1.Approve minutes of May meeting

2.Summary of final O&G submittals to the 2011 NEI v2

3.Summary of all comments and corrections/revisions received on draft V2 (tool and report)

4.Status of API comments on the tool

5. EPA O&G white papers

6. 2018 Modeling Platform / Growth factors

See Sections 4.2.4 and 4.2.5 in:

7.Preliminary discussion about the 2014 version of the tool (if time permits)

8.Next meeting July 10 at 2:00pm – 3:30pm, same call in number and confirmation number as before

The following are the minutes of the June meeting.

Approval of May minutes

Minutes approved without comment, other than a spelling correction.

Oil and Gas emissions submittals to NEI – Roy Huntley said a number of states updated their submittals for v2 of the 2011 NEI. The states who submitted v2 data were MI, CO, WY, TX, OK, and WV. TX requested that EPA augment the TX data with 6 SCCs, which is what they did for v1. For other states such as ND and NM, Roy will shortly be running V2 of the tool to generate data for the 2011 NEI v2. Roy confirmed that this version of the tool includes WRAP basin factors and data.

Comments and Corrections/revisions to the V2 tool and report – Roy said he did not receive any comments on the report. Some additional changes were made to V2 of the tool. Other than the ones between V1 and V2 that were discussed previously, some engine emission factors (see below)were revised per comments from OK, there was one SCC that was an error in the early V2 draft tool and which was corrected (bad = 2310123606; good = 2310023606), and some activity data and some basin factors for CBM for WV were changed (following discussions with Megan Murphy and Bob Betterton of WV).

The following is a copy/paste from an email from OK regarding some combustion emission factors. ERG changed the factors to the AP42 values.

SCC and Source

2310010100 Heater treaters

2310021100 Gas well heaters

2310023100 CBM well heaters

2310021400 Dehydrator reboilers (natural gas wells)

2310023400 Dehydrator reboilers (CBM wells)

The factors used in the Tool were derived from Profile 0003 from the SPECIATE database. These factors differ from the AP-42 factors as shown in the following table.

Pollutant / Emission Factor Used in Version 2.0 of the Tool
(lb/MMscf) / AP-42 Emission Factor
(lb/MMscf)
Benzene / 0.22 / 0.0021
Formaldehyde / 0.44 / 0.075
Toluene / 0.11 / 0.0034

Chris Nguyen of CA wanted to make sure EPA was not going to use the tool data for CA. Roy told him that EPA was going to use the data CA had already submitted. Chris said that he has new improved data for 2011, though he’s still not sure when he can submit to EPA. Roy told him that it is too late to make it into v2 of the 2011 NEI, but to send the data to EPA anyway. Chris got several questions about his updates as they represent an important opportunity to evaluate the tool. CA doesn’t use SCCs so Chris had to do a lot of mapping of their codes to EPA SCCs. Chris said he is using the EPA tool and revising the factors based on California-specific information from the districts. VOC increases compared to their NEI submission though not nearly as high as using the defaults in the tool. Roy thought that was very good to hear and requested that Chris send his tool changes to EPA so that EPA can incorporate them into the EPA version of the tool. This would help out on the 2014 tool estimates.

Tom Moore thought it fantastic that CA can provide information that reflects how there is going to be different emission rates for different processes in different parts of the country. This is going to be important for prioritizing the next level of work on the tool and should be on the agenda for a future meeting. For example, for 2014, how do we handle the new rules and regulations that are coming out of the states? Mark noted that making these variables more accessible to states with limited resources could perhaps be done with a Microsoft Access form interface in the next version of the tool.

Roy said he would like to see more consistent SCC’s and the same level of disaggregation of processes in the national inventory. Mark said we need to take a good look at the SCC table for the 2014 NEI. There are situations where some sources are split too much and others are lumped too much. Also, the crosswalk between point and area sources will become more and more important as more wellsites are inventoried as point sources.

Sherry Bogart of PA asked a clarifying question regarding whether EPA changed any PA data. Roy said that EPA is still using the data the PA sent in for v1 (which did include CBM emissions although these were not split out). There have been no changes to the PA data for the 2011 NEI.

Roy was asked about plans to release V2 of the tool publicly, and his plan is make it public around September 1 to coincide with the public release of 2011 NEI v2. It was pointed out that the draft V2 was accessible through the CHIEF website. Roy confirmed this and has removed the link to V2. The link to V1_4 has been restored. This was a mistake caused by miscommunication at EPA.

Mark Gibbs and Theresa Pella want ERG to walk the committee through the report before it is released in September. Mike and Regi are both out of the office for the July 10, 2014 meeting, so thiswill have to be done at the August 14, 2014 meeting. Mark suggested that the final draft be sent out in mid-July to give committee members sufficient time to review it before the August meeting. Ideas for improving the report were discussed. Among the changes needed is anunambiguous statement (preferably early in the report) that the NEI is a blend of data provided by the states (which may or may not make use of the tool) and EPA data (used only in the absence of a state submission), so that it is clear the tool is not equal to the NEI. However, Mike Pring pointed out that the report was never intended toprovide a detailed discussion about how the NEI is developed. The focus of the report was intended to be the tool itself. While that is true, there was a consensus that it would be helpful to include a discussion of where the tool would (and would not) be used in the NEI. Absent that sort of discussion in the report, it is likely that various parties will assume that the tool is synonymous with the oil & gas sector in the NEI. It was also suggested that any QA of the tool output should be included in the report. In particular, it would be helpful to include a comparison between the tool output and detailed state data (such as that available from WY and CO). Changes to the tool since V1 should also be included in the report.

EPA responses to API comments –EPA personnel are still internally reviewing Roy/ERG responses to API comments. EPA will probably meet with API mid-July. Roy cannot share this information with the committee until after the API meeting. Mark requested that these responses be shared with the committee after EPA meets with API to help states working with their industry groups. There was some discussion of the API comments and how we need to improve the gas speciation data. PA is pulling available data from their permit files on gas composition and so is WV. Tony Sadar of Allegheny County is convinced industry knows this information and that it should not be that difficult to get it. Perhaps a range of emission values per basin/area could be obtained from industry. Gas composition data from permit information as well as TANKS and GlyCALC runs might be low hanging fruit. Mark warned about needing to QC such data – for instance, incorrectly including a flash gas analysis can skew the average gas composition data. Lynn Deahl of KS pointed out that the importance of calculating percent composition on a weight basis (as opposed to a mole or volume basis, which is what is usually reported) when performing certain types of emissions estimates and that the values are definitely not the same. Perhaps someone knows of a guide/spreadsheet calculator for doing the conversions, including a handy list of molecular weights.

Roy gave an overview of the various offices within EPA that deal with Oil & Gas emissions.

  • OAP in DC: two groups who work on Subpart W reporting (facility GHG emissions) and the top-down national GHG inventory;
  • OAQPS in RTP:
  • Emission Inventory & Analysis Group – Roy’s group which works with the states on CAP & HAP inventories;
  • NSPS rule developers and NESHAP rule developers – these groups are in Roy’s building and he knows them better than the OAP folks in DC.
  • EPA Enforcement – Roy does not really work this group much at all.

Tom Moore wondered which of these groups might be proposing new requirements such as new rules which may be based on the methane white papers. He hoped that there would be an upfront discussion with states to help anticipate what are expected to be complex requirements that are coming down the road.

Julie asked if anyone in EPA has looked at the HAPs generated by the tool. Roy said not so much, although the NATA risk analysis is being processed and the 2011 NATA will be using the 2011 NEI v2, which contains lots of oil and gas production data that might not have been in previous inventories. The NATA process will evaluate risk due to HAPs, so we might get great feedback from NATA. Roy noted that he’s had questions about the increased NOx emissions that are now in the NEI because O&G emissions are being inventoried more comprehensively than previously but he has not had questions about increased toxic emissions so far. Concern was expressed that, similar to how EPA approached companies directly during the Risk & Technology Review (RTR) process, the NATA group might contact companies directly rather than state emission inventory staff first.

Mark noted that the emission inventory community is focused on deriving the best possible estimate of actual emissions. He expressed concern that other EPA groups might be more focused on remaining tied to certain emission factors because they had developed regulations based on those numbers.

The white papers developed by the rule writers at OAQPS were discussed. Dale Wells said CO commented on the leaks paper – atmospheric storage tanks with flashing emissions are always going to leak. The committee is interested in comments from industry and would like to read them as soon as possible. After the meeting, Roy contacted Chris Frantz of EPA who said the comments will be made available sometime this summer, but no firm date has been. Theresa will follow up with EDF about the status of the UT/EDF/Industry field studies. Julie will follow up with a group at Cornell which has studied emissions from abandoned wells.

Mark Gibbs summarized EPA’s 2018 O&G projections which are based on regional projections of production changes (perhaps too conservative?) and new controls applied only where there is expected production growth (perhaps also too conservative?). However, while it is important to understand changes in industry practices and the rapid development of new fields and basins, quantifying their effects on projected emissions for specific SCC’s in specific areas may be very tricky. Given this, the EPA approach may well be reasonable for a national modeling study. Mark asked if anyone is planning to submit comments on the 2018 modeling projections. Tom Moore said he will submit comments on WESTAR’s behalf. Atmospheric studies show a lot more VOC emissions than the bottom-up inventories indicate which, in contrast to such discrepancies for methane emissions, is more likely to be tied to underestimating O&G emissions. Tom noted that we probably understand the activity data pretty well at this time, but top-down studies force us to evaluate the uncertainty that remains in the emission factors.

Dale Wells commented on the recent work comparing top-down versus bottom-up inventories by NOAA et al.that was discussed at the GEIA Conference last week. There was fairly good agreement in dry gas like the Marcellus Shale. However, in wet gas basins like the Uinta in Utah and Denver-Julesburg in Colorado (Niobrara Shale), the top-down inventories indicate much greater VOC emissions (by factors from 2 to 4) than the bottom-up analyses. NOAA also looked at the Barnett Shale in Texas where there is both wet and dry gas, and will be doing more work on the Marcellus Shale and the Bakken Shale in North Dakota. Dale’s theory is that most of the difference in wet gas areas is caused by underestimating flashing emissions at condensate tanks. He believes flashing needs to occur in the separators so that the flash gas does not reach the tanks. Interestingly, in the Uinta basin NOx may be overestimated in the 2011 NEI v1 because there is lot less wellhead compression than the tool currently shows.

Julie mentioned that in her area, she is seeing desiccators used in the place of glycol dehydrators. Apparently, when the gas flow from wells decreases, operators are taking out glycol dehydrators (for which numbers per well may be overestimated in the tool in WV) and using desiccators instead. Industry personnel report that the desiccators should show no or very minimal emissions.Julie would be interested in hearing about other states experience with this equipment which, where present, could represent a major drop in emissions. In addition, she has observed that operators in her area are routing emissions from their pneumatic devices to flares. These findings illustrate the need to keep up with changing industry practices, particularly with respect to future projections.

Next meeting Thursday,July 10, same time, same phone number.