Posted on July 8, 2016
In May 2016, EPA finalized updates to its New Source Performance Standards (NSPS) for the oil and gas industry which amended 40 CFR Part 60, Subpart OOOO and added new requirements (Subpart OOOOa) to those established for Volatile Organic Compounds (VOCs) and sulfur dioxide (SO2) established for this industry sector in 2012. Importantly, the new requirements address reductions of greenhouse gas (GHGs) emissions, specifically methane. In its Executive Summary, EPA discussed the efforts by the agency to “complement” and “improve” the existing rules issued in 2012, stressing the agency’s efforts to engage states and stakeholders and solicit comments prior to its 2015 proposal or the rule. EPA also stressed it worked closely with the Bureau of Land Management to avoid conflicts and evaluated existing state and local programs to attempt to limit conflicts, where possible.
After promulgation of both the 2012 rule and 2013 amendments, the agency received petitions for reconsideration raising numerous issues, including the regulation of GHGs. EPA has addressed some of petitioners’ issues in 2015 amendments addressing storage vessels as well as in this rule, adding standards for methane and addressing storage vessel control device monitoring and testing; initial compliance requirements for bypass devices that divert emissions from control devices; recordkeeping requirements for repair logs for control devices which fail a visible emissions test, clarification of the due date for the initial annual report; emergency flare exemptions from routine compliance tests; leak detection and reporting for open-ended valves or lines; compliance period for leak detection and repair (LDAR) for newly affected process units; exemption to notification requirement for reconstruction of most types of facilities; and disposal of carbon from control devices. However, in a footnote, EPA makes clear it intends to complete its reconsideration process in a subsequent notice.
One interesting aspect of the 2016 rule publication is the extensive discussion of how the 1979 source category listing, “crude oil and natural gas production” is defined. The agency takes great pains to justify its broad authority over the industry to include not only production, but also processing, transmission and storage equipment. The EPA concludes that its category listing need not be revised to support the additions and amendments of this rule (even though it does clarify some wording), and sets out its justification for including the entire sector in its 2009 endangerment finding relating to GHGs.
EPA concluded that the Best System for Emissions Reduction (BSER) is the same for GHGs as it is for VOCs, so there are no changes required for equipment that was covered by the 2012 rule. Newly regulated sources covered by the 2016 rule include: heretofore unregulated hydraulically fractured oil well completions, pneumatic pumps, fugitive emissions from well sites and compressor stations; sources regulated under the 2012 regulation for VOCs for which GHGs are now also regulated (hydraulically fractured gas well completions and equipment leaks at natural gas processing plants); and certain equipment that is used across the source category for which subpart OOOO regulates emissions of VOCs from only a subset (pneumatic controllers, centrifugal compressors and reciprocating compressors), with the exception of compressors located at well sites.
In addition to emission reductions, LDAR utilizing optical gas imaging semi-annually is required for well sites and compressor stations. (Method 21 at a repair threshold of 500 ppm may be used.) Initial monitoring surveys must take place by June 3, 2017 or within 60 days of the startup of production, whichever is later. Repairs must be made within 30 days and a resurvey is required within 30 days of repair. Also, a monitoring plan that covers collection of fugitive emissions components is required to be developed and implemented for well sites and compressor stations. At natural gas processing plants, equipment leaks of methane (GHGs) are subject to the same requirements as those for VOCs. The compliance period begins on November 30, 2016.
And, the rule embraces “next generation” electronic reporting via EPA’s CDX, for enhanced accessibility and transparency to the public, as soon as the forms and systems are available. Professional engineers are required to provide certifications of technical infeasibility of connecting a pneumatic pump to an existing control device and to design closed vent systems.
Finally, there is a complicated discussion of EPA’s interpretation of UARG v. EPA, which merely results in EPA concluding that the rule should not affect applicability of Title V permit or PSD/NSR applicability determinations for “anyway” sources, even though, if not otherwise required to obtain and comply with a Title V permit, emissions of GHGs (methane) alone will not subject a source to Title V permit requirements.
Posted on August 10, 2015
Last year I published an article in Bloomberg BNA entitled “Protection of Judicial Review Watered Down in D.C. Circuit.” I focused on a recent D.C. Circuit ruling (UARG) I hoped would “turn out to be an unfollowed – and eventually forgotten – glitch.” The effect of the “glitch” is to delay interminably judicial review of final Clean Air Act (“CAA”) rule provisions that EPA never hinted might be included in a final rule – even though the un-foreshadowed provisions go into full force and effect.
The Court’s judges must have missed that BNA edition, because they have followed the same rationale at least twice more now – in their Mexichem opinion of May, 2015 and their “Transport Rule” (EME Homer) decision last week.
This regrettable situation arises from the Court’s new interpretation of a CAA provision (§307(d)(7)(B)) which is quoted in full in my BNA article. It begins with the hornbook proposition that you can’t attack a rule’s provision on judicial review on grounds that were not raised during the comment period. It then provides for a process known as a “petition for reconsideration.” If a party can show that it could not have raised an argument during the comment period, EPA must conduct a “reconsideration” process. EPA’s actions in response to the petition are then subject to judicial review. This provision has often been used where EPA supports a final rule with facts or rationale not included in the record when the public comment period was open.
Now consider the following hypothetical. Assume EPA proposes a CAA rule requiring boilers to install a certain type of control device. EPA’s final rule drops the control requirement and simply prohibits boilers from combusting coal, effective two years from the final rule’s issuance. EPA’s proposal never mentioned coal prohibition as an option, and no one suggested it in their comments. So most would assume that boiler owners could then file D.C. Circuit petitions for review and have slam-dunk arguments for vacatur.
As shown in my BNA article, the D.C. Circuit has on many occasions (as recently as December, 2013) done just that. But since then, EPA and DOJ lawyers have advanced what I think is a ludicrous position: when a party believes a final CAA rule provision was issued in violation of notice-and-comment requirements, it cannot pursue judicial review on that issue unless and until it first files a petition under §307(d)(7)(B) and waits for EPA to take final action on that petition.
Unfortunately, the D.C. Circuit has bought this position three times now. Here is how the D.C. Circuit summarized the point in EME Homer last week:
[P]etitioners argue that EPA violated the Clean Air Act’s notice and comment requirements by significantly amending the Rule between the proposed and final versions without providing additional opportunity for notice and comment. Because that argument is an objection to the notice and comment process itself, petitioners obviously did not and could not have raised it during the period for public comment. Under Subsection 7607(d)(7)(B), however, the only appropriate path for petitioners to raise this issue is through an initial petition for reconsideration to EPA.
Note the opinion in effect concedes just how absurd this is. The petitioners “obviously did not and could not” have raised this objection. How can one object to EPA’s failure to propose something that EPA failed to propose?
EPA almost always delays action on §307(d)(7)(B) petitions for years so in the hypothetical above, the coal prohibition would go into effect before judicial review could even begin. Boiler owners would either have to shut down operations or convert to non-coal burning facilities, at which point judicial review would become pointless. The effect: EPA stops coal burning at boilers by declining to propose such a requirement in the first place!
If you think EPA or the D.C. Circuit would out of fairness suspend application of rules in such situations, see the examples to the contrary in my BNA article and read the Mexichem opinion. If you think I am exaggerating about how long it takes for a §307(d)(7)(B) petition to be processed, see the examples in my BNA article. And consider that in last week’s EME Homer opinion, the Court concluded its discussion above by noting that at least one party had filed such a petition but that EPA had not yet acted upon it. That petition was filed in 2011.
Posted on July 10, 2013
In April of 2013 the Arkansas legislature put an end to the ad hoc policy of implementing the NAAQS through stationary source permitting based upon source specific NAAQS modeling. The Arkansas legislature did not need a crystal ball to predict the chaos that was about to occur when the new NAAQS (PM2.5, one hour SO2 and one hour NO2) were swept into the existing Arkansas regulatory program. Arkansas’ environmental agency, the Arkansas Department of Environmental Quality (ADEQ) has relied upon its stationary source permitting program to implement the NAAQS for years, as opposed to relying upon state implementation plan (SIP) development. ADEQ has required every permit applicant to submit air dispersion modeling, and thereby demonstrate that the source will not cause a NAAQS violation. By comparison, EPA generally requires only PSD permit applicants to submit NAAQS dispersion modeling, and requires the states to otherwise address NAAQS compliance through their SIPs.
When Arkansas’ SIP permit procedures were last updated in 2000, minor (non-Title V) sources, and “minor modifications” at major sources were not required to undertake NAAQS modeling. Arkansas’ policies regarding NAAQS modeling were generally in sync with the Clean Air Act and most other states. Over the ensuing years regulatory creep expanded Arkansas’ NAAQS modeling program to the point that nearly every stationary source permit application was involved. ADEQ permit engineers required NAAQS dispersion modeling for minor sources, for minor mods at major sources, and then for any permit renewal—even no change renewals, “just to make sure that the source is still OK.” For example, a facility that had operated in full permit compliance for decades, without any modifications, could face permit renewal problems for no reason other than background conditions or recent meteorological data changed the NAAQS modeling results. Suffice to say this development was unpopular, making permitting expensive, time consuming, and uncertain.
The uncertainty was predicted to become chaos in September of 2012 when ADEQ proposed to drop the new NAAQS into its existing SIP. ADEQ’s “plan” was that the new NAAQS would also be implemented through stationary source permitting, including ADEQ’s expansive NAAQS modeling policies. Of particular concern is the PM2.5 standard, which, at 12 ug/cm3, is already near or exceeded by the background levels measured at the majority of the ambient monitoring stations throughout the state—background that is rarely, if ever, the result of any stationary source activity, but more likely the result of rural road dust and other non-stationary sources.
It became apparent to the regulated community that each permit review following adoption of the new NAAQS would generate ad hoc findings of modeled exceedances of the new NAAQS. By implementing the NAAQS through stationary source permitting rather than SIP planning, ADEQ eliminated any evaluation of regional cause and effect, and precluded any consideration of comprehensive solutions that involve all contributing sources. Under ADEQ’s “plan,” the unwitting permit applicant is forced to stand alone and face the consequences of a failed NAAQS modeling exercise. Concerns raised by the regulated community fell on deaf ears.
The Arkansas legislature stepped in, and in April of 2013 it enacted Act 1302, which required ADEQ to stop “protecting the NAAQS” by requiring stationary source permit applicants to undertake dispersion modeling, except in enumerated circumstances. Act 1302 prohibits ADEQ from using modeling for stationary source permit decisions or requiring retrofit pollution control technology. With the exception of PSD and other limited situations, dispersion modeling can only be used when there is a source or pollutant-specific SIP requirement. The Clean Air Act requires states to develop a SIP “for maintenance and protection of the NAAQS,” and Act 1302 requires ADEQ to implement the NAAQS as required by the Clean Air Act. The legislature did not neuter the agency’s efforts to protect clean air (which was the agency’s unsuccessful lobbying position). The legislature just said quit implementing the NAAQS through ad hoc permit decisions based on source specific air dispersion modeling. The legislature told ADEQ to use its ambient monitoring network, area modeling, and other tools to evaluate NAAQS compliance, and where non-attainment occurs, do the comprehensive planning that is required by the Clean Air Act to address it. Act 1302 was carefully drafted to compliment the Clean Air Act, and serves as a good model for any state facing similar NAAQS implementation issues.
During the two months since Act 1302 has been the law in Arkansas the agency has gone through some needed growing pains. The proposed rulemaking to enact the new NAAQS in Arkansas is being re-evaluated in light of the requirements of Act 1302. Much of the regulatory creep that occurred over the past decade has been curtailed, such that minor sources, minor modifications and no change permit renewals are no longer being required to submit dispersion modeling or demonstrate NAAQS compliance.
There is nothing like the heavy hand of the legislature to bring reason back into agency decision making. It appears that ADEQ now recognizes (much like most other states) that modeling has its limitations, and these minor stationary source projects are not causing, nor are they likely to cause any NAAQS problems. There is still a lot of work to be done as the new NAAQS are adopted, and real SIP planning commences. But sometimes it takes a pre-emptive strike to get the process started on the right track.
Posted on June 25, 2013
Congress said EPA and the States are partners in implementing the Clean Air Act. It’s simple: EPA sets pollutant-by-pollutant standards for clean air (NAAQS) and each State develops and implements a state-specific plan to meet and maintain those NAAQS. Each partner is well-positioned and equipped to perform its assignment and Congress included appropriate “carrots and sticks” in the Act to ensure that both do their job. The Supreme Court has extolled Congress’s partnership approach and EPA routinely professes its deep appreciation of its State partners and their important role. So wassup with EPA suddenly demanding that thirty-six States delete rules about excess emissions during startup, shutdown and malfunction (SSM) that have been EPA-approved for 30 to 40 years?
On February 22, in response to a 2011 petition by Sierra Club, EPA proposed to “call” thirty-six state implementation plans (SIPs) because they contain affirmative defense, exemption, or director’s discretion rules for excess emissions during periods of SSM. EPA’s previous approval of the offending rules wasn’t even a speedbump. EPA also rejected any obligation to connect the offending rules with air pollution problems in the affected States. EPA’s legal position on how the States should enforce their CAA permits was enough to shuck the partnership and impose the federal will. And EPA didn’t even ask nicely. State requests for information about EPA’s consideration of their SIPs were ignored and States were given 30 days to comment on a proposal EPA took more than a year to develop. EPA gave its State partners another 45 days only after more than a dozen State Attorneys General jointly asked for more time and the Senate Committee considering the new Administrator’s confirmation made the same request.
When comments were filed on May 13, thirty affected States filed comments; none of them supported EPA’s proposed call of their SIP. Not even EPA’s regular supporters on issues like tougher NAAQS thought EPA’s dictation was a good idea. Complaints from EPA’s partners ranged from being wrongfully excluded from EPA’s evaluation of their SIP to EPA trampling on the States’ planning and implementation responsibilities to EPA creating a lot of work that could have been avoided if EPA had just talked to them. No amount of spin can make this look good for state–federal relationships.
So why? Well, Sierra Club did ask for it. Maybe because an obvious compliance impact is on emission limits with continuous monitoring and short averaging times like opacity. And maybe because coal-fired power plants always have opacity limits and deleting common excess emission rules will set those sources up for widespread enforcement litigation. Or, maybe the States and the previous EPAs had it wrong for all these years and someone finally straightened everyone else out. Like so many conundrums of this type, it might take some judges to give us the answer.
Pursuant to a settlement agreement with Sierra Club, EPA must finalize the SSM SIP Call by August 27, 2013.