Good morning. My name is Bob Slaughter. I am General Counsel and Director of Public Policy for the National Petrochemical & Refiners Association (NPRA). I am very pleased to be here this morning to address the need for reform of the "New Source Review/Prevention of Significant Deterioration" ("NSR") program under the Clean Air Act on behalf of both NPRA and the American Petroleum Institute ("API").
NPRA's membership includes virtually all U.S. refiners, as well as petrochemical manufacturers using processes similar to refineries. Our members own and/or operate almost 98 percent of U.S. refining capacity. NPRA includes not only the larger companies, but also many small and independent companies. API is a trade association that represents more than 400 member companies involved in all aspects of the petroleum industry including refining, exploration and production, transportation, and marketing industries. The NSR program significantly affects NPRA and API member companies.
II. OVERVIEW
The refining industry has dramatically reduced its direct and indirect emissions since Clean Air Act regulation began in the 1970s. Between 1980 and 1996, according to EPA's own figures, the refining industry decreased its criteria pollutant air emissions by 74%. Congress and EPA have required us to attain additional dramatic emissions reductions in the next few years.
We will meet these obligations. However, both our ability to meet them and our ability to efficiently make and deliver the products we refine to consumers is currently threatened by the likely prospect that EPA will claim that almost any operational change we make triggers "new source review" ("NSR") under the Clean Air Act.
Congress enacted the NSR program in the 1970s to ensure that sources that significantly increase their emissions must install technology to control that increase. You may well ask how an industry with the continuing record of dramatic emissions reductions which I have mentioned could be so affected by a program intended to control emissions increases.
The answer lies in the manner in which EPA now administers this program. EPA applies NSR to many changes that will never cause emissions increases, even to changes that will reduce emissions. Moreover, EPA's practice of defining critical elements of the program by guidance rather than through rulemaking -- or not defining them at all -- has created a situation where it is effectively impossible for even the most diligent refiner to determine when NSR applies and when it does not.
This state of affairs has created an urgent need for NSR reform. The policy consideration is this: EPA's reinterpretation of NSR threatens our ability to make the plant changes necessary to comply with important environmental requirements for stationary sources and fuel reformulation.
I would now like to address these points in more detail.
III. HOW THE NEW SOURCE REVIEW PROGRAM THREATENS FUTURE ENVIRONMENTAL PROGRESS
The refining industry now faces extensive new Clean Air Act regulations that will take effect in the near future. These include requirements both for control of refinery emissions, and for the reformulation of gasoline to remove sulfur and selected "air tonics". It seems certain in addition that EPA will require the reformulation of diesel fuel, and likely that Congress or EPA will consider requiring the phase-down or elimination of MTBE from gasoline.
Attached is a chart titled, "Cumulative Regulatory Impacts on Refineries: 2000~-2010" reflecting these requirements in more detail.
Implementing these upcoming programs is very important to EPA's environmental agenda. The refining industry's environmental progress to date is very impressive. Between 1980 and 1996, according to EPA's own figures, the refining industry decreased its criteria pollutant air emissions by 74%, while refining capacity decreased by only 16% (see attached chart titled "U.S. Refinery Emissions Reductions"). These figures underestimate our current emissions reductions, since they do not include the impact of many regulations issued under the 1990 amendments to the Clear Air Act. Nor do they reflect the significant emissions reductions that have been obtained through the use of reformulated gasoline produced by our industry. EPA expects emission reductions achieved by future fuel reformulation and stationary source emission requirements to be even greater. EPA estimates that just one of the upcoming product reformulation regulations, the Tier IV gasoline sulfur reduction requirements, will produce emission benefits equivalent to removing 164 million cars from the road.
EPA has recognized that refiners face tremendous logistical challenges in meeting the ambitious goals and deadlines of these important new regulations. To implement the regulations, refiners must make many infrastructure and process changes. For each change, refiners must determine whether NSR permitting and controls are required, and then obtain required permits before commencing any construction. Because it is now effectively impossible to determine when an NSR permit is required, and extremely time-consuming to obtain a permit, the current state of the NSR program directly threatens the industry's ability to meet Congress' deadlines for this suite of new regulations.
In order to meet Congress' ambitious goals and deadlines for upcoming Clean Air Act regulations, it is essential that refiners have a flexible and efficient permitting process. The current NSR program prohibits this and must be substantially reformed. Moreover, as discussed below, EPA's new interpretation of NSR applicability threatens continued environmental progress, as it penalizes refiners for making changes that decrease emissions.
IV. THE PROBLEM WITH EPA'S CURRENT APPROACH TO NSR APPLICABILITY
NSR is one of the most complicated regulatory programs ever created. EPA has recognized this and initiated the reform process to simplify and rectify the program. In this hearing, however, I want to focus on certain aspects of the program. EPA's current approach to NSR applicability makes it extremely difficult for refiners to determine when NSR permitting and controls are required and leaves refineries in enforcement jeopardy unless they consider NSR for any and all operational changes. As a result, the program is an untenable burden on state permitting authorities and refineries and threatens their ability to implement Congress' future environmental goals in a timely manner.
A. Background
Under the Clean Air Act and EPA's regulations, NSR is triggered by any "physical change or change in the method of operation" of a source that increases its emissions by a significant amount.\1\ If a physical/operational change does not itself significantly increase source emissions, or if the source "nets out" the change by offsetting emissions reductions in other places, then, under the law, NSR does not apply.
\1\ Clean Air Act 111(a)(4); 40 CFR 52.21(b)(2).
If a change does cause a significant emissions increase, NSR requires the source to get a permit before beginning construction of the change, install emissions control technology on the change, and perhaps meet other requirements as well. It takes eighteen months to two years on average to get an NSR permit.
EPA officials have recently made public statements that many changes at refineries over the past twenty years required NSR permits but that none were obtained. Since NSR is only triggered by an emissions increase, and given that the refining industry since 1980 has experienced dramatic emissions reductions, any such EPA claim of widespread NSR noncompliance would appear inconsistent with the basic intent of the Clean Air Act.
EPA has not disclosed information to support its claims of widespread refinery NSR noncompliance, and so we cannot comment on them specifically. However, EPA has reinterpreted its NSR rules in recent years so as to enable the Agency to allege that virtually any change a source might make requires NSR permitting and controls, even if emissions have not increased. In creating NSR, Congress intended that facilities that significantly increase emissions, by adding new equipment or making major changes, must install the latest pollution control equipment. NSR was never intended to impose new controls on older facilities simply because of their age and need for routine maintenance.
B. The Elements of EPA's Current Approach to NSR Applicability
1. The "Actual-to-Potential" Test
EPA uses the "actual-to-potential" test to determine whether a source has significantly increased its emissions. As explained below, the "actual-to-potential" test is bad public policy because it provides an incentive for sources to maximize their emissions, and punishes them for minimizing their emissions. The "actual-to-potential" test is also inconsistent with Congress' intent for the NSR program, because it requires a source to add controls when its emissions do not increase significantly or even when they decrease. Congress intended NSR to apply only when a source significantly increases its emissions. The "actual-to-potential" test is a result of EPA interpretation and should be altered or abandoned through the reform process.
In determining whether a "physical/operational change" at a source caused a significant emissions increase, EPA does not compare actual emissions before the change with actual emissions after the change. Instead, EPA compares actual emissions before the change with potential emissions -- that is, the maximum amount the source could emit -- after the change. According to EPA, NSR is triggered whenever the difference between "past actual" emissions and "future potential" emissions is "significant".
This "actual-to-potential" approach always overstates the emissions increase caused by a physical/operational change. There will always be a difference between "past actual" emissions and "future potential" emissions at any source that complies with its emissions limits. Sources must maintain a buffer between actual emissions and potential (permitted) emissions to avoid inadvertently exceeding the permitted limit. A source that cares about its environmental performance will go further and try to minimize its emissions at all times, and EPA should encourage this. However, EPA's "actual-to-potential" test punishes sources for doing so.
The "actual-to-potential" test penalizes efforts to maintain a compliance margin or minimize emissions and uses them to trigger NSR for changes that do not really increase emissions, or even decrease emissions. As a source lowers its actual emissions, the difference between those actual emissions and potential emissions gets greater. EPA counts that difference as an emissions increase that triggers NSR whenever that source makes a physical/operational change. Thus a source is rewarded for maximizing emissions and deterred from minimizing emissions. Additionally, under this approach, a process unit at a source can trigger NSR repeatedly even when its emissions do not increase at all or even decrease\2\.
\2\EPA policy forbids sources that engage in "emissions trading" from counting their compliance margin as an "emissions credit" when it would be advantageous to the source to do so. See Draft Economic Incentive Program Guidance (September 1999) at 81, 106-107. But EPA NSR policy counts that same compliance margin toward non-compliance with NSR and uses it to trigger permit requirements.
Although the "actual-to-potential" test is inconsistent with the intent of the statute, EPA requires that sources use this method and only this method to determine whether an emissions "increase" has occurred. EPA has found it to be a convenient way to require controls on more sources, whether or not their emissions have increased significantly.
2. What is a "Physical Change or Change in the Method of Operation"?
As we have shown, the "actual-to-potential" test creates phantom emission increases. As a result, almost any change labeled a "physical change or change in the method of operation" of a source will trigger NSR even if in reality it will not increase emissions at all, or even decreases emissions.
The question then become, what is a "physical change or change in the method of operation". EPA's application of the term is a moving target. Small repairs and improvements are needed constantly at complex sources like refineries. Under EPA's current approach, it is impossible to determine when such a repair or improvement will be counted as an NSR- triggering "physical/operational change", and when it will not. We know that EPA is increasingly aggressive in its claims that such repairs and improvements trigger NSR. However, that change in position has never been subject to public notice and comment, as the Administrative Procedure Act requires. Some of its elements have not even been issued as guidance. In some cases, we do not even know what they are.
Let me offer two illustrations of these points, picked from many possible candidates.
- EPA's rules provide that "routine maintenance repair and replacement" does not trigger NSR. EPA has never defined these terms, either in rulemaking or guidance. However, recently, EPA has begun to claim in enforcement actions and informal conversations that this exclusion never applies to changes that increase the efficiency of a unit, improve its reliability, or reduce its costs. Under that approach, repairing or maintaining a 1990 unit with year 2000 components that improve its performance could trigger NSR. Such an approach is both economically and environmentally counterproductive. It destroys the "total quality improvement" programs that businesses must adopt in today's competitive markets -- and that the Administration has endorsed because of their environmental benefits.
- EPA has always recognized that NSR is triggered when a single "physical change or change in the method of operation" causes an emissions increase. EPA has also always cautioned that if a source artificially splits a single project into two projects in order to avoid NSR, it will still treat that project as one. We agree. But EPA now claims that all changes at a plant should be aggregated together whenever they serve the "basic purpose" of the facility. Since changes that did not serve that "basic purpose" would not be made, this is a formula for aggregating all changes that a plant makes into one change. Once those changes have been aggregated, the "actual-to-potential" test makes it virtually certain NSR requirements will be triggered.
V. THE CONSEQUENCES OF EPA'S CURRENT APPROACH TO NSR
EPA's current approach to NSR applicability results in significant compliance uncertainty, overburdens state and refinery resources, and hinders future environmental progress.
1. Compliance Uncertainty
Under EPA's current approach, it has become nearly impossible for any refinery to determine which of its activities might trigger NSR and which will not; EPA's requirements are extremely unclear and a constantly moving target.
Refiners cannot rely on the current written guidance to determine when NSR is required because the existing guidance is unclear and often contradictory. It consists of over 4,000 pages of guidance documents, many of which are in draft form and contradict each other, and various EPA memoranda. Many of EPA's new positions on NSR applicability contradict the older guidance, and are not even in writing. Refiners often do not know EPA's latest position until it is incorporated into an enforcement action or initiative.
Refiners should be able to rely on state permitting authority decisions to determine when NSR is applicable, but it now appears they cannot. In most states, EPA has delegated the implementation of the NSR program to state permitting authorities. The state permitting authorities make permitting decisions for refineries, and regularly inspect refineries to ensure that state decisions are properly implemented. EPA also reviews and approves the states' programs, and periodically inspects the refineries themselves. However, compliance with state decisions does not necessarily accord with EPA's latest positions. In fact, EPA is currently conducting a widespread investigation of refineries regarding NSR permitting compliance as far back as 1980. In effect, EPA has called into question state NSR permitting decisions over the last 20 years. These decisions were not questioned during 20 years of state and EPA inspections.
2. Overburdening State Resources
Moreover, under EPA's current approach, hundreds of projects a year at a refinery might trigger NSR. No State has the resources to answer thousands of NSR permitting questions annually from all its major stationary sources, or to review its NSR permitting decisions over the past 20 years. Certainly, states have much better and more environmentally productive ways to invest their resources (e.g., expediting permitting for gasoline sulfur reduction requirements as EPA has agreed to do). State permitting may also be slowed down because states will proceed more cautiously for fear that they may be second-guessed by EPA. This may create permitting bottlenecks at the very time states need to proceed expeditiously to implement important upcoming regulations.
3. Overburdening Refineries
The end point of EPA's current position is universal NSR. However, no industrial economy could function if every change to a factory required a permit before construction could begin. This will be particularly burdensome for refineries given the operational changes necessary to comply with the blizzard of new fuel reformulation and stationary source regulations. EPA recognized that Congress did not intend universal NSR in its 1996 proposal for NSR reform, however EPA's new approach is achieving just that:
". . . section 111(a)(4) of the Act could -- read literally -- encompass the most mundane activities at an industrial facility (even the repair or replacement of a single leaky pipe, or an insignificant change in the way that pipe is utilized). However, the EPA has recognized that Congress did not intend to make every activity at a source subject to major new source requirements under parts C and D. As a result, the EPA has adopted several exclusions from the 'physical or operational change' component of the definition. For instance, the EPA has specifically recognized that routine maintenance, repair and replacement. . . [is not by itself] considered a physical or operational change in the method of operation within the definition of major modification." 61 Fed. Reg. 38250, 38253 (July 23, 1996).
4. Hindering Future Environmental Progress
As discussed in section III, EPA's current approach to NSR threatens the Agency's future environmental agenda by posing significant logistical challenges for implementing important upcoming regulations.
Additionally, the unnecessary costs of EPA's current approach to NSR will compete with resources needed to implement these regulations. Our resources are limited and the costs of these upcoming regulatory initiatives are high. Just one of these regulations, the Tier II/gasoline sulfur reduction requirements, is expected to nearly double the refining industry's environmental expenditures to approximately $8 billion annually. Expected requirements to reformulate diesel fuel could increase these costs by half again.
We simply do not see the logic for applying EPA regulatory reinterpretation to activities that do not increase emissions, or actually reduce emissions.
VI. REFINING INDUSTRY GOALS FOR NSR REFORM
The refining industry is encouraged by our current round of discussions with EPA on NSR reform and hope that this joint effort will continue and produce real reform. The discussions so far have been candid and useful, but we still do not know clearly what to expect from them.
We believe that any real reform must address both substantive and procedural issues. Real reform should ensure that NSR applies only if emissions actually increase significantly. The current system of perpetual exposure to NSR cannot be defended.
Real reform must alter or abandon the "actual-to-potential" test so that changes that do not increase emissions do not automatically trigger NSR. Real reform must also change EPA's current approaches to "routine maintenance, repair and replacement" and "aggregation", which work together with the "actual-to-potential" test to create exposure to NSR for virtually any change a plant makes. Perpetual NSR is unworkable, contrary to Congressional intent, and bad environmental policy.
Finally, real reform will address the need to expedite rather than hinder efforts to comply with federally-mandated environmental programs.
These changes should be subject to full public review and comment.
VII. EPA'S ENFORCEMENT INITIATIVE WILL UNDERMINE NSR REFORM.
I would like to conclude with a word about enforcement.
Over the past two years, EPA has been conducting a massive investigation of the refining industry, and several other industries, for purported "widespread" noncompliance of the NSR
program. Violations of NSR do occur, and the government should pursue them whenever they do. However, the refining industry believes EPA's allegations of widespread noncompliance are based on new and controversial reinterpretations of the NSR requirements that amount to rulemaking without notice and comment. By making fundamental changes to the NSR program through enforcement actions, EPA threatens to undermine the NSR reform process and an clarification of the program that reform can provide.
The reinterpretations that EPA wants to retroactively enforce would allow EPA to claim that virtually any source is subject to NSR. As previously discussed, this approach would be impossible to comply with, overburden state and industry resources, and undermine the implementation of future environmental regulations.
By questioning state permitting decisions and policy over the past 20 years, EPA will only further slow down the permitting process and divert state resources towards reviewing past decisions. This is inappropriate at a time when it is critical that state permitting authorities and refiners work together to expedite the permitting processes for important upcoming environmental regulations, such as the Tier II/ gasoline sulfur reductions requirements.
The decision criteria for many NSR issues are so opaque, and have changed so many times that, in our view, it is neither fair, nor just, nor sound public policy to make them the excuse for an aggressive enforcement program. The opportunity for public comment and congressional review of EPA's proposed reinterpretation of NSR is necessary to respect the due process rights of those who have to comply. If EPA wants to revise the NSR program, it should do so through the reform process.