February 26 , 2003

 

U.S. Environmental Protection Agency

EPA West (Air Docket)

1200 Pennsylvania Ave., NW

Room B108, Mail code: 6102T

Washington, DC 20460

Attn:  Docket ID # A-2002-04

OAR-2002-068

 

 

Ladies and Gentlemen:

 

My name is Victor B. Flatt, and I am the A.L. O’Quinn Chair in Environmental Law at the University of Houston.  I teach and do research in the area of environmental law, and in particular the area of environmental administration.  Due to my area of expertise, and my duties as a public servant, I wanted to submit a comment on the proposed EPA rule defining Routine Repair and Maintenance (RRAM) for purposes of New Source Review (NSR) of existing air pollution sources which undergo alterations.  As a law professor who has studied these issues over time, I believe that some of my comments might be germane to the viability and legality of this proposed rule.

 

Under Secs. 165(a), 172, and 173 of the Clean Air Act, annotated at 42 U.S.C.  Secs. 7475(a), 7502, and 7503, major new sources  or major modifications of existing sources of criteria air pollutants must submit to review as a new source and are required to install pollution control equipment which represent the Best Available Control Technology (BACT) for sources in attainment areas, and Lowest Achievable Emissions Rate in non-attainment areas. 42 U.S.C. Secs. 7475(a)(4) and 7503(a)(2).   A “new source” is defined in the Clean Air Act to include already existing sources which are modified after the effective date of these acts.  42 U.S.C. Sec. 7411(a)(2).  In pertinent part, a “modification” occurs if “any physical change” or “change in method of operation . . . increases the amount of any air pollutant emitted by such source.”  In general, a major source is defined as capable of emitting 100 tons of a regulated pollutant annually.  42 U.S.C. Sec. 7602(j).  With respect to the non-attainment program, “major” is defined as those sources which can emit or have the potential to emit either 100 tons or 250 tons (depending upon the type of source) of certain pollutants.  42 U.S.C. Sec. 7479(1).

 

On December 31, 2002, the EPA announced its final rule defining “major modification” for purposes of NSR, to be set out in 40 CFR, parts 50 and 51.  This rule attempts to define “major modification” as a result of “significant” emissions increase of a regulated NSR pollutant.  This rule specifically exempted routine repair and maintenance (RRAM) from activities which would trigger NSR for existing sources.  RRAM is the subject of this rulemaking, and is what these comments address.[1]

 

The definition of modification in 42 U.S.C. Sec. 7411 in many ways is quite explicit and clear.  However, there are some questions as to what constitutes a source for application or what activities cause emissions increases.   Historically, the EPA has not defined RRAM for purposes of triggering NSR for alterations of existing sources, preferring to work on a case by case basis.  In making its determination, the EPA has considered the nature, extent, purpose, frequency, and the cost of the work.  Those regulated entities that are unsure whether or not NSR and its accompanying technological requirements are triggered, may consult the reviewing authority for assistance in making the determination.

 

This proposal seeks to create classes of categories, defined by total expenditures as a percentage of capital cost, which would uniformly be considered RRAM.

 

My comments address two major considerations:  1) the need for this “clarification,” and 2) substantive and procedural problems with the proposal.

 

I.                    The need for a clarification of RRAM.

 

According to the call for comments, the EPA has undertaken this rulemaking because industry believes that the current approach is too uncertain, proving costly since industry may not undertake needed repair and maintenance because of uncertainty about what changes might be allowed without triggering NSR requirements.  According to the call for comments, the option of requesting an applicability determination is considered too costly by some in industry.  Such applicability determinations also require time and resources from the EPA as well.

 

If it is possible to avoid sacrificing legitimate NSR while introducing more specific guidelines to define RRAM, then it would make sense to do so and would not be inconsistent with the legislation that requires NSR for certain modifications of existing sources.  The problem is that the large diversity of industries and situations makes it almost impossible to avoid a case by case approach without making the definition of RRAM both over and under- inclusive.  Indeed, the current proposal does not truly eliminate the uncertainty of the procedure.  The current proposal posits that an owner or operator “would evaluate on a case-by-case test any activities that did not come within the allowance and that are not otherwise excluded, in order to determine whether they are RMRR.”  Furthermore, the proposal might exempt certain kinds of replacement or maintenance activity or other activity that is not legitimately considered a RRAM, even if it comes within the yearly budget allowance.

 

The truth is that there is no real way to categorically define all of the activities that could or could not be considered routine repair and maintenance. If that were possible, then Congress would have made that determination itself.  Similarly, it could have chosen some percentage expenditure, but it did not.  Instead, where Congress has explicitly spoken on the issue, it has stated that the term “modification” means an increase in the amount of “any air pollutant,” 42 U.S.C. Sec. 7411(a)(4), and that with respect to certain non-attainment areas, there shall be no de minimis exception for increases of volatile organic compounds from existing sources unless it is less than 25 tons over five years.  42 U.S.C. 7511a(c)(6). (emphasis added).

 

Essentially then, the proposed rule would simply exclude certain activities that fall under a specific budget, while still leaving uncertainty as to those activities beyond that budget or activities in which there is any question whether the activity is routine repair and maintenance.  However, this does nothing to eliminate the current uncertainty which the purported rule addresses.  This is because the “exception” to the annualized budget is no different than the current situation.  Under the current regulations, an owner or operator would only need to seek applicability review if the activity is not clearly routine repair and maintenance.  The definition of RRAM is not connected to the expense of the project, but rather the nature of the project, and presumably whether it would operate as an increase to emissions.  The new proposal does nothing to eliminate that.  If there is uncertainty whether an activity which falls within the annual budget is not really RRAM but instead an upgrade or major modification forbidden by the statute, then applicability review is still required.  The only change is that there will now be created a situation in which there is a safe harbor of activities (those under a certain budget) which will not trigger routine scrutiny as opposed to the current situation in which all activities might trigger such scrutiny.  Assuming that owners and operators follow the letter of the regulation then virtually all activities that are currently in question will still be in question under the proposed new rules.  Therefore the effect of the new rules is simply to allow larger investments in upgrades without scrutiny, subverting the purpose of the statute.  This brings us to part II.

 

II.                 Substantive and Procedural Problems with the RRAM

 

The proposed rule also suffers from substantive and procedural problems in that its implementation may be inconsistent with the Clean Air Act and the Administrative Procedures Act.

 

Secs. 165 and 173 of the C.A.A. specifically note that any change in pollution levels from an existing source triggers NSR and accompanying technological upgrades.  Nothing in the proposed definition based on cost of changes or maintenance address this clear language of Congress.

 

Moreover, this proposed definition of routine repair and maintenance would essentially exempt any modifications of the need to go through NSR.  That is because it allows for upgrades that will indefinitely keep existing sources in service without ever having to upgrade to new source requirements. 

 

This problem is recognized in the proposal, which notes that replacement of “existing process units” (which presumably would be an “upgrade” as opposed to a “routine maintenance”) would not qualify for the RRAM exceptions.  However, it does allow that “some partial replacements [of existing process units] would qualify for the proposed RRAM exclusion.  But partial replacements over time can easily have the same effect as an entire replacement of an “existing process unit” which the proposal notes would not be RRAM.  Under its own terms then, the proposed rule allows changes over the course of years to occur which it supposedly would not allow to occur in one year.  Either an activity is routine repair and maintenance or it is not; it should be immaterial if the activity occurs in the course of one year or five years.  It is true that many “partial replacements” may not be disguised upgrades but RRAM, but the only way to determine if piece by piece alterations are not cumulative upgrades is to have case by case analysis, which this proposal seeks to eliminate as routine procedure.

 

The proposed rule is also in direct violation of the enabling statute in noting that changes that would increase “efficiency” of the unit do not automatically qualify as a major modification.  This flies directly in the face of the words and spirit of the Clean Air Act, and cannot withstand analysis.

 

As the proposal itself points out, RRAM would usually be expected to encompass simple replacement of parts with identical ones.   This proposal seeks to expand this by allowing replacements of similar units or equipment, even if the use of an “improved” version increases the life or preserves the efficiency of the affected unit. (Section VII. A.)

 

The purpose of allowing existing sources to avoid the imposition of the pollution control equipment for new sources was to recognize the already fixed costs that had been incurred prior to the passage of the Clean Air Act.  However, it is clear from all contemporaneous debates that this was not meant to be a permanent situation.  It was believed that eventually as the older plants reached the end of their useful life, they would be replaced by new plants or significantly upgraded so that new state of the art, pollution control equipment would be installed.  If they were not eventually phased out or upgraded, there would not only still be dirty air, but it would also make new, pollution controlling plants, comparably inefficient.  Under the Clean Air Act, then, even allowing for one to one replacement of parts might be problematic.  The only justification for allowing even this is that eventually new plants, parts, and machines would be so much more efficient, that simply substituting one old fashioned part for another in an existing plant, would eventually cease to be economically logical, even when including the costs of significant pollution control upgrades.

 

The current proposal effectively eliminates the possibility of an existing plant ever having an economic incentive to upgrade by allowing these plants to install replacement equipment that is more efficient.  This means that existing plants can continue to modernize, bringing in comparable efficiencies to new plants while at the same time not being required to upgrade pollution control equipment.  As noted in the proposal itself (Section VII. B.),:

 

 “almost any component replacement can be expected to have some beneficial impact on the energy efficiency of the unit and, left unbounded, this approach could result in the replacement of an entire boiler with a new, more efficient boiler without state of the art pollution controls.”

 

This proposal admits that the “replacement” of an entire boiler should not be considered “routine,” and thus should trigger NSR, but the proposal notes that by allowing replacement of parts with more efficient and newer parts, that in fact this could occur.  Thus the proposal itself recognizes that it might be allowing “major modifications” under the proposed rule without triggering NSR, in violation of the enabling statute.  Any elimination of pollution control equipment upgrades for existing sources must be undertaken through the legislative process, and not implemented in the guise of an administrative change.

 

The problem is exacerbated by the lack of specificity in the proposal.  The proposal notes that RRAM allowance should be defined as a certain percentage of capital cost per year, but it fails to set out what that percentage will be or how “overall capital cost” will be determined.  In this case, the devil is truly in the details.  Depending upon how “capital cost” is determined and depending on what percentage is allowed, the number of activities that will qualify for RRAM could vary enormously, even to the point where almost any activity, not excluded as “major modification” by the owner itself will qualify under this “exception.”  In addition to the increasing scope with which this would allow more activities that should be considered “major modification” to avoid NSR, it also is procedurally problematic.

 

In order for this proposal to be implemented in final form, it first must be circulated as a draft in substantially the same form as it will appear as final.  Without this, there can be no effective comment as required by the APA, 5 U.S.C. Sec. 553, nor effective due process to those who will be affected by the rule.  Without sufficient detail, any movement to final rule implementation without more specificity regarding the calculation of “capital costs,” determination of percentage allowance, and determination of time event horizon for calculation would be an APA violation.

 

III.       Conclusion

 

This lack of specificity frames exactly what is wrong with the proposal.  The proposal is being put forward to supposedly solve the problem of uncertainty with respect to what will trigger NSR for repair and maintenance activities.  However, its effect cannot be evaluated without more detail as to how any annual allowance would be calculated.  Moreover, the proposal has not indicated how it would actually eliminate uncertainty since it still retains an exception for activities which are “not to be considered RRAM.”  Since the proposal does not effectively demonstrate how it will be an efficiency improvement over the current case by case analysis, but does indicate the likely possibility of currently regulated activities escaping NSR, its implementation must be considered arbitrary and capricious and an abuse of discretion, and thus a violation of the APA.  The only substantive proposal, that which anticipates allowing efficiency upgrades to occur with RRAM, violates the spirit and letter of the CAA.  Thus, this proposal should not be implemented.  The EPA should continue with the current case by case approach which is the only solution to analyzing the diversity of air pollutant sources which may trigger NSR.

 

 

Very truly yours,

 

Victor B. Flatt

A.L. O’Quinn Chair in Environmental Law

University of Houston Law Center

713-743-2155; vflatt@central.uh.edu

 


[1]               The final rule defining major modification has already been challenged in court, but the overturning of that rule will not affect the process governing this rule.  Thus my comments will address this rule as a self contained proposal.