SUMMARY

EPA's NSR ("New Source Review") rules, which for thirty years have been consistently applied only to new greenfield sources or major modifications of existing sources, are now being reinterpreted without any rulemaking change and applied to routine repair, replacement and maintenance activities at all existing sources, causing major disruption in routine maintenance schedules, curtailing power output, and dismembering whole Titles of the Clean Air Act. The rationale for the radical shift in interpretation is in the allegation that utilities are by illicit maintenance keeping afloat old plants that were "grandfathered" from any CAA controls and that are now threatening the nation's health. But the 1990 CAA Amendments mandated sweeping reductions for all power plants regardless of age through the use of highly efficient market incentives. The 1990 Act thus established a flexible market-based system that is working very efficiently to drive down pollution through 2010 and beyond, but that is now being repealed by administrative fiat and replaced by an outmoded, inefficient and counterproductive command and control regime.

I. How did we get here?

· The CAA, which has produced dramatic reductions in air pollution over the last three decades despite explosive economic growth, operates through two approaches. The first approach develops national health and environmental standards for the states to apply to the existing sources in their jurisdictions. DOE reports that the utility industry alone has spent more than $30 billion to achieve compliance with these health standards.

· The second approach applies the best current technology to new sources and major modifications of old sources that increase pollution levels where inclusion of such technology can be integrated in an efficient manner without highly disruptive retrofitting. The purpose is to prevent new pollution by new plants, both to preserve air quality in areas that attain health standards, and to avoid complicating ongoing plans to clean up existing plant and equipment in areas that do not.

· Because of delays and regulatory difficulties primarily associated with ozone attainment and a need to address acid rain not previously regulated, the Congress enacted the 1990 CAA Amendments ("1990 CAAA") to impose a sweeping array of new pollution reductions on power plants (and other pollution sources as well). These new programs included the acid rain program of Title IV, which mandates a 50% reduction in SO2 by 2010, and the interstate transport provisions of Title I, which are now being implemented to impose additional NOx controls in Midwestern power plants that may themselves be located in attainment areas, but that send pollution through tall smoke stacks to the neighboring states.

· These new programs adopt a different -- and highly successful -- approach that assigns and limits the absolute number of tons a plant can emit, leaving to the plant the decision as to how to reduce its tons, rather than assign a particular technology to the plant which it must build. Because the preexisting NSR program is technology-based, rather than ton-based, EPA issued a rulemaking in 1992 to reconcile the old with the new, as described more fully below. It is this 1990 CAAA and 1992 rulemaking which EPA is now blatantly violating -- by, for example, forcing utilities to accelerate reductions much faster than those mandated by Title IV of the 1990 CAAA.

· As indicated above, NSR was intended primarily to apply to new sources and can also apply to existing plants only when a large industrial source of air emissions, a refinery or a power plant makes a non-routine physical or operational change that results in or causes an emissions increase.

· Over the last thirty years, EPA's regulations and practice have excluded from NSR all "routine maintenance, repair and replacement" activities undertaken by power plants and other industries. Additionally, EPA surveyed utility maintenance projects, including "life extension projects," in the early 1990s and concluded that those did not trigger NSR. EPA also has published guidance in the Federal Register defining what was routine by reference to the standard practices of the relevant source category, in this case the utility industry. Likewise, EPA's regulations specifically exclude any increases in emissions associated with operating a facility more hours, unless such an increase is prohibited by a federally enforceable permit condition.

· EPA's practices interpreting the NSR rule were explicitly described to Congress by then-EPA Administrator Reilly and other Agency officials when Congress was considering the Clean Air Act Amendments of 1990. One of the reasons Congress adopted the Acid Rain provisions of Title IV to reduce SO2 by 50% (10 million tons) was because utility units typically operate for 65 years or longer without major modification and the NSR program would not obtain equivalent reductions. To help facilitate cost-effective compliance by the utility industry with both the ton-based 1990 CAAA and the pre-existing technology-based NSR program, EPA, after an extensive notice and comment process in 1992, promulgated a rule which explicitly laid out all of the NSR procedures applicable to the utility industry and confirmed that "pollution control" projects would not trigger NSR.

· In 1996, EPA initiated a rulemaking to revise the 1992 NSR rule, but never finished it. Instead, in 1999, EPA commenced a major enforcement initiative against virtually every coal-fired utility plant in the country for repair and replacement activities undertaken over the past 20 years. Under EPA's reinterpretation, virtually every maintenance, repair or replacement project undertaken by any utility plant could be considered non-routine. Any project that increases availability or efficiency or corrects problems causing forced shutdown of plants potentially triggers NSR. EPA abandoned its simple test for determining when maintenance practices are routine -- common industry practices -- and now applies a multi-factor (more than 20 different factors) weighing and balance test that only it can perform with any sort of regulatory certainty. Amazingly, even installation of pollution control equipment by utilities may now be viewed as an NSR-triggering event.

· Whatever policy merits EPA believes justify its new position on NSR applicability, EPA's efforts to achieve this through enforcement actions against utilities for projects undertaken decades ago is inconsistent with current law. If EPA believes this NSR reinterpretation is correct, it should only apply it after notice and comment rulemaking or ask Congress for new legislation to revise the 1990 CAAA.

· In justifying its enforcement actions, EPA claims that its sole goal is to avoid emission increases by power plants operating more hours than in the past. This point is so important that a more detailed explanation is in order. Under the Clean Air Act provisions, every power plant in the country is allowed to emit a certain quantity of various regulated pollutants, of which NOx and SOx are the two key ones. Each utility plant has a legally mandated emission rate -- a maximum amount of pollution that can be emitted per hour, per day, per month, or even annually, depending upon air quality and other consideration. But, any time a plant slows down because of a maintenance problem, it will necessarily be able, once repaired, to operate more hours -- and emit more -- than it did during the problem period -- even the emissions are well within the limits spelled out in the State SIP and the federal reductions required by Title IV. These various limits are spelled out in permits held by utility plants or in state implementation plans, and they reflect EPA-prescribed public health-driven ambient standards. These limits cannot be breached by power plants under any circumstances, and there is no claim that any of the plants subject to the EPA enforcement did exceed the permitted limit of emissions. However, every unit must be prepared to operate more hours within their tonnage limits in order to meet customer demand.

· EPA's definition of an emission increase is artificial and arbitrary. Power plants operate under extremely harsh conditions; every several years, as the plant equipment deteriorates, the plant's efficiency, availability and reliability go down. Eventually, the plant operator performs a set of routine maintenance procedures to restore and maintain the plant's efficiency, availability and reliability. To emphasize, throughout all of these changes, the plant never increases or exceeds its legally binding and public health-driven emission limits. EPA, however, compares a plant's actual emissions at the time it was operating in the recent past before a maintenance procedure with its future potential emissions following that procedure, assuming that the plant will, as a result of the project, operate every hour of every day in the year at maximum output. In other words, EPA's methods always predicts an emission increase even though none may occur, and even though the plant may not under any circumstances exceed the CAAA's mandated reductions.

II. EPA's Reinterpretation Discourages Needed Maintenance Procedures
and Reduces Generating Capacity

· EPA's reinterpretation is not only flawed as a matter of law, but it also undermines our energy supply. Because NSR is a costly and time-consuming process, EPA's current position discourages utilities from undertaking needed maintenance projects. This makes plants more reliant on deteriorating components, resulting in less efficient, less reliable and higher emitting power generation. For example, the efficiency of currently available steam boiler equipment deceases over time as plant components deteriorate. Boiler tubes, in particular, are subject to very harsh temperature, pressure, and chemical conditions, and leaks result. Short-term fixes include patching tubes where there are leaks, but eventually whole sections begin to wear out and must be replaced if the plant is to continue to operate. Yet EPA's reinterpretation of NSR could have such a routine and necessary activity declared non-routine.

· A plant operator typically will accept some level of deterioration in efficiency for a short period of time but must eventually undertake the repair and maintenance necessary to regain lost efficiency and to maintain unit availability. The timing of these projects depends in part on the demands being placed on the power plant to operate to meet energy supply needs. Unit unavailability can seriously impair a utility's ability to meet customer demand and nearly always results in running less efficient units. Operating inefficient units increase the amount of pollution emitted. Under the EPA Office of Enforcement and Compliance Assurance's new interpretation of the NSR rules, it is these projects, designed to maintain efficiency and availability, that are no longer regarded as "routine." EPA then assumes the unit will operate more hours than before the project and further assumes that the project, rather than customer demand, weather, or other unit outages, causes this increase. Once EPA thus determines that NSR will be triggered, the unit cannot even begin to proceed with the project without either going through the lengthy NSR permitting process, which takes a year or more, or without "capping" operations at historical levels. Thus, the unit must either wait or derate. Either alternative can have significant adverse consequences for the reliability of the country's electric supply. Waiting can idle a unit during peak demand for 12-24 months, more if intervenors challenge the permitting. Derating effectively confiscates capacity, even when the unit is permitted to operate at maximum output year-round.

· Over the next 3-5 years, thousands of megawatts of existing generating capacity will be lost if companies are not able to undertake these routine maintenance and repair projects, or if companies must accept caps on utilization to avoid lengthy NSR. In the longer term, EPA's new position would involve the loss of an even greater number of megawatts. The result of EPA's reinterpretation will be the decrease in available installed power plant capacity at a time when we already have a supply shortage -- something this nation, and the West in particular, can ill afford.

III. EPA's Reinterpretation Discourages Efficiency Improvements

· There are 300,000 megawatts of coal-fired generating capacity which is 55% of all electricity generated in the United States. Approximately 1,200 coal-fired generating units are in service. These generating units involve two distinct sets of operations: (1) a steam cycle (e.g., the boiler and related equipment), and (2) the turbine cycle (where the electricity is generated). In the past few years, there have been some very exciting innovations in the turbine technology area. For example, just one type of efficiency improvement project, the so-called Dense-Pack which enhances the efficiency of turbine blades, can result in a very significant improvement in the efficiency with which steam is turned into electricity.

· A more efficient turbine results in more electricity output from the same steam input, with no greater fuel use. For example if one assumes that most generating units could improve efficiency by between 2% and 4% (a very conservative estimate, based upon the actual operating experience of several units which have installed the Dense-Pack technology), this would mean an additional output of 6,000-12,000 megawatts of power in the near term, with significant decreases in emissions per unit of fuel burned. This increase in available installed capacity is the equivalent of building 20-40 new plants of 300 megawatts each with no new emissions.

· As an example, this type of efficiency improvement, if installed by the approximately 1,000 utility units (out of some 1,200 existing coal-fired utility plants) that can be most easily retrofitted with Dense-Pack technology, would reduce criteria pollutants that NSR was meant to address (NOx and SOx) substantially.

· However, under EPA's reinterpretation of its NSR rules, the installation of even this type of beneficial technology requires an elaborate, expensive and time-consuming permitting process, which results in the imposition of additional costly control technology requirements on existing plants, and therefore discourages the installation of new and more efficient technologies.

IV. Conclusion

Overall, the effect of EPA's recent position is to block routine maintenance, repair and efficiency improvement projects that could immediately expand generating capability without increasing fuel burning and will decrease by a significant percentage the total available installed capacity through caps on operations. Stated differently, EPA's reinterpretation of NSR is tantamount to shutting down dozens of utility units every year at a time when electricity supply is already so short as to be unreliable in many areas.