Reform of New Source Review – The Road Ahead
Posted on April 15, 2003
Posted By: Tobey Winters
New Source Review (“NSR”) was established as one of the basic administrative mechanisms to achieve the goals of the Clean Air Act of 1990. On the last day of 2002, the EPA issued both final and proposed regulations to reform NSR.
Aims and Objectives of NSR
NSR applies in two separate circumstances: 1) the requirement to prevent significant deterioration (“PSD”) of air quality in areas that meet air quality standards and 2) to help improve air quality in areas that are in “noncompliance” with the standards (“non-attainment”) – where the air is unhealthy. In general, the States (or localities designated by the States) administer the program, and have incorporated NSR regulations into their respective State Implementation Plans that specify emission limits.
The rationale for the NSR program is based in common sense. Air pollution controls (“APC”) are implemented most economically with new construction projects. However, the program also applies to major modifications to existing plants. NSR regulations are applied to the following pollutants: sulfur dioxides, carbon monoxide, nitrogen oxides, volatile organic compounds, and particulates. Depending on the level of the increased emissions, a new source, or major modification to an existing source, has to meet best available control technology (“BACT”) guidelines for each of these pollutants. In the case of NOx in a non-attainment area there is also the requirement to obtain a net reduction in overall emission levels for every ton of pollution emitted by the new source.
Essential Features of NSR
NSR is specific to large stationary emission sources. NSR is applied to long-lived capital assets like power plants, refineries, chemical plants, and pulp mills. Capital investment decisions in these facilities are made infrequently, but once made, often determine the remaining useful life of the asset. The pace of implementation has frustrated the EPA. The EPA has had significant success in reducing pollution in mobile sources where a new pollution control technology can be introduced for a class of vehicles and in less than 10 years achieve penetration in 90 percent of the vehicle fleet. By contrast, electric utilities have coal generation units approaching 50 years of operation that show few signs of economic obsolescence. This contrast explains in part the essential conflict between the desire for the EPA to accelerate the introduction of air pollution technologies through NSR on the one hand, and the plant operators’ interest in extending the life of existing generating units on the other.
The application of BACT to new construction has been key to the EPA’s enforcement. When applying the air pollution emission standard to a new or modified source, the permitting agency requires the use of BACT in PSD areas and “Lowest Achievable Emissions Rate” (LAER) in non-attainment areas. The essential differences in the two criteria are that in PSD areas the cost-effectiveness of the technology is considered in making decisions about BACT. For LAER determination in non-attainment areas, BACT need only demonstrate that the technology has been proven to work reliably in the application and process at a comparable scale. The cost effectiveness test is not applied in non-attainment areas.
BACT becomes “technology forcing” when air pollution control (APC) suppliers compete with new processes or on new applications. A customer anxious to get his project approved and willing to try a new technology accelerates this competition. The permitting agency facilitates the plant owner’s choice of a new technology by allowing flexibility in enforcement of the emission limitation or by giving the owner a fall back position in case performance is not met on the new less proven APC equipment.
Not only is BACT technology forcing, it is also a moving target. One example is illustrated by gas turbines. In the early 1980s BACT on a gas turbine for NOx was 42 parts per million (ppm) and steam or water injection was used to control NOx. The turbine manufacturers subsequently developed new low NOx combustors and reduced the levels to 25 ppm, and then 15 ppm and eventually 9 ppm in the early 1990s on large gas turbines. With back-end selective catalytic reduction (SCR) the NOx levels were reduced even further to 3-4 ppm. Because SCRs use ammonia to remove NOx, there is some excess ammonia that is released to the environment. The level of ammonia “slip” has been reduced from 5-10 ppm to 2 ppm. In addition, a new technology, SCONOx is able to obtain 2 ppm NOx limit with no ammonia use. By 2003 the South Coast Air Quality District in California determined that BACT on large gas turbines is now 2 ppm1 .
SCRs have also become the technology of choice for coal-fired plants. Catalyst contamination has and will continue to be an issue with oil and coal, but experience in Germany and Japan has been proven on US coals. NOx reduction levels continue to improve as the technology has proven itself in the late 1980s and early 1990s – first on the early new IPP units and then the transition to retrofit on existing coal units.
The environmental community applauds the technology forcing aspects of the NSR program. The essential features of BACT will not change with the EPA reforms.
Owners of existing plants are concerned about making the “major modification” cost effective including some certainty that the benefits of the modification are far more than the cost of the required APC investment. Arguably, the case-by-case approach applied to NSR on existing units over the years has done more to discourage than encourage APC investment in existing plants. The key issue for the EPA is to create an NSR program that encourages industry to make the APC investment.
The EPA has recognized for some time that the NSR program needed simplification, particularly since 28 different processes are being regulated under the program. The EPA instituted stakeholder meetings in 1993 and proposed major rule revisions in 1996 and identified additional revisions in 1998. Many of the issues needing simplification related to definitions and baselines for comparison. Many of the proposed changes revolved around definitions of “major” vs. “minor” modification, the definition of “significant” change, and the baseline measurement setting of emissions before and after “major” modification. These rules of the road clarifications would be needed in any simplification process.
A key issue from the economic perspective is the “netting out” provision in NSR, that is, the ability to expand or upgrade one process unit at a plant while making significant reductions in emissions within the process or at another process unit within the same plant. The netting out provision is the key to flexibility in managing capital investment decisions. From the regulator’s point of view, netting out adds complexity and state and local administrators are on record to eliminate the provision.2 From a broader perspective, however, the objective is to get overall emission reductions and flexibility encourages new investment. The following provisions in the EPA’s NSR modifications are key to flexibility and the broader goal of emission caps.
Plant-wide Applicability Limitations – or PALs
The intent of the EPA is to allow a generating station (or industrial facility) consisting of one or more generating units the ability to manage emissions within an overall plant-wide emissions cap rather than have to seek a NSR permit every time a modification is undertaken. The cap would be based on actual emissions from all emissions units at the station and not based on the “potential to emit” based on the capability of a specific process unit. 3
Clean Unit Status
The test of a clean unit is whether a unit has gone through a major NSR process and is complying with BACT or LAER. If a unit has not gone through the NSR process, it still may qualify as a “clean unit” if it passes a two part test which consists of meeting BACT or LAER and demonstration that the unit does not contribute to a PSD or other violation, such as visibility. The advantage of clean unit status is that it allows operational modifications to the plant for up to 10 years without additional review. A unit can lose its clean unit status, if changes are undertaken that would violate its emissions limitations set out in the original BACT or LAER permit, but otherwise operational flexibility is allowed for a period of ten years.
The criticism leveled at the PALs and Clean Unit provisions by the environmental community is that they do not require the reduction of emissions over time. This position ignores the linkage to the other programs and requirements of the State Implementation Plans. The critics’ view is that a “major modification” is a trigger mechanism to invest in BACT and that the designations of PALs and Clean Units might potentially allow major sources to avoid making the BACT investment.
Future reductions in emissions are forced by State Implementation Plans in a top down allocation. As utilities and areas are allocated smaller and smaller emission budgets in the future, by necessity, more APC investments will be required. Coupled with emissions trading, the plants that can most cost effectively invest in air pollution controls will make the largest and earliest APC investments. As the emissions budgets decrease, there is no “escape clause” in the NSR rules from the emission reduction requirements.
Routine Maintenance, Repair and Replacement
With the PALs and the Clean Unit provisions, managers will have the ability to get the most out of existing units as long as emissions do not increase. At the same time there will be increased ability to forecast the most cost effective APC investment decisions. These opportunities could be taken away with the EPA’s proposed changes to “Routine Maintenance, Repair and Replacement”. The EPA is proposing a stringent cost test on routine replacement. Unless a unit is a clean unit, plant operators under the proposed rules will be subject to an annual or multi-year budget (five or less) allowance that reflects how much can be spent on a unit without triggering NSR.
The purpose of the new replacement rules is to provide a bright line between an illegal re-building of an existing unit step by step (without adding BACT equipment) and the permitted practice of maintaining a unit in a safe and reliable condition. Although sensible in theory, the application of the proposed rule is anything but clear. The term “life extension” was banned from the utility manager’s lexicon precisely because what the plant manager sees as “routine replacements” (to reduce pre-mature or forced outage) became in the late 1990s precisely what EPA sees as replacements that delayed and avoided NSR or eventual unit retirement. Much like keeping an old car on the road, the dividing line between essential maintenance and life extension is never a bright line.
The preferred scenario is to make every base-load unit a “clean unit.” This path would allow utilities to make the BACT investment as emission budgets required and then have the flexibility to manage the generating asset flexibly for a period of time at the plant-wide level. The EPA regulations allowing for PALs and Clean Units appear to make this possible, but the proposed routine maintenance provisions retreat to command and control philosophy.
EPA is officially open to other criteria than cost for deciding about what constitutes routine repair and maintenance. For the utility industry rated capacity is a far better guide than cost, if the criterion was clarity on what constitutes acceptable vs. unacceptable kinds of component replacement expenditures. Comments to EPA are due by early May 2003.
Despite the half step forward represented by the NSR modifications, the future of NSR appears still to be in the hands of the courts. The EPA has not backed down from the lawsuits filed during the Clinton Administration. There are currently nine cases in federal court affecting 44 generating units and the nation’s largest coal generation based utilities. The case that is the furthest along (TVA v. EPA) heard oral arguments in May of 2002. Mediation was ordered and mediation failed in August of 2002. Some observers see the TVA case as the bell weather decision for future APC investment. EPA’s reported position is that the proposed NSR reforms are to be applied going forward and will not be affected by current court cases brought by the EPA and the Department of Justice.
It is interesting to speculate that the new NSR reforms are a signal to the courts from the Bush Administration. However, the routine maintenance provisions are proposed (not final) and potentially stringent in application, so it may be argued that the EPA’s NSR reform is really a message directed to Congress to pass the proposed Clear Skies multi-pollutant legislation. The EPA’s intentions are far from clear.
Clear Skies Act and the Clear Air Planning Act – NSR Provisions
NSR may be moving in the right the direction, but what is needed is a new level of discussion. Only Congress is in a position to raise the level of discourse.
Two proposed laws are in front of Congress this year. Both the President’s proposal (Clear Skies) and the Congressional proposal (Clear Air Planning Act) help clarify how NSR will affect existing units4 . The earlier democratic proposal introduced by Jim Jeffords is no longer a viable contender to the Administration’s position. The new replacement legislation sponsored by Jim Carper is more moderate, and the bill’s sponsor will lead the process as the ranking member of the Committee that deals with clean air issues. These two Acts deal with NSR in different ways.
Under the Administration’s Clear Skies plants would not have to go through case-by-case BACT analysis if they could meet specific NSPS limits5 . For coal fired boilers these are: NOx (1.0 lb/MW-hr), SO2 (2.0 lb/MW-hr), Mercury (0.15 lb/GW-hr), and particulate matter (0.2 lb/MW-hr). The mercury limit can be met in alternative ways that provide some flexibility. For particulates a more stringent standard applies to plants within 50 kilometers of a Federal Class 1 area (National Parks and Wilderness Areas). These plants must achieve 0.03 particulate levels in Class 1 areas (not 0.2) and are subject to the BART (Best Available Retrofit Technology) guideline under the EPA Regional Haze Rule.
The mercury standard can be met in two alternative ways: achieve an 80% reduction or install FGD and SCR. This flexibility might allow utilities alternatives to carbon injection or more reasonable levels of carbon injection.
If Congress passed the Clear Skies Act, the above standards might be a complete definition of a “clean unit” under NSR. However, for all the other units that do not meet the “clean unit” test, the alternative may be a highly restrictive definition of routine maintenance.
The Clear Air Planning Act is more straightforward than Clear Skies and the current proposed NSR reform. The Act simply defines a major modification as 1) either a component replacement that exceeds 50 percent of the capital construction cost of a comparable new unit of the same type, or 2) a change that results in an increase in the rate of emissions. The Act goes further and defines “Lowest Achievable Emission Rate” as any technology that does not exceed (in cost per ton) two times (2X) the cost guideline used for BACT. In addition, emission offsets are not required for new units. These are definitions any plant manager can interpret easily: a 50% cost test of a new unit and a predictable guide to LAER.
The Trade-Offs: Clear Skies v. Clean Air Planning Act
NSR under the Clean Air Planning Act would allow utilities to operate their plants without EPA involved in asset management decisions. However, the Act has provisions that will force more reductions in emissions sooner than Clear Skies. NOx emission budgets are less in 2008 (1.87 vs. 2.1 million tons) and the second stage reduction comes much sooner 2012 vs. 2018. NOx emissions are also based on megawatt hours rather than heat input, which provides more emission allowances to combined cycle gas plants relative to coal than the Clear Skies allocation based on heat input. In addition, the SO2 emissions in the Clean Air Planning Act come two years earlier (2008) and step down to 2.25 million tons by 2015 vs. 3 million tons by 2018 in Clear Skies.
Is the greater certainty and flexibility worth the additional investment in air pollution control to utilities? In a deregulated energy market for utilities whose future is coal generation the answer is probably yes.
The sticking point for the environmental community, however, is the inclusion of a cap on CO2. The Clean Air Planning Act incorporates CO2 emission reductions. The Administration has been adamant that CO2 is not on the table. The downside of CO2 caps is that there are no good technology options currently available to reduce CO2.
The consensus position in the scientific community is that CO2 reductions will be needed on a global basis in the future. The Administration is ideologically opposed to CO2 regulation, but the benefits of setting up a global system of CO2 trading may far outweigh the anxiety about the regulatory implications of CO2 emission caps. In addition, CO2 trading and CO2 caps may help put nuclear back on the list of new generating options.
More Regulatory Certainty Needed
Utilities are pulling back from many of the investments made in the 1990s, reducing capital expenditures and returning their focus to core businesses. Coal fired generation is one of the core areas. Capital expenditures on coal plants (SCRs, SO2 scrubbers, particulate and mercury removal) are driven by the emission budgets embedded in the State Implementation Plans and not by NSR.
Agreements have been hammered out in a number of states for specific plants and pollutants mostly as a result of lawsuits. These piecemeal court negotiated settlements are the outcomes of a broken process.
The future of NSR should be the application of BACT on new units. It only takes one new unit built somewhere in the US to re-define BACT, so much of the NSR enforcement mechanisms related to existing units is largely unnecessary to achieve clean air objectives.
What utilities (and the environmental community) need is a final resolution to the outstanding court cases and legislation modifying the Clean Air Act. At this juncture only Congress can put NSR on course. What is needed is a summit solution on Capitol Hill. EPA’s NSR regulatory reform by itself cannot move the ball ahead from any of the stakeholders’ perspectives.
1 South Coast Air Quality Management District, Public Notice for Gas Turbine BACT, http://www.aqmd.gov/back/PUBNOTICE-Gas_Turbines.htm.
2 S. William Becker, STAPPA/ALAPCO, December 16, 2002, in a briefing paper to the National Governor’s Association.
3 Cited in The Federal Register, Vol. 67, No. 251, December 31, 2002; p 80206.
4 Clear Skies Act of 2002, s2815 (Smith of NH) and H. R. 5266 (Barton); Clear Air Planning Act of 2002, s.3135, (Carper, Chaffee, Breaux and Baucus)
5 The Clear Skies alternatives are presented by Karen Burchardt and Carl V. Weilert, Burns & McDonnell, Status of Multi-Pollutant Control Legislation, presented at PowerGen 2002, December 9-11, Orlando, Florida.
Tobey Winters is Vice President, Research at Saw Mill Capital, an equity capital investment company focused on the global power market. Tobey works with the Saw Mill Capital team to provide needed capital resources to the supplier product and service companies that support the industry.
Prior to joining Saw Mill Capital, Tobey provided independent consulting on market and product issues to the industry including leading OEMs, energy technology companies
April, 18 2003
Edward Reid, Jr. says
This otherwise excellent presentation of the issues ignores one relatively recent phenomenon, which is also ignored by EPA: the evolution of merchant power generators.
The concept of "grandfathering" emissions from existing power plants in favor of tighter emissions controls on new power plants made excellent economic sense in an era in which all new power plants were built by the owners and operators of the existing plants. The approach is uniquely unsuited to a competitive generation environment.
The current system requires new entrants to invest in BACT or LAER level emissions controls and/or to purchase cleaner fuels at significant incremental ownership and operating costs. In non-attainment areas, emissions allowances must currently be purchased as well, frequently from competing power generators. These requirements represent significant barriers to market entry and impose anti-competitive cost burdens on the new market entrants. The proposed new approaches discusse above would improve on the existing approaches (in this regard) by not requiring the purchase of emissions allowances, except in the possible case of CO2 emissions caps.
The current approach also has the EPA-recognized "unintended consequence" of increasing the operating rates of existing, uncontrolled plants (which have lower ownership and operating costs in the absence of emission control system ownership and operation) while reducing the operating rates of the newer, lower emission plants (which have higher ownership and operating costs), under an economic dispatch generation regime.
There are two technically (though perhaps not politically) straightforward steps which could eliminate the market entry barriers and competitive disadvantages imposed by current legislation and regulation: a single, common, output-based emissions standard for all power plants; and, a single, common efficiency standard for all power plants. These standards would then be combined with emissions and efficiency trading programs. However, these trading programs would differ from the current trading programs in that the owners and operators of higher emission and lower efficiency plants would be required to purchase allowances from the owners and operators of lower emission and higher efficiency plants. This approach to trading programs would encourage emissions minimization and efficiency maximization in new power plant construction, since the cleanest and most efficient plants would earn the maximum number of saleable emissions allowances. This approach would impose the costs of emissions reductions and efficiency improvements on the noxious (pun intended) and inefficient, rather than on the clean and efficient, as is the case today. This appears to be an eminently sensible way forward in a competitive generation market.
These approaches are further developed in two papers available at no cost from the Center for the Advancement of Energy Markets (http://www.caem.org): Alternative Environmental Model(http://www.caem.org/website/pdf/Env_Model.pdf); and, Alternative Efficiency Model (http://www.caem.org/website/pdf/Eff_Model.pdf).