This is the first in a series of questions that highlights RFF’s Expert Forum on EPA’s Clean Power Plan.
RFF asks the experts: What is the definition of the “best system of emission reduction”?
The proposed Clean Power Plan originates from the US Environmental Protection Agency’s (EPA) authority under Section 111(d) of the Clean Air Act to require states to submit plans that establish and implement a standard of performance achievable through “application of the best system of emission reduction.” EPA chose to interpret the word “system” to encompass: (1) emissions rate reductions at coal plants, (2) shifting generation from coal to natural gas, (3) increasing generation from renewables, and (4) improving end use energy efficiency. A central issue in the environmental effectiveness and legal defensibility of the Clean Power Plan is whether this definition of “system” is within EPA’s statutory authority. EPA views the system as the network infrastructure that delivers electricity services to consumers, but others might hold that a system includes only technical operation of emitting plants.
What guidance can experts give to EPA on this issue?
“Given the lack of guidance in the statute and deference shown to agency’s interpretation of statutory ambiguity, I think EPA’s interpretation of ‘best system of emission reduction’ is likely to survive legal challenge, with two caveats.”
—Nathan Richardson, Assistant Professor, University of South Carolina School of Law (See full response.)
“[A] standard of performance must be based on ‘the application of the best system of emission reduction.’ In this case, the question is ‘the application of the system to what?’ EPA says, ‘to anything that produces or uses electricity.’ But the answer, according to the statute and almost 40 years of regulatory history, is ‘the type of facility being regulated.’
—Jeffrey Holmstead, Partner, Bracewell & Giuliani (See full response.)
“The phrase ‘best system of emission reduction’ plainly points toward a broad approach to cutting emissions. . . . [It] can certainly encompass all the techniques available at reasonable cost for reducing carbon pollution at power plants, not just end-of-pipe controls.”
—David Doniger, Director, Climate & Clean Air Program; and Benjamin Longstreth, Senior Attorney, Natural Resources Defense Council (See full response.)
Associate Professor, University of South Carolina School of Law
Visiting Fellow, Resources for the Future
EPA has sought input on its proposed approach to identifying the best system of reduction. There’s a common but important misunderstanding about “best system.” The language appears in the section of the act that defines performance standards, which are the tools the agency aims to use to regulate power sector emissions. Many people take this placement to mean that such standards let EPA and the states determine what the “best system” is and then direct emitters to apply that system. But that’s not right. The statute says that standards “reflect” the best system, not that they are the best system. In other words, regulators get to make their best guess at what emitters could do to cut their emissions (considering costs and other factors), but they don’t get to tell them what to do in practice. This is what makes performance standards different from command-and-control regulation. The standards are goals—real and legally enforceable ones—but not mandates about what about what actions emitters will take to achieve the goals.
A corollary of this is that “best system,” in my view, not only doesn’t give regulators any authority to require emitters to take specific actions, it doesn’t even really give them any guidance as to what kinds of emissions-cutting actions they can assume emitters will take (and, therefore, what the components of the “best system” actually are). For example, can regulators consider whether emitters can trade amongst themselves or whether they can work with customers to reduce demand—each of which might reduce the costs of emissions cuts? Despite the broad appeal of the words “best system”, I don’t think they answer that question. Guidance has to come from elsewhere—other parts of the statute that have actual, substantive weight to them. Unfortunately, Section 111 has nothing like that. I believe this is due to the technology-driven roots of the section; it was long assumed (and for a time explicitly stated) that standards would reflect the best technology available to emitters. Some argue that this interpretation should still hold and that, therefore, only demonstrated technological upgrades at the plant should be considered in setting the standard. I think that view is too narrow for a variety of reasons.
EPA’s view is to take guidance from the rest of Section 111, which says that states (under EPA guidance) must set performance standards “for” specified categories of sources. EPA interprets standards “for” sources to mean standards that assume (though, again, do not require) sources to take any action that would reduce their emissions. Under EPA’s interpretation, all four of its building blocks can be considered in setting standards since each reduces emissions from fossil power plants. Other measures that would help cut atmospheric carbon but would not cut fossil power emissions, like paying other sectors to reduce their emissions, or paying to plant trees, are out of bounds. I think that’s important, because it means that EPA’s broad interpretation isn’t unlimited. Given the lack of guidance in the statute and deference shown to the agency’s interpretation of statutory ambiguity, I think EPA is likely to survive challenge, with two caveats.
One caveat is that EPA’s proposal undermines its own interpretation by grouping coal and gas plants into separate categories. Its second building block explicitly assumes generation will shift from coal to gas. Doing so does cut overall emissions from fossil power, but not those from gas—gas emissions go up. The other four building blocks may have similar effects, especially to the extent that states allow trading. In other words, standards ostensibly “for” the gas category actually increase gas emissions. EPA can solve this problem simply by combining all fossil power into one legal category, as it already assumes for purposes of setting targets in the proposal. The agency has cryptically “co-proposed” such a merger of categories, but the traditional split categories remain the primary approach, for reasons that are unclear. The agency has extremely broad powers to define and redefine categories of sources. The fix here is easy—joining all fossil power in the same category—but the agency actually needs to do it or risk unnecessary legal vulnerability.
The other caveat is the view expressed by some judges, including those on the Supreme Court, that they will be less deferential to agency interpretations when those interpretations are based on sparse statutory text and assume broad powers for the agency. Think of this as the UARG exception to Chevron deference, which I elaborate on in this blog post. This is a risk for the agency, but it may be overblown. Opponents of broad EPA powers would surely prefer to have better arguments than a claimed special exception to Chevron.
On balance, I think EPA’s view on the “best system of emission reduction” and its four building blocks should survive.
Partner, Bracewell & Giuliani
EPA’s proposal stretches the term “standard of performance” far beyond the breaking point. Under the Clean Air Act, a “standard of performance” is a requirement (usually an allowable emission rate) that applies to an individual facility and is based on the “best system of emission reduction” that will ensure a “continuous emission reduction” from that type of facility. This is clear from the language of the act and almost 40 years of regulatory history. Even now, EPA agrees with this reading of the statute when it comes to new power plants. But when it comes to existing power plants, this term is somehow transformed into a requirement that applies to the state as a whole—a statewide allowable emission rate that varies dramatically from state to state based on EPA’s view of how the entire “electricity system” in each state, including both supply and demand, should be changed over the next 15 years. It is highly unlikely that EPA’s rather breathtaking new interpretation of a 40-year-old statutory provision will stand up in court.
Section 111 has been in place since 1977 and is quite straightforward. Before issuing any regulation under Section 111, EPA must first identify specific types of facilities (known as “sources”) that will be regulated. Then, under Section 111(b), EPA establishes a “standard of performance” that applies to any new source of that type. Over the last 37 years, EPA has set dozens of these standards for different pollutants from many types of sources. In every case, the standard of performance is a requirement (usually an allowable emissions rate) that each “new source” must meet.
Section 111(d) provides that, under certain circumstances, EPA may require states to set a “standard of performance” for “any existing source … to which a section 111(b) standard of performance would apply if such existing source were a new source.” It also states, however, that EPA “shall permit the State in applying a standard of performance to any particular source ... to take into consideration, among other factors, the remaining useful life of the existing source to which such standard applies.” Thus, EPA sets a standard of performance for any “new source” in the country, and individual states set a standard of performance for “any existing source” within their boundaries.
The statute certainly contemplates that a standard of performance must be met by each and every regulated source—new or existing. EPA agrees with this reading when it comes to new sources—even for carbon emissions from new power plants. But when it comes to carbon emissions from existing power plants, EPA claims that a “standard of performance” is something altogether different. Thus, instead of requiring states to establish a standard that applies to individual power plants, EPA is now proposing to require each state to meet a statewide emissions rate that is based on, among other things, (1) emissions that, in EPA’s view, can be “avoided” by programs designed to reduce the demand for electricity; (2) the amount of business that should be shifted from coal-fired plants to gas-fired plants; and (3) requirements that each state should adopt to compel the construction and use of wind and solar plants. This legally binding standard varies substantially from state to state depending on EPA’s view of how each state should change its current electricity system.
EPA justifies this approach based on a statutory provision that defines a “standard of performance” as “a standard for emissions of air pollutants which reflects the degree of emission limitation achievable through the application of the best system of emission reduction which (taking into account the cost of achieving such reduction and any nonair quality health and environmental impact and energy requirements) the Administrator determines has been adequately demonstrated.”
EPA focuses on the word “system,” which is certainly a broad term. And the statute does define a standard of performance, in part, as “the degree of emission limitation achievable through the application of the best system of emissions reduction.” The statute also provides that this system must ensure a “continuous emission reduction” from a source being regulated. But the key legal question in this case is not what a “system” may be. The statute says that a standard of performance must be based on “the application of the best system of emission reduction.” In this case, the question is “the application of the system to what?” EPA says, “to anything that produces or uses electricity.” But the answer, according to the statute and almost 40 years of regulatory history, is “the type of facility being regulated.” In the context of Section 111(d), this means to “any existing source,” as long as it ensures a “continuous emission reduction” from that source and that, “in applying a standard of performance to any particular source,” the state is able to “take into consideration, among other factors, the remaining useful life of the existing source to which such standard applies.”
For further information, read my July 30 testimony to the US House Committee on Science, Space and Technology.
David Doniger, Policy Director, Climate & Clean Air Program and
Benjamin Longstreth, Senior Attorney
Natural Resources Defense Council
EPA has a solid legal basis for using a system-based approach when it promulgates the emissions guideline that will govern carbon pollution standards for existing power plants. Under the Clean Air Act, EPA must determine the emissions limit that reflects the emissions reductions achievable using the “best system of emission reduction.” 42 USC Section 7411(a)(1). Fossil fuel power plants are part of an interconnected electricity system that matches supply and demand. As many states and utilities have demonstrated, carbon pollution can be reduced by increasing the efficiency of fossil fuel power plants, meeting demand with other lower-emitting sources, and reducing demand through energy efficiency. The “best system of emission reduction” properly takes advantage of all three techniques. Indeed, this is not the first time EPA has looked “beyond the fenceline” under Section 111(d). EPA’s 1995 standard for existing municipal waste combustors adopted such a system by allowing nitrogen oxide emissions credit trading between individual combustors. And in 2005, EPA adopted a cap-and-trade system for power plants’ mercury emissions under Section 111(d). (That approach failed because power plant emissions of mercury, a hazardous air pollutant, are properly subject to control under Section 112.)
The language of Section 111 supports a system-based approach for at least three reasons. First, the phrase “best system of emission reduction” plainly points toward a broad approach to cutting emissions. According to the American Heritage College Dictionary, a “system” is a “group of interacting, interrelated, or interdependent elements forming a complex whole.” A “system” of emissions reduction for the power sector can certainly encompass all the techniques available at reasonable cost for reducing carbon pollution at power plants, not just end-of-pipe controls.
Second, in 1990, Congress removed terms that might have limited the range of options available to EPA. Under the 1977 amendments, standards for new sources were required to reflect “technological systems of emission reduction” and standards for existing sources were required to reflect the “best system of continuous emission reduction.” In 1990, Congress amended this language to remove both the “technological” and “continuous” requirements, returning to language originally adopted in 1970. The effect of these amendments was to expand the range of compliance options EPA may consider in determining the “best system of emission reduction.”
Third, additional support for a system-based approach is found in Section 111(d)’s reference to adoption of a “procedure similar to that provided by section ” of the Clean Air Act. 42 USC Section 7411(d). Section 110 allows for use of system-based measures including “economic incentives such as fees, marketable permits, and auctions of emissions rights.” 42 USC Section 7410(a)(2)(A). Through this reference to section 110, Congress indicated that these system-based measures may also be used under Section 111(d). Accordingly, EPA may include them when it establishes power plant carbon emissions guidelines.
Given the integrated nature of the power system, it is arguable that EPA’s determination of the “best” system of emissions reductions must include the full suite of demonstrated and available measures to reduce emissions at fossil-fuel power plants. Greater reliance on lower and non-emitting energy sources and reduction of demand through energy efficiency programs are well-demonstrated methods of reducing emissions at fossil fuel fired power plants. It is reasonable for EPA to rely on a well-grounded forecast of the supply of additional renewables and energy efficiency, just as in other cases where EPA has relied on well-grounded forecasts of the supply of low-sulfur coal or end-of-pipe pollution control equipment. Accordingly, it is not merely permissible for EPA to rely on such system-wide pollution reduction methods. Rather, a rule that ignored these available reduction methods could not be defended as the “best system of emission reduction.”