10/11/2018 - Power Sector Rules - Regulatory Rollbacks

Legal Analysis of the Proposals to Repeal and Replace the Clean Power Plan

EELP has also produced a longer memorandum that should serve to be useful to people reviewing the proposed ACE Rule and writing comments on it. It provides an extensive step-by-step layout of the legal arguments advanced in the proposal to repeal the Clean Power Plan and the proposed ACE Rule with brief comments on the vulnerabilities of the arguments.

Memorandum on the proposal to repeal the Clean Power Plan and the proposed Affordable Clean Energy Rule

Points of Interest
  • Both the Clean Power Plan (CPP) repeal and the Affordable Clean Energy (ACE) proposals purport to show that the CPP’s interpretation of section 111 is impermissible on its face. However, EPA largely argues that its new interpretation is merely preferable to the CPP’s interpretation.
  • The proposals insist that the similarities between the language of section 111(d) and that of section 165 (establishing the Prevention of Significant Deterioration [PSD] permitting program) support the ACE proposal’s interpretation.
    • The comparison is presented in a way that ignores the context of the two provisions – PSD addresses individual sources permitting requirements and section 111(d) contemplates a uniform requirement for a category of sources – and drains section 111(d) of independent meaning, an approach generally disfavored by federal courts.
    • The ACE proposal uses its assimilation of sections 111(d) and 165 to arrive at 111(d) implementation requirements that essentially mimic the PSD permitting program. The proposal fails to address the significant differences between the grants of authority in sections 111(d) and 165, account for its departure from previous section 111(d) approaches or cite precedent for its new ACE approach.
  • The Regulatory Impact Analysis for the ACE proposal shows that it will have virtually no impact on CO2emissions, modest net benefits at best and net costs in several other scenarios.
    • Against that backdrop, the proposal’s justifications for determining that the “best system of emission reduction” provides only de minimisCO2reductions could also support declining to issue 111(d) guidelines at this time.
    • The ACE proposal is an outlier when measured against significant EPA and judicial precedent supporting best system of emission reduction determinations that achieve significant emissions reductions even at relatively high cost.
  • The analysis shows a 7.1%-9.2% increase in coal demand relative to the CPP in 2030, and a 4% increase in coal generation relative to business as usual.
    • These results reveal that EPA anticipates that the proposal will increase coal plants’ utilization and extend their useful lives.
    • If coal plants are used more and longer, SO2and NOx emissions are likely to increase as ACE’s revisions to the New Source Review program will reduce requirements to make pollution control upgrades.

In October 2017, EPA proposed repealing the Clean Power Plan because the agency is changing the interpretation of section 111(d)’s “best system of emissions reduction” (BSER), which EPA relied on when issuing the CPP. The CPP determined that BSER included the replacement of higher-emitting generation with lower-emitting generation supplied via the grid (“generation-shifting”). The repeal proposes that the “best” interpretation of Clean Air Act section 111(a)(1)’s “best system of emission reduction” is limited exclusively to measures that can be applied “inside the fence line”.

The repeal proposal argues – somewhat unclearly and at times inconsistently –  that the CPP’s interpretation of BSER is impermissible under the plain meaning of the statute. The proposal attempts to demonstrate that the CPP’s interpretation is impermissible mostly by way of arguments that affirmatively support the “inside the fence line”-only interpretation of BSER, at times suggesting that this interpretation is simply preferable. Those arguments rest on a very close parsing of words such as “at”, “for” and “to” in the statutory text, legislative history, a novel-seeming assimilation of section 111 and section 165, and general policy and federalism considerations.

When the repeal was proposed, EPA stated that it had not yet determined whether it would propose a successor rule under 111(d), leaving open the possibility that it would take no action beyond adopting the new interpretation and issuing a final repeal.

After issuing an Advanced Notice of Proposed Rulemaking in December 2017, EPA put its new interpretation into action when the ACE proposal was released in August 2018. ACE affects only coal-fired power plants with generating capacity of more than 25 megawatts (MW). It determines that BSER is a handful of “candidate technologies” that result in heat rate improvements at individual plants. Unlike the vast majority of emissions standards set under section 111, ACE does not specify an emissions rate that would benchmark the expected performance of the listed measures.  Instead, it charges states with the task of engaging with each of the affected sources in their respective jurisdictions to determine, on a source-by-source basis, which, if any, measure to apply and the schedule that the source must follow.

The preamble to the proposed ACE rule incorporates by reference the “inside the fence line”-only legal arguments made in the repeal proposal. It elaborates on one of those arguments in particular – the ostensible mirroring of the BSER language in section 111 and comparable language in the Clean Air Act’s PSD and Best Available Control Technology (BACT) provisions. The ACE preamble presses the case without acknowledging or addressing a variety of factors that clearly distinguish the programs. The ACE proposal’s assimilation of 111(d) and BACT extends well beyond its desire to import BACT interpretative constraints to 111(d). The ACE proposal’s approach to BSER (a list of technologies rather than a benchmark performance rate) and the level of responsibility it gives to states mimics PSD source-by-source permitting for new and modified sources, but for a numerous group of existing sources, and with fewer federal constraints on state discretion. Although the proposal acknowledges that it is modeling its approach to 111(d) on PSD, it does not offer any precedent for doing so. It also does not explain how the ACE approach compares to or differs from the approaches adopted in the Clean Air Mercury Rule and the Landfill Gas New Source Performance Standards, other 111(d) programs, except to allege inaccurately that all previous 111(d) rules limited BSER to “inside the fence line” measures.

The ACE rule would function as follows: EPA is proposing candidate technologies to achieve heat rate improvements (HRI) as the BSER. EPA will provide a range of reductions and costs that correspond with those candidate technologies. States will then use that information to develop standards of performance tailored to each electric generating unit (EGU) in their respective jurisdictions. Finally, states will set customized compliance schedules for their sources based on the standards for each EGU.

Here are the key passages from ACE’s preamble that capture the program it would create:

  • “To meet the requirements of the new proposed implementing regulations, EPA is proposing candidate technologies for HRI measures corresponding to a range of reductions and costs as information regarding the degree of emission reduction achievable through application of the BSER. Because affected EGUs in each state are different and the application of different HRI measures may take into account source-specific factors, EPA is providing expected ranges of HRIs.”
  • “EPA expects that states can use the information that EPA provides on the degree of emission limitation in developing standards of performance for affected EGUs as part of establishing a standard of performance for inclusion in a state’s plan pursuant to the requirements of section 111(d)(1).”
  • “…the ranges of HRIs are provided as guidance for states to use in evaluating the efficacy of implementing each measure identified as part of the BSER candidate technologies at each affected EGU. While the HRI potential range is provided as guidance for the states, the actual HRI performance for each of the candidate technologies will be unit-specific and will depend upon a range of unit-specific factors. The states will use the information provided by EPA as guidance, but will be expected to conduct unit-specific evaluations of HRI potential, technical feasibility, and applicability for each of the BSER candidate technologies.”
  • “Once a state evaluates the HRIs identified as part of the BSER in establishing a standard of performance for a particular affected EGU, it is within the state’s discretion to take certain factors concerning that source, such as remaining useful life, into consideration when determining how the standard of performance should be applied.”
  • Additionally, the new proposed implementing regulations require that an emission guideline identify information such as a timeline for compliance with standards of performance that reflect the application of the BSER. Seeproposed 40 CFR 60.22a. However, given the source-specific nature of this proposed emission guideline and reasonably anticipated variation between standards established for sources within a state, EPA believes it more appropriate that a state establish tailored compliance deadlines for its sources based on the standard ultimately determined for each source.”
New Source Review (NSR) Reform

ACE also proposes an amendment to the New Source Review program that would allow sources to avoid NSR permitting if their hourlyemissions rates did not increase post-modification, even in cases where annual emissions were projected to increase. The proposal erects this hourly test as a new first-step or gatekeeper in the NSR process. EPA acknowledges that this creates a more lenient NSR regime and argues that alleviating the burden of New Source Review is critical to facilitating sources’ adoption of the “candidate technologies” to improve their heat rates. As written, the proposal does not limit this additional step in the NSR application test to electric generating units that must adopt a “candidate technology” as a result of ACE. However, EPA does state that it “is soliciting comment on whether to confine the applicability of the hourly test to a smaller subset of the power sector”.

The use of a preliminary hourly emissions test all but explicitly concedes that the HRI measures will likely increase annual emissions. Without this “relief” from NSR, the plants installing “candidate technologies” would then need to install additional pollution controls for other pollutants like NOx and SO2, raising the overall compliance costs of the program or dissuading utilities and states from adopting HRI measures in a greater number of cases. The proposal does acknowledge that as many as 80 percent of coal fired power plants currently operate with sub-optimal NOx and SO2controls, below what would be required if they underwent NSR. The proposal offers only a brief explanation as to why maintaining this level of emissions under-control is acceptable notwithstanding the implied expectation that new investment in coal plants will increase their usage, extend their operating lives and yield higher levels of pollution.

Emissions and Benefit/Cost Analysis

EPA’s analysis shows a decline in CO2emissions of 33% below 2005 levels by 2030, as well as a significant decline in other pollutant emissions by 2030.  The vast majority of the decline is built into EPA’s projection of emissions in the business-as-usual (No CPP) scenario. The projection presumes that natural gas prices will decrease and remain low while coal-fired power plants continue to retire and renewable energy increases its penetration thanks to rapidly dropping costs. The projection also reflects continued progress in regional haze and ozone NAAQS implementation.

EPA’s analysis shows that ACE would cut CO2emissions from 2005 levels by no more than 1.5 percent by 2030 compared to business as usual. CO2emissions are projected to be 3% higher by 2030 than they would be under the CPP (which was re-modeled in accordance with the new business as usual projection).  Relative to business as usual, the ACE rule would reduce SO2by 0.7%, reduce NOx by 1% and reduce mercury by 0.5% by 2030. Meanwhile, EPA projects that U.S. retail electricity prices in 2025 will be 0.2% to 0.5% lower than they would have been under the CPP. Interestingly, the cost of the CPP (as re-modeled) relative to business-as-usual (No CPP) would be only about $0.7B per year in 2025 and 2030; and $0.4 B in 2035.

EPA presents several different approaches to benefit-cost analysis, showing results both for selected years 2025, 2030 and 2035, and in net present value terms and annualized terms for the period 2023-2037.

In some cases, the NPV and annualized net benefits of ACE’s CO2emissions impacts are either negative or, at best, exceedingly small, when compared to the CPP. In CO2-only comparisons with business-as-usual (No CPP), the analysis presents eight cases showing net costs and four showing net benefits.

When taking account of CO2, SO2, and NOx and emissions impacts, the analysis shows that ACE imposes net costs in every one of the 24 cases presented relative to the CPP. ACE also results in net costs across the board when comparing the costs and benefits of ACE to CPP over the period 2023-2037 in net present value and annualized terms. Taking account of the emissions impacts of the same suite of pollutants shows several cases whose ranges include both net costs and net benefits relative to business-as-usual (No CPP).

Finally, although EPA insists on taking a source-by-source approach and covers only coal-fired power plants, the CO2reductions attributed to ACE are delivered by the electricity system as a whole. Under at least one scenario, the analysis shows that more than half of the CO2reductions attributed to ACE occur at non-coal sources.

These negligible emissions reductions and cost-savings beg the question: why did EPA issue this proposal? The largest impact of ACE seems to be in coal demand, with EPA reporting that ACE will yield, relative to the CPP, a 4.5% to 5.9% increase in coal demand in 2025 and 7.1% to 9.2% increase in coal demand in 2030. Even so, compared to business as usual, ACE is projecting an increase in coal generation of 4.4% in 2025 and 4% in 2030. The increases relative to business as usual appear to be the result of additional investment in coal plants that would not otherwise occur without implementation of ACE and a more lenient version of NSR.