October 9, 2013
On Sept. 20, 2013, the U.S. Environmental Protection Agency (EPA) issued the first major action implementing the president’s Climate Action Plan, in the form of a proposed New Source Performance Standard (NSPS) rule to regulate emissions of carbon dioxide (CO2) from new fossil fuel-fired electric generating units (EGUs) under the Clean Air Act. The proposal has been applauded by some as an important first step by the administration to address climate change, and derided by others as an uneconomic and unlawful attack on the coal industry. EPA, however, specifically concluded that compliance with the rule will cost nothing, and accomplish nothing, because the rule is not predicted to result in any actually CO2 reductions.
The Emission Standards
The new proposal replaces EPA’s April 2012 NSPS power plant CO2 proposal, which is discussed here. The new proposal sets separate standards for new coal boilers, IGCC units and new gas combustion units. The standard for natural gas units (1,000 lb CO2/MWh) is the same as the prior proposal, except that combustion turbines are no longer exempt, but would rather be subject to a new standard set for smaller gas units (1,100 lb CO2/MWh) to account for cycling and efficiency issues. EPA predicts the new gas standards would readily be met by available new gas units, without incurring additional compliance costs. In turn, the standard for new coal units is set at 1,100 lb CO2/ MWh) and is based on a formal EPA conclusion that “partial” carbon capture and sequestration (CCS) is both technically feasible for coal units and has costs that are “reasonable.”
EPA’s CCS Decision
EPA’s key conclusion is that partial CCS for coal meets Section 111(b)’s standard that emission limits be based on the “best system of emission reductions … adequately demonstrated” (BSER). As part of its BSER determination, EPA evaluated four factors: (1) whether CCS is technically feasible; (2) whether its costs are “reasonable”; (3) the amount of emission reductions the technology would produce; and (4) whether selection of the technology would promote its development.
CCS is Feasible for Coal But Not Gas
First, EPA concluded that partial CCS for coal (about 40 percent capture) is technically “feasible” and is thereby “adequately demonstrated.” EPA based this decision on a “parts is parts” rationale, reasoning that each element of a CCS system has been implemented at least at some scale (pilot or commercial) in the U.S., and that there are commercial-scale coal-fired plants with CCS systems (i.e., with all the parts in one place) either under construction (two plants) or in development.
Having concluded that CCS is technically feasible for coal units, EPA also had to decide whether the same was true for gas units. It concluded that it was not, finding, tepidly, that it “was not clear that full or partial capture CCS is technically feasible.” It reasoned that there was only one gas CCS demonstration unit, but declined to perform the same “parts is part” analysis used for coal. EPA also reasoned that there are some differences between gas and coal units that could impact CCS efficacy, including lower CO2 concentrations in the exhaust gas and that at least some gas units cycle frequently.
While these two determinations will be more thoroughly evaluated in the coming weeks, EPA’s differing conclusions on the feasibility of CCS as between the two fuels raise some questions regarding thoroughness and consistency. The feasibility of CCS for coal will certainly be challenged given the dearth of operating CCS units. It also seems EPA did not apply the same feasibility standard for coal and gas such that if EPA’s CCS technical determination for coal units is valid, its conclusion that CCS is not feasible for gas units becomes more questionable.
Costs. EPA next concluded that the cost of partial CCS (but not full CCS) for coal units is “reasonable” under the statute. EPA observes that some companies are considering new nuclear units, as both an alternative to gas (fuel diversity) and a way to provide baseload power with a lower carbon footprint (implicit carbon price). It then concludes that since a new coal unit with partial CCS has costs comparable to a new nuclear plant, these costs are “reasonable” for utilities seeking fuel diversity or lower carbon power. Notably, EPA did not choose an apples-to-apples comparison since a new nuclear plant built to produce low carbon power has no carbon, while a partial CCS coal plant still has significant carbon emissions. Hence, EPA’s rational is questionable.
EPA also addressed fuel diversification concerns by noting that the partial CCS cost adder tacks on only about another 50 percent to the cost differential between new coal and gas (i.e., new coal costs $33/MWh more than a new gas unit, and partial CCS adds only another $18/MWh to the $33 cost delta). This cost adder was found to be “reasonable” because EPA concludes states and companies that want to build new coal in order to have fuel diversity are doing so for non-price purposes. EPA adds that these higher costs could be passed along to ratepayers, observing that a fuel diversity policy would be expected to be inelastic with respect to price. In other words, since a coal-based fuel diversity policy already costs more than the gas alternative, it is “reasonable” to require it to cost even more than it otherwise would.
In stark contrast to that reasoning, EPA concluded that adding partial CCS costs to new gas units would be “unreasonable.” Without citing any specific gas CCS costs, EPA concluded that since virtually all new fossil-fuel EGUs will be gas, “requiring CCS would have more of an impact on the price of electricity than the few projected coal plants with CCS.” It also noted that requiring CCS for gas could provide incentives to not replace older coal units with new gas, which, it concluded, would have adverse emission impacts. Although left unsaid, EPA’s concern about avoiding disincentives to replace existing coal units with new gas would become even more important if EPA issues an existing source power plant rule that raises the cost of operating existing coal units.
In effect, EPA concludes that CCS costs on new fossil fuel EGUs are unreasonable if those costs have to be borne by ratepayers in most states (gas), but are reasonable if those costs will be borne by ratepayers in only some states (coal). While this analysis may seem odd, EPA argues that case law under Section 111(b) allows it to assess that section’s cost consideration criterion on a “region wide or nationwide basis, and … [is] not limited to the individual source.” What is not clear is whether that case law supports the notion that costs that would otherwise be unreasonable if borne by many can somehow be reasonable if borne by only a few.
The third key factor in EPA’s BSER decision focused on whether requiring CCS would encourage the development of the technology. EPA concluded, without analysis, that a CCS mandate “will promote further development of the technology,” “because any new fossil fuel-fired utility boiler or IGCC unit will have to install partial CCS capture in order to meet the emissions standard.” EPA appears to be incorrect in this conclusion, at least according to the International Energy Agency (IEA). The issue is whether CCS can ever be a “pull me” technology (i.e., a technology that will be developed and implemented if it is mandated). In a 2012 report assessing policy options to accelerate CCS development, the IEA concluded that CCS is not a “pull me” technology, stating:
In other words, mandating CCS before it is ready does not encourage development and employment of the technology, but rather produces the opposite effect. Since the IEA’s conclusion about “forcing” emission reductions through high carbon prices would apply equally to efforts to “force” emission reductions through a NSPS technology mandate, we would expect the IEA would strongly disagree with EPA’s conclusion that mandating CCS will promote its development (assuming we can read the words “not credible” as expressing disagreement). Indeed, the report suggests that a too-soon CCS mandate is one of the worst policy choices that EPA could make if its objective is actually to combat climate change.
That a too-soon CCS mandate would retard development of the technology can also be explained by example. While some companies have decided, for their own reasons, and in the absence of any mandate, that it is worthwhile to attempt to develop commercial-scale CCS units, those decisions do not include a financial bet on the underlying generating asset, which forms the bulk of the investment. Once mandated, the viability of the underlying generating unit is then at risk, since it cannot be operated unless the CCS system can be completed and successfully operated within acceptable cost expectations. In other words, under a CCS mandate, no CCS = no power plant.
To place this “bet” in the context of EPA’s cost numbers, the coal plant would cost $92/MWh and the CCS system another $18/MWh. Thus, a “voluntary” CCS system risks $18/MWh (maybe 20 percent of its investment) on the CCS component, and if it fails or is not cost effective, the company still retains the viability of its $92/MWh coal unit investment. With CCS mandated, a company would have to bet the entire $110/MWh (100 percent of its investment) on the efficacy of CCS. That seems like a bet that companies will not take. It seems unlikely that rearranging the risks of CCS the way EPA proposes will plausibly fulfill Congress’s Section 111(b) goal to “stimulate the development of new technology.”
A Rule about Nothing
EPA states that its rule “will result in negligible CO2 emission changes, quantified benefits, and costs, by 2022,” which would seem to contradict EPA’s claim that the rule will “protect public health and address climate change.” The reason EPA predicts no costs and no benefits on the gas side is because EPA chose a CO2 limit for gas units that is readily met by new units. On the coal side, EPA expects no new coal units will be built, but if a few are, it expects that they would have included CCS anyway. It is not altogether clear why EPA, or any other agency, could have a rational basis to issue a rule that does absolutely nothing.
The real reason EPA may have issued a proposed rule that does nothing may have little to do with this rule, and everything to do with its upcoming CO2 rule for existing power plants under Section 111(d). EPA notes, in a single sentence, that the new proposal under Section 111(b) “will serve as a necessary predicate for regulation of existing sources” under Section 111(d). EPA is correct to say that the statute precludes it from issuing a rule covering existing sources under Section 111(d) unless a 111(b) rule applies to new sources. Hence, from a practical perspective, EPA has to issue this 111(b) rule, even if it does nothing, in order to issue the existing source rule called for in the president’s plan. It will be interesting to see whether a mere “pro forma” new source rule will satisfy the statute’s precondition to issuing a 111(d) rule.
Moreover, in this proposal, EPA has effectively decided to “get out of the way” of the market’s dash to gas. Does this suggest that EPA intends to obtain CO2 reductions from existing units in the forthcoming 111(d) rule by effectively mandating fuel-switching?
Clean Air Act Team
McGuireWoods LLP is a full service law firm with a focused practice in climate change and Clean Air Act matters. For more information on the final light duty vehicle rule or EPA's initiatives to regulate greenhouse gases, please contact the authors, or any member of our Clean Air Act team.