US EPA is presenting boiler operators with a real dilemma. The agency has proposed emission limits for some existing boilers without identifying a reliable path to achieving a compliant emission rate. This is the apparent result of the proposed Maximum Achievable Control Technology (MACT) standards for the control of hazardous air pollutants (HAPs) from industrial, commercial and institutional boilers and process heaters.9 When US EPA set emission limits for new and existing sources for its Gas-2 subcategory based on the best controlled similar sources, none of these sources actually used a control technology to achieve its emission rate. Similarly, most of the top-performing boilers and process heaters for dioxin/furans do not use a control technology or even an identifiable work practice that targets these compounds for emission reduction. As a result, there is no clear technological path for these sources to achieve compliance. US EPA claims its hands are tied by court interpretations of Section 112(d) of the Clean Air Act.

US EPA’s initial attempts to set MACT standards based on control technologies did not fare well in the courts. In a series of decisions culminating in the Brick MACT decision, the DC Circuit Court expressly rejected US EPA’s attempt to set limits based on what was achievable by a common control device or system. Sierra Club v. EPA, 479 F.3d 875, 880-81 (D.C. Cir. 2007) (citing Cement Kiln Recycling Coalition v. EPA, 255 F.3d 855 (D.C. Cir. 2001) and National Lime Ass’n v. EPA, 233 F.3d 625 (D.C. Cir. 2000)). The court admonished US EPA for ignoring nontechnology factors that affect emissions, such as raw materials, fuels, unit design and operating practices. The court accepted the premise advanced by environmental interests that strict numeric emission limits based on the best-performing sources will reflect the optimum combination of the full suite of emission reduction options available to sources in the source category. US EPA’s discretion was severely limited by the court’s determination that the Act unambiguously defines the MACT floor by reference to what is achieved in practice by the top 12 percent of the existing sources in a source category. As a result, US EPA is taking a strict numeric approach to setting the MACT limits in the Boiler MACT rule without regard to achievability. However, US EPA has other tools that could be useful in mitigating the dilemma of the unachievable Boiler MACT standard.

Judge Williams in his concurring opinion in the Brick MACT decision directed US EPA to the use of subcategories to address achievability. If US EPA limits the subcategory to truly similar sources, it increases the likelihood that emission rates achieved by the best performers will be achievable by the rest of the subcategory. Further subcategorization will help the Boiler MACT rule. The Gas- 2 subcategory, for instance, could separate units that burn landfill gas from those that burn coke oven gas. These fuels have significantly different emission profiles, and the lower-emitting landfill gas units are setting most of the emission rates for the current subcategory. This makes the emission limitations potentially unachievable for coke oven gas-fired sources. A subcategory for coke oven gas would make the MACT standards more achievable.

Unachievable standards for coke oven gas and other process gases create significant adverse environmental consequences. Process gases with Btu value must be combusted to ensure a safe, non-explosive environment and to oxidize the concentrations of harmful organic compounds. Over-regulation of the emissions from boilers combusting process gases will force these gases to be flared and replaced with fossil fuels, resulting in more emissions and less-efficient energy use. US EPA has excluded the waste heat recovery boiler on the basis that it recovers “normally unused energy and converts it to usable heat.” The same justification may apply to exclude process gas combustion from the source category.

US EPA should be encouraged to make choices that will mitigate the achievability problem. These include:

  • Best performing sources should be determined based on the achieved emission rates for all pollutants, not for each pollutant. Proposed source categories do not have a single source that is achieving all of the pollutant-specific standards simultaneously.  
  • Set standards only for those pollutants for which adequate reliable emissions data are available now and require testing in the rule to fill the data gaps for dioxin-furans and others to prepare for the next phase of emission standards.  
  • Use all available emissions information, not just stack test information, to develop emission profiles for all units in the source category so that the MACT standard is set based on the true top performing 12 percent of a category and not a single source when EPA has collected data from only nine sources in the categor y.  

The Clean Air Act as interpreted by the courts has set a rigid procedure for setting MACT standards that limits US EPA’s discretion. The rulemaking process offers an opportunity to educate the agency about its unachievable standards and to encourage US EPA to use the discretion that remains to avoid the consequences of unachievable standards. Setting “technology-forcing” regulations on the belief that innovations will emerge in time to meet the standards is a risky tactic in the current global economy. Why would a global company choose to invest capital in the United States on the hope that innovations will allow it to operate when it can invest in Brazil, India or China without that risk? US EPA has a responsibility to use the full measure of its discretion to set standards that offer a clear path to compliance for boilers and process heaters.