MEMORANDUM
TO:		Docket EPA-HQ-OAR-2006-0790
FROM:	Mary Johnson, EPA/OAR/OAQPS/SPPD
DATE:	December 2012
SUBJECT:	Beyond-the-Floor Analysis for Mercury
            and Carbon Monoxide
            Area Source Industrial/Commercial/Institutional Boilers

1.0 BACKGROUND
	This memorandum describes the EPA's beyond-the-floor analysis for mercury emissions and carbon monoxide (CO) emissions from new and existing area source coal-fired boilers with heat input capacity of 10 million British thermal units (MMBtu) per hour or greater.
      In our final rule action on reconsideration, we are amending the maximum achievable control technology (MACT)-based mercury emission limit for new and existing coal-fired boilers subject to subpart JJJJJJ of 40 CFR part 63. At the time of the March 2011 promulgation, the mercury limit for new and existing coal-fired boilers was 4.8 X 10[-6] pounds (lb) of mercury per MMBtu. That limit was based on the best performer of seven units for which data were available. The best performing boiler had a series of controls (dry limestone injection/dry scrubber/fabric filter/selective catalytic reduction/selective non-catalytic reduction (NH3 Or Urea Injection)) apparently for control of sulfur dioxide, particulate matter, and nitrogen oxides in series. All of the mercury data for the best-performing boiler were below the method detection limit. After promulgation, the EPA determined that the boiler on which the mercury limit was based is, in fact, a utility boiler and not part of the source categories being regulated under this action. Thus, the utility boiler and its emissions data should not have been considered in establishing the mercury emission limit for new and existing coal-fired boilers.
	An examination of the mercury emissions data for the remaining six units determined that the top performing boiler is now a unit with fabric filter control from Massachusetts that had measured emissions of 1.9 X 10[-6] lb of mercury per MMBtu. These emissions are above the method detection limit. Because the unit is from Massachusetts, the fuel variability factor of 10.9 for eastern bituminous coal is applicable. Using the data and the fuel variability factor of 10.9, the revised mercury emission limit for new and existing coal-fired boilers is 2.2 X 10[-5] lb of mercury per MMBtu.
	As discussed in the final reconsideration rule, we are amending the CO emission limit for new and existing coal-fired boilers from 400 parts per million (ppm) by volume on a dry basis, corrected to 3 percent oxygen, to 420 ppm by volume on a dry basis, corrected to 3 percent oxygen. It is anticipated that combustion controls, either a basic tune-up or a more advanced burner replacement or installation of a linkageless boiler management system, can achieve the MACT floor emission limit.

2.0 BEYOND-THE-FLOOR ANALYSIS
      As discussed in the preamble to the June 2010 proposed rule (75 FR 31896) and the preamble to the March 2011 final rule (76 FR 15554), the EPA determined that fuel switching was not an appropriate control technology for purposes of determining either the MACT floor level or beyond-the-floor level of control. As also discussed in the June 2010 and March 2011 preambles, we determined that an energy assessment requirement was an appropriate beyond-the-floor option for existing large boilers. These previous analyses continue to be applicable for mercury as well as the pollutants for which CO is a surrogate.
      In this beyond-the-floor analysis for mercury emissions, new and existing large coal-fired boilers would be required to comply with a mercury emission limit more stringent than the MACT floor-based emission limit of 2.2 X 10[-5] lb of mercury per MMBtu. To comply with a limit more stringent than the fabric filter-based MACT floor limit, it is expected that an affected boiler would need to employ fabric filter control along with activated carbon injection (ACI). Boilers already using a fabric filter would be expected to install an ACI system. Boilers without fabric filters would be expected to install a fabric filter and ACI system.
      The beyond-the-floor analysis for mercury emissions from major source boilers assessed the costs of and emission reductions associated with installation of ACI. The installation of ACI for control of mercury resulted in over $85,000 per additional pound of mercury emission reductions. It was determined that the beyond-the-floor option of installing ACI for mercury control is not economically feasible.[1] Based on the fact that the vast majority of area source boilers are small entities and that beyond-the-floor ACI control was not appropriate for major source boilers with higher levels of mercury and much larger economies of scale, we have also determined that the beyond-the-floor option of installing ACI for mercury control from area source coal-fired boilers is not economically feasible.
      In this beyond-the-floor analysis for CO emissions, new and existing large coal-fired boilers would be required to comply with a CO emission limit more stringent than the MACT floor-based emission limit of 420 ppm by volume on a dry basis, corrected to 3 percent oxygen. To comply with a limit more stringent than the MACT floor based limit, it is expected that an affected boiler may need to install an oxidation catalyst.
      The beyond-the-floor analysis for CO emissions from major source boilers assessed the costs of and emission reductions associated with installation of an oxidation catalyst. For units combusting solid fuel, the cost for reducing an additional ton of CO emissions is only $2,000 in annual costs. However, combustion of solid fuels results in significant amounts of dust and fly ash, which can plug or blind catalyst reactive sites necessary to oxidize CO. This reduces the efficiency of the oxidation catalyst. To maintain the effectiveness of the oxidation catalyst, the unit may require shutting down more frequently for cleaning or replacement of the catalyst, incurring additional costs on the facility. For these reasons, the beyond-the-floor option of installing an oxidation catalyst for CO control was determined to be technically infeasible. Other methods of reducing CO
emissions, such as upgrading new burners and overfire air systems, were also determined to be technically infeasible options available for solid fuel units. These combustion controls are not capable of
constantly achieving lower beyond-the-floor emission levels. Many existing units may have been designed with specific furnace heat release rates in order to be able to fire specific fuel blends, and may not be able to meet lower emission levels without complete overhaul or replacement of major boiler components at an unreasonably high cost.
        
