An "Inside the Fence" Analysis of CO2 Emissions from Fossil-Fueled Power Plants

Posted on March 28, 2016 by Eugene Trisko

On February 9, 2016, the Supreme Court issued a stay of U.S. EPA's Clean Power Plan (“CPP” or "Power Plan,” 80 Fed. Reg. 64,662, October 23, 2015) for reducing CO2 emissions from existing fossil-fueled electric generating units. The Court's action was unprecedented because challenges to the Power Plan by 27 states and numerous utility, business, and labor parties were still being heard before the U.S. Court of Appeals for the D.C. Circuit. West Virginia et al. v. EPA, DC Cir. No. 15-1363. The stay will remain in effect until the conclusion of all litigation against the rule. 

Among the core legal arguments against the Power Plan is EPA's reliance on "outside-the-fence" measures to reduce CO2 emissions. Section 111(d) of the Clean Act calls for EPA to set guidelines for states reflecting a standard of performance for "sources" based on the "Best System of Emission Reduction" ("BSER") that has been “adequately demonstrated.”  EPA defined BSER to include emission reduction actions that could be taken throughout the electric grid, such as limiting generation from coal units while increasing the output of existing natural gas combined-cycle units, and increasing reliance on new renewable energy sources. The data reviewed below show that the standard of performance established for coal-based generating units based on this BSER, 1,305 lbs. CO2/MWh, is not achievable in practice by any conventional coal unit. 

The Power Plan also calls for efficiency improvements at coal units that could reduce CO2 emissions. By adjusting units’ past heat rate data, EPA estimated that potential heat rate improvements of 2.1% to 4.3% were achievable for each of three regions in the U.S.  See 80 Fed. Reg. at 64,789. However, implementing these "inside-the-fence" measures would result in less than 100,000 tons of emission reductions - about two/tenths of one percent - of the overall 413-415 million ton CO2 emission reduction from base case levels projected to result from full implementation of the rule by 2030. See, EPA Tech. Sup. Doc., State Goal Computation, Table 5 (extrapolated to 48-state basis), Aug. 2015; CPP Reg. Impact Analysis at Table 3-5. 

An "inside-the-fence" analysis

EPA's methods for measuring the potential emission reductions achievable through efficiency improvements did not take into account the effects of different coal types on CO2 emissions.  Such "subcategorization" is specifically authorized by Section 111 of the Clean Air Act. This post seeks to open a line of inquiry into an alternative approach to achieving CO2 emission reductions based on the emission characteristics of the best-performing units in the coal fleet and taking into account differences in coal type.

A statistical analysis of CO2 emissions from coal plants was performed using the DOE/NETL 2007 coal plant public data base. This data base contains detailed coal type and emissions control and performance data for 2005. The objectives of the analysis were twofold:

1) To determine whether plants burning different grades of coal (bituminous, subbituminous, and lignite) have sufficiently different emission rates measured in pounds of CO2/MWh to consider subcategorization by coal type; and

2) To assess the potential CO2 emission reductions associated with applying a standard of performance based on the best-performing units in each coal category. 

The NETL data base was sorted to identify coal-fired units likely to remain in operation after implementation of EPA's 2012 Mercury and Air Toxics Standards (MATS) rule (77 Fed. Reg. 9,304, February 16, 2012).  Three screening criteria were applied: unit capacity of 400 MW or greater, current age of 50 years or less, and heat rate of 9,000 BTU/kWh or higher, typical of the performance of conventional pulverized coal boilers.

This sort produced 272 units, totaling 176.7 Gigawatts (GW) of capacity, grouped as follows:

·141 bituminous units, totaling 94.0 GW, with an average emission rate of 2,055 lbs. CO2/MWh;

·110 subbituminous units, totaling 69.5 GW, with an average emission rate of 2,214 lbs. CO2/MWh; and

·21 lignite units, totaling 13.1 GW, with an average emission rate of 2,425 lbs. CO2/MWh.

The total generating capacity represented by these 272 units is comparable to EPA’s projection of 174 to 183 GW of coal capacity remaining in service in 2030, following full implementation of the Power Plan. See, EPA CPP Reg. Impact Analysis at Table 3-12.

Regression analyses performed on the three plant groups assesses the relationship between heat rate (the independent variable) and CO2 emissions per MWh of generation (the dependent variable.) The results are summarized in Chart 1 for all 272 sampled units. The linear regression trend line confirms a moderate positive association between plant heat rate and CO2 emissions (i.e., units with lower heat rates tend to have lower CO2 emissions per MWh, and vice versa.)

Differences among the three coal types measured in average CO2 emission rates per MWh support subcategorization by coal type. As shown in Table 1, the sampled lignite units have an average CO2 emission rate 13% above the sample mean, and 18% above the average for bituminous coal units.  The average emission rate of bituminous units is 4% below the sample mean, while subbituminous coals have an average rate 3% above the sample mean. 

These differences among coal types could justify subcategorization similar to EPA’s MATS rule. MATS provides separate mercury emission limits for low-BTU lignite coals compared with the standard set for bituminous and subbituminous coals (defined by EPA as coals with a heat content of 8,300 lbs. of CO2 per million BTU, or greater.) See, 77 Fed. Reg. 9,304, 9,379.

Illustrative emission rate calculations

The three sample coal groups were analyzed for average CO2/MWh emission rates by quintile (i.e., lowest 20% emitting units, next lowest 20% emitting units, etc.)  Results of this subcategorization analysis are summarized in Table 1.  Assigning the average emission rate in CO2/MWh for the best-performing 20% units of each group of units to the other four quintiles (an approach similar to that prescribed by Congress for section 112 “MACT” standards) reduces the allowed emission rates for each subgroup, and the indicated levels of CO2 emissions measured in tons.

The overall reduction of CO2 emissions for the three coal types is 117 million tons based on 2005 emission rates and tonnages.  These data reflect NOx control retrofits in response to EPA’s 1998 NOx SIP Call, as well as scrubbers and other controls applied to meet CAA Title IV acid rain control limits. However, the data do not reflect additional retrofit control technologies added in response to the 2005 Clean Air Interstate Rule, as well as state laws and consent decrees.  The additional parasitic load associated with add-on controls would increase average heat rates (BTUs per kWh) by reducing net plant generation and increasing CO2 emission rates per MWh.

Additional research and applications

Additional analyses using more recent data are needed to assess the CO2 emission effects of retrofit controls applied since 2005, including those deployed in response to the MATS rule. This research could include additional subcategorization analyses based on metrics such as boiler age, size, and type.

If subcategorization by coal type or other criteria were applied to determine standards of performance for existing fossil-based generating units, states should be provided with flexible implementation mechanisms such as emissions trading and averaging "outside the fence." This would ensure that emission reduction targets could be achieved in a cost-effective manner, without mandating unachievable or uneconomic emission limits for specific units.

The findings of this preliminary analysis are also relevant to the determination of New Source Performance Standards (NSPS) in light of the substantial CO2 emission rate differences among different coal types. EPA chose not to subcategorize by coal type in its NSPS rulemaking under Section 111(b), and issued a uniform performance standard for coal-based generation units of 1,400 lbs. CO2/MWh.  Based on the sample unit data, meeting this standard implies a 42% reduction of CO2 emissions from lignite coals, and a 32% reduction for bituminous coals. Petitions for review of this standard also have been filed before the D.C. Circuit.  North Dakota et al. v. EPA, DC Cir. No. 15-1381.


*The author is an attorney in private practice ( who has specialized in Clean Air Act legislation and regulation since 1980. The coal quality and statistical regression data presented in this post were provided by the author to U.S. EPA staff in the pre-proposal stage of the development of the Clean Power Plan. The analysis set forth here is offered without prejudice to any legal positions by state or non-state petitioners before the D.C. Circuit in West Virginia et al. v. EPA or North Dakota, et al. v. EPA.


Table 1.  Summary of CO2 Emission Rates and Potential Reductions by Coal Type for

272 Unit Sample (176,679 MW) Assuming All Units

Meet Top-20% Average Emission Rate of Each Coal Type


Avg. Lbs. CO2/MWh


Avg. Lbs. CO2/MWh Top 20% of Units

Pct. Diff. vs. 2005 Avg.

2005 CO2 Emissions (Mil. Tons)

CO2 Emissions @ Top-20% Rate


CO2  Reduced

(Mil. Tons)































Chart 1. Regression Analysis of All 272 Coal Units,

Lbs. CO2/MWh vs. Heat Rate BTU/KWh