Technical Support Document (TSD)
for the Proposed Federal Implementation Plan Addressing Regional Ozone Transport for the
2015 Ozone National Ambient Air Quality Standard
Docket ID No. EPA-HQ-OAR-2021-0668
EGU NOx Mitigation Strategies Proposed Rule TSD
U.S. Environmental Protection Agency
Office of Air and Radiation
February 2022
Page 1 of 40
-------
A. Introduction: 2
B. Optimizing and Restarting Existing SCRs 3
1. Cost Estimates for Optimizing and Restarting Existing SCRs 3
2. NOx Emission Rate Estimates for Full Operation of Existing SCRs 8
C. Installing State-of-the-Art Combustion Controls 14
1. Cost Estimates for State-of-the-Art Combustion Control Upgrade 14
2. NOx Emission Rate Estimates for State-of-the-Art Combustion Control Upgrade 16
D. Cost Estimates for Optimizing and Restarting and Optimizing Idled Existing SNCR 18
E. Cost and Emission Rate Performance Estimates for Retrofitting with SNCR and Related Costs 22
F. Cost and Emission Rate Performance Estimates for Retrofitting with SCR and Related Costs 24
G. Generation Shifting 27
H. Feasibility Assessment: Implementation Timing for Each EGU NOx Control Strategy 29
I. Additional Mitigation Technologies Assessed but Not Proposed in this Action 33
1. Combustion Control and SCR Retrofits on Combined Cycle and Combustion Turbine Units 33
2. Mitigation Strategies at Small Units that Operate on High Electricity Demand Days (HEDD) 34
Appendix A: Historical Anhydrous Ammonia and Urea Costs and their Associated Cost per NOx ton
Removed in a SCR 40
emissions). These analyses inform the EPA's evaluation of costs and emission reductions in Step 3 of its
four step interstate transport framework.
NOx control strategies that are widely available for EGUs include:
• Returning to full operation any existing SCRs that have operated at fractional design capability;
A. Introduction:
The analysis presented in this document supports the EPA's proposed Federal Implementation Plan
Addressing Regional Ozone Transport for the 2015 Ozone National Ambient Air Quality Standard
(Cross-State Air Pollution Rule for the 2015 Ozone NAAQS). In developing this proposal, the EPA
considered all NOx control strategies that are widely in use by EGUs, listed below. This Technical
Support Document (TSD) discusses costs, emission reduction potential, and feasibility related to these
EGU NOx emission control strategies. Specifically, this TSD explores three topics: (1) the appropriate
representative cost resulting from "widespread" implementation of a particular NOx emission control
technology; (2) the NOx emission rates widely achievable by "fully operating" emission control
equipment; and (3) the time required to implement these EGU NOx control strategies (e.g., installing
and/or restoring an emission control system to full operation or shifting generation to reduce NOx
emissions). These analyses inform the EPA's evaluation of costs and emission reductions in Step 3 of i
four step interstate transport framework.
NOx control strategies that are widely available for EGUs include:
• Returning to full operation any existing SCRs that have operated at fractional design capability
• Restarting inactive SCRs and returning them to full operation;
• Restarting inactive SNCRs and/or returning to full operation any SNCRs that have operated at
fractional design capability;
Page 2 of 40
-------
• Upgrading combustion controls with newer, more advanced technology (e.g., state-of-the-art low
NOx burners);
• Installing new SCR systems;
• Installing new SNCR systems; and
• Shifting generation (i.e., changing dispatch) from high- to low-emitting or zero-emitting units.
To evaluate the cost for some of these EGU NOx reduction strategies, the agency used the capital
expenses, fixed and variable operation and maintenance costs for installing and fully operating emission
controls based on the cost equations used within the Integrated Planning Model (IPM) that were
researched by Sargent & Lundy, a nationally recognized architect/engineering firm with the EGU sector
expertise.1 From this research, EPA has created a publicly available Excel-based tool called the Retrofit
Cost Analyzer (Update 1-26-2022) that implements these cost equations.2 Application of the Retrofit Cost
Analyzer equations to the existing EGU fleet can be found in the docket.3 EPA also used the Integrated
Planning Model (IPM) to analyze power sector response while accounting for electricity market dynamics
such as generation shifting.
The costs presented in the TSD are in 2016 dollars, unless otherwise noted. For some cost estimates, EPA
provides multiple statistics to describe the cost. These include: the "emission weighted average," which is
the total cost of the mitigation strategy applied to the applicable units divided by the total tons of NOx
reduced; the "median," which is the cost of the mitigation strategy at the median, or 50th percentile, unit,
and the "90th percentile", which is the cost of the mitigation strategy at the 90th percentile unit.
B. Optimizing and Restarting Existing SCRs
1. Cost Estimates for Optimizing and Restarting Existing SCRs
Coal Steam:
EPA examined costs for full operation of SCR controls for units that already have this technology
installed. SCR systems are post-combustion controls that reduce NOx emissions by reacting the NOx with
a reagent (typically ammonia or urea). The SCR technology utilizes a catalyst to increase the conversion
efficiency and produces high conversion of NOx. Over time with use, the catalyst will degrade and
require replacement. The ammonia or urea reagent is also consumed in the NOx conversion process.
Fully operating an SCR includes maintenance costs, labor, auxiliary power, catalyst, and reagent cost.
The chemical reagent (typically ammonia or urea) is a significant portion of the operating cost of these
1 The underlying equations come from data and information in the following reports:
• "IPM Model - Updates to Cost and Performance for APC Technologies: SCR Cost Development
Methodology for Coal-fired Boilers" (February 2022) ("Coal-Fired SCR Cost Methodology" for short)
• "IPM Model - Updates to Cost and Performance for APC Technologies: SNCR Cost Development
Methodology for Coal-fired Boilers" (August 2021) ("Coal-Fired SNCR Cost Methodology" for short)
• "IPM Model - Updates to Cost and Performance for APC Technologies: SCR Cost Development
Methodology for Oil/Gas-fired Boilers" (August 2021) ("Gas-Fired SCR Cost Methodology" for short)
• "IPM Model - Updates to Cost and Performance for APC Technologies: SNCR Cost Development
Methodology for Oil/Gas-fired Boilers" (August 2021) (" Gas-Fired SNCR Cost Methodology" for short)
• "Combustion Turbine NOX Technology Memo" (January 2022)
• "Typical SCR and SNCR Schedule (Coal or Oil/Gas boilers)" (February 2022)
2 See https://www.epa.gov/airmarkets/retrofit-cost-analvzer for the "Retrofit Cost Analyzer (Update 1-26-2022)"
Excel tool
3 See the file "NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx"
for detailed cost estimates using the Retrofit Cost Analyzer for SCR and SNCR operation and installation.
Page 3 of 40
-------
controls. For a unit with an idled, bypassed, or mothballed SCR, all fixed operating and maintenance
(FOM) and variable operations and maintenance (VOM) costs such as auxiliary fan power, catalyst costs,
and additional administrative costs (labor) are realized upon resuming operation through full potential
capability.
EPA examined the costs to fully operate an SCR that was already being operated to some extent and the
costs to restart and fully operate an SCR that had been idled using the equations within the Retrofit Cost
Analyzer. There are VOM costs related to the consumption of reagent and degradation of the catalyst as
well as FOM costs related to maintaining and operating the equipment to be considered that pertain to
these two situations.
EPA examined three of the VOM costs illustrated in the Retrofit Cost Analyzer: reagent, catalyst, and
auxiliary power. Depending on circumstances, SCR operators may operate the system while achieving
less than "full" removal efficiency by using less reagent, and/or not replacing degraded catalyst which
allows the SCR to perform at lower reduction capabilities. For units where the SCR has been idled, there
would be no reagent or catalyst utilized. Consequently, the EPA finds it reasonable to include the costs of
both additional reagent and catalyst maintenance and replacement in representing the cost of optimizing
operating SCR systems and also to include these costs for restarting idled systems. In contrast, based on
the Retrofit Cost Analyzer equations, the auxiliary power component of VOM is largely indifferent to the
NOx removal. That is, auxiliary power is indifferent to reagent consumption, catalyst degradation, or NOx
removal rate. Therefore, for units where the SCR is operating, but may not be fully operating, the
auxiliary power VOM component has likely been incurred. For units where the SCR has been idled, this
cost component needs to be accounted for when assessing the cost to restart and fully optimize the SCR
control.
In addition, based on the Retrofit Cost Analyzer equations for FOM, units running their SCR systems
have incurred the complete set of FOM costs, regardless of reagent consumption, catalyst degradation, or
NOx removal rate. Thus, as was the case for the auxiliary power VOM cost component, the FOM cost
component is also not included in the cost estimate to achieve "full" operation for units that are already
operating. For units where the SCR has been idled, all of the FOM costs would need to be accounted for
when assessing the cost to restart and fully optimize the SCR control. In conclusion, EPA finds that only
the VOM reagent and catalyst replacement costs should be included in cost estimates for optimization of
partially operating SCRs, while the full suite of VOM and FOM costs should be included when assessing
the cost to restart and fully optimize the SCR control. EPA assessed these costs using both representative
units as well as assessment of the existing fleet of EGUs with these controls.
In an SCR, the chemical reaction consumes approximately 0.57 tons of ammonia or 1 ton of urea reagent
for every ton of NOx removed. During development of the Clean Air Interstate Rule (CAIR) and the
original CSAPR, the agency identified a marginal cost of $500/ton of NOx removed (1999$) with
ammonia costing $190/ton of ammonia, which equated to $108/ton of NOx removed for the reagent
procurement portion of operations. The remaining balance reflected other operating costs. Over the
years, reagent commodity prices have changed, affecting the operational cost in relation to reagent
procurement. For data on the relationship between reagent price and its associated cost regarding NOx
reduction, see Appendix A: "Historical Anhydrous Ammonia and Urea Costs and their Associated Cost
per NOx ton Removed in a SCR." These commodities are created in large quantities for use in the
agriculture sector. Demand from the power sector for use in pollution controls is small relative to the
magnitude used in agriculture. Fluctuations in price are expected and are demonstrated in the pricing data
presented in Appendix A. Some of these prices reflect conditions where demand and commodity prices
are high. Consequently, the reagent costs used by EPA in this proposal are representative. In the cost
Page 4 of 40
-------
estimates presented here, EPA uses the cost for urea, which is greater than ammonia costs, to arrive at a
conservative estimate. In the CSAPR Update, EPA used the default cost of $310/ton for a 50% weight
solution of urea in the Retrofit Cost Analyzer. With the updates to the Retrofit Cost Analyzer, the default
costs of the urea reagent also increased to $350/ton for a 50% weight solution of urea. In this action,
considering the most recent updates to the Retrofit Cost Analyzer, EPA assumed the cost of $350/ton for
a 50% weight solution of urea. Using the Retrofit Cost Analyzer (multiplying the VOM $/ton cost by the
ratio of the VOM cost for urea $/MWh to the total VOM cost $/MWh) results in a cost of around
$500/ton of NOx removed for the reagent cost alone.
EPA also estimated the cost of catalyst replacement and disposal in addition to the costs of reagent. EPA
identified the cost for returning a partially operating SCR to full operation applying the Retrofit Cost
Analyzer equations for all SCR-controlled coal-fired units that operated in 2021 in the United States on a
per ton of NOx removed basis. EPA updated the set of units based on the latest version of the NEEDS
database (October 2021). This assessment covered 226 units.4 EPA focused on a subset of 172 units that
had minimum "input" NOx emission rates of at least 0.14 lb/MMBtu).5 Here, EPA defines the term
"input" NOx rate, or "uncontrolled" NOx rate to be the emission rate of the unit following the combustion
process including the effects of all existing combustion controls, measured after it has left the boiler, and
where it would enter any post-combustion control equipment (if any). EPA used 0.14 lb/MMBtu because
it found that, as described in the combustion control evaluation of this document (Figure C.3), that
depending on unit type and fuel use, units emitting above that rate may have only state of the art
combustion controls installed. Input NOx rates below the 0.14 lb/MMBtu rate suggest that the SCR may
have been operating to some extent, thereby biasing the cost estimates. EPA was able to identify the costs
of each of the individual VOM and FOM cost components, including reagent, catalyst, and auxiliary fans
and thereby, for each unit, to estimate the costs to fully optimize an SCR assuming it is currently
operating to some extent and to restart and fully optimize an SCR assuming it currently has been idled.
Some of these expenses, as modeled by the Retrofit Cost Analyzer, vary depending on factors such as unit
size, NOx generated from the combustion process, and reagent utilized. The EPA performed multiple
assessments with this tool's parameters to investigate sensitivity relating to cost per ton of NOx removed.
Additionally, the agency modeled costs with urea, the higher-cost reagent for NOx mitigation (and the
reagent included in the Retrofit Cost Analyzer equations). The key input parameters in the cost equations
are the size of the unit, the "input" NOx rate, the NOx removal efficiency, the type of coal, and the
capacity factor.
In the analysis, we assumed these units burned the coal identified for the given unit in the NEEDS
database at a 56% capacity factor.6 We assumed that the SCRs operate with the NOx removal efficiency
needed for them to achieve the lower of their 2021 ozone season NOx rate or the coal steam SCR
optimized ozone season rate starting from the highest monthly NOx rate for the time-period 2009-2021.
This was selected as the "controlled" rate because it represented consistent and efficient operation of the
4 See the "NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx.
5 A NOx emission rate at or above 0.2 lb/MMBtu, and possibly as low as 0.14 lb/MMBtu for some plant
configurations, may be indicative of emissions from units where the SCR is not operating at all (See the discussion
about state-of-the-art combustion controls).
6 EPA evaluated costs of SCR operation at coal steam units utilizing the fleet wide coal steam capacity factor value
projected in the 2030 run year of the IPM Summer 2021 Reference Case
Page 5 of 40
-------
unit's SCR. For the input NOx rate, to identify an emission rate when the unit's SCR was not operating,
we identified each unit's maximum monthly emission rate from the period 2009-2021. 7
To estimate the cost to return a partially operating SCR to full operation, EPA examined only the sum of
the VOM reagent and catalyst cost components and for these units, EPA ranked the quantified VOM costs
for each unit and identified the cost at the 90th percentile level rank, which rounded to $900/ton of NOx
removed. EPA selected the $900/ton value because a substantial portion of units had combined reagent
and catalyst costs at or less than this value of NOx removed.
To estimate the cost to restart and fully optimize an idled SCR, EPA ranked the sum of the VOM and
FOM costs for each unit and identified the 90th percentile cost. When rounded, this was $l,700/ton of
NOx removed. EPA also identified the emission weighted average cost, which also rounded to $900/ton
of NOx removed.
To further evaluate the costs for returning a partially operating control to full operation and for restarting
a unit with an idled SCR, EPA applied the Retrofit Cost Analyzer equations for two "typical" units with
varying input NOx rates in a bounding analysis. The EPA used this to identify reasonable high and low
per-ton NOx control costs from reactivating an existing but idled SCR across a range of potential input
NOx rates.8 For a hypothetical 500 MW unit with a relatively high input NOx rate (e.g., 0.4 lb
NOx/MMBtu, 80% removal efficiency, 56% capacity factor, and 10,000 Btu/kWh heat rate), The urea
and catalyst costs were around $660/ton. The full VOM and FOM costs associated with restarting an idled
control were around $950/ton of NOx removed. Conversely, a unit with a lower input NOx rate (e.g.,
0.14 lb NOx/MMBtu and 60% removal) experienced a higher cost range revealing a urea and catalyst cost
of $l,150/ton. The full VOM and FOM costs associated with restarting an idled control were around
$2,220/ton of NOx removed.
Considering the fleetwide assessment and the bounding analysis, for coal-steam units with SCR, EPA
concludes that $900/ton NOx removed represents a reasonable estimate of the cost for operating SCR
post-combustion controls on coal steam units that are already operating to some extent based on current
market prices and typical operation.
Considering the fleetwide assessment and the bounding analysis, for coal-steam units with SCR,
consistent with the cost level used in the RCU, the EPA concludes that a cost of $l,600/ton of NOx
removed is a representative cost for the point at which restarting and fully operating idled SCRs becomes
widely available to EGUs. EPA notes that the majority of units identified as having SCR optimization
potential are already partially operating and best reflected by the $900/ton optimization cost for partially
operating units rather than this $1,600/ ton cost for fully idled units.
Oil/Gas Steam:
For existing oil/gas steam units with existing SCR controls, EPA conducted a similar fleetwide
assessment to that done for coal-steam units using the Retrofit Cost Analyzer.9 We assumed these units
burned the fuel identified in the NEEDS database at a 26% capacity factor.10 We assumed that the SCRs
7 For units where controls have always operated year-round, this method will likely underestimate the input NOx
rate.
8 For these hypothetical cases, the "uncontrolled" NOx rate includes the effects of existing combustion controls
present (e.g., low NOx burners).
9 See the "NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx.
10 EPA evaluated costs of SCR operation at oil/gas steam units utilizing the fleet wide oil/gas steam capacity factor
value projected in the 2030 run year of the 1PM Summer 2021 Reference Case which was 26%.
Page 6 of 40
-------
operate with the NOx removal efficiency needed for them to achieve the lower of their 2021 ozone season
NOx rate or the oil/gas steam SCR optimized ozone season rate starting from the highest monthly NOx
rate for the time-period 2009-2021. This was selected as the "controlled" rate because it represented
consistent and efficient operation of the unit's SCR. For the input NOx rate, to identify an emission rate
when the unit's SCR was not operating, we identified each unit's maximum monthly emission rate from
the period 2009-2021.11
To estimate the cost to return a partially operating SCR to full operation we examined only the sum of the
VOM reagent and catalyst cost components, while to assess the costs to restart and fully optimize an idled
SCR the full VOM and FOM costs were assessed. In this section, from the full set of existing oil/gas
steam unit, we focused on a subset of 16 units (of 20 existing oil/gas steam units with SCR) that had
minimum "input" NOx emission rates of at least 0.14 lb/MMBtu. EPA ranked the quantified VOM costs
for each unit and identified the cost at the 90th percentile level rank, which rounded to $500/ton of NOx
removed.
To estimate the cost to restart and fully optimize an idled SCR, EPA ranked the sum of the VOM and
FOM costs for each unit and identified the 90th percentile cost. When rounded, this was $l,000/ton of
NOx removed. We also identified an emission weighted average cost of $700/ton. EPA concludes that
$500/ton is a representative cost for units to optimize an SCR for an oil/gas steam unit that is currently
operating its control to some extent while $700/ton NOx removed is a representative cost to restart an
idled SCR on oil/gas steam units based on current market prices and typical operation.
Combined Cycle and Combustion Turbine:
Considering SCR controls for other types of units (i.e., combined cycle and combustion turbines), EPA
notes that these units would have similar operational costs to oil/gas steam units because most of the
additional cost would be for reagent, which is directly proportional to the number of tons removed.
Consequently, the costs to fully optimize an SCR that is already operating to some extent and the cost to
restart and fully optimize an idled SCR would be comparable to that for oil/gas steam.
Summary:
Thus, considering the cost estimates for coal, oil/gas steam, combined cycle, and combustion turbines,
EPA concludes that $900/ton NOx removed represents a reasonable (and likely a conservative) estimate
that encompasses the cost for fully operating SCR controls that are current operating to some extent
across all affected EGU types based on current market prices and typical operation. Similarly, for coal-
steam, oil/gas steam units, combined cycle units, and combustion turbines the costs to restart and fully
operate existing idled SCRs is typically less than $l,600/ton. Thus, EPA conservatively assumes that all
of these technologies will fully operate at this cost level.
In summary, EPA assumes that $900/ton of NOx removed is a broadly available cost point for units that
currently are partially operating SCRs to fully operate their NOx controls while $l,600/ton of NOx
removed is a broadly available cost point for units to restart and fully operate existing idled SCR.
11 For units where controls have always operated year-round, this method will likely underestimate the input NOx
rate.
Page 7 of 40
-------
2. NOx Emission Rate Estimates for Full Operation of Existing SCRs
Coal Steam:
EPA examined the ozone season average NOx rates for 176 coal-fired units in the contiguous U.S. with
an installed SCR over the time-period 2009-2021, then identified each unit's lowest, second lowest, and
third-lowest ozone season average NOx rate.12 EPA updated the set of units based on the latest version of
the NEEDS database (October 2021). EPA examined ozone season average NOx rates from 2009
onwards as this year marks the point at which annual NOx programs, rather than just seasonal programs,
became widespread in the eastern US with the start of CAIR in 2009. EPA captured this dynamic with its
baseline choice as this regulatory development could affect SCR operation (specifically, annual use of
SCR means more frequent change of catalyst and relative difficulty with scheduling timing when the unit
(or just the SCR) is not operating to allow for catalyst replacement and SCR maintenance).
The CSAPR Update and Revised CSAPR Update focused on the third-lowest ozone season NOx rates
achieved since 2009, reasoning that these emission rates are characteristic of a well-run and well-
maintained system and achievable on a routine basis. At that time, 2019 represented the most recent year
of full ozone-season data available. In the Revised CSAPR Update, EPA found that, between 2009 and
2019, EGUs on average achieved a rate of 0.079 lb NOx/MMBtu for the third-lowest ozone season rate.
Furthermore, EPA verified that in years prior to 2019, more than 95% of these same coal-fired units with
identified optimization-based reduction potential in the Revised CSAPR Update had demonstrated and
achieved a NOx emission rate of 0.08 lb/MMBtu or less on a seasonal and/or monthly basis.13 In the
Revised CSAPR Update, EPA selected 0.08 lb NOx/MMBtu as a reasonable representation for full
operational capability of an SCR.
For this proposed rule, EPA utilizes the same rationale and methodology for identifying the rate that it did
with the Revised CSAPR Update. As in prior rules, EPA assigns this rate as a "ceiling" on those units
with SCR optimization potential, while generally assigning to SCR-equipped units emissions rates
commensurate with actual historical performance. In other words, if an SCR-equipped EGU demonstrated
achievement of an emission rate below 0.08 lb/MMBtu, that rate was assigned to the unit in the budget-
setting process. EPA maintains that the timeline should include most-recent operational data (i.e., up
through 2021) and continue to extend back to 2009. Considering the emissions data over the full time-
period of available data that includes expected annual operation of SCRs at coal steam units (i.e., 2009-
2021) results in an equal weighted unit-average third-best rate of 0.071 lb/MMBtu.14 EPA notes that half
of the EGUs achieved a rate of 0.064 lb NOx/MMBtu or less over their third-best entire ozone season (see
Figure B.l). EPA verified that in years prior to 2021, the majority (over 90%) of these same coal-fired
units with identified optimization-based reduction potential in this rule had demonstrated and achieved a
NOx emission rate of 0.08 lb/MMBtu or less on a seasonal and/or monthly basis.15
After identifying this approach, the Agency examined each ozone season over the time period from 2009-
2021 and identified the lowest monthly average NOx emission rates for each year. Examining the third-
lowest historical monthly NOx rate, the EPA found that, on average EGUs achieved a rate of 0.062 lb
NOx/MMBtu. The third-lowest historical monthly NOx rate analysis showed that a large proportion of
units displayed NOx rates below 0.08 lb/MMBtu (see Figure B.2).
12 See "SCR_Historical_OS_Rates_2015proposal.xlsx" for details.
13 See "Optimizing SCR Units with Best Historical NOx Rates Final" in the docket for this rulemaking
14 See "SCR_Historical_OS_Rates_2015proposal.xlsx" for details.
15 See "Optimizing SCR Units with Best Historical NOx Rates Updated to 2021" in the docket for this rulemaking
Page 8 of 40
-------
EPA also considered if the analysis should be specific to coal rank, among other factors. The average for
bituminous coal units increased to 0.074 lb/MMBtu, while the average rate for subbituminous units
decreased to 0.068 lb/MMBtu. These rates are relatively similar and therefore coal type is not an
important factor when choosing a widely achievable NOx rate for an optimized SCR on coal steam units.
Figure B.l. '"Frequency" distribution plots for coal-fired units with an SCR showing their seasonal
average NOx emission rates (lb/MMBtu) during ozone seasons from 2009-2021. For each unit, the
lowest, second lowest, and third lowest ozone season average NOx rates are illustrated.
Three Lowest Seasonal Average Ozone-Season NOx Rates (2009 -
2021, for coal steam units with SCR)
¦ lowest (0.060 Ibs/MMBtu avg)
¦ 2nd lowest (0.065 Ibs/MMBtu avg)
¦ 3rd lowest (0.071 Ibs/MMBtu avg)
J?
¦A
(p A o8> -N ^
& & & <3° <$ &
Or Ov Qv 0) Qv Ov Q
.9
' O-
NOx Rate (Ibs/MMBtu)
Figure B.2. "Frequency" distribution plots for coal-fired units with an SCR showing their monthly
average NOx emission rates (lb/MMBtu) during ozone seasons from 2009-2021. For each unit, the
lowest, second lowest, third lowest, fourth lowest, and fifth lowest monthly average NOx rates are
illustrated.
70
60
50
40
30
20
10
0
Five Lowest Monthly Average Ozone-Season NOx Rates (2009 - 2021,
for coal steam units with SCR)
¦ ¦ L LlJlllll JlJ u
lowest (0.048 Ibs/MMBtu avg)
12nd lowest (0.057 Ibs/MMBtu avg)
i 3rd lowest (0.062 Ibs/MMBtu avg)
4th lowest (0.066 Ibs/MMBtu avg)
15th lowest (0.071 Ibs/MMBtu avg)
O Ov 0 Ov Ov ^ O- o- O' o- .o
*P xP *P *P Sp Kp *P . *P i^P O, sP > 0>
o ov ov ov ov ov q- ov ^ ^ o- o- o>- o- o- o o- cy ^
NOx Rate (Ibs/MMBtu)
Page 9 of 40
-------
Not only has the group of units with SCR optimization potential demonstrated they can perform at or
better than the 0.08 lb/MMBtu rate on average, but more than 90 percent of the individual units in this
group have met this rate on a seasonal and/or monthly basis based on their reported historical data. As an
example, Miami Fort Unit 7 was able to improve the performance of the SCR at its coal unit (consistent
with optimization assumptions) in the first year of CSAPR Update implementation (2017) and again in
the first period of the Revised CSAPR Update implementation (2021) when the updated budgets created
an incentive through higher allowance prices, (see Figure B.3)
Figure B.3. Example of Unit-level Emissions Rate Changes at a Given Capacity Factor Range.
Miami Fort Unit 7 Ozone Season Mid- to High-Load NOx Rates
00
1 0.6
£
V 0.4-
ro
cr
ox 0 2"
0
1
capacity factor 50-59%
capacity factor 60-69%
2017
2019
2021
2017
2019
capacity factor 70-79%
i+i
2021
2017
2019
2021
co 400 -
=5
o
X 300"
O)
¦£ 200-
CD
o. 100-
O
o-
EPA observed this pattern in other units identified in this rulemaking as having significant SCR
optimization emissions reduction potential. For instance, nearly 90% of the coal units identified as having
reduction potential from its existing SCR in 2021 had operated at an emission rate lower than the 0.08
lb/MMBtu optimized rate in prior years. In the Emissions Data TSD for the supplemental notice that EPA
recently released in a proceeding to address a recommendation submitted to EPA by the Ozone Transport
Commission under CAA section 184(c), EPA noted, "In their years with the lowest average ozone season
NOx emissions rates in this analysis, these EGUs had relatively low NOx emissions rates at mid- and
high-operating levels; moreover, there was little variability in NOx emissions rates at these operating
levels. However, during the 2019 ozone season, these EGUs had higher NOx emissions rates and greater
variability in NOx emissions rates across operating levels than in the past, particularly at mid-operating
levels/'16"
Finally, some stakeholders have suggested that maintaining emissions rates characteristic of optimized
SCR performance is difficult in an environment where coal units are dispatching less. EPA's review of
hourly data indicates that maintaining consistent SCR performance at lower capacity factors is possible.
For example, the unit-level performance data in Figure B.4 show the emissions rate at a plant, Brandon
16 See "Analysis of Ozone Season NOx Emissions Data for Coal-Fired EGUs in Four Mid-Atlantic States" (Docket
#: EPA-HQ-OAR-2020-0351) available at https://www.regulations.gov/document/EPA-HO-OAR-2020-0351-00Q4
Page 10 of 40
-------
Shores, staying relatively low (consistent with our optimization assumption of 0.08 lb/MMBtu) and stable
across a wide range of capacity factors. The use of the fleetwide average rate computed using each unit's
third best year to derive the SCR optimization rates further accommodates any heterogeneity in emissions
rate that may stem from a unit's operational decisions such as capacity factor or coal-rank choice.
Figure B.4. Example of Consistently Low Unit-level Emissions Rate During Periods of Varying Capacity
Factor
Brandon Shores
100%
ce
p
<
<
<
80%
60%
40%
20%
0%
0.35
0.30
0.25
0.20
LU
!§
0.15
X
O
2
0.10
0.05
0.00
14 7 101316192225283134374043464952555861646770
HOUR
-Average Capacity Factor
-Average NOx Rate
Based on all of the factors above, including the seasonal and monthly findings in Figures B.l and B.2, the
agency concludes an emission rate of 0.08 lb NOx/MMBtu is widely achievable by the portion of the
coal-fired EGU fleet with SCR optimization potential identified.
Oil/Gas Steam:
EPA conducted a similar analysis for oil/gas steam units. EPA examined the ozone season average NOx
rates for 23 oil/gas steam units in the contiguous U.S. with an installed SCR over the time-period 2009-
2021 and calculated an equal weighted unit-average third-best rate of 0.03 lb/MMBtu.17 EPA notes that
half of these EGUs achieved a rate of 0.014 lb NOx/MMBtu or less over their third-best entire ozone
season (see Figure B.5). EPA verified that in years prior to 2021, the majority (approximately 60%) of
these same oil/gas steam units with identified optimization-based reduction potential in 2021 data had
demonstrated and achieved a NOx emission rate of 0.03 lb/MMBtu or less on a seasonal and/or monthly
basis; all such units had achieved a NOx emission rate of 0.05 lb/MMBtu or less on a seasonal and/or
monthly basis. Based on this analysis, the agency concludes an emission rate of 0.03 lb NOx/MMBtu is
widely achievable by the portion of oil/gas steam EGU fleet with SCR optimization potential identified.
17 See "SCR_Historical_OS_Rates_2015proposal.xlsx" for details.
Page 11 of 40
-------
Figure B.5. Frequency distribution plots for oil/gas steam units with an SCR showing their seasonal
average NOx emission rates (lb/MMBtu) during ozone seasons from 2009-2021. For each unit, the
lowest, second lowest, and third lowest ozone season average NOx rates are illustrated.
10
9
8
7
6
5
4
3
2
1
0
Three Lowest Seasonal Average Ozone-Season NOx Rates (2009 -
2021, for oil/gas steam units with SCR)
¦ lowest (0.020 Ibs/MMBtu avg)
¦ 2nd lowest (0.023 Ibs/MMBtu avg)
¦ 3rd lowest (0.026 Ibs/MMBtu avg)
Ov
o «o .0 «p *o ,o .0
^ rsV r-So
^ rS> rS>
o- o- Q'
cr cP cr
o- c>- o-
NOx Rate (Ibs/MMBtu)
^ ^
Combined Cycle:
EPA conducted a similar analysis for combined cycle units. EPA examined the ozone season average
NOx rates for 853 combined cycle units in the contiguous U.S. with an installed SCR over the time-period
2009-2021 and calculated an equal weighted unit-average third-best rate of 0.012 lb/MMBtu.18 EPA
notes that half of the combined cycle units achieved a rate of 0.009 lb NOx/MMBtu or less over their
third-best entire ozone season (see Figure B.6). EPA verified that in years prior to 2021, the majority
(approximately 50%) of these same combined cycle units with identified optimization-based reduction
potential in 2021 data had demonstrated and achieved a NOx emission rate of 0.012 lb/MMBtu or less on
a seasonal and/or monthly basis; all such units had achieved a NOx emission rate of 0.066 lb/MMBtu or
less on a seasonal and/or monthly basis. Based on this analysis, the agency concludes an emission rate of
0.012 lb NOx/MMBtu is widely achievable by the portion of combined cycle EGU fleet with SCR
optimization potential identified.
18 See "SCR_Historical_OS_Rates_2015proposal.xlsx" for details.
Page 12 of 40
-------
Figure B.6. Frequency distribution plots for combined cycle units with an SCR showing their seasonal
average NOx emission rates (lb/MMBtu) during ozone seasons from 2009-2021. For each unit, the
lowest, second lowest, and third lowest ozone season average NOx rates are illustrated.
Three Lowest Seasonal Average Ozone-Season NOx Rates (2009 -
2021, for combined cycle units with SCR)
140
120
100
80
60
40
20
0
lowest (0.011 Ibs/MMBtu avg)
I 2nd lowest (0.011 Ibs/MMBtu avg)
I 3rd lowest (0.012 Ibs/MMBtu avg)
cO
iA>
o>
&
o-
<3>
&
<£>
er
c>-
o-
&
O' o- o-
NOx Rate (Ibs/MMBtu)
<
-------
Figure B.7. Frequency distribution plots for combustion turbine units with an SCR showing their
seasonal average NOx emission rates (lb/MMBtu) during ozone seasons from 2009-2021. For each unit,
the lowest, second lowest, and third lowest ozone season average NOx rates are illustrated.
Three Lowest Seasonal Average Ozone-Season NOx Rates (2009 -
2021, for combustion turbine units with SCR)
lowest (0.016 Ibs/MMBtu avg)
¦ 2nd lowest (0.018 Ibs/MMBtu avg)
¦ 3rd lowest (0.024 Ibs/MMBtu avg}
fir?
& O?" ^ /
-------
Figure C.l. Ozone Season NOx Rate (lb/MMBtu) Over Time for Coal-fired Units with Various
Combustion Controls*
NOx Control
For 2003 - 2008
For 2003 - 2008
For 2009-2020
For 209-2020
For 2021
For 2021
Technology
NOx Rate
Number of
NOx Rate
Number of
NOx Rate
Number of
(lb/MMBtu)
U nit-Years
(lb/MMBtu)
Unit-Years
(lb/MMBtu)
Unit-Years
Overfire Air
0.384
47h
0.291
647
0.260
17
Low NOx Burner
Technology (Dry
Bottom only)
0.351
1,062
0.266
1,208
0.217
32
Low NOx Burner
Technology w/
Overfire Air
0.306
464
0.225
742
0.203
29
Low NOx Burner
Technology w/
Closed-coupled OFA
0.266
341
0.219
361
0.162
14
Low NOx Burner
Technology w/
Separated OFA
0.222
451
0.187
635
0.151
24
Low NOx Burner
Technology w/
Closed-
0.207
460
0.166
886
0.138
52
coupled/Separated
OFA
* Source: Air Markets Program Data (AMPD), ampd.epa.gov, EPA, 2021
Current combustion control technology reduces NOx formation through a suite of technologies. Whereas
earlier generations of combustion controls focused primarily on either Low NOx Burners (LNB) or
Overfire Air (OFA), modern controls employ both, and sometimes include a second, separated overfire
air system. Further advancements in fine-tuning the burners and overfire air system(s) as a complete
assembly have enabled suppliers to obtain better results than tuning individual components. For this
proposed rule, the agency evaluated EGU NOx reduction potential based on upgrading units to modern
combustion controls. Combustion control upgrade paths are shown in Table 3-14 of the IPM version 6
Summer 2021 Reference Case documentation (see Chapters 3 and 5 of the IPM documentation for
additional information, and Table 2b below).21 The fully upgraded configuration for units with wall-fired
boilers is LNB with OFA. For units with tangential-fired boilers, the fully upgraded configuration is
LNC3 (Low NOx burners with Close-Coupled and Separated Overfire Air). For each unit, EPA's
understanding of the current NOx control configuration can be found in the "NOx Comb Control" column
of the NEEDS v6 () database file.22 EPA identified whether a unit has combustion control upgrade
potential by comparing the Mode 1 NOx Rate (lb/MMBtu) with the Mode 3 NOx Rate (lb/MMBtu)
within NEEDS. If the Mode 3 value is lower than the Mode 1 value, then the unit's combustion control
configuration does not match the state-of-the-art configuration outlined in Figure C.2. For these units,
EPA assumed a combustion control upgrade is possible based on the technology configurations identified
in the NOx post combustion control column.
21 https://www.epa.gov/airmarkets/documentation-epas-power-sector-modeling-platform-v6-summer-2021-
reference-case
22 See the NEEDS v.6 data file available in the docket and for download at https://www.epa.gov/airmarkets/national-
electric~energy-data~svstem~needs~v6
Page 15 of 40
-------
With the wide range of LNB configurations and furnace types present in the fleet, the EPA decided to
assess compliance costs based on an illustrative unit.23 The agency selected this illustrative unit because
its attributes (e.g., size, input NOx emission rate) are representative of the EGU fleet, and, thus, the cost
estimates are also representative of the EGU fleet. The EPA estimated costs for various combustion
control paths. The cost estimates utilized the equations found in Table 5-4 "Cost (2016$) ofNOx
Combustion Controls for Coal Boilers (300 MW Size)" from Chapter 5 of the IPM 5.13 documentation.24
For these paths, EPA found that the cost ranges from $450 to $l,220/ton NOx removed ($2011). EPA
examined slightly lower capacity factors (i.e., 70%) and found the costs increased, to a range of $550 to
$l,460/ton. At lower capacity factors (i.e., 56%), costs increased to a maximum of $l,800/ton for one
type of installation. Examining the estimates for all the simulations, the agency finds that the costs of
combustion control upgrades for units operating in a baseload fashion are typically comparable to the
costs for returning a unit with an inactive SCR to full operation (i.e., $l,600/ton). Consequently, EPA
identifies $l,600/ton as the cost level where upgrades of combustion controls would be widely available.
2. NOx Emission Rate Estimates for State-of-the-Art Combustion Control Upgrade
EPA derived its performance rate assumptions for combustion control upgrade from an assessment of
historical data where EPA reviewed similar boiler configurations with fully upgraded combustion controls
and their resulting emission rates. Specifically, EPA examined two types of coal steam units: 1) dry-
bottom wall-fired boilers and 2) tangentially-fired boilers. EPA looked at the current rate of existing units
of each firing configuration and that already had state-of-the-art combustion controls (SOA CC). EPA
estimated the average 2021 ozone season NOx emission rates for all such units by firing type. EPA did
not include any units that had post-combustion controls installed as their historical rate would be
indicative of not just combustion control potential, but also post-combustion control potential. For dry
bottom wall-fired coal boilers with "Low NOx Burner" and "Overfire," there were 34 units averaging
0.204 lb/MMBtu.25 For tangentially-fired coal boilers with "Low NOx Burner" and "Closed-
coupled/Separated OFA," there were 48 units averaging 0.138 lb/MMBtu.
Next, EPA identified the current boiler type for each unit in the fleet. It then applied the information
shown below in Figure C.2 regarding state-of-the-art configurations compared to that unit's reported
combustion control figuration to determine whether the unit had combustion control upgrade potential.
Starting with dry-bottom wall-fired boilers, EPA verified that the average performance rate identified
above for this boiler configuration with state-of-the-art combustion controls 1) resulted in reductions
consistent with the technology's assumed percent-reduction potential when applied to this subset of units,
23 For this analysis, EPA assumed a 500 MW unit with a heat rate of 10,000 Btu/kWh and an 85% annual capacity
factor. We assumed the unit was burning bituminous coal and had a NOx rate of 0.50 lb NOx / MMBtu initial rate
based on its existing NOx combustion controls and had a 42% NOx removal efficiency after the combustion control
upgrades. This 0.50 lb/MMBtu input NOx rate is comparable to the observed average rate of 0.48 lb/MMBtu for the
coal-fired wall-fired units from 2003-2008 that had not installed controls. There are very few remaining units that
lack combustion controls. One unit had a rate higher than 0.5 lb/MMBtu. Using 2019 data for wall-fired coal units
lacking combustion controls and comparing these rates against controlled units of the same type, EPA observes a
42% difference in rate. Similarly, EPA observes a 55% reduction for coal units with tangentially-fired boilers.
Despite the very small numbers of remaining units that lack combustion controls, to be conservative, EPA used the
42% reduction from wall-fired coal units.
24 Costs were converted to 2016$ from the tables original $2011$ values.
https://www.epa.gOv/sites/production/files/2015-07/documents/chapter_5_emission_control_technologies_0.pdf
25 In past rulemakings, EPA found the representative rate for state-of-the-art combustion controls to be 0.199
lb/MMBtu. Thus, this analysis re-affirms EPA's selection of 0.199 lb/MMBtu as the representative rate for state-of-
the-art combustion controls for most unit types.
Page 16 of 40
-------
and/or 2) had been demonstrated by both subbituminous and bituminous coal units with state-of-the-art
combustion controls in its 2019 dataset. EPA found an assumed emission rate of 0.199 lb/MMBtu to be
reasonable for dry-bottom, wall-fired boilers upgrading to state-of-the-art combustion controls.
EPA applied the same approach for tangentially-fired coal boilers. As shown in Figure C.3, this produced
a weighted average assumed emission rate of 0.147 lb/MMBtu. However, EPA observed that in 2019 no
bituminous burning units with this boiler type (and no post-combustion controls) had met the 0.147
lb/MMBtu average for this control configuration (which was heavily weighted by subbituminous units
with such combustion controls) (see Figure C.3). It also noted that the 0.147 rate lb/MMBtu would imply
a greater percent reduction for some bituminous units with upgrade potential than EPA identified as
representative for the technology. Therefore, given these two findings and the bituminous orientation of
the fleet with state-of-the-art combustion control upgrade potential covered in this action, EPA
determined that the 0.199 lb/MMBtu was also appropriate for tangentially-fired units in this action as that
rate satisfied both criteria.26
Figure C.2. State-of-the-Art Combustion Control Configurations by Boiler Type
Boiler Type
Existing NOx Combustion Control
Incremental Combustion Control
Necessary To Achieve "State-of-the-
Art"
Tangential Firing
Docs not include LNC1 and LNC2
LNC3
Tangential Firing
Tangential Firing
Tangential Firing
Includes LNC 1. but not LNC2
Includes LNC2. but not LNC3
Includes LNC1 and LNC2 or LNC3
Conversion from LNC2 to LNC3
Conversion from LNC1 to LNC3
Wall Firing, Dry Bottom
Docs not Include LNB and OFA
LNB + OFA
Wall Firing, Dry Bottom
Includes LNB. but not OFA
OFA
Wall Firing, Dry Bottom
Includes OFA, but not LNB
LNB
Wall Firing, Dry Bottom
Includes both LNB and OFA
-
Note: Low LNB =NOx Burner Technology, LNCl=Low NOx coal-and-air nozzles with close-coupled overfire air, LNC2= Low NOx Coal-and-
Air Nozzles with Separated Overfire Air, LNC3 = Low NOx Coal-and-Air Nozzles with Close-Coupled and Separated Overfire Air, OFA =
Overfire Air
Figure C.3. 2019 average NOX rate for units with state-of-the-art combustion controls for
tangentially-fired boilers (lb/MMBtu)
Coal Type
NOx Rate (lb/MMBtu)
Bituminous
0.199
Bituminous, Subbituminous
0.172
Subbituminous
0.134
Weighted Average
0.147
26 See the entry titled "State-of-the-art Combustion Control Data" in the docket for this rulemaking
Page 17 of 40
-------
D. Cost Estimates for Optimizing and Restarting and Optimizing Idled Existing SNCR
Coal Steam:
EPA sought to examine costs for full operation of SNCR. SNCR are post-combustion controls that
reduce NOx emissions by reacting the NOx with either ammonia or urea, without catalyst. Because the
reaction occurs without catalyst and is thereby a less efficient reaction, several times more reagent must
be injected to achieve a given level of NOx removal with SNCR than would be required to achieve the
same level of NOx removal with SCR technology. Usually, an SNCR system does not achieve the level of
emission reductions that an SCR can achieve, even when using large amounts of reagent. For the SNCR
analysis, as with the SCR analyses described above, the agency used the Retrofit Cost Analyzer equations
to perform a bounding analysis for examining operating expenses associated with a "generic" coal steam
unit returning an SNCR to full operation. EPA examined the costs to fully operate an SNCR that was
already being operated to some extent and the costs to restart and fully operate an SNCR that had been
idled.
Comparable to the discussion about fully optimizing an SCR, the cost to fully optimize an SNCR that is
currently operating to some extent is the VOM costs for additional reagent (urea and steam). Reagent
consumption represents the largest portion of the VOM cost component of unit operation and is directly
related to NOx removal. The other VOM components (e.g., to auxiliary power) as well as the FOM
components are assumed to be largely indifferent to additional NOx removal. For units with an idled, or
mothballed, SNCR returning to full operation, the owner incurs the full suite of VOM and FOM costs.
For this bounding analysis, the agency examined two cases: first, a coal steam unit with a high input NOx
rate 0.40 lb/MMBtu; second, a coal steam unit with a lower input NOx rate 0.20 lb /MMBtu - both
assuming a 25% removal efficiency for NOx.27'28 For the high-rate unit case, VOM and FOM costs were
calculated as approximately $2,100/ton NOx with about $l,600/ton of that cost associated with urea use.
For the low-rate unit case, VOM and FOM costs approached $3,500/ton NOx with nearly $2,700/ton of
that cost associated with urea procurement. Despite equivalent reduction percentages for each unit, the
cost dichotomy results from differences in the input NOx rates for the units and the type of boiler,
resulting in a modeled step-change difference in urea rate (either a 15% or 25% reagent usage factor).
EPA also examined SNCR cost sensitivity by varying NOx removal efficiency while maintaining the
input NOx emission rate. In this analysis, SNCR NOx removal efficiency was assumed to be 40% for the
first cost estimate and 10% for the second cost estimate. For a high-rate unit (with an input rate of 0.40 lb
NOx/MMBtu), the associated costs were $2,000/ton and $2,450/ton. For a low-rate unit (i.e., an input
rate of 0.20 lb NOx/MMBtu), the associated costs were $3,370/ton and $4,180/ton. These analyses
together illustrate that SNCR costs on a dollar per ton basis are more sensitive to a unit's input NOx rate
than the potential NOx removal efficiency of the SNCR itself.
27 For both cases, we examined a 500 MW unit with a heat rate of 10,000 Btu/kWh operated at a 26% annual
capacity factor while burning bituminous coal. The capacity factor used was based on the IPM-estimated operation
of units with SNCR. Furthermore, in the cost assessment performed here, the agency assumes SNCR NOx removal
efficiency to be 25%, which is consistent with the range of documented removal efficiency in IPM and historic data.
See the Retrofit Cost Analyzer (Update 1-26-2022) tool and the "Coal-Fired SNCR Cost Methodology."
28 Steam; Its Generation and Use 41ed (2005) J B Kitto; S C Stultz; Babcock & Wilcox Company, lists 20-30%
conversion of NOx as typical, with up to 50% possible under certain circumstances.
Page 18 of 40
-------
EPA conducted a fleetwide assessment of coal-fired units with existing SNCR (70 units total) using the
same set of assumptions used for the fleetwide assessment for units with SCR (i.e., the method for
determining input NOx rate, NOx removal efficiency, and capacity factor) using the Retrofit Cost
Analyzer.29 In that tool, we can examine the VOM and FOM cost components associated with operation
of the unit to approximate costs of fully optimizing a currently operating control and to restart and fully
optimize an idled control. Of the full set of units, 47 met the size and operating criteria set forth for SCR
units (i.e., 100 MW in size or great, 0.14 lb/MMBtu input NOx rate or greater, and at least some emission
reduction). EPA found that the median cost to fully optimize a currently operating control is $l,600/ton
while the cost to restart and fully optimize an idled control has a median cost of $3,500/ton and an
emission weighted cost of $2,100/ton.
In the Revised CSAPR Update, EPA concluded that $l,800/ton was a representative cost for units that are
currently operating their SNCR controls to some extent to fully operate those controls and $3,900/ton was
a representative cost for units that had idled their SNCR controls. Based on the fleetwide analysis, the
bounding analysis, and the Revised CSAPR Update analysis, EPA concludes that $l,800/ton is a
representative cost to fully operate SNCR for units that are currently operating an SNCR to some extent
and that a $3,900/ton cost is representative to turn on and fully operate controls for units whose SNCR
control has been idled.
Finally, EPA assessed the existing fleet of units with SNCR technology to assess whether, or not, the
units were typically operating their controls to some extent or whether the controls had been fully idled to
assess whether the cost breakpoint level should be identified as $l,800/ton or $3,900/ton. EPA examined
the fleet of coal-steam and oil/gas steam units that have SNCR optimization potential within the
geography and assessed whether the units are currently partially operating but not necessarily optimizing
their SNCRs or whether the units have idled their controls. EPA assessed whether SNCRs were partially
operating or were idled doing a two-part analysis. First, EPA compared each unit's 2021 reported rate to
this proposal's SNCR optimization emission rate for that unit. A difference less than 25% between
reported and optimized NOx rate for a given EGU is an indicator that its SNCR is already partially
operating as of 2021 (as 25% NOx removal is a representative average of what SNCRs may achieve when
going from no operation to full operation).30 To identify the optimized value for each unit and compare
that to 2021 baseline emission rates, EPA utilized the mode 2 rate from the NEEDS database (October
2021). As described in EPA's power sector IPM Modeling Documentation (Chapter 3), these unit-specific
NOx mode rates are calculated from historical data and reflect operation of existing post-combustion
controls.31 Four modes are identified for each unit to, among other things, identify their emission rates
with and without their post-combustion controls operating. Mode 2 for SNCR-controlled units is
intended to reflect the operation of that unit's post combustion control based on prior years when that unit
operated its control. The optimized SNCR emission rates assumed for each controlled unit are identifiable
in the NEEDS file "Mode 2 NOx rate (lb/MMBtu)" column.32 If a unit's 2021 emission rate was at or
lower than its "optimized" SNCR rate, than no additional reductions are expected from "optimizing" that
unit's post-combustion control.
29 See "NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx."
30 "Coal-Fired SNCR Cost Methodology"
31 https://www.epa.gov/system/files/documents/2021-09/chapter-3-power-system-operation-assumptions.pdf
32 See the NEEDS v.6 data file available in the docket and for download at https://www.epa.gov/ainiiarkets/national-
electric~energy-data~svstem~needs~v6
Page 19 of 40
-------
Comparing each unit's 2021 reported emissions rate to this proposal's SNCR optimization emission rate
(see Figure D. 1), the majority of affected EGUs with existing SNCR (44 of 49 units) have percent
differences less than 25%, suggesting that their SNCRs appeared to be partially operating at least to some
extent based on the first indicator alone. For the remaining 5 units, EPA compared the historical highest
rate for each unit (dating back to 2009) to its 2021 reported emission rate. For two of the units, the
reported 2021 rate was substantially lower than the unit's historical highest rate, so EPA assumes that the
SNCR at these units were also operating to some extent in 2021. For the remaining units, either the SNCR
has consistently been fully operating throughout the life of the unit (i.e., the time period of this analysis)
or the control may have been idled. Between the two indicators, EPA determined that nearly all SNCRs
with optimization potential were at least partially operating their controls during 2021. Consequently,
EPA concludes that a VOM reagent-centric cost of $l,800/ton is a reasonable representative cost for
emissions reductions for fully operating SNCR for units with this existing control.
Page 20 of 40
-------
Figure D.l. Ozone-season NOx Emission Rates (lb/MMBtu) for Units with SNCR Reduction Potential
Facility
State
ORIS
Boiler
2021 OS
NOx Rate
(lb/MMBtu)
SNCR
Optimization
OS Rate
(lb/MMBtu)
Percent Change
in Rate (2021 vs
Optimized Rate)
Historical
Max OS
NOx Rate
(2009-2021)
Percent Change
Between 2021
and Historical
Max
Yorktown Power Station
Virginia
3809
3
0.13
0.02
86%
0.25
46%
Grant Town Power Plant
West Virginia
10151
IB
0.32
0.16
51%
0.35
7%
Grant Town Power Plant
West Virginia
10151
1A
0.33
0.16
50%
0.35
7%
Manitowoc
Wisconsin
4125
9
0.08
0.05
43%
0.09
10%
Brame Energy Center
Louisiana
6190
3-2
0.04
0.03
32%
0.07
38%
Seward
Pennsylvania
3130
2
0.12
0.09
24%
Powerton
Illinois
879
61
0.11
0.08
20%
Edge Moor
Delaware
593
4
0.05
0.04
20%
Marion
Illinois
976
123
0.10
0.08
20%
Brame Energy Center
Louisiana
6190
3-1
0.04
0.03
18%
Powerton
Illinois
879
62
0.10
0.08
17%
Sikeston
Missouri
6768
1
0.12
0.10
14%
Whitewater Valley
Indiana
1040
1
0.32
0.28
13%
Whitewater Valley
Indiana
1040
2
0.34
0.30
12%
Grayling Generating Station
Michigan
10822
1
0.13
0.12
12%
Twin Oaks
Texas
7030
U1
0.10
0.09
12%
IPL - Harding Street Station
(EW Stout)
Indiana
990
60
0.04
0.04
10%
Limestone
Texas
298
LIM2
0.18
0.16
9%
Sioux
Missouri
2107
1
0.25
0.23
9%
Powerton
Illinois
879
51
0.11
0.10
9%
Spruance Genco, LLC
Virginia
54081
BLR04A
0.03
0.03
ox
00
Joliet 29
Illinois
384
82
0.10
0.09
ox
00
Spruance Genco, LLC
Virginia
54081
BLR03A
0.03
0.03
7%
Joliet 29
Illinois
384
81
0.10
0.09
7%
Joliet 29
Illinois
384
71
0.08
0.08
7%
Boswell Energy Center
Minnesota
1893
4
0.11
0.10
6%
Twin Oaks
Texas
7030
U2
0.10
0.09
6%
Spruance Genco, LLC
Virginia
54081
BLR04B
0.03
0.03
6%
Altavista Power Station
Virginia
10773
2
0.12
0.12
6%
Altavista Power Station
Virginia
10773
1
0.12
0.12
6%
Joliet 29
Illinois
384
72
0.08
0.08
5%
Southampton Power Station
Virginia
10774
1
0.13
0.12
4%
Southampton Power Station
Virginia
10774
2
0.13
0.12
4%
Barry
Alabama
3
4
0.27
0.26
4%
San Miguel
Texas
6183
SM-1
0.16
0.16
3%
Fort Martin Power Station
West Virginia
3943
1
0.29
0.28
3%
Hopewell Power Station
Virginia
10771
1
0.12
0.12
3%
Hopewell Power Station
Virginia
10771
2
0.12
0.12
3%
AES Warrior Run
Maryland
10678
001
0.07
0.07
3%
Seward
Pennsylvania
3130
1
0.11
0.11
3%
Nacogdoches Generating
Facility
Texas
55708
BFB-1
0.08
0.08
3%
Sioux
Missouri
2107
2
0.23
0.22
3%
New Castle
Pennsylvania
3138
4
0.07
0.07
1%
Laramie River
Wyoming
6204
2
0.14
0.14
1%
Edge Moor
Delaware
593
3
0.06
0.06
1%
Clinch River
Virginia
3775
2
0.13
0.13
1%
Big Cajun 2
Louisiana
6055
2B3
0.12
0.12
0%
New Castle
Pennsylvania
3138
3
0.07
0.07
0%
Rothschild Biomass
Cogeneration Facility
Wisconsin
58124
1
0.08
0.08
0%
Page 21 of 40
-------
Oil/Gas Steam:
Using the Retrofit Cost Analyzer, the agency conducted a bounding analysis to assess the costs for
optimizing Oil/Gas steam units that are not fully operating their controls and/or restarting and optimizing
idled SNCR controls. Again, assuming a 25% reduction in emission rate and for a unit similar to that
examined above for coal steam units (500 MW, tangential boiler, input NOx rate of 0.2 lb NOx/MMBtu,
10,000 Btu/kWh heat rate, with a 25% NOx removal efficiency) using natural gas with a 26% capacity
factor, the VOM and FOM costs were calculated as approximately $3,500/ton NOx with about $2,700/ton
of that cost associated with urea use. These costs are not particularly sensitive to the capacity factor of
the unit, with the urea-use cost insensitive to unit usage. The costs were also largely indifferent to the
selection of gas compared with oil. For the same unit, but with a lower input NOx rate of 0.14 lb/MMBtu,
the VOM+FOM costs increased to $3,600/ton while the urea cost remained at $2,700/ton. For the same
unit, but with a higher input NOx rate of 0.4 lb/MMBtu, the VOM+FOM costs decreased to $2,000/ton
while the urea cost dropped to $l,600/ton.
EPA also conducted a similar fleetwide analysis for oil/gas steam units with existing SNCR, which
included 22 units, 21 of which met the size and operating criteria (i.e., 100 MW in size or great, 0.14
lb/MMBtu input NOx rate or greater, and at least some emission reduction).33 EPA found that the median
cost to fully optimize a currently operating control is $l,600/ton while the cost to restart and fully
optimize an idled control has a median cost of $l,900/ton and an emission weighted cost of $2,000/ton.
Based on the fleetwide analysis and the bounding analysis, EPA estimates that $l,600/ton is a
representative cost to fully operate SNCR for oil/gas steam units that are currently operating an SNCR to
some extent and that a $3,500/ton cost is representative to turn on and fully operate SNCR for oil/gas
steam units whose control has been idled.
Summary:
EPA finds that $l,800/ton is a representative cost to fully operate SNCR for both coal and oil/gas steam
units that are currently operating an SNCR to some extent and that a $3,900/ton cost is representative to
turn on and fully operate controls for SNCR units whose controls have been idled.
E. Cost and Emission Rate Performance Estimates for Retrofitting with SNCR and
Related Costs
SNCR technology is an alternative method of NOx emission control that incurs a lower capital cost
compared with an SCR, albeit at the expense of greater operating costs and less NOx emission reduction.
Some units with anticipated shorter operational lives or with low utilization may benefit from this control
technology. The higher cost per ton of NOx removed reflects this technology's lower removal efficiency
which necessitates greater reagent consumption, thereby escalating VOM costs. The agency examined
the costs of retrofitting coal steam and oil/gas steam units with SNCR technology using the Retrofit Cost
Analyzer. The agency did not consider retrofitting SNCR on combustion turbine or combined cycle units
as, except for a single combustion turbine at the Delaware City Plant, there are no EGUs of these plant
types in NEEDS equipped with SNCR.
33 See the "NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx.
Page 22 of 40
-------
Coal Steam:
For SNCR retrofits on coal steam units, the agency conservatively set the NOx emission reduction rate at
25% - the same assumption used in the Revised CSAPR Update Rule and in EPA's power sector
modeling.34 For a similar illustrative unit examined for combustion control upgrades (500 MW, tangential
boiler, 10,000 Btu/kWh heat rate, bituminous fuel) and input emission rate of 0.2 lb NOx/MMBtu, with a
capacity factor of 56%, the cost is $6,200/ton.35 When the capacity factor is 26%, the costs increase to
$9,700/ton. At a higher capacity factor (i.e., 80%), the costs decrease, going to $5,300/ton, respectively.
Next, EPA examined the cost of SNCR installation at small units (25 MW and 100 MW in size). The
costs at the three capacity factors (26%, 56%, and 80%) for the 25 MW unit were $38,300/ton,
$19,500/ton, and $14,600/ton. The costs for the 100 MW unit at the three capacity factors (26%, 56%,
and 80%)36 were $19,300/ton, $10,700/ton, and $8,400/ton.
Next, using a similar set of assumptions as in the previous paragraph (i.e., type of coal and capacity
factor),37 EPA examined the remaining coal-fired fleet that lack SNCR or other NOx post-combustion
control to estimate an emission weighted average and median cost of SNCR installation (on a $/ton basis)
using the Retrofit Cost Analyzer.38 In this case the input NOx rate was the 2023 baseline NOx rate
engineering analysis rate (see Section B in the Ozone Transport Policy Analysis Proposed Rule TSD for
details) and the "controlled" NOx rate was usually the input NOx rate reduced by 25% (circulating
fluidized bed boilers were assigned reductions of 50%). Costs were estimated for units that had a
minimum input NOx rate of at least 0.14 lb/MMBtu. In this instance, the emission weighted average cost
is $6,000/ton and the median value is $6,300/ton. EPA repeated these calculations for the subset of coal-
fired units less than 100 MW and found an emission weighted average cost of $6,700/ton and a median
value of $9,600/ton.39
EPA also looked at the size of coal steam units that had a SNCR or SCR. Looking at the existing coal
fleet in NEEDS for the Summer 2021 Reference Case, there are 146 GW (318 units) with a NOx post-
combustion control: 25 GW (91 units) with SNCR and 121 GW (227 units) with SCR. There are
approximately 52 GW of coal units without a NOx post-combustion control. The average size of units
with SNCR or SCR is about 275 MW and 530 MW, respectively, indicating that not only is SCR a more
common NOx post-combustion control, covering about 60% of the coal fleet capacity, but also that it
tends to be installed on larger coal-fired units. Looking at the subset of units with post-combustion
controls that are less than 100 MW, EPA found that post-combustion controls favored SNCR. There are
30 such units with SNCR and 4 with SCR, indicating that SNCR seems to be the overwhelming choice of
NOx post-combustion control for smaller units. Finally, EPA notes SNCR retrofits on coal steam units
34 "Coal-Fired SNCR Cost Methodology"
35 The input emission rate for a tangentially-fired unit with bituminous coal and state of the art combustion controls
is 0.2 (see Figure C.3).
36These capacity factors were chosen because they provide a wide range of capacity factors. Furthermore, 26% and
56% are the operating fleet-wide oil-gas steam and coal steam unit capacity factors in 2030 in the Summer 2021
IPM Reference Case, and therefore represent the projected average capacity factor for those plant types.
37 Units were assumed to use the coal identified in NEEDS, were assumed to operate the control throughout the year,
and were assumed to have a capacity factor of 56%.
38 See the Excel file
NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx for additional
details.
39 See the NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx for
details.
Page 23 of 40
-------
have a decreasing removal efficiency as unit size increases, which further explains the observed
prevalence on smaller units.40
Oil/Gas Steam:
Using the Retrofit Cost Analyzer, the agency conducted a similar analysis for SNCR retrofit on Oil/Gas
Steam units, again assuming a 25% reduction in emission rate.41 For a unit similar to that examined above
for coal steam units (500 MW, tangential boiler, 0.2 lb NOx/MMBtu, 10,000 Btu/kWh heat rate, with a
25% NOx removal efficiency) using natural gas with a 26% capacity factor, the cost is $8,400/ton of NOx
removed. At a capacity factor of 56%, the cost is $5,600/ton. At higher capacity factors (i.e., 80%), the
costs decrease, going to $4,900/ton. The costs were largely indifferent to the selection of gas compared
with oil. For the same unit, but with a lower starting NOx rate of 0.14 lb/MMBtu, the costs at the three
capacity factors increased to $10,500, $6,600, and $5,600/ton. Next, EPA examined the cost of SNCR
installation at small units (25 MW and 100 MW in size with an input NOx rate of 0.2 lb/MMBtu). The
costs at the three capacity factors (26%, 56%, and 80%) for the 25 MW unit were $31,400/ton,
$16,200/ton, and $12,300/ton, respectively. The costs for the 100 MW unit were $16,100/ton, $9,200/ton,
and $7,400/ton, respectively.
Finally, using a similar method for determining an input NOx rate and a NOx removal efficiency
described above for coal steam units, EPA examined the remaining oil/gas steam fleet that lack SNCR or
other NOx post-combustion control to estimate an emission weighted average and median cost of SNCR
installation (on a $/ton basis) using the Retrofit Cost Analyzer.42 Costs were estimated for units that had a
minimum input NOx rate of at least 0.14 lb/MMBtu and had athree-year average (2019-2021) of at least
150 tons of ozone season NOx (i.e., approximately 1 ton per day during the ozone season). Units were
assumed to have a NOx removal rate of 25% and assumed to operate the control throughout the year and
had an assumed capacity factor of 26% (the fleet-wide oil/gas steam capacity factor from the Summer
2021 IPM reference case). In this instance, the emission weighted average cost is $8,600/ton and the
median value is $9,600/ton.
F. Cost and Emission Rate Performance Estimates for Retrofitting with SCR and Related
Costs
For coal-fired and oil/gas steam units, an SCR retrofit is the state-of-the-art technology used to achieve
the greatest reductions in NOx emissions. The agency examined the cost for newly retrofitting a unit with
SCR technology. Based on the updated Retrofit Cost Analyzer,43 EPA assumed a new state-of-the-art
SCR retrofit could achieve 0.05 lb/MMBtu emission rate44 performance on a coal steam unit (regardless
40 See Table 5-4 in the Documentation for EPA's IPM Summer 2021 Reference Case.
41 "Gas-Fired SNCR Cost Methodology"
42 See the NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx for
details
43 See the Retrofit Cost Analyzer (Update 1-26-2022)
44 In prior modeling used for CSAPR rulemakings, EPA had assumed new SCR retrofits could achieve a 0.07
lb/MMBtu NOx emission rate for bituminous coal and 0.05 lb/MMBtu for other coal rank (e.g., subbituminous).
EPA had previously highlighted the upper end of this range as a representative rate based on the achievable NOx
Page 24 of 40
-------
of type of coal used) and 0.03 lb/MMBtu for an oil-/gas steam unit. The same assumption was used in the
EPA's power sector modeling for this proposed rule. Historically, many coal-fired units with SCR
retrofits after 2010 achieved NOx emission rates near or below 0.05 lb/MMBtu, further supporting that
new SCRs can achieve such an emission rate. Additionally, most oil/gas steam units equipped with SCR
have achieved NOx emission rates under 0.03 lb/MMBtu.
Coal Steam:
To understand the effect of input NOx rate on costs, using the Retrofit Cost Analyzer equations, the EPA
performed a bounding analysis to identify reasonable high and low per-ton NOx control costs for adding
SCR post-combustion controls across a range of potential input NOx rates.45 For a hypothetical unit 500
MW in size with a relatively low input NOx rate (e.g., 0.2 lb NOx/MMBtu, bituminous fuel, 2 lb
S02/MMBtu, 60% removal efficiency, 56% capacity factor, and 10,000 Btu/kWh heat rate), the capital
cost was about $151,000,000. For a similar unit with an input NOx rate of 0.4 lb/MMBtu and 80% NOx
removal efficiency, the total capital cost was $160,000,000. The cost on a per-ton basis varies with the
assumptions concerning the operation of the unit and the book life of the loan (or lifetime of the
equipment). Assuming an annual capital recovery factor of 0.143, NOx rate of 0.2 lb/MMBtu and
removal efficiency of 60% and annual operation, the cost per ton was $16,400/ton ($14,700/ton for the
capital cost, $300/ton for the FOM cost, and $l,400/ton for the VOM cost). For the unit with the NOx
rate of 0.4 lb/MMBtu and removal efficiency of 80%, the costs were $6,800/ton ($5,800/ton for the
capital cost, $100/ton for the FOM cost, and $850/ton for the VOM cost).
For this proposed rule, EPA examined the remaining coal-fired fleet that lack SCR to estimate a median
cost of SCR installation (on a $/ton basis). In this case the "input" NOx rate was, generally, the 2023
baseline NOx rate engineering analysis rate (see Section B in the Ozone Transport Policy Analysis
Proposed Rule TSD for details). For units with existing SNCR controls, for the "input" NOx rate, we
identified each unit's maximum monthly emission rate from the period 2009-2021. Costs were estimated
for units that had "input" NOx rates of at least 0.14 lb/MMBtu prior to installation of the post-combustion
control. With the control installed, the output NOx rate was reduced to a value of 0.05 lb/MMBtu.
Furthermore, we assumed annual operation of the control and assumed a capacity factor of 56% (the
capacity factor for coal units from the Summer 2021 IPM v.6 reference case). In this instance, these
assumptions produce an emission weighted average of $ll,000/ton, a median value of $13,700/ton and a
90th percentile value of $20,900/ton. EPA repeated these calculations for the subset of coal-fired units less
than 100 MW and found a median value of $15,500/ton and a weighted average of $11,900/ton.46
In states linked in 2026, there are 21 non-CFB coal steam units at least 100 MW that are currently
equipped with SNCR that reported emissions. In 2021, those units had a combined 17,000 tons of OS
NOx emissions, with a weighted average NOx rate of 0.17 lb/MMBtu. If those were to have a achieved a
0.05 lb/MMBtu emission rate, commensurate with retrofitting a new SCR, they could reduce 12,000 tons
emission rate for units burning bituminous coal, though units burning subbituminous and lignite coal could achieve
lower emission rates. Improvement in the catalyst and additional experience constructing and operating SCRs has
resulted in higher confidence in guaranteeing lower emission rates, even for units burning bituminous coal. See "
Coal-Fired SCR Cost Methodology" and "Gas-Fired SCR Cost Methodology" for details..
45 For these hypothetical cases, the "input" NOx rate includes the effects of existing combustion controls (e.g., low
NOx burners).
46 See the NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx for
details of this calculation.
Page 25 of 40
-------
of OS NOx emissions, with several units achieving reductions of over 1,000 tons of OS NOxeach.
Because these units are already achieving some level of NOx emission reduction with an SNCR, the
dollar per ton cost would be greater for these units than comparable units that lacked SNCRs. If an
existing SNCR achieves 25% removal efficiency, upgrading to an SCR with 90% removal efficiency
would add an incremental 65% removal efficiency. Assuming the SCR retrofit costs about the same for a
unit whether or not it already has an SNCR, then the SCR retrofit cost would properly be spread over only
the incremental tons of reductions. This would lead to the per ton cost increasing by 38%, not accounting
for cost savings because SCRs have a lower per ton O&M cost than SNCRs.
For a hypothetical unit 500 MW in size with a relatively low input NOx rate (e.g., 0.17 lb NOx/MMBtu,
bituminous fuel, 2 lb S02/MMBtu, 70% removal efficiency, 56% capacity factor, and 10,000 Btu/kWh
heat rate), the capital cost was about $155,000,000. Assuming an annual capital recovery factor of 0.143,
NOx rate of 0.17 lb/MMBtu and removal efficiency of 70% and annual operation, the cost per ton was
$17,000/ton ($15,200/ton for the capital cost, $300/ton for the FOM cost, and $l,500/ton forthe VOM
cost).Repeating the fleet-wide analysis of SCR retrofit cost, but limited to just units with SNCR that could
be considered for SCR retrofit (at least 100 MW in size and non a circulating fluidized unit), results in an
emission weighted average of $13,400/ton, a median value of $14,100/ton and a 90th percentile value of
$19,000/ton. These costs are not simply 38% higher than the fleet-wide analysis for all units without SCR
because this subset of units has different characteristics than the wider fleet.
Oil/Gas Steam:
To understand the effect of input NOx rate on costs, using the Retrofit Cost Analyzer equations, the EPA
performed a bounding analysis to identify reasonable high and low per-ton NOx control costs for adding
SCR post-combustion controls across a range of potential input NOx rates.47 For a hypothetical unit 500
MW in size with an input NOx rate of 0.14 lb NOx/MMBtu, 79% removal efficiency, 26% capacity
factor, and 10,000 Btu/kWh heat rate using natural gas, the capital cost was about $60,000,000. For a
similar unit with an input NOx rate of 0.2 lb/MMBtu and 85% NOx removal efficiency, the total capital
cost was $61,000,000. The cost on a per-ton basis varies with the assumptions concerning the operation
of the unit and the book life of the loan (or lifetime of the equipment). Assuming an annual capital
recovery factor of 0.143, NOx rate of 0.14 lb/MMBtu and removal efficiency of 79% and annual
operation, the cost per ton was $14,700/ton ($13,600/ton for the capital cost, $300/ton for the FOM cost,
and 800/ton for the VOM cost). For the unit with the NOx rate of 0.2 and removal efficiency of 85%, the
costs were $9,900/ton ($9,000/ton forthe capital cost, $200/ton forthe FOM cost, and $700/ton forthe
VOM cost).
For this proposed rule, EPA examined the remaining oil/gas steam fleet that lack SCR to estimate a
median cost of SCR installation (on a $/ton basis). In this case the "input" NOx rate was, generally, the
2023 baseline NOx rate engineering analysis rate (see Section B in the Ozone Transport Policy Analysis
Proposed Rule TSD for details). For units with existing SNCR controls, for the "input" NOx rate, we
identified each unit's maximum monthly emission rate from the period 2009-2021. Costs were estimated
for units that had input NOx rates of at least 0.14 lb/MMBtu prior to installation of the post-combustion
control and decreasing to rates of 0.03 lb/MMBtu following control installation. EPA relied on the same
methodology for determining SCR optimization at these units to inform its estimate of emission rate
47 For these hypothetical cases, the "input" NOx rate includes the effects of existing combustion controls (e.g., low
NOx burners).
Page 26 of 40
-------
performance for SCR operation at new retrofits, which agreed with the emission rate achievable with an
SCR retrofit described in "Gas-Fired SCR Cost Methodology." Furthermore, we assumed annual
operation of the control and assumed a capacity factor of 26% (the capacity factor for coal units from the
Summer 2021 IPM v.6 reference case). In this instance, these assumptions produce an emission weighted
average of $7,700/ton, a median value of $10,000/ton and a 90th percentile value of $15,300/ton.48 EPA
repeated these calculations for the subset of oil/gas steam units less than 150 tons per ozone season and
found aweighted average of $15,600/ton.49
G. Generation Shifting
For this proposed rule, EPA modeled generation shifting to units with lower NOx emission rates only
within the same state as a proxy for estimating the amount of generation that could be shifted in the near-
term (i.e., by 2023 and 2026). For 2023, by limiting economic builds in IPM to baseline projected levels
we further assume that such generation shifting only occurs within and among all generators (including
non-fossil sources) that are already in operation and connected to the grid in EPA's IPM baseline. Under
these circumstances, shifting generation to lower NOx- or zero-emitting EGUs, similar to operating
existing post-combustion controls, uses investments that have already been made, and can significantly
reduce EGU NOx emissions relatively quickly. For example, natural gas combined cycle (NGCC)
facilities can achieve NOx emission rates of 0.0095 lb/MMBtu, compared to existing coal steam facilities,
which emitted at an average rate of 0.12 lb/MMBtu of NOx across the 22 states included in the CSAPR
Update in 2019. Similarly, generation could shift from coal units lacking post-combustion controls to coal
units with SCR-or SNCR post-combustion controls. Shifting generation to lower NOx-emitting EGUs
would be a cost-effective, timely, and readily available approach for EGUs to reduce NOx emissions. In
2026, IPM builds were not constrained to baseline levels, resulting in the potential for additional
reductions as a result of building additional lower emitting capacity beyond the levels selected under the
baseline, consistent with the assumption of new control installations in budget setting. EPA considers that
the amount of generation shifting modeled to occur within each particular linked state in response to the
selected control strategy represented by $l,800/ton and $10,000/ton reflects the generation shifting that
can occur in the 2023 and 2026 ozone seasons, respectively.50 Figures G. 1, G.2, and G.3 below illustrate
the relatively low amount of generation-shifting assumed in EPA's analysis relative to historical levels.
48 See the "NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx"
49 See the "NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx"
50 EPA conducted IPM runs to determine emission reductions associated with generation shifting at $10,000/ton, but
ultimately determined that $11,000/ton should be the representative cost of retrofitting an SCR on a coal steam unit
and associated NOx mitigation strategies. Therefore, the $ll,000/ton cost threshold incorporates the $10,000/ton
generation shifting. To accurately describe the analysis conducted, EPA is labeling the generation shifting conducted
as $10,000/ton in this section of the TSD. Since that NOx price did not induce significant amounts of generation
shifting, given the other mitigation strategies included in the model run, EPA does not believe that the results would
have changed appreciably if a $11,000/ton price on NOx was included instead.
Page 27 of 40
-------
Figure G.l. Regional Coal and Gas Summer Generation Changes Base to Cost Threshold Case
(2023, GWh)
Region
Coal Adj.
Coal
Coal
Coal
Combined
Combined
Combined
Combined |
Base Case
($1,800/
Change
Percent
Cycle Adj.
Cycle
Cycle
Cycle
ton)
Change
Base Case
($1,800/
Change
Percent
ton)
Change
IMISO
85.803
84.784
-1 u|';
-1.19%
|o| 891
102.260
68
0.36% I
\Y
0
0
0
0.00%
28.807
-r
-0.06% I
IMM
51.276
5U.4XX
-788
-1.54%
157.994
I5X.T,<>
¦;45
0.60% I
SERC
50.540
5u.4'U
-46
-0.09%
i 4
i -.:xi
1 .<: IX
0.77%
ERCOT
i:.1¦;<.<.
12.077
-SX<>
-6.86%
92.320
T,.280
%1
1.04%
WECC
29.058
28.889
-169
-0.58%
88.115
88.949
834
0.95%
Figure G.2. Regional Coal and Gas Summer Generation Changes Base to Cost Threshold Case (2026,
GWh)51
Region
Coal Adj.
Base Case
Coal
($10,000/
ton)
Coal
Change
Coal
Percent
Change
Combined
Cycle Adj.
Base Case
Combined
Cycle
($10,000/
ton)
Combined
Cycle
Change
Combined
Cycle
Percent
Change
MISO
71.889
65.975
-5.914
-8.23%
101.799
102,383
584
0.57%
\ YISO
0
0
0
0.00%
27.100
X".
"73
2.85%
PJM
40.5XX
^J.090
-1.4'JX
-3.69%
li.MXt)
161.880
5()()
o U%
SI-'.KC
40. U5
37.748
-:.5<><.
-6.43%
133.603
136.091
:.4xx
1 X<>%
ERCOT
6.203
2.345
- S5X
-62.20%
81.016
X4.":4
"07
4 5X"„
WECC
26,707
24.319
-2.388
-X :.5<.-
95.151
:.5X4
:
Figure G.3. Historical Rate of Generation Change for Coal and Combined Cycle Units
Coal
GWh
2016
Coal
GWh
2017
Coal
GWh
2018
Coal | Average
GWh annual
2019 | change
Combined
Cycle
GWh
2016
Combined
Cycle
GWh
2017
Combined
Cycle
GWh
2018
Combined
Cycle
GWh
2019
Average
annual
change
Total
linked
states
450,217
425,755
400,888
328,390 -9.8%
354,841
337,916
393,282
422,744
6.4%
51 Emission reductions derived from generation shifting will be captured in the dynamic budgets in all cases. For the
pre-set budget years, and the illustrative 2026 values shown above, it is estimated and incorporated through an
additional calculation step. For dynamic budget years, it will be directly incorporated through the inclusion of
updated heat input data reflecting observed, post-compliance generation shifting - therefore the need for an
"estimation" is mooted.
Page 28 of 40
-------
H. Feasibility Assessment: Implementation Timing for Each EGU NOx Control Strategy
Restarting and Optimizing Existing Post-Combustion Controls
EPA evaluated the time it would take for EGUs to turn on idled SCRs and SNCRs. The EGU sector is
very familiar with restarting SCR and SNCR systems. Based on past practice and taking account of the
steps needed to restart the controls (e.g., re-stocking reagent, bringing the system out of protective lay-up,
performing inspections), returning these idled controls to operation is possible by 2023. In states
included in the NOx Budget Trading program, 2005 EGU NOx emission data suggest that 126 out of 156
coal-fired units with SCR operated the systems in the summer ozone season, likely for compliance with
the program, while idling these controls for the remaining seven non-ozone season months of the year.52
Similarly, in 2005, 20 of 72 coal-fired units with SNCR in the NOx Budget Trading program operated in
both the ozone season and non ozone season and had ozone season NOx rates more than 25% below their
non ozone season NOX rates suggesting that the SNCR was operated during the ozone season and may
not have been operated during non ozone season.53 To comply with the seasonal NOx limits, these SCR
and SNCR controls were regularly taken out of and then returned to service within seven months.
Therefore, EPA believes this SCR and SNCR optimization mitigation strategy is available for the 2023
ozone season.
EPA also found there are very few units that appear to have fully idled their SCRs. EPA assessed the
number of coal-fired units with SCR that are currently operating with ozone-season emission rates greater
than or equal to 0.2 lb/MMBtu suggesting that their units may not be operating their NOx post-
combustion control equipment. EPA finds that only 8 units in the contiguous United States (of which
four are in states that are "linked" at or above 1% in this proposed rule) fit this criterion.54 EPA's
assumptions that this mitigation technology is available for the 2023 ozone-season is further bolstered
given that the previous rulemakings (i.e., CSAPR Update and the Revised CSAPR Update) identified
turning on and optimizing existing SCR as a cost-effective control technology available by the next ozone
season after those rules were promulgated. Many sources successfully implemented that strategy in
response to those rules, and it appears that only a low number of units in the regions covered by these
rules may have entirely ceased operating these controls subsequently. Similarly, for SNCRs (as shown
above where the cost of operating an existing SNCR control is described), most units are currently
operating these controls, meaning that timing considerations need to focus on additional reagent
acquisition.
Full operation of existing SCRs that are already operating to some extent involves increasing reagent (i.e.,
ammonia or urea) flow rate, and maintaining and replacing catalyst to sustain higher NOx removal rate
operations. As with restarting idled SCR systems, EGU data demonstrate that operators have the
capability to fully idle SCR systems during winter months and return these units to operation in the
summer to comply with ozone season NOx limits. The EPA believes that this widely demonstrated
behavior also supports our finding that fully operating existing SCR systems currently being operated,
which would necessitate fewer changes to SCR operation relative to restarting idled systems, is also
52 Units with SCR were identified as those with 2005 ozone season average NOx rates that were less than 0.12 / lb
MMBtu and 2005 average non-ozone season NOx emission rates that exceeded 0.2 lb/MMBtu.
53 For this analysis, we identified units that had heat input greater than zero in both seasons and had non-ozone
season NOx rates above 0.20 lb/MMBtu.
54 See the "2021 NOx Rates for 234 SCR Coal Units.xlsx" file in the docket for details. Of the four units in states
linked at or above 1%, three are in Missouri and one is in Pennsylvania.
Page 29 of 40
-------
feasible for the 2023 ozone season. Regarding full operation activities, existing SCRs that are only
operating at partial capacity still provide functioning, maintained systems that may only require increased
chemical reagent feed rate up to their design potential and catalyst maintenance for mitigating NOx
emissions. Units must have adequate inventory of chemical reagent and catalyst deliveries to sustain
operations. Considering that units have procurement programs in place for operating SCR, this may only
require updating the frequency of deliveries. This may be accomplished within a few weeks. The vast
majority of existing units with SCRs covered in this action fall into this category. To the extent that there
are any existing units with fully idled SNCR systems, the same timing considerations and conclusions
apply. Moreover, since SNCR systems do not require catalyst, and may only need additional ammonia or
urea reagent, the timing may even be faster than that for SCR systems. EPA concludes that these SNCR
systems could be fully operational within a few weeks.
Moreover, hourly unit-level data, such as that shown in Figure B.3, clearly show that SCR performance
can improve within a 2-month time frame. Specifically, when controls are partially operating (as EPA has
demonstrated is the case in nearly all units with optimization potential), the data shows the hourly
emission rate varying significantly (reflective of SCR performance) over hours that occur well within two
months of one another. For instance, the size of the rectangle in Figure B.3 showing hourly NOx rates for
the unit when the control is partially operating in 2017 and 2018 reflect the 25th-75th percentile
hours. The top left graphic shows emission rates varying between approximately 0.22 b/MMBtu to 0.07
lb/MMBtu in 2017 for instance. This variation, reflective of SCR performance, is occurring within a two-
month time span, indicating the ability for quick improvements in control performance even controlling
for load levels.55
Combustion Control Upgrades
Combustion controls, such as LNB and/or OFA, represent mature technologies requiring a short
installation time - typically, four weeks to install along with a scheduled outage (with order placement,
fabrication, and delivery occurring beforehand and taking a few months). Construction time for installing
combustion controls was examined by the EPA during the original CSAPR development and is discussed
in the TSD for that rulemaking entitled, "Installation Timing for Low NOx Burners (LNB)", Docket ID
No. EPA-HQ-OAR-2009-0491-0051.56 Industry has demonstrated retrofitting LNB technology controls
on a large unit (800 MW) in under six months (excluding permitting). This TSD is in the docket for the
CSAPR Update and for this rulemaking.
Retrofitting Post-Combustion Control Technologies
SNCR is a mature, post-combustion technology that requires about 12 months from award through
commissioning (not including permitting) at a single boiler. Conceptual design, permitting, financing, and
bid review require additional time. A month-by-month evaluation of SNCR installation is discussed in
EPA's "Engineering and Economic Factors Affecting the Installation of Control Technologies for
55 Unit-level hourly emission rate data at www.epa. gov/ampd. See also "Miami Fort Hourly Emission Rate at
Capacity Factor of 50%-80%" in the docket for this rulemaking. This file shows the emission rate changes occuring
withing two months of one another.
56 http://www.epa.gov/airmarkets/airtransport/CSAPR/pdfs/TSD_Installation_timing_for_LNBs_07-6-10.pdf
Page 30 of 40
-------
Multipollutant Strategies," located in the docket for this rulemaking.57'58 The analysis in this exhibit
estimates the installation period from contract award as within a 10-13-month timeframe. The exhibit
also indicates a 16-month timeframe from start to finish, inclusive of pre-contract award steps of the
engineering assessment of technologies and bid request development. EPA is therefore assuming SNCRs
can be retrofit in 16 months, consistent with the more complete timeframe estimated by the analysis in the
exhibit.
SCR is a mature, post-combustion technology, for which EPA makes a fleetwide retrofit timing
assumption of about 3 years. Unit-specific installation timing depends on site-specific engineering
considerations and on the number of installations being considered. A month-by-month evaluation of
SCR installation is discussed in EPA's "Engineering and Economic Factors Affecting the Installation of
Control Technologies for Multipollutant Strategies," located in the docket for this rulemaking.59 A
detailed analysis for a single SCR system and a seven-SCR system can be found in Exhibits A-3 and A-4
of the report. For a single SCR, the analysis in this exhibit estimates the installation period from contract
award in a 16-month timeframe. The exhibit also indicates a 21-month timeframe from start to finish,
inclusive of pre-contract award steps of the engineering assessment of technologies and bid request
development, though there are instances of SCR retrofits taking as little as 9 months from contract award
to completion (about 13 months allowing for work prior to contract award).60 At a facility with multiple
SCR units, the facility can stagger installation to minimize operation disruptions. For a seven-unit SCR
system, the analysis in this exhibit finds that seven SCR units could be installed at a single facility in 35
months, or roughly an incremental 2.5 months per additional retrofit. EPA notes that most of the SCR
retrofits that may be expected in response to this proposed rule would happen at facilities where only one
or two boilers would need to be retrofit. EPA believes the timing estimates for SCR installation have not
changed significantly since the publication of this report.
In addition to these prior reports and installation schedules cited above and in prior CSAPR Rules, EPA
assessed the schedules again as it prepared this proposal. In the summer of 2021, a third party engineering
and consulting firm provided an assessment based on the latest data of SCR cost, performance, and timing
assumptions. This timing assessment is consistent with the timing schedules described in EPA's analysis.
In particular, the findings noted that a SCR project could go from capital investment to commercial
operation in 30-36 months, with faster schedules possible if motivated by a deadline or compliance date.61
The availability of skilled labor - specifically, boilermakers - is an important consideration for, and
potential constraint to, the installation of a significant amount of emission controls. Significant analysis
on boilermaker labor was conducted by the EPA to verify that sufficient boilermaker labor would be
available to support the installation of pollution control retrofits required by Clean Air Interstate Rule
57 "Engineering and Economic Factors Affecting the Installation of Control Technologies for Multipollutant
Strategies" (EPA) http://nepis.epa.gov/Adobe/PDF/P1001G0Q.pdf
58 Exhibit A-5 in "Final Report: Engineering and Economic Factors Affecting the Installation of Control
Technologies for Multipollutant Strategies" shows the timeline for retrofitting an ACI system, which has an
equivalent timeline to an SNCR.
59 "Engineering and Economic Factors Affecting the Installation of Control Technologies for Multipollutant
Strategies" (EPA) http://nepis.epa.gov/Adobe/PDF/P1001G0Q.pdf
60 The AES Somerset station completed an SCR retrofit on a 675 MW boiler in 9 months from contract award. The
Keystone plant retrofit SCR on two 900 MW boilers are estimated to have taken just 17 to 19 months, even allowing
for a relatively long 6 to 8 months of preliminary engineering and contract negotiation. Two 600 MW units at New
Madrid were retrofit with SCR, with the first unit being retrofit in about 20 months from contract award.
61 See "Typical SCR and SNCR Schedule (Coal or Oil/Gas boilers)" in the docket for this rulemaking.
Page 31 of 40
-------
(CAIR), and can be found in the Boilermaker Labor Analysis for the Final Clean Air Interstate Rule
TSD.62 At the time of CAIR implementation, the US boilermaker population was estimated to be 28,000
boilermakers, based on International Brotherhood of Boilermakers (IBB) membership in 2003. EPA
analysis showed that the boilermaker population of 28,000 provided more than enough available
boilermaker labor to install the incremental 24 GW of SCR retrofits projected to be required by the rule in
a three-year time period, in addition to 49 GW of scrubbers over the same time period. As of September
28, 2021, the IBB reports its active membership at 47,615 members, a 70% increase in membership over
the past 18 years.63 In analyzing the labor availability necessary for the installation of 73 GW of SCR on
142 units and 123 GW of SNCR on 497 units that was projected under the NOx SIP Call, EPA found that
approximately 13,000 to 19,000 workers would be capable of completing all SCR and SNCR installations
over a two to three year period.64 Given that there are now more boilermakers available, and EPA is
projecting 32 GW (14 GW of coal steam units; 18 GW of oil/gas steam units) of incremental SCR
retrofits as part of this proposed rule,65'66 EPA believes the current number of boilermakers will provide
sufficient labor availability to install all of the SCR retrofits in this proposed rule in the proposed
implementation timeline.
The air pollution control industry has demonstrated that they are able to install significant amounts of air
pollution control equipment in short periods of time. Based on EPA data and analysis, SCRs went online
for the first time on approximately 41.7 and 15.6 GW of coal-fired capacity in 2003 and 2002,
respectively, as a result of the NOx SIP Call.67 In addition to the large number of SCRs installed during
this time, boilermakers were also working on an unusually high number of natural-gas fired power plants
during the same period. In the three decades prior to 2000, an average of 5 GW of new natural gas
capacity was constructed in the United States per year. However, from 2000 to 2004 there was on average
41 GW of new natural gas-fired power plant builds, for a total of 205 GW. Furthermore, the peak year for
new gas power plant builds during this period, 60 GW in 2003, coincides with the peak year for SCR
start-ups for the NOx SIP call.68 EPA believes this demonstrates a robust maximum rate of SCR start-ups
after the NOx SIP Call of 41.7 GW, which is more than enough to support the capacity of SCR retrofits in
this proposed rule.
While the proposed rule does not require installation of controls on any particular source, instead leaving
choice of compliance strategy to facility owners and operators, EPA projects roughly 32 GW of SCR
retrofits to occur. Accordingly, EPA considers it appropriate to evaluate the feasibility of installation of
the projected 32 GW of SCR retrofits on the proposed implementation timeline. EPA believes that, given
the timing estimates of 21 months for a single SCR installation to 35 months for SCR retrofits on seven
62 "Boilermaker Labor Analysis for the Final Clean Air Interstate Rule TSD"
https://archive.epa.gov/airmarkets/programs/cair/web/pdf/finaltech05.pdf
63 "Boilermakers AFL-CIO FORM LM-2 LABOR ORGANIZATION ANNUAL REPORT 2020-2021"
https://olmsapps.dol.gov/query/orgReport.do?rptId=782863&rptForm=LM2Form
64 "Feasibility of Installing NOx Control Technologies by May 2003" https://www.regulations.gov/document/EPA-
HQ-OAR-2001 -0008-0125
65 See the Regulatory Impact Analysis for this proposal
66 EPA is also projecting 0.6 GW of SNCR being retrofit on units. As this is a very small amount and SNCR
construction is much simpler than SCR, it does not affect EPA's conclusion on there being sufficient labor
availability.
67 "Boilermaker Labor Analysis for the Final Clean Air Interstate Rule TSD"
https://archive.epa.gov/airmarkets/programs/cair/web/pdf/finaltech05.pdf
68 "Typical Installation Timelines for NOx Emissions Control Technologies on Industrial Sources"
https://cdn.ymaws.eom/www.icac.com/resource/resmgr/ICAC_NOx_Control_Installatio.pdf
Page 32 of 40
-------
units at a single facility, a significant increase in boilermaker labor since 2003, and historical analysis that
shows a maximum rate of SCR retrofits after the NOx SIP Call of 41.7 GW in a single year, despite
exceptionally high new gas power plant builds occurring in that same year, the SCR installation timeline
in the proposed rule is feasible and consistent with past EPA analyses.
Generation Shifting
For this proposed rule, EPA examined generation shifting expected to occur when the covered fleet is
subjected to a marginal cost commensurate with the identified mitigation technologies. As discussed
above, this amount of generation shifting reflects the change is system dispatch as the fleet is faced a
different set of economic incentives. The amount of generation shifting could occur quickly as the
existing fleet readily makes unit dispatch decisions in the near term, real time based on market signals.
I. Additional Mitigation Technologies Assessed but Not Proposed in this Action
1. Combustion Control and SCR Retrofits on Combined Cycle and Combustion Turbine Units
Combined Cycle:
While many newer and/or larger combined cycle units are equipped with SCRs, some smaller, older units
do not have a post-combustion control. Based on sample cost estimates, EPA's estimate of the cost of
retrofitting an SCR on a 90 MW natural gas combined cycle unit (operating at an 65% capacity factor and
a 0.05 lb/MMBtu NOx emissions rate) to have an overnight project cost of $5.5 million dollars with an
annual O&M cost of $265,000, or roughly $12,000/ton.69
Evaluating the fleet of combined cycle units in states linked in 2026, there are 45 units, or about 4 GW of
capacity, that do not have an SCR, had a NOx emissions rate over 0.05 lb/MMBtu, and a capacity factor
above 10%. If those units were retrofit with SCR, they would have a reduction potential 3,100 tons, or
about 70 tons per unit.
Combustion Turbine:
Roughly 75% of Combustion Turbines use either water injection, Dry Low NOx Burners (DLN), or Ultra
Low NOx Burners (ULN). In some circumstances, these technologies can be difficult to retrofit, given
requirements for either access to sufficient water or space for upgraded combustors. It is estimated that
retrofitting DLN or ULD on a 50 MW combustion turbine that only operated in the summer at a 15%
capacity factor and NOx emission rate of 0.15 lb/MMBtu would have a project cost of roughly $2.2
million dollars with an annual O&M cost of $53,000. That translates to a cost of roughly $21,600/ton.70
Evaluating the fleet of combustion turbine units in states linked in 2023, there are 11 units, or about 780
MW of capacity, that: do not have combustion controls and could achieve a lower emission rate by
retrofitting them; and operated at a capacity factor greater than 10% on average for the 2019-2021 ozone
69 See the NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx and
the report "Combustion Turbine NOx Technology Memo."
70 See the NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx and
the report "Combustion Turbine NOx Technology Memo."
Page 33 of 40
-------
seasons. Those units have combined 2021 ozone season NOX emissions of 286 tons. If those units were
to retrofit with DLN/ULN, they would have a reduction potential of 143 tons, or about 13 tons per unit.71
While almost 20% of the combustion turbine fleet has an SCR, many units do not have SCRs and operate
at relatively low capacity factors and/or emission rates. EPA estimates the cost of retrofitting an SCR on a
50 MW combustion turbine unit (operating at a 15% capacity factor during the ozone season, and
negligibly the rest of the year, and a 0.15 lb/MMBtu NOx emissions rate) to have an overnight project
cost of $ 14 million dollars with an annual O&M cost of $62,000. The relatively low utilization of the unit
and high project cost imply a cost of roughly $102,000/ton.72
2. Mitigation Strategies at Small Units that Operate on High Electricity Demand Days (HEDD)
Within the universe of combustion turbine EGUs, EPA also examined the subset of units that typically
operate in a "peaking" fashion (i.e., those with low seasonal capacity factors and operating primarily on
days of peak electricity demand). EPA did not identify a broadly available mitigation strategy across this
class of units that was within the range of cost effectiveness values of other EGU strategies included in
the proposal at step 3, or that had relatively favorable air quality benefit to cost ratios (as discussed in
preamble section VI.D). 890 out of 1117 combustion turbines included in EPA's 2023 analysis for linked
states operated at a capacity factor of 15% or less across the 2021 ozone season (i.e., at levels suggesting
a peaking unit) and averaged less than 0.1 ton of NOx per day. The vast majority emitted less than 5 tons
for the season and did not have a day where they exceeded a single ton of emissions. This heavy
weighting towards low overall emissions and low emission-reduction opportunities from this subset of
units suggests a uniform mitigation strategy applied to all units would deliver relatively little reductions
relative to the amount of sources and relative to the average cost. However, EPA recognizes that within
the heterogenous universe of combustion turbine EGUs, there may be instances where mitigation
measures are cost effective for a given unit. In addition, these units generally operate as part of the
interconnected electricity grid with other EGUs. Therefore, EPA includes this class of units in its CSAPR
trading programs and proposes the same approach in this proposed rule. This creates an economic
incentive for emissions reductions to be realized from this class of units to the extent such opportunities
exist.
Types of mitigation strategies that could be installed on peaking units range from combustion controls
such as water injection technology to post-combustion controls such as SCR retrofit. As discussed in 1.1,
using representative values characteristic of uniform control treatment, these costs could range from
$24,000/ton (combustion controls) to $115,000/ton (post-combustion control) on average when applied to
the combustion turbine fleet. A mitigation strategy of broadly assumed generation shifting away from
these sources is both limited by the fact that 1) many of these units operate at peak demand times and may
71 Using a tons per season rather than a capacity factor cutoff, the fleet of combustion turbine units in states linked in
2023, there are 32 units, that: do not have combustion controls and could achieve a lower emission rate by
retrofitting them; and emitted 10 or more tons on average for the 2019-2021 ozone seasons. Those units have
combined 2021 ozone season NOx emissions of 995 tons.
72 See the NOx_Control_Retrofit_Cost_Tool_Fleetwide_Assessment_Proposed_CSAPR_2015_NAAQS.xlsx and
the report "Combustion Turbine NOx Technology Memo."
Page 34 of 40
-------
serve critical reliability purposes, and 2) the overall reduction potential is limited in non-peak times as
these units generally do not operate during those periods.
However, EPA's regulatory coverage and allowance holding requirements for these units will extract
emission reduction across this generator type at the limited units where it may be cost effective, even
when a mitigation strategy is not cost-effective across the fleet segment more broadly. The opportunity
cost of an allowance surrendered to cover emissions is anticipated to incentivize any mitigation behavior
where viable in this universe of units and cost-effective at the level of stringency of the overall EGU
control strategy (i.e., around $ll,000/ton).
EPA also identifies two features of these units that make this segment well suited towards a market
incentive approach as opposed to assuming any uniform at-the-source mitigation strategies or unit-
specific rate limitations. These are: 1) reliability considerations and 2) existing and ongoing state
regulatory developments.
Peaking units play a unique role in ensuring grid reliability, and in certain areas, as a result of their
function, also may contribute relatively large emissions to ozone levels at nearby receptors. The days that
are conducive to ozone in the summer tend to have high temperatures, and as a result, are associated with
substantial additional electricity demand from air conditioning (among other reasons). To meet this
incremental demand, particularly in some areas where there are noted transmission constraints, small
units that have relatively high emission rates initiate operation. These units are often simple cycle
combustion turbines or oil-fired boilers. They are usually small and only operate a few hours out of the
summer when they are critical to meeting demand. The generation they provide is likely critical to
ensuring grid stability and power system reliability during these high-demand times.
EPA recently evaluated unit-level operations on HEDD for the States in the Revised CSAPR Update. In
the 12 states affected by the CSAPR Update Revision, EPA identified a total of 1,096 units that operated
during the 2019 ozone season. Of these, 102 units exhibited capacity factors that fell below 10% for the
period. The majority of these units (94 out of 102) were combustion turbine units—29 of which were
fueled by oil and 65 were fueled by natural gas. While the 102 identified units, called "peaker units" here
(in reference to their use during "peak" electricity demand), operated in relatively few hours during the
2019 ozone season, an average of 13% of gross generation from these units occurred in higher energy
demand hours, which we define as the top 1% of hours with the highest regional electric load. For 18 of
these units, electricity production in higher energy demand hours accounted for at least 20% of their total
generation for the 2019 ozone season.
With their relatively high emission rates, relatively small seasonal capacity factors, and tendency to
operate on HEDD, the emissions from these units could have substantial emissions and air quality
impacts on high ozone days. An assessment of emissions intensities for the units relative to the state and
regional average emission rate indicates that the emission rates of these units can be up to 118 times their
respective state averages. In the 12-state region, 50 units across 17 facilities had emission intensity values
substantially higher than the state average. Dividing the unit-level 2019 ozone season NOx rate by the
average 2019 ozone season NOx rate for the state indicated that the emission rates for these units were at
least 20 times that of their respective state averages for the 2019 ozone season.
In a separate analysis, EPA identified six states located in the northeastern US in which significant air
quality problems may persist on HEDD—Pennsylvania, New Jersey, New York, Delaware, Connecticut,
and Maryland. For a better understanding of the emissions impact of combustion turbine unit operations,
we compared the peak hour generation and emissions activities of natural gas and oil units located in these
Page 35 of 40
-------
states on sample HEDD and low energy demand days (LEDD). The sample days were chosen from a
selection of 15 days in the 2019 ozone season with the highest and lowest cumulative daily gross load for
all EGUs in CAMD's data sets in the six states. The sampled HEDD and LEDD days and peak hours used
in the analysis are July 30, 16:00, and May 18, 18:00, respectively.
When comparing gross generation between the two days, we observe that combustion turbine natural gas
and oil units generate more in days and hours of higher energy demand. On the sampled LEDD,
combustion turbine natural gas and oil units in the six northeastern states provided a total of 3,953 MWh
of electricity over the course of the day—673 MWh of which was produced in the peak hour (Figure 1.1),
contributing to 539 lbs, or 9% of the total peak hour NOx emissions in the six states (Figure 1.2).
Comparatively, on the HEDD, gross generation from combustion turbine units amounted to 28,263 MWh
over the course of the day. Generation in the peak hour reached 2,207 MWh (Figure 1.1), and contributed
to 4,881 lbs, or 19% of total peak hour NOx emissions (Figure 1.3).
For the sampled HEDD day, the largest shares of peak hour NOx emissions from combustion turbine
units originate in New York and Pennsylvania (Figure 1.3). A unit-level assessment of peakers in New
York state indicates that while these units are highly emissions-intensive, they provide relatively minimal
generation in peak hours. Specifically, combustion cycle natural gas and oil units in New York contribute
to 1,359 lbs, or 19%, of the state's total peak hour NOx emissions on the sample HEDD, while only
providing 1,186 MWh, or 8%, of generation. On this sample HEDD, the Glenwood and Holtsville
facilities, in particular, account for 4% of total peak hour oil generation in New York but contribute to
31% of the total peak hour NOx emissions from oil units (Figure 1.4). With peak hour NOx rates of 0.44
lb/MMBtu and 0.58 lb/MMBtu, respectively, the Glenwood and Holtsville facilities are relatively
emissions intensive; however, these units only dispatch in hours and days with higher energy demand. In
2019, Glenwood operated a total of 31 hours, 19 of which fell in the ozone season, while Holtsville ran a
total of 403 hours. Of these, 222 hours fell in the ozone season.
These units have the highest emissions, and the most downwind impact, on particular HEDDs and at
particular receptors that are in relatively close proximity to the peaking units. As discussed above, they
otherwise do not have emissions levels, impacts, or cost-effective emission reduction opportunities that
would warrant imposition of additional control requirements (other than inclusion in the trading program)
at step 3 under the good neighbor provision for the 2015 ozone NAAQS. Moreover, in states with
relatively higher concentrations of peaker units that are in closer proximity to out of state receptors
identified in this proposed rule, EPA observes that mitigation measures have often been implemented.
These states, New York and New Jersey, have already adopted regulations to reduce summertime NOx
emissions from peaking units, and in doing so have conducted thorough assessments of the degree of
emissions reduction that is achievable from these units, particularly in light of their critical function in
ensuring grid reliability.
The New York Department of Environmental Conservation (NY-DEC), for example, adopted a rule in
January 2020 to limit emissions from combustion turbines that operate as peaking units. This rule.
Subpart 227-3,73 entitled "Ozone Season NOx Emission Limits for Simple Cycle and Regenerative
Combustion Turbines." applies to simple cycle combustion turbines (SCCTs) with anameplate capacity
of 15 MW or greater that supply electricity to the grid. The regulation, more commonly referred to as the
"Peaker Rule", contains two compliance dates with increasingly stringent NOx limits, as follow s: by May
73 Subpart 227-3 is found within Chapter III. Air Resources, Part 227, of Title 6 of New York Codes. Rules and
Regulations (NYCRR).
Page 36 of 40
-------
1, 2023, all SCCTs subject to the rule must meet a NOx emission limit of 100 ppmvd.74 and by May 1,
2025, gas-fired SCCTs must meet a NOx emission limit of 25 ppmvd. and distillate or other liquid-fueled
units must a limit of 42 ppmvd. In lieu of meeting these limits directly. New York's rule offers two
alternative compliance options. The first compliance option allows owners and operators to elect an
operating permit condition that would prohibit the source from operating during the ozone season. The
second option allows owners and operators to adhere to an output-based NOx daily emission rate that
includes electric storage and renewable energy under common control with the SCCTs with which they
would be allowed to average.
New York performed comprehensive studies to assess the reliability implications of removing these units
from the system. The 2020 Reliability Needs Assessment Report, carried out by the NYISO in November
2020, found that the deactivation of all peaking units impacted by New York's Peaker Rule without
replacement solutions results in transmission and resource adequacy deficiencies that leave the bulk
power system unable to serve the forecasted load in New York City throughout the study period (up to
2030).75 While units can be retrofitted with new NOx controls to comply with the rule, many of the older
units in New York are not configured in a manner conducive to the retrofit pollution control technology.
These units are therefore likely to elect to retire rather than pay the high installation costs for new water
injection systems. Units that elect to retire are required to submit a Generator Deactivation Notice to the
NYISO. If they do so, the NYISO will perform a Generator Deactivation Assessment to ensure that the
retirement of the respective unit does not create reliability issues for the system. Units identified by this
assessment as being necessary for grid reliability receive a two-year extension on their compliance
deadlines.
New Jersey addressed the issue of NOx emissions on peak energy demand days with a rule that defines
peak energy usage days, referred to as high electric demand days (HEDD), and sets operational
restrictions for HEDD units that operate on these days.76'77 New Jersey's HEDD Rules came into effect on
May 19, 2009, and imposed emissions control requirements for units with NOx rates exceeding 0.15
lb/MMBtu and lacking identified control technologies.78
Due to their relatively small size, large number of units, potentially high control costs, low emission
reduction potential, reliability considerations, and existing regulations, EPA does not identify any
uniform, cost-effective mitigation strategy for this class of units as a whole in its EGU NOx mitigation
strategy assessment for this proposal. EPA recognizes heterogeneity in receptor impacts and emissions
reduction potential across these types of units, but in the one region where emissions reductions at such
units may be warranted due to uniquely large impacts at nearby out-of-state receptors (i.e., in the New
York City area), the immediate two upwind states have already adopted measures to reduce emissions
from these units. EPA does not have sufficient information on the current record that would warrant
assuming application of additional mitigation technologies beyond those already in place or otherwise
incentivized or required. Nonetheless, EPA does include these units, along with all other EGUs meeting
74 Parts per million on a dry volume basis at 15% oxygen. Using a heat rate of 15,000 BTUs/KWh identified in New
York's background materials for the rule and the alternative 3 lbs NOx/MWh limit equating to 100 ppmvd, the
equivalent emission rate would be approximately 0.2 lb/MMBtu and significantly lower for the full implementation
of 25 ppmvd. See https://www.dec.txv.gov/does/air pdf/siprevision2273.pdf
75 https://www.nyiso.com/documents/20142/2248793/2020-RNAReport-Nov2020.pdf
76 N.J.A.C. §7:27-19
77 HEDD units are EGUs greater or equal to 15MW that commenced operation prior to May 1, 2005, and that
operated less than or equal to an average of 50% of the time during the ozone seasons of 2005 and 2007.
78 CTs are required to have water injection or selective catalytic reduction (SCR) controls; steam units must have
either an SCR or noncatalytic reduction (SNCR)
Page 37 of 40
-------
applicability criteria, in this proposed rule as covered sources (as in prior CSAPR trading programs).
Inclusion in the trading program will create an economic incentive for these units to take advantage of
economical emission reduction opportunities wherever available.
Hourly .gross generatwci by comfcustwri turbine natural gas and oil units in hk northeastern states on LEDD,
»!*nuniGji »a»i
£
5
E ™
- ¦ I
J I 4
11 it if i» n M ti n it
Hourly iroii t*ne»*&on by
-------
Figure 1.3. Percentage share of peak hour NOx emissions by unit type across region and states on an
sampled HEDD (Source: Air Markets Program Data (AMPD), ampd.epa.gov, EPA, 2020)
Peak hour NOx emissions by unit-type across states and region on HEDD1,
lb
5 2 woo
I
E
41
S
z
*5
J 10000
CT share of
emissions.
O
¦ Corrtburfiric Turtine ¦Other
¦
MID
Region arid States
1) Ji% 30"' (16:Q0f is used as the exemplary HEDD day and peak haur.
Figure 1.4. Peak hour generation and emissions on HEDD by oil units in NY as a percentage (Source:
Air Markets Program Data (AMPD), ampd.epa.gov, EPA, 2020)
Peak hour gross generation on a HEDD from oil units in NY state1-2
MWh
Peak hour NOx emissions on a HEDD from oil units in NY state.
Pounds
2%
5% to
J 4% 2*
Bayswater Peaking Facility
¦ Glenwood
Hawkeye Energy Greenport, LLC
¦ Holtsville Fadlity
Northport
Ravenswood Generating Station
Roseton Generating LLC
Page 39 of 40
-------
Appendix A: Historical Anhydrous Ammonia and Urea Costs and their Associated Cost
per NOx ton Removed in a SCR
Minimum Cost to Operate Anhydrous NH3 & Urea costs ($/ton) [from USDA]
Year
NH3 (anh)
NHaCost
Urea cost
Urea Cost
Cost
/ ton NOx
/ ton NOx
2009
$ 562
$320
$425
$425
2010
$ 548
$312
$424
$424
2011
$ 801
$457
$543
$ 543
2012
$ 808
$461
$746
$ 746
2013
$ 866
$494
$508
$ 508
2014
$ 739
$421
$533
$ 533
2015
$ 729
$416
$472
$472
2016
$ 588
$335
$354
$ 354
2017
$ 501
$286
$328
$ 328
2018
$ 517
$295
$357
$ 357
2019
$ 612
$349
$433
$433
2020
$499
$284
$375
$ 375
2021
$727
$414
$543
$543
Average price from the first reporting period in July of each year.
Source: Illinois Production Cost Report (GX_GR210) USDA-IIL, Dept of Ag Market News Service,
Springfield, IL
www.ams .usda.gov/mnreports/gx_gr210 .txt
www.ams.usda.gov/LPSMarketNewsPage
Page 40 of 40
------- |