Emergency Application — North Dakota, et al., Applicants v. Environmental Protection Agency, et al.
Supreme Court briefAug 16, 2024
Ask Donna
What actually matters in this document.
Text
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 12 of 18
ELIMINATION OF THE MERCURY SUBCATEGORY FOR
LIGNITE CAUSES IMMEDIATE AND IRREPARABLE
HARM TO THE NORTH DAKOTA LIGNITE INDUSTRY
AND TO BASIN ELECTRIC
28.
EPA established the lignite subcategory for mercury because
lignite units and lignite coal are markedly different than bituminous and
subbituminous coals. Lignite has a higher mercury content in many
instances and presents greater variability than other coals. The higher
sulfur content found in lignite fuels inhibits the ability of injected
sorbents to reduce mercury emissions at lignite plants. The mercury
content also results in higher levels of SO3 formed, which significantly
limits the mercury emission reduction potential of emission controls at
lignite plants.
29.
Basin Electric has used the same technology (combination of
sorbent injection plus a chemical additive (oxidizing agent)) as its
primary mercury control strategy since the MATS rule came into effect
and is not aware of more effective control technology.
30.
There is no evidence that the units at Antelope Valley and
Leland Olds could achieve compliance with the New Mercury Limitation
on a sustained basis with the currently installed equipment as is required
to meet a 30-day rolling basis while operating at full load.
-11(Page 383 of Total)
305a
USCA Case #24-1119
31.
Document #2058570
Filed: 06/07/2024
Page 13 of 18
The MATS RTR sets a mercury limitation for lignite units
without any technical basis that it can be met on a continuous basis, in
general, and provides no compliance margin to account for the variability
in unit performance and emissions control capabilities from unit to unit.
32.
Basin Electric is irreparably harmed by the final MATS RTR
because it is unknown if Antelope Valley and Leland Olds' existing
mercury controls can achieve the New Mercury Limitation of 1.2 lb/Tbtu
on a sustained basis at full load.
33.
The Final Rule places Basin Electric in an impossible position,
given the Rule's impending compliance date. Noncompliance with the
Clean Air Act is not an option.
34.
To have any possibility of meeting the New Mercury
Limitation, Basin Electric must modify the existing system at both
Antelope Valley and Leland Olds to produce a higher injection rate and
make the systems more robust. Even though EPA has not demonstrated
that the New Mercury Limitation will provide any health benefits, Basin
Electric must complete this modification project to lower the emission
rate. The modification costs and ongoing operation expenses are
significant.
(Page 384 of Total)
Specifically,
these
technologies
will
require
over
-12-
306a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 14 of 18
$4,000,000.00 in capital expenditures upfront for the four units
collectively, as well as increased labor costs for installation, operation,
and maintenance of the technology and equipment and associated
training, along with additional sorbent injection, will result in increased
operating costs over the long term. We must begin expending these
dollars immediately, and certainly before the resolution of this case, in
order to meet the deadlines set out in the Final Rule.
35.
Costs to comply with the New Mercury Limitation are
exorbitant and damage Basin Electric. Costs will be passed along to its
member cooperatives and end users who are harmed via higher
electricity prices. The capital and operational costs to Basin Electric, its
member cooperatives, and end users cannot be recouped.
THE NEW FPM LIMITATION WILL CAUSE IMMEDIATE AND
IRREPARABLE HARM TO THE ELECTRIC COOPERATIVES
AND TO BASIN ELECTRIC
36.
EPA's New fPM limit of 0.010 lb/MMBtu will require upgrades
at Leland Olds and Laramie River.
37.
Basin Electric's harm is immediate. Basin Electric would need
to begin engineering and constructing, at a minimum, ESP upgrades at
Leland Olds and Laramie River as soon as possible to have any
-13(Page 385 of Total)
307a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 15 of 18
opportunity to meet the new compliance date for the MATS RTR. If ESP
upgrades are required, Basin Electric would need 36 months to complete.
It is likely that the 36-month estimate will be further protracted due to
the lack of contractors available to perform the work.
38.
If ESP upgrades were not sufficient, baghouse technology
would be required. If a baghouse is required, Basin Electric would need
approximately 48 months to convert to baghouse technology.
39.
Costs of compliance with the New fPM Limitation are overly
burdensome, for the following reasons.
40.
ESP retrofits are expensive. They may cost an estimated
$67,262 per fPM ton removed. See Cichanowicz Technical Report.
41.
Baghouse installation is extremely costly. It is estimated to
cost $282,715 per fPM ton removed. See Cichanowicz Technical Report.
42.
Electric cooperatives have limited financial resources to
undertake projects of this magnitude coincident with other environmental
compliance projects.
43.
To comply with the MATS RTR, Basin Electric is forced to
take measures that immediately increase compliance and operational
costs. The MATS RTR impacts Basin Electric's ability to supply
(Page 386 of Total)
-14-
308a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 16 of 18
affordable, reliable energy to its customers. Added costs will place
upward pressure on rates for rural customers, particularly when
combined with the effects of EPA's other recent electric utility sectorfocused rules.
THE MATS RTR CREATES GRID RELIABILITY CONCERNS
44.
Lignite power plants, which provide a significant source of
electric power in North Dakota, are important to the regional economy.
45.
Thus, the Final Rule, with its reversal of EPA's position on
lignite-fired sources, impacts North Dakota more profoundly than other
areas of the country. These concentrated impacts affect the ability of the
North Dakota utilities to maintain adequate generation resources.
46.
Most (if not all) of the lignite plants in North Dakota must
make some changes as result of the Final Rule. These changes will
require an immense amount of coordination between different regulated
facilities and likely involve serious risks to the reliability of electric grids
providing power to the region while the removal equipment at each of the
impacted facilities are taken offline to undergo the additions and
upgrades required by the Final Rule.
-15(Page 387 of Total)
309a
USCA Case #24-1119
47.
Document #2058570
Filed: 06/07/2024
Page 17 of 18
The North American Electric Reliability Corporation has
predicted continued future shortfalls in North Dakota.' The MATS RTR
intensifies an already tenuous, overburdened grid in transition.
SUMMARY OF HARM TO BASIN ELECTRIC
48.
Basin Electric is harmed because it must immediately
commence costly compliance testing and project development to evaluate
whether it can meet the MATS RTR emissions limits and applicable
compliance deadline.
49.
The MATS RTR could potentially cause Antelope Valley,
Leland Olds and Laramie River which are dispatchable, reliable
generating resources, to operate differently at a substantial cost and
permanent loss to Basin Electric.
50.
Even if the MATS RTR is overturned, the direct costs to Basin
Electric, its member cooperatives, and end users cannot be recouped once
spent. These damages are permanent.
[Signature Follows on Next Page]
1 NERC, 2024 Summer Reliability Assessment (May 2024),
https://www.nerc.com/pa/RAPA/ra/Reliability%20Assessments%20DL/NERC_SRA_2024.pdf.
-16(Page 388 of Total)
310a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 18 of 18
I declare under penalty of perjury that the foregoing is true and
correct.
in A. McCollam
Dated:
(Page 389 of Total)
6 iciz-oz y
311a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 2 of 204
ROBERT MCLENNAN
DECLARATION OF HARM IN SUPPORT OF MOTION FOR A STAY
PENDING REVIEW
-1(Page 128 of Total)
312a
USCA Case #24-1119
3.
Document #2058570
Filed: 06/07/2024
Page 3 of 204
4.
-2(Page 129 of Total)
313a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 4 of 204
1
6.
FERC, MISO, https://www.ferc.gov/industries-data/electric/electric-powermarkets/miso.
1
-3(Page 130 of Total)
314a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 5 of 204
7.
9.
-4(Page 131 of Total)
315a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 6 of 204
10.
MILTON R. YOUNG STATION
11.
12.
2 2
-5(Page 132 of Total)
316a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 7 of 204
13.
2
14.
15.
-6(Page 133 of Total)
317a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 8 of 204
16.
17.
See
-7(Page 134 of Total)
318a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 9 of 204
A8achment A
Seeid.
MATS RTR RULE REVISIONS
18.
19.
20.
-8(Page 135 of Total)
319a
USCA Case #24-1119
21.
Document #2058570
Filed: 06/07/2024
Page 10 of 204
LIGNITE COMBUSTION
22.
2
See A8achment A
23.
LEC Comments filed June 23, 2024, https://downloads.regulations.gov/EPA-HQ-OAR2018-0794-5957/attachment_1.pdf
2
-9(Page 136 of Total)
320a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 11 of 204
24.
25.
-10(Page 137 of Total)
321a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 12 of 204
See A8achment A
ELIMINATION OF THE MERCURY SUBCATEGORY FOR LIGNITE
CAUSES IMMEDIATE AND IRREPARABLE HARM TO THE NORTH
DAKOTA LIGNITE INDUSTRY AND TO MINNKOTA
26.
See
-11(Page 138 of Total)
322a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 13 of 204
A8achment B
3
Id.
27.
-12(Page 139 of Total)
323a
USCA Case #24-1119
28.
Document #2058570
Filed: 06/07/2024
Page 14 of 204
Attachment C.
Attachment D
-13(Page 140 of Total)
324a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 15 of 204
29.
-14(Page 141 of Total)
325a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 16 of 204
30.
See Attachment A
A8achment A)
-15(Page 142 of Total)
326a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 17 of 204
MRY Unit
Average Hourly Mercury Emissions Value Achieved at Full
Load (Sorbent Trap Data)
18 ppm MProve and Non-Brominated PAC
Unit 1
2.17
Unit 2
1.61
31.
MRY Unit
Average Hourly Hg Emissions
Value Achieved at Full Load
(Sorbent Trap)
Brominated PAC
Average Hourly Hg Emissions
Value Achieved at Full Load
(Sorbent Trap)
Non-Brominated PAC
Unit 1
2.57
2.17
-16(Page 143 of Total)
327a
USCA Case #24-1119
32.
Document #2058570
Filed: 06/07/2024
Page 18 of 204
33.
-17(Page 144 of Total)
328a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 19 of 204
A8achment A
Figure -1 — MRY Unit 1
Existing System Mercury Removal Performance Capabilities using
Brominated PAC
-18(Page 145 of Total)
329a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 20 of 204
-19(Page 146 of Total)
330a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 21 of 204
-20(Page 147 of Total)
331a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 22 of 204
See Attachment
A
-21(Page 148 of Total)
332a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 23 of 204
Example MRY Unit 2 Cost Underestimations Summary Table
-1
EPA
Example
Hypotheti
cal
800 MW
EPA
Assumed
MRY U2
Costs
447 MW
Est.
Actual
MRY U2
Costs
447 MW
1
$2.6 M
$0.3 M
$1.9 M
Current Hg Removed
1,295 lb
77 lb
149 lb
Parameter
Current Hg Compliance (4.0 lb/TBtu) Cost
Current C/E ($ per lb Hg Removed)
2,004
3,845
12,754
Hg Control System Annualized Capital
Cost
Not
included
Not
included
$472k 2
BPAC Cost @ 5 lb/MMacf
$7.5 M
Not
included
$0.6 M
$1.3 M 3
$0.2 M
$1.6 M 4
M-Prove Cost
Future Hg Compliance (@ 5 lb/MMacf)
Cost
$7.5 M
$0.8 M
$3.4 M
Future Hg Removed
(EPA Assumed @ 1.2 lb/TBtu)
1,447 lb 5
110 lb
216 lb
Future C/E ($ per lb Hg Removed)
5,083
7,040
15,678
22,217
28,176
14,360
Incremental C/E ($ per lb Hg Removed)
Note 1 – EPA example only based on sorbent. EPA assumed current compliance cost includes sorbent and chemical fuel additive.
Est. actual cost based on 2023 MRY Unit 2 usage rate & pricing for both sorbent and chemical additive.
Note 2 – Cost of $5.0 million dollars from S&L project database was annualized using a capital recovery factor calculated based
on annual interest rate of 7% (pre-tax marginal rate of return on private investment, EPA Cost Manual Section 5) and 20 year
evaluation period (EPA Cost Manual Section 6).
Note 3 – Cost based on EPA assumed rate but using 2023 MRY BPAC pricing.
Note 4 – Cost based on 2023 MRY Unit 2 usage rate & pricing instead of assuming same as sorbent costs.
Note 5 – Based on calculated value for EPA example inlet Hg of 1,542 lbs (current Hg coal content) – 95 lbs (future emitted
amount). However, the EPA example identifies 1,468 lb for the incremental cost effectiveness calculation.
-22(Page 149 of Total)
333a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 24 of 204
-23(Page 150 of Total)
334a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 25 of 204
-24(Page 151 of Total)
335a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 26 of 204
THE NEW fPM LIMITATION WILL CAUSE IMMEDIATE AND
IRREPARABLE HARM TO THE NORTH DAKOTA UTILITIES AND
TO MINNKOTA
See
Attachment E
Attachment E
-25(Page 152 of Total)
336a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 27 of 204
Attachment
E
Id
Id.
-26(Page 153 of Total)
337a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 28 of 204
See A8achment B
See A8achment B
-27(Page 154 of Total)
338a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 29 of 204
THE MATS RTR CREATES GRID RELIABILITY CONCERNS
DUE TO EARLY RETIREMENTS OF COAL-FIRED UNITS
-28(Page 155 of Total)
339a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 30 of 204
3
4
New Source Performance Standards for Greenhouse Gas Emissions From New,
Modified, and Reconstructed Fossil Fuel-Fired Electric Generating Units; Emission
Guidelines for Greenhouse Gas Emissions From Existing Fossil Fuel-Fired Electric
Generating Units; and Repeal of the Affordable Clean Energy Rule, 89 Fed. Reg. 39798
(May 9, 2024); Hazardous and Solid Waste Management System: Disposal of Coal
Combustion Residuals From Electric Utilities; Legacy CCR Surface Impoundments, 89
Fed. Reg. 38950 (May 8, 2024); Supplemental Effluent Limitations Guidelines and
Standards for the Steam Electric Power Generating Point Source Category, 89 Fed. Reg.
40198 (May 9, 2024); National Emission Standards for Hazardous Air Pollutants: Coaland Oil-Fired Electric Utility Steam Generating Units Review of the Residual Risk and
Technology Review, 89 Fed. Reg. 38508 (May 7, 2024).
4
NERC, 2024 Summer Reliability Assessment (May 2024),
https://www.nerc.com/pa/RAPA/ra/Reliability%20Assessments%20DL/NERC_SRA_202
4.pdf.
3
-29(Page 156 of Total)
340a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 31 of 204
A8achment F
-30(Page 157 of Total)
341a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 32 of 204
5
See, e.g., Hanchey, “Mortality Surveillance During Winter Storm Uri, United States –
2021,” Disaster Med Public Health Prep (Dec. 2023),
https://pubmed.ncbi.nlm.nih.gov/37974501/; Sharma, “Winter Storm Elliott death toll
climbs to 56 as thousands still without power in -40 temperatures,” Yahoo News (Dec.
26, 2022),
https://www.yahoo.com/news/winter-storm-elliot-power-outages-154557710.html.
5
-31(Page 158 of Total)
342a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 33 of 204
SUMMARY OF HARM TO MINNKOTA
-32(Page 159 of Total)
343a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 34 of 204
-33(Page 160 of Total)
344a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 35 of 204
-34(Page 161 of Total)
345a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 36 of 204
Table A: MRY 1 and 2 Mercury Compliance Costs
Activity
Cost
Notes
Future mercury testing to
determine lowest achievable rate
$600,000
This is a minimum value.
Inlet Hg Monitor
$150,000
To track coal quality
WFGD Additive Dosing System
$750,000
To attempt to reduce mercury
emissions further
WFGD Oxidizing Reduction
Potential (ORP) Monitoring
System
$7,500
For WFGD dosing system
feedback
MRY Unit 2 Capital Costs:
Mercury New PAC Silo and
$5,000,000
injection equipment capital cost to
reach the lowest achievable rate
Based on industry data from
similar projects; This is the total
project cost without financing
costs.
MRY Unit 2 Operating & Maintenance (O&M) Costs:
WFGD Additive costs (based on
annual operation)
$1,412,000
Based on MRY usage rate and
supplier pricing
Mercury control additional PAC
costs (based on annual operation)
$1,300,000
Based on EPA hypothetical 5.0
lb/MMacf injection rate for 800
MW unit
-35(Page 162 of Total)
346a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 37 of 204
Activity
Cost
Notes
Mercury control additional
Potassium Iodide costs (based on
annual operation)
$1,600,000
Cost based on 2023 MRY Unit 2
usage rate & pricing instead of
assuming same as sorbent costs.
Cost is $1.4 million more than
estimated by EPA.
Incremental Mercury Control
O&M cost
$2,412,000
This is the cost in excess of the
current O&M costs. This
estimate is based on current
compliance of approximately
$1.9 million.
Total MRY 2 Costs
$8,919,500
Per MW (440MW) = $18,978
MRY 1 Projected Costs
$4,880,000
MRY has 235 MW. Based on the
cost per MW from itemized
costs for MRY 2
Total for MRY 1 and MRY 2
$13,799,500
Capital & O&M Costs:
Table B: MRY 2 fPM Compliance Costs
Activity
Cost
Notes
fPM Feasibility Study
$175,000
Based on roughly budgetary
estimates from Southern
Environmental , Inc.
Low cost: MRY 2 ESP
Rebuild Capital Cost
$36,326,000
Based on S&L's conceptual
cost estimates and inputs
from Southern
Environmental, Inc.
-36(Page 163 of Total)
347a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 38 of 204
Activity
Cost
Notes
Low cost: MRY 2 ESP
Rebuild Incremental O&M
Cost
$530,000
Incremental costs accounts
for costs incurred above
what is currently paid for by
station for existing PM
compliance (i.e. ESP power
consumption, fly ash
disposal, etc.)
Low cost: MRY 2 ESP
Rebuild Outage Cost
$1,421,000
High cost: New MRY 2
Baghouse
$242,083,000
Low cost: MRY 2 Baghouse
Incremental O&M Cost
$4,047,000
Low cost: MRY 2 Baghouse
Outage Cost
$507,000
Total fPM Cost Range:
Based on S&L's conceptual
cost estimating
Incremental costs accounts
for costs incurred above
what is currently paid for by
station for existing PM
compliance (i.e. ESP power
consumption, fly ash
disposal, etc.)
High – $246,812,000
Low – $38,452,000
Table C: Minnkota’s Total MRY Mercury and fPM Compliance Costs
Activity
Cost
Notes
MRY Total Mercury Costs
for MRY 1 and MRY 2
$13,799,500
From Table above, O&M
based on 1 year
MRY Total fPM Costs for
MRY 2
High – $246,812,000
Low – $38,452,000
From Table above, O&M
based on 1 year
-37(Page 164 of Total)
348a
USCA Case #24-1119
Document #2058570
Activity
Cost
Total Compliance Cost to
MRY
High – $260,611,500
Low – $52,251,500
Filed: 06/07/2024
Page 39 of 204
Notes
-38(Page 165 of Total)
349a
USCA Case #24-1119
(Page 166 of Total)
Document #2058570
Filed: 06/07/2024
Page 40 of 204
350a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 41 of 204
ATTACHMENT A
(Page 167 of Total)
351a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 42 of 204
Minnkota Power Cooperative, Inc.
Milton R. Young Station Units 1 and 2
Mercury Testing Results for the
MATS Residual Risk and
Technology Review
Rev. 1
May 22, 2023
Project No.: A14559.013
S&L Nuclear QA Program Applicable:
Yes
No
55 East Monroe Street
Chicago, IL 60603-5780 USA
312-269-2000
www.sargentlundy.com
(Page 168 of Total)
352a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
Filed: 06/07/2024
Page 43 of
204
Rev.
1
May 22, 2023
A14559.013
1.INTRODUCTION
1.1. PURPOSE
Sa
Sargent
& Lundy (S&L) was retained by Minnkota Power Cooperative, Inc. (Minnkota) to support the evaluation
of mercury (Hg) emissions reductions in response to the pre-published rule to amend the National Emission
Standards for Hazardous Air Pollutants (NESHAP) for Coal-and Oil-Fired Electric Utility Steam Generating
Units (EGUs), commonly known as Mercury and Air Toxics Standards (MATS) published on April 24, 2023
that would require additional Hg emissions reductions on the Milton R. Young (MRY) Station Units 1 and 2. As
part of this evaluation, S&L assisted Minnkota in the coordination of a Hg control test campaign to determine
if it is feasible to achieve incremental Hg emission reduction on a lignite-fired unit without a fabric filter that is
sufficient to meet a 1.2 lb/TBtu Hg emission rate on a continuous basis.
1.2. FACILITY BACKGROUND
The MRY station is located approximately seven (7) miles southeast of Center, North Dakota or forty (40)
miles northwest of Bismarck, North Dakota on ND Highway 25 at 3401 24th Street SW, Center, North Dakota
58530. MRY station provides energy to the Midcontinent Independent System Operator (MISO) system. MRY
station consists of two (2) units. Both MRY units are lignite-fired Babcock and Wilcox (B&W) cyclone boilers.
Both boilers fire North Dakota lignite coal supplied from BNI Coal, Ltd.’s Center Mine located in close proximity
to the plant. The MRY Unit 1 single wall cyclone boiler (Caroline type, radiant natural circulation) was placed
into service in 1970 and has a typical output capacity rating of 257 MWg (gross). The MRY Unit 2 opposed
wall cyclone boiler (Carolina type, radiant pump assisted natural circulation) was placed into service in 1977
and has a typical output capacity rating of 470 MWg (gross). Both units utilize selective non-catalytic reduction
(SNCR) and separated overfire air (SOFA) systems for NOx control, fuel additive (or halide) injection system
and non-halogenated (or non-brominated) powdered activated carbon (PAC) for Hg control, dry electrostatic
precipitators (ESP) for PM emissions control, and wet flue gas desulfurization (WFGD) systems for sulfur
dioxide (SO2) control.
1.2.1.Current Hg Control System Specifications
The existing Hg control system is designed to control Hg emissions below 4.0 lb/TBtu using a combination of
M-Prove halide injection and non-halogenated PAC. The M-Prove is directly applied on the coal belt prior to
reaching coal silos, whereas the non-halogenated PAC is injected into the duct downstream of the air preheater (APH). Additional information on the design of the existing fuel additive and PAC injection systems for
MRY Units 1 and 2 are summarized below:
•
MRY Common Non-brominated PAC Storage Silo:
o PAC Utilized: Cabot DARCO® Hg-H non-halogenated PAC
o Single storage silo with three (3) outlet cones or discharge connections. Each cone is
connected to a feeder train (A, B, and C).
o Feeder Train A is dedicated to MRY Unit 1
o Feeder Trains B and C are dedicated to MRY Unit 2
o Storage Volume: 4,200 cu.ft. (Nominal)
o Capacity: 105,000 lbs. (based on PAC density of 25 lbs/cu.ft.)
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 169 of Total)
1
353a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
A14559.013
o
Filed: 06/07/2024
Page 44 of
204
Rev.
1
May 22, 2023
Storage duration: Approximately 18 days based on silo capacity of 105,000 lbs. and total
combined PAC consumption rate of 244 lb/hr (MRY Unit 1 at 86 lb/hr and MRY Unit 2 at 158
lb/hr)
•
MRY Unit 1 (257 MWg)
MR
o Fuel Additive: ARQ (formerly ADA) M-Prove
Average M-Prove application rate: 6.0 ppm
Maximum M-Prove dosage pump rate: 18.0 ppm
o Non-brominated PAC Injection:
Maximum Train A PAC injection at 100% feeder rate: 1.43 lb/min (approximately 86
lb/hr or 1.06 lb/MMacf)
Transport piping limited to 192 lb/hr (2.37 lb/MMacf) to avoid pluggage issues
PAC injected into flue gas using eight (8) lances located across the APH outlet duct.
The lance depths vary from 18” – 54” to provide even distribution of PAC into the flue
gas stream
•
MRY Unit 2 (470 MWg)
o Fuel Additive: ARQ (formerly ADA) M-Prove
Average M-Prove application rate: 8.0 ppm
Maximum M-Prove dosage pump rate: 18.0 ppm
o Non-brominated PAC Injection:
Maximum Train B and C PAC injection at 100% feeder rate: 2.64 lb/min
(approximately 158 lb/hr or 1.12 lb/MMacf)
PAC injected into flue gas using eight (8) lances located across each of the North and
South APH outlet ducts for a total of sixteen (16) lances.
The lance depths vary from 15” – 78” to provide even distribution of PAC into the flue
gas stream
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 170 of Total)
2
354a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Filed: 06/07/2024
Milton R. Young Station Units 1 and 2
Page 45 of
204
Rev.
1
May 22, 2023
A14559.013
2.TEST CAMPAIGN SUMMARY
The MRY Units 1 and 2 test campaign was completed in phases to control testing variables and to
accommodate vendor availability, and scheduled outages. Testing included:
• November 23
23, 2023 to November 24, 2023: Maximizing MRY Unit 1 capabilities of the existing MProve fuel additive system and non-halogenated PAC injection (at 100% feeder rate) to evaluate if the
current system can meet 1.2 lb/TBtu.
•
December 19, 2023 to December 20, 2023: Maximizing MRY Unit 2 capabilities of the existing MProve fuel additive system and non-halogenated PAC injection (at 100% feeder rate) to evaluate if the
current system can meet 1.2 lb/TBtu.
•
March 19, 2024 to March 23, 2024: Utilizing a rental bulk bag unloading (BBU) system provided by
Motus Group tied into the existing MRY Unit 1 PAC conveying lines and injection lances to inject
brominated PAC (or BPAC), ARQ’s FastPAC Platinum®, at varied injection rates ranging from 100
lb/hr (or 1.23 lb/MMacf) to a maximum of 185 lb/hr (2.28 lb/MMacf) to stay below the transport piping
pluggage limit. The majority of this testing also included maximizing MRY Unit 1 capabilities of the
existing M-Prove fuel additive system; however, test runs on March 22 and March 23 included BPAC
injection with no fuel additive usage. Individual coal samples were taken and analyzed by a 3rd party
lab for determination of inlet Hg coal content.
•
March 28, 2024 to April 1, 2024: Individual coal samples were taken and analyzed by a 3rd party lab
for determination of inlet Hg coal content.
This testing was not able to be completed during the proposed rule’s short comment period of only 60 days.
Due to timing of boiler cleaning outages, time required to develop a test protocol and schedule, and
coordination with multiple vendors, rental equipment availability, various site activities, and unplanned unit
upsets/outages, a much longer duration was needed.
2.1. INCREMENTAL HG REMOVAL TEST RESULTS
The Hg emissions achievable based on maximizing current design capabilities using non-brominated PAC
and M-Prove without any modifications is summarized below for both MRY Units 1 and 2.
Table 2-1 — MRY Units 1 and 2 Existing System Capabilities
MRY Unit 1
MRY Unit 2
18 ppm M-Prove and
18 ppm M-Prove and
100% Non-brominated PAC
100% Non-brominated PAC
MWg
242
469
46
PAC Injection Rate
lb/MM
MMacf
1.06
1.12
Avg. Sorbent Trap Hg Emissions
lb/TBtu
2.17
1.61
Parameter
Unit Load during testing
Units
Based on maximizing injection capabilities of the existing systems (without any modifications), the test results
show that MRY Unit 1 and MRY Unit 2 cannot achieve the proposed MATS limit of 1.2 lb/TBtu.
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 171 of Total)
3
355a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
A14559.013
Filed: 06/07/2024
Page 46 of
204
Rev.
1
May 22, 2023
2.2. BROMINATED PAC PERFORMANCE
The proposed rule assumes a 90% Hg removal efficiency is feasible from all lignite units, even those equipped
with an ESP.
• In the Beyond-the-Floor memo (Docket ID No. EPA-HQ
HQ-OAR-2009-0234), it states that “[g]reater than
90 percent control can be achieved at lignite-fired units at a 2.0 lb/MMacf injection rate for units with
installed fabric filter and using treated (i.e., brominated) activated carbon or at an injection rate of 3.0
lb/MMacf for units using treated activated carbon with installed ESPs.”
•
According to the proposed MATS rule, EPA reiterates that “[i]n the beyond-the-floor analysis in the
final MATS rule, we noted that the results from various demonstration projects suggest that greater
than 90 percent Hg control can be achieved at lignite-fifired units using brominated activated carbon
sorbent at an injection rate of 2.0 lb/MMacf for units with installed FFs for PM control and at an injection
rate of 3.0 lb/MMacf for units with installed ESPs for PM control.”
The Final Rule relies on the same assumption. In EPA’s 2024 Technology Memorandum, EPA finds, “In the
beyond-the-floor analysis in the final MATS rule, we noted that the results from various demonstration projects
suggest that greater than 90 percent Hg control can be achieved at lignite- fired units using brominated
activated carbon sorbent at an injection rate of 2.0 lb/MMacf for units with installed Faric Fililters for PM control
and at an injection rate of 3.0 lb/MMacf for units with installed ESPs for PM control. . . all units (in 2022) would
have needed to control their Hg emissions to less than 95 percent to meet an emission standard of 1.2 lb/TBtu.
Based on this, we expect that the units could meet the proposed, more stringent, emission standard of 1.2
lb/TBtu by utilizing brominated activated carbon at the injection rates suggested in the beyond-the-floor
memorandum from the final MATS rule.”
During the MRY Unit 1 March testing, MRY secured a temporary rental injection skid. The materials of
construction of the existing PAC silo (common to MRY Units 1 and 2) is not currently compatible to store
halogenated PAC. The silo would require an internal coating to prevent corrosion (but could otherwise be
reused). The temporary rental injection skid avoided
ed corrosion to the existing silo, but also allowed for
decoupling MRY Unit 1 from the common PAC storage silo to prevent interfering with MRY Unit 2 Hg control
operation.
To achieve a dosage rate of 3.0 lb/MM
MMacf, an injection rate of 245 lb/hr would be required which would exceed
the existing MRY Unit 1 Train A PAC injection/transport system limit of 192 lb/hr (2.37 lb/MMacf). The
maximum BPAC injection rate tested was limited to 185 lb/hr (2.28 lb/MMacf) to avoid line pluggage.
The Hg emissions reductions achievable based on maximizing the use of BPAC (without any fuel additives)
supplied via a temporary rental injection system tied into the existing transport piping/lances is summarized
below for MRY Unit 1. A higher PAC injection rate was not possible due to maximum capability of the existing
transport piping while preventing pluggage.
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 172 of Total)
4
356a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Filed: 06/07/2024
Milton R. Young Station Units 1 and 2
Page 47 of
204
Rev.
1
May 22, 2023
A14559.013
Table 2-2 — MRY Unit 1 Existing System Capabilities using Brominated PAC
Parameter
Units
Unit Load during testing
MRY Unit 1
185 lb/hr BPAC
MWg
257.1
.1
PAC Injection Rate
lb/MM
MMacf
2.28
2.
28
Avg. Sorbent Trap Hg Emissions
lb/TBtu
2.57
At the current injection capabilities of the existing system (i.e. requiring minimal modifications/retrofit of the
existing equipment), BPAC cannot be applied to reduce Hg emissions to 1.2 lb/TBtu.
2.3. MRY MERCURY REMOVAL EFFICIENCY
2.3.1.Lignite Coal Mercury Content
To calculate an overall mercury removal efficiency needed to control to 1.2 lb/TBtu, the coal Hg inlet must be
defined.
•
EPA reported the “Hg Inlet” level based on the maximum Hg content of the range of feedstock coals
that the EPA assumes is available to each of the plants in the Integrated Planning Model (IPM).
o With respect to MRY, EPA reported “Hg inlet”:
MRY Units 1 and 2: 7.81 lb/TBtu
•
According to the proposed rule, EPA estimated the 2021 Hg inlet concentration from actual 2021 fuel
usage and 2021 Hg emissions reported to the EPA. However, based on the 2024 Technical Memo,
EPA updated the information based on 2022 information.
o With respect to MRY, EPA “Estimated Hg inlet” content documented in 2023 and 2024
Technical Memo is summarized in the table below:
Table 2-3 — EPA Estimated North Dakota Lignite Coal Hg Inlet
•
2023 Technical
Memo
2024 Technical
Memo
(Estimated 2021
Hg Inlet)
(Estimated 2022
Hg Inlet)
Parameter
Units
MRY Unit 1
lb/TBtu
7.78
9.70
MRY Unit 2
lb/TBtu
7.79
9.70
However, recent test information and other resources for the North Dakota lignite fired at MRY has
indicated that significantly higher inlet Hg is experienced at MRY:
o Within the BNI Coal, Ltd.’s Center Mine, the Kinneman Creek (KC) and Hagel (HA) beds are
targeted for the coal supply for MRY. Based on the 2021 BNI coal data (constructed from
Carlson reports), the avg. coal Hg content is approximately 16 lb/TBtu for KC and 15 lb/TBtu
for HA
HA.
o The variability of the projected lignite coal quality received from the Center Mine from 2025
through 2036 is shown in the following table.
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 173 of Total)
5
357a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Filed: 06/07/2024
Milton R. Young Station Units 1 and 2
Page 48 of
204
Rev.
1
May 22, 2023
A14559.013
Table 2-4 — Forecasted 2025 – 2036 Center Mine Ultimate Coal Analyses (As-Received)
Fuel Parameter
Units
Average
Minimum
Maximum
Mercury Content
ppm
0.091
0.053
0.184
Higher Heating Value (HHV)
Btu/lb
6,625
6,489
6,739
Estimated Hg Emission
lb/TBtu
8.41
4.79
17.42
o
Industry experience has shown that lignite coal deposits vary significantly in quality, including
fuel combustion performance, mineral content, and Hg content, resulting in a coal that can
change on a day-to
to-day basis depending on the coal seam being mined at the time. This
variability was demonstrated by the range of coal analyses from MRY Unit 1 recent short-term
testing in 2024 (average = 10.1 lb/TBtu, with individual results ranging from 4.9 – 18.6 lb/TBtu
over the course of five (5) days of testing). Individual coal samples and how they varied across
coal feeders, per day are shown in following table.
Table 2-5 — MRY Unit 1 Coal Sampling Analysis
Coal Hg Inlet (lb/TBtu)
Date
19-Mar-24
19
24
20-Mar-24
20
24
20-Mar-24
20
24
21-Mar-24
21
24
21-Mar-24
21
24
22-Mar-24
22
24
22-Mar-24
22
24
23-Mar-24
23
24
28-Mar-24
28
24
1-Apr-24
24
Sample
Feeder #1
Feeder #3
Feeder #4
Feeder #5
Feeder #7
#1@ 0730 hrs
14.5
13.0
-
-
-
#2@ 1600 hrs
-
-
11.1
8.2
8.0
#3@ 0100 hrs
12.5
10.5
-
-
-
#1@ 0730 hrs
6.2
7.9
-
-
-
#2@ 1600 hrs
-
-
7.2
10.1
18.5
#3@ 0100 hrs
10.9
8.1
-
-
-
#1@ 0730 hrs
14.1
7.9
-
-
-
#2@ 1600 hrs
-
-
18.6
4.9
7.1
#3@ 0100 hrs
7.2
7.1
-
-
-
#1@ 0700 hrs
10.4
13.4
-
-
-
#2@ 1600 hrs
-
-
6.9
11.0
11.4
#3@ 0100 hrs
9.2
7.8
-
-
-
#1@ 1030 hrs
10.2
8.3
-
-
-
#2@ 1500 hrs
-
-
14.9
11.9
9.5
#1@ 0930 hrs
16.3
8.0
-
-
-
#2@ 1300 hrs
-
-
6.0
12.1
12.3
#3@ 1500 hrs
10.2
6.9
-
-
-
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 174 of Total)
6
358a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
Filed: 06/07/2024
Page 49 of
204
Rev.
1
May 22, 2023
A14559.013
2.3.2.Required Mercury Removal Based on Lignite Coal Mercury Content
Based on the recent Hg fuel analyses, Hg control higher than 90% would actually be required based on the
range of inlet coal Hg content expected to control to 1.2 lb/TBtu (i.e. keeping the outlet value calculated by the
EPA constant). Note that control to this value does not offer any operating margin for potential exceedances
that may occur due to response delays associated with coal variability. The following table identifies the
required Hg control needed based on several different coal Hg content references. Based on these
estimations, any Hg control approach would need to be able to accommodate a wide range of inlet Hg in order
to optimize operating costs long-term.
Table 2-6 — Hypothetical Hg Emissions and Control Performance Based on Coal
Analyses
Est. Hg Control at
4.0 lb/TBtu
Est. Hg Control at
1.2 lb/TBtu
(%)
(%)
7.81
48.8
84.6
9.70
58.6
87.6
Average
Av
10.1
60.4
88.1
Maximum
18.6
78.5
93.5
Minimum
4.9
18.4
75.5
Average
Av
8.41
52.4
85.7
Maximum
17.42
77.0
93.1
Minimum
4.79
16.5
75.0
Fuel Hg Content Reference
Coal Hg Inlet
(lb/TBtu)
EPA Technical Memo
2023 Table 11
Docket ID. No: EPA-HQ
HQ-OAR-2018-0794 1
2024 Table 10
Docket ID. No: EPA-HQ
HQ-OAR-2018-0794 2
2024 MRY Unit 1 Test Campaign
Center Mine Forecast
2.3.3.Projected Mercury Removal Based 3.0 lb/MMacf BPAC
Based on the maximum BPAC rate that MRY Unit 1 was able to test due to current system limitations (1
(185
lb/hr or 2.28 lb/MMacf), the figure below plots the estimated percent removal at the higher injection rate of 3.0
lb/MMacf BPAC using all measurements from the MRY Unit 1 March testing (with and without fuel additive
usage). The plotted values demonstrate a trend line in which BPAC cannot even achieve 80% Hg removal
efficiency.
1
Benish S. et al.
l. (January 2023). 2023 Technology Review for the Coal- and Oil-Fired EGU Source Category.
Environmental Protection Agency.
2
Benish S. et al. (January 2024). 2024 Update to the 2023 Proposed Technology Review for the Coal- and Oil-Fired
EGU Source Category. Environmental Protection Agency.
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 175 of Total)
7
359a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
Filed: 06/07/2024
Page 50 of
204
Rev.
1
A14559.013
May 22, 2023
Figure 2-1 — MRY Unit 1 Existing System Mercury Removal Performance
Capabilities using Brominated PAC
This result is contrary to EPA’s assumption that BPAC at a rate of 3.0 lb/MM
MMacf can be used to result in a 90%
removal efficiency. The plotted curve shown in the figure shows a leveling off such that increasing the amount
of sorbent results in diminishing improvement in Hg control. The projected curve based on the test campaign
results shows this leveling off taking place somewhere less than 80% capture.
Although the plotted values do not support a conclusion that the new Hg 1.2 lb/TBtu limit can be met, further
investigation into other Hg control options in combination with upgrading/optimizing existing Hg control
equipment would be required to determine the lowest mercury emission rate in lb/TBtu that can be achieved
on a long-term basis, considering the range of fuel Hg variability and other technological challenges inherent
in capturing Hg resulting from lignite that have
ve been documented to occur. Some proposed options for
additional Hg control include:
•
Increased fuel additive rate
•
Improved reliability of fuel additive concentration in relation to real-time coal firing rates
•
Implementation of inlet Hg monitor for improved feedback control of Hg control systems
•
Improved lance design to achieve ideal distribution of PAC at all typical unit operating conditions
•
Application of WFGD re-emission control additive
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 176 of Total)
8
360a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
A14559.013
Filed: 06/07/2024
Page 51 of
204
Rev.
1
May 22, 2023
Further analysis, engineering, testing and equipment modifications would be necessary to determine if these
options would improve Hg control. However, it is clear that adding more brominated PAC, as was assumed in
the Final Rule, is not adequate, given the properties of lignite, compliance margin necessary, and limitation of
mine mouth facilities in regards to fuel staging (i.e. must use coal received from mine; unable to fire only certain
coals that have a more ideal or predictable range of Hg content during a 30-day rolling average).
It should be noted that the achievable Hg emission rate should not be construed to represent an enforceable
regulatory or proposed permit limit. Corresponding permit limits must consider normal operating fluctuations
and coal variability and take into account a minimum additional 20% margin for these fluctuations. Since a
combination of new and/or upgraded control systems would be expected to be required, obtaining a guarantee
from a single vendor to ensure that the unit achieves compliance below the permit limit will be challenging.
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 177 of Total)
9
361a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
Filed: 06/07/2024
A14559.013
Page 52 of
204
Rev.
1
May 22, 2023
3.EPA COST VALIDITY
3.1.1.Current Hg Compliance Cost Effectiveness (4.0 lb/TBtu)
With respect to MRY, EPA estimated the cost effectiveness for current 2021 Hg emissions is shown below in
an excerpt from Table 12 in 2023 Technology Review for the Coal- and Oil-Fired EGU Source Category
(Docket ID. No: EPA-HQ
HQ-OAR-2018-0794).
).
Response: Flaws in EPA’s cost analysis for current compliance:
• Est. Hg In (lb) & Hg Out (lb)
o Table 12 would appear to have flipped MRY Unit 1 and Unit 2 in the table, utilizing the higher
MRY Unit 2 operating conditions (heat input, hg loading, etc.) for the smaller sized Unit 1 and
vice versa.
•
PAC Injection Rate:
o Table 12 Avg. Sorbent (lb/hr) – EPA noted MRY Unit 1: 19.0 lb/hr and MRY Unit 2: 43.0 lb/hr
to achieve controlled Hg rate of 3.2 lb/TBtu.
o Minnkota PAC sorbent injection rates to achieve controlled Hg rate of 3.85 lb/TBtu for MRY
Unit 1 is expected to be 86 lb/hr and for MRY Unit 2 is 158 lb/hr
hr.
•
Cost of PAC:
o Table 12 non-brominated PAC sorbent cost – EPA assumed a cost of $0.83/lb.
o In the 2024 Technical Memo, EPA adjusted this cost down to $0.80/lb.
o Based on MRY operational costs for 2023, non-brominated PAC sorbent cost is $0.86/lb.
o Based on MRY operational costs for 2023, actual non-brominated PAC costs for achieving
current compliance with 4.0 lb/TBtu indicated MRY Unit 1: $119,813 and MRY Unit 2:
$329,328
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 178 of Total)
10
362a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
Filed: 06/07/2024
A14559.013
•
Page 53 of
204
Rev.
1
May 22, 2023
Cost of Fuel Additive:
o Table 12 Est. 2021 Additive Cost – EPA noted that "Additive costs are unknown. For this
analysis, the EPA assumed the additive costs are the same, annually, as the sorbent costs.”
And lists costs as MRY Unit 1: $227,410 and MRY Unit 2: $147,267
o Based on MRY operational costs for 2023, actual fuel additive costs for achieving current
compliance with 4.0 lb/TBtu indicated MRY Unit 1 $715,157 and MRY Unit 2: $1,574,793.
o Based on the actual 2023 fuel additive usage rates and costs, EPA’s underestimate results in
$487,747 and $1,347,383 that should have been included in the cost analysis for MRY Units
1 and 2, respectively.
3.1.2.Future Hg Compliance Cost Effectiveness (1.2 lb/TBtu)
EPA calculated unit-level cost-effectiveness to meet the proposed, more stringent, emissions standard using
brominated activated carbon at an injection rate of 5.0 lb/MMacf for units with an ESP for PM control or at an
injection rate of 2.5 lb/MMacf for units with fabric filter for PM control.
With respect to MRY, the EPA estimated the cost effectiveness (assuming 2021 operational characteristics)
is shown below in an excerpt from Table 13 in 2023 Technology Review for the Coal- and Oil-Fired EGU
Source Category (Docket ID. No: EPA-HQ
HQ-OAR-2018-0794):
EPA’s incremental cost-effectiveness per the 2024 Update to the 2023 Proposed Technology Review for the
Coal- and Oil-Fired EGU Source Category (Docket ID. No: EPA-HQ
HQ-OAR-2018-0794) is based on a model
800 MW Gulf Coast lignite-fired EGU with a heat rate of 11,000 Btu/kWh operating at an 80% capacity factor
and a Hg concentration of 25.0 lb/TBtu, resulting in an incremental cost-effectiveness of $28,176 per pound
of Hg controlled. It assumes that the unit currently meets a Hg emission standard of 4.0 lb/TBtu using an
injection rate of 2.5 lb/MMacf of non-brominated activated carbon at a sorbent cost of $0.80/lb and that the
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 179 of Total)
11
363a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
Filed: 06/07/2024
A14559.013
Page 54 of
204
Rev.
1
May 22, 2023
unit can meet a Hg emission standard of 1.2 lb/TBtu using an injection rate of 5.0 lb/MMacf of brominated
activated carbon at a sorbent cost of $1.15/lb.
•
•
•
Note that the example does not include fuel additives or any equipment upgrade costs.
EPA made following changes to the calculations between 2023 and 2024 Technical Memo’s:
o EPA updated the Gulf Coast Hg concentration from 14.9 lb/TBtu (2023) to 25.0 lb/TBtu (2024).
This resulted in the baseline annual uncontrolled Hg emissions to change from 919 lb Hg to
Th
1,542 lb Hg.
o EPA corrected the formula for conversion of sorbent injection rate from lb/MMacf to lb/hr by
adjusting the conversion factor from (520 R / 785 R) to (785 R / 520 R). The conversion factor
was applied incorrectly in 2023 Technical Memo.
o EPA added an additional factor to update the formula for conversion of sorbent injection rate
from lb/MMacf to lb/hr which was not previously accounted for in 2023 Technical Memo.
For comparison with the values calculated by the EPA in Table 13, it should be noted that the 2024
calculated cost effectiveness of the 800 MW example used by the EPA to meet 1.2 lb/TBtu, without
fuel additives, is $5,083 per pound of Hg controlled.
Response: Flaws in EPA’s cost analysis for future compliance with 1.2 lb/TBtu:
• Est. Hg In (lb) & Hg Out (lb)
o See previous responses on Table 12 for flipped MRY Unit 1 and MRY Unit 2 unit
information/sizing and cost of fuel additive.
•
BPAC Injection Rate:
o EPA’s cost analysis assumes lignite units with an ESP can achieve 1.2 lb/TBtu, which has not
been demonstrated. The injection level has a direct bearing on the operational costs because
it dictates the amount of BPAC necessary to reduce Hg emissions. Therefore, cost
calculations are hypothetical because no project data demonstrates what the injection level
would be, if 1.2 lb/TBtu is feasible.
o Although the overall feasibility of complying with the proposed Hg limit is undetermined, the
testing confirms that based on maximizing injection capabilities of the existing systems, MRY's
current equipment configuration cannot achieve 1.2 lb/TBtu.
•
Cost of BPAC:
o Table 13 brominated PAC sorbent cost – EPA assumed of $1.15/lb.
o MRY Unit 1 test campaign brominated PAC cost = $1.25/lb.
•
Missing capital costs:
o Irrespective of feasibility, EPA calculated cost-effectiveness shown in Table 13 does not
include capital costs for modifying, upgrading and/or adding new equipment that would be
necessary for the MRY Station due to limitations of existing equipment.
o Modification to the existing PAC injection system, would include,
e, but not be limited to, the
following:
The materials of construction of the existing PAC silo (common to MRY Units 1 and
2) is not currently compatible to store halogenated PAC. The silo would require an
internal coating to prevent corrosion in order to store brominated PAC.
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 180 of Total)
12
364a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
A14559.013
Filed: 06/07/2024
Page 55 of
204
Rev.
1
May 22, 2023
New feeding equipment, transport piping and injection lances would be required to
accommodate a higher injection rate.
As the existing PAC storage silo is shared by MRY Units 1 and 2, the higher injection
rate required for achieving 3.0 lb/MMacf for both units would reduce the total storage
duration to less than seven (7) days of storage. Due to the weather experienced at
the site and the remote location, seven (7) days of storage is recommended for each
unit. Improved equipment redundancy would also likely be required to accommodate
the range of coal Hg expected to be experienced in the future. Therefore, it is likely
that the existing equipment would be dedicated to MRY Unit 1, and a separate silo
would be required for MRY Unit 2 to ensure adequate supply, turndown flexibility, and
reliability is achieved to maintain compliance with a defined Hg emission limit.
As such, a new MRY Unit 2 system would be required to achieve higher injection rates of
PAC. An analogous project to install Hg control equipment at a 500 MW coal-fired unit in 2021
21
costs roughly $5.0 million dollars, based on S&
S&L internal mercury control database, actual
project costs from recent relevant projects, and adjusted for MRY specific design.
o
Overall, the cost-effectiveness calculated is still a substantial under-estimation for the incremental Hg control
on MRY Units 1 and 2.
• To provide an example, hypothetical MRY Unit 2 costs are summarized in the following table to
underscore the magnitude of dollars that EPA failed to include in its calculations and that must be
expended by Minnkota.
• Note the table below does not include or account for any costs associated with MRY Unit 1 system
upgrades.
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 181 of Total)
13
365a
Minnkota Power Cooperative, Inc.
USCA
Case #24-1119
Document #2058570
Milton R. Young Station Units 1 and 2
Filed: 06/07/2024
Page 56 of
204
Rev.
1
May 22, 2023
A14559.013
Table 3-1 — Example MRY Unit 2 Cost Underestimations Summary
EPA Example
Hypothetical
800 MW
EPA Assumed
MRY U2 Co
Costs
447
47 MW
Est. Actual
MRY U2 Costs
447
47 MW
Current Hg Compliance (4.0 lb/TBtu) Cost 1
$2.6 M
$0.3 M
$1.
1.9 M
Current Hg Removed
1,295 lb
77 lb
149 lb
Current C/E ($ per lb Hg Removed)
2,004
3,845
12,754
Not included
Not included
$472k 2
$7.5 M
$0.6 M
$1.3 M 3
Not included
$0.2 M
$1.6 M 4
Future Hg Compliance (@ 5 lb/MMacf) Cost
$7.5 M
$0.8 M
$3.4 M
Future Hg Removed
(EPA Assumed @ 1.2 lb/TBtu)
1,447 lb 5
110 lb
216 lb
5,083
7,040
15,678
Parameter
Hg Control System Annualized Capital Cost
BPAC Cost @ 5 lb/MMacf
M-Prove Cost
Future C/E ($ per lb Hg Removed)
22,217
28,176
14,360
Incremental C/E ($ per lb Hg Removed)
Note 1 – EPA example only based on sorbent. EPA assumed current compliance cost includes sorbent and chemical
fuel additive. Est. actual cost based on 2023 MRY Unit 2 usage rate & pricing for both sorbent and chemical additive.
Note 2 – Cost of $5.0 million dollars from S&L project database was annualized using a capital recovery factor
calculated based on annual interest rate of 7% (pre-tax marginal rate of return on private investment, EPA Cost
Manual Section 5) and 20 year evaluation period (EPA Cost Manual Section 6).
Note 3 – Cost based on EPA assumed rate but using 2023 MRY BPAC pricing.
Note 4 – Cost based on 2023 MRY Unit 2 usage rate & pricing instead of assuming same as sorbent costs.
Note 5 – Based on calculated value for EPA example inlet Hg of 1,542 lbs (current Hg coal content) – 95 lbs (future
emitted amount). However, the EPA example identifies 1,468 lb for the incremental cost effectiveness calculation.
Mercury Testing Results for the MATS Residual Risk and Technology
Review
(Page 182 of Total)
14
366a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 57 of 204
ATTACHMENT B
(Page 183 of Total)
367a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 58 of 204
Technical Comments on
National Emission Standards for Hazardous Air Pollutants: Coal- and Oil-fired
Electric Utility Steam Generating Units Review of Residual Risk and Technology
Prepared by
J. Edward Cichanowicz
Consultant
Saratoga, CA
James Marchetti
Consultant
Washington, DC
Michael C. Hein
Hein Analytics, LLC
Whitefish, MT
Prepared for the
National Rural Electric Cooperative Association
American Public Power Association
America’s Power
Midwest Ozone Group
NAACO
National Mining Association
Power Generators Air Coalition
June 19, 2023
(Page 184 of Total)
368a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 59 of 204
1.
Summary of Flaws in EPA’s Approach ............................................................................. 1
2.
Introduction.................................................................................................................... 3
3.
Description of EPA Reference PM Database .................................................................... 5
3.1
Coal Fleet Inventory ............................................................................................................. 5
3.2
Database Characteristics ...................................................................................................... 6
3.2.1 Selection of Sample Year, Quarter.......................................................................................... 6
3.2.2 Number of Samples ................................................................................................................ 7
3.2.3 PM Data Selection and Analysis.............................................................................................. 8
3.2.4 Example Cases ........................................................................................................................ 9
3.3
Conclusions ........................................................................................................................ 10
4.
Coal Fleet PM Emissions Characteristics .........................................................................12
4.1.1
4.1.2
4.1.3
5.
PM Rate of 0.015 lbs/MBtu .................................................................................................. 13
PM Rate of 0.010 lbs/MBtu .................................................................................................. 13
PM Rate of 0.006 lbs/MBtu .................................................................................................. 13
CRITIQUE OF COST-EFFECTIVENESS CALCULATIONS ........................................................14
5.1
EPA Evaluation ................................................................................................................... 14
5.1.1 EPA Study Inputs ................................................................................................................... 14
5.1.2 EPA Results ........................................................................................................................... 16
5.2
Industry Study.................................................................................................................... 17
5.2.1 Revised Cost Inputs .............................................................................................................. 17
5.2.2 Cost Effectiveness Results .................................................................................................... 19
5.3
Conclusions ........................................................................................................................ 21
6.
Mercury Emissions: Lignite Coals ...................................................................................22
6.1
North Dakota Mines and Generating Units ......................................................................... 22
6.2
Texas Gulf Coast Mines and Generating Units .................................................................... 27
6.3
Role of Flue Gas SO3 .......................................................................................................... 30
6.3.1 EIA Hg, Sulfur Relationship ................................................................................................... 30
6.3.2 SO3: Inhibitor to Hg Removal ................................................................................................ 31
6.4
EPA Cost Calculations Ignore FGD ....................................................................................... 32
6.5
Conclusions ........................................................................................................................ 33
7.
Mercury Emissions: Non-Low Rank Fuels .......................................................................34
7.1
Hg Removal ........................................................................................................................ 34
7.2
Role of Fuel Composition and Process Conditions .............................................................. 36
7.2.1 Coal Variability ...................................................................................................................... 36
7.2.2 Process Conditions................................................................................................................ 37
7.3
Conclusions: Mercury Emissions - Non-Low Rank Coals ...................................................... 38
8.
EPA IPM RESULTS: EVALUATION AND CRITIQUE .............................................................39
8.1
IPM 2030 Post-IRA 2022 Reference Case: A Flawed Baseline .............................................. 39
8.1.1 Analytical Approach .............................................................................................................. 39
(Page 185 of Total)
ii
369a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 60 of 204
8.1.2 Coal Retirements .................................................................................................................. 40
8.1.3 Coal CCS ................................................................................................................................ 44
8.1.4 Coal to Gas Conversions (C2G) ............................................................................................. 44
8.2
Summary ........................................................................................................................... 44
Appendix A: Additional Cost Study Data ................................................................................45
Appendix B: Example Data Chart ...........................................................................................48
*****************************************
List of Tables
Table 5-1. Summary of EPA Results ............................................................................................ 16
Table 5-2. ESP Rebuild Costs: Four Documented Cases ............................................................ 18
Table 5-3. Summary of Results: Industry Study........................................................................... 20
Table 6-1. Hg Variability for Select North Dakota Reference Stations ....................................... 26
Table 6-2. Hg Variability for Select Texas Reference Stations .................................................... 29
Table 8-1. Coal Retirement Errors ................................................................................................ 40
Table 8-2. IPM Coal Retirement Errors: 2028 Post-IRA 2022 Reference Case Run .................. 41
Table 8-3. IPM Coal Retirement Errors: 2030 Post IRA 2022 Reference Case Modeling Run... 42
Table 8-4 Units in the NEEDS to Be Operating in 2028 .............................................................. 42
Table 8-5 Units IPM Predicts CCS By 2030 ............................................................................... 43
Table 8-6 Units IPM Erroneously Predicts Switch to Natural Gas ............................................. 43
Table A-1. Technology Assignment for 0.010 lbs/MBtu PM Rate: Industry Study .................... 46
Table A-2 Technology Assignment for 0.006 lbs/MBtu PM Rate: Industry Study .................... 47
*****************************************
List of Figures
Figure 3-1. Inventory of EPA-Project 2028 Fleet by Control Technology Suite .......................... 6
Figure 3-2. Numbers of Quarters Sampled by EPA for Use in PM Database ................................ 7
Figure 3-3. Coronado Generating Station: 20 Operating Quarters ............................................... 10
Figure 4-1. Fraction of Units Exceeding Three PM Rates: By Control Technology .................. 12
Figure 6-1. Mercury Content Variability for Eight North Dakota Lignite Mines ........................ 23
Figure 6-2. Fuel Sulfur Content Variability for Eight North Dakota Lignite Mines .................... 23
Figure 6-3. Fuel Alkalinity/Sulfur Ratio for Eight North Dakota Mines ..................................... 24
Figure 6-4. Spatial Variation of Hg in a Lignite Mine ................................................................. 25
Figure 6-5. Mercury Variability for Two Gulf Coast Sources: Mississippi, Texas...................... 27
Figure 6-6. Sulfur Variability for Mississippi, Texas Lignite Mines19.1 .................................... 28
Figure 6-7. Fuel Alkalinity/Sulfur Ratio for Mississippi, Texas Lignite Mines........................... 28
Figure 6-8. Lignite Hg and Sulfur Content Variability: 2021 EIA Submission ........................... 30
Figure 6-9. Sorbent Hg Removal in ESP in Lignite-Fired Unit: Effect of Injection Location ..... 32
Figure 7-1. Mean, Standard Deviation of Annual Hg Emissions: 2018 ....................................... 35
Figure 7-2. Mean, Standard Deviation of Annual Hg Emissions: 2018 ....................................... 35
Figure 7-3. Annual Average of Fuel Hg, Sulfur Content in Coal ................................................. 36
Figure A-1. Unit ESP Investment (per EPA’s Cost Assumptions): PM of 0.010 lbs/MBtu ....... 45
(Page 186 of Total)
iii
370a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 61 of 204
Summary: Flaws in EPA’s Approach
1. Summary of Flaws in EPA’s Approach
The following is a summary of flaws in EPA’s analysis, further described in detail in this report.
Particulate Matter (PM) Database
EPA’s database of PM emissions is inadequate. EPA attempts to capture typical PM emissions
by acquiring samples from 3 years – 2017, 2019, and 2021. For the vast majority of the units –
80% - EPA uses only 2 of the potentially available 12 quarters (in those 3 years; up to 20
quarters from 2017 to 2021) of data to construct the PM database. Further, of these limited
samples. EPA cites the lowest to reflect a target PM emissions rate. EPA cites the use of the
“99th percentile” PM rate in lieu of the average compensates for variability; but this approach
accounts for variability within a single (“the lowest”) quarter. It fails to account for long-term
variability, which is affected by changes in fuel and process conditions, among others.
Lack of Design and Compliance Margin
EPA recognizes the need for margin in both design and operation (for compliance) of
environmental control equipment, but ignores this concept in developing this proposed rule. The
need for design margin is recognized in a 2012 OAQPS memo 1 addressing the initial
developments of this very same rule, while margin for operation is considered in evaluating
CEMS calibration2 for this proposed rule. Neither design nor operating margin is considered in
setting target PM standards, resulting in underestimation of number of units affected and total
costs to deploy control technology. For some owners of fabric filter-equipped units, the revised
rate of 0.010 lbs/MBtu eliminates any operating margin.
Inadequate Cost for ESP Rebuild
Of three categories of ESP upgrades considered by EPA, the cost for the most extensive – a
complete rebuild to add collecting plate area – is inadequate. Four such major ESP rebuild
projects have been implemented for which costs are reported in the public domain – and not
acknowledged by EPA. Incorporating these results elevates the range of cost from EPA’s
estimate of $75-100/kW to $57-213/kW. Consequently, the “average” cost for this action used
in the cost per ton ($/ton) evaluation increases from $87/kW to $133/kW.
Hutson, N., National Emission Standards for Hazardous Air Pollutants (NESHAP) Analysis
of Control Technology Needs for Revised Proposed Emission Standards for New
Source Coal-fired Electric Utility Steam Generating Units, Memo to Docket No. EPA-HQ-OAR—20090234, November 16, 2012. Hereafter Hutson 2012.
2
Parker, B., PM CEMS Random Error Contribution by Emission Limit, Memo to Docket ID No. EPAHQ-OAR-2018-0794, March 22, 2023. Hereafter Parker 2023.
1
(Page 187 of Total)
1
371a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 62 of 204
Summary: Flaws in EPA’s Approach
Inadequate $/ton Removal Cost
As a consequence of under-predicting capital required for ESP “rebuild,” and not recognizing the
need for a design and operating margin, EPA under-predicts the number of units requiring
retrofit and incurred cost. As a result, in contrast to the annual cost of $169.7 M projected by the
Industry Study described in this report, EPA estimates a range from $77.3 to $93.2 M. Further,
the Industry Study estimates the cost per ton ($/ton) of fPM to be $67,400, 50% more than the
maximum cost estimated by EPA - $44,900 /ton.
Faulty Lignite Hg Rate Revision
EPA’s proposal to lower the Hg emission rate for lignite-fired units to 1.2 lbs/TBtu is based on
improper interpretation of Hg emissions data – both in terms of the mean rate and variability.
EPA’s projection that 85 and 90% Hg removal would be required for the proposed rate is
incorrect, with up to 95% Hg removal required for some units – a level of Hg reduction not
feasible in commercial systems. In addition to the variability of Hg content in lignite, EPA
ignores the deleterious role of flue gas SO 3 in lignite-fired units, which compromises sorbent
performance and effectiveness – even though this latter barrier is recognized and cited by EPA’s
contractor for the IPM model.3
Faults in IPM Modeling
IPM creates a flawed Baseline scenario that does not adequately measure the impacts of the
proposed rule. Most notably, IPM err in the number of coal units that would be retired in both
2028 and 2030; as a consequence, EPA underestimates the number of units subject to the
proposed rule. Also, IPM unrealistically retrofitted 27 coal units with carbon capture and storage
(CCS) in 2030. Consequently, IPM modeling results of the Baseline likely understate the
compliance impacts of the proposed rule.
3
IPM Model – Updates to Cost and Performance for APC Technologies: Mercury Control Cost
Development Methodology, Prepared by Sargent & Lundy, Project 12847-002, March 2013.
(Page 188 of Total)
2
372a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 63 of 204
Introduction
2.
Introduction
The Environmental Protection Agency (EPA) is proposing to amend the National Emissions
Standards for Hazardous Air Pollutants (NESHAP) for Coal- and Oil-fired Electric Utility Steam
Generating Units (EGUs), otherwise known as the Mercury and Air Toxics Standards (MATS).
The specific emissions limits being revised address the filterable particulate matter (fPM)
standard (which is the surrogate standard for non-mercury (Hg) metal HAPs); the Hg standard
for lignite-fired units; fPM measurement methods for compliance; and the definition of startup.
This report provides a review and evaluation of EPA’s approach to selecting the revised fPM
standard, the capital and annual costs for achieving the proposed revised standard, and the cost
per ton ($/ton) to control non-Hg metal HAPs; and a critique of EPA’s basis for proposing an Hg
limit of 1.2 lbs/TBtu for lignite-fired units. This document also provides information supporting
EPA’s decision to retain the present Hg limit for bituminous and subbituminous coal.
The proposal to lower fPM and Hg limits is premised on EPA’s interpretation of data related to
the cost and capabilities of PM and Hg emission control technologies. EPA reports to have
conducted realistic assessments of PM and Hg emissions and control technology capabilities in
support of their analysis. EPA’s assumptions are reported in the
MATS_RTR_Proposal_Technology Review Memo4 where EPA describes the PM database they
developed, the cost and control capabilities of upgrades to electrostatic precipitators (ESPs) and
fabric filters, and their understanding of the key factors that affect Hg emissions in bituminous,
subbituminous, and lignite coal - and how the latter are alike or differ.
Many of EPA’s assumptions are contrary to data in their possession or strategies previously
adopted by EPA, but not considered. EGUs have been reporting fPM compliance data to EPA
since MATS became applicable to them – i.e., for the vast majority of EGU, April 2015 or April
2016 for units that obtained a one-year extension. However, EPA’s effort to “mine” fPM
emissions data from prior years provides a sparse, inadequate database that does not reflect
operating duty nor account for inevitable variability; further EPA misinterprets this information.
No design or operating margins are considered in setting fPM (the same is true for lignite Hg
emission rates). The cost to upgrade ESPs to meet the proposed limits is inadequate for the most
significant modification EPA envisions – the complete ESP Rebuild. The cost to deploy
enhanced operating and maintenance (O&M) actions on existing fabric filers is inadequate.
Regarding revised Hg limits for lignite coal, EPA does not recognize the differences in lignite
versus Powder River Basin (PRB) subbituminous coal that effect Hg control. EPA draws an
incorrect analogy between PRB and lignite, improperly assuming the Hg removal by carbon
sorbent observed with PRB can be replicated on lignite.
Benish, S. et. al., 2023 Technology Review for the Coal- and Oil-Fired EGU Source Category,
Memo to Docket ID No. EPA-HQ-OAR-2018-0794. January 2023. Hereafter RTR Tech Memo.
4
(Page 189 of Total)
3
373a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 64 of 204
Introduction
The remaining sections of this report detail the findings summarized in Section 1, and are as
follows:
•
•
•
•
•
•
•
Section 3 describes EPA’s approach to assembling their fPM database, and the flaws and
weaknesses in their approach.
Section 4 evaluates the fPM rates assigned by the database for the EPA analysis.
Section 5 evaluates EPA’s cost bases for the proposed fPM revised standard, and
compares these to the realistic assumptions used in the Industry Study described in the
paper.
Section 6 addresses EPA’s proposal to lower Hg from lignite-fired units to 1.2 lbs/TBtu,
delineating the shortcomings in EPA’s approach and assumptions.
Section 7 provides historical data for Hg emission from non-low rank fuels, showcasing
the inherent variability in the 30-day rolling average.
Section 8 reviews the IPM modeling analysis conducted by EPA to support this rule.
Appendix B presents examples of PM emission timelines for a limited number of units 5
that show how EPA’s sparse database does not capture the authentic “PM signature” of
the units.
5
We reviewed data for a limited number of units because the comment period was very short and did not
allow adequate time to undertake a more thorough review. EPA has all the data and in our opinion should
have conducted such an analysis for every unit at issue.
(Page 190 of Total)
4
374a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 65 of 204
Description of EPA Reference PM Database
3.
Description of EPA Reference PM Database
Section 3 describes the PM database assembled by EPA which serves as the basis for the
proposed NESHAP rule. Section 3 first describes the coal fleet inventory reflected, and then
identifies shortcomings of this database concerning (a) selection of the sample year and quarter,
(b) number of samples considered, and (c) data analysis.
3.1 Coal Fleet Inventory
EPA projects that a total of 275 generating units will be operating at the compliance date of
January 1, 2028, representing a reduction from the present (2023) operating inventory of
approximately 450 units. EPA identified the 275 units based on their estimate of unit retirements
and units planning to switch to natural gas by the compliance date. EPA accounted for these
assets not as individual units, but in terms of the number of reporting monitors to the Clean Air
Markets Division. As 27 units employ common stack reporting, the data presented by EPA in the
draft rule and RTR Tech Memo consider 248 discrete data points that reflect the 275 units. This
analysis will adopt the same reporting methodology.
EPA’s selection of 275 units contains 22 units that have publicly disclosed plans to retire or
switch to natural gas by the compliance date of January 1, 2028. For the purposes of this
analysis, these units are retained in the database so the results can be more readily compared.
Figure 3-1 depicts the installed inventory projected by EPA, presented according to the suite of
control technology. The first two bars (from the left) report units equipped with ESPs as the
primary PM control device in the following configurations: a total of 54,116 MW for an ESP
followed by a wet FGD; and a total of 16,346 MW with an ESP only. The next 3 bars describe
the total inventory equipped with a fabric filter in the following three configurations: 12,194
MW with the fabric filer as the sole device; 20,206 MW with a fabric filter followed by a wet
FGD, and 19,995 MW where the fabric filter is preceded by a dry FGD process. Consequently,
the bulk of the inventory (70,462 MW) will employ an ESP as part of the control scheme, with
52,395 MW employing a fabric filter for PM. Given the role of wet FGD in PM emissions – in
most cases such devices will reduce PM by approximately 50% - more than half (74,322 MW)
employ wet FGD as the last control step.
(Page 191 of Total)
5
375a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 66 of 204
Description of EPA Reference PM Database
60,000
Bituminous
Subbituminous
Lignite
Waste Coal
Environmental Control Inventory
50,000
40,000
30,000
20,000
10,000
0
ESP/Wet FGD
ESP Only
FF Only
FF/Wet FGD
FF/Dry FGD
Wet ESP
Figure 3-1. In
Inventory of EPA-Project 2028 Fleet by Control Technology Suite
3.2 Database Characteristics
Several characteristics of EPA’s database severely compromise the quality of the analysis. These
are the (a) selection of sampling year and quarter and (b) number of samples used.
3.2.1 Selection of Sample Year and Quarter
EPA does not describe the rationale for the limited data selected. The selection of three reference
years (2017, 2019, and 2021) from at least 5-6 years of data readily available to EPA, and the
sampling periods within each year (typically the 1 st or the 3rd quarter even though all quarters are
generally available) are not discussed. EPA extracts data from the year 2021 using a different
approach from the years 2019 and 2017 without explanation. EPA states for 2021 that 2 quarters
of data are utilized (always the 1st and the 3rd). For 2019, EPA reports utilizing data from
“quarters three and occasionally four” while for 2017 EPA reports data acquired from “variable
quarters.”6
The rationale for the irregular selection of quarters is not stated. For 2021, the first and third
quarters are selected with no technical basis. For 2019, the selection of quarters three and
“occasionally” four does not replicate the time periods selected for 2021. For 2017, there is no
description of the quarters or selection criteria.
EPA ignores a rich field of data that could support a much more robust and reasonable analysis.
6
RTR Tech Memo, page 2.
(Page 192 of Total)
6
376a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 67 of 204
Description of EPA Reference PM Database
3.2.2 Number of Samples
The number of discrete data points in EPA’s Reference Database – defined by the number of
operating quarters – is extremely limited. EPA’s description of the sampling approach7 is as
follows:
Quarterly data from 2017 (variable quarters) and 2019 (quarters three and occasionally four)
were first reviewed because data for all affected EGUs subject to numeric emission limits had
been previously extracted from CEDRI. In addition, the EPA obtained first and third quarter
data for calendar year 2021 for a subset of EGUs with larger fPM rates (generally greater than
1.0E-02 lb/MMBtu for either 2017 or 2019).
Figure 3-2 shows most monitor locations — 193 of the 245 — are characterized by only 2
quarters of data, which is inadequate compared to the 16 or 20 EPA has access to. The
distribution of quarters selected by EPA according to either CEMS or stack test measurement for
all 245 locations is shown. The second largest category is 33 units characterized by 4 quarters.
Figure 3-2. Numbers of Quarters Sampled by EPA for Use in PM Database
7
RTR Tech Memo, page 2.
(Page 193 of Total)
7
377a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 68 of 204
Description of EPA Reference PM Database
Additional depictions of the data (not shown) reveal that only nine units are described by data in
2017, and 187 units by data from 2019. Only 41 units are described by data in 2021; the lack of
data in 2021 was intentional as EPA considered this year only if data from 2017 or 2019 showed
the unit exceeding the 0.010 lbs/MBtu proposed limit. 8 In other words, EPA looked at 2021 only
when it was trying to find an emission rate less than 0.010 lbs/MBtu for a unit.
3.2.3 PM Data Selection and Analysis
EPA does not explain the methodology chosen to reflect each quarters’ emission rate, using at
least two methods, depending on the year. EPA followed a four-step process to construct its
database to select the “base rate” for each unit. The process is described as follows:
Step 1: Quarter Selection. EPA looked at 2-4 (usually 2) quarters for each unit. EPA states:
“Quarterly data from 2017 (variable quarters) and 2019 (quarters three and occasionally four)
were first reviewed …. In addition, the EPA obtained first and third quarter data for calendar
year 2021 for a subset of EGUs with larger fPM rates (generally greater than 1.0E-02 lb/MMBtu
for either 2017 or 2019).”9
As noted previously, EPA considered Q1 and Q3 2021 data solely to find a PM rate lower than
0.010 lb/MMBtu, and further explained: “The quarterly 2021 data summarizes recent emissions
and also reflect the time of year where electricity demand is typically higher and when EGUs
tend to operate more and with higher loads.”10
Step 2. Select Single Quarter. From the candidate quarters identified in Step 1, EPA selected a
single value, using criteria specific for each tests methodology:
•
•
PM CEMS: for quarters in 2017 and 2019, EPA selected the 30-day average observed on
the last day of the quarter; for quarters in 2021, EPA determined the average of the 30day rolling averages observed in that quarter.
Stack Tests: EPA took the average of the multiple (usually 3) test runs.
Step 3. Select Lowest Quarter. EPA selected the “lowest quarter” PM rate from the quarters
selected in Step 2.
Step 4. Determine PM of 99th Percentile. For this lowest quarter per Step 3, EPA calculated the
statistical percentile values as observed over the entire quarter. The methodology varied on
whether PM CEMS or stack test data was provided. For PM CEMS, the percentiles were
calculated for all 30-day rolling averages in the quarter. For stack tests, the percentiles were
calculated for the typically 3 test runs.
8
Personal communication: Sarah Benish to Liz Williams, April 28, 2023. “Data for 2021 was mined
only for the EGUs that showed 2017 or 2019 fPM data above 1.0E-02 lb/MMBtu. We did not mine 2021
PM data for EGUs not expected to be impacted by the proposed fPM limit.”
9
RTR Memo, page 2.
10
Ibid.
(Page 194 of Total)
8
378a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 69 of 204
Description of EPA Reference PM Database
The results are reported in Appendix B of the Technology Review Memo. The 99 th percentile
rate was chosen as the “base rate,” supposedly to account for variability within the “lowest
quarter.”
EPA does not describe why data selected was restricted to the years 2017, 2019, and 2021. EPA
does not explain why 2021 data was limited to the 1 st and 3rd quarters, 2019 data was limited to
the 3rd and occasionally the 4th quarter, while 2017 data from variable quarters could be utilized.
Of concern is the limited subset of data used for this analysis – Figure 3-2 showed that for 80%
of the units the lowest is selected from only two samples. EPA states “By using the lowest
quarter’s 99th percentile as the baseline, the analyses account for actions individual EGUs have
already taken to improve and maintain PM emissions.” 11 EPA states employing the PM rate at
the 99th percentile –reflecting approximately the highest data within that quarter – remedies any
bias.12
There is no basis for this statement. EPA is assuming that because a unit emitted fPM during a
single quarter at a particular level, the lowest such level must necessarily reflect “actions
individual EGUs have already taken to improve and maintain PM emissions,” and therefore each
EGU must be able to replicate that rate in every quarter going forward, indefinitely. Also, EPA
ignores the unavoidable variability in emission rates: the “actions individual EGUs have already
taken to improve and maintain PM emissions” are not the only factor that determines fPM
emissions rate. The factors that affect fPM rates are numerous and include but are not limited to
the following: coal quality (e.g., chemical composition and ash content) which varies within a
single mine; variation in temperature within an ESP; content of SO 3 and trace constituents that
determine ash electrical resistivity; physical conditions (spacing) of collecting plates and
emitting electrodes; effectiveness of the rapping “hammers” that dislodge collected ash from the
collecting plates; and physical properties of the collected ash layer that define ash reentrainment. Further, boiler operation will influence ESP performance, most notably unit duty
(i.e., relatively stable operating level for a “baseload” unit versus more load changes for an
intermediate unit or a unit operating in peaking mode), operating level, and load “ramp” rate.
Achieving the “least emission” rate observed during a quarter that EPA selected is not
necessarily feasible at other times and under other conditions.
3.2.4 Example Cases
Figure 3-3 presents an example that demonstrate the shortcomings of EPA’s approach. Figure 33 presents PM data from Coronado Generating Station Units 1 and 2 reflecting all operating
quarters from 2017 through 2021. Both the average PM rate and the 99 th percentile from each
quarter are presented for 20 quarters of operation over the 4-year period. Figure 3-3 also
identifies the two samples EPA selected from 2017 Q3 and 2019 Q3 as representative of low
fPM rate, with the latter as the “least” – and the 99th-percentile reporting 0.0086 lbs/MBtu.
Figure 3-3 shows EPA’s two samples do not capture the full character of Coronado operating
duty (with the red dotted line denoting the PM rate selected as representative of the units’
11
12
RTR Tech Memo, page 4.
Ibid.
(Page 195 of Total)
9
379a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 70 of 204
Description of EPA Reference PM Database
capabilities to control PM). These quarters as selected by EPA are far from representative of unit
operations or capabilities: among 20 quarters for which data are available, the units’ 90th
percentile fPM rates exceed the 0.0086
86 lbs/MBtu rate EPA selected for 16 quarters. Ten out of
th
20 quarters showed 90 percentile fPM rates exceeded the proposed standard of 0.010 lb/MBtu.
0.025
Mean PM Rate
99th Percentile Rate
PM Emission Rate (lbs/MBtu)
0.020
0.015
EPA Reference
Quarter
EPA Reference
Quarter
0.010
0.005
0.000
17Q1 17Q2 17Q3 17Q4 18Q1 18Q2 18Q3 18Q4 19Q1 19Q2 19Q3 19Q4 20Q1 20Q2 20Q3 20Q4 21Q1 21Q2 21Q3 21Q4
2017
2018
2019
2020
2021
Figure 3-3. Coronado Generating Station: 20 Operating Quarters
Coronado Units 1/2 show how selecting the least PM rate of any quarter, and adopting the 99 th
percentile PM rate within that quarter, does not capture the variability in fP
fPM emission rates,
which are affected by the variability of coal and operating conditions, among others. These
examples demonstrate that EPA used best-case fPM data from both compliance measures
(continuous monitor and performance test data).
Additional examples are presented in the Appendix B to this report.
3.3
Conclusions
•
EPA’s database is sparse and does not fully capture operating duty. Of the 275 units and
approximately 250 monitoring locations, the vast majority – 80
80% - are characterized by
only two samples.
•
Selecting the lowest quarter - “one” of what in most cases are “two” samples - fails to
capture the operating profile of the unit, and presents a serious deficiency in representing
(Page 196 of Total)
10
380a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 71 of 204
Description of EPA Reference PM Database
operations. EPA’s approach of considering the 99th percentile within a quarter is
inadequate to assess variability, particularly that induced by fuel composition, as such
fuel changes are observed over a characteristic time of years and not several months.
•
The use of statistical means within one quarter does not capture the multi-month
variances in coal composition, seasonal load, and process conditions that are not
constrained to 3-month events.
•
An improved, robust database would allow observing variation between– as opposed to
within – operating quarters, to better reflect variations and uncertainties in operating duty
and fuel supply.
(Page 197 of Total)
11
381a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 72 of 204
Coal Fleet PM Emissions
4. Coal Fleet PM Emissions Characteristics
Section 4 characterizes
es the coal-fired fleet selected to represent the PM emissions
The emission control technologies on the 275 units projected by EPA to be operating in 2028
present a variety of approaches to lower fPM emission limits – with implications for upgrades
and actions that would be required to meet a revised standard for fPM. This subsection presents
the distribution of control technology by ability to operate below the revised PM limits for the
units in EPA’s database. By necessity, this analysis uses EPA’s database (both for a discussion
of expected or achievable fPM emission rates and the units projected to operate in 2028 and
later), and such use does not represent an endorsement or acceptance of EPA’s approach. As
discussed above, EPA’s analysis of expected/achievable fPM emission rates is inadequate. And
as discussed later in this report, EPA’s selection of units that would continue to operate after
2028 is flawed: it contains multiple errors; and EPA’s post-IRA IPM analysis is inaccurate.
Figure 4-1 is used to present our analysis.
Figure 4-1. Fraction of Units Exceeding Three PM Rates: By Control Technology
Figure 4-1 presents for five control technology configurations the percentage of units that emit
(according to EPA’s chosen “base rate”) above the following PM emission limits: 0.015
lbs/MBtu, 0.010 lbs/MBtu, and 0.006 lbs/MBtu. The control technologies are (a) dry FGD with a
fabric filter, (b) ESP followed
ed by a wet FGD, (c) fabric filter alone (employing low sulfur coal or
multi-unit station-averaging to meet an SO2 limit), (d) wet ESP as the last control device, (e) ESP
(Page 198 of Total)
12
382a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 73 of 204
Coal Fleet PM Emissions
alone (employing low sulfur coal or multi-unit station-averaging to meet an SO 2 limit), and (f)
fabric filter followed by a wet FGD.
In Figure 4-1, the proportion of units in the inventory that exceed the contemplated fPM rate is
proportional to the height of the bar; a higher bar implies a greater fraction of units in the
inventory exceed the contemplated fPM rate. Thus:
4.1.1 PM Rate of 0.015 lbs/MBtu
Units in three categories exceed this highest contemplated rate – those with an ESP alone, a dry
FGD followed by a fabric filter, and an ESP followed by a wet FGD. The latter category of
ESP/wet FGD benefits in that actions within the absorber tower – although not designed to
removed fPM – can under some conditions remove fPM. Data describing PM removal via wet
FGD is sparse but suggests 50% removal can be observed.
4.1.2 PM Rate of 0.010 lbs/MBtu
The number of units in each of the three preceding categories exceeding this rate increases –
there is no change for the category of ESP-alone, but the number of units exceeding this rate
more than triple for dry FGD/fabric filter and ESP/wet FGD. No units with fabric filter/wet FGD
or a wet ESP emit at greater than this rate.
4.1.3 PM Rate of 0.006 lbs/MBtu
The number of units exceeding a rate of 0.006 lbs/MBtu increases with this most stringent
contemplated rate. More than 1/3 of the units with ESP/wet FGD and ¼ of ESP- only cannot
meet this rate, with fabric filters either operating with dry FGD (20%) or alone (16%) not
achieving this target. Almost 20% of those with fabric filter/wet FGD units emit greater than this
value.
In conclusion, within six major categories of control technology, units equipped with fabric
filters achieve the lowest PM rates. Units with ESPs – either operating alone or with a wet FGDrepresent the highest fraction of their population that exceed the strictest contemplated rate.
Units with fabric filters – operating alone, or as part of a wet or dry FGD arrangement – are
among the lowest exceeding the strictest contemplated PM rate. As noted previously, this
analysis used EPA’s database (as reflected in Appendix B of the RTR Tech Memo) out of
necessity, and such use does not represent an endorsement or acceptance of EPA’s approach.
(Page 199 of Total)
13
383a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 74 of 204
Critique of Cost-Effectiveness Calculations
5. CRITIQUE OF COST-EFFECTIVENESS CALCULATIONS
Section 5 addresses the cost effectiveness ($/ton basis) estimated to reduce the PM emission rate
to EPA’s proposed limit of 0.010 lbs/MBtu, and the alternative limit of 0.006 lbs/MBtu. EPA
has conducted this calculation with inputs based on analysis by Sargent & Lundy (S&L) 13 and
Andover Technology Partners (ATP).14 EPA’s results are presented in both Table 3 of the
proposed rule and in Table 7 of the RTR Tech Memo.
This section reviews EPA’s calculation methodology, critiques inputs of the EPA Study, and
presents results of an Industry Study that utilizes realistic costs. Results from EPA’s evaluation
and the Industry Study addressing the 0.010 lbs/MBtu and 0.006 lbs/MBtu PM rates are
compared.
5.1 EPA Evaluation
5.1.1 EPA Study Inputs
The EPA study used both the PM database described in Section 3 and cost and technology
assumptions derived by the above-mentioned S&L and ATP references. As noted in Section 2,
EPA’s sparsely-populated database is inadequate from which to base a revised PM rate that
represents a significant reduction in PM emissions but is achievable in long-term duty.
The analyses by S&L and ATP provide capital cost for three categories of ESP upgrades,
improvements to fabric filter operating and maintenance (O&M) and associated costs, capital
requirement for fabric filter retrofit and associated O&M cost. Most of the analysis is premised
on the costs and PM removal performance of ESP upgrades as defined by S&L. It should be
noted S&L did not provide specific projects with publicly available data as the basis of their
assumptions.
The most significant shortcoming of EPA’s assumptions is low capital estimates for the most
significant ESP upgrade - the “ESP Rebuild” scenario. In contrast to the generalizations of the
S&L memo, Table 5-2 reports publicly documented costs incurred for “ESP Rebuild.” Equally
significant, EPA ignores the inherent variability of fPM and FGD process equipment by not
utilizing a design or operating margin in selecting the value of fPM rates that would require
operator action. This is counter to EPA’s prior acknowledgement of the use of margin in the
initial rulemaking for MATS15 and recent observations as to CEMS calibration. 16 It is also
contrary to basic operation goals: no source operates at the applicable standard; a compliance
13
PM Incremental Improvement Memo, Project 13527-002, Prepared by Sargent & Lundy, March 2023.
Hereafter S&L PM Improvement Memo.
14
Analysis of PM Emission Control Costs and Capabilities, Memo from Jim Staudt (Andover
Technology Partners) to Erich Eschmann, March 22, 2023. Hereafter ATP 2023.
15
Hutson 2012.
16
Parker 2023.
(Page 200 of Total)
14
384a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 75 of 204
Critique of Cost-Effectiveness Calculations
margin is always necessary, at least to account for unavoidable variability of performance in the
real world. By ignoring the need for margin, EPA’s evaluation under-predicts the number of
units that would be retrofit with new or upgraded control technology to meet the target rate.
These and other critiques of EPA’s approach are discussed subsequently.
Shortcomings in EPA inputs compromise the results of their analysis. These shortcomings, as
well as other observations, are summarized as follows:
ESP Upgrade. Three categories of ESP upgrade are proposed by EPA. The most significant
shortcoming relates to the “ESP Rebuild” category in which - as described by S&L – additional
plate area is added to the ESP. The addition of collecting surface area will require major changes
to – or demolition and complete rebuilding of – the gas flow confinement that houses the existing
collecting plates. Also, these process changes require specialized labor for fabrication and
installation that may be limited in availability. The costs suggested by S&L (without citation of
references) - $75-100/kW –are low when compared to publicly disclosed costs from similar
projects.
Fabric Filter O&M. Fabric-filter-equipped units that emit greater than 0.010 lbs/MBtu are
assumed to adopt enhanced O&M practices. These enhanced practices consist of (a) upgrading
filter material to higher quality fabrics, such PTFE, and (b) increasing the replacement frequency
so that filters are replaced on a 3-year basis. The cost premium for this action, based on analysis
by ATP, does not consider the additional manpower costs for the more frequent replacement.
Fabric Filter Construction. EPA’s range of capital cost for retrofit of fabric filter technology is
consistent with industry experience.
Design/Compliance Margin. A premise of environmental control system design is accounting for
variability due to many factors, including, for example, variations in fuel composition, operating
load, and process conditions. Such variability is generally addressed by a design/compliance
margin – selecting a target emission rate less than mandated by a standard. The concept of
design/compliance margin is broadly applied in the industry, and was acknowledged in a 2012
EPA memo summarizing the range of margin adopted by various process suppliers, with a
minimum cited as 20-30%.17 EPA did not adopt a design/compliance or operating margin in
selecting fPM emission rates for a revised fPM standard in this evaluation, despite the fact that
elsewhere in the record of this proposal EPA acknowledges a typical “operational target” of 50%
of the limit.18 Because of its assumption of no design/compliance margin whatsoever, EPA
presumes that units that report an operating fPM of 0.010 lbs/MBtu – based on EPA’s sparse
database - require no investment to meet the proposed standard of 0.010 lb/MBtu.
Hutson, N., National Emission Standards for Hazardous Air Pollutants (NESHAP) Analysis
of Control Technology Needs for Revised Proposed Emission Standards for New
Source Coal-fired Electric Utility Steam Generating Units, Memo to Docket No. EPA-HQ-OAR—20090234, November 16, 2012.
18
Parker 2023.
17
(Page 201 of Total)
15
385a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 76 of 204
Critique of Cost-Effectiveness Calculations
Separate from the preceding issues, EPA did not disclose the capacity factors assumed in the
analysis. The capacity factor can be inferred from the tons of PM removed as reported in
Appendix B of the RTR Tech Memo; this requires acquiring heat input and net plant heat rate
from AMPD and EIA data.
5.1.2 EPA Results
Table 5-1 presents results of EPA’s evaluation.
Table 5-1. Summary of EPA Results
EPA Study
Annual Cost $/ton
($M/y)
fPM
(average)
Non-Hg
metallic HAPS
Removed
(tons)
$/ton
non-Hg metallic
HAP
($000s)
Target: 0.010 lbs/MBtu
20
2,074
77.3-93.2
37,30044,900
6.34
12,200-14,700
Target: 0.006 lbs/MBtu
65
6,163
633
103
24.7
25,600
Unit
Affected
Tons fPM
Removed
Proposed Limit: 0.010 lbs/MBtu. EPA estimates 20 units in the entire inventory are required to
retrofit some form of ESP upgrade. The number of units with existing fabric filters required to
enhance O&M is not identified, nor is their cost. EPA estimates a range in annual cost to
implement the ESP and fabric filter O&M enhancement of $77.3 to 93.2 M/yr, with the range
determined by the range in cost and performance of each option as described by S&L. 19 This
total annualized cost translates into an average fPM removal cost effectiveness of $37,300 $44,900 per ton of fPM and $12.2M -$14.7 M per ton of total non-Hg metallic HAPs. These
steps remove a total of 2,074 tons of fPM (6.34 tons of total non-Hg metallic HAPs) annually.
EPA did not consider in its analysis the potential impact of the capital cost of major controls
construction or upgrades (i.e., ESP rebuilds for most of the 20 units; new Fabric Filters for the
two Colstrip units) on the viability of the units at which such rebuilds would occur. Appendix
Figure A-1 presents the capital required for each unit as designated by EPA for upgrade –
requiring an investment likely prohibitive for continued operation.
Potential Limit: 0.006 lbs/MBtu. EPA estimates 65 units in the entire inventory are required to
retrofit a fabric filter or deploy enhanced O&M to an existing fabric filter. EPA estimate an
annual cost of $633 M/yr will be incurred, at an average cost effectiveness of $103,000 per ton
19
S&L PM Improvement Memo.
(Page 202 of Total)
16
386a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 77 of 204
Critique of Cost-Effectiveness Calculations
of fPM and $25.6 M per ton of total non-Hg metallic HAPs. These steps remove a total of 6,163
tons of fPM (24.7 tons of total non-Hg metallic HAPs) annually.
5.2 Industry Study
The Industry Study alters several assumptions to reflect actual, documented cost data and the
necessity of a design/compliance margin. Table 5-2 presents these results.
5.2.1 Revised Cost Inputs
The modified cost inputs necessary to reflect authentic conditions ESP upgrade and fabric filter
operation are discussed as follows.
ESP Upgrades. The three categories of ESP upgrades are assessed as follows.
Minor Upgrades (Low Cost). Both the cost range and PM removal efficiency for this activity as
estimated by S&L are adopted for this analysis. ESPs requiring Minor Upgrade are assigned a
$17/kW cost to derive an average of 7.5% removal of fPM.
Typical Upgrades (Average Cost). Both the cost range and PM removal efficiency for this
activity as estimated by S&L are adopted for this analysis. ESPs requiring Typical Upgrade are
assigned a $55/kW cost to derive an average of 15% fPM removal.
ESP Rebuild (High Cost). The cost range for this activity as estimated by S&L does not reflect
that reported publicly for four projects that represent the “ESP Rebuild” category. Two projects
were completed at the AES Petersburg station – the complete renovation of the ESPs on Units 1
and 420 for which S&L provided engineering services. The cost for this work has been publicly
reported in 2016-dollar basis. Two additional major ESP upgrades were implemented by
Ameren at the Labadie station unit in 2014 – with costs publicly reported. 21
Table 5-2 summarizes the cost incurred for the four major ESP retrofits, including costs in the
year incurred and escalated (using the Chemical Engineering Process Cost Index) 22 to 2021.
Table 5-1 shows a cost range of $57-209/kW, with 3 of the 4 units incurring a cost exceeding
$100/kW. These costs significantly exceed EPA’s maximum for this range.
20
State of Indiana – Indian Public Utility Commission, Cause No. 44242, August 14, 2013. See
Appendix, electronic page 50 of 51.
21
Ameren Missouri Installs Clean Air Equipment at its Labadie Energy Center;
https://ameren.mediaroom.com/news-releases?item=1351
22
https://www.chemengonline.com/pcihome#:~:text=Since%20its%20introduction%20in%201963,from%20one%20period%20to%20another.
(Page 203 of Total)
17
387a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 78 of 204
Critique of Cost-Effectiveness Calculations
Table 5-2. ESP Rebuild Costs: Four Documented Cases
Owner/Station
AES/Petersburg
AES/Petersburg
Ameren Labadie
Ameren Labadie
Unit
1
4
1
2
Basis Year
2016
2016
2014
2014
2021 ($/kW)
117
57
192
209
Consequently, the range of ESP rebuild costs is adjusted to $57-209/kW, and the mean value of
$133/kW (2021 basis) selected to represent this category of upgrade. 23
FF O&M. A fabric filter O&M cost was derived for existing units, based on the assumption by
S&L that filter material will be upgraded, as well as the frequency of filter replacement. An
increase in cost – reflected as fixed O&M – of $515,000 is estimated for a 500 MW unit. This
cost premium is comprised of higher material cost of $425,000 to upgrade filter material to PTFE
fabric and an additional $90,000 for installation labor. This cost premium as is assigned to
existing units based on generating capacity, and using a conventional “6/10 th” power law.
The revised Industry Study costs are based on (a) gas flow volume treated, (b) surface area of
filter required based on the unit design, (c) unit cost of filter (e.g. $ per ft 2 of cleaning surface),
and (d) replacement rate of filter material. Gas flow treated for each unit was determined using
the quantitative relationships derived by S&L for fabric filter cost evaluation developed for the
IPM model.24 Filter surface area was not defined for each unit as dependent on the specific
air/cloth ratio; rather a fleet air/cloth ratio of 5 – a mean value between conventional and pulsejet design concepts – is selected. The unit cost for fabric was selected (at $4.00/ft 2) per ATP
analysis. Per S&L’s IPM fabric filter costing procedure25 and the EPA-sponsored review of filter
material cost,26 the increase in cost for enhanced O&M is derived. The cost to upgrade material,
accelerate filter replacement (from 5 to 3 years) and supporting cages (from 9 to 6 year) intervals
is estimated as $425K per year for a reference 500 MW unit.
Fabric Filter Capital Cost. EPA proposed a capital cost to retrofit a fabric filter as $150$360/kW. The cost range offered by EPA is consistent with industry experience and is used in
this study.
EPA did not share the incremental operating cost incurred by the retrofit fabric filters. The
Industry Study adopted fixed and variable operating costs from the previously cited S&L fabric
filter cost estimating procedure. For the assigned inputs, the S&L evaluation projects a fixed
23
Colstrip Units 3 and 4 are equipped with legacy FGD that combine removal of SO2 and PM in a wet
venturi; there is not an ESP option to upgrade. Fabric filer retrofit is the only option; as Colstrip
represents an atypical case the costs are reported in the category of Major ESP upgrade.
24
IPM Model – Updates to Cost and Performance for APC Technologies: Particulate Control Cost
Development Methodology, Project 13527-001, Sargent & Lundy, April 2017. Hereafter S&L Fabric
Filter 2017.
25
Ibid.
26
ATP report.
(Page 204 of Total)
18
388a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 79 of 204
Critique of Cost-Effectiveness Calculations
O&M of $0.27/kW-yr and a variable operating cost of 0.48 $/MWh. The variable O&M cost is
mostly comprised of filter replacement at the accelerated rate described, and auxiliary power.
Design/Compliance Margin. EPA in two public documents address – and apparently recognize –
the need for design/compliance margin. 27 The use of design/compliance margin was
acknowledged in a 2012 EPA memo summarizing the range adopted by various suppliers, citing
a minimum of 20-30%.28 For the proposed limit of 0.010 lbs/MBtu, the minimum of 20% is
used as a design target for ESP upgrades. Thus, the Industry Study applied ESP upgrade and
fabric filter O&M enhancements to attain 0.008 lbs/MBtu, in lieu of EPA’s target of 0.010
lbs/MBtu. It should be noted this 20% margin is the least of those considered; if the highest
operating margin of 50% suggested by EPA in the record of this rule was used the units requiring
upgrade and the cost would have been even higher.
As noted by EPA, the sole reliable compliance means for a 0.006 lbs/MBtu PM rate is a fabric
filter. Fabric filters historically exhibit low variability due to their inherent design; thus, the
operating margin is slightly relaxed to 0.005 lbs/MBtu. Consequently, the Industry Study
assumed ESP-equipped units emitting greater than 0.005 lbs/MBtu will retrofit a fabric filter to
insure 0.006 lbs/MBtu is attained. Units with existing fabric filters operating at greater than
0.005 lbs/MBtu will adopt improved operation and maintenance, as previously described.
5.2.2 Cost Effectiveness Results
Revised costs from the Industry Study are projected for the proposed fPM limit of 0.010
lbs/MBtu, and the alternative rate of 0.006 lbs/MBtu. Table 5-4 presents these results.
Proposed Limit: 0.010 lbs/MBtu. Results derived in the Industry Study are reported for all three
categories of ESP upgrade in Table 5-1. A total of 26 units are required to upgrade ESPs – 11
deploying Minor, 7 deploying Typical, and 8 deploying Major upgrades. 29 In addition, 11 units
equipped with fabric filters are required to enhance O&M activities. The totality of these actions
each year incur an operating cost of $169.7 M/yr, and remove 2,523 tons of PM.
27
Hutson, 2012 and Parker, 2023.
Hutson, N., National Emission Standards for Hazardous Air Pollutants (NESHAP) Analysis
of Control Technology Needs for Revised Proposed Emission Standards for New
Source Coal-fired Electric Utility Steam Generating Units, Memo to Docket No. EPA-HQ-OAR—20090234, November 16, 2012. at 1 (discussing mercury); 2 (discussing PM).
29
The two Colstrip units are equipped with an early generation FGD process which does not include an
ESP, thus the concept of an ESP upgrade is irrelevant. Consistent with EPA’s assumption, the Colstrip
units are assumed to retrofit a fabric filter as the only option to meet a limit of 0.010 lbs/MBtu.
28
(Page 205 of Total)
19
389a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 80 of 204
Critique of Cost-Effectiveness Calculations
Table 5-3. Summary of Results: Industry Study
Non-Hg
$/ton
Technology Annual Tons
metallic HAPS
fPM
fPM
Cost
(Units
($M/y) Removed average Removed (tons)
Affected)
$/ton
non-Hg metallic HAP
($000s)
Target: 0.010 lbs/MBtu
ESP Minor 20.9
100
(11)
ESP
34.7
282
Typical (7)
ESP Major 113.6
1,665
†
(8)
FF O&M
0.4
475
(11)
Total or
169.7
2,523
Average
Target: 0.006 lbs/MBtu
FF O&M
1.23
652
(23)
FF Retrofit 1,955.4 6,269
(52)
Total or
1,956.6 6,921
Average
209,340
0.31
67,470
122,926
0.86
40,216
68,228
5.1
21,662
869
1.45
284
67.3
7.71
22,000
1,887
2.61
617
311,900
25.13
102,000
282,715
27.74
92,470
† Includes 2 fabric filters retrofit to Colstrip Units 3 and 4. See footnote #23.
The incurred cost per ton varies significantly by ESP upgrade category. For the ESP Minor
upgrade, the average cost effectiveness is approximately $67,470,000 per ton of non-Hg metal
HAP for 0.31 of tons removed ($209,340 per ton of fPM for 100 tons of fPM removed). The
cost-effectiveness cost effectiveness for the ESP Typical upgrade average $40,216,000 per ton of
non-Hg metal HAP for 0.86 tons removed ($122,956 tons of fPM for 282 tons of fPM removed).
The Major upgrade removes the most non-Hg metal HAP – 5.1 tons – (1,665 tons of fPM) for an
average cost effectiveness of $21,662,000 per ton of non-Hg metal HAP ($68,228 per ton of
fPM). The most cost-effective control evaluated is enhanced fabric filter O&M, which removes
1.45 tons of non-Hg metal HAP at a cost-effectiveness of $284,230/ton (475 tons of fPM at a
cost-effectiveness of $869/ton).
These actions cumulatively remove a total of 2,523 tons of PM for an average cost effectiveness
of 22,000,000 per ton of non-Hg metal HAP ($67,262 per ton of fPM) removed, a 50% increase
compared to the cost estimated by EPA.
Appendix Table A-1 reports the units to which the Industry Study assigned ESP upgrades, and
defines the category of upgrade to meet the proposed fPM limit of 0.010 lbs/MBtu.
(Page 206 of Total)
20
390a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 81 of 204
Critique of Cost-Effectiveness Calculations
Possible Lower Limit: 0.006 lbs/MBtu. The Industry Study projects 52 ESP-equipped units
would be required to retrofit a fabric filter, removing 25.13 tons of non-Hg metal HAP (6,269
tons of fPM) for an average cost effectiveness of $102,000,000 per ton of non-Hg metal HAP
($311,900 per ton of fPM). In addition, 23 existing units equipped with fabric filters would have
to adopt enhanced O&M, removing an additional 2.61 tons of non-Hg metal HAP (652 tons of
fPM) for an average of cost of $617,195/ton of non-Hg metal HAP ($1,887/ton of fPM). These
actions cumulatively remove a total of 27.74 tons of non-Hg metal HAP (6,921 tons of fPM) for
an average cost effectiveness of $92,470,000/ton non-Hg metal HAP ($282,715/ton of fPM)
removed. These costs are a factor of almost three times that projected by EPA.
Appendix Table A-2 reports the units to which the Industry Study assigned fabric filter retrofits
and enhancements of operating and maintenance procedures, to meet the alternative fPM limit of
0.006 lbs/MBtu.
5.3 Conclusions
•
EPA’s cost study is deficient in terms of the number of ESP-equipped units required to
retrofit improvements, the capital cost assigned for the most significant Major ESP
improvement, and estimates of $/ton cost-effectiveness incurred. EPA, by ignoring the
need for a design and operating margin cited in at least two of their publications (Hutson,
2012 and Parker, 2023) under-predicts the number of units that would require retrofits.
•
This study – using the minimum margin cited by EPA in previous publications – projects
a much higher annual cost for capital equipment to meet the proposed 0.010 lbs/MBtu $169.7 M versus EPA’s maximum estimate of $93.3 M. To meet the alternative PM rate
of 0.006 lbs/MBtu, this study projects 50% more units (87 versus 65) must be retrofit
with fabric filters or implement enhanced O&M to an existing fabric filter, incurring an
annual cost of $1.96 B versus EPA’s estimate of 633 M/yr – a three-fold increase.
•
As a consequence, this study predicts the cost effectiveness to meet 0.010 lbs/MBtu will
average $22,000,000 per ton of non-Hg metal HAP removed ($67,262 per ton of fPM), a
50% premium to EPA’s estimate of $12,200,000 - $14,700,000/ton of non-Hg metal HAP
($37,300 – $44,900/ton of fPM) removed. This study projects the cost to meet the
alternative rate of 0.006 lbs/MBtu will average $92,470,000/ton non-Hg metal HAP
($282,715/ton fPM) removed, almost a factor of three higher than EPA’s estimate of
$103,000/ton.
(Page 207 of Total)
21
391a
USCA Case #24-1119
Document #2058570
Filed: 06/07/2024
Page 82 of 204
Mercury Emissions: Lignite Coals
6.
Mercury Emissions: Lignite Coals
Section 6 addresses EPA’s proposed action to reduce the limit for Hg for lignite-fired units to 1.2
lbs/TBtu. (the following Section 7 addresses EPA’s proposal to retain the present emission limit
of 1.2 lbs/TBtu for units firing bituminous and subbituminous coals (i.e., non-low rank fuels).)
This section critiques EPA’s basis for proposing the lignite Hg emission rate of 1.2 lbs/MBtu,
while supporting the proposal to retain the existing rate for non-low rank coals.
EPA states the following in support of their proposal regarding lignite:
“…..ash from lignite and subbituminous coals tends to be more alkaline (relative to that from
bituminous coal) due to the lower amounts of sulfur and halogen and the presence of a more
alkaline and reactive (non-glassy) form of calcium in the ash. The natural alkalinity of the
subbituminous and lignite fly ash can effectively neutralize the limited free halogen in the flue
gas and prevent oxidation of the Hg0.
Both lignite and subbituminous coal do contain less sulfur than bituminous coal, but other major
differences in composition exist that EPA does not recognize. These are Hg content and its
variability, the sulfur content, and the alkalinity of inorganic matter. EPA’s failure to recognize
these differences manifests itself as (a) assuming activated carbon sorbent effectiveness observed
on subbituminous coal (specifically PRB) extends to lignite, and (b) ignoring variability in Hg
content, as well as the role of sulfur trioxide (SO3), which compromises achieving 90%+ Hg
removal as required to attain 1.2 lbs/TBtu.
Fuel properties are described separately for the North Dakota and Gulf Coast (Texas and
Mississippi) lignite mines.
6.1 North Dakota Mines and Generating Units
Figures 6-1 to 6-4 present data provided by lignite suppliers from North Dakota mines that
describe the variability for Hg and other constituents key to Hg removal. These figures present
data as a “box and whisker” plot, which portrays the mean value, the 25 th and 75th percentile of
the observed data, and the near-minimum (5%) and near-maximum (95%) extremities. Figure 61 shows the variability of Hg and Figure 6-2 the variability of sulfur content. Figure 6-3 shows
variability of fuel alkalinity compared to sulfur content – specifically, the ratio of calcium (Ca)
and sodium (Na) to sulfur – i.e., the (Ca + Na)/S metric.
(Page 208 of Total)
22
392a
This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.