Date:		September 9, 2011

Subject:	Confirmation of Operation of Air Pollution Control Devices
(APCDs) for the Top 10 Sources in the Coal-fired Mercury (Hg), Hydrogen
Chloride (HCl), and Particulate Matter (PM) Floors 

From:		Jeffrey Cole (RTI International)

To:  		Bill Maxwell

		OAQPS/SPPD/ESG (D243-01)

		U.S. Environmental Protection Agency

		Research Triangle Park, NC 27711

Introduction

The EPA WAM requested that RTI confirm what air pollution control
devices (APCDs) were actually in operation on the top 10 sources during
performance tests used in the coal-fired MACT floor analysis for the Hg,
HCl, and PM floors. 

Background

Through a review of early public comments, the WAM received indications
that some APCDs that were listed as being installed on a given unit may
not have been operating during the ICR performance stack testing. The
commenter asserted that the EPA should review these control
configurations for accuracy, and then instead of choosing the lowest
reported emission rate for the new source MACT standard, the EPA should
choose the controlled source near the lowest emitting source that was
actually controlled with the control configuration most likely to be
placed on a newly built EGU. The commenter believed this approach makes
the new MACT floor more realistic and achievable. To aide in this
evaluation, RTI was requested to confirm the APCDs operating during the
Hg, HCl, and PM ICR performance testing for the top 10 units in Hg, HCl
and PM floors and make recommendations as to which top performer
reflects the air pollution controls configuration most likely required
to permit a new EGU.

Technical Approach for Coal-Fired Units

RTI contacted the operators of the top 9 of the 131 units in the PM
floor.  The 10th unit’s (San Juan, Unit 3) APCDs that were in
operation during Part III ICR testing were confirmed based on the
configuration listed in Part I of the ICR and the controls used during
Part III testing were noted in the ERT files. For units in the top
10 of the 131 units in the HCl floor and in the top 10 of the 40 units
in the Hg floor, RTI determined the APCDs that were in operation during
ICR testing by comparing the APCD configurations listed in Part I of the
ICR with the controls used during Part III testing as noted in the ERT
files.

There was a change to unit-ranking within PM floor datasets examined
based on information submitted by operators since proposal. In the PM
floor, the previous best unit tested, Bonanza Unit 1, dropped out of the
top 10 units in the floor upon incorporation of a corrected submittal.
This had the effect of shifting each unit in the list up one ranking and
moving San Juan Unit 3 into the top ten for lowest PM emissions. 

Note that there was one unit in the top ten of the Hg floor units and
two units in the top ten HCl floor units that were ranked based solely
on Part II data. The controls that were in operation during the MACT
floor performance tests were consistent with the controls listed in
their Part I submissions.  The Part I controls for those units are
listed in Tables 2 and 3. 

In order to determine the range of control technologies that would be
expected on a new EGU to comply with existing major source permitting
requirements and to evaluate the impacts of these required control
technologies on achievability of the MACT emission standards, a review
of the RACT/BACT/LAER Clearinghouse was conducted for EGU’s permitted
since January 1, 2006. This review indicated the following predominant
technologies:

Fabric filters were the predominant BACT for PM; twenty-seven of the
twenty-nine sources were for permitted fabric filters, and two were
permitted for ESPs.

Fifty percent of the listed sources were permitted for dry scrubbers and
fifty percent were permitted for wet scrubbers as BACT for SO2. 

For sulfuric acid mist (a primary constituent in condensable PM), the
predominant BACT was dry sorbent injection either integral to the spray
dryer absorber or upstream of a wet scrubber.

SCRs were the predominant BACT for NOx control.

Sorbent injection or activated carbon injection was predominant for
control of mercury.

Based on this review of the RACT/BACT/LAER Clearinghouse, the air
pollution control configuration that would be expected on a new EGU
permitted under the NSR program would be expected to have the following
four major components:

Either a fabric filter or a high efficiency ESP;

Either a wet or dry FGD system;

Activated carbon or other sorbent injection;

Selective Catalytic Reduction (SCR).

Summary of Findings

Table 1 shows the confirmed APCDs in operation during MACT floor total
PM performance tests for the revised top 10 coal-fired units. 

Table 2 the confirmed APCDs in operation during MACT floor HCl
performance tests for the top 10 coal-fired units. 

Table 3 shows the confirmed APCDs in operation during MACT floor Hg
performance tests for the top 10 coal-fired units.

Particulate Matter (PM)

In the proposed rule, the EPA chose to use Dunkirk Unit 1 (ORIS Code
2554) for the new source total PM input-based floor (lb/MMBtu) and
Bonanza Unit 1 (ORIS Code 7790) for the new source total PM output-based
floor (lb/MWh). As stated above the Bonanza Unit 1’s original ranking
was based on an erroneous submittal and, once corrected, it dropped out
of the top 10 units in the total PM floor. Based on the comments
received it is clear that Dunkirk Unit 1’s control configuration (see
highlighted row in Table 1) was different during the MACT floor
performance tests than the control configuration reported in Part I of
the ICR because no SNCR or dry sorbent injection system was operating
during the ICR stack tests. Therefore, RTI does not recommend that
either of these units continue to be considered for the new floors. In
addition, we also believe that setting a new MACT based on a unit with
only a fabric filter would not be reflective of either the best
available control technology (BACT) or 112(g) case-by-case MACT that
would be required under existing major source preconstruction permitting
programs:

There are only two units within the top ten with the air pollution
control configurations that are typical of the predominant BACT
technologies:

Springerville Unit 4 deploys a dry FGD system. This source is equipped
with an SCR, spray dryer absorber, and a fabric filter. 

Havana Boiler 9 deploys a dry FGD system. This source is equipped with
an ESP, SCR, ACI, dry scrubber, and a fabric filter.

Because wet scrubbers are effective in removing PM larger than five
microns from flue gas, ESPs followed by wet scrubbers can also approach
the PM emission rates achievable by dry scrubbers with fabric filters as
reflected by the presence of Roxboro Unit 2 in the top 10. These air
pollution control strategies are driven by state mercury control
emissions standards and national fine particulate ambient standards.
Requirements to deploy these technologies can directly impact a
source’s ability to meet the surrogate total PM limit due to the
significant dust loading added to the flue gas by the controls. The
layout of dust from these systems can be problematic during startups and
shutdowns when the particulate control device may be bypassed (e.g.,
fabric filters) or when the PM control device may be below design
temperatures (e.g., ESPs). Additionally, these control technologies can
pose challenges during malfunctions (e.g., punctured bags) as dust loads
remain high while operators determine the location of the problem. For
these reasons, RTI recommends selection of Havana Boiler 9 as the source
within the top 10 for PM that reflects the air pollution controls
typical for new EGUs because this source deploys both ACI and dry
sorbent.

Acid Gases

The EPA chose to utilize Logan Unit 1’s (ORIS Code 10043) emission
averages in the proposed rule for the new HCl floor. This still seems
like a good choice, as it is equipped with an SCR, spray dryer absorber
and a fabric filter, which are consistent with the predominant controls
found in the RACT/BACT/LAER Clearinghouse. 

Mercury

Finally, the EPA chose Nucla Unit 1 (ORIS Code 527) in the proposed rule
for the Hg floor because the two Spruance Genco Units (Gen 2 and 3, ORIS
Code 54081) are an unusual furnace design that is likely atypical of
future new EGU’s. Nucla Unit 1 still appears to be an appropriate
choice, as it is equipped with an SCR, spray dryer absorber and a fabric
filter. These controls are consistent with the predominant controls
found in the RACT/BACT/LAER Clearinghouse. However, in our review of
comments, we have seen concerns that Nucla Unit 1 is not a good choice
for a new floor unit because, in the commenter’s opinion, this unit
has a “unique” fuel that cannot be reproduced and that yields
emission rates that are unachievable by other sources. It is worth
noting that Logan Unit 1 (ORIS Code 10043) achieved virtually the same
Hg emission rate as Nucla Unit 1 during MACT floor performance testing.
Logan Unit 1 is a conventional, wall-fired unit (as compared to
Nucla’s FBC), and it is controlled with an SCR, spray dryer absorber
and a fabric filter. This unit would also be a good choice as a
representative unit for newly constructed EGUs.

Table 1. The Top 10 Units in the Total PM MACT Floor for Coal-fired
EGUs

ORIS code	Plant Name	Unit Number	Unit Type	Boiler Type	Boiler Count
Total Capacity

MWe	Total max heat input (MMBtu/hr)	Unit heat rate (MMBtu/MWh)	Fuel
type(s)	APCD Controls listed at Proposal	Confirmed APCD Controls

2554	Dunkirk Generating Plant	1	Conventional Boiler	Tangential firing	1
85	922.2	10.85	subbituminous	SNCR, DSI, FF	FF

492	Martin Drake	Unit 5 - Coal	Conventional Boiler	Wall firing - front
firing	1	55	570	10.36	subbituminous	FF, WFGD	FF, WFGD

568	Bridgeport Station	BHSEMU3OS3-#2	Conventional Boiler	Tangential
firing	1	403	4100	10.17	subbituminous	ESP, ACI, FF	ESP, ACI, FF

2712	Roxboro Steam Electric Plant	Rox_Cfg_2c	Conventional Boiler
Tangential firing	1	703	6893	9.81	bituminous	SCR, ESP, WFGD	SCR, ESP,
WFGD

8223	Springerville	4	Conventional Boiler	Wall firing - opposed firing	1
450	4200	9.33	bituminous	SCR, DFGD, FF	SCR, DFGD, FF

891	Havana	Boiler 9	Conventional Boiler	Wall firing - unspecified	1	481
5030	10.46	bituminous	ESP, SCR, ACI, DFGD, FF	ESP, SCR, ACI, FF

2554	Dunkirk Generating Plant	4	Conventional Boiler	Tangential firing	1
195	1836	9.42	subbituminous	SNCR, DSI, FF	FF

7097	J K Spruce	1	Conventional Boiler	Tangential firing	1	580	6361	10.97
bituminous	FF, WFGD	FF, WFGD

2324	Reid Gardner	1	Conventional Boiler	Wall firing - front firing	1	111
1215	10.95	bituminous	FF, WFGD	FF, WFGD

2451	San Juan	Unit 3	Conventional Boiler	Wall firing - opposed firing	1
544	5758	10.58	subbituminous	FF, ACI, WFGD	FF, ACI, WFGD



Table 2. The Top 10 Units in the HCl MACT Floor for Coal-fired EGUs

ORIS code	Plant Name	Unit Number	Unit Type	Boiler Type	Boiler Count
Total Capacity

MWe	Total max heat input (MMBtu/hr)	Unit heat rate (MMBtu/MWh)	Fuel
type(s)	APCD Controls listed at Proposal	Confirmed APCD Controls

10043	Logan Generating Plant	Unit1	Conventional Boiler	Wall firing -
opposed firing	1	241.7	2116	8.75	bituminous	SCR, DFGD, FF	SCR, DFGD, FF

54081	Spruance Genco, LLC	GEN3	Conventional Boiler	Stoker - underfeed	2
57.4	750	13.07	bituminous	DFGD, FF	DFGD, FF

54081	Spruance Genco, LLC	GEN2	Conventional Boiler	Stoker - underfeed	2
57.4	750	13.07	bituminous	DFGD, FF	DFGD, FF

3130	Seward	SEW-1	Fluidized bed firing	Fluidized bed firing	2	585	6200
10.60	coal refuse (culm or gob)	SNCR, FF	SNCR, FF

3130	Seward	SEW-2	Fluidized bed firing	Fluidized bed firing	2	585	6200
10.60	coal refuse (culm or gob)	SNCR, FF	SNCR, FF

52071	Sandow Station	5A	Fluidized bed firing	Fluidized bed firing	1
282.35	2960	10.48	lignite	SNCR, ACI, DFGD, FF	SNCR, ACI, DFGD, FF

56224	TS Power Plant	TSPower	Conventional Boiler	Wall firing - opposed
firing	1	242	2112	8.73	subbituminous	SCR, ACI, DFGD, FF	SCR, ACI, DFGD,
FF

3179	Hatfield's Ferry Power Station	001	Conventional Boiler	Wall firing
- opposed firing	1	590	5700	9.66	bituminous	ESP, WFGD	ESP, WFGD

52071	Sandow Station	5B	Fluidized bed firing	Fluidized bed firing	1
282.35	2960	10.48	lignite	SNCR, ACI, DFGD, FF	SNCR, ACI, DFGD, FF

108	HOLCOMB	SGU1	Conventional Boiler	Wall firing - opposed firing	1	387
3389	8.76	subbituminous	DFGD, FF	DFGD, FF

Table 3. The Top 10 Units in the Mercury MACT Floor for Coal-fired EGUs

ORIS code	Plant Name	Unit Number	Unit Type	Boiler Type	Boiler Count
Total Capacity

MWe	Total max heat input (MMBtu/hr)	Unit heat rate (MMBtu/MWh)	Fuel
type(s)	APCD Controls listed at Proposal	Confirmed APCD Controls

54081	Spruance Genco, LLC	GEN2	Conventional Boiler	Stoker - underfeed	2
57	750	13.07	bituminous	DFGD, FF	DFGD, FF

54081	Spruance Genco, LLC	GEN3	Conventional Boiler	Stoker - underfeed	2
57	750	13.07	bituminous	DFGD, FF	DFGD, FF

527	Nucla	001	Fluidized bed firing	Fluidized bed firing	1	110	1032	9.38
bituminous	FBC, SNCR, FF	FBC, SNCR, FF

10043	Logan Generating Plant	Unit1	Conventional Boiler	Wall firing -
opposed firing	1	242	2116	8.75	bituminous	SCR, DFGD, FF	SCR, DFGD, FF

3130	Seward	SEW-1	Fluidized bed firing	Fluidized bed firing	2	585	6200
10.60	coal refuse (culm or gob)	SNCR, FF	SNCR, FF

2527	AES Greenidge	Unit 4	Conventional Boiler	Tangential firing	1	112
1117	9.97	bituminous, petroleum coke	SCR, ACI, DFGD, FF	SCR, ACI, DFGD,
FF

54035	Roanoke Valley I	Boiler 1	Conventional Boiler	Wall firing - front
firing	1	182	1700	9.34	bituminous	DFGD, FF	DFGD, FF

2526	AES Westover, LLC	8	Conventional Boiler	Tangential firing	1	86	875
10.17	bituminous	SCR, DFGD, FF	SCR, DFGD, FF

50976	Indiantown Cogeneration, L.P.	001	Conventional Boiler	Wall firing
- opposed firing	1	361	3422	9.48	bituminous	SCR, DFGD, FF	SCR, DFGD, FF

50888	Northampton Generating Company, L.P.	GEN1	Fluidized bed firing
Fluidized bed firing	1	121	1146	9.47	coal refuse (culm or gob)	FBC,
SNCR, FF	FBC, SNCR, FF



  PAGE   \* MERGEFORMAT  6 

  PAGE   \* MERGEFORMAT  1 

