Fish and Wildlife Service and National Park Service 

Department of Interior Comments  

State of Maine Draft Regional Haze Rule State Implementation Plan

July 23, 2010

The Fish and Wildlife Service (FWS) and National Park Service (NPS)
would like to thank the State of Maine for their efforts towards an
effective and interactive consultation with the Federal Land Managers on
their Regional Haze Implementation Plan (RHSIP).  Furthermore, we
appreciate the State’s efforts in addressing our comments on the
preliminary draft document that we received back in 2009.  Through
cooperative efforts, the draft RH SIP is one of the best examples of a
comprehensive and well written draft of the MANE-VU States.  

We would like to further commend the State for addressing most, if not
all, of the “MANE-VU Ask” (“Ask”) identified by the MANE-VU
Regional Planning Organization (RPO).  On numerous occasions, the FWS
and NPS have expressed concerns with the lack of “Ask”
implementation language in many of MANE-VU State RHSIPs.  These
omissions have led to less than ideal expectations on our part that
controls will be realized overall in the MANE-VU Region. The State of
Maine not only appears to have addressed the “Ask”, but has done an
exemplary job of describing the State’s intent to meet Maine’s share
of emission controls.

Overall Comments

As mentioned above, we have concern over the success of the MANE-VU
Region realizing the overall emission reductions expected by the
“Ask”.  Although Maine is actively implementing controls of the
“Ask”, many States are not.  There is a good possibility that
reasonable progress goals that States’ are setting based on full
implementation of these “Ask” will not be achieved without honestly
discerning which emission reductions will take place and which ones will
not.  

It is our recommendation that Maine consider providing discussion and
additional plots of the reasonable progress goals without the “Ask”
that are based on either “on the way (OTW)/on the books (OTB)” or
“better than on the way” scenarios.  An additional presentation,
taking more realistic emission reductions into consideration, would
offer a better representation of the span of control being implemented
in the region.  Although, we recognize that it is not within Maine’s
power to get other States to comply with the “Ask”, none the less,
the State should access whether or not Pennsylvania, Ohio,
Massachusetts, and New York will meet their share of the “Ask”, and
communicate and incorporate these findings.  This is important given
that these States were identified as key contributors to Maine’s Class
I areas.

Specific Comments

Chapter 1, Introduction

Section 1.8:  The State identifies a suite of analysis methods to
produce a weight of evidence approach to basic source apportionment. 
Although we commend the weight of evidence approach, the FWS and NPS
does not consider MANE-VU’s application of the CALPUFF model as within
recommended modeling practices.  As such, the use of non-standard models
or configurations, customarily require a performance evaluation that
demonstrates beneficial use, which was not presented in Maine’s SIP.

Chapter 8, Emission Inventory

We commend Maine for being the first MANE-VU State to implement the low
sulfur fuel strategy.  

It would be helpful to the reader if an explanation could be given as to
why SO2 emissions for EGUs increase between the 2018 OTB/OTW inventory
and the final modeling inventory.

We recommend that Maine add text to clarify which emission reductions
assumed in the final modeling inventory (Table 8-4) are being
implemented.  Added language would explain questions such as: is the
final modeling inventory for non-EGUs point and area source SO2
emissions accurate for Maine’s actual implemented controls, and are
all the Best Available Retrofit (BART) emissions included in these
inventories?

Chapter 10, Best Available Retrofit Technology 

An overarching concern is that it is not clear from the Maine BART
documents (posted 6/29/10) how it applied any of the five factors in the
BART analyses in making its BART determinations. For example, MAINE does
not appear to have given much weight to the visibility benefits that
could be realized from the control strategies evaluated. At least, it is
not clear how Maine applied this factor in developing its BART
conclusions. As we shall discuss later, there appears to be great
inconsistency among the methods used by Maine to assess and evaluate
costs and benefits that would result from the various control strategies
chosen by Maine as representing BART.

The individual company BART determinations were not found in the record.
 Please add an Appendix to the State Implementation Plan for Regional
Haze and include these documents to aid third-party reviewers to deal
with the complete record.  

The core purpose of the BART program is to improve visibility in our
Class I areas, and BART is not necessarily the most cost-effective
solution. Instead, BART represents a broad consideration of technical,
economic, energy, and environmental (including visibility improvement)
factors. We believe that it is essential to consider both the degree of
visibility improvement in a given Class I area as well as the cumulative
benefits of improving visibility across all of the Class I areas
affected. 

There are at least six Class I areas impacted by Maine’s BART sources.
We believe that it is appropriate to consider both the degree of
visibility improvement in a given Class I area as well as the cumulative
effects of improving visibility across all of the Class I areas
affected. The same metric should not be used to evaluate the effects of
reducing emissions from a BART source that impacts only one Class I area
as for a BART source that impacts multiple Class I areas. Also,
evaluating impacts at one Class I area, while ignoring others that are
similarly significantly impaired should not be done. Emissions savings
from a source are benefits that will be spread well beyond only the
most-impacted Class I area, and should be considered. While Maine
presented data describing improvements to visibility at a specific Class
I area that would result from the various control scenarios it
investigated, the State has not explained how it incorporated this
information on impacts upon all Class I areas into its BART decision. 

For example, Wyoming evaluated cumulative visibility improvement for
both its BART and reasonable progress determinations—following are
excerpts from those Wyoming determinations (with emphasis added): 

 for enhanced ESP (Post-Control Scenario A) was 1.070 Δdv from Unit 1,
0.199 Δdv from Unit 2, 1.068 Δdv from Unit 3, and 0.892 Δdv from Unit
4. 

-Control Scenario A for each unit, was 3.558 Δdv from Unit 3 and 1.963
Δdv from Unit 4.

Visibility impacts were addressed in a comprehensive visibility analysis
covering three visibility impairing pollutants and the associated
control options. The cumulative visibility improvement as compared to
the baseline across Wind Cave NP and Badlands NP achieved with new LNB
with OFA at the 30-day limit of 0.23 lb/MMBtu (based on the 98th
percentile modeled results) was 0.14 Δdv from each of the three units.
The expected visibility improvement over the course of a full annual
period would be even greater due to the annual BART limit that is based
on 0.19 lb/MMBtu.

Visibility impacts were addressed in a comprehensive visibility analysis
covering all three visibility impairing pollutants and associated
control options. The cumulative 3-year averaged 98th percentile
visibility improvement from the baseline summed across both Class I
areas achieved with LNB with advanced OFA, wet FGD, and existing ESP
with FGC (Post-Control Scenario A) was 1.716 Δdv from Unit 1 and 1.934
Δdv from Unit 2. 

Visibility impacts were addressed in a comprehensive visibility analysis
covering all three visibility impairing pollutants and associated
control options. The cumulative 3-year averaged 98th percentile
visibility improvement from the baseline summed across both Class I
areas achieved with LNB with advanced OFA, upgrading the existing dry
FGD, and a new full-scale fabric filter, Post-Control Scenario A for
Unit 1, was 0.996 Δdv. 

Oregon considered cumulative benefits for the Boardman Power Plant SCR
addition for reasonable progress:

Table 22: Visibility Modeling Results (percent improvement) Total
visibility impacts (sum of 98th percentile for all Class I areas)

The BART guidelines recommend analyzing visibility improvement for the
highest impacted Class I area with the assumption that any improvement
in the worse impacted area would result in improvement in the lesser
impacted areas. However, since the Boardman Plant significantly impacts
14 Class I Areas within 300 kilometers, the Department tried to include
other parameters that would assess the significance of the improvements
for all Class I areas impacted. Therefore, the Department added the
number of Class I areas with impacts greater than 1.0 delta deciview,
the total delta deciviews for all Class I areas (98th percentile), and
the average delta deciview for all Class I areas (98th percentile). As
can be seen in Table 21, any one of the parameters is fairly
representative of the other parameters perhaps with the exception of
WFGD. Given these results, the Department does not believe that adding
additional parameters, such as total deciview days, would result in any
other conclusions and would probably just add confusion to the analysis
(e.g., more days of impacts than are in a year). Using the results of
the visibility modeling, the cost effectiveness of the control
technologies is recalculated by relating the costs to deciview
improvement (Mt. Hood and all Class I areas) as shown in the following 2
tables.

Maine has ignored the other Class I areas where a given BART source is
also causing or contributing to visibility impairment. The dollar cost
per increment of visibility improvement would be substantially lower if
full consideration is given to all affected Class I areas that would
benefit from emission reductions. While we recognize that EPA has
provided no guidance on this issue of assessing visibility benefits that
would result in multiple Class I areas when emissions are reduced from a
given BART source, we commend Wyoming and Oregon for their initiative in
addressing the issue. We also recognize that there is no “perfect”
method for addressing cumulative benefits, but we firmly believe that
Maine must show how it considered the cumulative impact of the BART
sources the affected Class I areas. We have suggested an approach to
Maine that is consistent with available information and with the
approach used by Wyoming and Oregon, and again request that MEDEP show
how it has considered the cumulative benefits of potential BART
reductions.

Based upon our reviews of BART analyses across the U.S., we believe that
cost-per-deciview ($/dv) of visibility improvement is the most-common
and most-useful parameter for assessing the cost-effectiveness of
strategies to improve visibility in Class I areas. Our compilation of
BART analyses across the U.S. reveals that the average cost/dv proposed
by either a state or a BART source is $13 - $20 million, with a maximum
of almost $50 million/dv proposed by Colorado at the Martin Drake power
plant in Colorado Springs. 

Section 10.7.2:  In the Best Available Retrofit Technology (BART)
determination section for FPL Energy Wyman, LLC, three unlabeled tables
identify visibility benefits based on 1st and 8th high values (page
110).  Our understanding is that the quality and quantity of meteorology
used during the BART determinations fall within the recommended modeling
practices.  Maine voluntarily agreed to limit evaluations to 1st high
values in lieu of generating 3 years of quality meteorological input.
Please communicate that the State did not use the 8th high to base their
BART conclusion. 

Comments on the BART determinations for individual facilities that are
subject to BART follow.  We are focusing our comments on the BART
determinations for the Verso Androscoggin and FPL Wyman facilities
because they have larger impacts than the other BART sources. We are
also providing comments on the other BART sources.  

Verso Androscoggin Paper Mill

The Power Boiler #1 and #2, and the Waste Fuel Incinerator (WFI) units
at Verso Androscoggin Paper Mill are BART eligible.  Both SCR and SNCR
are evaluated for each of these units as BART options for controlling
NOx emissions.  In each case, we have concerns with the cost estimation
methodologies used:  annual reagent and catalyst costs are significantly
above what should be expected, capital recovery factor calculations use
annual interest rates nearly double the standard of EPA’s OAQPS Air
Pollution Cost Control Manual, and recovery periods only half as long,
and there are unexplained differences between the company’s proposal
and the Maine cost estimates.  In summary, our data indicates that both
SCR and SNCR should be considered as viable NOx BART conclusions for
these units, and that lower sulfur residual oils should be more fully
evaluated as an SO2 BART option for the Power Boilers. We also have
several questions regarding the SO2 BART conclusion for the WFI.  Please
see the detailed comments contained in Attachment 1 for specifics.

FPL Energy Wyman Station

Power Boilers #3 and #4 at the FPL Energy Wyman Station are
BART-eligible units.  The State’s SO2 BART analysis appears to be the
only BART analysis conducted by Maine in which cost-effectiveness was
not evaluated in terms of annual cost/ton of pollutant removed. Instead,
Maine appears to have relied solely upon annual cost/deciviews (dv) of
visibility improvement. While we encourage the use of the $/dv metric,
it was not properly calculated nor applied in this case. Using the data
available in the BART analyses, we assessed the cost per ton of SO2
reduced by the BART options, as well as corrected $/dv calculations. 
Based upon the results, we believe that it is reasonable to conclude
that 0.5% - 0.3% sulfur fuels are BART for the FPL boilers.  See the
discussion in Attachment 2 to these comments for further details.

 

SAPPI SD Warren Paper Mill

We are confused as to the BART status for Power Boiler #1 at the SAPPI
SD Warren Paper Mill. While the company-prepared BART analysis
(September 2009) did not mention this unit, the January 21, 2010, Maine
BART analysis does identify and analyze BART controls for Power Boiler
#2.  The final Maine BART analysis for the facility, posted on June 29,
2010, is again silent on this unit.  Please explain the BART eligibility
status for the SAPPI SD Warren Paper Mill Power Boiler #1, and include
any appropriate BART determination in the final SIP.  Supporting
information for this comment is included in Attachment 3 to these
comments.

Domtar Industries, Inc. – Woodland Mill 

The State did a good job of reflecting the five-factor BART protocol in
the Departmental Findings of Fact and Order for Domtar Industries, Inc.
– Woodland Mill.  Section II makes reference from the company BART
determination to a Dry Electrostatic Precipitator BART alternative
estimated to cost $4,640 per ton of particulate matter removed.  It also
makes reference to the Selective Non-Catalytic Reduction (SNCR) BART
alternative estimated by the company to cost $7,360 per ton of NOx
removed.  If the detailed information correctly supports the values
shown above, then it may be reasonable to conclude that the cost per ton
of removal was excessive.  Maine seemed to rely solely upon the MANE-VU
visibility data to evaluate in a general way the visibility impact of a
given unit on nearby Class I areas, but individual modeling of each BART
alternative was not performed.  In the case of the Woodland Mill it
seemed that by concluding that an alternative was not cost-effective on
a cost per ton basis, Maine believes that the specific cost of
visibility improvement was not necessary.  Normally, the visibility cost
step is performed, even if cost per ton is deemed to be excessive. 
Existing SO2 controls on Power Boiler #9 and the Lime Kiln seemed to be
considered ‘top controls’, so further cost analysis was not
necessary.        

Lincoln Paper

As we understand Maine proposes that Lincoln be allowed to burn 2% S
fuel oil.  Additionally, SO2 emissions from the recovery boiler shall be
controlled to 141 ppmv (dry basis) @ 8% O2 on a 24-hour block average
basis when firing only black liquor or when firing a combination of
black liquor and oil. The recovery boiler fires #6-fuel oil.  Oil fired
in the recovery boiler alone shall not exceed 0.7% sulfur by weight or
2.0% sulfur by weight when firing a combination of black liquor and oil.
The recovery boiler is fired with fuel oil for startup purposes (in
order to initiate Black Liquor Solids (BLS) firing) in addition to
shutdowns and other events which require the addition of oil firing.

Maine should explain why use of lower sulfur (0.7% S) fuel (that is
already used when the recovery furnaces fires 100% #6 oil) would incur a
capital cost that made use of that fuel all the time too expensive. The
Lincoln Paper BART determination is deficient because it does not
evaluate the use of 0.7% fuel oil at all times.

Dragon Cement

Based upon the 6/29/10 BART analysis, Maine has determined that for NOx,
Dragon shall operate an SNCR (Selective Non-Catalytic Reduction) system
to reduce NOx emissions from the calciner to achieve a 45% control
efficiency. NOx emissions from the kiln system shall be limited to 350.0
lb/hr on a 90 day rolling average and 1533.0 tons/year on a 12 month
rolling total basis. 

 

We concur.

Katahdin Paper

Maine is limiting emissions from Katahdin paper's only BART source to <
250 tpy to exempt it from BART.  

We concur.

Red Shield Old Town Fuel & Fiber

Maine is limiting emissions from Red Shield's Old Town Fuel & Fiber's
only BART sources to 

< 250 tons/year to exempt it from BART.  

We concur.

Rumford Paper

Maine is limiting emissions from Rumford Paper to less than 250 tons per
year so as to exempt it from BART.  

Please assure that Departmental Findings of Fact and Order or other
federally enforceable documents are promulgated to define the emission
limitations and place them in the official BART record. 

Verso Bucksport paper mill

MEDEP is capping the Verso Bucksport paper mill out of BART, but did not
post the actual permit that does so. 

Please post the pertinent permit.

Chapter 11, Reasonable Progress Goals

By setting reasonable progress goals based on the “Ask”, rather than
the OTW/OTB inventory, the MANE-VU States have made it more difficult to
demonstrate that they have implemented the controls necessary to meet
the reasonable progress goals.  It would be helpful for Maine to discuss
whether or not the OTW/OTB controls were sufficient to meet the uniform
rate of progress at the Maine Class I areas.

Section 11.3, Additional Reasonable Controls:  On page 133, the
statement is made that MANE-VU States have up to 10 years to implement
reasonable controls.  We believe this to be incorrect statement.  It is
our understanding that the regional haze rule requires the controls to
be in progress (e.g., BART determination or rule requirement) when the
RHSIP is submitted as final. 

Table 11-5 summarizes SO2 emissions in 2002 and 2018 modeling inventory
for 12 sources that were assumed to be required to install BART
controls.  A similar table which summarizes actual BART reductions for
Maine sources would be extremely helpful.

MANE-VU indicated that emissions were backfilled in the final inventory
calculations in order to fully meet the Clean Air Act Interstate Rule
(CAIR) cap.  When this backfill method was applied to sources outside of
MANE-VU emission rates for some sources were overestimated, ignoring
State rules and consent decrees.  Please explain in more detail how
Maine consulted with these non-MANE-VU States and how the results from
this consultation were reconciled in making these emission control
decisions.

Chapter 12, Long Term Strategy

Section 12.7.2:  Please identify whether the State implements a smoke
management plan.  If so, identify whether the program is voluntary or
mandatory and whether the impacts to the Class I areas are considered
during the process.

Table 12-1 lists non-CAIR BART facilities that were modeled.  Please
confirm the modeled emissions are consistent with the actual BART
determinations.

It should be stated earlier in the document that Maine will be fully
meeting the “Ask” by 2018.  Providing a statement to this effect at
the beginning of the document will address reader questions earlier in
the review of RHSIP.

"

&

.

¦

¯

Æ

÷

ù

.

/

0

/

0

A

B

0

?

@

A

B

›

¿

Å

È

Ï

h‡`

ᤀ.  This section links reasonable progress for visibility to the
Prevention of Significant Deterioration requirements.  

 For example, MEDEP calculates cost/ton for the Androscoggin paper mill
but does not calculate cost/dv. However, for the Wyman power plant,
MEDEP calculates cost/dv but does not calculate cost/ton.

 DEPARTMENT OF ENVIRONMENTAL QUALITY  AIR QUALITY DIVISION  BART
Application Analysis  AP-6040  May 28, 2009  NAME OF FIRM: PacifiCorp 
NAME OF FACILITY: Jim Bridger Power Plant

 DEPARTMENT OF ENVIRONMENTAL QUALITY AIR QUALITY DIVISION BART
Application Analysis AP-6041 May 28, 2009 NAME OF FIRM: PacifiCorp NAME
OF FACILITY: Dave Johnston Plant

 DEPARTMENT OF ENVIRONMENTAL QUALITY AIR QUALITY DIVISION BART
Application Analysis AP-6047 May 28, 2009 NAME OF FIRM: Basin Electric
Power Cooperative NAME OF FACILITY: Laramie River Station

 DEPARTMENT OF ENVIRONMENTAL QUALITY  AIR QUALITY DIVISION  BART
Application Analysis  AP-6042  May 28, 2009  NAME OF FIRM: PacifiCorp 
NAME OF FACILITY: Naughton Power Plant

 DEPARTMENT OF ENVIRONMENTAL QUALITY AIR QUALITY DIVISION  BART
Application Analysis  AP-6043  May 28, 2009  NAME OF FIRM: PacifiCorp 
NAME OF FACILITY: Wyodak Plant

 DEQ BART Report for the Boardman Power Plant Updated December 19, 2008

 See  HYPERLINK "http://www.wrapair.org/forums/ssjf/bart.html"
http://www.wrapair.org/forums/ssjf/bart.html  

 For example, PacifiCorp has stated in its BART analysis for its Bridger
Unit #2 that “The incremental cost effectiveness for Scenario 1
compared with the baseline for the Bridger WA, for example, is
reasonable at $580,000 per day and $18.5 million per deciview.”

 PAGE   

 PAGE   7 

