
[Federal Register Volume 77, Number 181 (Tuesday, September 18, 2012)]
[Rules and Regulations]
[Pages 57863-57919]
From the Federal Register Online via the Government Printing Office [www.gpo.gov]
[FR Doc No: 2012-20918]



[[Page 57863]]

Vol. 77

Tuesday,

No. 181

September 18, 2012

Part III





Environmental Protection Agency





-----------------------------------------------------------------------





40 CFR Part 52





Approval and Promulgation of Implementation Plans; State of Montana; 
State Implementation Plan and Regional Haze Federal Implementation 
Plan; Final Rules

  Federal Register / Vol. 77 , No. 181 / Tuesday, September 18, 2012 / 
Rules and Regulations  

[[Page 57864]]


-----------------------------------------------------------------------

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Part 52

[EPA-R08-OAR-2011-0851, FRL 9719-9]


Approval and Promulgation of Implementation Plans; State of 
Montana; State Implementation Plan and Regional Haze Federal 
Implementation Plan

AGENCY: Environmental Protection Agency (EPA).

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: The Environmental Protection Agency (EPA) is promulgating a 
Federal Implementation Plan (FIP) to address regional haze in the State 
of Montana. EPA developed this FIP in response to the State's decision 
in 2006 to not submit a regional haze State Implementation Plan (SIP) 
revision. The FIP satisfies requirements of the Clean Air Act (CAA or 
``the Act'') that require states, or EPA in promulgating a FIP, to 
assure reasonable progress towards the national goal of preventing any 
future and remedying any existing man-made impairment of visibility in 
mandatory Class I areas. In addition, EPA is approving one of the 
revisions to the Montana SIP submitted by the State of Montana through 
the Montana Department of Environmental Quality on February 17, 2012, 
specifically, the revision to the Montana Visibility Plan that includes 
amendments to the ``Smoke Management'' section, which adds a reference 
to Best Available Control Technology (BACT) as the visibility control 
measure for open burning as currently administered through the State's 
air quality permit program. This change was made to meet the 
requirements of the Regional Haze Rule. EPA will act on the remaining 
February 17, 2012 revisions in the State's submittal in a future 
action.

DATES: This final rule is effective October 18, 2012.

ADDRESSES: EPA has established a docket for this action under Docket ID 
No. EPA-R08-OAR-2011-0851. All documents in the docket are listed on 
the www.regulations.gov Web site. Although listed in the index, some 
information is not publicly available, e.g., Confidential Business 
Information (CBI) or other information whose disclosure is restricted 
by statute. Certain other material, such as copyrighted material, is 
not placed on the Internet and will be publicly available only in hard 
copy form. Publicly available docket materials are available either 
electronically through www.regulations.gov, or in hard copy at the Air 
Program, Environmental Protection Agency (EPA), Region 8, 1595 Wynkoop 
Street, Denver, Colorado 80202-1129. EPA requests that if at all 
possible, you contact the individual listed in the FOR FURTHER 
INFORMATION CONTACT section to view the hard copy of the docket. You 
may view the hard copy of the docket Monday through Friday, 8 a.m. to 4 
p.m., excluding Federal holidays.

FOR FURTHER INFORMATION CONTACT: Scott Jackson, Air Program, Mailcode 
8P-AR, Environmental Protection Agency, Region 8, 1595 Wynkoop Street, 
Denver, Colorado 80202-1129, (303) 312-6107, or Jackson.Scott@epa.gov.

SUPPLEMENTARY INFORMATION: 

Definitions

    For the purpose of this document, we are giving meaning to certain 
words or initials as follows:
     The words or initials Act or CAA mean or refer to the 
Clean Air Act, unless the context indicates otherwise.
     The initials A/F mean or refer to air-to-fuel.
     The initials ALM mean or refer to Ammonia Limiting Method
     The initials ARM mean or refer to Administrative Rule of 
Montana.
     The initials ARP mean or refer to the acid rain program.
     The initials ARS mean or refer to Air Resources 
Specialists.
     The initials ASOFA mean or refer to advanced separated 
overfire air.
     The initials BACT mean or refer to Best Available Control 
Technology.
     The initials BART mean or refer to Best Available Retrofit 
Technology.
     The initials CAA mean or refer to the Clean Air Act.
     The initials CAM mean or refer to compliance assurance 
monitoring.
     The initials CAMD mean or refer to EPA Clean Air Markets 
Division.
     The initials CAMx mean or refer to Comprehensive Air 
Quality Model.
     The initials CBI mean or refer to confidential business 
information.
     The initials CCM mean or refer to EPA Control Cost Manual.
     The initials CCOFA mean or refer to close-coupled overfire 
air system.
     The initials CDS mean or refer to circulating dry 
scrubber.
     The initials CGA mean or refer to gas cylinder audit.
     The initials CELP mean or refer to Colstrip Energy Limited 
Partnership.
     The initials CEMS mean or refer to continuous emissions 
monitoring systems.
     The initials CEPCI mean or refer to Chemical Engineering 
Plant Cost Index.
     The initials CFAC mean or refer to Columbia Falls Aluminum 
Company.
     The initials CFB mean or refer to circulating fluidized 
bed.
     The initials CKD mean or refer to cement kiln dust.
     The initials CMAQ mean or refer to Community Multi-Scale 
Air Quality modeling system.
     The initials CPMS mean or refer to continuous parametric 
monitoring system.
     The initials CO mean or refer to carbon monoxide.
     The initials CPI mean or refer to Consumer Price Index.
     The initials CRF mean or refer to Capital Recovery Factor.
     The initials CSAPR mean or refer to Cross-State Air 
Pollution Rule.
     The initials DAA mean or refer to Dry Absorbent Addition.
     The initials DPCS mean or refer to digital process control 
system.
     The initials D-R mean or refer to Dresser-Rand.
     The initials DSI mean or refer to dry sorbent injection.
     The initials EC mean or refer to elemental carbon.
     The initials EGU mean or refer to Electric Generating 
Units.
     The words EPA, we, us or our mean or refer to the United 
States Environmental Protection Agency.
     The initials ESP mean or refer to electrostatic 
precipitator.
     The initials FCCU mean or refer to fluid catalytic 
cracking unit.
     The initials FGD mean or refer to flue gas 
desulfurization.
     The initials FGR mean or refer to flue gas recirculation.
     The initials FIP mean or refer to Federal Implementation 
Plan.
     The initials FLMs mean or refer to Federal Land Managers.
     The initials HAR mean or refer to hydrated ash 
reinjection.
     The initials HDSCR mean or refer to high-dust selective 
catalytic reduction.
     The initials HC mean or refer to hydrocarbons.
     The initials gr/scf mean or refer to grains per standard 
cubic foot.
     The initials IMPROVE mean or refer to Interagency 
Monitoring of Protected Visual Environments monitoring network.
     The initials IPM mean or refer to Integrated Planning 
Model.
     The initials IWAQM refer to Interagency Workgroup on Air 
Quality Modeling.
     The initials LDSCR mean or refer to low-dust selective 
catalytic reduction.
     The initials LEA mean or refer to low excess air.
     The initials LNBs mean or refer to low NOX 
burners.

[[Page 57865]]

     The initials LSD mean or refer to lime spray drying.
     The initials LSFO mean or refer to limestone forced 
oxidation.
     The initials LTS mean or refer to Long-Term Strategy.
     The initials MACT mean or refer to maximum achievable 
control technology.
     The initials MATB mean or refer to Montanan's Against 
Toxic Burning.
     The initials MDEQ mean or refer to Montana's Department of 
Environmental Quality.
     The initials MDF mean or refer to medium density 
fiberboard.
     The initials MISO mean or refer to Midwest Independent 
Transmission System Operator.
     The initials MDU mean or refer to Montana-Dakota Utilities 
Company.
     The initials MEL mean magnesium-enhanced lime.
     The initials MKF mean or refer to mid-kiln firing of solid 
fuel.
     The words Montana and State mean the State of Montana.
     The initials MSCC mean or refer to Montana Sulphur and 
Chemical Company.
     The initials NAAQS mean or refer to National Ambient Air 
Quality Standards.
     The initials NC mean or refer to North Carolina.
     The initials ND mean or refer to North Dakota.
     The initials NEI mean or refer to National Emission 
Inventory.
     The initials NESHAP mean or refer to National Emission 
Standards for Hazardous Air Pollutants.
     The initials NH3 mean or refer to ammonia.
     The initials NOX mean or refer to nitrogen 
oxides.
     The initials NP mean or refer to National Park.
     The initials NPS mean or refer to National Parks Service.
     The initials NSCR mean or refer to non-selective catalytic 
reduction.
     The initials NSPS mean or refer to New Source Performance 
Standards.
     The initials NWR mean or refer to National Wildlife 
Reserve.
     The initials OMB mean or refer to the Office of Management 
and Budget.
     The initials OC mean or refer to organic carbon.
     The initials OFA mean or refer to overfire air.
     The initials PC mean or refer to pulverized coal.
     The initials PH/PC mean or refer to preheater/precalciner.
     The initials PM mean or refer to particulate matter.
     The initials PM2.5 mean or refer to particulate matter 
with an aerodynamic diameter of less than 2.5 micrometers (fine 
particulate matter).
     The initials PM10 mean or refer to particulate matter with 
an aerodynamic diameter of less than 10 micrometers (coarse particulate 
matter).
     The initials PMCD mean or refer to particulate matter 
control device.
     The initials ppb mean or refer to parts per billion.
     The initials ppm mean or refer to parts per million.
     The initials PRB mean or refer to Powder River Basin.
     The initials PSAT mean or refer to Particulate Matter 
Source Apportionment Technology.
     The initials PSD mean or refer to Prevention of 
Significant Deterioration.
     The fraction Q/D means quantity of emissions over 
distance.
     The initials RAA mean or refer to relative accuracy audit.
     The initials RATA mean or refer to relative accuracy test 
audit.
     The initials RAVI mean or refer to Reasonably Attributable 
Visibility Impairment.
     The initials RICE mean or refer to Reciprocating Internal 
Combustion Engines.
     The initials RMC mean or refer to Regional Modeling 
Center.
     The initials ROFA mean or refer to rotating opposed fire 
air.
     The initials RP mean or refer to Reasonable Progress.
     The initials RPG or RPGs mean or refer to Reasonable 
Progress Goal(s).
     The initials RPOs mean or refer to regional planning 
organizations.
     The initials RRI mean or refer to rich reagent injection.
     The initials RSCR mean or refer to regenerative selective 
catalytic reduction.
     The initials SCOT mean or refer to Shell Claus Off-Gas 
Treatment.
     The initials SCR mean or refer to selective catalytic 
reduction.
     The initials SDA mean or refer to spray dryer absorbers.
     The initials SIP mean or refer to State Implementation 
Plan.
     The initials SMOKE mean or refer to Sparse Matrix Operator 
Kernel Emissions.
     The initials SNCR mean or refer to selective non-catalytic 
reduction.
     The initials SO2 mean or refer to sulfur dioxide.
     The initials SOFA mean or refer to separated overfire air.
     The initials SRU mean or refer to sulfur recovery unit.
     The initials TAC mean or refer to Texas Administrative 
Code.
     The initials TESCR mean or refer to tail-end selective 
catalytic reduction.
     The initials TCEQ mean or refer to Texas Commission on 
Environmental Quality.
     The initials tpy mean tons per year.
     The initials TSD mean or refer to Technical Support 
Document.
     The initials URP mean or refer to Uniform Rate of 
Progress.
     The initials USFWS mean or refer to U.S. Fish and Wildlife 
Service.
     The initials VOC mean or refer to volatile organic 
compounds.
     The initials WA mean or refer to Wilderness Area.
     The initials WEG mean or refer to WildEarth Guardians.
     The initials WEP mean or refer to Weighted Emissions 
Potential.
     The initials WETA mean or refer to Western Environmental 
Trade Association.
     The initials WRAP mean or refer to the Western Regional 
Air Partnership.
     The initials YELP mean or refer to Yellowstone Energy 
Limited Partnership.

Table of Contents

I. Background
II. Basis for Our Final Action
III. Final Action
IV. Issues Raised by Commenters and EPA's Responses
    A. Comments on Modeling
    B. General Comments on BART
    C. Comments on Cement Kilns
    D. Comments on Ash Grove
    E. Comments on Holcim
    F. Comments on CFAC
    G. Comments on Colstrip Units 1 and 2
    H. Comments on Corette
    I. Comments on Reasonable Progress and Long Term Strategy
    J. Comments on Colstrip 3 and 4
    K. Comments on Devon Energy
    L. Comments on Montana Dakota Utilities
    M. Comments on Montana Sulphur and Chemical Company
    N. Comments on Health, Ecosystem Benefits, Other Pollutants, and 
Coal Ash
    O. General Comments Supporting Our Proposal or for Stricter 
Controls
    P. General Comments That The Proposal Is Too Stringent
    Q. Comments on Visibility Improvement and Other Causes of Haze
    R. Comments on Cost, Economic Impact, Jobs and Price to 
Consumers
    S. Comments About Other Forms of Energy
    T. Other Miscellaneous Comments
V. Changes From Proposed Rule and Reasons for the Changes
    A. Emission Limits for Corette
    B. Changes to 40 CFR 52.1396(c)(2)--Emission Limitations for 
Cement Kilns:
    C. Change to 40 CFR 52.1396(d)--Compliance date:
    D. Change to 40 CFR 52.1396(e)(3)--CEMS for cement kilns:
    E. Change to 40 CFR 52.1396(e)(4)(ii)--Compliance determination 
methods for SO2 and NOX at cement kilns:
    F. Change to 40 CFR 52.1396(f)(1) and (f)(2)--Compliance 
determinations for

[[Page 57866]]

PM BART limits at EGUs and cement kilns:
    G. Change to 40 CFR 52.1396(f)(2)--Compliance determinations for 
cement kiln PM BART limits:
    H. Change to 40 CFR 52.1396(h)(6)--Recordkeeping requirements 
for cement kilns:
    I. Change to 40 CFR 52.1396(i)--Reporting:
    J. Change to 40 CFR 52.1396(i)(1) and (i)(2)--Reporting for CEMS 
for SO2 and NOX:
    K. Changes to 40 CFR 52.1396 for Devon Energy, Blaine County 
1 Compressor Station
VI. Statutory and Executive Order Reviews

I. Background

    We signed our notice of proposed rulemaking on March 20, 2012, and 
it was published in the Federal Register on April 20, 2012. In that 
notice, we proposed a FIP to address regional haze in the State of 
Montana for the first implementation period (through 2018) including 
determinations of Best Available Retrofit Technology (BART) for 
specific sources subject to that requirement. 77 FR 23988. Montana did 
not submit a SIP, knowing that as a consequence EPA would be required 
to propose and finalize a FIP. A detailed explanation of the CAA's 
visibility requirements and the Regional Haze Rule as it applies to 
Montana was provided in the notice of proposed rulemaking and will not 
be restated here. In that notice, we also proposed to approve a 
revision to the Montana SIP submitted by the State of Montana through 
the Montana Department of Environmental Quality on February 17, 2012. 
The State's submittal contained revisions to the Montana Visibility 
Plan that included amendments to the ``Smoke Management'' section, 
which adds a reference to Best Available Control Technology (BACT) as 
the visibility control measure for open burning as currently 
administered through the State's air quality permit program. EPA's 
rationale for proposing approval of the revisions to the Montana 
Visibility Plan that included amendments to the ``Smoke Management'' 
section was described in detail in the proposal and will not be 
restated here. We note that in the future, Montana retains the option 
of submitting a SIP meeting the requirements of the Regional Haze Rule, 
to replace the FIP.

II. Basis for Our Final Action

    We have fully considered all significant comments on our proposal, 
and, except as noted in section V, below, have concluded that no other 
changes from our proposal are warranted. Our action is based on an 
evaluation of Montana's Visibility SIP submittal and our FIP against 
the regional haze requirements at 40 CFR 51.300--51.309 and CAA 
sections 169A and 169B. All general SIP requirements contained in CAA 
section 110, other provisions of the CAA, and our regulations 
applicable to this action were also evaluated. The purpose of this 
action is to ensure compliance with these requirements. Our authority 
for action on Montana's Visibility SIP submittal is based on CAA 
section 110(k). Our authority to promulgate our FIP is based on CAA 
section 110(c).

III. Final Action

    With this final action we are approving Montana's submittal 
containing revisions to the ``Smoke Management'' section of Montana's 
Visibility Plan that was submitted by the State through the Montana DEQ 
on February 17, 2012. The SIP includes amendments to the ``Smoke 
Management'' section, which adds a reference to BACT as the visibility 
control measure for open burning as currently administered through the 
State's air quality permit program as meeting the requirement of 40 CFR 
308(d)(3)(v) to consider smoke management techniques for agricultural 
and forestry management purposes including plans as they currently 
exist within the state for these purposes. We are promulgating a FIP 
for the remaining parts of the regional haze requirements. Table 1 
shows the control technologies, associated cost, and emission 
reductions for each source that is subject to the FIP.

                Table 1--Control Technologies, Cost, Emissions Reductions and Cost-Effectiveness
----------------------------------------------------------------------------------------------------------------
                                                                                  Annual NOX/SO2
                                                  Total capital       Total         emissions          Cost
            Source              Technology \1\      cost ($)       annualized       reductions     effectiveness
                                                                    cost ($)          (tpy)           ($/ton)
----------------------------------------------------------------------------------------------------------------
Ash Grove Cement.............  LNB + SNCR......       1,191,632       2,238,893  1,088 NOX......           2,058
Holcim, Inc..................  SNCR............       1,312,800         650,399  556 NOX........           1,170
Colstrip Unit 1..............  SOFA + SNCR.....      13,380,673       3,278,964  2,097 NOX......           1,564
Colstrip Unit 2..............  Lime Injection +      28,000,000       4,093,200  4,486 SO2......             912
                                Additional
                                Scrubber Vessel.
Colstrip Unit 2..............  SOFA + SNCR.....      13,380,673       3,256,127  2,072 NOX......           1,571
Colstrip Unit 2..............  Lime Injection +      28,000,000       4,093,200  4,129 SO2......             991
                                Additional
                                Scrubber Vessel.
Devon Energy, Blaine County    NSCR............              --         105,000  335 NOX........             282
 1 Compressor
 Station, Engine 1.
Devon Energy, Blaine County    NSCR............              --         105,000  335 NOX........             282
 1 Compressor
 Station, Engine 2.
                              ----------------------------------------------------------------------------------
    Cumulative Total Annual    ................  ..............      13,727,583
     Cost.
----------------------------------------------------------------------------------------------------------------
-- Total Capital Cost was not calculated.
\1\ The technology listed is the technology evaluated as BART, but sources can choose to use another technology
  or combination of technologies to meet established emission limits. Also where additional control technologies
  are not required, existing controls may still be necessary to meet established emission limits.

IV. Issues Raised by Commenters and EPA's Responses

    This action addresses comments on the Montana Regional Haze FIP. 
The publication of EPA's proposed rule on April 20, 2012 resulted in a 
60-day public comment period that ended on June 19, 2012. We held four 
public hearings for this proposal. Two hearings were held in Helena, 
Montana on Tuesday, May 1, 2012 and two hearings were held in Billings, 
Montana on

[[Page 57867]]

Wednesday, May 2, 2012. During the public comment period we received 
numerous written comments from individual citizens, members of various 
organizations, and also from Ash Grove Cement (Ash Grove), Columbia 
Falls Aluminum Corporation (CFAC), EarthJustice, the U.S. Fish and 
Wildlife Service (USFWS), Holcim Inc. (Holcim), Montana Dakota 
Utilities (MDU), Montana Sulphur and Chemical Company, the National 
Parks Service (NPS), the owners of Colstrip Units 1-4, the State of 
Montana, and WildEarth Guardians (WEG). We have reviewed the comments 
and provided our responses below. Transcripts from the public hearings 
and full copies of the comment letters are available in the docket for 
review.

A. Comments on Modeling

    Comment: PPL and others stated that the proposed BART at Colstrip 1 
and 2 for both NOX and SO2 would result in no 
reasonably anticipated visibility benefit, even assuming that EPA's 
emissions reduction estimates and modeling are correct. In one specific 
comment, the commenter stated:

    A projected 0.066 dv is not a visibility improvement that `may 
reasonably be anticipated to result from the use' of additional 
scrubber vessels at Colstrip Units 1 and 2. 42 U.S.C. 7491(g)(2). 
Such an insignificant projected visibility change is beyond the 
modeling capability of the CALPUFF model version EPA used and is far 
below the threshold for human perceptibility.

    Response: We disagree that any controls required by our action must 
demonstrate a perceptible visibility improvement. In a situation where 
the installation of BART may not result in a perceptible improvement in 
visibility, the visibility benefit may still be significant. The 
Regional Haze Rule states:

even though the visibility improvement from an individual source may 
not be perceptible, it should still be considered in setting BART 
because the contribution to haze may be significant relative to 
other source contributions in the Class I area. Failing to consider 
less-than-perceptible contributions to visibility impairment would 
ignore the CAA's intent to have BART requirements apply to sources 
that contribute to, as well as cause, such impairment.

70 FR 39129.
    Visibility impacts below the thresholds of perceptibility cannot be 
ignored because regional haze is produced by a multitude of sources and 
activities which are located across a broad geographic area. As stated 
in our proposal, with respect to Colstrip 1 and 2, we weighed the 
relatively low costs for lime injection with the additional scrubber 
vessel against the anticipated visibility impacts and determined that 
the cost was justified by the visibility improvement. Similarly, we 
weighed the relatively low cost of separated overfire air (SOFA) + 
selective noncatalytic reduction (SNCR) against the anticipated 
visibility benefit and determined that the cost was justified by the 
visibility benefit.
    We respond to the modeling capabilities of CALPUFF in a response to 
a later comment.
    Comment: A commenter asserted that EPA's modeling assumes constant 
levels of ammonia and failed to consider monitoring data showing that 
ammonia levels are lower during the winter months.
    Response: EPA recognizes that there can be seasonal variability in 
ambient ammonia concentrations and that it is preferable to use ambient 
ammonia measurements when such data are available rather than using 
default background ammonia concentrations. Ammonia monitoring data is 
not available in Montana, however, ammonia monitoring data is available 
in western North Dakota at the Beulah monitoring site. Theodore 
Roosevelt NP, located in western North Dakota, is impacted by Montana 
BART sources and EPA determined that it would be more appropriate to 
use the North Dakota ammonia monitoring data instead of using CALPUFF 
default ammonia concentrations. Therefore EPA used monthly average 
measured ammonia concentrations shown in Table 2 that were measured by 
North Dakota at their Beulah monitoring site.\1\ The monthly average 
ammonia concentrations values were derived from data collected during 
years 2001-2002 and the ambient data were filtered to eliminate data 
from wind directions associated with sources causing a local bias. 
North Dakota concluded in its regional haze modeling analysis that 
these monthly average ammonia values are generally representative of 
background ammonia concentrations in western North Dakota. As a result, 
we did not assume a constant level of ammonia as asserted by the 
commenter, and we did represent seasonal variability in ammonia 
concentrations.
---------------------------------------------------------------------------

    \1\ Protocol for BART-Related Visibility Impairment Modeling 
Analyses in North Dakota (Final), North Dakota Department of Health, 
Division of Air Quality, 1200 Missouri Avenue Bismarck, ND (Nov 
2005), p 32-33.
---------------------------------------------------------------------------

    Additionally, EPA used the POSTUTIL \2\ program with the Ammonia 
Limiting Method (ALM) to post-process the CALPUFF output to correct the 
assumption of constant ammonia availability in the model. The CALPUFF 
model represents multiple plumes that can overlap. The default model 
approach assumes that background ammonia is fully available to form 
nitrate in each plume. The ALM method corrects this assumption by 
partitioning the ammonia between overlapping plumes. Therefore, EPA has 
fully accounted for non-constant ammonia levels by using monthly 
measured background ammonia and by using the ALM in the analysis of 
CALPUFF model results.
---------------------------------------------------------------------------

    \2\ POSTUTIL is a part of the suite of programs associated with 
the CALPUFF modeling system and is used to repartition ammonia in 
overlapping puffs. The model is available at: http://www.src.com/calpuff/calpuff1.htm.

           Table 2--Monthly Ammonia Background Concentrations
------------------------------------------------------------------------
                                                                 Value
                            Month                                (ppb)
------------------------------------------------------------------------
Jan..........................................................       1.22
Feb..........................................................       1.23
Mar..........................................................       1.60
Apr..........................................................       1.94
May..........................................................       2.29
Jun..........................................................       1.63
------------------------------------------------------------------------
Jul..........................................................       1.65
Aug..........................................................       1.69
Sep..........................................................       0.98
Oct..........................................................       1.04
Nov..........................................................       1.37
Dec..........................................................       1.06
------------------------------------------------------------------------

    Comment: A commenter stated that EPA failed to acknowledge 
uncertainty in the CALPUFF model at short distances, and the commenter 
further argues that model uncertainty increases at distances greater 
than 200 km and has a tendency to over predict impacts at greater 
distances.
    Response: The Interagency Workgroup on Air Quality Modeling (IWAQM) 
Phase 2 report (EPA, 1998) \3\ reviewed model performance evaluations 
of CALPUFF as a function of distance from the source and concluded 
that:
---------------------------------------------------------------------------

    \3\ Interagency Workgroup on Air Quality Modeling (IWAQM) Phase 
2 Report and Recommendations for Long-Range Transport Impacts. EPA-
454/R-98-019. U.S. Environmental Protection Agency. Research 
Triangle Park, NC (``IWAQM Phase II Report'') (1998), p 18.

    Based on the tracer comparison results presented in Section 4.6, 
it appears that CALPUFF provides reasonable correspondence with 
observations for transport distances of over 100 km. Most of these 
comparisons involved concentration values averaged over 5 to 12 
hours. The CAPTEX comparisons, which involved comparisons at 
receptors that were 300 km to 1000 km from the release, suggest that 
CALPUFF can overestimate surface concentrations by a factor of 3 to 
4. Use of

[[Page 57868]]

the puff splitting option in CALPUFF might have improved these 
comparisons, but there are serious conceptual concerns with the use 
of puff dispersion for very long-range transport (300 km and 
beyond). As the puffs enlarge due to dispersion, it becomes 
problematic to characterize the transport by a single wind vector, 
as significant wind direction shear may well exist over the puff 
dimensions. With the above thoughts in mind, IWAQM recommends use of 
CALPUFF for transport distances of order 200 km and less. Use of 
CALPUFF for characterizing transport beyond 200 to 300 km should be 
---------------------------------------------------------------------------
done cautiously with an awareness of the likely problems involved.

    Therefore, we modeled Class I areas within 300 km of each BART 
sources but did not model impacts at distances exceeding 300 km.
    EPA has acknowledged that there is uncertainty in the CALPUFF model 
predicted visibility impacts. However, the CALPUFF model can both 
underpredict and overpredict visibility impacts. For example, in a 
presentation for the 2010 annual Community Modeling and Analysis System 
conference, Anderson et al. (2010) \4\ found that the CALPUFF model 
frequently predicted lower nitrate concentrations compared to the CAMx 
photochemical grid model which has a much more rigorous treatment of 
photochemical reactions. EPA recognized the uncertainty in the CALPUFF 
modeling results when EPA made the decision, in the final BART 
Guidelines, to recommend that the model be used to estimate the 98th 
percentile visibility impairment rather than the highest daily impact 
value. While recognizing the limitations of the CALPUFF model in the 
BART Guidelines Preamble, EPA concluded that, for the specific purposes 
of the Regional Haze Rule's BART provisions, CALPUFF is sufficiently 
reliable to inform the decision making process. The Preamble states:
---------------------------------------------------------------------------

    \4\ Anderson, B., K. Baker, R. Morris, C. Emery, A. Hawkins, E. 
Snyder ``Proof-of-Concept Evaluation of Use of Photochemical Grid 
Model Source Apportionment Techniques for Prevention of Significant 
Deterioration of Air Quality Analysis Requirements'' Presentation 
for Community Modeling and Analysis System (CMAS) 2010 Annual 
Conference, (October 11-15, 2010) can be found at http://www.cmascenter.org/conference/2010/agenda.cfm.

    Because of the scale of the predicted impacts from these 
sources, CALPUFF is an appropriate or a reasonable application to 
determine whether such a facility can reasonably be anticipated to 
cause or contribute to any impairment of visibility. In other words, 
to find that a source with a predicted maximum impact greater than 2 
or 3 deciviews meets the contribution threshold adopted by the 
States does not require the degree of certainty in the results of 
the model that might be required for other regulatory purposes. In 
the unlikely case that a State were to find that a 750 MW power 
plant's predicted contribution to visibility impairment is within a 
very narrow range between exemption from or being subject to BART, 
the State can work with EPA and the FLM to evaluate the CALPUFF 
results in combination with information derived from other 
appropriate techniques for estimating visibility impacts to inform 
the BART applicability determination. Similarly for other types of 
BART eligible sources, States can work with the EPA and FLM to 
determine appropriate methods for assessing a single source's 
---------------------------------------------------------------------------
impacts on visibility.

77 FR 39123.
    Therefore, given that the IWAQM guidance provides for the use of 
the CALPUFF model at receptor distances of up to 200 to 300 km, and 
given that EPA has already addressed uncertainty in the CALPUFF model, 
we believe it is reasonable to use CALPUFF to evaluate visibility 
impacts up to 300 km.
    Comment: A commenter stated that the CALPUFF model cannot 
accurately predict visibility changes at the levels EPA predicted for 
Holcim using indirect firing alone (0.125 deciview) or even for the 
additional improvement from the combination of SNCR + indirect firing 
as compared to SNCR alone. The commenter believes that the EPA 
predicted visibility improvement of 0.424 deciview for the combination 
of SNCR + indirect firing is within the uncertainty range of the 
CALPUFF model and cannot reliably predict visibility improvements.
    Response: We disagree. EPA has previously addressed the issue of 
uncertainty in the CALPUFF model. EPA recognized the uncertainty in the 
CALPUFF modeling results when EPA made the decision in the final BART 
Guideline to recommend that the model be used to estimate the 98th 
percentile visibility impairment rather than the highest daily impact 
value. While recognizing the limitations of the CALPUFF model in the 
Preamble, EPA concluded that, for the specific purposes of the Regional 
Haze Rule's BART provisions, CALPUFF is sufficiently reliable to inform 
the decision making process. 70 FR 39123. We continue to maintain that 
it is appropriate to use CALPUFF for BART modeling for Holcim and other 
Montana BART sources.
    Comment: Some commenters stated that we should have modeled impacts 
to additional Class I areas. Some commenters stated that EPA should 
have modeled visibility impacts on Class I areas at a distance of up to 
500 km from the BART source and some commenters specified certain Class 
I areas that they thought should be included in the modeling for a 
particular source.
    Some commenters stated that the Western Regional Air Partnership 
(WRAP) subject to BART modeling indicated impacts from BART sources to 
additional Class I areas that we did not assess. One commenter stated 
that when assessing the impacts from the Big Stone I facility in the 
South Dakota SIP, EPA evaluated visibility as far away as Badlands 
National Park (NP), 470 km, Theodore Roosevelt NP, 555 km, and Boundary 
Waters Wilderness Area (WA) and Voyageurs NP, 431 and 438 km, 
respectively, and the commenter stated that, EPA should evaluate 
visibility impacts at more distant Class I areas for the Montana FIP.
    Response: We modeled all Class I areas within 300 km of the BART 
source. As discussed in a response to a previous comment, the IWAQM 
Phase 2 report concluded that CALPUFF can overestimate surface 
concentrations at distances of 300 to 1,000 km by a factor of 3 to 4. 
Therefore, IWAQM recommends use of CALPUFF for transport distances of 
approximately 200 km or less. Use of CALPUFF for characterizing 
transport beyond 200 to 300 km should be done cautiously with an 
awareness of the likely problems involved. Therefore, we modeled Class 
I areas within 300 km of each BART source. We did not model impacts at 
distances exceeding 300 km.
    In the case of the Big Stone I facility in South Dakota, there were 
no Class I areas within a distance of 300 km of the source. Therefore, 
the State and the facility agreed in their modeling protocol to 
evaluate visibility impacts at more distant sources by using a non-
regulatory option in CALPUFF called ``puff splitting''. As discussed in 
the IWAQM Guidance,\5\ the use of the puff splitting option in CALPUFF 
might improve model performance at long distances, but there are also 
serious conceptual concerns with the use puff splitting to represent 
puff dispersion for very long-range transport at distances of more than 
300 km. EPA concurred with South Dakota on this approach for Big Stone 
I because there were no Class I areas within 300 km of the source, and 
EPA approved the South Dakota SIP using these modeling results. In the 
case of Montana, there are several Class I areas less than 300 km from 
each BART source, and EPA based its analysis on CALPUFF visibility 
model results for these areas.
---------------------------------------------------------------------------

    \5\ IWAQM Phase 2 report, p. 27.
---------------------------------------------------------------------------

    EPA did not use the non-regulatory puff splitting option in CALPUFF 
to model more distant sources because of

[[Page 57869]]

the greater uncertainty in model results at distances of more than 300 
km, as we have explained in previous responses.
    While WRAP performed CALPUFF modeling at Class I areas more distant 
than 300 km from Colstrip, WRAP also recognized the larger uncertainty 
in the model results for distances greater than 300 km. and included 
the following caveat in their modeling protocol:

    Relevant guidance suggests that the CALPUFF model is generally 
applicable at distances from 50 km to 300 km downwind and may be 
used for distance less than 50 km when complex flows exist on a case 
by case basis. [citation omitted] Class I areas in the west 
generally are located in complex terrain resulting in complex flows. 
Consequently, the BART screening modeling conducted by the RMC will 
include results for potential BART eligible sources that reside 
within 50 km of a Class I area. The WRAP RMC BART screening modeling 
may also apply CALPUFF to downwind distances greater than 300 km. 
When providing results to the States, the downwind distance between 
the BART source and the Class I area will be included, and a 
recommendation from the RMC as to the utility of applying the 
results for Class I areas less than 50 km and greater than 300 km 
from the source. The individual States will need to make their own 
regulatory assessment of the applicability of the model results at 
those distances less than 50 km and greater than 300 km.\6\
---------------------------------------------------------------------------

    \6\ CALMET/CALPUFF Protocol for BART Exemption Screening 
Analysis for Class I areas in the Western United States Available at 
http://pah.cert.ucr.edu/aqm/308/bart/WRAP_RMC_BART_Protocol_Aug15_2006.pdf.

    It also should be noted that WRAP found smaller visibility impacts 
at the distances of more than 300 km compared to Class I areas at 
distances of less than 300 km.\7\ The BART Guidelines explain that if 
the highest modeled effects are observed at the nearest Class I area, 
it may not be necessary to model other Class I areas. The BART 
Guidelines state:
---------------------------------------------------------------------------

    \7\ Summary of WRAP RMC BART Modeling for Montana, Draft 
5 May 30, 2007. More information can be found at http://pah.cert.ucr.edu/aqm/308/bart.shtml.

    One important element of the protocol is in establishing the 
receptors that will be used in the model. The receptors that you use 
should be located in the nearest Class I area with sufficient 
density to identify the likely visibility effects of the source. For 
other Class I areas in relatively close proximity to a BART-eligible 
source, you may model a few strategic receptors to determine whether 
effects at those areas may be greater than at the nearest Class I 
area. For example, you might choose to locate receptors at these 
areas at the closest point to the source, at the highest and lowest 
elevation in the Class I area, at the IMPROVE monitor, and at the 
approximate expected plume release height. If the highest modeled 
effects are observed at the nearest Class I area, you may choose not 
to analyze the other Class I areas any further as additional 
---------------------------------------------------------------------------
analyses might be unwarranted.

70 FR 39170.
    Comment: Commenters stated that EPA should have added the 
visibility impacts at each Class I area to assess cumulative visibility 
impacts.
    Response: Contrary to the commenter's assertion, we did assess 
cumulative visibility impacts. In our analysis of visibility impacts, 
we considered the visibility improvement at all Class I areas within 
300 km of the subject BART unit. For example, in our analysis of BART 
control options for Corette, we considered the visibility improvement 
at all Class I areas within 300 km (Gates of the Mountains WA, North 
Absaroka WA, Red Rock Lakes WA, Teton WA, UL Bend WA, Washakie WA, and 
Yellowstone NP). 77 FR 24042 and 77 FR 24046. In our proposal, for each 
of the BART sources we assessed the visibility improvement at each 
Class I area within 300 km of the source associated with the controls 
under consideration, as well as the number of days with a greater than 
0.5 deciview impact at each of these Class I areas. Therefore, our 
proposed rule did not ignore the visibility improvement that would be 
achieved at areas other than the most impacted Class I area, and we 
disagree with the assertions that we did not consider the impacts at 
multiple Class I areas. We did, however, in the proposed rule focus on 
the visibility benefits at those Class I areas with the most meaningful 
visibility impacts in determining whether NOX or 
SO2 controls should be determined to be BART. We took a 
similar approach for all the Montana BART units. We did not ignore the 
visibility benefits at the other Class I areas but did not consider the 
benefits sufficient to warrant a change in our determination as to the 
appropriate level of control.
    Comment: USFWS stated that for the three SO2 control 
alternatives, EPA made judgments on cost per deciview based on only the 
most impacted Class I area, Washakie WA and that USFWS continued to 
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. USFWS stated that it does not make sense to use the 
same metric 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 and that it does not make sense to 
evaluate impacts at one Class I area, while ignoring others that are 
similarly significantly impaired. USFWS stated that if emissions from 
Corette are reduced, the benefits will be spread well beyond only the 
most impacted Class I area, and this must be accounted for. USFWS 
stated that, in the context of the multiple Class I areas that are 
affected by Corette, the Lime Spray Dryer (LSD) SO2 control 
alternative, the cumulative Class I area impact is $12.7 million per 
deciview of visibility improvement and costs $4,981 per ton of 
SO2 removed USFWS stated that LSD should be considered as 
being a viable candidate for BART for Corette. USFWS made similar 
comments regarding NOX controls for Corette.
    Response: We disagree. In our analysis of visibility impacts, we 
considered the visibility improvement at all Class I areas within 300 
km of the subject BART unit. As explained in the response to the 
previous comment, in our analysis of BART control options for Corette, 
we considered the visibility improvement at all Class I areas within 
300 km. In our proposal, for each of the BART sources we assessed the 
visibility improvement at each Class I area within 300 km of the source 
associated with the controls under consideration, as well as the number 
of days with a greater than 0.5 deciview impact at each of these Class 
I areas. Therefore, our proposed rule did not ignore the visibility 
improvement that would be achieved at areas other than the most 
impacted Class I area, and we disagree with the assertions that we did 
not consider the impacts at multiple Class I areas. We did, however, in 
the proposed rule focus on the visibility benefits at those Class I 
areas with the most meaningful visibility impacts in determining 
whether NOX or SO2 controls should be determined 
to be BART. We did not ignore the visibility benefits at the other 
Class I areas but did not consider the benefits sufficient to warrant a 
change in our determination as to the appropriate level of control. As 
we explained in other responses, we did not use the $/deciview ratio as 
a basis for our decision.
    Comment: EarthJustice's consultant Air Resources Specialists (ARS) 
performed additional analysis on possible visibility benefits of SCR at 
Colstrip Units 1 and 2 combined with the benefits of BART controls on 
SO2 emissions. The commenter stated that the ARS analysis 
``demonstrates that EPA's analysis of visibility benefits of selective 
catalytic reduction (SCR) controls is incomplete and inadequate.'' The 
commenter also stated, ``the evidence demonstrates that with SCR and 
SO2 controls, the visibility impairment at UL Bend WA and 
Theodore Roosevelt NP attributable to

[[Page 57870]]

Colstrip would be virtually eliminated, the very goal of the CAA haze 
requirements.''
    The commenter also stated that when SCR + SOFA is coupled with a 
dry scrubber/baghouse, it is likely that Corette would no longer have 
any noticeable impact on haze in any Class I area, and this result 
complies with the Congressional directive to eliminate haze in Class I 
areas.
    Response: We disagree that our analysis was incomplete or 
inadequate. We analyzed visibility benefits for both SO2 and 
NOX emissions reductions following procedures established in 
the BART Guidelines, and we proposed emissions reductions consistent 
with the five factor analysis. The Regional Haze Rule has a goal that 
anthropogenic visibility impairment be eliminated by 2064; however, it 
does not require that all anthropogenic contributions to visibility 
impacts be fully eliminated in the near term, nor is that the goal of 
the BART element of the Regional Haze program. 40 CFR 51.308 
(e)(1)(ii)(A) requires that EPA consider the cost of compliance; the 
energy and nonair quality environmental impacts; any pollution control 
equipment in use at the source; the remaining useful life of the 
source; and the degree of improvement which may be reasonably 
anticipated to result from the use of such technology. Visibility 
improvement is only one of the five factors that are required to be 
considered. Our proposed BART controls achieve significant reductions 
in contributions to visibility impairment while also considering other 
components of the five factor analysis.
    Comment: EarthJustice stated that, ``ARS concluded that the 
incremental benefit of SCR compared to SNCR at Colstrip Units 1 and 2 
is larger when viewed in combination with the SO2 emission 
controls at either emission rate.''
    Response: ARS estimated the relative improvement in SCR compared to 
SNCR for the case with baseline SO2 emissions and for the 
case with our proposed BART SO2 emissions. The ARS analysis 
showed that the incremental improvement in SCR compared to SNCR was 
almost identical for the 98% worst days regardless of the level of 
SO2 emissions used. For example, in EPA's analysis the 
incremental improvement of SCR over SNCR for Theodore Roosevelt NP was 
0.27, 0.23, and 0.28 deciview, respectively, for 2006, 2007 and 2008. 
The ARS analysis found incremental improvements of 0.28, 0.26, and 0.28 
deciview, respectively, for 2006, 2007 and 2008. Moreover, ARS did not 
perform additional CALPUFF simulations for this analysis, but only 
combined estimates of extinction contributions from different CALPUFF 
simulations.
    Comment: EarthJustice stated that that we aggregated Colstrip Units 
1 and 2 for assessing visibility benefits of SNCR, but arbitrarily kept 
our assessment of benefits of SCR segregated by unit.
    Response: We disagree. Modeling was performed in the same manner 
for SCR as for SNCR. The modeling protocol, results, and final report 
were available in the docket. Our evaluation of the visibility benefits 
was made in consideration of all of the modeling results, which 
includes a visibility improvement assessment for application of SCR at 
Colstrip Units 1 and 2 individually, as well as an assessment of the 
total visibility benefit from application of SCR at both units 
collectively.
    Comment: A commenter stated that we failed to examine the 
collective visibility benefit of SCR in combination with SO2 
upgrades at Colstrip Units 1 and 2.
    Response: We examined the individual benefits of NOX and 
SO2 controls to be able to assess the difference between 
pollutant-specific control options. Our evaluation of the visibility 
benefits was made in consideration of all of the modeling results.
    Comment: EarthJustice stated that their contractor (ARS) performed 
AERMOD simulations to evaluate the impacts of Colstrip SO2 
emissions relative to the 1-hour average SO2 National 
Ambient Air Quality Standard (NAAQS) and reported modeled violations of 
the SO2 NAAQS.
    Response: EPA will address compliance with the 1-hour average 
SO2 NAAQS separately from Regional Haze requirements. It is 
beyond the scope of this rulemaking. It will be addressed by EPA at a 
later date.
    Comment: Holcim commented that EPA discarded all prior modeling and 
developed a new modeling analysis in 2011. Holcim stated that EPA did 
not explain why it used a new modeling analysis and that EPA's BART 
conclusions are therefore based on modeling that is not transparent and 
not available for review.
    Response: We disagree. As we explained in our proposal, we used 
CALPUFF modeling to evaluate emissions control scenarios that were 
consistent with the application of control scenarios for the Montana 
sources that were subject to BART. We did this because we were unable 
to obtain the modeling files from some of the sources and we wanted 
each source to be modeled consistently. The modeling protocol, final 
report, and all related files were available for review in the docket.
    Comment: The Western Environmental Trade Organization (WETA) 
commented that the EPA recently approved the SIP for regional haze 
developed by the State of North Dakota. WETA explained that the North 
Dakota plan relied on extensive modeling that demonstrated emissions 
control technology installations at certain facilities would result in 
insignificant improvement in visibility. WETA requested that the EPA 
develop a visibility plan for Montana that offers the same flexibility 
and cost-effective standards included in North Dakota's plan.
    Response: WETA did not explain what flexibility it was seeking; 
therefore, we are not able to evaluate whether such flexibility could 
be accommodated. To the extent that WETA is stating that our proposed 
requirements are not cost-effective, we disagree. To the extent that 
WETA is stating that we are being inconsistent with decisions we made 
for regional haze in North Dakota, we disagree. We have responded to 
more specific comments on the cost-effectiveness of controls elsewhere.
    Comment: The commenter stated that EPA's proposed BART 
determinations for Colstrip Units 1& 2 are erroneous because EPA's 
modeling failed to include actual air quality measurements, including 
visual quality measurements, in its inputs to its regional haze model. 
The commenter further stated that real air quality data for Class I 
areas is critical to determining what the degree of visibility 
improvement may be in a given Class I area.
    Response: EPA used ambient monitoring data to evaluate the CMAQ and 
CAMx grid model simulations that were used for modeling the uniform 
rate of progress toward natural visibility conditions. However, the 
commenter appears to be referring specifically to the CALPUFF model 
simulations used to evaluate visibility impacts of BART sources. The 
BART Guidelines require that visibility impacts from BART sources be 
evaluated in comparison to natural visibility conditions. The 
procedures used to estimate natural visibility conditions are described 
in the ``Guidance for Estimating Natural Visibility Conditions Under 
the Regional Haze Rule.'' \8\ It would be

[[Page 57871]]

inappropriate to use ambient monitoring data for current degraded 
visibility conditions in the evaluation of BART source visibility 
impacts. EPA previously considered and responded to the comment that 
current visibility conditions should be used in BART source evaluations 
in 40 CFR part 51, appendix Y, promulgated at 70 FR 39104. EPA 
considered the approach of assessing a BART-eligible source's impacts 
on visibility by using current or near-term future conditions, and EPA 
determined that BART visibility impacts should be evaluated in 
comparison to natural background visibility. In the final rulemaking 
EPA wrote:
---------------------------------------------------------------------------

    \8\ Guidance for Estimating Natural Visibility Conditions Under 
the Regional Haze Rule, U.S. Environmental Protection Agency, 
September 2003. Can be found at: http://www.epa.gov/ttncaaa1/t1/memoranda/rh_envcurhr_gd.pdf.

    Using existing conditions as the baseline for single source 
visibility impact determinations would create the following paradox: 
The dirtier the existing air, the less likely it would be that any 
control is required. This is true because of the nonlinear nature of 
visibility impairment. In other words, as a Class I area becomes 
more polluted, any individual source's contribution to changes in 
impairment becomes geometrically less. Therefore the more polluted 
the Class I area would become, the less control would seem to be 
needed from an individual source. We agree that this kind of 
calculation would essentially raise the ``cause or contribute'' 
applicability threshold to a level that would never allow enough 
emission control to significantly improve visibility. Such a reading 
would render the visibility provisions meaningless, as EPA and the 
States would be prevented from assuring ``reasonable progress'' and 
fulfilling the statutorily-defined goals of the visibility program. 
Conversely, measuring improvement against clean conditions would 
---------------------------------------------------------------------------
ensure reasonable progress toward those clean conditions.

70 FR 39124.
    Therefore, EPA correctly used estimates of natural visibility 
conditions in our evaluation of BART source visibility impacts, and we 
disagree with the comment that we failed to appropriately use air 
quality data at Class I areas.
    Comment: EarthJustice stated that they do not agree with EPA's 
approach to use the fifth factor in determining the degree of 
visibility improvement from emissions control technologies where EPA 
adds an additional incremental benefit factor with an apparent but 
unstated threshold for improvement sufficiency that is contrary to the 
purpose and direction of the CAA.
    Response: We disagree that we only evaluated visibility benefit on 
an incremental basis and that we used a threshold for improvement 
sufficiency. In the proposed FIP, we included tables showing the 
visibility improvement for control options as compared to baseline 
conditions. Incremental improvement can be easily calculated from the 
data in the tables, however, we did not calculate this separately for 
each option. In addition, our modeling protocol, modeling report and 
tables of results were included in the docket.
    Comment: Commenters stated that we used incorrect baselines for 
modeling impacts from sources at Corette and Colstrip.
    Response: We explain our rationale for the chosen baseline periods 
in responses to other comments.

B. General Comments on BART

    Comments: Montana Department of Environmental Quality (MDEQ) stated 
that EPA should have used a dollar-per-deciview ($/deciview) metric 
rather than the $/ton metric to evaluate BART and reasonable progress. 
MDEQ argued that the use of deciviews is consistent with the Regional 
Haze Rule, which expresses Reasonable Progress Goals (RPGs), baseline 
visibility, current visibility conditions and natural conditions in 
deciviews. MDEQ also referenced both the BART Guidance and the 
Reasonable Progress Guidance to support this argument.
    The NPS stated that one of the options suggested by the BART 
Guidelines to evaluate cost-effectiveness is cost/deciview and that the 
NPS believes that visibility improvement must be a critical factor in 
any program designed to improve visibility. The NPS stated that 
compared to the typical control cost analysis in which estimates fall 
into the range of $2,000-$10,000 per ton of pollutant removed, spending 
millions of dollars per deciview to improve visibility may appear 
extraordinarily expensive, but that the NPS compilation of BART 
analyses across the United States reveals that the average cost per 
deciview proposed by either a state or a BART source is $14-$18 
million, with a maximum of $51 million per deciview proposed by South 
Dakota at the Big Stone I power plant. The NPS noted that even though 
it has no Class I areas, Nebraska Department of Environmental Quality 
has chosen $40 million/deciview as a cost criterion, which is also 
above the national average. The NPS compared its estimates for annual 
cost of adding SOFA + SCR to EPA's estimates for visibility impacts and 
stated that the cost-effectiveness of adding SOFA + SCR to improve 
visibility at the five Class I areas modeled by EPA is less than $10 
million/deciview and significantly less than the $14-$18 million/
deciview national average of BART proposals and determinations.
    Response: For BART, the BART Guidelines require that cost 
effectiveness be calculated in terms of annualized dollars per ton of 
pollutant removed, or $/ton. 70 FR 39167. MDEQ and the NPS are correct 
in that the BART Guidelines allows for the $/deciview ratio as an 
additional cost effectiveness metric that can be employed along with $/
ton for use in a BART evaluation. However, the use of this metric 
further implies that additional thresholds or notions of acceptability, 
separate from the $/ton metric, would need to be developed for BART 
determinations. We have not used this metric for BART purposes because 
(1) It is unnecessary in judging the cost effectiveness of BART, (2) it 
complicates the BART analysis, and (3) it is difficult to judge. The $/
deciview metric has not been widely used and is not well-understood as 
a comparative tool. In our experience, $/deciview values tend to be 
very large because the metric is based on impacts at one Class I area 
on one day and does not take into account the number of affected Class 
I areas or the number of days of improvement that result from 
controlling emissions. In addition, the use of the $/deciview suggests 
a level of precision in the CALPUFF model that may not be warranted. As 
a result, the $/deciview can be misleading. We conclude that it is 
sufficient to analyze the cost effectiveness of potential BART controls 
using $/ton, in conjunction with an assessment of the modeled 
visibility benefits of the BART control. Within the context of 
reasonable progress, the Guidance for Setting Reasonable Progress Goals 
Under the Regional Haze Program, states that ``[y]ou should evaluate 
both average and incremental costs.'' \9\ This is consistent with the 
approach under BART. As commenters note, the guidance then stated that 
``simple cost effectiveness estimates based on a dollar-per-ton 
calculation may not be as meaningful as a dollar-per-deciview 
calculation, especially if the strategies reduce different groups of 
pollutants.'' However, the guidance makes this statement on the basis 
that ``different pollutants differently impact visibility impairment.'' 
That is, for example, a one ton reduction in SO2 would have 
a greater visibility benefit than a one ton reduction of coarse mass. 
As only SO2 and NOX controls were evaluated for 
the reasonable progress point sources, the use of the $/deciview is not 
particularly

[[Page 57872]]

relevant or informative. In addition, we did not use the $/deciview 
metric for our evaluation of reasonable progress controls for largely 
the same reasons as stated above for BART controls.
---------------------------------------------------------------------------

    \9\ Guidance for Setting Reasonable Progress Goals Under the 
Regional Haze Program, U.S. Environmental Protection Agency, June 1, 
2007, p.5-2.
---------------------------------------------------------------------------

    Comment: The NPS stated that we used inconsistent criteria in 
selecting BART controls.
    Response: We disagree. As explained later, pursuant to 40 CFR 
51.308(e)(1)(ii)(A) we considered the following five factors in our 
analysis: The cost of compliance; the energy and nonair quality 
environmental impacts; any pollution control equipment in use at the 
source; the remaining useful life of the source; and the degree of 
improvement which may be reasonably anticipated to result from the use 
of such technology. The Regional Haze Rule defines BART as the best 
system of continuous emission control technology available and 
associated emission reductions achievable, as determined through an 
analysis of these five factors. The NPS is correct in that the BART 
Guidelines allows for the $/deciview ratio as an additional cost 
effectiveness metric that can be employed along with $/ton for use in a 
BART evaluation of the five statutory factors. 70 FR 39126 to 70 FR 
39127. While the Regional Haze Rule may not prevent us from 
establishing a bright line for some of the factors such as cost-
effectiveness and visibility, we are not required to do so, and have 
not done so for this action as the cost and visibility factors are both 
weighed in making control decisions. Also, while the BART Guidelines 
allows for the $/deciview ratio as an additional cost effectiveness 
metric that can be employed along with $/ton for use in a BART 
evaluation, we have not used this metric in our evaluation. As 
explained in our determinations for each source, the cost effectiveness 
of controls on a dollar per ton basis and the visibility benefit of 
those controls were the two factors that had the most influence over 
our decision.
    Comment: MDEQ stated that in the North Dakota Regional Haze SIP/
FIP, coal-fired utilities with much greater estimated visibility impact 
were required to install controls similar to those required at Colstrip 
1 and 2.
    Response: We disagree that certain BART determinations from the 
North Dakota Regional Haze SIP/FIP are appropriate comparisons to our 
BART determinations in this FIP. Our determination on the 
NOX BART determinations at Milton R. Young Station Units 1 
and 2 and Leland Olds Station Unit 2\10\ is explained in our final 
action for regional haze for North Dakota. 77 FR 20893. Our BART 
determinations were made on a source- specific basis in consideration 
of the five statutory factors.
---------------------------------------------------------------------------

    \10\ We presume these units are the ``coal-fired utilities'' to 
which MDEQ is referring.
---------------------------------------------------------------------------

    Comment: MDEQ stated that we ``accept, discard or include new cost 
information without reason or justification.'' MDEQ supported this 
claim by arguing that we used Integrated Planning Model (IPM) data in 
one instance, but used costs provided by sources and an outside 
consultant instead of IPM data for the North Dakota Regional Haze SIP/
FIP.
    Response: The BART Guidelines provide some flexibility in how to 
calculate and consider costs. 70 FR 39127. Generally, we followed a 
reasonable and supported approach. We have responded to specific 
comments regarding our cost analysis in other responses.
    Comment: MDEQ stated that the averaging times and compliance 
demonstrations for Colstrip 1 & 2, Corette and Devon Energy are not 
practically enforceable, and therefore counter to the BART Guidelines. 
MDEQ stated that the 30-day rolling average particulate matter (PM) 
emission limits for Colstrip 1, Colstrip 2 and Corette, and the 
NOX limit for Devon are not enforceable with an annual stack 
test.
    Response: We disagree with some aspects of this comment and have 
made changes in the final FIP to clarify requirements in response to 
other aspects of this comment. In the proposed FIP, we concluded that 
annual stack tests, along with emissions monitoring in accordance with 
the applicable Compliance Assurance Monitoring (CAM) plan are 
sufficient to determine compliance with BART PM limits. 77 FR 24099 
(April 20, 2012). In its comments, MDEQ provides no evidence to the 
contrary aside from the general statements about practical 
enforceability described in the comment above. With regard to the Devon 
Energy Reasonable Progress determination, we have revised the 
monitoring, recordkeeping and reporting requirements in the final FIP. 
We have also clarified in a correction notice that the PM limits listed 
at 40 CFR 52.1396 are not based on a 30-day average. 77 FR 29270.
    Comment: MDEQ noted that Cross-State Air Pollution Regulation 
(CSAPR) trading programs were recently determined by EPA to be an 
alternative to source-by-source BART determinations. 77 FR 33642 (April 
20, 2012). MDEQ argued that, because CSAPR is a health-based standard, 
``EPA in the East is advocating the position that Montana has taken for 
our own state: Realize the benefits (including visibility) from health-
based standards and make compliance with those standards the 
demonstration for BART.''
    Response: Emissions trading programs and other alternative programs 
can be used in place of source specific BART controls ``as long as the 
alternative provides greater reasonable progress towards improving 
visibility than BART.'' 77 FR 33644. Because Montana is not within the 
geographic areas covered by CSAPR, and because the State did not submit 
an emissions trading program or alternative program that was subject 
to, let alone satisfied, the ``greater reasonable progress'' test, EPA 
does not agree that compliance with other standards may replace a BART 
demonstration for sources subject to BART in Montana.
    Comment: A commenter claimed that our elimination of best emission 
controls based on incremental benefit is not legally supportable and 
that EPA's analyses do not satisfy the purpose or the regulatory 
requirements for BART determinations. The commenter stated that we 
applied additional filters with unstated thresholds or standards in our 
consideration of BART and that those filters eliminate or significantly 
diminish the weight and importance of the required five factors. The 
commenter stated that EPA used an incremental benefit test and reached 
a subjective conclusion.
    Response: We disagree that our determinations are not legally 
supportable. Pursuant to 40 CFR 51.308(e)(1)(ii)(A) we considered the 
following five factors in our analysis: The cost of compliance; the 
energy and nonair quality environmental impacts; any pollution control 
equipment in use at the source; the remaining useful life of the 
source; and the degree of improvement which may be reasonably 
anticipated to result from the use of such technology. The Regional 
Haze Rule defines BART as the best system of continuous emission 
control technology available and associated emission reductions 
achievable, as determined through an evaluation of the five statutory 
factors. 70 FR 39126 to 70 FR 39127. While the Regional Haze Rule may 
allow us to establish a bright line for some of the factors such as 
cost-effectiveness and visibility, we are not required to do so, and 
have not done so for this action.
    Comment: MDEQ commented that EPA makes a case for ordering the 
installation of control equipment for measurable emissions reductions 
absent a visibility improvement goal to achieve reasonable progress as 
measured in deciviews. MDEQ stated that one of the

[[Page 57873]]

factors to consider when determining BART is any existing pollution 
control technology in use at the source and that EPA may be 
interpreting this provision to mean BART requires the installation of 
any new pollution control technology that is useful for reducing 
emissions generally. MDEQ stated that the statute and the Regional Haze 
Rule are both clear that a BART determination focuses on existing 
pollution controls and that the suitability of additional controls for 
co-beneficial purposes that may be tangentially related to the National 
Goal is not part of the analysis. MDEQ stated that overall purpose of 
any SIP, including Montana's, is the control of emissions to comply 
with the NAAQS as set forth in 42 U.S. Code (USC) Section7410 and that 
the purpose of the Regional Haze Rule is to control emissions that 
cause or contribute to visibility impairment in Class I Federal areas. 
MDEQ stated that, ``Montana is adamant on this point because it forms 
the basis for its reluctant renunciation of authority over Montana's 
BART program.'' MDEQ stated that, ``the consideration of a co-benefit 
strategy is not without merit, but the imposition of BART is set forth 
very clearly in statute and rule. MDEQ stated that the determination of 
BART has everything to do with visibility impairment and improvement, 
not the attainment or maintenance of the NAAQS.'' MDEQ suggested that, 
``EPA limit the BART criteria to that set forth in the rule at 40 CFR 
51.308(e) and refuse to propose new controls that are not calculated to 
fulfill BART criteria.''
    Response: We disagree that we have misinterpreted the BART 
provision to consider any existing pollution control technology at the 
source. We point out that considering existing pollution control 
technology in use at the source does not preclude the consideration of 
new technology. As listed in the BART Guidelines, Step 1 of the ``Five 
Basic Steps of a Case-by-Case BART Analysis'' is ``Identify All 
Available Retrofit Technologies.'' 70 FR 39164. A footnote to the word 
``All'' in this step of the BART Guidelines reads as follows; ``In 
identifying `all' options, you must identify the most stringent option 
and a reasonable set of options for analysis that reflects a 
comprehensive list of available technologies. It is not necessary to 
list all permutations of available control levels that exist for a 
given technology--the list is complete if it includes the maximum level 
of control each technology is capable of achieving.'' 70 FR 39164. Our 
analysis for each Montana source subject to BART included each of the 
``Five Basic Steps of a Case-by-Case BART Analysis,'' as well as a 
complete five-factor analysis which included consideration of ``any 
existing pollution control technology in use at the source.'' Existing 
pollution control technology was considered when identifying available 
control options, when establishing a baseline for determining 
visibility impacts or for determining annual emission reductions for 
available control options. Existing pollution control technology also 
was considered in establishing emission limits. With regard to MDEQ's 
comment that we interpreted this provision to mean BART requires the 
installation of any new pollution control technology that is useful for 
reducing emissions generally, we point out that in many cases our BART 
determinations did not require additional pollution control technology 
to be installed for BART.
    We also disagree that we have interpreted BART to require the 
installation of any new pollution control technology that is useful for 
reducing emissions generally, that we used criteria other that those 
listed at 40 CFR 51.308(e)(1)(ii)(A), or that we proposed new controls 
that are not calculated to fulfill BART criteria. As stated in other 
responses, pursuant to 40 CFR 51.308(e)(1)(ii)(A) we considered the 
five factors in our analysis.. The Regional Haze Rule defines BART as 
the best system of continuous emission control technology available and 
associated emission reductions achievable, as determined through an 
evaluation of the five statutory factors. 70 FR 39126 to 70 FR 39127. 
As stated in another response, at no point in the proposed FIP did we 
discuss public health impacts as a consideration in our analyses, as 
they were not. As stated elsewhere, we agree that the Regional Haze 
Rule is not a health-based standard, and that we are not authorized to 
consider public health impacts in promulgating our FIP for purposes of 
this action.
    Comment: The NPS commented that EPA determined that the incremental 
visibility improvement from a control option must exceed 0.5 deciview 
at a given Class I area if costs exceed $5,000/ton in order to qualify 
as BART and stated that the NPS disagrees with this approach. The NPS 
stated that while the BART Guidelines do recommend estimation of 
incremental costs, it makes no mention of an incremental visibility 
improvement test. The NPS explained that if applied linearly, EPA's 
cost estimate of $3,235/ton for SCR would correspond to a visibility 
improvement of 0.32 deciview, not 0.5 deciview to justify SCR. The NPS 
stated that EPA concluded the benefit of SCR at Theodore Roosevelt NP 
is 0.4 deciview and that therefore, by EPA criteria SCR is BART for 
each Units 1 and 2.
    Response: We disagree. We have not determined that the incremental 
visibility improvement from a control option must exceed 0.5 deciview 
at a given Class I area if costs exceed $5,000/ton in order to qualify 
as BART. As stated in other responses, while the Regional Haze Rule may 
allow us to establish a bright line for some of the factors such as 
cost-effectiveness and visibility, we are not required to do so, and 
have not done so for this action.

C. Comments on Cement Kilns

    Comment: A commenter stated that we must not exempt cement kilns 
from BART for PM. The commenter described baseline visibility impacts 
from Ash Grove and Holcim and stated that the high degree of visibility 
impairment warrants analysis of whether PM emission limits should be 
lower to reflect BART.
    Response: We disagree that we have exempted cement kilns from BART 
for PM. In our proposal, Table 35 shows that Ash Grove's greatest 
baseline visibility impact is 4.446 deciviews at Gates of the Mountains 
WA and Table 49 shows that Holcim's greatest baseline visibility impact 
is .980 deciview at Gates of the Mountains WA. 77 FR 24011 and 77 FR 
24017. While we agree with the commenter that the baseline impacts are 
significant, the PM contribution to this overall baseline impact is 
small. In our proposal, Table 38 shows that for Ash Grove, coarse PM 
only contributes 0.84% and fine PM only contributes 4.77% to the 
overall baseline visibility impact of 4.446 deciviews. 77 FR 24013. 
Table 64 shows that for Holcim, coarse PM only contributes 5.79% and 
fine PM only contributes 12.61% to the overall baseline visibility 
impact of .980 deciview. 77 FR 24022. By contrast, our BART 
determination for Ash Grove for NOX is anticipated to 
achieve a visibility improvement of 1.248 deciviews and our BART 
determination for Holcim is anticipated to achieve a visibility 
improvement of 0.424 deciview. Any visibility improvement that could be 
achieved with improvements to the existing PM controls would be 
negligible. BART for PM was based on using the existing control 
equipment and the emission limit established in each facility's Title V 
permit. The PM BART limits for Ash Grove and Holcim were listed in our 
proposal at 77 FR 24098 and are

[[Page 57874]]

codified by our final action at 40 CFR 52.1396.

D. Comments on Ash Grove

    Comment: Ash Grove stated that they did not object to EPA's 
conclusion that BART should be based on the installation of low 
NOX burner (LNB) and SNCR. However, the company stated that 
they objected to the assumptions made about what SNCR can achieve. Ash 
Grove stated that they explained in the prior correspondence that the 
company did not believe that it is appropriate to assume that the 
Montana City kiln can achieve 50% control effectiveness. Ash Grove 
stated that, as the data in Table 10 of the preamble clearly showed, 
only one of the three kilns at Ash Grove's Midlothian plant is able to 
achieve 50% control effectiveness while the other two kilns had an 
average control efficiency of 37.7% and 40.5%.
    Ash Grove also believes that no other credible evidence is provided 
for our conclusion as to SNCR NOX control efficiency. Ash 
Grove stated that we referenced studies from other industry sectors and 
a marketing brochure from Cadence stating that ``control efficiency of 
up to 50% can be achieved on long wet kilns'' and that this quote 
states the upper end of what might be achievable. Ash Grove indicated 
that the brochure does not state that 50% control efficiency can be 
achieved on all long wet kilns, that Cadence's experience with SNCR on 
long wet kilns is what is shown in Table 10, Ash Grove indicated that 
Cadence was Ash Grove's partner in developing the Midlothian SNCR, 
which, according to Ash Grove, are the only long wet kilns in the 
United States with any track record of SNCR use. Ash Grove indicated 
that even after years of optimization on the Midlothian kilns, the data 
show that it has not been possible to bring all three kilns up to a 50% 
control efficiency and that the Midlothian NOX reduction 
data reflect not only the benefits of SNCR, but also the mid-kiln 
firing of tires, use of a mid-kiln fan and other technologies that are 
not available to the Montana City kiln, but that were implemented 
concurrent with the SNCR installation/optimization at Midlothian to 
reduce NOX emissions. Ash Grove explained that in 
considering the Midlothian data, one needs to account for the direct 
control efficiency these technologies provide, in addition to the 
synergistic effects of using more than one control device/technique at 
a time at Midlothian and that these benefits would not be available at 
Montana City and should not be assumed.
    Ash Grove summarized that they continued to believe that a SNCR 
system at Montana City cannot be assumed to reach greater than 35% 
control efficiency and that EPA has produced no credible evidence in 
the record for supporting a different conclusion. Ash Grove stated that 
they recognized that their initial BART submittal listed 50% control as 
achievable for the combination of a low NOX burner and SNCR 
at the Montana City kiln but since then they have realized they cannot 
get to that level on all three kilns at Midlothian. Ash Grove stated 
that they are willing to not contest the 8.0 lb/ton clinker limit, and 
they anticipate that compliance could require additional control 
technologies/strategies; therefore, they need the maximum time 
allowable to find ways to consistently maintain NOX at or 
below that level.
    Response: We disagree that SNCR cannot achieve a 50% control 
effectiveness at Ash Grove. The data from Ash Grove's Midlothian, Texas 
kilns in Table 10 of the proposed FIP, 77 FR 24003, show the SNCR 
control effectiveness achieved. The data were not intended to imply 
that this is the upper bound of what might be achieved. Ash Grove did 
not submit any information demonstrating that this was the maximum 
reduction that could have been achieved. It was not necessary to 
achieve greater reductions from the Midlothian Texas kilns to comply 
with the required emission limit. In Texas, SNCR was used at Midlothian 
to comply with the emission limit established at Texas Administrative 
Code (TAC) 117.3110(a)(1)(B) of 4.0 lb/ton clinker. TAC 117.3110(b) 
allowed an owner or operator of a long wet kiln to comply with the 4.0 
lbs/ton clinker emission limit on the basis of a weighted average for 
multiple cement kilns. Thus, it was not necessary for each individual 
kiln to achieve the maximum percentage reduction possible; one or more 
kilns could emit more than 4.0 lbs/ton clinker as long as the weighted 
average complied with the emission limit.
    Ash Grove has not submitted any data to demonstrate that SNCR was 
optimized in an attempt to achieve the greatest control efficiency 
possible. For the Midlothian kilns, from June-August 2009, the emission 
rate from kiln 1 was 3.7 lbs/ton clinker and the emission rate from 
kiln 2 was 4.8 lbs/ton clinker and from June through August 2010, the 
emission rate from kiln 1 was 2.6 lbs./ton clinker, the emission rate 
from kiln 2 was 4.8 lbs/ton clinker, and the emission rate from kiln 3 
was 4.4 lbs/ton clinker. These emission rates are significantly higher 
than the emission rates from June to August 2008 (an average of 1.8 
lbs/ton clinker for kiln 1, 2.7 lbs/ton clinker for kiln 2, and 2.7 
lbs/ton clinker for kiln 3). An increase in NOX emissions 
over time would not be expected if the SNCR were being optimized.
    With regard to Ash Grove's claim that we need to account for the 
direct control efficiency of other technologies that are not available 
to the Montana City Kiln, the tire-derived fuel system was already 
being used at Midlothian in 2006 and is already accounted for in the 
2006 baseline to which the 2008 post-SNCR emissions are compared.\11\ 
Thus, no further adjustment is necessary. Ash Grove has not provided 
data to demonstrate that a synergistic effect has occurred between mid-
kiln firing of tires and SNCR at Midlothian.
---------------------------------------------------------------------------

    \11\ Letter from Molly Cagle to Chance Goodwin, Initial Control 
Strategy Development for DFW Ozone Nonattainment Area, July 30, 
2010, p. 1.
---------------------------------------------------------------------------

    Ash Grove has not submitted data to support their claim that only 
35% reduction can be achieved with SNCR at the Montana City kiln. All 
of the Midlothian kilns were able to achieve greater than 35% reduction 
with SNCR and there is no information to demonstrate that SNCR was 
optimized to its maximum potential. The BART Guidelines state, ``In 
assessing the capability of the control alternative, latitude exists to 
consider special circumstances pertinent to the specific source under 
review, or regarding the prior application of the control alternative. 
However, you should explain the basis for choosing the alternate level 
(or range) of control in the BART analysis. Without a showing of 
differences between the source and other sources that have achieved 
more stringent emissions limits, you should conclude that the level 
being achieved by those other sources is representative of the 
achievable level for the source being analyzed.'' 70 FR 39166. Ash 
Grove has not demonstrated the differences between their Montana City 
kiln and the Midlothian kilns.
    With regard to Ash Grove's statement that we have not produced 
credible evidence in the record for supporting a greater than 35% 
control effectiveness for SNCR, we provided a detailed explanation in 
our proposed FIP at 77 FR 24003. Ash Grove has used SNCR at its 
Midlothian kilns where it was shown to achieve the reductions ranging 
from 37.7% to 62.5% and these are within the range of control 
effectiveness demonstrated at other kilns. Considering that control 
effectiveness is greater when initial NOX concentrations are 
greater, and that the baseline NOX emissions of the Montana 
City kiln are

[[Page 57875]]

significantly greater than the Midlothian kilns, the Montana City kiln 
would be expected to achieve even greater control effectiveness when 
compared to the Midlothian kilns. 77 FR 24003 and 77 FR 24004.
    Ash Grove's comment that they are willing to not contest the 8.0 
lb/ton clinker limit is noted. With regard to Ash Grove's comment that 
they anticipate that compliance could require additional control 
technologies/strategies and that therefore they need the maximum time 
allowable to find ways to consistently maintain NOX at or 
below that level, we disagree that additional control technologies/
strategies are necessary; however, the final FIP does not require 
specific control technologies/strategies to be used. The final FIP 
allows for the maximum time available to comply with the 8.0 lb/ton 
clinker limit.
    Comment: Ash Grove stated that the company supported the 
conclusions as to what constitutes BART for SO2. Ash Grove 
noted that in the preamble we stated that there is so little 
improvement in visibility associated with implementing add-on 
SO2 controls that there is no basis for requiring such 
controls under BART. Ash Grove stated that Clean Air Act Section 
169A(g)(2) clearly states that ``the degree of improvement in 
visibility which may reasonably be anticipated to result'' must be used 
in evaluating potential BART controls. Ash Grove concluded that given 
the nominal improvement in visibility predicted from add-on controls, 
there is no basis under BART for requiring the addition of such 
controls. Ash Grove stated that the BART program has a very narrow 
statutory focus in that it exclusively addresses visibility improvement 
and that absent a material increase in visibility, the company believes 
that we would have been arbitrary and capricious if we had required 
add-on controls for SO2 utilizing our BART authority. Ash 
Grove stated that the company supported our ultimate conclusion.
    Response: The comment is noted. The final FIP makes no changes to 
the conclusions regarding SO2 controls for Ash Grove.
    Comment: Ash Grove stated that the company supported our conclusion 
that existing PM controls (an electrostatic precipitator (ESP)) 
constitute BART and that ESPs are well-accepted controls for wet kilns. 
Ash Grove stated that their compliance with the filterable particulate 
standard in the process weight rule applicable to the kiln is an 
appropriate limit for ensuring that the ESP is properly operating and 
that annual compliance demonstrations will ensure ongoing compliance. 
Ash Grove stated that they believe that this approach is particularly 
appropriate where the contribution of PM emissions to visibility 
impairment is nominal.
    Response: The comment is noted. The final FIP makes no changes to 
the conclusions regarding PM controls for Ash Grove.
    Comment: Ash Grove requests clarification on whether they must 
comply with BART limits for SO2 and PM within five years of 
the effective date of the rule, as specified in the proposed regulatory 
text at 40 CFR 52.1396(d), or within 180 days for SO2 and 30 
days for PM, as suggested by the preamble to the proposed rule. If the 
intent is to require compliance sooner than five years from the 
effective date, then Ash Grove requests that the rule be renoticed, and 
that if EPA will not allow five years from the effective date, then Ash 
Grove requests that the BART compliance date for these pollutants be 
30/180 days after the effective date, or the Portland cement National 
Emission Standards for Hazardous Air Pollutants (NESHAP) compliance 
date, whichever is later, in order to coordinate with the 
implementation of EPA's Portland cement NESHAP and New Source 
Performance Standard (NSPS) requirements, including installation and 
certification of continuous emission monitoring systems (CEMS). Ash 
Grove stated that the monitoring that EPA is imposing as part of the 
concurrent Portland cement Maximum Achievable Control Technology (MACT) 
rulemaking is very complicated and must be able to work in concert with 
what EPA imposes under this BART rulemaking. Ash Grove also stated that 
critical components of Ash Grove's envisioned monitoring scheme, such 
as installation of clinker weigh belts or the development of slurry 
conversion mechanisms, cannot be implemented within the 180 day period 
after the effective date.
    Response: We agree with aspects of this comment, but disagree with 
others. We agree that there is an omission in the proposed 40 CFR 
52.1396(d). We failed to specify, in the rule language itself, the 
compliance deadline for SO2 of 180 days after the effective 
date of the FIP, and the compliance deadline for PM of 30 days after 
the effective date of the FIP. These deadlines were mentioned in the 
rule preamble. We have corrected the rule language at 40 CFR 52.1396(d) 
to specify these deadlines. For both SO2 and NOX, 
we further clarify that the 180-day deadline is applicable only where 
installation of additional controls is not necessary to comply with the 
BART limit; otherwise the compliance deadline is five years after the 
effective date of our rule.
    We disagree that the compliance deadline should be 30/180 days 
after the FIP effective date, or the Portland cement NESHAP compliance 
date, whichever is later. With regard to ``whichever is later,'' EPA 
does not have the option of specifying an open-ended compliance 
deadline for BART. In our FIP proposal at 77 FR 23993, we explained 
that ``Once EPA has made its BART determination, the BART controls must 
be installed and in operation as expeditiously as practicable, but no 
later than five years after the date of the final FIP. CAA section 
169(g)(4) and 40 CFR 51.308(e)(1)(iv).'' Ash Grove's comment does not 
dispute this explanation.
    Further, Ash Grove has not presented any specific reason for us to 
wait on the Portland cement MACT rulemaking before imposing PM and 
SO2 monitoring requirements for purposes of BART. First in 
regard to SO2 monitoring, the proposed amendments to the 
Portland cement MACT and NSPS rules do not include any changes to the 
SO2 CEMS monitoring requirements. In the proposed 
amendments, EPA is proposing to correct the emission rate calculation 
formula for SO2 in NSPS Subpart F, at 40 CFR 60.64(c), but 
since we are making the same correction in our final FIP rule (see our 
response below to the comment on NOX and SO2 
emission rate calculation), this is not a valid reason to wait until 
the Portland cement MACT and NSPS amendments are finalized before 
imposing SO2 monitoring in the FIP.
    Further, the proposed amended Portland cement MACT and NSPS rules 
require a SO2 CEMS only if the kiln is subject to an 
SO2 limit under NSPS. Ash Grove has not indicated that their 
kiln in Montana is subject to an SO2 limit under NSPS, and 
even if it is, the proposed amended Portland cement MACT and NSPS rules 
will not impose any different requirements for an SO2 CEMS 
than those in existing NSPS rules at 40 CFR 60.63(f), which are cross-
referenced by our proposed regulatory text at 40 CFR 52.1396(e)(3). Ash 
Grove has also not presented any specific reason, such as vendor 
unavailability or site-specific complications, why it should take more 
than 180 days to replace and certify their SO2 CEMS. We have 
already stated in our FIP proposal that 180 days would allow time for 
monitoring systems to be certified if necessary. We are clarifying that 
CEMS will have to be certified for BART purposes independent of NSPS 
requirements.
    Second, in regard to PM monitoring, the proposed amendments to the

[[Page 57876]]

Portland cement MACT and NSPS rules require a PM continuous parametric 
monitoring system (CPMS), whereas the existing Portland cement MACT and 
NSPS rules require a PM CEMS. Since our FIP proposal does not require 
PM CPMS nor PM CEMS, the proposed amendments to the Portland cement 
MACT and NSPS rules do not affect the FIP and are not a valid reason to 
extend the 30-day compliance deadline for PM in the FIP.
    With regard to Ash Grove's statement that critical components of 
the monitoring scheme, such as installation of clinker weigh belts or 
the development of slurry conversion mechanisms, cannot be implemented 
within the 180 day period after the effective date of the FIP, Ash 
Grove has not presented any specific reason why it should take longer 
than 180 days. With regard to Ash Grove's statement that the clinker 
monitoring must work in concert with the MACT rulemaking, our proposed 
regulatory text at 40 CFR 52.1396(e)(4)(ii) cross-references 40 CFR 
60.63(b) for clinker production monitoring requirements. The proposed 
amendments to the Portland cement MACT and NSPS rules do not change the 
requirements in the existing section 60.63(b) for determining the 
amount of clinker produced. Only minor language clarifications are 
proposed, and there is no change to actual requirements.
    We note that Ash Grove has no issue with the proposed PM BART 
emission limit. However, in preparing responses to Ash Grove's comments 
on other aspects of our proposed FIP, we identified a typographical 
error in our emission limit table for cement kilns. We made a 
correction to the emission limit table for cement kilns at 
52.1396(c)(2), to clarify that the PM emission limit for Ash Grove is 
in lb/hr, not lb/ton clinker. Only the PM emission limit for Holcim is 
in lb/ton clinker. Similarly, we have clarified 40 CFR 52.1396(f)(2) to 
indicate that the emission rate of particulate matter shall be reported 
in lb/hr for Ash Grove, and in lb/ton clinker for Holcim. Ash Grove is 
not required to monitor clinker production for purposes of 
demonstrating compliance with the PM BART limit. We have also included 
in 40 CFR 52.1396(f)(2) the equation for calculating lb/ton clinker for 
PM at Holcim, rather than cross-reference 40 CFR 52.1396(e)(4)(ii), 
which pertains to SO2 and NOX, not PM.
    Comment: Ash Grove does not object to the requirement in our 
proposed regulatory text at 40 CFR 52.1396(e)(3) to maintain, calibrate 
and operate SO2 and NOX CEMS on the cement kiln 
stack. Ash Grove requests, to be consistent with other requirements to 
which they are subject, that the language be revised and proposed 
creating an exception during CEMS breakdown, repairs, calibration 
checks, and zero and span adjustments.
    Response: We agree it is appropriate to address the language for 
consistency purposes. Rather than use the language proposed by Ash 
Grove, we are incorporating language from 40 CFR 60.63(g), which says,

    You must operate the monitoring system and collect data at all 
required intervals at all times the affected source is operating, 
except for periods of monitoring systems malfunctions, repairs 
associated with monitoring system malfunctions, and required 
monitoring system quality assurance or quality control activities 
(including, as applicable, calibration checks and required zero and 
span adjustments).

    We have revised the regulatory text at 40 CFR 52.1396(e)(3) 
accordingly. 40 CFR 60.63(g).
    Comment: Ash Grove also believes it is critical that the facility 
have adequate time to install, shake down and calibrate the necessary 
CEMS equipment. The facility currently lacks a flow meter and does not 
have certified CEMS. As a result, Ash Grove anticipates that it must 
replace its CEMS system, including the data acquisition and handling 
system (DAHS) as part of Portland cement MACT implementation. Ash Grove 
stated that this effort cannot be completed until the Portland cement 
MACT requirements are finalized, as Ash Grove understands that the 
NESHAP monitoring provisions are in flux. Therefore, Ash Grove believes 
that the BART CEMS requirements must be implemented at the same time 
that the Portland cement MACT CEMS requirements are implemented and not 
before.
    Response: We disagree. See our response on compliance deadlines 
above. EPA does not have the option of specifying an open-ended 
compliance deadline for BART. Further, Ash Grove has not presented any 
specific reason, such as vendor unavailability or site-specific 
complications, why it should take longer than 180 days to install a 
flow meter and replace the CEMS system with a certified system. This 
comment has not resulted in any change to our proposal.
    Comment: Ash Grove supports the approach whereby the CEMS data are 
utilized to demonstrate compliance with the NOX and 
SO2 BART limits. However, Ash Grove believes there is a 
material error in the formula used in the proposed regulatory text at 
40 CFR 52.1396(e)(4)(ii). The formula expresses the concentrations of 
SO2 and NOX in grains per standard cubic foot 
(gr/scf). Ash Grove noted that a CEMS would not normally generate 
SO2 or NOX concentrations in gr/scf, but in parts 
per million (ppm), consistent with the requirements of 40 CFR 60, 
Appendix B, Performance Specification 2. Ash Grove recognizes that this 
formula was likely intended to match Equation 3 in the 2010 revised 
Subpart F NSPS. While Ash Grove appreciates the effort to maintain 
consistency between the requirements, Ash Grove believes that Equation 
3 in the Subpart F NSPS is in error and will be corrected in the 
upcoming public notice addressing Subpart F. Ash Grove provided a 
suggested formula to replace the formula stated in the proposed 
regulatory text.
    Response: We agree for the reasons stated by Ash Grove that the 
formula for calculating the emissions should express SO2 and 
NOX concentrations in ppm, not in gr/scf. We have corrected 
40 CFR 52.1396(e)(4)(ii) accordingly; however, rather than use the 
language proposed by Ash Grove, we have used the formula and associated 
language found in the proposed amendments to the Portland cement NSPS. 
77 FR 42397.
    Comment: Ash Grove noted that the proposed regulatory text at 40 
CFR 52.1396(f) would require that Ash Grove perform EPA Method 5, 5B, 
5D or 17, 40 CFR Part 60, Appendix A, to demonstrate compliance with 
the PM limit and that the test consist of three runs with each run at 
least 120 minutes long and each run collecting a minimum sample of 60 
dry standard cubic feet. Ash Grove supports the approach of identifying 
the specific source test methods in the rule. However, Ash Grove does 
not support specifying the duration of each test run and the minimum 
sample size. Ash grove stated that this BART FIP is being implemented 
with the intent that it will control emissions for many years to come. 
Ash Grove stated that placing this type of detailed data into the rule, 
rather than letting the test duration and sample size be determined 
based on the test method as it exists at the time of the test, invites 
future confusion and trouble. Therefore, Ash Grove suggested that EPA 
specify the test methods but delete the other language relating to the 
test duration and sample size.
    Response: We disagree. The test method does not determine the test 
duration and sample size, but cross-references other rules in this 
regard. EPA Method 5 states in subsection 8.2.4, ``Select a total 
sampling time greater than or equal to the minimum total sampling time 
specified in the test

[[Page 57877]]

procedures for the specific industry, such that (1) the sampling time 
per point is not less than 2 minutes (or some greater time interval as 
specified by the Administrator), and (2) the sample volume taken 
(corrected to standard conditions) will exceed the required minimum 
total gas sample volume.'' Methods 5B and 5D cross-reference Method 5 
for sampling time and sampling volume. Method 17 does not cross-
reference Method 5 for sampling time and sampling volume, but does not 
specify anything different. We consider three test runs, with each run 
at least 120 minutes long, and each run collecting a minimum sample of 
60 dry standard cubic feet, to be appropriate and necessary for 
purposes of the Montana Regional Haze FIP. We note that this has been 
specified in PM stack testing requirements in other regional haze FIPs. 
(See, for example, Proposed Final FIP by EPA Region 9 for the Four 
Corners Power Plant, 76 FR 52387, August 22, 2011.) This comment has 
not resulted in any change to our proposal.
    Comment: Ash Grove stated that the proposed regulatory text at 40 
CFR 52.1396(h)(6) would require that they maintain, among other things, 
records required by Part 75. Ash Grove is not subject to part 75 as 
that applies only to electrical generating units. Ash Grove believes 
that this reference to Part 75 was just a ``catch-all'' and not 
intended to impose any obligations under Part 75 upon cement kilns 
otherwise not subject to Part 75. However, due to the potential for 
misunderstanding and the lack of relevance of the Acid Rain provisions 
to cement kilns, Ash Grove requested that the reference to Part 75 be 
deleted.
    Response: We agree. Since the proposed monitoring requirements for 
cement kilns, at sections 52.1396(e)(3) and (4), and at section 
52.1396(f)(2), do not cross-reference Part 75, there are no applicable 
Part 75 recordkeeping requirements under our FIP proposal. Therefore, 
the reference to Part 75 on recordkeeping, at 40 CFR 52.1396(h)(6), is 
not necessary and has been removed.
    Comment: Ash Grove stated that the proposed regulatory text at 40 
CFR 52.1396(i) would require that Ash Grove submit quarterly excess 
emission reports and CEMS performance reports. Ash Grove currently is 
subject to similar reporting requirements under the Title V and NESHAP 
programs. However, in both of those programs the reports are submitted 
semi-annually, not quarterly. Ash Grove sees no purpose gained by 
submitting the reports quarterly and the additional administrative 
burden is significant. Therefore, Ash Grove requested that EPA revise 
this reporting requirement to make it consistent with the similar 
reports submitted under Title V and NESHAP programs, i.e., semiannual 
reports.
    Response: We agree. We used provisions in NSPS Subparts A and F 
applicable to cement kilns as a model for the CEMS-related reporting 
requirements for cement kilns in our FIP proposal. The general 
provisions of NSPS Subpart A, at 40 CFR 60.7(c), require semiannual 
excess emission reports and monitoring systems performance reports, 
except when more frequent reporting is specifically required by an 
applicable subpart, or if the Administrator, on a case-by-case basis, 
determines that more frequent reporting is necessary to accurately 
assess the compliance status of the source. NSPS Subpart F for cement 
kilns does not specify more frequent reporting. Therefore, we have 
revised the required reporting frequency to semiannual in 40 CFR 
52.1396(i)(1) and (i)(2) for cement kilns. The required reporting 
frequency for EGUs remains quarterly.
    Comment: Ash Grove requested that EPA revise its proposed 
regulatory text at 40 CFR 52.1396(i)(2)(ii) requiring the company to 
submit Relative Accuracy Audits (RAAs) and Cylinder Gas Audits (CGAs). 
Ash Grove does not object to the idea of submitting Relative Accuracy 
Test Audits (RATAs) as those are documented in a highly formalized test 
report prepared by a third party testing contractor. However, the RAAs 
and CGAs are not documented in the same type of formal third party 
report. Ash Grove believes that it is adequate to certify that the 
audits have been performed as part of the semiannual reports.
    Response: We disagree. Our proposed regulatory text at 40 CFR 
52.1396(e)(3) states that the CEMS shall be used to determine 
compliance with the emission limitations in section 52.1396(c), for 
each unit, in combination with data on actual clinker production. For 
cement kilns, 40 CFR section 52.1396(i)(2)(ii) requires submittal of 
results of any CEMS performance tests required by 40 CFR part 60, 
appendix F, Procedure 1, which is titled ``Quality Assurance 
Requirements for Gas Continuous Emission Monitoring Systems Used for 
Compliance Determination.'' Under Section 7 of Procedure 1 (Reporting 
Requirements), it is not adequate to merely certify that the RAAs and 
CGAs have been performed. Section 7 requires submittal of a Data 
Assessment Report for each quarterly audit, which must include 
``Assessment of CEMS data accuracy and date of assessment, as 
determined by a RATA, RAA or CGA described in Section 5, including * * 
*, the A [accuracy] for the RAA or CGA, the RM [reference method] 
results, the cylinder gases certified values, the CEMS responses, and 
the calculations results as defined in Section 6.'' This information 
must be included in the semiannual reports referenced in our response 
to the previous comment above. We consider this information appropriate 
and necessary. This comment has not resulted in any change to our FIP 
proposal.
    Comment: Ash Grove requested that EPA drop the requirement proposed 
in 40 CFR 52.1396(k)(2) to provide semiannual progress reports on 
construction of SO2 and NOX control equipment. 
Ash Grove does not object to filing notification of commencement of 
construction as this obligation is consistent with what Ash Grove is 
used to under the NSPS and state new source review program. However, 
semiannual construction progress reports are not something that Ash 
Grove is typically set up to generate and there seems to be little 
gained from such reports. Therefore, Ash Grove requested that this 
requirement be dropped from the rule.
    Response: We disagree. We consider construction progress reports 
necessary as part of ensuring that BART sources meet their five-year 
compliance deadlines. Since installation of substantial equipment may 
be involved, there could be unforeseen construction delays that we 
would want to be aware of well before the five-year deadline. We do not 
consider this reporting a burdensome requirement, as our FIP proposal 
does not specify any particular level of detail for these progress 
reports. This comment has not resulted in any change to our FIP 
proposal.
    Comment: Ash Grove noted that the BART limits are identified as 
applying at all times, including startup, shutdown and malfunction. 
Although the preamble states that the proposed limits allow ``for a 
sufficient margin of compliance,'' Ash Grove argued that these limits 
do not take into account the impact of sudden and unforeseen effects 
attributable to malfunctions. As compliance with all three limits 
(i.e., SO2, PM and NOX) could be affected by a 
malfunction, Ash Grove believes that it is appropriate for EPA to 
provide the same affirmative defense in the event of a malfunction as 
is provided in the Portland cement MACT rules. Specifically, Ash Grove 
requested that EPA incorporate the same affirmative defense provided in 
40 C.F.R. 63.1344 to address malfunctions.
    Response: EPA disagrees with this comment. As stated in our 
proposal, to determine the BART NOX limit for Ash

[[Page 57878]]

Grove, we first applied the efficiency of the selected controls, LNB + 
SNCR at 58%, to the 99th percentile 30-day rolling average 
NOX emission rate at this facility for May 26, 2006 through 
September 8, 2008, resulting in a figure of 7.82 lb/ton clinker. 77 FR 
at 24007 n.45. We then set the BART limit above this, at 8.0 lb/ton 
clinker. Ash Grove provides no data to show that, at this facility, 
this limit cannot be achieved due to malfunctions, or that our use of 
the 99th percentile 30-day rolling-average NOX emission rate 
in combination with the additional margin (from 7.82 to 8.0 lb/ton 
clinker) provides an insufficient margin of compliance.
    For SO2, we did not select any additional controls for 
BART. We based the BART SO2 limit on the 99th percentile 30-
day rolling average SO2 emission rate at this facility for 
May 26, 2006 through September 8, 2008, 11.02 lb/ton clinker, and set 
the BART limit at 11.5 lb/ton clinker. 77 FR at 24013 n.73. Ash Grove 
provides no data to show that, at this facility, this limit cannot be 
achieved due to malfunctions, or that our use of the 99th percentile 
30-day rolling average SO2 emission rate at this facility in 
combination with the additional margin (from 11.02 to 11.5 lb/ton 
clinker) provides an insufficient margin of compliance.
    We also did not select any additional controls for PM. Ash Grove 
currently has an electrostatic precipitator for PM control and is 
subject to a process weight-based PM10 emission rate set out 
in Montana's approved SIP and Ash Grove's title V operation permit. We 
set the BART limit, based on use of the current control technology, at 
the existing emission rate. Ash Grove has not provided any data to show 
that it is not able to meet the existing limit due to malfunctions. As 
a result, we continue to maintain that the NOX, 
SO2, and PM BART limits for Ash Grove provide for a 
sufficient margin of compliance, including taking into account 
malfunctions.
    With respect to the Portland cement MACT standard, we note that the 
MACT standard applies across the entire source category, while the BART 
limits imposed in this FIP reflect application of the five statutory 
BART factors to a particular facility, Ash Grove. Ash Grove does not 
explain why, in this circumstance, the existence of the affirmative 
defense in the MACT standard necessarily implies an affirmative defense 
is required for the BART limits, which as discussed above, for 
NOX and SO2 are based in part on actual emissions 
from Ash Grove, and for PM are based on an existing limit for Ash 
Grove. We therefore disagree that the affirmative defense provided for 
in 40 CFR section 63.1344 should be also provided for in this FIP.
    Comment: The opening sentence of the proposed regulatory text at 40 
CFR 52.1396(i) states ``All reports under this section, with the 
exception of 40 CFR 53.1395(n) and (o) shall be submitted * * *'' Ash 
Grove believes that this cross-reference is in error, as Ash Grove is 
not aware of there being a 40 CFR 53.1395(n) or (o). Ash Grove believes 
this was intended to cite to 40 CFR 52.1396(n) and (o).
    Response: We agree this was an error. We have corrected the 
language to cite to section 52.1396(n) and (o), instead of section 
53.1395(n) and (o).

E. Comments on Holcim

    Comment: Montanans Against Toxic Burning (MATB) applauded our 
proposed retrofit of the Holcim kiln to include LNB and SNCR.
    Response: We acknowledge MATB's support.
    Comment: MATB believes that we should reanalyze the fuel-switching 
option for the Holcim cement kiln. Specifically, they stated that 
petroleum coke inputs should be reduced, which they believe would lead 
to significant reductions in SO2 emissions. They also stated 
that our analysis may be skewed by what MATB describes as Holcim's 
``low-ball'' estimates of its sulfur emissions. MATB believes that a 
review of Holcim's past monitoring data could lead to a different 
conclusion.
    Response: We disagree that it is necessary to reanalyze fuel 
switching options for Holcim. In our analysis, we used annual 
SO2 emissions as reported to the National Emissions 
Inventory and we have no reason to believe that these were 
underestimated. The annual emissions (50.2 tpy) are so minimal that 
fuel switching options resulting in increased annual cost would not be 
considered cost-effective on a dollar per ton basis. In addition, the 
visibility improvement from fuel switching is very low at 0.015 
deciview for fuel switching option 1 and 0.024 deciview for fuel 
switching option 2.
    Comment: MATB commented that a ``real-time hourly'' standard for 
NOX and SO2, rather than the 30-day rolling 
averages based on clinker production proposed, is needed to assure 
compliance with protective limits. MATB explained that with the 30-day 
rolling averages, spikes due to malfunction or improper operation will 
``disappear'' in the averaging process.
    Response: We assume that by ``real-time hourly'' standard, the 
commenter means an emission limit in pounds per hour. We disagree that 
we should establish an hourly standard rather than a 30-day rolling 
average limit based on clinker production. As we explained in our 
proposal (77 FR 24007), we chose an output-based standard because it 
avoids rewarding a source for becoming less efficient, i.e., requiring 
more feed to produce a unit of product. An output-based standard 
promotes the most efficient production process. With regard to 30-day 
versus hourly averaging time, EPA's BART guideline calls for BART 
emission limits to be expressed as 30-day rolling averages for 
electrical generating units. 70 FR 39172. We believe this is 
appropriate for other BART units as well. The proposed limit allows for 
a sufficient margin of compliance for a 30-day rolling average limit 
that would apply at all times, including startup, shutdown, and 
malfunction. 77 FR 24018.
    Comment: MATB believes that more oversight, transparency, and 
accountability are needed when it comes to reporting and record 
keeping.
    Response: We are confident that the information used to make our 
decision is accurate. With regard to reporting and recordkeeping 
requirements under the FIP, the commenter has not explained what 
oversight, transparency and accountability is lacking and what more is 
needed in this regard. That said, section 114 of the CAA allows EPA and 
the State to ask for monitoring data and reports as necessary. These 
documents are available to the public unless the information is claimed 
to be confidential business information.
    Comment: MATB commented that the efficiency of Holcim's ESP is 
incorrect as stated in EPA's analysis, and does not operate during most 
malfunctions. These malfunctions can last 24 hours or more. 
Additionally, MATB stated that EPA's analysis fails to consider PM 
during periods of startup, shutdown and malfunction and considering the 
frequent upsets with the Trident kiln, that cause its ESP to be turned 
off, an additional control measure at Holcim is essential. MATB 
encouraged us to analyze the addition of a fabric filter.
    Response: We disagree that it is necessary to evaluate the 
installation of a fabric filter at Holcim. In our proposal, we 
explained that PM emissions from Holcim did not significantly 
contribute to visibility impairment. We used actual emission rates to 
model the visibility impact from Holcim. Because the baseline 
visibility impact from PM was low, improvements to the existing PM 
control device would not be significant.
    Comment: The commenter stated that an annual three-hour stack test 
is

[[Page 57879]]

inadequate to monitor PM emission limit compliance.
    Response: We disagree. The proposed requirements for demonstrating 
compliance with PM emission limits include more than just an annual 
three-hour stack test. ``In addition to annual stack tests, owner/
operator shall monitor particulate emissions for compliance with the 
BART emission limits in accordance with the applicable Compliance 
Assurance Monitoring (CAM) plan developed and approved in accordance 
with 40 CFR part 64.'' 77 FR 24099. The requirements include the 
following:
     40 CFR 64.3(a) requires that a monitoring parameter be 
selected by the owner/operator as an indicator of emission control 
performance for the control device.
     40 CFR 64.3(b) requires that an indicator range for that 
parameter be selected ``such that operation within the range provides a 
reasonable assurance of ongoing compliance with emission limitations or 
standards for the anticipated range of operating conditions.''
     40 CFR 64.7(d) requires the owner/operator, upon detecting 
an excursion or exceedance of the CAM indicator range, to restore 
operation of the emitting unit and emission control device to its 
normal or usual manner of operation as expeditiously as practicable, in 
accordance with good air pollution control practices for minimizing 
emissions.
     40 CFR 64.8 says the Administrator or permitting authority 
may require the owner/operator, in the event of repeated excursions or 
exceedances of the CAM indicator range, to develop and implement a 
Quality Improvement Plan, to correct any control device performance 
problems.
    Further, 40 CFR 52.11396(l) states, ``At all times, owner/operator 
shall maintain each unit, including associated air pollution control 
equipment, in a manner consistent with good air pollution control 
practices for minimizing emissions'' This applies to all sources in the 
FIP.
    Comment: MATB explained that there are inconsistencies in EPA's 
proposed NOX and SO2 emissions limits, and there 
appears to be a mistake on Table 53 dealing with fuel-switching 
options.
    Response: These inconsistencies were corrected in the FR notice 
dated May 17, 2012. 77 FR 29270.
    Comment: Holcim commented that that the output-based standards we 
proposed reward a source for operating inefficiently. Holcim indicated 
that our proposed FIP is unfairly stringent with respect to Holcim as 
compared to Ash Grove. They stated that the kiln types and capacities 
of the two plants are substantially equal, but that Holcim's emissions 
profiles are notably different. Holcim stated that they use proper kiln 
design and best combustion practices to control NOX 
emissions at their plant, and that Ash Grove has NOX 
emissions that are 42% higher than NOX emissions from the 
Holcim plant. Holcim further stated that our proposed FIP rewards Ash 
Grove with a NOX BART emission limit that is 60% higher than 
Holcim's proposed NOX BART emission limit. Holcim pointed 
out that their kiln has substantially lower current NOX 
emission rates, lower current visibility impacts, and a lower 
subsequent visibility improvement, yet our FIP requires substantially 
tighter emission limits for NOX and SO2.
    Holcim commented that, based on EPA's analysis, the proposed 
NOX limit would require Holcim to invest a total of $5.6 
million in SNCR and indirect firing, which would result in an 
improvement in visibility at Gates of the Mountains WA that is 
significantly less than the 1.0 deciview perceptibility threshold and 
that our proposed FIP would require only a $1.19 million capital 
investment from Ash Grove, even though Ash Grove's impact on Gates of 
the Mountains WA is more than double the impact from Holcim. Holcim 
also stated that we estimated that Ash Grove's NOX emissions 
caused degradation in visibility of greater than 0.5 deciview at Gates 
of the Mountains WA on approximately 33% of the days in the baseline 
period while Holcim impacted Gates of the Mountains WA at greater than 
0.5 deciview only on approximately 4% of the days during the baseline 
period. Holcim stated that EPA's approach would reward Ash Grove's 
higher emissions and inefficient operation by creating an economic 
disadvantage for Holcim in a highly competitive market.
    Response: We disagree. Our explanation in the proposed FIP 
regarding the output-based standard was provided to explain the 
difference between a standard expressed in quantity of pollutant per 
amount of feed and quantity of pollutant per amount of product 
produced. As explained in our proposal, an output-based standard avoids 
rewarding a source for becoming less efficient, i.e., requiring more 
feed to produce a unit of product. 77 FR 24007. Our explanation did not 
imply that both sources should have exactly the same emission rate. The 
NOX standards for both Holcim and Ash Grove were determined 
by applying the control efficiency of the selected technologies to the 
current emission rates at each facility. This is the most appropriate 
method to determine emission limits for these two sources. As explained 
in other responses, we are not requiring Holcim to convert to indirect 
firing in the final FIP, so the information comparing capital 
investment is no longer relevant. In the final FIP, we have determined 
the emission rate for Ash Grove by applying the control effectiveness 
of LNB + SNCR (58%) to the current emission rate and as explained in 
other responses we have revised the emission rate for Holcim by 
applying the control effectiveness of SNCR (50%) to the current 
emission rate. In both cases, we have determined the emission rate 
based on the control effectiveness of the control technology that was 
chosen based on the five statutory BART factors listed at CAA section 
169A(g)(2) and 40 CFR 51.308(e)(1)(ii)(A). The five statutory factors 
include the costs of compliance and visibility improvement; therefore, 
these factors were evaluated and considered in the selection of 
controls. Applying the control effectiveness of the technology that was 
identified based on the five statutory factors to the current emission 
rates for each source is a logical method for determining emission 
rates. This same methodology was used for determining the emission 
rates for both sources.
    We note that in the final FIP, Ash Grove will reduce an estimated 
1,088 tons per year of NOX using LNB+SNCR at a total annual 
cost of $2,238,893, but Holcim will only reduce an estimated 556 tons 
per year of NOX at a total annual cost of $650,399. Ash 
Grove will be reducing 946 tons per year of NOX through the 
operation of SNCR, but Holcim will only be reducing 556 tons per year 
through the operation of SNCR.\12\ We provide this information to 
demonstrate that overall, more emissions will be reduced by Ash Grove 
and to also illuminate the fact that annual cost will be greater for 
Ash Grove. The cost of reagent is proportional to the amount of 
pollutant removed; therefore, annual reagent cost will be significantly 
greater for Ash Grove.
---------------------------------------------------------------------------

    \12\ See Table 11, FR 77 24004, and Table 22, 77 FR 24007 for 
Ash Grove. Holcim's baseline NOX emissions are 1,112 tpy. 
Revised emissions reduction for SNCR only for Holcim is 556 tpy and 
cost is $1,170/ton.
---------------------------------------------------------------------------

    We are not requiring additional controls for SO2 for 
either Holcim or Ash Grove and the SO2 limits for each 
facility were determined based on current emission rates. This 
determination was based on an evaluation of the five statutory factors 
and the SO2 emission rates were determined in the same 
manner for both

[[Page 57880]]

facilities. There is no necessity for additional SO2 control 
at either facility; the current controls were considered to be BART.
    As for Holcim's comment that the proposed FIP rewards Ash Grove's 
higher emissions and inefficient operation by creating an economic 
disadvantage for Holcim in a highly competitive market, the BART 
Guidelines do allow for the consideration of unusual circumstances that 
justify taking into consideration the conditions of the plant and the 
economic effects of requiring the use of a given control technology. 
The BART Guidelines state:

    [t]hese effects would include effects on product prices, the 
market share, and profitability of the source. Where there are such 
unusual circumstances that are judged to affect plant operations, 
you may take into consideration the conditions of the plant and the 
economic effects of requiring the use of a control technology. Where 
these effects are judged to have a severe impact on plant operations 
you may consider them in the selection process, but you may wish to 
provide an economic analysis that demonstrates, in sufficient detail 
for public review, the specific economic effects, parameters, and 
reasoning.

70 FR 39171. Holcim did not provide information for us to consider in 
such an analysis.
    The BART Guidelines also state, ``[a]ny analysis may also consider 
whether other competing plants in the same industry have been required 
to install BART controls if this information is available.'' 70 FR 
39171. In this case, Ash Grove is required to install BART controls. We 
have considered each plant individually, and based on the BART analyses 
both Holcim and Ash Grove plants are required to install BART controls.
    Comment: Holcim argued that the Texas kilns cited by EPA in the FIP 
are not representative and two of the three kilns did not achieve 50% 
NOX reduction. Holcim cited several site-specific factors 
that impact SNCR performance that they state EPA did not adequately 
consider, including turbulent mixing, heat transfer, spray droplet 
size, spray drop evaporation, devolatilization and others. Holcim also 
argued that the carbon monoxide (CO) levels at the Trident kiln are 
much lower than the CO levels at the Texas kilns, which will adversely 
impact NOX reductions and ammonia slip at the Trident kiln 
relative to the Texas kilns. Holcim additionally argued that EPA did 
not adequately consider NOX emissions variability in setting 
the limit because of the limited time frame considered for the data 
from the Texas kilns.
    Response: We disagree. EPA has assumed that 50% reduction is 
possible with SNCR; however, this does not rule out the possibility 
that Holcim may determine that other means, such as mid-kiln firing, 
may be better than SNCR alone in terms of cost or other factors for 
achieving 50% NOX reduction. In any event, 50% 
NOX reduction is achievable with SNCR and this is supported 
by the data cited in the proposed FIP. We address this in more detail 
in a response to Ash Grove.
    Holcim also noted that SNCR performance depends upon a wide range 
of site-specific factors. They list rate-limiting processes, including 
turbulent mixing, heat transfer, spray droplet size, spray drop 
evaporation, devolatilization and others. As detailed in a contractor's 
report in the docket, we have considered these factors and none of them 
causes us to change our decision. In brief, spray droplet size is a 
factor the SNCR system designer can control and tailor to the needs of 
the system. Turbulent mixing may or may not be within the SNCR system 
designer's ability to control, but in any case our selection of SNCR 
does not depend on optimal turbulent mixing.
    With respect to CO concentration, if the CO at the Trident kiln is 
much lower than at the Texas kilns referred to in the comments, as 
Holcim describes, this simply means that the SNCR reagent should be 
introduced at a point in the process where the gas temperature is 
higher than the injection point used at the Texas kilns where the CO 
levels are higher. This may in fact improve SNCR performance.
    With regard to NOX emission variability raised by 
Holcim, first, the data used by EPA in Table 10 of the proposed FIP 
cover a three month period which should be adequate time to address 
normal operating changes that would impact NOX. Second, SNCR 
can be used to mitigate variability in NOX emissions. This 
is confirmed by the data on the Midlothian kilns that is in the 
proposed FIP and as described in response to comments from Ash Grove. 
For every kiln, the standard deviation in the monthly NOX 
emission rate was lower after the application of SNCR than before, 
indicating a lower variation in NOX emissions.
    Comment: Holcim argued that a detached plume may result from 
operation of the SNCR in the winter months, which will make it 
necessary to not operate the SNCR system or to allow a condition where 
visibility is adversely impacted to continue. The detached plume could 
be the result of the formation of ammonium salt reactions with sulfate 
or chlorides.
    Response: We disagree. As discussed by Miller,\13\ there are 
several factors that could contribute to a visible detached plume, and 
these include moisture, temperature, and availability of the 
constituents that contribute to the plume--ammonia, sulfates and 
chlorides. Ammonia slip from the SNCR process can be well controlled in 
a cement kiln, and the SNCR system would not affect the amount of 
ammonia contributed by raw materials.
---------------------------------------------------------------------------

    \13\ Miller, F. M., ``Management of Detached Plumes in Cement 
Plants'' 2001 IEEE-IAWPCA Cement Industry Technical Conference 
Vancouver, British Colombia, Canada April 2001.
---------------------------------------------------------------------------

    Sulfates and chlorides are largely the result of impurities in the 
raw materials, and ammonia can be contributed by raw materials. 
Holcim's SO2 emissions are low indicating low levels of 
sulfates in the exhaust. Therefore, the risk of an ammonium sulfate 
plume, even with ammonia present, is small. The presence of chlorides 
will depend upon the raw materials and whether the chlorides become 
bound to alkaline material before being emitted up the stack.
    Chlorides, if present, will typically preferentially be bound to 
alkaline material that is present rather than be emitted. Holcim did 
not provide any information on stack chloride emission levels at this 
site to support their concerns about detached plume from ammonium 
chloride.
    Because of the importance of impurities in the raw materials in 
contributing to the chemical constituents that form a plume, the 
experience at one kiln cannot be directly applied to another without 
more information. Therefore, while there may be a risk of a visible 
plume at the Trident kiln, Holcim has not provided enough data to 
indicate that addition of an SNCR system would increase this risk 
significantly. Furthermore, a localized plume would not necessarily 
impact a Class I area and Holcim has not provided any information 
indicating such an impact.
    Comment: Holcim indicated that EPA failed to consider the 
NOX control technology already installed at the Trident 
plant. Holcim explains that they changed the burner at Trident in May 
2009 to a multichannel LNB design as part of the company's burner 
system modification for NOX control, as detailed in Holcim's 
2007 BART analysis.
    Holcim stated that EPA's BART analysis ignored the installation of 
the multichannel LNB at the Trident plant, in contravention of EPA's 
obligation to consider ``any existing pollution control

[[Page 57881]]

technology in use at the source'' as part of the five-factor BART 
analysis. 42 U.S.C. 7491(g)(2). Holcim's BART analysis was prepared and 
submitted in 2007, before the multichannel LNB technology was 
installed.
    Holcim explains that they originally installed a multichannel 
burner in April 2008 but it caused operational issues and was removed 
in July 2008. The multichannel burner was redesigned, installed in May 
2009, and has operated continuously since that time. According to 
Holcim, the multichannel design allows the fuels to be separated into 
different channels and enables Holcim to more precisely control the 
amount of air passing through each of the channels. Consequently, 
Holcim says, they can better control the flame characteristics in the 
kiln, which results in higher thermal efficiency of the kiln and 
improved product quality.
    Holcim stated that they also anticipated that the multichannel 
design would reduce NOX and SO2 emissions. Holcim 
acknowledges that the effects of the technology are difficult to 
quantify. Based on a comparison of NOX emissions pre- and 
post-installation of the LNB technology where the fuel mix was 
generally the same, Holcim says the plant's NOX emissions 
decreased by approximately 13% with the installation of the 
multichannel LNB. In addition to the multichannel LNB, Holcim stated 
that they installed an indirect firing system for the petroleum coke 
system.
    Holcim notes that EPA used a baseline for the Trident plant of 
years 2008 through 2011, a period of time that already includes the 
effects of the LNB technology at the plant. Holcim stated that EPA 
assumed in its BART proposal for the Trident plant that the combination 
of LNB and indirect firing would achieve a NOX reduction of 
15%. However, Holcim stated that a 13% reduction in NOX 
emissions has already been achieved through prior installation of the 
multichannel LNB. Holcim states there is no basis to assume that 
indirect firing would improve NOX emissions reductions at 
Trident and that additional NOX reductions can only be 
obtained through installation of SNCR. As a result, Holcim concludes 
that EPA's analysis of the cost-effectiveness and visibility impact for 
the installation of indirect firing is, ``clearly erroneous and should 
be disregarded''.
    Response: We agree with aspects of this comment, but disagree with 
others. As described in more detail below, Holcim has not provided 
enough information to demonstrate that the installed multi-channel 
burner that Holcim installed is in fact a low NOX burner. In 
any case, the baseline used for the BART analysis included emissions 
averaged over a four year period (2008-2011), which would have included 
the time that the multi-channel burner was installed. We have decided 
that the incremental cost of indirect firing and a low NOX 
burner is not justified and have revised the BART emission limit 
accordingly.
    We agree that our BART proposal, did not consider installation of 
the new burners that Holcim describes as ``multichannel LNB'' in its 
March 20, 2008 letter to Vickie Walsh of the MDEQ. As the June 9, 2009 
letter from Holcim to EPA notes, ``a low NOX burner 
modification would require low primary air and, thus, a conversion of 
Trident's firing system from a direct to an indirect system.'' Based on 
the information we have, it appears that the Trident kiln has not 
installed an indirect firing system for coal and therefore the 
multichannel burner does not meet the definition of LNB in Holcim's 
letter. The burner is not capable of operating at low primary air 
levels on pulverized coal and cannot achieve the NOX 
reductions that an indirect firing system would achieve.
    However, we disagree that we must credit the newly installed burner 
with a 13% reduction in NOX emissions, because we are 
lacking validation data that such a reduction has been achieved. Holcim 
has only presented summary information to support the claim of 13% 
reduction and has not provided the underlying data to validate its 
claim. Our examination of NOX emissions data provided by 
Holcim on March 2, 2012, covering the period from 2008 through 2011 
(referenced in our proposal at 77 FR 24018, footnote 93), does not 
reveal any evidence of sustained NOX emission reduction 
after May of 2009. We have used data from the time period 2009-2011, 
after the new burner was installed, in calculating baseline emissions. 
77 FR 24014, Table 39, footnote 1. This baseline accurately reflects 
current conditions and is appropriate for comparison to available 
control scenarios.
    Nevertheless, since a switch to indirect firing to accommodate 
installation of LNB, as described in our FIP proposal, would have an 
unreasonably high incremental cost-effectiveness of $8,029/ton, with 
minimal visibility benefits (see our response below), we are not 
requiring a switch to indirect firing and LNB as BART in the final FIP. 
We also are clarifying that we intended this option to include 
switching to indirect firing and a LNB. We have recalculated the 
proposed BART limit for NOX to reflect a 50% reduction in 
NOX emissions from that baseline by addition of SNCR alone, 
rather than the 58% reduction we previously used, which reflected 
switching to indirect firing and adding a LNB plus SNCR.
    In recalculating our proposed BART emission limit for 
NOX, we continue to rely on the estimate of baseline 
NOX emissions in lb/ton clinker provided in Holcim's 2012 
submittal, cited in our proposal at 77 FR 24018, footnote 93. That 
submittal listed a 99th percentile 30-day rolling average 
NOX emission rate of 12.6 lb/ton clinker, for the period 
2008-2011. Applying a 50% reduction to the 99th percentile figure 
yields 6.3 lb/ton clinker. To allow for a sufficient margin of 
compliance for a 30-day rolling average limit that would apply at all 
times, including startup, shutdown and malfunction (as explained in our 
proposal at 77 FR 24018), we are setting the BART limit at 6.5 lb/ton 
clinker in our final rule.
    Since the estimated baseline NOX emissions have not 
changed from our proposal, and since our estimate of 50% NOX 
reduction for SNCR alone has not changed from our proposal, our 
estimate of 556 tons per year of expected NOX reduction for 
SNCR alone has also not changed from our proposal.
    Comment: Holcim stated that EPA underestimated the costs of 
installing and maintaining a SNCR system. Holcim stated that the 
company calculated the direct annual costs of SNCR to be $443,341 and 
the indirect annual costs for SNCR to be $227,538, and that these 
calculations employed a 15-year amortization schedule, as requested by 
EPA in 2007.\14\ Holcim noted that EPA's estimated direct annual costs 
and indirect annual costs for SNCR are lower than Holcim's estimates by 
approximately 67% and 46%, respectively and suggested that the 
difference may be at least in part due to EPA's use of a 20-year period 
in the proposal.
---------------------------------------------------------------------------

    \14\ August 2009 Submittal (EPA-R08-OAR-2011-0851-0038); Letter 
from Callie A. Videtich to Ned Pettit (Nov. 26, 2007) (EPA-R08-OAR-
2011-0851-0038).
---------------------------------------------------------------------------

    Holcim stated that it is unclear how EPA derived its numbers and 
that EPA provided no explanation in the FIP proposal. Holcim requested 
clarification of EPA's method for calculating these costs and urged EPA 
to instead use the cost calculation numbers provided by Holcim.
    Also, Holcim stated that if EPA reviews selective catalytic 
reduction (SCR) for cement kilns in subsequent reasonable progress 
planning periods, and determines that Holcim must install SCR instead 
of SNCR at that time then

[[Page 57882]]

the 20-year amortization for SNCR costs would not accurately reflect 
the annual costs of installing SNCR. Holcim also stated that since the 
company conducted its original analysis, Holcim has installed SNCR at 
its plant in Hagerstown, Maryland in 2011, which also has a long kiln. 
Holcim stated that the total capital costs for the SNCR installation at 
Hagerstown were approximately $1,920,000, including the cost of 
commissioning and spare parts and that, in addition, Hagerstown 
budgeted $591,000 for 2012 operating costs ($1.35 per metric ton of 
clinker or $1.23 per metric ton of cement). Holcim stated that actual 
operating costs for 2012 through the end of April have been $179,000 
($1.40 per metric ton of clinker or $1.28 per metric ton of cement). 
Holcim anticipates that similar capital and operating costs would apply 
to the installation of SNCR at Trident. Holcim requested that EPA use 
these updated figures in its analysis of the costs of SNCR at Trident.
    Response: We agree with aspects of this comment, but disagree with 
others. We note that the letter to which Holcim refers requested that 
Holcim reanalyze annualized costs using a 15-year amortization period 
for a scrubber, not SNCR. We agree that EPA underestimated the cost of 
SNCR and that clarification on cost is needed, but we disagree with the 
statement that EPA provided no explanation in its proposal on how EPA 
derived its cost numbers. We also disagree with the statement that EPA 
provided no explanation for use of a 20-year amortization period. We 
also disagree with the statement that SNCR costs at the Trident kiln 
should be similar to Holcim's Hagerstown kiln.
    We agree that we underestimated the cost of SNCR and that 
clarification is needed. The underestimate arose from our omission of 
cost of reagent. In Holcim's August 12, 2009 submittal, two versions of 
a SNCR cost spreadsheet were included. In one version, Holcim redacted 
the line item for reagent cost, on the basis of a Confidential Business 
Information (CBI) claim. This was the version we used for our proposal. 
However, in its cover letter for the August 12, 2009 submittal, Holcim 
stated that it later retracted its CBI claim. So the submittal included 
a second version of the same SNCR cost spreadsheet, in which the 
reagent line item now appears. The reagent cost is listed by Holcim in 
this second version at $379,183.
    We have recalculated the annual costs of SNCR to include the cost 
of reagent. Relying on the second version of the cost spreadsheet in 
Holcim's August 12, 2009 submittal, we now calculate the annual costs 
other than capital recovery at $526,471 and the total annual cost, 
including capital recovery, at $650,399. Using an estimated emission 
reduction of 556 tons per year of NOX, as we did in our 
proposal (which is a 50% reduction from the NOX emissions 
baseline of 1,112 tons per year), we have recalculated the cost-
effectiveness of SNCR at $1,170/ton. At this cost-effectiveness, we 
still consider SNCR to be BART for NOX. Holcim has given us 
no reason to think otherwise.
    We disagree with the statement that EPA provided no explanation in 
its proposal on how EPA derived its cost numbers. We explained that we 
relied on cost estimates supplied by Holcim for capital costs and 
annual costs of SNCR, with the exception of the Capital Recovery Factor 
(CRF) used. 77 FR 24015. We included a footnote to Table 44 to explain 
that we relied on Holcim's capital cost estimate for SNCR. We included 
a second footnote to that table to explain what CRF we used. We also 
included a footnote to Table 45 to explain that we relied on Holcim's 
estimate of direct annual operating costs. 77 FR 24016.
    We disagree with the statement that EPA provided no explanation for 
use of a 20-year amortization period. As explained at 77 FR 24015, we 
relied on Holcim's estimates of SNCR capital cost and annual costs, 
with the exception of the capital recovery factor (CRF). We acknowledge 
that we wrote to Holcim in 2007 to recommend 15-year amortization, and 
that our decision since then to use 20-year amortization instead needs 
clarification. We now clarify that after reviewing EPA national 
guidance on CRFs in more detail since 2007, we determined that it would 
be more appropriate to use a CRF consistent with 20 years for the 
useful life of the kiln and associated SNCR controls. As explained 
below, our use of a 20-year period was not arbitrary.
    The guidance we relied on was EPA's Air Pollution Control Cost 
Manual (CCM), which says, in regard to SNCR, that ``In general, 
indirect annual costs (fixed costs) include the capital recovery cost, 
property taxes, insurance, administrative charges, and overhead. 
Capital recovery cost is based on the anticipated equipment lifetime 
and the annual interest rate employed. An economic lifetime of 20 years 
is assumed for the SNCR system.'' EPA Air Pollution Control Cost 
Manual, Sixth Edition, EPA/452/B-02-001, January 2002, Section 4.2, 
Chapter 1, page 1-37. We explained in our FIP proposal that without 
commitments for an early shutdown, EPA cannot consider a shorter 
amortization period. 77 FR 24014. For consistency in comparing control 
options for NOX and SO2 for all Montana BART 
sources, our FIP proposal uses a 20-year equipment life in all the BART 
analyses (provided that the equipment life of each control option is 20 
years or more). The CRF for a 20-year equipment life and 7% discount 
rate (the latter being recommended in Office of Management and Budget 
(OMB) Circular A-4, which we cited at 77 FR 24016) is 0.0944. As shown 
in Table 44 at 77 FR 24016, we multiplied Holcim's estimated capital 
cost of $1,312,800 by this CRF to yield a capital recovery cost of 
$123,928.
    With regard to Holcim's comment that a 20-year amortization would 
misrepresent actual costs in the event that SCR rather than SNCR were 
to be required in the next planning period, we cannot anticipate every 
event that might happen in the future and we are not required to do so 
in establishing an amortization period.
    We disagree with the statement that SNCR costs at the Trident kiln 
should be similar to Holcim's Hagerstown kiln. The Trident kiln is much 
smaller than the Hagerstown kiln. The Trident kiln is permitted at 
425,000 tons per year of clinker production. Montana Air Quality Permit 
0982-11, Condition II.B.6. The Hagerstown kiln is rated at 
630,114 tons per year of clinker production capacity. Prevention of 
Significant Deterioration (PSD) Permit Application for Approval, Holcim 
Hagerstown, October 30, 2008. Also, the Hagerstown kiln--a dry kiln--
likely has different emission rates than the Trident kiln. Without more 
information, it is not possible to determine how much of the claimed 
$1,920,000 capital cost of the Hagerstown kiln SNCR system, as well as 
operating costs, would be costs that are permissible for inclusion in a 
BART cost estimate. For these reasons, without more information, the 
costs of the SNCR system at the Hagerstown kiln are not useful for 
estimating the costs at the Trident kiln. Therefore, we continue to 
rely on the SNCR capital cost estimate of $1,312,800 and operating cost 
estimate of $147,288 for Trident, already supplied to us by Holcim in 
the August 2009 submittal. We also note that, even with a capital cost 
of $1,920,000, it appears SNCR would remain cost-effective; Holcim has 
provided no reason why our BART selection would change. This comment 
has not resulted in any changes to our regulatory text for 
NOX BART.
    Comment: Holcim indicated that EPA underestimated the costs of 
installing indirect firing at Trident. Holcim stated that the company 
did not include indirect firing in its 2007 BART analysis

[[Page 57883]]

and did not consider indirect firing to be an appropriate technology to 
evaluate to achieve NOX reductions at Trident. Holcim stated 
that at EPA's request, the company submitted an estimate to EPA of the 
costs of installing indirect firing at Trident.\15\ Holcim stated that 
in EPA's own analysis, the Agency ``inexplicably and arbitrarily'' 
eliminated a significant portion of the costs from Holcim's analysis. 
Nonetheless, even using EPA's underestimated costs for the installation 
of indirect firing and mistaken assumption that indirect firing could 
reduce NOX emissions at Trident by 15%, neither the average 
cost-effectiveness of indirect firing nor the incremental cost-
effectiveness of indirect firing warrant a determination that indirect 
firing should be selected as BART.
---------------------------------------------------------------------------

    \15\ Letter from Greg Gannon to Laurel Dygowski, June 9, 2009. 
(See EPA-R08-OAR-2011-0851-0038).
---------------------------------------------------------------------------

    Holcim pointed out that EPA is proposing to require that Holcim 
install both SNCR and indirect firing at Trident based on its analysis 
of the average cost-effectiveness of installing both technologies 
together. Holcim stated that the overwhelming majority of 
NOX emissions reductions and improvements in visibility 
would result from the installation of SNCR alone and that by ignoring 
the incremental costs of SNCR + indirect firing, and focusing solely on 
the average cost effectiveness, Holcim states that EPA tries to make 
the costs of SNCR + indirect firing appear reasonable. Holcim stated 
that the average cost-effectiveness for the installation of SNCR at 
Trident is well within the range of what EPA has considered for BART, 
but that EPA estimated the average cost effectiveness of indirect 
firing to be $4,279/ton, which is far outside the range of what EPA has 
considered to be reasonable for BART. With such high costs for indirect 
firing, the incremental cost-effectiveness of SNCR + indirect firing as 
compared to SNCR alone is $8,029/ton. Holcim stated that EPA should 
consider both the average and incremental cost effectiveness of its 
BART analysis for Trident. Holcim stated that, although EPA clearly 
identified the incremental cost effectiveness of SNCR + indirect 
firing, EPA ``inexplicably ignored this unreasonable figure in 
concluding that the combination of technologies constitutes BART for 
Trident''. Holcim stated that the incremental cost effectiveness of 
SNCR + indirect firing is unreasonable given the slight to nonexistent 
improvement in visibility that it would achieve and that EPA should 
eliminate this combination of technologies as BART.
    Holcim further stated that, based on modeling, the installation of 
indirect firing and SNCR at Trident, even if it could achieve EPA's 
claimed 58% reduction in NOX emissions, would result in an 
improvement of visibility of only 0.424 deciview in Gates of the 
Mountains WA and that this does not constitute a significant or 
perceptible improvement in visibility. Holcim stated that EPA's 
conclusion is even more unjustifiable considering the actual percentage 
reduction that Trident could be expected to achieve with the 
installation of SNCR of approximately 35% on an annual average basis.
    Finally, Holcim stated that the average cost effectiveness 
estimates for indirect firing alone ($4,279/ton) and for SNCR + 
indirect firing ($1,528/ton) are well above what EPA used as a cost-
effectiveness threshold for NOX in the Cross-State Air 
Pollution Rule (CSAPR), which EPA promulgated last year to address 
health-based standards. Holcim stated that the company does not 
understand why EPA believes it appropriate to use a higher cost 
threshold for an aesthetic standard than it has for a health-based 
standard.
    Response: We agree with aspects of this comment, but disagree with 
others. We agree that an incremental cost effectiveness of $8,029/ton, 
for LNB/indirect firing + SNCR, versus SNCR alone makes LNB/indirect 
firing + SNCR unreasonable for BART at the Trident kiln.
    As explained in a previous response above, we have removed 
switching to indirect firing and a LNB from consideration as an option 
for further reducing NOX emissions and are treating any 
NOX emission reduction that may have been achieved from 
installation of a new burner as part of the emissions baseline. We have 
recalculated the proposed BART limit for NOX to reflect a 
50% reduction in NOX emissions from that baseline by 
addition of SNCR alone, rather than the 58% reduction we previously 
used, which reflected a switch to indirect firing and a LNB plus SNCR. 
The recalculated NOX BART limit is 6.5 lb/ton clinker.
    We disagree, however, with the statement that EPA analyzed for 
indirect firing as a separate control option. We did not. Throughout 
our proposal, we refer to the control option as LNB and are now 
clarifying that this option was intended to include switching to 
indirect firing and a LNB. We explained at 77 FR 24015 that the capital 
cost estimate of $4,385,307 for LNB includes the cost of converting 
from a direct to an indirect firing system to accommodate LNB, 
including installation of a baghouse, additional explosion prevention, 
pulverized coal storage, and dosing equipment. We cited Holcim's 
additional response of August 2009 as the source of this information.
    We disagree with the statement that SNCR could be expected to 
achieve only a 35% reduction in NOX emissions. See our 
response to Holcim's comment above.
    We also disagree with the statement that any controls required by 
our action must demonstrate a perceptible visibility improvement. In a 
situation where the installation of BART may not result in a 
perceptible improvement in visibility, the visibility benefit may still 
be significant. The July 6, 2005 BART Guidelines state:

    even though the visibility improvement from an individual source 
may not be perceptible, it should still be considered in setting 
BART because the contribution to haze may be significant relative to 
other source contributions in the Class I area. Thus, we disagree 
that the degree of improvement should be contingent upon 
perceptibility. Failing to consider less-than-perceptible 
contributions to visibility impairment would ignore the CAA's intent 
to have BART requirements apply to sources that contribute to, as 
well as cause, such impairment.

70 FR 39129. Visibility impacts below the thresholds of perceptibility 
cannot be ignored because regional haze is produced by a multitude of 
sources and activities which are located across a broad geographic 
area.
    With regard to Holcim's comment comparing the cost-effectiveness of 
controls required under the CSAPR, with cost-effectiveness of controls 
required under the Regional Haze Rule and the BART Guidelines, we 
reject the comparison. The two rules address different requirements of 
the CAA.
    Comment: Holcim agreed with EPA's proposal that no additional 
controls constitute BART for SO2 at Trident but objected to 
the imposition of a 30-day SO2 limit. In Holcim's view, 
imposing a 30-day limit is neither reasonable nor necessary. Holcim's 
Trident plant relies on inherent scrubbing to achieve its extremely low 
SO2 emissions. EPA's modeling confirms that SO2 
emissions from Trident have effectively zero visibility impact. Trident 
could more than double its current SO2 emissions and still 
not have any reliably predictable impact on visibility (less than 0.1 
deciview). Even if all SO2 emissions from Trident were 
eliminated, visibility would improve in Gates of the Mountains WA by 
less than 0.05 deciview; less than one-twentieth of a perceptible 
change in visibility. See

[[Page 57884]]

77 FR at 24021. Id. at 24021, Table 63. Holcim stated that the kiln 
could not increase its emissions sufficiently to affect visibility 
without exceeding its currently enforceable limit. Consequently, Holcim 
stated that there simply is no need to impose short term SO2 
limits to protect visibility.
    Second, Holcim stated that because Trident relies on inherent 
scrubbing to control SO2, the plant has no real control over 
the short-term emissions variability that results from the natural 
variability in limestone from its quarry. The emissions variability 
would never rise to a level that could affect visibility, but it could 
cause Trident to exceed the proposed 30-day limit. Thus, the only 
effect of the 30-day limit would be to impose unnecessary regulatory 
burdens on the plant and expose it to potential penalties for short-
term emissions variability, over which Holcim has no control and which 
would not impact visibility.
    Holcim also commented that EPA is proposing to impose an 
SO2 limit that is not based on the installation of retrofit 
control technology or a process change and that offers no improvement 
in visibility. Holcim stated that because the proposed limit is based 
on current emissions and will not improve visibility, it cannot be 
considered BART; the CAA and EPA's own BART Guidelines require that, in 
determining BART, the Administrator consider the degree of improvement 
in visibility which may reasonably be anticipated to result from the 
use of such technology. Holcim requested that EPA eliminate its 
proposed 30-day SO2 limit as it does not represent BART and 
would impose unnecessary regulatory burdens and new compliance risks 
while serving no visibility purpose.
    Response: We disagree. The July 6, 2005 BART Guidelines state that 
``* * * you must establish an enforceable emission limit for each 
subject emission unit at the source and for each pollutant subject to 
review that is emitted from the source.'' 70 FR 39172. Our FIP proposal 
states that ``States, or EPA if implementing a FIP, must address all 
visibility-impairing pollutants emitted by a source in the BART 
determination process. The most significant visibility impairing 
pollutants are SO2, NOX and PM.'' 77 FR 23993. 
Similarly, the BART Guidelines identify SO2, NOX 
and PM as visibility-impairing pollutants. 70 FR 39160. Since these 
pollutants are subject to review, emission limits must be established. 
This comment has not resulted in any changes to our proposal. We note 
that Holcim has not provided any specific data to demonstrate that they 
may exceed the emission limit established for SO2.
    Comment: Holcim disagreed with EPA's proposal to impose an emission 
limit for PM at Trident of 0.77 lb/ton clinker. Holcim stated that the 
proposed limit, which is based on Trident's current emissions, is 
unjustified because it would result in no visibility impact and that as 
the company had already explained, the selected BART must consider the 
degree of improvement in visibility. Holcim stated that adding a 
duplicative applicable requirement to Trident's Title V permit would 
serve no purpose other than to ``create the potential for multiple 
penalties if the requirement were violated.''
    Response: See the previous response.

F. Comments on CFAC

    Comments: CFAC requested that EPA conduct a BART analysis for their 
facility now, rather than in the future, so that CFAC has more 
information for planning a restart. The NPS commented similarly. CFAC 
also commented that not knowing what the BART controls may or may not 
be for their facility makes it difficult to know whether those controls 
could be installed within the five-year timeframe. Another commenter 
stated that we must either set BART limits for CFAC in the FIP, or we 
must require plant shutdown as part of the FIP.
    Response: We disagree that it is necessary to conduct the BART 
analysis at this time. The information necessary to complete such a 
BART analysis is not available until CFAC's future operational plans 
are known. The requirements for CFAC at 40 CFR 52.1396(n) are 
sufficient at this time. With regard to CFAC's comment that not knowing 
what the BART controls may or may not be for their facility makes it 
difficult to know whether those controls could be installed within the 
five-year timeframe, the BART Guidelines state that we must require 
compliance with emission limits no later than five years following the 
final FIP. 70 FR 39172. CFAC can provide the necessary information to 
EPA to conduct a BART analysis at any time.

G. Comments on Colstrip Units 1 and 2

    Comment: A commenter stated that PPL's modeling files related to 
the June 2008 Addendum to PPL Montana's Colstrip BART Report should be 
placed in the docket.
    Response: We requested the modeling files from PPL and PPL 
responded that they could not locate those files. We based our 
decisions on the more recent modeling described at 77 FR 24002.
    Comment: Commenters stated that they object to our proposed BART 
determinations for NOX and SO2 because it would 
impose emission limits based on SNCR and an additional scrubber vessel, 
respectively. Commenters stated that EPA's proposed BART analysis for 
Colstrip Units 1 and 2 is inconsistent with our statutory obligations 
and our own Guidelines. Commenters suggested that our BART 
determinations contain significant errors. Commenters stated that we 
did not properly or correctly consider the costs of the proposed 
controls, the incremental cost-effectiveness of the controls, and the 
lack of any reasonably expected visibility improvements resulting from 
the proposed controls. Instead of the BART proposed by EPA, commenters 
supported the installation of SOFA for NOX control with an 
emission limit of 0.20 lb/MMbtu, and lime injection for SO2 
control with an emission limit of 0.20 lb/MMBtu (both as a 30-day 
rolling average).
    Response: In proposing our BART determinations, we met the 
statutory requirements under section 169A of the CAA and also followed 
the BART Guidelines. Based on our consideration of the five statutory 
BART factors, we continue to find that BART for NOX is 
SOFA+SNCR with an emission limit of 0.15 lb/MMBtu (30-day rolling 
average). Similarly, based on our consideration of the five statutory 
BART factors, we continue to find that BART for SO2 is lime 
injection and an additional scrubber vessel with an emission limit of 
0.08 lb/MMBtu (30-day rolling average). Each specific issue raised by 
the commenters is addressed in a separate response to comments.
    Comment: Several commenters asserted that EPA's costs for SNCR on 
Colstrip Units 1 and 2 were inaccurate and that SNCR is not cost 
effective. Commenters asserted that this was due to a number of errors, 
including use of an incorrect baseline, overstating the emission 
benefits that can be achieved with SNCR, and using improper cost 
estimation techniques. The commenters submitted their own cost 
estimates challenging those reported by EPA.
    Response: EPA estimated a cost effectiveness for SNCR+SOFA of about 
$1,550/ton. This estimate has been confirmed after the proposal through 
information supplied by SNCR vendors.\16\ For this control combination, 
Nalco Mobotec Inc. (Mobotec) estimated a cost effectiveness of roughly 
$1,395/ton, while Fuel Tech Inc. (Fuel Tech) estimated a cost 
effectiveness of $1,642/

[[Page 57885]]

ton. The average vendor cost effectiveness of $1,518/ton is slightly 
lower than what was previously estimated by EPA. Likewise, EPA 
estimated a cost effectiveness for SNCR (after SOFA) of about $3,300/
ton. For SNCR (after SOFA) Nalco Mobotec estimated a cost effectiveness 
of roughly $2,800/ton, while Fuel Tech estimated a cost effectiveness 
of $3,500/ton.\17\ The average vendor cost effectiveness of $3,150/ton 
is slightly lower than what was previously estimated by EPA.
---------------------------------------------------------------------------

    \16\ Memo from Jim Staudt, Andover Technology Partners, to Doug 
Grano, July 10, 2012.
    \17\ Id.
---------------------------------------------------------------------------

    Further, the cost effectiveness of SNCR is of course highly 
dependent on the emission benefits that the control technology can 
achieve. The discrepancy between our cost effectiveness and that 
supplied by the commenters is largely driven by this factor. We address 
this issue, as well as other issues raised by commenters in regard to 
our SNCR cost estimates for Colstrip Units 1 and 2, separately below.
    Comment: Two commenters claimed that EPA used an incorrect baseline 
of 2008-2010 for Colstrip pollutant emissions in our BART analyses. One 
commenter stated that the BART Guidelines require a baseline for BART 
analyses of 2000-2004, while another stated it requires a baseline of 
2001-2003. Both of these baseline periods were prior to the 
installation of additional combustion controls at Colstrip Units 1 and 
2. In addition, one commenter claimed that the 2008-2010 baseline 
emissions are not representative as they reflect a period of economic 
downturn.
    Response: We disagree with these comments. The BART Guidelines 
require you to choose a representative baseline period, but do not 
specify that this period must be 2000-2004 or 2001-2003:

    The baseline emissions rate should represent a realistic 
depiction of anticipated annual emissions for the source. In 
general, for the existing sources subject to BART, you will estimate 
the anticipated annual emissions based upon actual emissions from a 
baseline period.

70 FR 39167.
    As we discussed in our proposed rule, in 2007 PPL installed 
additional combustion controls on Colstrip Units 1 and 2 in order to 
meet new Acid Rain Program emission limits. As these controls were not 
installed to meet BART requirements, we find that it is appropriate to 
reflect them in the baseline emissions.
    Furthermore, annual heat input data contained in the CAMD emissions 
system shows the baseline period of 2008-2010 is representative of the 
operation of the Colstrip Unit 1 and 2. For example, the 2000-2010 
annual heat input for Colstrip Unit 1 ranged from a low of 24,003,758 
MMBtu/yr in 2006 to a high of 30,770,151 MMBtu/yr in 2004. The 2008-
2010 annual average heat input used by EPA in our BART analysis of 
26,578,089 MMBtu/yr falls about in the middle of this range. Therefore, 
the baseline period chosen by EPA is a realistic depiction of the heat 
input (and thereby annual emissions) of the Colstrip Units 1 and 2.
    Finally, the 2000-2004 annual average heat input (the period that 
one commenter asserted we should have used), was 26,966,516 MMBtu/yr, 
and only slightly higher than the heat input used by EPA of 26,578,089 
MMBtu/yr. Therefore, even if we had used the 2000-2004 heat input, it 
would not have affected the BART analysis in a meaningful way.
    Comment: Commenters asserted that EPA overstated the emissions 
benefit of SNCR and that it cannot achieve the level of control 
claimed. The commenters stated that SNCR cannot achieve a 25% emission 
reduction. They also stated that SNCR (in combination with combustion 
controls) cannot achieve an emission limit of 0.15 lb/MMBtu on a 30-day 
rolling average.
    PPL based their assertions on analyses which they obtained from 
SNCR vendors, Nalco Mobotec, Inc. and Fuel Tech Inc. They stated that 
these analyses show that the lowest feasible emissions limit for these 
units on a 30-day rolling average would be in the range of 0.17 to 0.18 
lbs/MMBtu. PPL estimates that only a 10% reduction in NOX 
emissions could be achieved since ammonia slip must be limited to 0.5 
ppm.
    NPS questioned whether SNCR can achieve 0.15 lb/MMBtu on a 30-day 
rolling average due to the sensitivity of SNCR to boiler operation, 
size, and configuration. NPS did not provide any data or information to 
support their concerns other than to state that a query of the CAMD 
emissions system revealed only two EGUs that are consistently meeting 
0.15 lb/MMBtu on monthly basis.
    Response: We disagree that we have overstated the emissions benefit 
of SNCR. Neither the vendor analyses nor the SNCR performance data 
contained in the CAMD emissions system support a conclusion that we 
overstated the emission benefits of SNR.
    The vendor analyses provided by PPL confirm the assumptions made by 
EPA regarding the emissions benefits that can be achieved with SNCR. 
Both vendors indicate that a control efficiency of 25%, as assumed by 
EPA, can be achieved. For example, Fuel Tech indicates that a ``10 ppm 
ammonia slip would result in ~25% NOX reduction.'' \18\ 
Similarly, Mobotec indicates that ``[a]t 7 ppm of ammonia slip, 
NOX emissions could be reduced up to 25%, provided there 
would be no impact on the performance of the air preheater, or other 
plant systems.'' \19\ We realize that the control efficiency of SNCR is 
highly dependent on the level of ammonia slip. However, we find no 
reason that an ammonia slip level of 5 to 10 ppm is unacceptable for 
the Colstrip Unit 1 and 2. These levels of ammonia slip are typical for 
SNCR. In fact, Fuel Tech stated that ``[i]n the coal-fired Utility 
market segment, the SNCR systems have been historically designed for a 
minimum of 5 ppm ammonia slip with some lower sulfur applications with 
NH3 slip levels of 10 ppm.'' \20\ (We address the potential 
impacts from ammonia slip in a separate response to comments).
---------------------------------------------------------------------------

    \18\ Letter from Dale T Pfaff, Fuel Tech, Inc. to Gordon 
Criswell, PPL Montana, May 29, 2012.
    \19\ Letter from Gary Tonnemacher, Mobotec, to Gordon Criswell, 
PPL Montana, May 25, 2012.
    \20\ Fuel Tech, May 29, 2012.
---------------------------------------------------------------------------

    Further, we note that the control efficiencies provided by the 
vendors are for operation at full load, and that higher control 
efficiencies can be achieved at lower loads. For instance, Mobotec 
relates that ``[h]igher NOX reductions can be achieved at 
mid to low load heat inputs, possibly up to 40%.'' \21\ Given that the 
Colstrip Unit 1 and 2 frequently operate at below full load, it is 
likely that on an annual basis SNCR can achieve better than the 25% 
emission reduction assumed by EPA.
---------------------------------------------------------------------------

    \21\ Mobotec, May 25, 2012.
---------------------------------------------------------------------------

    PPL has erred in stating that the control efficiency of SNCR is no 
more than 10% since ammonia slip levels must be limited to 0.5 ppm. The 
commenter bases this claim on what they believe to be a precedent set 
in the Centralia Power Plant BART determination. However, the Centralia 
BART determination prepared by Washington stated that, ``TransAlta's 
cost analysis uses a urea-based SNCR system providing a nominal 25% 
reduction in NOX levels with a 5 ppm ammonia slip.'' \22\ 
And as established by the vendor analyses discussed above, much higher 
emission reductions than 10% can be achieved with SNCR at ammonia slip 
levels of 5 to 10 ppm.
---------------------------------------------------------------------------

    \22\ BART Determination Support Document for Transalta Centralia 
Generation LLC Power Plant, Centralia, Washington, Prepared by 
Washington State Department of Ecology, Revised November 2011, p. 
14; Region 10 clarified the typographical error in their Federal 
Register notice via email from Steve Body to Aaron Worstell dated 
July 26, 2012.

---------------------------------------------------------------------------

[[Page 57886]]

    Similarly, the performance data contained in CAMD emissions system 
only serves to reinforce the assumptions made by EPA regarding the 
emission benefits of SNCR. Based on our review of the CAMD emissions 
data, there are many EGUs equipped with SNCR (with combustion controls) 
that are achieving an emission rate of 0.15 lb/MMBtu or lower on a 
monthly basis. One unit in particular, Boswell Unit 4, is very 
comparable to the Colstrip Unit 1 and 2. Boswell Unit 4, like the 
Colstrip Unit 1 and 2, burns sub-bituminous coal and is tangentially 
fired. In addition, Boswell Unit 4 had a baseline annual emission rate 
(with LNB and CCOFA, but prior to the installation of SNCR and SOFA) 
similar to the Colstrip Unit 1 and 2 of approximately 0.35 lb/MMBtu. 
Since the installation of full combustion controls and SNCR, the 
Boswell Unit has achieved a monthly emission rate of below 0.15 lb/
MMBtu. For example, between April 2011 and April 2012, the most recent 
full year of emissions data available in the CAMD emissions system, the 
monthly emission rates for Boswell Unit 4 were between 0.11 and 0.14 
lb/MMbtu. This is a strong indicator of the performance rates that can 
be expected for Colstrip Units 1 and 2.
    We acknowledge that a range of performance rates are currently 
being achieved with SNCR, and are in some cases not as low as at 
Boswell Unit 4. However, without a showing that there are circumstances 
unique to the Colstrip Unit 1 and 2 that would prevent SNCR from 
achieving the same reductions as at Boswell Unit 4, we find no reason 
that an emission limit higher than 0.15 lb/MMBtu on a 30-day rolling 
average is warranted. This is consistent with the BART Guidelines:

    Without a showing of differences between the source and other 
sources that have achieved more stringent emissions limits, you 
should conclude that the level being achieved by those other sources 
is representative of the achievable level for the source being 
analyzed.

70 FR 39166.
    Finally, due to the smaller size of Colstrip Unit 1 and 2 (333 MW 
each), we expect that SNCR would be more effective than at Boswell Unit 
4 (525 MW). This is because the effectiveness of SNCR on large boilers 
is somewhat reduced as the relatively larger cross-section of the 
boiler makes distribution of the reagent difficult.
    For the reasons stated here, we find no basis in claims that we 
overestimated the emission benefits for SNCR.
    Comment: Commenters stated that EPA did not properly consider the 
incremental cost-effectiveness of SNCR at Colstrip Units 1 and 2. 
Commenters stated that EPA improperly assessed the costs of SNCR when 
combined with SOFA, and not as an individual technology. Commenters 
stated that the incremental cost of adding SNCR to SOFA outweighs the 
benefits. One commenter cited portions of the BART Guidelines that 
address consideration of incremental costs between competing 
technologies.
    Response: We disagree with these comments. We addressed why these 
control technologies were analyzed together in our proposed rule:

    The post-combustion control technologies, SNCR and SCR, have 
been evaluated in combination with combustion controls. That is, the 
inlet concentration to the post-combustion controls is assumed to be 
0.20 lb/MMBtu. This allows the equipment and operating and 
maintenance costs of the post-combustion controls to be minimized 
based on the lower inlet NOX concentration.

77 FR 22043.
    If we had not combined the control technologies, then the cost 
effectiveness would have been more favorable to SNCR. This is because 
the inlet to the SNCR would reflect the current annual baseline 
emissions (e.g., 0.308 lb/MMbtu for Colstrip Unit 1, 2008-2010), as 
opposed to the anticipated post-combustion (i.e., with SOFA) rate of 
0.20 lb/MMBtu assumed by EPA. This would lead to larger emission 
reductions being achieved by SNCR, and thereby, more favorable cost 
effectiveness.
    Regardless, EPA did disclose the costs of SNCR alone (after SOFA) 
in our proposed rule. Consider for example our BART analysis for 
Colstrip Unit 1. See 77 FR 24025-24027 and spreadsheet entitled ``EPA 
SNCR Cost Colstrip Unit 1 Final'' located in the docket. The total 
annual cost of SNCR given in our proposed rule was $2,188,569, while 
the emission reductions were 664 tpy. This results in a cost 
effectiveness of $3,291/ton, essentially the incremental cost 
effectiveness between SNCR+SOFA and SOFA as given in Table 77 of the 
proposed rule. EPA selected SNCR as BART in consideration of these 
costs, all of which were presented to the public in our proposed rule.
    Comment: Various commenters stated that EPA disregarded, or did not 
properly account for, issues associated with ammonia slip from SNCR 
systems. The commenters expressed concerns about both potential 
operational and environmental impacts. In regard to potential 
operational impacts, commenters expressed concerns about fouling of the 
air preheater. In regard to potential environmental impacts, commenters 
expressed concerns related to a visible wet plume, greenhouse gases, 
and toxic emissions.
    Response: We disagree with these comments. In our proposed rule, we 
explicitly considered issues associated with ammonia slip from SNCR 
systems. For example:

    As Colstrip Unit 1 burns sub-bituminous PRB coal having a low 
sulfur content of 0.91 lb/MMBtu (equating to a SO2 rate 
of 1.8 lb/MMBtu), [citation omitted] it was not necessary to make 
allowances in the cost calculations to account for equipment 
modifications or additional maintenance associated with fouling due 
to the formation of ammonium bisulfate. These are only concerns when 
the SO2 rate is above 3 lb/MMBtu.[citation omitted] 
Moreover, ammonium bisulfate formation can be minimized by 
preventing excessive NH3 slip. Optimization of the SNCR 
system can commonly limit NH3 slip to levels less than 
the 5 parts per million (ppm) upstream of the pre-air heater.

77 FR 24025.
    This observation has been verified by the vendor analyses submitted 
by PPL. For example, Fuel Tech stated that ``[s]ince the Colstrip 1&2 
coal has low sulfur, there is less concern of ammonium bisulfate 
formation and its associated air preheater pluggage issues.'' \23\
---------------------------------------------------------------------------

    \23\ Fuel Tech, May 29, 2012.
---------------------------------------------------------------------------

    We also find that concerns about the potential for adverse 
environmental impacts, such as a visible wet plume, toxic ammonia 
emissions, or greenhouse gas emissions, are unfounded or exaggerated. 
As previously discussed, optimization of the SNCR system would limit 
ammonia slip to acceptable levels (i.e., 5-10 ppm). Moreover, as noted 
in the BART determination for the Transalta Centralia Power Plant in 
Washington, ammonia in the gas stream is further removed when a wet 
scrubber is present.\24\ Since the Colstrip Units 1 and 2 utilize wet 
scrubbers, no additional plume visibility or other local impacts would 
be anticipated.
---------------------------------------------------------------------------

    \24\ BART Determination Support Document for Transalta Centralia 
Generation LLC Power Plant, Centralia, Washington, Washington State 
Department of Ecology, revised November 2011, p. 13.
---------------------------------------------------------------------------

    While we did not quantify increases in greenhouse gases associated 
with SNCR in our proposed rule, we did quantify the additional amount 
of coal that is needed to account for the loss in thermal efficiency 
and found it to be insignificant. For example:

    SNCR reduces the thermal efficiency of a boiler as the reduction 
reaction uses thermal energy from the boiler.[citation omitted] 
Therefore, additional coal must be burned to make up for the 
decreases in power generation. Using CCM calculations we

[[Page 57887]]

determined the additional coal needed for Unit 1 equates to 77,600 
MMBtu/yr.

77 FR 24026.
    We note that 77,600 MMBtu/yr is only 0.3% of the 2008-2010 annual 
average heat input for Colstrip Unit 1. The increase in CO2 
emissions would be proportional (that is, 0.3%). The formation of other 
greenhouse gases, such as nitrous oxide, would be highly dependent upon 
the reagent used, the amount of reagent injected and the injection 
temperature. Regardless, we note that the potential for CO2 
increases also exists for SCR, the technology favored by some 
commenters. This is due to the energy penalty associated with the large 
pressure drop across the SCR reactor. Therefore, consideration of 
greenhouse gases would not have necessarily favored SNCR over SCR.
    Comment: MDEQ stated that EPA failed to provide analysis or 
consideration of the impact SNCR installation may have on mercury 
controls at Colstrip 1 & 2. MDEQ stated that this failure ignores 
factor 3 of the five-factor analysis, ``Any existing pollution control 
technology in use at the source.'' MDEQ contended that the application 
of SNCR will require these units to displace the sorbent injection 
systems which limit mercury emissions, and that this displacement will 
compromise the Montana Mercury Rule.
    Response: We disagree with this comment. SNCR should have no impact 
on mercury capture in the scrubber or with mercury capture from sorbent 
injection and will neither improve nor harm any efforts at Colstrip 
Units 1 and 2 to comply with Montana's Mercury Rule. There is no reason 
why Colstrip Units 1 and 2 cannot utilize both SNCR and sorbent 
injection (if sorbent injection is what PPL chooses to use at Colstrip 
1 and 2). Injection points for SNCR and for sorbent injection are at 
different locations--the furnace for SNCR and the downstream ductwork 
for sorbent injection. A review of EPA's National Electric Energy Data 
System (NEEDS) reveals that are currently 17 utility boilers equipped 
with both SNCR and activated carbon injection systems.\25\ The list of 
facilities includes units ranging from 65 MW to 405 MW and burning both 
bituminous and subbituminous coals. Therefore, there is no basis for 
the assertion that these two pollution control systems cannot be used 
together on the same facility.
---------------------------------------------------------------------------

    \25\ Memo from Jim Staudt, Andover Technology Partners, to Doug 
Grano, July 13, 2012, p. 9.
---------------------------------------------------------------------------

    Comment: MDEQ stated that EPA lacks consideration of Montana's 
existing SIP requirements. For instance, sources required to add 
controls would have to provide ``de minimis'' notifications under ARM 
17.8.745, or potentially a resource-intensive demonstration that the 
additional control would not contribute to a violation of an air 
quality standard. Additionally, MDEQ stated that some of the proposed 
controls might require either a minor source permit or a major 
modification under the NSR program. MDEQ expressed particular concern 
over EPA's lack of analysis of PPL's estimated increase in ammonia 
slip.\26\ MDEQ suggested that increases in ammonia slip could lead to 
increases in PM2.5 emissions at Colstrip 1 & 2, potentially 
requiring the unit(s) to submit a ``politically controversial, legally 
complex, and technically challenging'' NSR major modification permit. 
MDEQ also stated that an NSR major modification would significantly 
alter the time and cost required to implement the proposed BART.
---------------------------------------------------------------------------

    \26\ September 23, 2011 PPL submittal titled ``NOX 
Control Update to PPL Montana's Colstrip Generating Station BART 
Report.''
---------------------------------------------------------------------------

    Response: We disagree with these comments. MDEQ has not provided 
any data or information to substantiate that our BART determinations 
would interfere with existing SIP requirements, including those for 
permitting. They have only speculated that these might be concerns. In 
addition, these concerns would not negate our obligation to prescribe 
BART controls. We have addressed concerns related to ammonia slip in a 
separate response to comments.
    Comment: Commenters stated that EPA asserted, with no analysis, 
that the energy needs associated with installation SNCR or SCR on the 
Colstrip Unit 1 and 2 are minimal and neither the additional energy 
requirements nor the nonair quality environmental impacts associated 
with disposal of the ash waste or transportation of the chemical 
reagents or catalysts warranted eliminating either SCR or SNCR.
    Response: We disagree with this comment. We provided analysis of 
the energy impacts associated with SNCR and SCR in our proposed rule. 
For example, for the application of SNCR to Colstrip Unit 1 we 
``determined the additional coal needed for Unit 1 equates to 77,600 
MMBtu/yr.'' 77 FR 24026. Similarly, we determined that SCR requires 
``additional electric power to meet fan requirements equivalent to 
approximately 0.3% of the plant's electric output.'' [citation omitted] 
77 FR 24026. We also provided analysis of the non- air-quality impacts 
associated with SNCR and SCR in our proposed rule. See for example 77 
FR 24026. We did not find it necessary to quantify these impacts as 
they are negligible. Also, the nonair quality impacts would be no 
different than those at numerous other boilers where SNCR or SCR have 
been successfully applied. Regardless, the commenters did not present 
any data or information that establishes that the energy or nonair 
quality impacts of SNCR or SCR would make these control options 
unacceptable.
    Comment: NPS stated that allowing five years from promulgation of 
the rule to install SNCR on Colstrip Units 1 and 2 is excessive since 
it can be installed in less than one year.
    Response: We agree that SNCR in some cases can be installed in less 
than one year. However, the BART Guidelines require compliance with the 
BART emission limit as expeditiously as possible but in no event later 
than five years after promulgation of the FIP. 40 CFR 51.308(e)(1)(iv). 
Our FIP is consistent with that requirement.
    Comment: The NPS agreed with EPA that an annual emission rate of 
0.05 lb/MMBtu is achievable with SCR.
    Response: Comment noted.
    Comment: EarthJustice stated that EPA incorrectly rejected SCR as 
BART for NOX pollutant control for Colstrip Units 1 and 2. 
They asserted that EPA's analysis was biased against the selection of 
SCR as BART. They also asserted that we manipulated data, made 
assumptions, and performed calculations where the results are specified 
but the calculation itself is absent from the public record.
    Response: We disagree with these comments. Our selection of 
SNCR+SOFA, and not SCR+SOFA, as BART was based on our objective 
consideration of the five statutory factors. Moreover, all of our 
analyses and assumptions were supported by the docket which was 
available for public review.
    Comment: EarthJustice stated EPA underestimated the NOX 
reductions that can be achieved with SCR technology. They stated that 
major SCR catalyst vendors routinely guarantee at least 90% removal 
efficiency for SCR systems.
    Response: We disagree. EarthJustice has incorrectly assumed that a 
90% control efficiency can be achieved in all applications regardless 
of the input NOX emission rate or other parameters. The 
baseline annual emission rate for Colstrip BART units is around 0.31 
lb/MMBtu (annually). After the installation of SOFA, the emission rate 
is expected to be 0.20 lb/MMBtu (annually). Therefore, a 90% control 
efficiency for SCR would correspond to a controlled emission rate of 
0.02 lb/MMBtu

[[Page 57888]]

(annually). We find that this is an unrealistic expectation of the 
level of control that can be achieved with SCR.
    Comment: EarthJustice stated that EPA incorrectly used the 
Integrated Planning Model (IPM) for the direct capital costs of SCR for 
Colstrip Units 1 and 2 and that we failed to explain why we did so. 
They stated that the BART Guidelines require that the CCM be used for 
BART cost analyses, except for the site-specific cost of the equipment 
itself which will vary depending on site-specific conditions. 
EarthJustice also stated that our use of IPM led to the double counting 
of installation costs.
    Response: We disagree with these comments. We explained our 
rationale for using IPM for direct costs for SCR in the proposed rule:

    We relied on a number of resources to assess the cost of 
compliance for the control technologies under consideration. In 
accordance with the BART Guidelines (70 FR 39166 (July 6, 2005)), 
and in order to maintain and improve consistency, in all cases we 
sought to align our cost methodologies with the EPA's Control Cost 
Manual (CCM).[citation omitted] However, to ensure that our methods 
also reflect the most recent cost levels seen in the marketplace, we 
also relied on control costs developed for the Integrated Planning 
Model (IPM) version 4.10.[citation omitted] These IPM control costs 
are based on databases of actual control project costs and account 
for project specifics such as coal type, boiler type, and reduction 
efficiency. The IPM control costs reflect the recent increase in 
costs in the five years proceeding 2009 that is largely attributed 
to international competition. Finally, our costs were also informed 
by cost analyses submitted by the sources, including in some cases 
vendor data.

77 FR 24024.

    As noted in the proposed rule, our use of IPM was intended to 
ensure that the direct capital costs reflect the most recent cost 
levels seen in the marketplace. Therefore, we disagree that this led to 
an overestimation of the costs of SCR. Also as noted in the proposal, 
while we did use IPM for direct capital costs, the remainder of our 
analysis for SCR conformed to the CCM.
    EarthJustice is mistaken in asserting that our use of IPM led to 
the double counting of installation costs. EarthJustice is also 
mistaken in asserting that ``in the Cost Control Manual, those 
installation costs [direct installation costs] are included as indirect 
capital costs.'' Direct installation costs are treated in the same way 
whether using the CCM or IPM. That is, both provide direct capital 
costs that are inclusive of the direct installation costs. For example, 
the CCM states:

    Direct capital costs (DCC) include purchased equipment costs 
(PEC) such as SCR system equipment, instrumentation, sales tax and 
freight. This includes costs associated with field measurements, 
numerical modeling and system design. It also includes direct 
installation costs such as auxiliary equipment (e.g., ductwork, 
fans, compressor), foundations and supports, handling and erection, 
electrical, piping, insulation, painting, and asbestos removal.\27\ 
(emphasis added)
---------------------------------------------------------------------------

    \27\ CCM, Section 4, Chapter 2, p. 2-41.

    Similarly, the IPM documentation states the bare module costs 
include equipment, installation, buildings, foundations, electrical, 
and the retrofit factor.\28\ Since we used the bare module capital 
costs to replace the direct capital costs in the CCM calculations, we 
did not double count direct installation costs. For example, for 
Colstrip Unit 1 we used the bare module capital cost of $55,578,137 
(2010 dollars) as input for the direct capital cost.
---------------------------------------------------------------------------

    \28\ IPM, Chapter 5, Appendix 5-2A, p. 2.
---------------------------------------------------------------------------

    Comment: EarthJustice stated that EPA overestimated capital costs 
of SCR on Colstrip Units 1 and 2 by using an inflated capital recovery 
factor (CRF) that is not based on accurate, available, site-specific 
information and by underestimating the lifetime of SCR. EarthJustice 
asserted that EPA should have used a CRF based on a 5% interest rate 
and an equipment life of 30 years
    Response: We disagree that the CRF used by EPA led to an 
overestimation of capital costs for SCR. In our cost analysis for 
Colstrip Units 1 and 2, we used an interest (discount) rate of 7% for 
all control options. This is consistent with guidance contained in the 
Office of Management and Budget, Circular A-4, for regulatory 
analysis.\29\ In regard to the equipment life assumed by EPA for SCR, 
the BART Guidelines state:
---------------------------------------------------------------------------

    \29\ Office of Management and Budget, Circular A-4, Regulatory 
Analysis, http://www.whitehouse.gov/omb/circulars_a004_a-4/.

    For example, the methods for calculating annualized costs in 
EPA's OAQPS Control Cost Manual require the use of a specified time 
period for amortization that varies based upon the type of control. 
If the remaining useful life will clearly exceed this time period, 
the remaining useful life has essentially no effect on control costs 
and on the BART determination process. Where the remaining useful 
life is less than the time period for amortizing costs, you should 
---------------------------------------------------------------------------
use this shorter time period in your cost calculations.

70 FR 39169 (emphasis added).

    And in regard to SCR, the CCM states:
    Capital recovery cost is based on the anticipated equipment 
lifetime and the annual interest rate employed. An economic lifetime 
of 20 years is assumed for the SCR system. The remaining life of the 
boiler may also be a determining factor for the system lifetime.\30\ 
(emphasis added)
---------------------------------------------------------------------------

    \30\ CCM, Section 4, Chapter 2, p. 2-48.
---------------------------------------------------------------------------

    The equipment life assumed by EPA is consistent with that specified 
by the CCM for SCR (that is, 20 years). In addition, the consistent use 
of a 7% interest rate and 20 year equipment life ensures that the costs 
are comparable between all of the control options considered (provided 
that each option has an equipment life of at least 20 years). It also 
ensures that the costs are comparable to other BART analyses where 
similar assumptions have been made. However, we acknowledge that there 
may be circumstances where it is reasonable to assume a shorter or 
longer equipment life. In particular, it may be appropriate to consider 
a shorter equipment life where the owner plans to shut a unit down in 
less than 20 years.
    Further, assuming a 30 year economic life would not change our 
conclusions regarding BART for Colstrip Units 1 and 2. For example, for 
Colstrip Unit 1 we have recalculated the cost-effectiveness amortizing 
over 30 years. The resulting cost effectiveness for SCR+SOFA is $2,879/
ton, as compared to the cost effectiveness of $3,195/ton amortizing 
over 20 years which we cited in our proposed rule. We find that the 
cost of SOFA+SCR is reasonable regardless of the assumed equipment 
life. However, we find that the limited visibility benefits would 
continue to preclude our selection of SCR+SOFA as BART.
    Comment: EarthJustice claimed that EPA skewed the cost 
effectiveness results away from SCR for Colstrip Units 1 and 2 by 
overestimating the operations and maintenance costs associated with SCR 
by approximately $600,000. In particular, EarthJustice questioned our 
costs for maintenance, catalyst replacement, and reagent use.
    Response: We disagree. While EarthJustice has suggested alternate 
assumptions that could be made when estimating each of the operation 
and maintenance costs (that is, direct annual costs) noted, they have 
not substantiated that their assumptions are superior to those used by 
EPA. Moreover, they have not substantiated that EPA erred in making any 
of the cost assumptions related to operations and maintenance. They 
have only pointed out instances in which they would make different 
assumptions. Therefore, we see no reason that our cost assumptions for 
O&M should be supplanted by those that EarthJustice would otherwise 
choose in order to arrive at lower cost effectiveness.
    Regardless, if we were to incorporate each of the changes to the 
O&M costs

[[Page 57889]]

suggested by EarthJustice, it would not change our BART determination. 
For example, for Colstrip Unit 1, reducing the O&M costs of SCR by 
$600,000 would only moderately lower the cost effectiveness of SNR+SOFA 
from $3,195/ton to $3,019/ton. Though we find that both of these costs 
are reasonable, we continue to find that there is insufficient 
visibility benefit (0.404 deciview for Unit 1 and 0.423 deciview for 
Unit 2 at the most improved Class I area) to support the selection of 
SCR as BART.
    Comment: EarthJustice stated that EPA made multiple errors in our 
SCR cost analysis for Colstrip Units 1 and 2. EarthJustice claims that 
EPA made errors in relation to the baseline NOX emissions, 
the control efficiency of SCR, the cost estimation method for direct 
capital costs (CCM vs. IPM), specific operation and maintenance costs, 
and the calculation of indirect annual costs (by way of the CRF). 
EarthJustice provided their own cost estimates for SCR, addressing the 
errors which they claimed EPA made. EarthJustice's cost effectiveness 
is 55-65% lower than the values calculated by EPA, making SCR+SOFA 
significantly more cost effective.
    Response: We disagree that we made multiple errors in our SCR cost 
analysis for SCR for Colstrip Units 1 and 2 which led to inaccurate 
cost effectiveness. Each of the errors which EarthJustice claims EPA 
made has been addressed in separate responses. Therefore, we find that 
the cost effectiveness for SCR in the proposed rule was accurate and a 
correct basis for rejecting SCR as BART (in consideration of the 
remaining statutory BART factors).
    Comment: The NPS commented that EPA has placed undue weight on the 
incremental cost effectiveness of SOFA+SCR at Colstrip Units 1 and 2.
    Response: We disagree. In our proposed rule, we estimated the 
incremental cost effectiveness of SCR+SOFA (over SNCR+SOFA) to $5,770/
ton and $5,887/ton, respectively. These costs far exceed the 
corresponding average cost effectiveness of $3,195/ton and $3,235/ton. 
Given these costs, we continue to find that SCR+SOFA is not justified 
by the visibility improvement that would be provided.
    Comment: Some commenters stated that EPA properly concluded that 
SCR does not constitute BART for Colstrip Units 1 and 2, but that EPA 
incorrectly analyzed the capital costs and cost-effectiveness of SCR. 
Commenters stated that EPA failed to consider SCR costs estimates which 
PPL submitted in February 2012.\31\ Commenters also stated that EPA's 
reliance on outdated information is not consistent with its own 
guidance to use engineering estimates and that EPA should modify its 
rationale in the final rule to conclude that, when the actual costs of 
the technology are taken into consideration, SCR is not a cost-
effective technology. In particular, commenters noted that EPA 
estimates the capital cost of the SCR at $78 million and rejects PPL's 
cost estimate of $190 million
---------------------------------------------------------------------------

    \31\ Letter from David Bowen, Burns & McDonnell, to James 
Parker, PPL Montana, February 7, 2012.
---------------------------------------------------------------------------

    Response: We disagree that we incorrectly analyzed the capital 
costs and cost-effectiveness of SCR. We did not accept the SCR cost 
estimates submitted by PPL in February 2012 that were based on cost 
estimates provided to PPL by a consultant. EPA rejected these cost 
estimates for a number of reasons.
    First, the cost estimates provided to PPL by the consultant do not 
represent site-specific costs. The BART Guidelines state that ``[t]he 
basis for equipment cost estimates also should be documented, either 
with data supplied by an equipment vendor (i.e., budget estimates or 
bids) or by a referenced source (such as the OAQPS CCM Fifth Edition, 
February 1996, EPA 453/B-96-001).'' 70 FR 39166. Since the costs 
submitted by PPL were simply adapted from another (undisclosed) utility 
boiler, and are not specific to Colstrip Units 1 and 2, they should not 
be considered a budgetary bid as described in the BART Guidelines. In 
fact, PPL's consultant represents the costs as a ``feasibility capital 
cost estimate'' and not as a budgetary bid.\32\
---------------------------------------------------------------------------

    \32\ Bowen letter.
---------------------------------------------------------------------------

    Second, the capital costs for SCR claimed in PPL's February 2012 
submittal are far in excess of the range of capital costs documented by 
various studies for actual installations. Five industry studies 
conducted between 2002 and 2007 have reported the installed unit 
capital cost of SCRs, or the costs actually incurred by owners, to 
range from $79/kW to $316/kW (2010 dollars).\33\ These studies show 
actual capital costs are much lower than estimated by PPL for Colstrip 
Units 1 and 2 ($571/kW for each unit; 2011 dollars). Moreover, the 
capital costs surveyed by the studies represent a range of retrofit 
difficulties, including very difficult retrofits having significantly 
impeded construction access, extensive relocations, and difficult 
ductwork transitions. Therefore, to the extent that similar retrofit 
difficulties may exist for Colstrip Units 1 and 2, the high end of the 
range documented in the reports is representative.
---------------------------------------------------------------------------

    \33\ Dr. Phyllis Fox, Revised BART Cost-Effectiveness Analysis 
for Tail End Selective Catalytic Reduction at Basin Electric Power 
Cooperative Leland Olds Station Unit 2. Report Prepared for U.S. 
EPA, RTI Project Number 0209897.004.095, March 2011.
---------------------------------------------------------------------------

    Third, we are concerned about the disparity among the various cost 
estimates submitted by PPL. Between August 2007 and February 2012, PPL 
submitted four separate SCR cost estimates for the Colstrip Unit 1 and 
2. In the first SCR cost estimate, submitted in August 2007, PPL 
estimated capital costs of $25,282,233 ($76/kW), total annual costs of 
$7,289,482 and a cost effectiveness of $2,272/ton (each unit; 2007 
dollars).\34\ In the second SCR cost estimate, submitted in June 2008, 
PPL estimated capital costs of $29,581,465 ($88/kW), total annual costs 
of $7,987,179 and a cost effectiveness of $1,735/ton (each unit; 2008 
dollars).\35\ PPL's first and second cost estimates were generally 
performed in conformance with EPA's CCM. The lower cost effectiveness 
in the second submittal was driven primarily by a change in the assumed 
maximum control level (from 0.15 lb/MMBtu to 0.06 lb/MMbtu), and 
thereby greater annual emission reductions. In the third SCR cost 
estimate, submitted in September 2011, PPL estimated capital costs of 
$152,508,328 ($457/kW), total annual costs of $16,733,719 and a cost 
effectiveness of $7.405/ton (each unit; 2011 dollars).\36\ The third 
cost estimates were largely based on control costs developed for the 
Integrated Planning Model.\37\ PPL assumed a retrofit factor of 2 when 
using the IPM approach. We note that this retrofit factor, equating to 
100% over the IPM base model capital costs, was unsupported and far in 
excess of the range described in the IPM documentation: ``Retrofit 
difficulties associated with an SCR may result in capital cost 
increases of 30 to 50% over the base model.'' \38\ In the fourth SCR 
cost estimate, submitted in February 2012, PPL estimated capital costs 
of $190,000,000 ($571/kW), total annual

[[Page 57890]]

costs $19,956,767, and a cost effectiveness of $8,884/ton (each unit; 
2011 dollars).\39\ The fourth cost estimate was also largely based on 
control costs taken from IPM, but was augmented by capital cost 
estimates provided to PPL by a consultant. In all, the capital costs 
varied by a factor of more than seven ($76/kW to 571/kW), and the cost 
effectiveness varied by a factor of more than 5 ($1,735/ton and $8,884/
ton). The large disparity between PPL's February 2012 cost estimates 
and those in their previous submittals led us to question their 
accuracy.
---------------------------------------------------------------------------

    \34\ BART Assessment Colstrip Generating Station, prepared for 
PPL Montana, LLC, by TRC (``Colstrip Initial Response''), August 
2007, Table A4-8(c).
    \35\ Addendum to PPL Montana's Colstrip BART Report Prepared for 
PPL Montana, LLC; Prepared by TRC, (``Colstrip Addendum''), June 
2008, Table 5.3-3.
    \36\ NOX Control Update to PPL Montana's Colstrip 
Generating Station BART Report Prepared for PPL Montana, LLC, by 
TRC, September 2011, Table 2-3b.
    \37\ Documentation for EPA Base Case v.4.10 Using the Integrated 
Planning Model, August 2010, EPA 430R10010.
    \38\ IPM, Chapter 5, Appendix 5-2A, p. 1.
    \39\ Letter from Mark M. Hultman, P.E., TRC, February 9, 2012.
---------------------------------------------------------------------------

    Finally, PPL's February 2012 cost estimates contained cost items 
that are either speculative in nature or not well documented. For 
example, they include capital costs for duct and boiler reinforcement 
even though the potential for boiler implosion was not evaluated by 
PPL's consultant.\40\
---------------------------------------------------------------------------

    \40\ Bowen letter, p. 2.
---------------------------------------------------------------------------

    For the reasons stated above, EPA finds that no changes to the BART 
determinations or to the FIP are needed in response to this comment.
    Comment: Various commenters objected to EPA`s BART determinations 
for Colstrip 1 and 2. EarthJustice urged EPA to require selective 
SCR+SOFA as the best system of continuous emission control to meet a 
0.05 lb/MMBtu NOX emission limit applicable on a 30-day 
rolling average basis. NPS also recommended that we require SCR+SOFA. 
PPL supported a BART emissions rate for NOX of 0.20 lb/MMBtu 
on a 30-day rolling average basis, reflecting the installation of SOFA.
    Response: Based on our consideration of the five statutory BART 
factors, we continue to find that BART for NOX at each of 
the Colstrip Unit 1 and 2 is an emission limit of 0.15 lb/MMBtu (30-day 
rolling average) achievable with SNCR+SOFA.
    Comment: PPL stated that EPA's proposed emission limit for PM of 
0.10 lb/MMbtu on a 30-day rolling average for each of the Colstrip Unit 
1 and 2 is flawed. PPL asserted that the current PM limit is 0.10 lbs/
MMBtu as an annual average, based on a compliance assurance monitoring 
plan together with annual stack testing. In order to accommodate the 
shorter averaging period, the PPL suggested that the 30-day rolling 
average emission limit proposed in the FIP be increased to 0.12 lb/
MMBtu.
    Response: We disagree with some aspects of this comment, but agree 
with others. PPL has erred in stating that the current PM limit is 0.10 
lb/MMBtu as an annual average. The Final Title V Operating Permit 
(OP0513-06) indicates that the emission limit is 0.10 lb/
MMbtu, but does not provide an averaging period. The Title V permit 
requires that compliance with the emission limit be demonstrated by a 
Method 5 or Method 5B stack test once per year. As these stack test 
methods typically consist of three sampling runs of at least 120 
minutes in duration, and are not long-term continuous measurements, it 
is not possible to average the emissions over 30-days or a year. For 
this reason, we corrected the proposed PM emission limits in a 
correction notice. 77 FR 29270. We clarified that that emission limits 
for NOX and SO2, but not PM, shall apply on a 30-
day rolling average.
    As we are not requiring that PM emission limits apply on a 30-day 
rolling average, PPL's suggestion that the emission limit be increased 
to 0.12 lb/MMBtu is no longer relevant. The PM emission limits will 
remain unchanged from those in the proposed rule which are identical to 
those in the Title V permit.
    Comment: EarthJustice stated that EPA's exemption of Colstrip Units 
1 and 2 from BART for PM is improper and unsupported. EarthJustice 
asserts that EPA has not complied with its statutory and regulatory 
obligations to determine BART for PM emissions from Colstrip Units 1 
and 2 in that EPA simply made a declaration and skipped the statutory 
process. EarthJustice stated that the existing venturi scrubbers are 
not best technology and have not been considered such for a long time 
because particle scrubbers do not remove particulates sufficient to 
comply with basic CAA requirements. In addition, EarthJustice stated 
that EPA should have considered more effective technologies, such as 
baghouses.
    Response: We disagree. As with existing SO2 controls, we 
do not find that it is necessary to consider the replacement of 
existing PM controls with new controls. This is particularly true for 
PM as the existing controls for Colstrip Units 1 and 2 currently reduce 
emissions by more than 98%. Moreover, the contribution to the baseline 
visibility impact from PM is very small (e.g., for Colstrip Unit 1, 
less than 4% of 0.922 deciview, or 0.037 deciview). The most visibility 
improvement that could be expected, even if all PM were eliminated, is 
0.037 deciview. The visibility improvement that could be expected with 
upgrades to the existing PM controls is only a fraction of 0.037 
deciview. Therefore, it was reasonable for us to conclude that the 
existing controls represent BART.
    In addition, EarthJustice has conflated the most stringent controls 
with BART. BART is not necessarily the most stringent controls, but the 
best system of continuous emission reduction taking into consideration 
the five statutory factors.
    Comment: NPS stated that they disagree with the PM emissions that 
we used in modeling the visibility impacts for Colstrip Units 1 and 2. 
They stated that the PM emissions data provided by PPL is more 
representative because it included both condensable and filterable PM 
emissions, while the PM data used by EPA did not measure condensable 
PM.
    Response: The difference in the approach used to characterize PM 
emissions for visibility modeling purposes is negligible. Moreover, as 
the PM emissions were held constant for all of the control scenarios 
that EPA modeled, they had no impact on our BART determinations for 
NOX and SO2.
    Comment: EarthJustice stated that EPA made the same error in 
calculating baseline emissions in its SO2 BART determination 
for Colstrip Units 1 and 2 as it did in its NOX BART 
determination. EarthJustice asserted that EPA should have used a 
baseline of 2001-2003.
    Response: We disagree with this comment. As discussed in a separate 
response to comments, we have established a baseline which provides a 
realistic depiction of anticipated annual emissions for the source. For 
example, the 2008-2010 baseline we used for Colstrip Unit 1 reflects 
annual average emissions of 5,548 tons/yr. By comparison the annual 
average emissions for 2000-2010, 5,504 tons/yr, were only slightly 
lower.
    Comment: PPL stated that EPA's estimate of the performance that can 
be achieved with lime addition on Colstrip Units 1 and 2 was wrong. The 
commenter stated that EPA's assumed emission rate for SO2 of 
0.15 lbs/MMBtu was overly optimistic, and that a rate of 0.20 lbs/MMBtu 
on a 30-day rolling average basis is achievable.
    Response: We disagree with this comment. The emission rate which 
EPA assumed for limestone lime addition (injection) on Colstrip Units 1 
and 2 was 0.15 lb/MMBtu on an annual basis, not on a 30-day rolling 
average basis. This was based on PPL's amended BART submittal of August 
of June 2008.\41\ We did not specify a 30-day rolling average

[[Page 57891]]

emission limit for limestone injection since we did not select it as 
BART.
---------------------------------------------------------------------------

    \41\ Colstrip Addendum, p. 4-1.
---------------------------------------------------------------------------

    Comment: PPL commented that installation of an additional scrubber 
vessel is technically impracticable, if not infeasible, due to space 
constraints and the potential for equipment scaling.
    Response: First, addition of a fourth scrubber vessel for each of 
Colstrip units 1 and 2 does not appear to be impracticable due to space 
constraints. PPL's argument that there is no space availability for an 
additional scrubber vessel is not supported by its own consultant. In 
addition, the site visit conducted by EPA \42\ verified and the site 
plan provided by PPL shows ample space for locating additional 
equipment. A satellite image of units 1 and 2 located in the 
docket.\43\ In fact, PPL's consultant, Burns & McDonnell was able to 
find space for a new vessel with associated ductwork: ``[t]here is 
sufficient space behind the stacks for installation of the fourth 
scrubber module, ID fan, ductwork and accessories.'' \44\ As URS 
pointed out, this might require an additional booster fan, which is 
included in the Burns & McDonnell estimate.\45\
---------------------------------------------------------------------------

    \42\ On September 27, 2011 Aaron Worstell and Vanessa Hinkle 
conducted a site visit at Colstrip.
    \43\ Staudt memo, p. 4.
    \44\ Report on the Fourth Scrubber Module Cost Estimate for PPL, 
Burns and McDonnell, p. 4-3.
    \45\ Letter from Jonas Klingspor, URS Corporation, to Gordon 
Criswell, PPL Montana, June 15, 2012.
---------------------------------------------------------------------------

    Second, an additional scrubber vessel may not be necessary to avoid 
scaling. It is possible to inject lime and mitigate the risk of scaling 
through addition of a forced oxidation system or by use of chemical 
additives that mitigate scaling. The current system uses natural 
oxidation. Forced oxidation will enable higher lime injection rates 
while avoiding scaling. Forced oxidation systems will require blowers 
and piping, and agitators that could be retrofit on the existing 
scrubber vessels at what is likely to be a much lower cost than the 
cost of a new absorber vessel. An alternative to forced oxidation is 
use of chemical additives that address scaling. These additives are 
available from companies such as Nalco Chemical Company.
    We find that it is acceptable for PPL to reduce emissions by means 
other than installing an additional scrubber vessel, provided that the 
emission limit of 0.08 lb/MMBtu on a 30-day rolling average is met.
    Comment: PPL stated that EPA overstated the emissions benefit of an 
additional scrubber vessel.
    Response: PPL argues that an additional vessel would not in fact 
reduce emissions because velocity through the existing scrubber vessel 
tray will be reduced. As noted in responses to other comments, an 
additional scrubber vessel may not be necessary to achieve 95% 
SO2 capture. Nevertheless, with regard to addition of 
another scrubber vessel and the impact on SO2 reduction, PPL 
relies on a June 15, 2012, letter from Jonas Klingspor of URS 
Corporation that states the reduced gas velocity would reduce 
SO2 reduction. The URS letter and PPL, however, overlook the 
fact that the openings in the tray for the existing vessels could be 
reduced to restore gas velocity to the original level.
    URS provided estimates of emission rates possible under different 
conditions. The analyses performed by URS were limited either by 
increased scaling (the lowest rate of 0.13 lb/MMBtu with three vessels) 
or lower absorber gas velocity (0.16 lb/MMBtu with four vessels). Since 
URS did not evaluate addition of a forced oxidation system or any other 
means to address scaling, it is likely that a significantly lower 
emission rate than 0.13 lb/MMBtu is possible while using three vessels. 
And, addition of a fourth scrubber vessel, with tray openings in the 
three original vessels adjusted to maintain gas velocity, in 
combination with a forced oxidation system would certainly increase 
SO2 capture performance even more.
    Regardless, if PPL uses the additional scrubber vessel as a spare 
in a manner similar to that for Colstrip Units 3 and 4, then gas flow 
will remain unchanged. In this mode of operation, the spare scrubber 
vessel helps allow for maintenance that is needed due to the scaling 
caused by the additional lime. Without the spare vessel, the unit must 
be shut down to perform the maintenance. This is the mode of operation 
proposed by PPL in their August 2007 submittal.
    Comment: Commenters stated that an additional scrubber vessel costs 
far more than EPA proposed and is therefore not cost-effective. 
Commenters stated that it was inappropriate for EPA to rely on outdated 
costs for an additional scrubber vessel in our proposed rule. PPL 
provided cost estimates obtained from Burns & McDonnell \46\ showing 
higher costs than estimated by EPA.
---------------------------------------------------------------------------

    \46\ Burns and McDonnell, p. 1-1.
---------------------------------------------------------------------------

    Response: Foremost, we note that the costs that we cited for an 
additional scrubber vessel in our proposed rule were costs provided by 
PPL in their BART submittals of August 2007 and June 2008. PPL did not 
explain why the cost estimates submitted by PPL during the comment 
period are more than two and a half times their original cost 
estimates.
    The cost estimated by Burns & McDonnell of adding a single module 
to treat 25% of the flue gas is unreasonable, equating to around $213/
kW ($71 million divided by 333,000 kW),- or the equivalent of $853/kW 
when adjusting for the fact that only one fourth of the flue gas is 
being treated. To put this in perspective, this is more costly on a $/
kW basis than the typical cost of a complete limestone forced oxidation 
wet FGD system (around $500/kW) that would provide over 95% removal for 
100% of the flue gas.\47\ Also, according to the 2010 EIA Form 860 
Enviroequip data, the original scrubber structure with three modules 
for Colstrip Unit 1 cost $34 million in 1975 (slightly over $100/kW). 
Using the Chemical Engineering Plant Cost Index (CEPCI) to escalate to 
2011 dollars, the cost in today's dollars would be about $109 million 
($34 million times 585.7/182.4, or about $327/kW). This would suggest 
the cost of an additional vessel to be on the order of $27 million, or 
about 38% of what Burns & McDonnell estimated and consistent with what 
EPA has previously estimated. Moreover, the difference in cost between 
EPA's estimate and what Burns & McDonnell has estimated is far too 
large to be explained by the additional ductwork and fans associated 
with the retrofit, which PPL asserts are necessary. Additionally, Table 
4-1 of the documentation from Burns & McDonnell has several costs that 
are questionable or high ($900,000 for Owner's Project Management and 
$400,000 for Owner's Legal Counsel and $3.4 million in Escalation) and 
others that are very high and therefore require better explanation 
($8.1 million for furnish and erect packages plus the estimates for 
Mechanical, Electrical and Civil and Structural Construction that total 
over $12 million). Engineering costs as well as many other costs are 
typically determined as a percentage of the other costs, therefore the 
effect of overestimation of one cost is compounded because it 
contributes to overestimation of other costs. Because the estimate by 
Burns & McDonnell is so much higher than what is reasonably expected 
and includes several unsubstantiated and questionable cost elements. In 
any event, an additional scrubber vessel may not be necessary if a 
forced oxidation system or other means to control scaling is used on 
the existing three scrubber vessels. PPL may determine that other means 
may be

[[Page 57892]]

better than adding an additional scrubber vessel in terms of cost or 
other factors for achieving the BART emission rate.
---------------------------------------------------------------------------

    \47\ IPM, Chapter 5, Table 5-4 shows a range of illustrative $/
kW costs.
---------------------------------------------------------------------------

    Comment: Commenters stated that EPA did not properly consider the 
incremental cost-effectiveness of additional scrubber vessels at 
Colstrip Units 1 and 2. Commenters stated that while the average cost-
effectiveness of lime injection and an additional scrubber vessel is 
$912/ton, the incremental cost-effectiveness of a scrubber vessel is 
$2,379/ton, nearly three times higher.
    Commenters also stated that it was improper for EPA to evaluate 
lime injection and an additional scrubber vessel together. Commenters 
stated that the incremental cost of adding an additional scrubber 
vessel to lime injection outweighs the benefits. In particular, they 
noted that use of lime injection alone would cost $1,883,200, while the 
addition of a scrubber vessel adds $2,217,000 to the total cost. By 
contrast, they noted that the SO2 reductions achieved from 
the addition of the scrubber vessel are 929 tpy, while the use of lime 
injection alone results in emission reductions of 3,557 tpy.
    Response: We agree with this comment in part. We miscalculated the 
incremental cost effectiveness of an additional scrubber vessel at 
Colstrip Unit 1 (which we stated to be $1,975/ton), but not at Colstrip 
Unit 2 ($2,410/ton). The correct incremental cost effectiveness for an 
additional scrubber vessel at Colstrip Unit 1 is $2,380/ton, not 
$1,975/ton as given in our proposed rule.
    However, we disagree that it was improper to evaluate lime 
injection with an additional scrubber vessel together. We also disagree 
that cost of the additional scrubber vessel outweighs the benefits. For 
example, for Colstrip Unit 2, individually the total annual cost of an 
additional scrubber vessel is $2,210,000, while the emission reduction 
is 917 tons per year. This results in a cost effectiveness of $2,410, 
essentially the same as the incremental cost effectiveness between the 
two control options. The visibility improvement from lime injection 
alone is 0.225 deciview (at Theodore Roosevelt NP), while the 
improvement from lime injection with an additional scrubber vessel is 
0.280 deciview (at Theodore Roosevelt NP). We continue to find that the 
cost is reasonable given the visibility benefits and that lime 
injection with an additional scrubber vessel represents BART.
    Comment: PPL commented that in proposing SNCR, EPA appears to rely 
on its determination that relevant Class I areas are currently above 
the Regional Haze Glide Path (RHGP). 77 FR 24,038. The RHGP is an 
important factor for the reasonable progress goals, but it is not one 
of the five statutory factors specified for EPA to consider in its BART 
analysis. Furthermore, as discussed above, there is no incremental 
benefit in visibility from installation of SNCR that would affect the 
area improvement in visibility relative to the glide path.
    Response: We agree with some aspects of this comment and disagree 
with others. We agree that the Regional Haze glidepath is not one of 
the five statutory factors specified for EPA to consider in its BART 
analysis. We based our decision solely on the five statutory factors.
    Comment: EarthJustice stated that EPA settled for minor adjustments 
for SO2 pollutants from Colstrip Units 1 and 2 instead of 
proper BART controls. In particular, EarthJustice stated that EPA 
failed to examine a full suite of options for SO2 BART, 
including replacement of the existing scrubbers with state-of-the-art 
scrubbers that could remove 98% of the SO2 from Colstrip 
Units 1 and 2.
    In addition, EarthJustice claimed that EPA failed to consider all 
feasible upgrades to the existing venturi scrubbers, including the use 
of magnesium enhanced lime. EarthJustice stated that significant 
emission reductions could be achieved via these upgrades, even without 
the installation of an additional scrubber vessel. EarthJustice held 
that an emission limit of 0.06 lb/MMbtu can be achieved with these 
upgrades.
    Response: We disagree that we should have considered replacement of 
the existing controls. As noted in our proposed rule, for example:

    The Colstrip Unit 1 venturi scrubber currently achieves greater 
than 50% removal of SO2. For units with preexisting post-
combustion SO2 controls achieving removal efficiencies of 
at least 50%, the BART Guidelines state that upgrades to the system 
designed to improve the system's overall removal efficiency should 
be considered.

77 FR 24028.
    The BART Guidelines only recommend evaluating constructing a new 
FGD system ``[f]or coal-fired EGUs with existing post-combustion 
SO2 controls achieving less than 50 percent removal 
efficiencies.'' 70 FR 39171. Therefore, it was appropriate for us to 
not consider new state-of-the art scrubbers, or for that matter, any 
replacement technology.
    As noted in a separate response, we agree that it may not be 
necessary to add an additional scrubber vessel in order to achieve an 
emission limit of 0.08 lb/MMBtu on a 30-day rolling average. We 
acknowledge that it may be possible to achieve the emission limit with 
modifications to the existing scrubbers, such as a forced oxidation 
system or by use of chemical additives that mitigate scaling. However, 
these alternative approaches would likely be at a lower cost than an 
additional scrubber vessel. Given that equivalent emission reductions 
would be achieved at lower costs, the cost effectiveness would be even 
more reasonable. Accordingly, we are extending flexibility to PPL to 
meet the emission limit using the lowest cost approach.
    Regardless of whether PPL chooses to meet the emission limit with 
an additional scrubber vessel or modifications to the existing scrubber 
vessels, we continue to find that an emission limit of 0.08 lb/MMBtu, 
and not 0.06 lb/MMBtu as suggested by the commenter, is appropriate. As 
noted in the proposed rule, this is based on the level of performance 
being achieved by Colstrip Units 3 and 4 which already employ scrubbing 
systems similar to that being contemplated for Colstrip Units 1 and 2.
    The use of MEL is addressed in a separate response to a similar 
comment from EarthJustice in regard to Colstrip Units 3 and 4.

H. Comments on Corette

    Comment: EarthJustice indicated that EPA`s decision not to impose 
BART on Corette violates the statutory requirements for BART and is not 
supported by the facts. EarthJustice stated that EPA engaged in the 
same kind of non-BART result oriented process for Corette as it did for 
Colstrip. They asserted that EPA's approach is no more legitimate or 
compliant with the haze requirements in the case of Corette. Based on 
their own BART analyses, they determined that BART for Corette is 
installation of a dry scrubber and baghouse for the control of 
SO2 and PM emissions, and SCR+SOFA for NOX.
    Response: We disagree with this comment. Our selection of BART for 
Corette was based on our objective consideration of the five statutory 
factors. We continue to find no additional controls are necessary for 
Corette. Below, we address specific issues raised by EarthJustice in 
regard to our BART determination for Corette.
    Comment: EarthJustice stated that, as with Colstrip Units 1 and 2, 
we used an improper baseline in our BART evaluation of 2008-2010. 
EarthJustice asserted that using these years artificially depresses the 
emissions baselines, which in turn makes visibility improvement appear 
less than they

[[Page 57893]]

actually are and thereby makes BART alternatives look less cost-
effective than they actually are.
    Response: See response to similar comments made by EarthJustice in 
regard to Colstrip Units 1 and 2. Here again, as required by the BART 
Guidelines, we used a baseline that is reflective of actual operations. 
We acknowledge that the 2008-2010 emissions for both SO2 and 
NOX were in fact somewhat lower than the long-term trend. 
For example, the 2000-2010 SO2 emissions were 3,129 tpy, 
while the 2008-2010 emissions were 2,723 tpy. Similarly, the 2000-2010 
NOX emissions were 1,748 tpy, while the 2008-2010 emissions 
were 1,625 tpy. Nonetheless, the difference in the baseline emissions 
would not have impacted the cost-effectiveness calculations in an 
appreciable manner.
    Comment: EarthJustice stated that EPA understated the cost 
effectiveness of SCR+SOFA.
    Response: See response to similar comment made by EarthJustice in 
regard to Colstrip Units 1 and 2.
    Comment: EarthJustice stated that EPA's cost-effectiveness 
calculations for SO2 controls for Corette contain a number 
of incorrect assumptions. In particular, EarthJustice stated that much 
lower emission reductions can be achieved with LSD (90% with low sulfur 
coal) than assumed by EPA. Also, EarthJustice stated that EPA's 
approach of using IPM for capital costs resulted in a double counting 
of installation costs.
    Response: We disagree. See response to similar comment made by 
EarthJustice in regard to Colstrip Units 1 and 2.
    As we have noted previously, EarthJustice has erred in assuming 
that a given control efficiency can be achieved in all applications 
regardless of the input emission rate or other parameters. The level of 
performance assumed by EPA for LSD (0.065 lb/MMBtu annually) is 
generally reflective of what can be achieved with this technology.
    Further, we used IPM based calculations for both capital costs and 
O&M costs for SO2 controls at Corette. (This is unlike for 
NOX controls, where we used IPM based capital costs to 
reflect recent market trends). Therefore, we could not have double 
counted the installation costs for SO2 controls (from IPM 
and the CCM).
    Comment: EarthJustice stated that EPA wrongly exempted Corette from 
BART for PM.
    Response: See response to a similar comment made by EarthJustice in 
regard to PM BART for Colstrip Units 1 and 2.
    Comment: PPL stated that they support our conclusions with respect 
to BART for Corette that further controls are not justified.
    Response: Comment noted. The final FIP does not require additional 
controls for Corette.
    Comment: Commenters stated that they disagree with EPA's cost 
analysis for NOX and SO2 control technologies at 
Corette and that EPA incorrectly concluded that a number of the control 
technologies are cost-effective. Commenters noted that PPL submitted a 
five factor BART analysis for Corette in August 2007, and later 
supplemented with the analysis with updated information in June 2008 
and September 2011.\48\ Commenters stated that in view of the 
information that PPL provided, EPA incorrectly concluded that SOFA, 
SOFA+SNCR, and SOFA+SCR are ``all cost effective technologies'' (77 FR 
24043) and that the proposed FIP also incorrectly concluded that dry 
sorbent injection (DSI) for SO2 is cost-effective at $3,940/
ton. 77 FR 24047.
---------------------------------------------------------------------------

    \48\ NOX Control Update to PPL Montana's J.E. Corette 
Generating Station BART Report, September 2011, Prepared for PPL 
Montana, LLC by TRC, at ES-1 (``NOX Control Update''); 
SO2 Control Update to PPL Montana's J.E. Corette 
Generating Station BART Report, August 2011, Prepared for PPL 
Montana, LLC by TRC, at ES-1 (``SO2 Control Update'')
---------------------------------------------------------------------------

    Commenters stated that as documented in PPL's 2011 submissions, the 
company used the IPM control technology cost estimation techniques, 
which are more robust than those used in previous BART reports 
submitted by PPL.\49\ Commenters stated that with respect to 
NOX, PPL determined the cost-effectiveness of SNCR to be 
approximately $13,544/ton (as compared to EPA's $2,596 for SOFA+SNCR) 
and the cost-effectiveness for SCR to be $8,457/ton of additional 
NOX controlled (as compared to EPA's $4,491 for SOFA + 
SCR).\50\ The company stated that for SO2 controls, the 
updated analysis determined that the cost-effectiveness of DSI is 
$10,920/ton (as compared to EPA's $3,940/ton).\51\ Commenters stated 
that the proposed FIP failed to consider that the installation of DSI 
would most likely require upgrades to the existing particulate controls 
to achieve the SO2 reductions that EPA evaluated and that 
EPA relied on the outdated and inaccurate CCM to develop these 
estimates.
---------------------------------------------------------------------------

    \49\ See NOX Control Update to PPL Montana's J.E. 
Corette Generating Station BART Report, September 2011, Prepared for 
PPL Montana, LLC by TRC, at ES-1 (``NOX Control 
Update''); SO2 Control Update to PPL Montana's J.E. 
Corette Generating Station BART Report, August 2011, Prepared for 
PPL Montana, LLC by TRC, at ES-1 (``SO2 Control 
Update'').
    \50\ NOX Control Update, at ES-3.
    \51\ SO2 Control Update, at 14.
---------------------------------------------------------------------------

    Response: We disagree. See our response to similar comments made by 
PPL in regard to cost analyses for Colstrip Units 1 and 2. PPL's cost 
estimates for Corette included many of the same incorrect methods and 
assumptions that the company used when developing cost estimates for 
Colstrip Units 1 and 2. In particular, PPL used unsupported retrofit 
factors that were well in excess of the range described in the IPM 
documentation.
    Also, we disagree that installation of DSI would most likely 
require upgrades to the existing particulate controls to achieve the 
SO2 reductions that EPA evaluated. In fact, DSI using trona 
would ``typically either improve performance or have little impact, 
even at high injection rates.'' \52\ It would not require the 
replacement of the existing ESP with a new baghouse as reflected in 
PPL's cost effectiveness estimate of $10,920/ton.\53\ Therefore, we 
find that EPA's cost estimate of $3,490 is accurate.
---------------------------------------------------------------------------

    \52\ United Conveyer Corporation Dry Sorbent Injection FAQ 
(http://unitedconveyor.com/dsi_systems/).
    \53\ Ref 2: SO2 Control Update to PPL Montana's J.E. 
Corette Generating Station BART Report, Prepared for PPL Montana, 
LLC, by TRC, August 2011, p. ES-2.
---------------------------------------------------------------------------

    Comment: Commenters stated that our proposed SO2 and 
NOX emission limits for Corette were flawed. One commenter 
stated that EPA must increase the limits to no less than 0.81 lb/MMBtu 
for SO2 and 0.46 lb/MMBtu for NOX in order to 
account for compliance over a 30-day rolling average. By contrast, 
another commenter stated that our proposed emission limits were too 
high and would actually result in increased emissions.
    Response: Based on these comments, we have reassessed our 
SO2 and NOX emission limits for Corette. As we 
have not prescribed any additional controls for Corette, the emission 
limits should reflect emission rates currently being achieved with 
existing controls. In order to establish appropriate emission limits, 
we have conducted a statistical analysis of the monthly emissions data 
contained in the CAMD emissions system. For the period 2000-2010, the 
99th percentile monthly SO2 emission rate was 0.548 lb/
MMbtu. Similarly, the 99th percentile monthly NOX emission 
rate was 0.335 lb/MMBtu. In our final action, we are establishing 
emission limits slightly above these 99th percentile emission rates in 
order to allow a sufficient margin for compliance. This is because the 
emission limits must apply at all times,

[[Page 57894]]

including during startup, shutdown, and malfunction. The revised 
emission rates are 0.57 lb/MMBtu for SO2 and 0.35 lb/MMBtu 
for NOX, both on a 30-day rolling average. We have revised 
the emission limits for Corette contained in section 52.1396(c)(1) 
accordingly. Our complete analysis of SO2 and NOX 
emission limits for Corette can be found in the docket.0.5480.3350.57 
We have addressed the emission limit for PM at Corette in a separate 
response to comments.
    Comment: PPL stated that EPA's PM emission limit for Corette was 
flawed. PPL noted that over the past five years, stack test results 
have shown that PM emissions have ranged from 0.059 lb/MMBtu to 0.252 
lb/MMBtu. PPL stated that an emission limit of 0.30 lb/MMBtu would be 
necessary to account for a 30-day rolling average.
    Response: We agree, in part. In our proposed rule, we incorrectly 
specified a PM emission limit of 0.10 lb/MMBtu on a 30-day rolling 
average. In consideration of the stack test data provided by PPL, we 
have determined that that a limit of 0.26 lb/MMBtu is more appropriate. 
In addition, and as discussed in response to a similar comment made by 
PPL in regard to Colstrip, we find that it is not feasible to require 
compliance with this emission limit on a 30-day rolling average. Again, 
this is because compliance is shown using stack methods such as Method 
5 and 5B. These stack test methods typically consist of three sampling 
runs of at least 120 minutes in duration, and are not long-term 
continuous measurements. As such, it is not possible to average the 
emissions over 30 days or a year.
    Accordingly, we are revising our FIP to reflect a PM emission limit 
for Corette of 0.26 lb/MMBtu. We are also removing the 30-day averaging 
period requirement for the PM emission limit at Corette. More 
specifically, we are revising section 52.1396(c)(1) to clarify that 
emission limits for NOX and SO2, but not PM, 
shall apply on a 30-day rolling average. Note that we are retaining the 
requirement that compliance with the PM emission limit shall be 
monitored in accordance with the CAM plan.
    As we are not requiring that the PM emission limit applies on a 30-
day rolling average, PPL's suggestion that the emission limit be 
increased to 0.30 lb/MMBtu is no longer relevant.
    Comment: The USFWS commented that there are at least two other 
similarly-sized installations implementing lime spray drying (LSD) for 
SO2 control that justify the positions taken by EPA in the 
proposed BART determination. USFWS stated that in justifying emission 
limits of small units burning clean coal, Newmont Nevada is a 200 MW 
plant that attains a 30-day rolling average 0.065 lb/MMBtu 
SO2 emission limit with an SO2 control efficiency 
of 93.1% and that capital cost of LSD units is corroborated by Great 
River Energy's 188 MW Stanton 1 plant costing $79,514,000.
    Response: We acknowledge that the USFWS has provided information 
from two other similarly-sized installations which are implementing LSD 
for SO2 corroborating our LSD cost estimates for Corette. 
However, as noted in our proposed rule, the cost of controls is not 
justified by the visibility improvement (0.253 deciview).
    Comment: The USFWS stated that the capital costs proposed by EPA 
for dry sorbent injection (DSI) and LSD should be considered as 
maximums, because the costs should only decrease due to significant 
curtailment of construction of air pollution control devices during the 
economic downturn and cancellation or postponement of many coal burning 
electrical generation units. The USFWS stated that quantified estimates 
of the decreases could provide for firm reductions in the capital cost 
estimates, but it is agreed that they would be difficult to affirm with 
confidence at this time.
    Response: We agree that any changes in cost associated with 
economic downturn would be difficult to affirm with confidence at this 
time.
    Comment: The USFWS stated that the paragraph following Table 123 
states that EPA considers $4,659 per ton of SO2 emissions 
reduction using DSI as reasonable, but that $5,442 per ton for LSD is 
not cost effective. The USFWS stated that other proposed SO2 
BART determinations resulting in cost efficiency in the range of 
Corette include PacifiCorp's Dave Johnston, WY-$4,743; Northshore 
Mining's Silver Bay Power, MN-$7,309 and Xcel Energy's Taconite Harbor, 
MN-$5,300 and as stated above, the capital cost of an LSD unit on Great 
River Energy's 188 MW Stanton 1 plant is $79,514,000. USFWS 
stated that such a total capital cost incorporated as the cost of LSD 
at Corette would result in a cost per ton of SO2 removed of 
$4,891 and that the LSD alternative might then also be considered by 
EPA as being cost effective along with DSI.
    Response: We disagree. We continue to find that the cost of LSD for 
Corette is not justified by the visibility improvement. Moreover, the 
capital cost that we estimated for LSD is specific to Corette, and we 
see no reason to supplant that cost with costs from Taconite Harbor or 
other individual facilities.
    Comment: The USFWS stated that regarding the cost-effectiveness of 
visibility improvement for SO2 controls, the second 
paragraph after Table 123 in the draft proposed BART determination 
states, `` * * * the cost of controls is not justified by the 
visibility improvement'' and that this proposed conclusion warrants 
further scrutiny. The USFWS stated that implementation of the DSI 
alternative results in a 0.176 deciview improvement at Washakie WA, the 
highest impacted Class I area, at a cost of $3.4 million per deciview 
of improvement and that this is a very reasonable cost for visibility 
improvement. The UFWS stated that the cost of visibility improvement 
for SO2 controls proposed in other BART determinations for a 
single most-impacted Class I area include: Colorado Springs Utilities, 
Martin Drake, CO-$49.9 million/deciview; PacifiCorp, Wyodak, WY-$44.7 
million/deciview; PacifiCorp, Jim Bridger, WY-$37.1 million/deciview; 
PG&E, Boardman, OR-$35.2 million/deciview; and Dominion, Brayton Point, 
MA-$33.9 million/deciview; Northshore Mining, Silver Bay Power, MN-
$26.2 million/deciview; Dominion, Salem Harbor, MA-$25.1 million/
deciview; Great River Energy, Stanton 1, ND-$21.9 million/
deciview; PacifiCorp, Naughton, WY-$18.2 million/deciview; PacifiCorp, 
Dave Johnson, WY-$16.7 million/deciview. The USFWS stated that the 
conclusion from the above is that since the cost per ton of 
SO2 removal and the cost per deciview of visibility 
improvement are both reasonable, DSI should be considered as a feasible 
and cost-effective SO2 control alternative and be accepted 
as BART for the PPL Montana, J.E. Corette Generating Station.
    Response: We disagree. The total annual cost of DSI for Corette, as 
cited in our proposed rule was $5,363,896, while the greatest 
visibility improvement was 0.176 deciview (Washakie WA). This results 
in cost of $30 million per deciview, not $3.4 million per deciview. We 
continue to find that the cost of LSD for Corette is not justified by 
the visibility improvement.
    Comment: The USFWS commented that Table 110 states the visibility 
improvement associated with each of the three NOX control 
alternatives and by dividing respective Total Annual Costs by their 
visibility improvements, they result in cost per deciview of visibility 
improvement from $16.7 million to $17.8 million at the Washakie WA, the 
highest impacted Class I area.

[[Page 57895]]

The USFWS stated that when these values are compared to other single 
Class I area impacts for some other NOX BART proposals as 
summarized below, it would indicate that they each could be considered 
as reasonable. The USFWS stated that when total annual cost for each of 
the three NOX control alternatives is divided by the 
respective visibility improvement for all affected Class I areas (as 
discussed above for SO2) they result in cost per deciview of 
visibility improvement from $4.7 million to $5.0 million, which is a 
very reasonable visibility cost. USFWS stated that since the cost per 
ton of NOX removal and the cost per deciview of visibility 
improvement are both reasonable, at least the Separated Over-fire Air 
(SOFA)-only or, preferably SOFA plus Selective Non-Catalytic Reduction 
(SNCR) should definitely be considered as feasible and cost-effective 
NOX control alternatives and be accepted as BART for 
Corette.
    Response: We disagree that SOFA or SOFA+SNCR should be accepted as 
BART for Corette. The BART Guidelines require that cost effectiveness 
be calculated in terms of annualized dollars per ton of pollutant 
removed, or $/ton. 70 FR 739167. The BART Guidelines list the $/
deciview ratio as an additional cost effectiveness metric that can be 
employed along with $/ton for use in a BART evaluation. However, we did 
not use this metric for the reasons that were explained in other 
responses. As we stated in the proposed FIP, we weighed costs against 
the anticipated visibility impacts and we explained that any of the 
control options would have a positive impact on visibility; however, 
the cost of controls was not justified by the visibility improvement. 
As we have explained elsewhere, in our proposal, we considered the 
visibility improvement at all Class I areas within 300 km of the 
subject BART unit.
    In addition, we note that the UFWS seems to have miscalculated the 
dollars per deciview values for the NOX control options.
    Comment: The USFS stated the BART determinations for Corette are 
not consistent with previous BART demonstrations that have been made 
for other facilities in Montana, as well as with decisions EPA has 
approved in other SIPs. And that EPA has identified control options for 
both NOX and SO2 that are technically feasible 
and cost effective. USFS stated that it is their understanding that EPA 
has also determined that the visibility improvement does not justify 
the cost of the additional controls.
    Response: We disagree. As the commenter has noted, we rejected 
additional controls for Corette since the visibility improvement does 
not justify the cost of controls. Moreover, the USFWS has not 
identified how this is inconsistent with other BART determinations in 
Montana or elsewhere.
    Comment: WEG stated that EPA arbitrarily rejected requiring SCR as 
BART for NOX emissions from Corette and that we stated in 
the proposed FIP that the control technology would be cost-effective 
and achieve greater visibility benefits--in favor of no additional 
controls. WEG stated that the EPA's proposed BART determination is 
inconsistent with the CAA and the Agency's own record. WEG stated that 
that under the factors required to be considered by EPA in determining 
BART under the CAA, SCR would constitute BART. WEG stated that EPA 
found that SCR for Corette would not be cost-prohibitive and that the 
Agency also identified no energy and nonair quality impacts that would 
mitigate against the use of SCR, or any remaining useful life issues 
that would preclude the use of SCR. WEG stated that with regard to 
visibility improvement, the EPA further found that SCR, as opposed to 
doing nothing, would achieve greater visibility improvements and that 
given that SCR represents ``the best system of continuous emission 
control technology available'' (40 CFR 51.308(e)(1)(ii)), there appears 
to be no reason to dismiss SCR as BART for Corette. WEG stated that the 
EPA asserted that SCR for Corette ``is not justified by the visibility 
improvement.'' Yet, the proposed FIP indicates that with the use of 
SCR, visibility improvements in the most impacted Class I area, the 
Washakie WA, would be 264%, an enormous improvement from current 
conditions. WEG stated that SCR would have a visibility improvement of 
0.264 deciview and that SCR would reduce visibility impairment at seven 
different Class I areas, and that SCR would cumulatively improve 
visibility amongst the seven impacted Class I areas by 0.939 deciview. 
77 FR 24042.
    WEG stated that such cumulative visibility improvements do not 
appear to be unreasonable, but that in this case, the EPA appears to 
believe that the level of visibility improvement is not significant 
enough to justify the use of SCR. WEG stated that the proposed FIP 
provides no information or analysis to indicate that EPA's belief is 
not anything more than an arbitrary claim and that there is no 
explanation as to why the EPA believed the level of improvement with 
the use of SCR was somehow discountable or insignificant. WEG stated 
that the EPA's logic is further belied by the fact that the FIP will 
fail to achieve meaningful reasonable progress in attaining natural 
visibility conditions in Class I areas in Montana and that given the 
prospect of such dismal progress in achieving natural visibility, it is 
reasonable to presume that any improvement in visibility, no matter how 
small, would be significant. WEG stated that the EPA failed to provide 
any information or analysis in the proposed FIP or the supporting 
record suggesting otherwise. WEG stated that although it is true that 
EPA is allowed to consider the degree in improvement in visibility in 
determining BART, there is no indication that this factor could be 
interpreted to allow the Agency to make arbitrary determinations that a 
264% improvement in visibility under a plan that already contains 
unreasonable RPGs is insignificant or otherwise not worthy of 
regulatory action under the CAA's regional haze program.
    Response: We disagree. We did not arbitrarily reject SCR. Our 
proposal clearly laid out the bases for our proposed BART determination 
for NOX for Corette. Our regulations define BART as an 
emission limitation based on the degree of reduction achievable through 
the application of the best system of continuous emission reduction for 
each pollutant which is emitted by an existing stationary facility. The 
emission limitation must be established, on a case-by-case basis, 
taking into consideration the technology available, the costs of 
compliance, the energy and nonair quality environmental impacts of 
compliance, any pollution control equipment in use or in existence at 
the source, the remaining useful life of the source, and the degree of 
improvement in visibility which may reasonably be anticipated to result 
from the use of such technology. The BART analysis identifies the best 
system of continuous emission reduction taking into account:
    (1) The available retrofit control options, (2) Any pollution 
control equipment in use at the source (which affects the availability 
of options and their impacts), (3) The costs of compliance with control 
options, (4) The remaining useful life of the facility, (5) The energy 
and nonair quality environmental impacts of control options (6) The 
visibility impacts analysis. 70 FR 39163.
    As the final BART Guidelines explain, both the 2001 proposal and 
the 2004 reproposal requested comments on two options for evaluating 
the ranked options. The first option was similar to

[[Page 57896]]

the process that WEG implies should have been followed, where the most 
stringent control option must be chosen as long as it does not impose 
unreasonable costs of compliance or energy and nonair quality 
environmental impacts would justify selection of an alternative control 
option. 70 FR 39130. The second option was:

    An alternative decision-making approach that would not begin 
with an evaluation of the most stringent control option. For 
example, States could choose to begin the BART determination process 
by evaluating the least stringent technically feasible control 
option or by evaluating an intermediate control option drawn from 
the range of technically feasible control alternatives. Under this 
approach, States would then consider the additional emissions 
reductions, costs, and other effects (if any) of successively more 
stringent control options. Under such an approach, States would 
still be required to (1) display all of the options and identify the 
average and incremental costs of each option; (2) consider the 
energy and nonair quality environmental impacts of each option; and 
(3) provide a justification for adopting the technology selected as 
the ``best'' level of control, including an explanation of its 
decision to reject the other control technologies identified in the 
BART determination.

    In the final guidelines, EPA ``decided that States should retain 
the discretion to evaluate control options in whatever order they 
choose, so long as the State explains its analysis of the CAA 
factors.'' 70 FR 39130. The BART Guidelines state that we ``have 
discretion to determine the order in which you should evaluate control 
options for BART'' and that we ``should provide a justification for 
adopting the technology that you select as the ``best'' level of 
control, including an explanation of the CAA factors that led you to 
choose that option over other control levels.'' 70 FR 39170.
    We explained our analysis of the five factors and explained that 
the CAA factors that led to our decision were cost-effectiveness and 
visibility improvement. The cost-effectiveness of SOFA + SCR was 
determined to be $4,491/ton and the visibility improvement at the most 
impacted Class I area, Washakie WA, was 0.264 deciview. The impact at 
additional Class I areas was shown in Tables 123 and 124. 77 FR 24042. 
When we weighed the costs against the anticipated visibility 
improvement for Corette the cost of controls was not justified by the 
limited visibility improvement. 77 FR 24043.
    With regard to WEG's claim that SCR would result in a visibility 
improvement of 264%, WEG used a fundamentally flawed approach to 
calculate visibility improvements. Using WEG's approach, a 0.1 deciview 
change would produce a 1000% improvement in visibility compared to a 
0.01 deciview change. In fact, the change would be 0.09 deciview or 
about 1% relative to natural visibility conditions. The approach that 
WEG used to calculate percent visibility improvement is mathematically 
incorrect. WEG compared a 0.264 deciview change to a zero deciview 
change and arbitrarily called this a 264% improvement in visibility. To 
get a more accurate estimate, you can use the rule of thumb that 0.5 
deciview is approximately equivalent to a 5% change in perceived 
visibility. The 0.264 deciview change would be approximately a 2.6% 
improvement in visibility relative to natural visibility conditions. 
WEG makes the same mistake on page 3 in the comment on Colstrip where 
they state: ``with the use of SCR, visibility improvements in the most 
impacted Class I areas would be around 50% greater than with the use of 
SNCR.'' Here they compared 0.784 deciview with SCR to 0.518 deciview 
with SNCR, and concluded that SCR provides a 50% visibility improvement 
over SNCR. Again, using the rule of thumb, this would be about a 2.6% 
difference in perceived visibility between SCR and SNCR relative to 
natural visibility conditions.
    The BART Guidelines state that to make the net visibility 
improvement determination you should, ``assess the visibility 
improvement based on the modeled change in visibility impacts for the 
pre-control and post-control emission scenarios. You have flexibility 
to assess visibility improvements due to BART controls by one or more 
methods. You may consider the frequency, magnitude, and duration 
components of impairment.'' 70 FR 39170. The BART Guidelines also state 
that, ``Comparison thresholds can be used in a number of ways in 
evaluating visibility improvement (e.g. the number of days or hours 
that the threshold was exceeded, a single threshold for determining 
whether a change in impacts is significant, or a threshold representing 
an x percent change in improvement.'' 70 FR 39170. Our proposal shows 
the baseline visibility impact in deciviews, the visibility improvement 
in deciviews, the number of Class I areas impacted within 300 km, and 
fewer days impacted more than 0.5 deciview in Tables 123 and 124 and 
these are more appropriate metrics for evaluating visibility impact.
    We disagree with WEG's statement that the FIP will fail to achieve 
meaningful reasonable progress in attaining natural visibility 
conditions in Class I areas in Montana and that given the prospect of 
such dismal progress in achieving natural visibility, it is reasonable 
to presume that any improvement in visibility, no matter how small, 
would be significant. We have explained in other responses that 40 CFR 
51.308(d)(1)(ii) states that, ``if the State establishes a reasonable 
progress goal that provides for a slower rate of improvement in 
visibility that the rate that would be needed to attain natural 
conditions by 2064, the State must demonstrate, based on the factors in 
paragraph (d)(1)(i)(A) of this section, that the rate of progress for 
the implementation plan to attain natural conditions by 2064 is not 
reasonable; and that the progress goal adopted by the State is 
reasonable. The State must provide the public for review as part of its 
implementation plan an assessment of the number of years it would take 
to attain natural conditions if visibility improvement continues at the 
rate of progress selected by the State as reasonable.'' We explained in 
other responses how we have met those requirements.

I. Comments on Reasonable Progress and Long Term Strategy

    Comment: A commenter stated that based on the WRAP emissions 
inventory and air quality modeling, EPA proposed reasonable progress 
goals for the 20% worst visibility days for the Montana Class I areas 
that are significantly less (16-51%) than the uniform rate of progress 
by 2018 and that no Montana Class I area is projected to achieve 
natural visibility conditions by 2064. The commenter stated that EPA 
projects that, at best, the national goal will not be met for 135 years 
at Cabinet Mountains WA and, at worst, for 437 years at the Medicine 
Lake WA.
    The commenter stated that the WRAP inventory indicates that point 
sources contribute 71% of Montana's total SO2 emissions, yet 
point source SO2 emissions in Montana are projected to be 
reduced by less than 1% by 2018 (this includes SO2 
reductions for BART for Colstrip Units 1 and 2). This change in point 
source emissions inventory is considerably less than projected by other 
states in Region 8, yet EPA has determined that no additional 
SO2 controls are reasonable. The commenter stated that the 
WRAP inventory projects that point source NOX emissions 
would be reduced by 3% (23,000 tons per year), primarily due to 
estimated NOX reductions at Colstrip and that EPA's RP 
analyses determined that $282 per ton for NOX reduction at 
Devon Energy was cost effective, but NOX controls for all

[[Page 57897]]

other facilities were not cost effective. Several controls were below 
the cost of $4,659 for SO2 controls at Corette Generating 
Station that EPA determined were cost effective for BART. Given the 
lack of progress in improving visibility at the Class I areas, EPA 
needs to reconsider the cost effectiveness of point source 
SO2 and NOX controls.
    Response: We disagree that we should reconsider the cost 
effectiveness of point source controls given the lack of progress in 
improving visibility at the Class I areas. In determining the measures 
necessary to make reasonable progress and in selecting RPGs for 
mandatory Class I areas within Montana, we took into account the 
following four factors into consideration: costs of compliance; time 
necessary for compliance; energy and nonair quality environmental 
impacts of compliance; and remaining useful life of any potentially 
affected sources. CAA section 169A(g)(1) and 40 CFR 51.308(d)(1)(i)(A). 
In the FIP, we demonstrated how these four factors were considered. 40 
CFR 51.308(d)(1)(ii) allows for a slower rate of improvement in 
visibility than the URP, as long as it is demonstrated that based on 
these four factors, it is not reasonable to achieve the URP and that 
the selected RPG is reasonable. CAA section 169A(g)(1) and 40 CFR 
51.308(d)(1)(i)(A). We respond to specific critiques of our four-factor 
analyses elsewhere. To the extent that the commenter is stating that 
cost-effectiveness is a fixed value and must be the same whether a 
source is subject to BART or RP, we disagree. While the Regional Haze 
Rule may allow us to establish a bright line for some of the factors 
such as cost-effectiveness and visibility, we are not required to do 
so, and have not done so for this action.
    Comment: A commenter stated that oil and gas development has 
increased markedly in Montana and neighboring states since the initial 
inventory projections provided by the WRAP in 2007 and that EPA should 
compare the most recent (Phase III) oil and gas emissions inventory to 
that used in the WRAP source apportionment modeling and discuss the 
implications of future oil and gas development for visibility at 
Montana Class I areas.
    Response: We disagree that we should reevaluate the oil and gas 
inventory and discuss the implications of future oil and gas 
development for visibility at Montana Class I areas at this time. 40 
CFR 51.308(d)(3)(iii) requires us to document the technical basis, 
including modeling, monitoring and emissions information on which we 
relied. It also requires that we identify the baseline emission 
inventory on which our strategies are based. As stated in the proposal, 
an emissions inventory for each pollutant was developed by WRAP for 
Montana and these inventories were used as inputs to photochemical 
modeling that was used to determine the 2018 reasonable progress goal. 
77 FR 24047 and 77 FR 24054. 40 CFR 51.308(d)(3)(iii) allows us to rely 
on the technical analysis developed by the WRAP, which we have done. We 
recognize that emission inventories are dynamic, but at this time it is 
not necessary to reevaluate the emission inventories. The Regional Haze 
Rule recognizes the need for periodic progress evaluation and requires 
progress reports to be submitted every five years. 40 CFR 51.308(g)(4) 
requires this report to include, ``[A]an analysis tracking the change 
over the past five years in emissions of pollutants contributing to 
visibility impairment from all sources and activities within the 
state.'' As we explained in our proposal, we will update the statewide 
emissions inventories periodically or as necessary and review emissions 
information from other states and future emissions projections.
    Comment: MDEQ stated that EPA fails to consider the potential 
benefits of the Mercury Air Toxics Standard, the new NOX and 
SO2 NAAQS, the forthcoming Boiler MACT, and other rules that 
will significantly impact PM2.5, SO2 and 
NO2 emissions in its LTS.
    Response: We are sensitive to the challenges of coordinating 
compliance with a variety of rules. However, to the extent that MDEQ is 
implying that we should have considered the potential benefits of 
possible future regulations in our LTS, we disagree. As explained in 
our proposed FIP, in order to establish RPGs for the Class I areas in 
Montana and to determine the controls needed for the LTS, we followed 
the process established in the Regional Haze Rule. The anticipated 
visibility improvement in 2018 in all Montana Class I areas accounting 
for all existing enforceable federal and state regulations already in 
place was considered. 77 FR 24055. With regard to regulations that are 
not yet final, we cannot speculate on unknown reductions from 
anticipated future federal or state regulations prior to those actions 
completing the full regulatory process. None of the Montana sources 
have notified us that they will be reducing emissions as a result of 
future regulation and we have no basis for estimating what those 
emissions may be. Without an enforceable commitment, we cannot assume 
that additional reductions will be achieved and we cannot account for 
them in our LTS for the Regional Haze FIP. MDEQ has not provided 
information to indicate that anything in the Regional Haze FIP will 
interfere with the requirements of other regulations. In fact, where 
additional controls are required, we would expect that the lower 
emission limit would make it easier to comply with future regulations 
that also require lower emission limits. We note that the Regional Haze 
FIP requires compliance with a specific emission limit and not 
necessarily the installation of a specific control technology and that 
sources have a full five years after the finalization of the FIP to 
comply with any emission limit that would require the installation of 
additional control technology.
    Comment: MDEQ suggested that we include all smoke emissions from 
open burning and wildfires in the natural background estimates and 
recalculate URP and RPGs in each of the State's Class I areas with 
these adjusted background levels. MDEQ perceived fire to be the major 
contributing factor to the State's visibility impairment, and claimed 
that EPA does not make a realistic allowance for smoke contributions to 
haze in Montana.
    Response: We agree that industrial facilities are not the only 
causes of haze, but we disagree that we should make adjustments to the 
inventories, the URP, or the RPGs. Our action considered the many 
contributors to haze including industrial facilities. It is not 
appropriate to consider open burning as natural background because open 
burning is anthropogenic. In our proposal, the emissions inventory 
appropriately included natural (non-anthropogenic) wildfire and 
anthropogenic sources such as open burning. 77 FR 24093. In developing 
a LTS, 40 CFR 51.308(d)(3)(iv) requires us to consider all 
anthropogenic sources. More specifically, 40 CFR 51.308(d)(3)(v)(E) 
requires the LTS to address smoke management techniques for 
agricultural and forestry management techniques. We note that our 
proposed action also proposed to approve the revisions to the paragraph 
titled ``Smoke Management'' of Title 17, Chapter 8, Subchapter 6, Open 
Burning as meeting the requirement in 40 CFR 308(d)(3)(v)(E) because 
the plan control emissions from these sources by requiring BACT and 
takes into consideration the visibility impacts on mandatory Class I 
areas.
    Regardless of the contribution from smoke emissions, 40 CFR 
51.308(d)(3)(iv) states, ``The State must identify all anthropogenic 
sources of visibility impairment considered by the State in developing 
its long-term strategy. The State should consider major and minor 
stationary sources,

[[Page 57898]]

mobile sources, and area sources.'' In this case, we acted in the place 
of Montana and were required to abide by the same requirement to 
consider point sources. 40 CFR 51.308(d)(1)(ii) states that, ``if the 
State establishes a reasonable progress goal that provides for a slower 
rate of improvement in visibility that the rate that would be needed to 
attain natural conditions by 2064, the State must demonstrate, based on 
the factors in paragraph (d)(1)(i)(A) of this section, that the rate of 
progress for the implementation plan to attain natural conditions by 
2064 is not reasonable; and that the progress goal adopted by the State 
is reasonable. The State must provide the public for review as part of 
its implementation plan an assessment of the number of years it would 
take to attain natural conditions if visibility improvement continues 
at the rate of progress selected by the State as reasonable.'' In this 
case, we are acting in the place of Montana. In determining the 
measures necessary to make reasonable progress and in selecting RPGs 
for mandatory Class I areas within Montana, we evaluated major and 
minor point sources according to the four factors required by 40 CFR 
51.308 (d)(1)(i)(A) (costs of compliance; time necessary for 
compliance; energy and nonair quality environmental impacts of 
compliance; and remaining useful life of any potentially affected 
sources CAA section 169A(g)(1) and 40 CFR 51.308(d)(1)(i)(A)). In 
addition, 40 CFR 51.308(e) requires states to make a BART determination 
for each BART-eligible source and in that determination, the state must 
consider the five statutory factors.
    The requirements of 40 CFR 51.308(d)(3)(iv) and 40 CFR 51.308(e) 
are not dependent on the showing of a certain amount of impairment from 
point sources.
    EPA recognized that variability in natural sources of visibility 
impairment causes variability in natural haze levels as described in 
its ``Guidance for Estimating Natural Visibility Conditions Under the 
Regional Haze Rule.'' \54\ The preamble to the BART Guidelines (70 FR 
39124) describes an approach used to measure progress toward natural 
visibility in Mandatory Class I areas that includes a URP toward 
natural conditions for the 20% worst days and no degradation of 
visibility on the 20% best days. The use of the 20% worst natural 
conditions days in the calculation of the URP takes into consideration 
visibility impairment from wild fires, windblown dust and other natural 
sources of haze.\55\ 70 FR 39124. The Guidance for Estimating Natural 
Visibility Conditions also discusses the use of the 20% best and worst 
estimates of natural visibility, provides for revisions to these 
estimates as better data becomes available, and discusses possible 
approaches for refining natural conditions estimates.\56\
---------------------------------------------------------------------------

    \54\ Guidance for Estimating Natural Visibility Conditions Under 
the Regional Haze Rule, U.S. Environmental Protection Agency, 
September 2003. http://www.epa.gov/ttncaaa1/t1/memoranda/rh_envcurhr_gd.pdf">http://www.epa.gov/ttncaaa1/t1/memoranda/rh_envcurhr_gd.pdf, page 1-1 (Guidance for Estimating Natural 
visibility Conditions). The guidance states that, ``Natural 
visibility conditions represent the long-term degree of visibility 
that is estimated to exist in a given mandatory Federal Class I area 
in the absence of human-caused impairment. It is recognized that 
natural visibility conditions are not constant, but rather they vary 
with changing natural processes (e.g., windblown dust, fire, 
volcanic activity, biogenic emissions). Specific natural events can 
lead to high short-term concentrations of particulate matter and its 
precursors. However, for the purpose of this guidance and 
implementation of the regional haze program, natural visibility 
conditions represents a long-term average condition analogous to the 
5-year average best- and worst-day conditions that are tracked under 
the regional haze program.''
    \55\ The preamble further stated that, ``with each subsequent 
SIP revision, the estimates of natural conditions for each mandatory 
Federal Class I area may be reviewed and revised as appropriate as 
the technical basis for estimates of natural conditions improve.''
    \56\ Guidance for Estimating Natural Visibility Conditions, p.3-
1 to 3-4.
---------------------------------------------------------------------------

    For the evaluation of visibility impacts for BART sources, EPA 
recommended the use of the natural visibility baseline for the 20% best 
days for comparison to the ``cause or contribute'' applicability 
thresholds. This estimated baseline is reasonably conservative and 
consistent with the goal of attaining natural visibility conditions. 
While EPA recognizes that there are natural sources of haze, the use of 
the 20% worst natural visibility days is inappropriate for the ``cause 
or contribute'' applicability thresholds. For example, if BART source 
visibility impacts were evaluated in comparison to days with very poor 
natural visibility resulting from nearby wild fires or dust storms, the 
BART source impacts would be significantly reduced relative to these 
poor natural visibility conditions and would not be protective of 
natural visibility on the best 20% days.
    Comment: MDEQ insisted that visibility issues in the Western U.S. 
are less stationary source driven than in the Eastern U.S., and that 
greater understanding of this difference has developed since Congress 
passed the Visibility Protection Act of 1977 and the visibility statute 
of the CAA Amendments of 1990.
    Response: To the extent that MDEQ is implying that we are not 
required to analyze controls for stationary sources, we disagree. As 
explained in other responses, 40 CFR 51.308(d)(3)(iv) requires us to 
identify all anthropogenic sources of visibility impairment considered 
in developing our long term strategy. It specifically states that we 
should consider major and minor stationary sources, mobile sources, and 
area sources. Please see the language of 40 CFR 51.308(e) in the 
response to the previous comment. The requirements of 40 CFR 
51.308(d)(3)(iv) and 40 CFR 51.308(e) are not dependant on the showing 
of a certain amount of impairment from point sources.
    Comment: A commenter stated that BART sources such as Corette 
should also be considered under reasonable progress and that this would 
be consistent with actions EPA has approved in other SIPs. The 
commenter stated that EPA is using visibility improvement as measured 
by Q over D values as an indirect measure of the benefit of additional 
controls under reasonable progress and that it is their understanding 
that this is not supported under the Regional Haze Rule as reasonable 
progress decisions do not consider visibility improvement. The 
commenter requested that control options considered technologically 
feasible and cost effective under BART also be considered under 
reasonable progress.
    Response: We disagree that BART sources need to be re-evaluated for 
the purposes of reasonable progress and that, under the Regional Haze 
Rule, reasonable progress determinations may not consider visibility 
improvement. Our RP Guidance states, ``Since the BART analysis is 
based, in part, on an assessment of many of the same factors that must 
be addressed in establishing the RPG, it is reasonable to conclude that 
any control requirements imposed in the BART determination also satisfy 
the RPG-related requirements for source review in the first RPG 
planning period. Hence you may conclude that no additional emissions 
controls are necessary for these sources in the first planning 
period.'' \57\ The EPA has concluded that, based on the similarity of 
many of the same factors for both BART and reasonable progress, that no 
additional emissions controls are necessary for BART sources for this 
planning period. The commenter has given us no basis to change that 
conclusion: Regardless of whether any states have chosen to reevaluate 
BART sources for reasonable progress, the

[[Page 57899]]

Regional Haze Rule does not require states to do so. With regard to the 
statement about using visibility improvement to evaluate additional 
controls under reasonable progress, EPA's reasonable progress guidance 
states: ``In determining reasonable progress, CAA section 169A(g)(1) 
requires States to take into consideration a number of factors. 
However, you have flexibility in how to take into consideration these 
statutory factors and any other factors that you have determined to be 
relevant.'' \58\ The potential reduction in quantity over distance (Q/
D) is a factor that we consider to be relevant because the goal of the 
Regional Haze Rule is to improve visibility. The commenter has not 
cited any authority supporting the position that visibility 
improvements may not be considered in reasonable progress 
determinations and therefore has given us no basis to change our use of 
this factor.
---------------------------------------------------------------------------

    \57\ Guidance for Setting Reasonable Progress Goals Under the 
Regional Haze Program, U.S. Environmental Protection Agency, 
(``Reasonable Progress Guidance'') (June 1, 2007) p.4-2--4-3.
    \58\ Reasonable Progress Guidance, p.5-1.
---------------------------------------------------------------------------

    Comment: A commenter stated that the proposal fails to achieve 
reasonable progress. The commenter explained that the proposal will 
leave visibility in the parks and WAs that are affected by Montana 
sources impaired for hundreds of years into the future, nonetheless, we 
propose no additional emission reductions from Montana's stationary 
sources.
    Response: We disagree that the FIP fails to achieve reasonable 
progress. 40 CFR 51.308(d)(1)(ii) states:

    If the State establishes a reasonable progress goal that 
provides for a slower rate of improvement in visibility than the 
rate that would be needed to attain natural conditions by 2064, the 
State must demonstrate, based on the factors in paragraph 
(d)(1)(i)(A) of this section, that the rate of progress for the 
implementation plan to attain natural conditions by 2064 is not 
reasonable; and that the progress goal adopted by the State is 
reasonable. The State must provide the public for review as part of 
its implementation plan an assessment of the number of years it 
would take to attain natural conditions if visibility improvement 
continues at the rate of progress selected by the State as 
reasonable.

    In determining the measures necessary to make reasonable progress 
and in selecting RPGs for mandatory Class I areas within Montana, we 
took into account the following four factors into consideration: Costs 
of compliance; time necessary for compliance; energy and nonair quality 
environmental impacts of compliance; and remaining useful life of any 
potentially affected sources. CAA section 169A(g)(1) and 40 CFR 
51.308(d)(1)(i)(A). In the FIP, we demonstrated how these four factors 
were considered and we also provided, in Table 197, an assessment of 
the number of years it would take to attain natural conditions if 
visibility improvement continues at the rate of progress that we 
selected was reasonable. We respond to specific critiques of our four-
factor analyses elsewhere.
    Comment: A commenter stated that EPA failed to evaluate controls on 
all BART-subject sources to meet reasonable progress requirements and 
that EPA stated that the BART analyses for these facilities are similar 
to the requisite reasonable progress analysis. 77 FR at 24059. The 
commenter stated that EPA has ensured that Montana will not achieve 
reasonable progress toward natural visibility conditions at Class I 
areas affected by Colstrip and Corette and that EPA's approach is 
flawed legally and factually. The commenter stated that EPA's approach 
fails to distinguish between the purposes of BART and the long-term 
strategy under the Regional Haze Rule and that while both are 
mechanisms to help states achieve reasonable progress, BART is applied 
to a given source--for the purpose of eliminating or reducing 
visibility impairment caused or contributed to by that source. 42 
U.S.C. section 7491(b)(2)(A). The commenter stated that rather than 
focusing on specific sources, the development of a long-term strategy 
requires EPA to look at existing visibility impairment--after emissions 
reductions due to BART and other strategies are accounted for--and 
attribute responsibility for eliminating that impairment among sources 
and categories. 40 CFR 51.308(d)(1). The commenter stated that in this 
way, the states and EPA maintain flexibility to determine the most 
effective and efficient way to eliminate haze pollution when technology 
mandates on specified sources have not done the job. The commenter 
stated that therefore, measures within a long-term strategy are 
required to achieve reasonable progress above and beyond BART and that 
by categorically eliminating all BART-subject sources from its 
reasonable progress analysis, EPA has failed to meet its obligation to 
determine whether emissions reductions from these sources beyond those 
required by BART are necessary to achieve the national goal of 
eliminating visibility impairment.
    Response: We disagree that BART sources need to be re-evaluated for 
the purposes of reasonable progress. Our reasonable progress guidance 
states:

    Since the BART analysis is based, in part, on an assessment of 
many of the same factors that must be addressed in establishing the 
RPG, it is reasonable to conclude that any control requirements 
imposed in the BART determination also satisfy the RPG-related 
requirements for source review in the first RPG planning period. 
Hence you may conclude that no additional emissions controls are 
necessary for these sources in the first planning period.\59\
---------------------------------------------------------------------------

    \59\ Reasonable Progress Guidance, p. 4-2--4-3.

    The commenter has given no reason for us to change this position.
    Comment: A commenter stated that EPA's approach essentially 
duplicates all of the errors from its BART analysis in its reasonable 
progress analysis and that in particular, EPA's incremental visibility 
justification for dismissing the most stringent pollution control 
technologies is especially inappropriate in the reasonable progress 
framework. The commenter stated that incremental visibility improvement 
is not included among the four factors to be considered in establishing 
reasonable progress measures. 40 CFR 51.308(d)(1)(i)(A). The commenter 
stated that if this justification is applied to eliminate the most 
effective pollution-reduction measures at every source--especially the 
largest and oldest sources that are subject to BART--then Montana may 
never make reasonable progress toward achieving natural visibility 
conditions.
    Response: We disagree that there are errors in our approach for 
BART and reasonable progress for the same reasons we have discussed 
previously. Pursuant to 40 CFR 51.308(e)(A) for our BART analyses, we 
considered the following five factors in our analysis: The appropriate 
level of BART control; the cost of compliance; the energy and nonair 
quality environmental impacts; any pollution control equipment in use 
at the source; the remaining useful life of the source; and the degree 
of improvement which may be reasonably anticipated to result from the 
use of such technology. We agree that visibility improvement is not one 
of the four factors required by CAA section 169A(g)(1) and 40 CFR 
51.308(d)(1)(i)(A), however, it (along with other relevant factors) can 
be considered when determining controls that should be required for 
reasonable progress. Our reasonable progress guidance states: ``In 
determining reasonable progress, CAA section 169A(g)(1) requires States 
to take into consideration a number of factors. However, you have 
flexibility in how to take into consideration these statutory factors 
and any other factors that you have determined to be relevant.'' \60\ 
For certain potentially affected sources, we considered Q/D and 
potential reductions in Q/D, which are relevant to

[[Page 57900]]

the goal of the Regional Haze Rule, improving visibility.
---------------------------------------------------------------------------

    \60\ Reasonable Progress Guidance, p. 5-1.
---------------------------------------------------------------------------

    Comments: A commenter stated that EPA failed to require that 
Colstrip Units 1 and 2 and Corette make emissions reductions that were 
relied upon by the WRAP, EPA, and states neighboring Montana in 
establishing reasonable progress goals, and that if EPA fails to revise 
its BART determinations for Colstrip Units 1 and 2 and Corette, EPA 
must require additional reductions of visibility-impairing pollutants 
in its long-term strategy. Another commenter stated that EPA should 
have required SCR+SOFA as BART for Colstrip Units 1 and 2 and should 
have required SOFA+SCR and a dry scrubber/baghouse for Corette, but 
even if EPA were to justify its contrary BART finding in response to 
these comments, EPA should have required SCR+SOFA and a dry scrubber/
baghouse at these units as part of its long term strategy. The 
commenter explained that where sources within a state contributes to 
visibility within another state's Class I area or areas, the state has 
an obligation to adopt controls necessary to ensure it achieves its 
share of the pollution reductions that are required to meet the 
reasonable progress goals set for the subject Class I area.
    Response: We do not agree that we must revise our BART 
determinations for Colstrip Units 1 and 2 and Corette. We have stated 
in other actions addressing regional haze that a plan that provides for 
emission reductions consistent with the assumptions underlying the WRAP 
modeling will ensure that a State is not interfering with measures 
designed to protect visibility in other states. See e.g. 76 FR 491, 
496-497 (Jan. 5, 2011). Similarly, a plan that is consistent with the 
assumptions underlying the modeling used to establish RPGs in a state 
likely will include the measures necessary to achieve those RPGs. 
However, there is no requirement that a SIP (or FIP) adopt the 
assumptions underlying the models as enforceable requirements. The air 
quality models used to support the regional haze SIPs are extremely 
complex, and due to the time consuming nature of performing the 
modeling, this work was performed early in the process. The emissions 
projections by the RPOs, relied upon in the air quality modeling, 
incorporated the best available information at the time from the 
states, and utilized the appropriate methods and models to provide a 
prediction of emissions from all source categories into the future. 
There was an inherent amount of uncertainty in the assumed emissions 
from all sources, including emissions from BART-eligible sources, as 
the final control decisions by all of the states were not yet complete. 
The WRAP used their best estimates of what regional haze SIPs would 
achieve as inputs for the modeling. In the end, reductions resulting 
from BART determinations based on the statutory factors may differ from 
those estimates.
    One relevant requirement cited by the commenter, at 40 CFR 
51.308(d)(3)(ii), is that EPA must demonstrate that it has included all 
measures necessary to obtain its share of the emission reductions 
needed to meet the RPGs for Class I areas where it causes or 
contributes to impairment. Montana's neighboring Class I states 
originally set the reasonable progress goals in their SIP based on 
emission reductions expected to be achieved through application of 
presumptive BART and other emission reductions qualified for that 
purpose. These neighboring states had the opportunity to comment on the 
regional haze FIP, and did not ask for additional emission reductions. 
We also note that the RPGs are not enforceable goals. Neighboring 
states will have the responsibility to consider whether other 
reasonable control measures are appropriate to ensure reasonable 
progress during subsequent periodic progress reports and regional haze 
SIP revisions as required by 40 CFR 51.308(f)-(h), and may at that time 
consider asking EPA for additional emission reductions.
    With respect to Colstrip Units 1 and 2, we note that our FIP 
achieves SO2 emissions reductions well beyond those assumed 
in the WRAP PRP18b emissions inventory. Specifically, at Units 1 and 2, 
assuming operation at 85% of capacity, our FIP achieves reductions of 
7,538 tpy of SO2, which is 1,504 tpy better than indicated 
by the PRP18b projections. By way of comparison, again assuming 
operation at 85% of capacity, our FIP achieves reductions of 6,652 tpy 
of NOX for Colstrip Units 1 and 2, which is 1,709 tpy below 
that indicated by the PRP18b projections. Because the additional 
SO2 reductions are close to the shortfall in NOX 
reductions at Colstrip Units 1 and 2, and as SO2 may have a 
greater impact than NOX on visibility in Montana, we find 
that the overall emissions reductions achieved at Colstrip Units 1 and 
2 will result in similar visibility improvement to the emissions 
reductions assumed in the WRAP PRP18b projections.
    With respect to Corette, the commenter has overstated the 
discrepancy between the emissions associated with our BART 
determination and the PRP18b projections, because the commenter has 
compared WRAP projections based on annual emissions with emissions 
limits that are on a 30-day rolling average. In addition, we note that 
we have revised the NOX and SO2 emission limits 
for Corette in our FIP to be somewhat more stringent than what we 
proposed (and more reflective of actual emissions with existing 
controls). Finally, the WRAP projections do not reflect application of 
SOFA+SCR or a dry scrubber/baghouse to Corette. Therefore, the 
projections do not support the commenter's position that these controls 
are required.
    Moreover, there are NOX reductions at other BART sources 
that are greater than assumed by WRAP. At Ash Grove and Holcim, the 
total reductions from our FIP are significantly more relative to the 
PRP18b projections that the WRAP used. In conclusion, our FIP contains 
additional emission reductions at BART sources that largely offset any 
shortfall at Colstrip Units 1 and 2 and Corette.
    Comment: A commenter stated that our reasonable progress goals are 
unreasonable, unsupported, and effectively contrary to the CAA's 
requirements that we assure reasonable progress in achieving natural 
visibility conditions in Class I areas. The commenter stated that the 
proposed RPGs, at a minimum, double the timeframe required to achieve 
natural visibility conditions for every Class I area in Montana and 
that this is not reasonable. The commenter also stated that the 
reasonable progress goals are unreasonable based on the statutory 
factors that must be considered by EPA under 42 U.S.C. 7491(g)(1), and 
that we provided two reasons for asserting that the reasonable progress 
goals are reasonable: That our four factor analyses resulted in limited 
opportunities for reasonable progress controls for point sources and 
that significant visibility impairment is caused by non-anthropogenic 
sources in and outside Montana. The commenter stated that with regard 
to the latter issue of non-anthropogenic sources in and outside of 
Montana, this is not a statutory factor that EPA is allowed to consider 
in establishing RPGs.
    Response: We disagree. It is not necessarily unreasonable for the 
RPGs to reflect a longer period of time than the URP. The URP is simply 
calculated by dividing the difference between the present visibility 
conditions and natural visibility conditions by the number of years 
between the baseline and 2064. It assumes a steady rate of progress and 
does not take into account the four statutory factors for determining 
reasonable progress or any additional factors that warrant 
consideration. As a

[[Page 57901]]

result, the RPGs, which do reflect consideration of these factors, may 
well vary from the URP.
    In determining reasonable progress controls, EPA did consider the 
statutory factors for determining reasonable progress set out in 42 
U.S.C. 7491(g)(1). To the extent that the commenter argues with our 
evaluation of these factors, we respond to specific comments on our 
evaluation of these factors elsewhere.
    The commenter is correct that consideration of non-anthropogenic 
sources in and outside of Montana is not one of the statutory four 
factors that must be considered under 42 U.S.C. 7491(g)(1). However, 
EPA's reasonable progress guidance states: ``In determining reasonable 
progress, CAA section 169A(g)(1) requires States to take into 
consideration a number of factors. However, you have flexibility in how 
to take into consideration these statutory factors and any other 
factors that you have determined to be relevant.'' \61\ The data 
demonstrating that significant visibility impairment is caused by non-
anthropogenic sources in and outside Montana is relevant because it 
diminishes the potential improvement that might be realized through 
controlling an individual point source within Montana. Therefore, it 
was proper for EPA to consider this additional factor.
---------------------------------------------------------------------------

    \61\ Reasonable Progress Guidance, p. 5-1.
---------------------------------------------------------------------------

    Comment: A commenter stated that based on the four factors set 
forth under the CAA, it appears that EPA grossly overstated its 
assertion that there are only limited opportunities for reasonable 
controls for point sources. The commenter stated that this is 
particularly the case with regard to NOX emissions from 
coal-fired EGUs in Montana. The commenter stated that our proposal 
disclosed that for every coal-fired EGU assessed under the four-factor 
analysis for determining RPGs, including Colstrip units 3 and 4, 
Colstrip Energy, and the Lewis and Clark Station, that cost-effective 
SCR control technology could achieve greater NOX emissions 
reductions and greater visibility improvements than under our FIP. The 
commenter stated that despite this, we rejected SCR as a control option 
and ultimately adopted no NOX emission controls for these 
four sources. The commenter stated that we also rejected SCR as BART 
for Colstrip Units 1 and 2 and the Corette coal-fired EGUs, even though 
we found SCR to be a cost-effective and reasonable technology, we 
rejected it in favor of weaker controls. The commenter concluded that 
we did not show that any of the four factors would mitigate against 
additional control and stronger RPGs. The commenter stated that our 
assertion that there would be no degradation is not reasonable or 
legally justified and that we must establish our reasonable progress 
goals based on all coal-fired EGUs using SCR to reduce NOX 
emissions.
    Response: We disagree that the four factor analyses for EGUs that 
are potentially affected reasonable progress sources mandate the 
addition of SCR and that visibility, although not one of the four 
statutory factors that are required to be considered, cannot be 
considered in determining appropriate controls under reasonable 
progress. EPA's reasonable progress guidance states: ``In determining 
reasonable progress, CAA section 169A(g)(1) requires States to take 
into consideration a number of factors. However, you have flexibility 
in how to take into consideration these statutory factors and any other 
factors that you have determined to be relevant.'' \62\ For example, 
the potential reduction in Q/D is a factor that we consider to be 
relevant because the goal of the Regional Haze Rule is to improve 
visibility at Class I areas. We note that the commenter, in citing 
potential visibility improvement at the facilities mentioned, undercuts 
their own argument that the four statutory RP factors by themselves, 
without consideration of other factors, demonstrate that EPA ``grossly 
overstated'' its conclusion that there are only limited opportunities 
for reasonable controls for point sources. Commenter misstated EPA's 
conclusions by stating that EPA ``found SCR to be a cost-effective and 
reasonable technology'' for the BART EGUs. While we did state that the 
cost on a dollars per ton basis was cost-effective, we also explained 
that the cost of SOFA + SCR was not justified by the visibility 
improvement. 77 FR 24027, 77 FR 24035, and 77 FR 24043. The commenter 
misstated the requirements of the Regional Haze Rule. In examining 
potentially affected sources for possible controls and setting RPGs, 
EPA is not required to ``show that any of the four factors would 
mitigate against additional controls and stronger reasonable progress 
goals.'' Instead, EPA is required to consider the four statutory 
reasonable progress factors. In addition, EPA may consider additional, 
relevant factors such as visibility improvement from controls. To the 
extent that the comment argues with our determinations for particular 
potentially affected sources, we respond to specific criticisms 
elsewhere. With regard to commenter's statement that our basis for 
determining there would be no degradation on the least impaired days 
was unreasonable and not legally justified, we note that the commenter 
did not identify any flaw in our data or methodology in deriving Table 
198 in the proposal. We therefore disagree with the statement.
---------------------------------------------------------------------------

    \62\ Reasonable Progress Guidance, p.5-1.
---------------------------------------------------------------------------

    Comment: PPL commented that to try to address visibility impairment 
only within the universe of point sources subject to potential EPA 
regulation within the United States is not reasonable and will not lead 
to achievement of Reasonable Progress Goals (RPGs). PPL stated further 
that EPA, in conjunction with other federal and state agencies and the 
FLMs, should re-evaluate some of the conclusions as to the 
uncontrollable nature of several listed significant contributors of 
SO2 and NOX. PPL stated that application of the 
BART analysis excludes consideration of a number of factors, including 
outside domain sources. PPL pointed out that the RPGs in the proposed 
FIP do not take into account the contribution of international 
emissions to the visibility, and do not address challenges faced by the 
state of Montana.
    Response: To the extent that PPL commented that we are addressing 
visibility impairment only within the universe of point sources subject 
to potential EPA regulation within the United States, that we did not 
consider other sources of emissions, we disagree. As explained 
elsewhere, our action considered the many contributors to haze 
including all anthropogenic sources as required by 40 CFR 
51.308(d)(3)(iv) and smoke management techniques for agricultural and 
forestry management techniques as required by 40 CFR 
51.308(d)(3)(v)(E). In our proposal, the emissions inventory 
appropriately included natural (non-anthropogenic) wildfire and 
anthropogenic sources such as open burning and international emissions. 
We proposed approve the revisions to the smoke management section of 
Montana's Visibility SIP as meeting the requirement in 40 CFR 
308(d)(3)(v)(E).
    Comment: The NPS commented that EPA used inconsistent criteria in 
selecting reasonable progress controls.
    Response: We disagree. As explained in other responses, in 
determining the measures necessary to make reasonable progress and in 
selecting RPGs for mandatory Class I areas within Montana, we took the 
following four factors into consideration: costs of compliance; time 
necessary for compliance; energy and nonair quality

[[Page 57902]]

environmental impacts of compliance; and remaining useful life of any 
potentially affected sources. CAA section 169A(g)(1) and 40 CFR 
51.308(d)(1)(i)(A). As also explained in other responses, we also 
considered potential visibility improvement in a general sense by 
considering the potential reduction in haze causing pollutants and also 
the distance from the source to the nearest Class I area. For Colstrip 
3 and 4, we also considered visibility modeling results and have 
explained the reasoning for that decision in another response.

J. Comments on Colstrip Units 3 and 4

    Comment: Some commenters agreed with EPA's conclusion not to 
require additional emissions controls at Colstrip Units 3 and 4. 
Commenters asserted that, given the aggressive pollution control 
technologies already in place, EPA properly concluded that additional 
controls for Reasonable Progress are not appropriate.
    Response: We acknowledge the commenters' support for our decision 
not to require additional emission controls on Colstrip Units 3 and 4 
in this planning period. Whether additional emission reductions from 
reasonable progress sources, including Colstrip Units 3 and 4, are 
necessary will be re-evaluated in subsequent planning periods.
    Comment: Various commenters stated that we underestimated the costs 
of SNCR for Colstrip Units 3 and 4.
    Response: We disagree that we underestimated the costs of SNCR for 
Colstrip Unit 3 and 4. For a further explanation, see our response to 
similar comments made in relation to SNCR costs for Colstrip Unit 1 and 
2.
    Comment: Commenters stated that they disagree with EPA's cost 
analysis for NOX control technologies for Colstrip Units 3 
and 4. In particular, commenters stated that we underestimated the 
capital costs and cost-effectiveness of these controls. Commenters 
referenced cost estimates submitted by PPL in September 2011 and 
February 2012, which show much higher capital costs and cost-
effectiveness than those estimated by EPA.
    Response: We disagree. We have rejected PPL's cost estimates for 
NOX control options for Colstrip Units 3 and 4 for the same 
reasons that we rejected them for Colstrip Units 1 and 2. See previous 
responses to comments.
    Comment: NPS stated that EPA modeled baseline visibility impacts at 
five Class I areas from Colstrip Units 3 & 4 using 2008-2010 emissions, 
while PPL modeled visibility impacts using 2001-2003 emissions. NPS 
agreed with the PPL modeling approach because it is consistent with EPA 
guidance to use the 2001-2003 pre-control emissions.
    Response: See our response to a similar comment made in regard to 
the baseline emissions used for Colstrip Units 1 and 2.
    Comment: NPS stated that after EPA concluded its statutory four-
factor analysis of Colstrip 3 and 4, it created a new, ``Optional 
Factor: Modeled Visibility Impacts'' fifth factor, only for Colstrip 3 
& 4. NPS further stated that this ``optional'' fifth factor is not 
required by statute or regulation, and that EPA only used it on one 
reasonable progress source (2 units) and did not explain what criteria 
it used to evaluate it.
    Response: As we explained elsewhere, our RP Guidance allows for 
consideration of additional factors such as visibility impacts or 
benefits. Given the large annual emissions of NOX and 
SO2 from Colstrip Units 3 and 4 compared to other reasonable 
progress sources, we found that it was reasonable to model the 
visibility benefits and consider them when evaluating controls.
    Comment: NPS stated that EPA has not provided criteria used in 
making the determination of what ``Costs of Compliance'' are 
reasonable, and its determinations vary significantly across Montana 
facilities.
    Response: As we have explained elsewhere, while the Regional Haze 
Rule and BART Guidelines allow states to establish thresholds for cost-
effectiveness, we are not required to do so and have not done so for 
this action. Also, our Reasonable Progress determinations were made 
based not just on the cost of compliance, but with consideration of the 
four factors along with additional information that was pertinent.
    Comment: EarthJustice stated that EPA must set NOX 
emission limits for Colstrip Units 3 and 4 based on SCR to help achieve 
reasonable progress. EarthJustice stated that EPA's analysis is skewed 
to underestimate the benefits of SCR, both in terms of control 
effectiveness and visibility improvement, and overestimates the costs. 
EarthJustice made claims regarding our cost analysis for Colstrip Units 
3 and 4 that were very similar to the claims they made regarding 
Colstrip Units 1 and 2.
    Response: We disagree. Below we address each of EarthJustice's 
arguments that support their assertion that SCR must be required for 
Colstrip Units 3 and 4.
    Comment: EarthJustice stated that EPA underestimated the control 
effectiveness of SCR.
    Response: See our response to similar comment made by EarthJustice 
in regard to Colstrip Units 1 and 2.
    Comment: EarthJustice stated that EPA overestimated the cost of 
SCR.
    Response: See our response to similar comment made by EarthJustice 
in regard to Colstrip Units 1 and 2.
    Comment: EarthJustice claimed that the visibility benefit of SCR on 
Units 3 and 4 is substantial and therefore SCR should be required. 
EarthJustice noted that EPA modeled visibility benefits of SNCR and SCR 
and found a visibility benefit of 0.273 dv per unit from application of 
SCR. EarthJustice stated that application of SCR at both units would 
approximately halve the units' emissions of visibility impairing 
pollutants and would reduce the number of days of visibility impairment 
at Theodore Roosevelt NP to just 2 days and would eliminate visibility 
impairment caused by Units 3 and 4 at four other Class I areas. 
EarthJustice stated that, in light of this, we lacked a basis for our 
determination to not impose SCR at Colstrip Units 3 and 4. EarthJustice 
noted that, in North Dakota, we imposed LNB on two units at Antelope 
Valley Station based on a combined visibility benefit of 0.39 deciview, 
which we stated was significant even on a unit-by-unit basis of 0.2 
deciview.
    Response: We disagree that SCR should be required based solely on 
the modeled visibility benefits. As we explained in our proposal, we 
considered the four factors and the modeled visibility benefits of 
controls and determined that no additional controls should be required 
for this planning period. 77 FR 24066. Also, we stated that 
specifically, for SCR, the modeled visibility benefits (0.273 deciview 
and 0.260 deciview) were not sufficient for us to consider it 
reasonable to impose SCR in this planning period. 77 FR 24066. In 
making this determination, we noted that SCR was the more expensive 
option ($4,574/ton at Unit 3 and $4,607/ton at Unit 4). The cost of 
compliance is one of the four statutory factors, and EarthJustice has 
not provided a reason why it should be ignored. For the same reason, we 
reject the comparison with our North Dakota action. There, the cost-
effectiveness of LNB at Antelope Valley Station was $586/ton for Unit 1 
and $661/ton at Unit 2. 76 FR 58631. We explicitly considered these 
costs in making our determination to impose LNB. Here, the cost-
effectiveness of SCR at Colstrip Units 3 and 4 is far above the

[[Page 57903]]

cost-effectiveness of LNB at Antelope Valley Units 1 and 2. Thus, the 
comparison gives us no basis to change our determination that SCR 
should not be required in this planning period.
    Comment: EarthJustice stated that EPA should set more stringent 
SO2 emission limits at Colstrip Units 3 and 4 to help 
achieve reasonable progress. EarthJustice stated that EPA incorrectly 
found that no additional upgrades are feasible and that 98% 
SO2 removal to meet an SO2 emission limit of 0.05 
lb/MMBtu at Units 3 and 4, which is readily achievable at little 
expense using MEL.
    Response: EarthJustice cites a 1984 paper presented at the American 
Power Conference to support their argument of a lower emission rate. 
Colstrip 3 had only started operation in 1984 and Colstrip 4 did not 
commence operation until 1986,\63\ the data cited by EarthJustice 
cannot be more than short-term tests of Unit 3 that are not 
representative of longer term performance. Annual emissions from 1985 
and 1990 emissions from CAMD can be found in the docket. At the time 
these scrubbers were built, wet MEL scrubbers and wet caustic scrubbers 
were the only scrubbers that could deliver high capture rates (over 
90%) with reasonable reliability. Scrubber technology has improved and 
other, less expensive, reagents are now preferred. Although Colstrip 
Units 3 & 4 used MEL in the past, MEL is not readily available in the 
region near the Colstrip plant. MEL is produced from a blending of 
dolomitic lime with high calcium lime to achieve a lime with a 
magnesium content of 3-6% or so. The lime is produced by calcination of 
limestone. Dolomitic limestone is limestone with a significant amount 
of dolomite, or calcium magnesium carbonate. Because there are no 
dolomitic limestone deposits near the Colstrip plant, the dolomitic 
lime must be sourced from remote locations. This increases the cost of 
the lime (that is made from the dolomitic limestone). According to 
Carmeuse, a supplier of MEL, the closest source of dolomitic lime is 
1,000 miles away from the Colstrip plant and transportation would cost 
$0.12 per mile per short ton plus a 24% fuel surcharge to 
transport,\64\ or close to $150/short ton just for transportation of 
the reagent. Because the lime would be blended in closer to the plant 
with high calcium lime at perhaps an 8:1 ratio (reducing magnesium 
content from about 40% to about 4-5% this would result in an increased 
reagent cost of $15-$20 per ton. Assuming a high-calcium lime cost of 
about $95/ton,\65\ this raises the cost of reagent by close to 20% 
assuming constant reduction. Reagent use might be improved somewhat for 
a given reduction level, but considering this is a unique scrubber 
design, it is difficult to assess what the impact may be. Regardless, 
reliance on a reagent source that is 1,000 miles away may cause 
operating risks during the winter months if delivery was interrupted.
---------------------------------------------------------------------------

    \63\ See EIA Form 860 data.
    \64\ Email from Bob Roden, Carmeuse, to Jim Staudt, Andover 
Technologies, July 31, 2012.
    \65\ Sargent & Lundy, ``IPM Model--Revisions to Cost and 
Performance for APC Technologies, SDA FGD Cost Development 
Methodology FINAL'', Prepared for US EPA, August 2010 see table 2.
---------------------------------------------------------------------------

    We also note that EarthJustice did not provide site-specific cost 
information, for us to evaluate MEL. The cost of compliance is one of 
the factors required to be considered by CAA section 169A(g)(1) and 40 
CFR 51.308(d)(1)(i)(A). Based on all four factors, we continue to find 
that the level of performance of the current SO2 removal 
system for Colstrip Units 3 and 4 is satisfactory for this planning 
cycle. We will re-evaluate additional SO2 controls for 
Colstrip Units 3 and 4 in the next planning cycle.
    Comment: PPL stated that EPA properly concluded that RPGs do not 
require additional emissions controls on Colstrip Units 3 and 4 and 
that existing emissions controls at Units 3 and 4 already limit 
emissions to levels below the presumptive BART limit. PPL stated that 
EPA's RP conclusion should not be affected by EPA's ultimate 
determination with respect to BART requirements for Colstrip Units 1 
and 2 and that no further controls are warranted based on conclusions 
regarding the extent of existing emissions controls and the cost-
ineffectiveness of further controls.
    Response: PPL did not provide specific information for us to 
consider in making a change to our FIP. In any case, we have not 
required additional controls for Colstrip Units 3 and 4 in our final 
FIP.

K. Comments on Devon Energy

    Comment: MDEQ stated that we failed to provide information or 
analysis of any visibility benefit that would result from the 
application of NSCR for Devon Energy. MDEQ suggested that we must 
consider visibility benefits as part of the Devon Energy reasonable 
progress analysis, as the BART Guidelines include evaluation of 
visibility impacts ``which would also appear to be required under the 
reasonable progress guidelines.''
    Response: The four reasonable progress factors are the costs of 
compliance, the time necessary for compliance, the energy and nonair 
quality environmental impacts of compliance, and the remaining useful 
life of any potentially affected sources CAA section 169A(g)(1) and 40 
CFR 51.308(d)(1)(i)(A). Our Reasonable Progress Guidance states: ``In 
determining reasonable progress, CAA section 169A(g)(1) requires States 
to take into consideration a number of factors. However, you have 
flexibility in how to take into consideration these statutory factors 
and any other factors that you have determined to be relevant.'' \66\ 
As stated in our proposal at 77 FR 24069, for Devon, we considered Q/D 
and potential reductions in Q/D, which are relevant to the goal of the 
Regional Haze Rule, improving visibility.
---------------------------------------------------------------------------

    \66\ Reasonable Progress Guidance, p. 5-1.
---------------------------------------------------------------------------

    Comment: MDEQ commented that EPA should review the NOX 
limit for Devon with respect to its averaging time and compliance 
determining method for practical enforceability.
    Response: In the final FIP, we have made changes to the language in 
40 CFR 52.1396 to clarify the requirements for Devon Energy.

L. Comments on Montana-Dakota Utilities

    Comment: Montana-Dakota Utilities (MDU) commented that the company 
did not disagree with our Reasonable Progress determination. MDU stated 
that, for EPA's reference, paragraph 3 on page 1 of the Sargent & Lundy 
IPM model method document cautions as follows with respect to the 
application of the model to smaller units:

    The costs for retrofitting a plant smaller than 100 MW increase 
rapidly due to the economy of size. The older units which comprise a 
large proportion of the plants in this range generally have more 
compact sites with very short flue gas ducts running from the boiler 
house to the chimney. Because of the limited space, the SCR reactor 
and new duct work can be expensive to design and install. 
Additionally, the plants might not have enough margins in the fans 
to overcome the pressure drop due to the duct work configuration and 
SCR reactor and therefore new fans may be required.

    MDU stated that Lewis & Clark Station is a small, 52 MW net 
capacity unit. In addition, MDU believes that the fan margin is not 
present at Lewis & Clark Unit 1 to overcome the pressure drop as 
discussed in the Sargent & Lundy guidance.
    Response: MDU has not provided the information that would be 
necessary for

[[Page 57904]]

us to determine whether or not to agree with the implied point of this 
comment, which seems to be that EPA underestimated the cost of SCR. 
First, MDU has not indicated whether there are, in fact, space 
limitations at Lewis & Clark Station that would cause installation of 
an SCR reactor and associated ductwork to be more expensive than the 
cost estimate in our analysis. Second, MDU has not indicated whether 
the additional pressure drop from installation of SCR at Lewis & Clark 
Station would, in fact, require installation of new fans, and if so, 
whether or not our cost analysis failed to factor in the cost of new 
fans.
    Comment: MDU indicated that EPA uses a Retrofit Factor value of 1 
for Lewis & Clark Station Unit 1 in the IPM Model calculation (factor B 
in the EPA cost sheets) which indicates an average retrofit cost, 
however, a higher value would be expected for Lewis & Clark since it is 
a small facility (as discussed/cautioned above by Sargent & Lundy) and 
could be difficult to retrofit. A more appropriate value between 1.3 
and 2.0 is therefore recommended.
    Response: We disagree. MDU has not provided any data or information 
to substantiate that a retrofit factor other than 1 is warranted for 
Lewis & Clark Station. The IPM capital cost calculations for retrofits 
already account for unit size. We note that capital cost does not vary 
linearly with size in IPM. Instead, in the capital cost formula in IPM, 
the cost varies exponentially with unit size (a least squares fit). The 
IPM document states, ``The least squares curve fit was based upon an 
average of the SCR retrofit projects.'' IPM Model--Revisions to Cost 
and Performance for APC Technologies, SCR Cost Development Methodology, 
Final, Sargent & Lundy, August 2010, Chapter 5, Appendix 5-2A, page 4-
5.
    We also disagree with the statement that a more appropriate 
retrofit factor should be 1.3 to 2.0. The aforementioned IPM document 
states that, ``Retrofit difficulties associated with an SCR may result 
in capital cost increases of 30 to 50% over the base model.'' 
Therefore, the highest retrofit factor that might be considered would 
be 1.5.
    This comment has not resulted in any change to our FIP proposal or 
to our cost calculations for SCR.
    Comment: MDU stated that the model ``Type of Coal'' input indicates 
``PRB'', but should be ``Lig,'' since Lewis & Clark burns lignite coal. 
That stated, the ``Coal Factor'' value in the cell below ``Type of 
Coal'' indicates lignite coal was actually considered. As such, this 
recommendation is clerical in nature.
    Response: As shown in the ``Given/Assumptions'' spreadsheet in our 
SCR cost analysis, we used a heating value of 6,714 Btu/lb, which we 
considered to be representative of lignite coal. PRB coal would have a 
much higher heating value.
    Comment: MDU stated that EPA used a NOX input emission 
rate to the SCR of 0.26 lb/MMBtu, which is the low load emissions rate 
of low NOX burners (LNB) and Separated Overfire Air (SOFA) 
that MDU estimated in Table C.2-1 of Appendix C.2 of the Emissions 
Control Analysis for Lewis & Clark Station Unit 1. The 0.25 lb/MMBtu 
for LNB/SOFA at high load is a more appropriate rate to use as the 
inlet to an SCR. While this does not result in a significant change to 
the overall conclusions in the report, it is nonetheless important 
because the EPA-derived cost was based on full load operation, as 
opposed to lower load.
    Response: We disagree with the statement that we obtained the 
emission rate of 0.26 lb/MMBtu from the low-load scenario presented in 
Table C.2-1 of Appendix C.2 of MDU's Emissions Control Analysis. 
Instead, as indicated in the ``Given/Assumptions'' spreadsheet of our 
SCR cost analysis, we obtained the rate of 0.26 lb/MMBtu from Table 
C.2-6 of MDU's analysis. Table C.2-6 is not identified by MDU as a low-
load scenario.
    Comment: MDU stated that, from the IPM model guidance, EPA did not 
include factors N through V in the model calculations for operating 
costs for Lewis & Clark Station's evaluations. Although factors N 
through R and T through V are utility costs that were not needed in 
EPA's evaluation, the catalyst cost (factor S) was applied based on an 
alternative source. EPA references ``Cichanowicz (Jan 2010)'' with a 
cost of $170/ft\3\ as compared to the IPM value of $8,000/m\3\ 
($226.53/ft\3\ in 2009$) and MDU's value of $214.29/ft\3\. MDU 
recognized that a range of potential costs exist, and believes that 
either the IPM value or the value MDU provided would be more 
appropriate for EPA to use since they are based on industry and vendor 
data respectively and are expected to represent a more site specific 
value as opposed to a literature based value.
    Response: We disagree. The Cichanowicz document we used provided 
actual catalyst costs observed over time. It demonstrates that catalyst 
costs continue to decline. In fact, based on the trend displayed in the 
graph on page 6-6 of the document, it is likely that catalyst costs in 
upcoming years will be even lower than the $6,000/m\3\ assumed in our 
FIP proposal. Current Capital Cost and Cost-Effectiveness of Power 
Plant Emissions Control Technologies, J. Edward Cichanowicz, Prepared 
for Utility Air Regulatory Group, January 2010, page 6-6, Figure 6-6. 
This comment has not resulted in any change to our FIP proposal or to 
our cost calculations for SCR.
    Comment: Similarly, to item e above, MDU noted that the cost EPA 
associated with aqueous ammonia ($0.12/lb) is lower than the cost MDU 
estimated of $0.70/lb. MDU recognized that a range of ammonia costs 
exists, that the price of ammonia fluctuates over time, and that the 
price is related to natural gas prices. As such, if SCR were to be 
considered in the future, MDU would ask that site specific, local, as 
delivered cost be evaluated at that time.
    Response: We disagree. In its own SCR cost spreadsheet, MDU did not 
indicate the basis for its estimate of $0.70/lb. We used $0.12/lb based 
on data provided to us by control technology vendors on cost of aqueous 
ammonia. This comment has not resulted in any change to our FIP 
proposal or to our cost calculations for SCR.
    Comment: MDU stated that, through the FR correction, EPA changed 
the language on 77 FR 24071 to state that an 85% control efficiency was 
used instead of the initially quoted 95% control efficiency for SDA and 
baghouse. MDU believes this correction was in error. Table 172 in the 
FR lists the control efficiency as 85% for SDA and baghouse and this 
value should be corrected to 95% control efficiency for SDA and 
baghouse as the textual representation in the FR was correct.
    Response: We disagree. We made the correction from 95% to 85% 
because MDU's Emissions Control Analysis dated June 2011, at Table 1 on 
page 14, shows an expected SO2 emission reduction of 850.3 
tons per year, for SDA with baghouse. The baseline SO2 
emissions listed in the table are 1,002.1 tons per year. This amount of 
reduction represents 85% control efficiency. We presented these figures 
at 77 FR 24071, Table 172. MDU later wrote to us on February 10, 2012, 
to say that 70-90% control is the generally anticipated range of 
SO2 control for this control option, and that 95% control 
was also assumed and represented a screening level assumption for a 
high degree of SO2 control. In its February 10, 2012 
submittal, MDU did not indicate that Table 1 of their June 2011 
submittal should be revised, so we used the figures presented in MDU's 
Table 1.
    Comment: In Table 172 of the proposed FIP (77 FR 24071), EPA 
provides a 10% control effectiveness for

[[Page 57905]]

both DSI with baghouse and existing scrubber mod; however, MDU stated 
that this value should be changed to 70% to reflect the overall 
reduction and not the incremental reduction as shown in Table 1 of 
MDU's Emissions Control Analysis for Lewis & Clark Station Unit 1.
    Response: We disagree. We stated that we did use 70% overall 
SO2 control effectiveness for DSI with baghouse, as well as 
for existing scrubber mod, in our analysis. 77 FR 24071. However, we 
also stated that existing SO2 controls at Lewis & Clark 
Station, consisting of a flooded disc wet scrubber, have achieved up to 
60% control under certain operating conditions. 77 FR 24070. We 
obtained this information from MDU's analyses. 77 FR 24070, footnote 
265. MDU's Emissions Control Analysis dated June 2011, at Table 1 on 
page 14, lists an expected emissions reduction of 100.2 tons per year 
for DSI with baghouse, and the same amount of reduction for existing 
scrubber mod. This is a 10% reduction from the baseline emissions of 
1,002.1 tons per year listed in that table. We relied on these figures 
from MDU in listing a control effectiveness of 10% for DSI with 
baghouse, as well as a control effectiveness of 10% for existing 
scrubber mod. For all control options analyzed in our FIP proposal, we 
present control effectiveness in terms of the reduction that might be 
achieved from baseline emissions. In this case, the baseline emissions 
already reflected a 60% level of SO2 control.
    Comment: EarthJustice argued that EPA should require Lewis and 
Clark to switch from lignite fuel to natural gas as a reasonable 
progress measure. The unit already uses natural gas for startup, there 
is a natural gas supply close by, and thus switching to natural gas is, 
in commenter's view, quite feasible and cost effective for Lewis and 
Clark station. Switching to natural gas should be required in the FIP 
to help achieve reasonable progress, as this measure would virtually 
eliminate the unit's SO2 and PM emissions and would also 
reduce NOX emissions. Although EPA dismissed fuel switching 
as not cost effective, commenter argues that EPA vastly understated the 
cost effectiveness of this measure.
    Commenter first stated that EPA has overstated the costs of 
switching to natural gas, in large part because it has underestimated, 
and in some cases ignored, the tremendous cost savings that would 
result from not operating the facility's scrubber, multi-cyclone dust 
collector, and coal preparation systems. EPA also relied on inflated 
estimates for natural gas and natural gas supply pipelines provided by 
MDU, which owns Lewis and Clark.
    Commenter also stated that EPA has improperly calculated the 
emissions reductions achievable from fuel switching. EPA failed to take 
into account the fact that the use of natural gas would replace the 
existing SO2 and PM controls. Commenter stated that, in view 
of the 54 kilometer distance from Lewis and Clark to the closest Class 
I area, filterable PM must be considered. Thus, EPA should have 
accounted for the pollution reductions that would be achieved with 
natural gas from uncontrolled levels of SO2 and PM. Properly 
calculated, fuel switching would eliminate 24,000 tons per year of 
SO2, NOX and filterable PM. As EPA noted, Lewis 
and Clark's remaining emissions would be ``negligible.''
    Commenter concluded that, even using EPA's inflated cost estimate, 
when uncontrolled rates of SO2 and PM are used as the 
baseline, the cost effectiveness of switching to natural gas at Lewis 
and Clark station is $909/ton of SO2, NOX and PM 
removed. This measure is highly cost effective and should be required 
to help achieve reasonable progress.
    Response: We disagree. Although we do not believe it was 
necessarily an error for us to rely on MDU's estimate of the price of 
natural gas, we acknowledge that price estimates for natural gas can 
vary, and that the $3.07/Mscf price of natural gas cited on page 129 of 
the commenter's Technical Support Document, obtained from the Energy 
Information Administration (EIA), is substantially lower than MDU's 
estimate of $7.91/Mscf. However, even if we rely on the price cited by 
the commenter, the cost of a fuel switch would still be excessive. 
Using $3.07/Mscf, along with MDU's estimate of 3,282,876 Mscf of 
natural gas which would be needed to fuel Lewis and Clark station year-
round solely on natural gas (not disputed by the commenter), we 
calculate the annual cost of natural gas at $10,078,429. MDU estimated 
the annual cost of coal at $5,754,732. The annual fuel cost 
differential would therefore be $4,324,197. To this result we add the 
annualized cost of constructing a natural gas pipeline ($1,699,200), as 
we did in our FIP proposal.\67\ This yields a total annual cost of 
$6,023,397. Dividing this result by an expected SO2 emission 
reduction of 1,002 tons per year yields cost effectiveness of $6,011/
ton. Based on this cost and other factors for Lewis and Clark station 
described in our FIP proposal at 77 FR 24072, we would still eliminate 
fuel switching as a control option for SO2.
---------------------------------------------------------------------------

    \67\ Commenter's speculation that the existing pipeline could be 
upgraded does not provide sufficient basis for us to supplant MDU's 
estimated cost for a new pipeline with some other cost. We note 
that, even if the upgrade were feasible and had zero cost, the cost 
effectiveness of the SO2 reductions would still be well 
over $4,000/ton.
---------------------------------------------------------------------------

    We disagree with the statement that a fuel switch would yield 
``tremendous'' cost savings from not operating the facility's scrubber, 
multi-cyclone dust collector, and coal preparation systems. Commenter 
has not quantified the cost savings. We have no reason to believe they 
would be ``tremendous.'' We believe the cost savings would be minimal 
in comparison to other components of our cost calculations for a fuel 
switch. The cost savings would likely consist primarily of avoidance of 
electricity and maintenance costs for the equipment cited by the 
commenter.
    Also, we disagree with the statement that we should have calculated 
reductions from uncontrolled levels of SO2 and PM. In every 
cost analysis of control options for our FIP, we calculate reductions 
from an emissions baseline which is the current actual annual 
emissions, consistent with the approach laid out in the 2005 Regional 
Haze Rule, at 70 FR 39167, for calculating cost effectiveness of 
control options. Commenter's citation to a 2008 letter sent by EPA in 
the course of developing initial information for a FIP ignores the 
basis for the action we actually proposed.
    We also disagree with the statement that a ``proper cost analysis'' 
would result in cost-effectiveness of $909/ton. Commenter apparently 
calculated $909/ton based on reduction from uncontrolled emissions, for 
the sum of three pollutants (PM, SO2 and NOX). We 
have explained above why we do not use uncontrolled emissions as the 
baseline. We also explained in our proposal that, in our reasonable 
progress determinations, we were not evaluating controls for PM for 
potentially affected sources, based on our analysis of the emissions 
inventory and results from BART modeling. 77 FR 24055-56. Commenter has 
not disputed those bases; commenter merely notes the 54 kilometer 
distance to Theodore Roosevelt NP. Given these flaws, the commenter's 
cost analysis provides no basis for us to reconsider our decision.
    Comment: Commenter noted that, although MDU proposed upgrades to 
its existing SO2 and NOX pollution controls, EPA 
failed even to require these measures to help achieve reasonable 
progress. See 77 FR 24074. Commenter stated that MDU's proposal is 
vastly inferior to fuel switching at reducing haze pollution, but MDU's

[[Page 57906]]

proposed controls are the bare minimum that EPA should have required 
for reasonable progress.
    Commenter noted that MDU proposed to improve SO2 removal 
to 70% by optimizing the existing particulate scrubber and lime 
injection system with a proposed limit of 0.45 lb/MMBtu. EPA estimated 
the cost effectiveness of this modification at $1,383/ton 
SO2 removed. MDU also proposed SOFA and low NOX 
burners (upgraded) to achieve a NOX emission rate of 0.25 
lb/MMBtu. EPA estimated the cost effectiveness of this option as 
$1,213/ton of NOX removed. Commenter stated that, although 
the emissions reductions from these measures are modest, they are 
highly cost effective and are the minimum that EPA should have required 
from Lewis and Clark to achieve reasonable progress.
    Response: We disagree. MDU's proposal to improve SO2 and 
NOX emission control was contained in its June 2011 
Emissions Control Analysis, which was submitted in response to a CAA 
section 114 information request from us. Under the Regional Haze Rule, 
we are not bound by controls that a source has proposed when we make 
our reasonable progress determination based on the four statutory 
factors.
    With regard to the statement that cost-effectiveness of $1,383/ton 
for SO2 and $1,213/ton for NOX is ``highly cost-
effective'' and should result in a requirement for emissions 
reductions, commenter has not provided a basis for this conclusion. As 
explained in our FIP proposal at 77 FR 24072 (for SO2) and 
24074 (for NOX), in making our reasonable progress 
determination for Lewis and Clark Station, we considered the following 
four reasonable progress factors: cost of compliance, the time 
necessary for compliance; the energy and nonair quality environmental 
impacts of compliance; and the remaining useful life of the source. We 
also took into account the following additional factors: size of the 
facility, the baseline Q/D of the facility, and the potential reduction 
in Q/D from the controls. Commenter has not disputed the 
appropriateness of using the four reasonable progress factors and other 
factors in our proposal.
    Comment: WEG commented that the determination in the proposed rule 
that no additional SO2 controls are required on Lewis & 
Clark Station is unreasonable. WEG notes that two highly effective 
control options are available (fuel switch to natural gas at 99% 
control effectiveness and SDA with baghouse at 85% control 
effectiveness) and should be further considered.
    Response: We disagree. EPA did not evaluate control options for 
Regional Haze FIP development solely based on emission control 
effectiveness. As indicated in EPA's analysis, the cost of fuel 
switching is estimated at $21,875 per ton of pollutant removed and the 
cost of SDA with baghouse is estimated at $11,825 per ton of pollutant 
removed. 77 FR 24072, Table 173. EPA has already explained that this 
cost is excessive. WEG has not provided a reason to not consider the 
cost excessive. Besides the cost of compliance, EPA also explained that 
other factors were taken into consideration in determining whether 
additional SO2 controls should be required at Lewis & Clark 
Station, those being the time necessary for compliance, the energy and 
nonair quality environmental impacts of compliance, the remaining 
useful life of the facility, the size of the facility, the baseline Q/D 
of the facility, and the potential reduction in Q/D from the controls. 
WEG did not provide a reason to re-evaluate these other factors.
    Comment: WEG comments that EPA should re-examine its decision to 
eliminate all control options for NOX and move to require 
HDSCR + SOFA/LNB at Lewis & Clark Station. WEG notes that this control 
option has a high control effectiveness of 87.5% and considers the cost 
of $4,853 per ton of pollutant removed to be reasonable. To rule it out 
alongside a fuel switch to natural gas, which has a much higher cost of 
$41,934 per ton of pollutant removed, lacks reason. WEG stated that the 
cost and visibility benefits of HDSCR + SOFA/LNB should be considered 
individually, and the control option should be implemented because of 
the great emissions reduction it achieves, and because the FIP is far 
from attaining a Uniform Rate of Progress (URP) akin to the regulatory 
rate. WEG also stated that the final analysis of control options took 
into account only ``the most cost effective option (SOFA/LNB)'' when 
weighing cost against overall reductions in emissions.
    Response: We disagree. EPA did consider control options 
individually. At Step 5 of its NOX analysis, EPA mentioned 
cost of HDSCR + SOFA/LNB in the same sentence as cost of a fuel switch 
only because those two options happened to be the most expensive. 77 FR 
24074. Besides the cost of compliance, EPA also explained that other 
factors were taken into consideration in determining whether additional 
NOX controls should be required at Lewis & Clark Station, 
those being the time necessary for compliance, the energy and nonair 
quality environmental impacts of compliance, the remaining useful life 
of the facility, the size of the facility, the baseline Q/D of the 
facility, and the potential reduction in Q/D from the controls. At Step 
5, EPA explained how these factors were considered with respect to all 
control options, not just SOFA/LNB. In the case of HDSCR + SOFA/LNB, 
EPA explained that this control option was eliminated on the basis of 
not only cost, but also on the basis of the small size of the facility 
and the relatively small baseline Q/D of the facility. WEG has not 
provided a reason to re-evaluate these other factors. With regard to 
URP, that comment was addressed in a previous response.

M. Comments on Montana Sulphur and Chemical Company

    Comment: MSCC commented that the company agrees with the conclusion 
in the proposed FIP that additional controls are not required at this 
time. MSCC also stated it does not believe we should have considered it 
to be a BART-eligible source. The company referenced several letters 
and discussions with MDEQ that were previously submitted and had as 
part of development of the regional haze plan for Montana.
    Response: Because the commenter ultimately agrees with the final 
conclusion and controls are not required for MSCC, at this time, we 
find the comment to be non-substantive.

N. Comments on Health, Ecosystem Benefits, Other Pollutants, and Coal 
Ash

    Comment: Several commenters stated that haze pollution 
significantly impacts human health and ecosystem health. Specifically, 
commenters asserted that haze pollution, including haze pollutants 
NOX, SO2 and PM, contributes to heart attacks, 
asthma attacks, chronic bronchitis and respiratory illness, decreased 
lung function, increased hospital admissions, and even premature death. 
Another commenter stated that NOX and SO2 can 
combine to create photochemical smog and ozone, which can exacerbate 
health problems.
    Some commenters cited a 2010 Clean Air Task Force report in stating 
that the Colstrip coal-fired power plant put 31 people at risk of 
premature death, 48 people at risk of a heart attack, 47 people at risk 
of acute bronchitis, and 534 at risk of an asthma attack each year.\68\ 
Several commenters encouraged EPA to finalize the regional haze 
proposal citing their own health

[[Page 57907]]

problems, or the health problems of family members.
---------------------------------------------------------------------------

    \68\ Several commenters cited numbers that were similar to 
these, but did not match them exactly.
---------------------------------------------------------------------------

    Some commenters stated that the negative health impacts of this 
pollution disproportionately harm vulnerable populations, specifically 
the young and elderly, and that this disproportionate harm potentially 
makes this a case of environmental justice. A commenter claimed that 
Colstrip causes a dark shadow on snow and takes human lives. One 
commenter stated the rate of asthma in children in Rosebud County is 
the third highest of all counties in the State, while another stated 
the rate of birth abnormality in the area downwind of Colstrip is much 
higher (34%) than in most other counties in Montana (10%). One 
commenter stated that over 10% of Montana high school students were 
estimated to have asthma in 2009. A commenter surmised that a 50% 
reduction in pollution from Colstrip would help human health more than 
eliminating pollutants from all other Montana sources.
    Some commenters expressed a willingness to pay more for power in 
support of pollution control technology, with others stating that we 
should all pay the full cost of energy and not pass it on as healthcare 
costs. Another commenter stated that the cost of pollution controls, 
especially at Colstrip, was small when compared to the health-related 
benefits. Other commenters stated that the sources should not be 
allowed to externalize the costs of their pollution onto the people, 
who must pay for them in the form of health-related costs.
    Some commenters stated that haze pollution negatively impacts 
ecosystem health. Commenters expressed concern for the effects of haze 
pollution on plants and water bodies. Some commenters specifically 
expressed concern over acid deposition from SO2 and 
NOX emissions, which they argued can leach into drinking 
water sources and harm crops. One commenter attributed high levels of 
mercury in some Montana back country lakes to coal-fired power plant 
emissions.
    Other commenters supported EPA's position that consideration of 
health benefits is not relevant under the regional haze program.
    One commenter stated that we should regulate coal ash at Colstrip. 
Another commenter expressed concern about acid rain, and one commenter 
stated that various pollutants such as dioxin and formaldehyde were 
byproducts of coal pollution.
    Response: We acknowledge the commenters' concerns regarding the 
negative health impacts of haze-causing emissions. We agree that the 
same PM2.5 emissions that cause visibility impairment can 
cause respiratory problems, decreased lung function, aggravated asthma, 
bronchitis, and premature death. We also agree that the same 
NOX emissions that cause visibility impairment also 
contribute to the formation of ground-level ozone, which has been 
linked with respiratory problems, aggravated asthma, and even permanent 
lung damage. We agree that these pollutants may have negative impacts 
on vegetation, and reduce crop yields. However, for purposes of this 
action, we are not authorized to consider these impacts in promulgating 
our FIP, and we have not done so. However, to the extent that this FIP 
will lead to reductions in these pollutants, there will be co-benefits 
for public health.
    We recognize the importance of considering environmental justice; 
for this action, we are finalizing emission limitations that will 
result in emissions reductions that will benefit potential 
environmental justice communities. Therefore, this action will have no 
high adverse and disproportionate impact on potential environmental 
justice communities.
    Mercury is not a visibility impairing pollutant, and was therefore 
not included in our analysis. We also are not authorized to regulate 
coal ash in this action.
    Comment: Some commenters noted that regional haze is not a health-
based standard, and that there are other recently enacted rules that 
protect human health.
    Response: We agree that the Regional Haze Rule was not intended to 
address health concerns. Regional Haze is not a health-based standard.

O. General Comments Supporting Our Proposal or for Stricter Controls

    Comment: NPCA and MATB commended EPA's required controls for the 
Ash Grove and Holcim cement kilns. The Northern Cheyenne Tribe 
expressed support of our proposal as a whole.
    Response: We acknowledge the support provided by these commenters.
    Comment: Overall, we received more than 47,000 comment letters from 
members representing various organizations and concerned citizens 
requesting that EPA mandate more stringent and effective controls, most 
notably SCR, on eligible Montana sources. These comments were received 
at the public hearings in Billings and Helena, Montana, by Internet, 
and through the mail. Many of these commenters argued that SCR is 
required at over 200 facilities in the U.S., and that SCR should 
therefore also be required at the coal-fired plants in Montana. A mass 
mailer from WEG claimed that SCR was shown to be cost-effective, but is 
not required. Several comments more generally stated that EPA should 
require the most modern, effective pollution controls on Montana 
sources, but did not specifically discuss the desired requirements. The 
Montana Conservation Voters pointed out that pollution from Colstrip 
will be three times higher than if SCR were required.
    Response: Although we acknowledge the commenters' encouragement 
that we adopt even stricter standards, the standards discussed in our 
proposal are appropriate considering the costs and visibility 
improvement.
    Comment: One commenter pointed out that Colstrip emits more 
pollutants than the nine next largest haze producers, combined.
    Response: The commenter did not explain specifically what they were 
requesting.
    Comment: A commenter pointed out that Colstrip 3 and 4 are as 
highly polluting as Colstrip 1 and 2, and thought that Colstrip 3 and 4 
should also be required to install additional controls.
    Response: As explained in our proposal, the modeled visibility 
benefits are not sufficient for us to consider it reasonable to impose 
additional controls for Colstrip units 3 and 4 for this planning 
period. 77 FR 24066 and 77 FR 24067.
    Comment: One commenter stated that the upgrading of pollution 
controls on coal-burning facilities also helps mitigate the effects of 
climate change. A separate commenter requested that EPA's plan consider 
CO2 because of its impacts on climate change, while another 
stated that coal should no longer be burned, as such action would slow 
global climate change.
    Response: While we understand the commenters' concerns with respect 
to climate change, consideration of climate change is outside the scope 
of this action. CO2 is a greenhouse gas (GHG) and is not 
considered a visibility impairing pollutant. However, EPA implements 
regulations that address GHGs in order to protect the public and the 
environment from the negative impacts of climate change.

P. General Comments That the Proposal Is Too Stringent

    Comment: Various commenters generally stated they did not support 
the proposed rulemaking. Their reasons included: It will negatively 
affect the local economy; it will negatively affect the coal power 
plant industry; electricity costs will increase; health

[[Page 57908]]

concerns are exaggerated; direct and indirect jobs/businesses would be 
adversely affected; the costs outweigh the benefits; Colstrip is 
already significantly regulated; there are no air quality issues in 
Colstrip; and it will not result in noticeable visibility improvements. 
One commenter insisted our proposal is part of a broader anti-coal plan 
to shut down coal plants, while another stated that Congress should 
legislate national energy policy rather than involving federal 
agencies. One commenter stated that PPL is very committed to clean air 
and environmental stewardship and another stated that Colstrip is 
already heavily regulated and additional controls are unnecessary. One 
commenter stated that mismanagement of forests causes more haze and 
that Colstrip provides good jobs and has a good compliance record.
    Response: We acknowledge these general comments that opposed our 
proposed action as being too stringent. We provide responses that 
address some of these issues elsewhere in this action. This action is 
based on the statutory and regulatory requirements for regional haze 
which we have followed.

Q. General Comments on Visibility Improvement and Other Causes of Haze

    Comment: Some commenters stated that any controls required by our 
action must demonstrate a perceptible visibility improvement and some 
stated that the reductions in the proposal will not produce perceptible 
visibility improvement. Other commenters said that there were no haze 
issues in Montana and that the change in visibility is subjective. The 
Montana Chamber of Commerce commented that our FIP is not based on 
sound science, accurate measures, or proven measures that will solve 
the problem.
    Some commenters stated that gravel roads and forest fire are the 
real causes of haze.\69\ WETA commented that under the FIP, haze would 
not be effectively reduced and EPA's regional haze plan should consider 
all established sources of emissions and not just industrial 
facilities. Another commenter suggested that money to clean up 
pollution should be spent in urban areas where there are real problems, 
not in rural areas like Montana. An individual submitted information 
comparing Montana emissions from different sources.
---------------------------------------------------------------------------

    \69\ One commenter also mentioned idling trucks, oil refineries 
and farms as causes of haze.
---------------------------------------------------------------------------

    One commenter noted that the proposed rule delays, by hundreds of 
years, in some cases, achievement of the 2064 natural visibility goal. 
Numerous commenters stated that EPA should not forego cost-effective 
pollution controls when more progress is clearly needed to protect air 
quality. Some commenters stated that there is currently haze at 
Yellowstone that was not visible years ago.
    With regard to Colstrip, a commenter said that shutting down 
Colstrip would not clear the haze and that areas outside Montana, 
including Oregon, Washington, and China influence the haze at 
Yellowstone. Another commenter stated that there is no haze in the town 
of Colstrip and that the wind does not blow in the directions of 
Yellowstone and Roosevelt.
    Response: We disagree that any controls required by our action must 
demonstrate a perceptible visibility improvement. In a situation where 
the installation of BART may not result in a perceptible improvement in 
visibility, the visibility benefit may still be significant. The 
Regional Haze Rule states ``even though the visibility improvement from 
an individual source may not be perceptible, it should still be 
considered in setting BART because the contribution to haze may be 
significant relative to other source contributions in the Class I area. 
Failing to consider less-than-perceptible contributions to visibility 
impairment would ignore the CAA's intent to have BART requirements 
apply to sources that contribute to, as well as cause, such 
impairment.'' 70 FR 39129. Visibility impacts below the thresholds of 
perceptibility cannot be ignored because regional haze is produced by a 
multitude of sources and activities which are located across a broad 
geographic area.
    We agree that industrial facilities are not the only causes of 
haze. Our action considered the many contributors to haze including 
industrial facilities. In this action, we also proposed changes to 
Montana's Visibility SIP that would require BACT for open burning.
    Even though some Class I areas will not attain natural visibility 
conditions by 2064, our action requires the controls that were 
determined to be effective according to our evaluation. For those 
sources subject to BART, we evaluated: (1) Cost of compliance, (2) the 
energy and nonair quality environmental impacts of compliance, (3) any 
existing pollution control technology in use at the source, (4) 
remaining useful life of source, and (5) degree of improvement in 
visibility which may reasonably be anticipated to result from the use 
of such technology and we determined which controls should be required 
according to that evaluation. In determining the measures necessary to 
make reasonable progress and in selecting RPGs for mandatory Class I 
areas within Montana, we took into account the following four factors: 
(1) Costs of compliance, (2) time necessary for compliance, (3) Energy 
and nonair quality environmental impacts of compliance; and (4) 
remaining useful life of any potentially affected sources. CAA section 
169A(g)(1) and 40 CFR 51.308(d)(1)(i)(A).
    For Colstrip, we evaluated visibility improvement at all Class I 
areas within 300 km. As stated above we evaluated other sources of 
haze, including but not limited to, gravel roads and forest fires. The 
most impacted Class I areas were Theodore Roosevelt NP and UL Bend WA. 
While sources outside Montana do contribute to haze in the Class I 
areas within Montana, that does not preclude our obligation to evaluate 
Colstrip Units 1 and 2 according to the five BART factors and to 
evaluate Colstrip Units 3 and 4 according to the four reasonable 
progress factors and to require additional controls where necessary.

R. Comments on Cost, Economic Impact, Jobs and Price to Consumers

    Comment: Some commenters stated that the proposed rule would have a 
negative economic impact and a negative impact on job creation and 
growth. Some commenters stated that PPL might shut down Colstrip Units 
1 and 2 as a result of this action. One commenter explained that 
shutting down power plants removes jobs, and prevents other businesses 
from using the energy from the power plant, causing a domino effect. A 
commenter submitted documents describing Colstrip's positive economic 
and community impact. Another commenter said that specifically, Montana 
has a large percentage of low income and senior citizens who would be 
majorly burdened by an increase in utility cost and another commenter 
said that the cost would also be very burdensome for the small business 
community in the area. The Southeastern Montana Development Corporation 
stated that the economic impact of this action would be devastating to 
consumers. One commenter said that the costs were prohibitively 
expensive and another said that the costs could put the plants at risk 
for future investments due to lack of economic viability. A commenter 
suggested that the initial cost of investment at Colstrip 1 and 2, 
including the cost of debt and capital, would be in excess of $82 
million and that the capital cost, plus operating cost of $377 million 
could result in a 19.6% increase in the cost of production. Another 
commenter suggested that the cost of electricity could increase by a

[[Page 57909]]

factor of 20 in 3-4 years. One commenter urged us to consider the 
indirect ways that controls on Colstrip 3 & 4 could affect electric 
rates. Numerous commenters stated that the reason EPA was not requiring 
SCR was to save polluters money.
    Other commenters said that the health costs of pollution and 
economic benefit from tourism should be considered. One commenter said 
that the health related costs from Colstrip are estimated to be $230 
million annually. Another commenter stated that air pollution controls 
are cost effective based on an EPA report. One commenter said that 
pollution hinders the Billings economy because the city's economic 
vitality is linked to high quality life-styles, while another noted 
that haze diminishes tourists' scenic vistas.
    Some commenters pointed out that the proposed rule would create 
jobs. One commenter stated that complying with the rule would create 
good, high-paying jobs for Montana's skilled work force, including 
boilermakers, laborers and pipefitters. Numerous commenters stated that 
nearly 1,000 full-time jobs could be created at Colstrip from 
installing pollution control equipment. One commenter said that the 
Colstrip plant will not shut down just because added technology is 
required.
    Many commenters expressed a willingness to pay more for power in 
support of pollution control technology. Others similarly stated that 
we should all pay the full cost of energy and not pass it on to 
healthcare. Some commenters stated that they thought PPL could afford 
to pay for additional controls based on the company's profit. A report 
submitted by Power Consulting, Inc. found that the typical residential 
customer's bill would increase by 55 to 89 cents if SCR were required 
on Colstrip unit 4. The overall conclusion from that report was that 
the impact of a required SCR retrofit on customer's rates would be 
small enough that it would not disrupt household budgets nor cause a 
significant impact on the Montana economy.
    Response: EPA's evaluation of capital and annual expenses 
associated with implementation of the FIP shows such expenses to be 
justified by the degree of improvement in visibility in relationship to 
the cost of implementation. BART requires that we evaluate: (1) Cost of 
compliance, (2) the energy and nonair quality environmental impacts of 
compliance, (3) any existing pollution control technology in use at the 
source, (4) remaining useful life of source, and (5) degree of 
improvement in visibility which may reasonably be anticipated to result 
from the use of such technology. In determining the measures necessary 
to make reasonable progress and in selecting reasonable progress goals 
for mandatory Class I areas within Montana, we must take into account 
the following four factors: (1) Costs of compliance, (2) time necessary 
for compliance, (3) Energy and nonair quality environmental impacts of 
compliance; and (4) remaining useful life of any potentially affected 
sources. CAA section 169A(g)(1) and 40 CFR 51.308(d)(1)(i)(A). The cost 
of electricity to consumers and the overall impact on the economy is 
outside the scope of our evaluation for this action.
    Although we did not consider the potential positive benefits to 
local economies in making our decision, we do expect that improved 
visibility would have a positive impact on tourism-dependent local 
economies. Also, the retrofits required are large construction projects 
that will take up to five years to complete. These projects will 
require well-paid, skilled labor which can potentially be drawn from 
the local area and support local growth.
    Comment: A commenter stated that EPA should have included, as 
associated per-unit costs, consideration of the ``wider market 
consequences'' of a potential shutdown of generating capacity at 
Colstrip 1 and 2. The commenter says that, ``[i]f the cost of 
production resulting from this rule * * * exceeds the market value of 
power, PPL may make a decision to shutter the plant.'' The commenter 
also states that, ``[b]ased on an analysis of production cost data, 
there is at least some chance that Colstrip Units 1 and 2 would become 
uneconomical as a result of mandated upgrades.'' Specifically, 
commenter estimated that the ``all-in'' cost of production of 
electricity post-controls is $25.591 per megawatt-hour, a 19.6% 
increase over the current $21.40 per megawatt-hour cost of production 
reported in Federal Regulatory Commission filings. Commenter stated 
that, compared to current market prices from a regional trade 
publication,\70\ Colstrip 1 and 2 would often be uneconomical at that 
estimated cost.
---------------------------------------------------------------------------

    \70\ Commenter cited the trade publication ``Clearing Up,'' 
which commenter stated reports on prices at the Mid-Columbia trading 
club.
---------------------------------------------------------------------------

    The commenter also argued that a closure at Colstrip 1 and 2 would 
decrease available electrical generation in the northwestern U.S. The 
commenter stated that we wrongly failed to consider these factors of 
potential plant closure and the subsequent constriction of power supply 
in our analyses.
    Response: Analyzing the wider market consequences of a potential 
shutdown of generating capacity at Colstrip 1 and 2 involves many 
complicated factors and it is unclear from the information provided by 
the commenter that Colstrip Units 1 and 2 would, in fact, shut down. As 
noted previously, we have received conflicting information regarding 
potential rate increases. Specifically, a report submitted by Power 
Consulting, Inc. found that the typical residential customer's bill 
would increase by 55 to 89 cents if SCR were required on Colstrip unit 
4. The BART Guidelines allow for the consideration of unusual 
circumstances that justify taking into consideration the conditions of 
the plant and the economic effects of requiring the use of a given 
control technology. The BART Guidelines state:

    [t]hese effects would include effects on product prices, the 
market share, and profitability of the source. Where there are such 
unusual circumstances that are judged to affect plant operations, 
you may take into consideration the conditions of the plant and the 
economic effects of requiring the use of a control technology. Where 
these effects are judged to have a severe impact on plant operations 
you may consider them in the selection process, but you may wish to 
provide an economic analysis that demonstrates, in sufficient detail 
for public review, the specific economic effects, parameters, and 
reasoning.

70 FR 39171. The commenter has not provided any basis that unusual 
circumstances exist here. Nor has the commenter providing any 
information that indicates a shutdown will occur that we could have 
taken into account in our analysis. The owners of Colstrip Units 1 and 
2 have made no indication that there are unusual circumstances present 
that warrant taking wider market consequences into consideration.

S. Comments About Other Forms of Energy

    Comment: We received comments regarding alternative forms of 
energy. Some commenters believed that wind energy would create more 
jobs while others believed that it would not create as many jobs 
compared to coal fired power plants. Some commenters stated that wind 
energy was cheaper to produce while one commenter pointed out that the 
government subsidizes wind energy. One commenter believed that the wind 
farm in Judith Gap produces energy more cheaply compared to the 
Colstrip coal plant. One commenter stated that our energy

[[Page 57910]]

should be focused on renewable sources rather than coal and another 
commenter stated that the most important thing we can do to slow global 
warming is to stop burning coal.
    Response: While we do generally acknowledge that many kinds of 
renewable energy do not produce haze-causing pollutants, and 
transitioning to those sources of energy could lead to visibility 
improvements. In this action we are required to review specific 
retrofit options for specific sources subject to BART or the sources 
analyzed under reasonable progress. Renewable energy technology is not 
a retrofit option for these sources and is outside the scope of our 
determinations and regulatory requirements in this action.

T. Other Miscellaneous Comments

    Comment: One commenter asked whether EPA was concerned that 
requiring these facilities to install emissions control equipment to 
address fine particles and precursors might impact the effectiveness of 
equipment installed to address other pollutants.
    Response: The control technologies that are required will not 
negatively impact the effectiveness of equipment installed to address 
other pollutants.
    Comment: One commenter asked whether the agency was concerned that 
the technologies prescribed to address particles and precursors might 
also impact the efficiency and reliability of kilns, boilers, 
generators and other essential equipment.
    Response: The control technologies required will not negatively 
impact the efficiency and reliability of kilns, boilers, generators and 
other essential equipment. As required under BART, we evaluated the 
energy impacts for each control option considered. 70 FR 39168 and 70 
FR 39169. These impacts are discussed in the relevant sections of the 
proposed rule and in all cases are minor. In addition, as required 
under BART, we evaluated the technical feasibility for each control 
option considered. Where we have selected additional controls, the 
controls are shown to be technically feasible at similar facilities. 
Issues associated with the reliability of the emission units, if any, 
are resolvable.
    Comment: MDEQ requested that EPA extend the comment period to sixty 
days from the date of the publication of corrections, or July 16, 2012.
    Response: The comment period for our proposal closed on June 19, 
2012. We carefully considered the request for an extension to the 
comment period. We took into consideration how an extension might 
affect our ability to consider comments received on the proposed action 
and still comply with our consent decree deadlines. We do note that our 
May 1, 2012, public hearing in Helena, Montana and May 2, 2012, public 
hearing in Billings, Montana were well attended and provided an 
opportunity for people to comment on our proposal. We also note that 
the corrections published May 17, 2012, (77 FR 29270) primarily amended 
typographical errors.\71\
---------------------------------------------------------------------------

    \71\ We corrected some technical information in the Holcim 
SO2 BART analysis. See 77 FR 29270.
---------------------------------------------------------------------------

    Comment: MDEQ suggested that EPA issue a request for additional 
comment to clarify the scope of the proposed FIP. MDEQ asserted that 
such a clarification is necessary to prevent confusion among the public 
regarding the Regional Haze Rule's prevention and correction of adverse 
health effects, about which EPA received multiple comments. MDEQ warned 
that ``the level of this misperception threatens to pervert not only 
the National Goal, but, ostensibly, the public health goals of Section 
110.''
    Response: We do not agree that the scope of the proposed FIP 
requires clarification. At no point in the proposed FIP did we discuss 
public health impacts as a consideration in our analyses, as they were 
not. As stated elsewhere, we agree that the Regional Haze Rule is not a 
health-based standard, and that we are not authorized to consider 
public health impacts in promulgating our FIP for purposes of this 
action. However, we have not been presented any information from the 
public to indicate that there is confusion that that reduction of 
visibility impairing pollutants also provides health benefits.
    Comment: One commenter stated that the Cheyenne Reservation was 
given Class I air quality designation and that according to that 
designation there is not supposed to be any degradation of that air.
    Response: The Regional Haze Rule requires analysis for the 156 
mandatory Class I areas listed at 40 CFR Part 81. The Cheyenne 
Reservation is not one of these federally mandated Class I areas.
    Comment: WEG stated that EPA overlooked, in two respects, the 
requirement of section 110(l) of the Act to prevent interference with 
attainment or maintenance of the NAAQS. First, WEG stated that EPA has 
not demonstrated that this FIP adequately safeguards the 2006 
PM2.5 NAAQS, the 2008 ozone NAAQS, the 2010 1-hour 
NO2 NAAQS, and the 2010 1-hour SO2 NAAQS. In 
particular, WEG noted that the FIP emissions limitations are generally 
expressed as 30-day rolling averages, which, in WEG's view, do not 
adequately protect short-term NAAQS such as the 2010 1-hour 
SO2 and NO2. Second, WEG argued that several BART 
emissions limitations are relaxations that may impact the NAAQS. As an 
example, WEG cited another portion of its comments in which WEG argued 
that the BART emissions limitations for Corette will allow actual 
emissions from Corette to increase. WEG concluded that EPA must conduct 
a 110(l) demonstration in order to protect public health and not 
interfere with maintenance and attainment of the NAAQS.
    Response: EPA disagrees with WEG. In relevant part, section 110(l) 
provides that EPA shall not approve a revision of a plan if the 
revision would interfere with any applicable requirement concerning 
attainment and reasonable further progress or any other applicable 
requirement of the CAA. First, WEG does not explain how section 110(l) 
applies to EPA's initial promulgation of a FIP for certain regional 
haze requirements when there is no existing SIP to meet those 
requirements. Second, to the extent that section 110(l) applies, EPA's 
promulgation of this FIP satisfies its requirements. It is EPA's 
consistent interpretation of section 110(l) that a SIP revision does 
not interfere with attainment and maintenance of the NAAQS if the 
revision at least preserves the status quo air quality by not relaxing 
or removing any existing emissions limitation or other SIP requirement. 
EPA does not believe that a full attainment or maintenance 
demonstration for each NAAQS is required for every SIP revision under 
section 110(l).
    In this case, the FIP imposes new emissions limitations on a number 
of existing sources, and it does not relax any existing emissions 
limitations or other SIP requirements. WEG's statement that actual 
emissions at Corette and other BART sources might rise to the BART 
limit misses the point: In the absence of the BART limit (or any other 
limit), those actual emissions could increase much more. In other 
words, imposing an emissions limitation where one did not exist before 
is necessarily a more stringent requirement, regardless of actual 
emissions. Nor does WEG explicitly identify any existing emissions 
limitation or other SIP requirement that is relaxed by the FIP. For 
that matter, nothing in the proposal, or in the preamble or regulatory 
text for this rule, purports to modify any existing SIP-approved 
emissions limitation or other SIP requirement. Thus, even if there were 
such a requirement--and WEG has identified none--it would not be

[[Page 57911]]

relaxed by this FIP. EPA therefore concludes that, to the extent that 
section 110(l) is applicable to this FIP, its requirements are 
satisfied.
    Comment: Commenter stated that the input of Montana residents 
should be given more weight than the input of special interest groups 
that receive support from outside the State. Commenter also requested 
that future hearings be held in areas of impact.
    Response: Any commenter who submits a comment on the proposed FIP, 
either orally or written, during the public comment period is entitled 
to do so. EPA takes all comments into consideration in making its final 
decision on the FIP. If future hearings are required for any reason, we 
will do the best we can to ensure access is available to all those who 
wish to participate.

V. Changes From Proposed Rule and Reasons for the Changes

A. Emission Limits for Corette

    We proposed a PM emission limit of 0.10 lb/MMBtu for Corette at 40 
CFR 52.1396(c). We inadvertently stated that we were imposing an 
emission limit of 0.10 lb/MMBtu in the preamble to our proposed FIP (77 
FR 24047) and also at 40 CFR 53.1396(c)(1). PPL commented that the 
emission limit in the proposed FIP was flawed and PPL provided 
additional information indicating that over the past five years, stack 
test results have shown that PM emissions have ranged from 0.059 lb/
MMBtu to 0.252 lb/MMBtu. We have changed the emission limit in the 
final regulatory requirements at 40 CFR 1396(c)(1). In the final FIP, 
we are establishing a PM emission limit of 0.26 lb/MMBtu.
    We proposed a SO2 emission limit of 0.70 lb/MMBtu and a 
NOX emission limit of 0.40 lb/MMBtu for Corette at 40 CFR 
52.1396(c). In the final FIP, we are establishing a SO2 
emission limit of 0.57 lb/MMBtu and a NOX emission limit of 
0.35 lb/MMBtu. We have made this change as a result of the comments we 
received. One commenter stated that EPA must increase the limits to no 
less than 0.81 lb/MMBtu for SO2 and 0.46 lb/MMBtu for 
NOX in order to account for compliance over a 30-day rolling 
average. By contrast, another commenter stated that our proposed 
emission limits were too high and would actually result in increased 
emissions.
    Based on these comments, we have reassessed the SO2 and 
NOX emission limits for Corette. In order to establish 
appropriate emission limits, we conducted a statistical analysis of the 
monthly emissions data contained in the CAMD emissions system. For the 
period 2000-2010, the 99th percentile monthly SO2 emission 
rate was 0.548 lb/MMbtu. Similarly, the 99th percentile monthly 
NOX emission rate was 0.335 lb/MMBtu. In our final action, 
we are establishing emission limits slightly above these 99th 
percentile emission rates in order to allow a sufficient margin for 
compliance. This is because the emission limits must apply at all 
times, including during startup, shutdown, and malfunction. The revised 
emission rates are 0.57 lb/MMBtu for SO2 and 0.35 lb/MMBtu 
for NOX, both on a 30-day rolling average. We have revised 
the emission limits for Corette contained in section 52.1396(c)(1) 
accordingly.

B. Changes to 40 CFR 52.1396(c)(2)--Emission Limitations for Cement 
Kilns

    In response to a comment from Holcim that EPA failed to consider 
the NOX control technology already installed at the Trident 
cement plant, and that EPA failed to give proper weight to the 
excessively high average cost-effectiveness ($4,279/ton) and 
incremental cost-effectiveness ($8,029/ton) of a switch to indirect 
firing and a Low-NOX Burner (LNB), we have removed switching 
to indirect firing and a LNB from consideration as an option for 
further reducing NOX emissions and are treating any 
NOX emission reduction that may have been achieved from 
installation of a new burner as part of the emissions baseline. We have 
recalculated the BART limit for NOX to reflect a 50% 
reduction in NOX emissions from that baseline by addition of 
SNCR alone, rather than the 58% reduction we previously used, which 
reflected a switch to indirect firing and LNB plus SNCR. The 
recalculated NOX BART limit is 6.5 lb/ton clinker. We have 
replaced the NOX emission limit of 5.5 lb/ton clinker from 
our proposal with 6.5 lb/ton clinker, on a 30-day rolling average.
    Also, during our evaluation of comments on PM BART from Ash Grove, 
we found that the table of emission limits for cement kilns, at section 
52.1396(c)(2) of our proposal, needed to clarify that the PM emission 
limit for Ash Grove is in lb/hr, not lb/ton clinker. Only the PM 
emission limit for Holcim is in lb/ton clinker. The column header for 
PM emission limits for both cement kilns erroneously said ``lb/ton 
clinker.'' We have corrected this error by changing the header from 
``PM Emission Limit (lb/ton clinker)'' to ``PM Emission Limit.'' We did 
not change the text of the PM emission limit for Ash Grove, as it is 
already clear in that text that the limit is in lb/hr. However, at the 
bottom of the column, we have clarified the PM emission limit for 
Holcim to say ``0.77 lb/ton clinker'' rather than``0.77 lb/ton.''

C. Change to 40 CFR 52.1396(d)--Compliance Date

    In response to a comment from Ash Grove which identified the 
failure of our regulatory text at 40 CFR 52.1396(d) to specify the 
SO2 and PM compliance dates described in the preamble to our 
proposed rule, we have revised 40 CFR 52.1396(d) to read as follows:

    The owners and operators of the BART sources subject to this 
section shall comply with the emissions limitations and other 
requirements of this section as follows, unless otherwise indicated 
in specific paragraphs: Compliance with PM limits is required within 
30 days of the effective date of this rule. Compliance with 
SO2 and NOX limits is required within 180 days 
of the effective date of this rule, unless installation of 
additional emission controls is necessary to comply with emission 
limitations under this rule, in which case compliance is required 
within five years of the effective date of this rule.

D. Change to 40 CFR 52.1396(e)(3)--CEMS for Cement Kilns

    In response to a comment from Ash Grove Cement that this section 
should be revised to include an exception from CEMS data collection 
during CEMS breakdowns, repairs, calibration checks and zero and span 
adjustments, we have added the following language from 40 CFR part 60, 
subpart F, New Source Performance Standards for cement kilns, at 40 CFR 
60.63(b):

    You must operate the monitoring system and collect data at all 
required intervals at all times the affected source is operating, 
except for periods of monitoring systems malfunctions, repairs 
associated with monitoring system malfunctions, and required 
monitoring system quality assurance or quality control activities 
(including, as applicable, calibration checks and required zero and 
span adjustments).

    Also, during our evaluation of comments from Ash Grove on CEMS 
requirements, we found that section 52.1396(e)(3) inadvertently failed 
to cross-reference the requirements for CEMS for cement kilns at 40 CFR 
60.63(g). Section 52.1396(e)(3) only cross-referenced 60.63(f). There 
are important requirements for cement kiln CEMSs at 40 CFR 60.63(g), as 
well as important CEMS requirements at 60.63(h) which are cross-
referenced only by 60.63(g) and not by 60.63(f). We have therefore 
added ``and (g),'' such that the first sentence of section 
52.1396(e)(3) now reads as follows:

    At all times after the compliance date specified in paragraph 
(d) of this section, the owner/operator of each unit shall maintain,

[[Page 57912]]

calibrate, and operate a CEMS, in full compliance with the 
requirements found at 40 CFR 60.63(f) and (g), to accurately measure 
concentration by volume of SO2 and NOX 
emissions into the atmosphere from each unit.

E. Change to 40 CFR 52.1396(e)(4)(ii)--Compliance Determination Methods 
for SO2 and NOX at Cement Kilns

    In response to a comment from Ash Grove that the formula at section 
52.1396(e)(4)(ii) of our proposal incorrectly expresses the 
concentrations of SO2 and NOX in grains per dry 
standard cubic foot, rather than in parts per million, we have deleted 
the equation E = (CsQs)/(PK) from this section, as well as the 
definitions of terms in that equation, and replaced it with the 
following equation, which appears in the proposed amendments to 40 CFR 
part 60, subpart F, New Source Performance Standards for cement kilns, 
published in the Federal Register on July 18, 2012:
[GRAPHIC] [TIFF OMITTED] TR18SE12.068


Where:

ED = 30 kiln operating day average emission rate of 
NOX or SO2, lb/ton of clinker;
Ci = Concentration of NOX or SO2 
for hour i, ppm;
Qi = volumetric flow rate of effluent gas for hour i, 
where
Ci and Qi are on the same basis (either wet or 
dry), scf/hr;
Pi = total kiln clinker produced during production hour 
i, ton/hr;
k = conversion factor, 1.194 x 10-7 for NOX 
and 1.660 x 10-7 for SO2
n = number of kiln operating hours over 30 kiln operating days, n = 
1 to 720.
For each kiln operating hour for which you do not have at least one 
valid 15-minute CEMS data value, use the average emissions rate (lb/
hr) from the most recent previous hour for which valid data are 
available.

F. Change to 40 CFR 52.1396(f)(1) and (f)(2)--Compliance Determinations 
for PM BART Limits at EGUs and Cement Kilns

    In response to a verbal comment from Holcim, in a meeting with EPA 
in June of 2012 on the proposed FIP, that BART sources should be 
allowed to retain the PM stack testing schedule already established 
under State permits, we have added the following sentence, after the 
sentence in sections 52.1396(f)(1) and (f)(2) that requires the first 
annual PM performance stack test for PM within 60 days after the PM 
compliance deadline:

    The results from a stack test meeting the requirements of this 
paragraph that was completed within 12 months prior to the 
compliance deadline can be used in lieu of the first stack test 
required. If this option is chosen, then the next annual stack test 
shall be due no more than 12 months after the stack test that was 
used.

    The meeting between Holcim and EPA is documented in the docket for 
this rulemaking.

G. Change to 40 CFR 52.1396(f)(2)--Compliance Determinations for Cement 
Kiln PM BART Limits

    Consistent with our clarification of the table of PM emission 
limits for cement kilns at 40 CFR 52.1396(c)(2), we have clarified 40 
CFR 52.1396(f)(2), to indicate that the emission rate of PM shall be 
reported in lb/hr for Ash Grove and in lb/ton clinker for Holcim. We 
have also clarified that the average of the results of three test runs 
for PM shall be used for demonstrating compliance. Specifically, we 
have added the following language after the third sentence of section 
52.1396(f)(2):

    The average of the results of three test runs shall be used for 
demonstrating compliance. For Ash Grove, the emission rate of 
particulate matter shall be computed for each run in pounds per hour 
(lb/hr). For Holcim, the emission rate (E) of particulate matter 
shall be computed for each run in lb/ton clinker, using the 
following equation: * * *
    We have also revised section 52.1396(f)(2) in response to a comment 
from Ash Grove that the equation at 40 CFR 52.1396(e)(4)(ii), cross-
referenced by this section 52.1396(f)(2), for calculating emissions in 
lb/ton clinker, is not valid for calculating SO2 and 
NOX emissions, but is only valid for calculating PM 
emissions. Therefore, we have moved this equation from section 
52.1396(e)(4)(ii) to section 52.1396(f)(2). We have also changed the 
pollutant in the equation to PM. We have also clarified (as explained 
above) that the equation is to be used for calculating PM in lb/ton 
clinker only for Holcim, not for Ash Grove (which, as explained above, 
is subject to a PM emission limit in lb/hr, not in lb/ton clinker). 
Below is the equation we have now inserted into section 52.1396(f)(2), 
immediately after the revised text described above:

E = (CsQs)/(PK)

Where:

E = emission rate of PM, lb/ton of clinker produced
Cs = concentration of PM in grains per standard cubic foot (gr/scf)
Qs = volumetric flow rate of effluent gas, where Cs and Qs are on 
the same basis (either wet or dry), scf/hr
P = total kiln clinker production rate, tons/hr, and
K = conversion factor, 7000 gr/lb

    We have also deleted the cross-reference to section 
52.1396(e)(4)(ii) for this equation.

H. Change to 40 CFR 52.1396(h)(6)--Recordkeeping Requirements for 
Cement Kilns

    In response to a comment from Ash Grove that the reference to ``40 
CFR Part 75'' should be deleted because Part 75 applies only to 
electrical generating units, not to cement kilns, we have deleted that 
reference. We note that since the monitoring requirements for cement 
kilns in the FIP, at 40 CFR 52.1396(e)(3) and (4), and at 40 CFR 
52.1396(f)(2), do not cross-reference Part 75, there are no applicable 
Part 75 recordkeeping requirements in the FIP. Section 52.1396(h)(6) 
now reads as follows:

    Any other records required by 40 CFR part 60, subpart F, or 40 
CFR part 60, Appendix F, Procedure 1.

I. Changes to 40 CFR 52.1396(i)--Reporting

    In response to a comment from Ash Grove that the first sentence of 
this section mistakenly references 40 CFR 53.1395(n) and (o), rather 
than 52.1396(n) and (o), we have made the correction.

J. Change to 40 CFR 52.1396(i)(1) and (i)(2)--Reporting for CEMS for 
SO2 and NOX

    In response to a comment from Ash Grove that the reporting 
frequency for CEMS excess emission reports and CEMS performance reports 
for cement kilns should be changed from quarterly to semiannual, 
because reporting requirements under other programs (Title V and 
NESHAP) only require semiannual reporting, we have changed the 
frequency to semiannual, but have kept the frequency at quarterly for 
EGUs.
    We note that the general provisions of NSPS subpart A, at 40 CFR 
60.7(c), which we used as a template for our FIP provisions for CEMS 
reporting, require semiannual excess emission reports and monitoring 
system performance reports, except when more frequent reporting is 
specifically required by an applicable subpart, or if the 
Administrator, on a case-by-case basis, determines that more frequent 
report is necessary to accurately assess the compliance status of the 
source. NSPS subpart F for cement kilns does not specify more frequent 
reporting.
    Therefore, we have deleted ``quarterly'' from the first sentence of 
section 52.1396(i)(1) and from the first sentence of section 
52.1396(i)(2). After the first sentence in each of those

[[Page 57913]]

sections, we have inserted the following sentence: ``Reports shall be 
submitted quarterly for EGUs and semiannually for cement kilns.''

K. Changes to 40 CFR 52.1396 for Devon Energy, Blaine County 1 
Compressor Station

    In the final FIP, we are clarifying testing requirements, 
monitoring, recordkeeping and reporting requirements, and emission 
limitations for Devon Energy, Blaine County 1 Compressor 
Station. We made these changes in response to a comment stating that 
the requirements for this source were not practically enforceable.
    We have changed the text at 40 CFR 52.1396(c)(3) to read, ``The 
owners/operators of LP, Blaine County 1 Compressor Station 
shall not emit or cause to be emitted from each 5,500 horsepower 
Ingersoll Rand 616 natural gas-fired compressor engine installed at the 
facility, total NOX in excess of 21.8 lbs/hr (average of 
three stack test runs).'' We have made this change to clarify that the 
emission limit of 21.8 lbs/hr applies to each of the 5,500 horsepower 
Ingersoll Rand 616 natural gas-fired compressor engines installed at 
the facility and that the emission rate will be determined by averaging 
the results of three stack test runs.
    We have changed the text at 40 CFR 52.1396(e)(5) to read, ``The 
owner/operator of Blaine County 1 Compressor Station shall 
install a temperature-sensing device (i.e. thermocouple or resistance 
temperature detectors) before the catalyst in order to monitor the 
inlet temperatures of the catalyst for each engine. The owner/operator 
shall maintain the exhaust temperature at the inlet to the catalyst for 
each engine at a minimum of least 750 [deg]F and no more than 1250 
[deg]F in accordance with the catalyst manufacturer's specifications. 
Also, the owner/operator shall install gauges before and after the 
catalyst for each engine in order to monitor pressure drop across the 
catalyst, and that the owner/operator maintain the pressure drop within 
 2'' water at 100% load plus or minus 10% from the pressure 
drop across the catalyst measured during the initial performance test. 
The owner/operator shall follow the manufacturer's recommended 
maintenance schedule and procedures for each engine and its respective 
catalyst. The owner/operator shall only fire each engine with natural 
gas that is of pipeline-quality in all respects except that the 
CO2 concentration in the gas shall not be required to be 
within pipeline-quality.'' We have made this change to clarify that it 
is the exhaust temperature that must be maintained at a minimum of at 
least 750 [deg]F and no more than 1250 [deg]F in accordance with the 
catalyst manufacturer's specifications, and not the engine temperature 
that must be kept within this temperature range. We are also making 
this change to clarify that the temperature range must be kept in 
accordance with the catalyst manufacturer's specifications and not the 
engine manufacturer's specifications.
    We have added a new section, 40 CFR 52.1396(j) which includes 
testing requirements for Blaine County 1 Compressor Station. 
This section was inadvertently omitted from the proposed FIP, but is 
necessary to ensure adequate testing is performed to ensure compliance 
with the NOX emission limit for Blaine County 1 
Compressor Station.
    We have changed 40 CFR 52.1396(k)(1) to read: ``The owner/operator 
shall measure NOX emissions from each engine at least semi-
annually or once every six-month period to demonstrate compliance with 
the emission limits. To meet this requirement, the owner/operator shall 
measure NOX emissions from each engine using a portable 
analyzer and a monitoring protocol approved by EPA.'' We have changed 
the first sentence from referring to engines to refer to each engine to 
clarify that NOX emissions must be measured from each 
engine.
    We have added a new paragraph at 40 CFR 52.1396(k)(9) to read, 
``The owner/operator shall keep records of all deviations from the 
emission limit or operating requirements (e.g., catalyst inlet 
temperature, pressure drop across the catalyst) for each engine. The 
records shall include: The date and time of the deviation, the name and 
title of the observing employee and a brief description of the 
deviation and the measures taken to address the deviation and prevent 
future occurrences.'' We have made this change to ensure that adequate 
records are kept by the owner or operator of Blaine County 1 
Compressor Station to demonstrate compliance with the required emission 
limit and appropriate operation of the NSCR system.
    We have changed the text of 40 CFR 52.1396(k)(10) to correct a 
typographical error and to add to the requirements that the owner/
operator of Blaine County 1 Compressor Station must maintain 
records of deviations from operating requirements for a period of at 
least five years and that these records must be made available upon 
request by EPA.

VI. Statutory and Executive Order Reviews

A. Executive Order 12866: Regulatory Planning and Review and Executive 
Order 13563: Improving Regulation and Regulatory Review 13563

    This action will finalize a SIP approval for a revision to 
Montana's Smoke Management plan and a source-specific Regional Haze FIP 
for imposing federal controls to meet BART requirements for PM, 
NOX and SO2 emissions on five specific units at 
four sources in Montana (Ash Grove, Holcim, Colstrip Units 1 and 2, and 
Corette) and imposing controls to meet RP requirements for 
NOX emissions at one additional source (Devon) in Montana. 
The net result of the FIP action is that EPA is proposing direct 
emission controls on selected units at five sources. The sources in 
question are two large electric generating plants (one plant includes 
two units), two cement plants, and one gas compressor station. This 
action also imposes notification requirements on CFAC and M2Green 
Redevelopment LLC. This type of action is exempt from review under 
Executive Orders 12866 (58 FR 51735, October 4, 1993) and 13563 (76 FR 
3821, January 21, 2011).

B. Paperwork Reduction Act

    This action does not impose an information collection burden under 
the provisions of the Paperwork Reduction Act, 44 U.S.C. 3501 et seq. 
Under the Paperwork Reduction Act, a ``collection of information'' is 
defined as a requirement for ``answers to * * * identical reporting or 
recordkeeping requirements imposed on ten or more persons * * *. '' 44 
U.S.C. 3502(3)(A). Because the FIP applies to just seven sources, the 
Paperwork Reduction Act does not apply. See 5 CFR 1320(c).
    Burden means the total time, effort, or financial resources 
expended by persons to generate, maintain, retain, or disclose or 
provide information to or for a federal agency. This includes the time 
needed to review instructions; develop, acquire, install, and utilize 
technology and systems for the purposes of collecting, validating, and 
verifying information, processing and maintaining information, and 
disclosing and providing information; adjust the existing ways to 
comply with any previously applicable instructions and requirements; 
train personnel to be able to respond to a collection of information; 
search data sources; complete and review the collection of information; 
and transmit or otherwise disclose the information.

[[Page 57914]]

    An agency may not conduct or sponsor, and a person is not required 
to respond to a collection of information unless it displays a 
currently valid Office of Management and Budget (OMB) control number. 
The OMB control numbers for our regulations in 40 CFR are listed in 40 
CFR Part 9.

C. Regulatory Flexibility Act

    The Regulatory Flexibility Act (RFA) generally requires an agency 
to prepare a regulatory flexibility analysis of any rule subject to 
notice and comment rulemaking requirements under the Administrative 
Procedure Act or any other statute unless the agency certifies that the 
rule will not have a significant economic impact on a substantial 
number of small entities. Small entities include small businesses, 
small organizations, and small governmental jurisdictions.
    For purposes of assessing the impacts of today's rule on small 
entities, small entity is defined as: (1) A small business as defined 
by the Small Business Administration's (SBA) regulations at 13 CFR 
121.201; (2) a small governmental jurisdiction that is a government of 
a city, county, town, school district or special district with a 
population of less than 50,000; and (3) a small organization that is 
any not-for-profit enterprise which is independently owned and operated 
and is not dominant in its field.
    After considering the economic impacts of this action on small 
entities, I certify that this action will not have a significant 
economic impact on a substantial number of small entities. The Regional 
Haze FIP that EPA is finalizing consists of imposing federal controls 
to meet BART and RP requirements for PM, NOX and 
SO2 emissions on specific sources as described above in 
section A. None of these sources are owned by small entities, and 
therefore are not small entities.

D. Unfunded Mandates Reform Act (UMRA)

    This rule does not contain a Federal mandate that may result in 
expenditures of $100 million or more for state, local, and tribal 
governments, in the aggregate, or the private sector in any one year. 
Table 1 notes that the cumulative total annual costs for this action 
are $13.7 million. Thus, this rule is not subject to the requirements 
of sections 202 or 205 of UMRA.
    This rule is also not subject to the requirements of section 203 of 
UMRA because it contains no regulatory requirements that might 
significantly or uniquely affect small governments.

E. Executive Order 13132: Federalism

    This action does not have federalism implications. It will not have 
substantial direct effects on the states, on the relationship between 
the national government and the states, or on the distribution of power 
and responsibilities among the various levels of government, as 
specified in Executive Order 13132, because it merely addresses the 
State of Montana not meeting its obligation to adopt a SIP that meets 
the regional haze requirements under the CAA. Thus, Executive Order 
13132 does not apply to this action. In the spirit of Executive Order 
13132, and consistent with EPA policy to promote communications between 
EPA and state and local governments, EPA specifically solicited comment 
on this rule from state and local officials. A summary of each comment 
and EPA's response to those comments is provided in section IV of this 
preamble.

F. Executive Order 13175: Consultation and Coordination With Indian 
Tribal Governments

    This action does not have tribal implications, as specified in 
Executive Order 13175 (65 FR 67249, November 9, 2000). This action 
applies to only seven sources in Montana. Thus, Executive Order 13175 
does not apply to this rule. Although Executive Order 13175 does not 
apply to this action, EPA did send letters, dated October 7, 2011, to 
each of the Montana tribes explaining our regional haze FIP action and 
offering consultation. We did not receive any written or verbal 
requests from the Montana tribes for more information or for 
consultation. As a follow-up to our letter, we invited all of the 
tribes to a January 5, 2012 conference call. The call was attended by 
tribal Air Program Managers and one Environmental Director from tribes 
from four reservations. We also met with the Montana tribes prior to 
the start of the public hearings held in Helena and Billings, Montana. 
EPA specifically solicited additional comment on this rule from tribal 
officials and we received comments and responded to them in section IV 
of this preamble.

G. Executive Order 13045: Protection of Children From Environmental 
Health Risks and Safety Risks

    Executive Order 13045: Protection of Children from Environmental 
Health Risks and Safety Risks (62 FR 19885, April 23, 1997), applies to 
any rule that: (1) Is determined to be economically significant as 
defined under Executive Order 12866; and (2) concerns an environmental 
health or safety risk that we have reason to believe may have a 
disproportionate effect on children. EPA interprets EO 13045 as 
applying only to those regulatory actions that concern health or safety 
risks, such that the analysis required under section 5-501 of the EO 
has the potential to influence the regulation. This action is not 
subject to EO 13045 because it implements specific standards 
established by Congress in statutes. However, to the extent this rule 
limits emissions of NOX, SO2, and PM, the rule 
will have a beneficial effect on children's health by reducing air 
pollution.

H. Executive Order 13211: Actions Concerning Regulations That 
Significantly Affect Energy Supply, Distribution, or Use

    This action is not subject to Executive Order 13211 (66 FR 28355 
(May 22, 2001)), because it is not a significant regulatory action 
under Executive Order 12866.

I. National Technology Transfer and Advancement Act

    Section 12 of the National Technology Transfer and Advancement Act 
(NTTAA) of 1995 requires federal agencies to evaluate existing 
technical standards when developing a new regulation. To comply with 
NTTAA, EPA must consider and use ``voluntary consensus standards'' 
(VCS) if available and applicable when developing programs and policies 
unless doing so would be inconsistent with applicable law or otherwise 
impractical.
    The EPA believes that VCS are inapplicable to this action. Today's 
action does not require the public to perform activities conducive to 
the use of VCS.

J. Executive Order 12898: Federal Actions To Address Environmental 
Justice in Minority Populations and Low-Income Populations

    Executive Order 12898 (59 FR 7629, February 16, 1994), establishes 
federal executive policy on environmental justice. Its main provision 
directs federal agencies, to the greatest extent practicable and 
permitted by law, to make environmental justice part of their mission 
by identifying and addressing, as appropriate, disproportionately high 
and adverse human health or environmental effects of their programs, 
policies, and activities on minority populations and low-income 
populations in the United States.
    We have determined that this rule will not have disproportionately 
high and adverse human health or environmental effects on minority or

[[Page 57915]]

low-income populations because it increases the level of environmental 
protection for all affected populations without having any 
disproportionately high and adverse human health or environmental 
effects on any population, including any minority or low-income 
population. This rule limits emissions of NOX, 
SO2, and PM from five sources in Montana.

K. Congressional Review Act

    The Congressional Review Act, 5 U.S.C 801 et seq., as added by the 
Small Business Regulatory Enforcement Fairness Act of 1996, generally 
provides that before a rule may take effect, the agency promulgating 
the rule must submit a rule report, which includes a copy of the rule, 
to each House of the Congress and to the Comptroller General of the 
United States. Section 804 exempts from section 801 the following types 
of rules (1) rules of particular applicability; (2) rules relating to 
agency management or personnel; and (3) rules of agency organization, 
procedure, or practice that do not substantially affect the rights or 
obligations of non-agency parties. 5 U.S.C 804(3). EPA is not required 
to submit a rule report regarding today's action under section 801 
because this action is a rule of particular applicability. This rule 
finalizes a FIP for seven sources.

L. Judicial Review

    Under section 307(b)(1) of the CAA, petitions for judicial review 
of this action must be filed in the United States Court of Appeals for 
the appropriate circuit by November 19, 2012. Pursuant to CAA section 
307(d)(1)(B), this action is subject to the requirements of CAA section 
307(d) as it promulgates a FIP under CAA section 110(c). Filing a 
petition for reconsideration by the Administrator of this final rule 
does not affect the finality of this action for the purposes of 
judicial review nor does it extend the time within which a petition for 
judicial review may be filed, and shall not postpone the effectiveness 
of such rule or action. This action may not be challenged later in 
proceedings to enforce its requirements. See CAA section 307(b)(2).

List of Subjects in 40 CFR Part 52

    Environmental protection, Air pollution control, Intergovernmental 
relations, Incorporation by Reference, Nitrogen dioxides, Particulate 
matter, Reporting and recordkeeping requirements, Sulfur dioxide, 
Volatile organic compounds.

    Dated: August 15, 2012.
Lisa P. Jackson,
Administrator.

    40 CFR part 52 is amended as follows:

PART 52--[AMENDED]

0
1. The authority citation for part 52 continues to read as follows:

    Authority:  42 U.S.C. 7401 et seq.

Subpart BB--Montana

0
2. Section 52.1370 is amended by revising paragraph (c)(27)(i)(H) to 
read as follows:


Sec.  52.1370  Identification of plan.

* * * * *
    (c) * * *
    (27) * * *
    (i) * * *
    (H) Appendix G-2, Montana Smoke Management Plan, effective April 
15, 1988, is removed and replaced by Sec.  52.1395.
* * * * *

0
3. Add section 52.1395 to read as follows:


Sec.  52.1395  Smoke management plan.

    The Department considers smoke management techniques for 
agriculture and forestry management burning purposes as set forth in 40 
CFR 51.308(d)(3)(v)(E). The Department considers the visibility impact 
of smoke when developing, issuing, or conditioning permits and when 
making dispersion forecast recommendations through the implementation 
of Title 17, Chapter 8, subchapter 6, ARM, Open Burning.

0
4. Add section 52.1396 to read as follows:


Sec.  52.1396  Federal implementation plan for regional haze.

    (a) Applicability. This section applies to each owner and operator 
of the following coal fired electric generating units (EGUs) in the 
State of Montana: PPL Montana, LLC, Colstrip Power Plant, Units 1, 2; 
and PPL Montana, LLC, JE Corette Steam Electric Station. This section 
also applies to each owner and operator of cement kilns at the 
following cement production plants: Ash Grove Cement, Montana City 
Plant; and Holcim (US) Inc. Cement, Trident Plant. This section also 
applies to each owner or operator of Blaine County 1 
Compressor Station. This section also applies to each owner and 
operator of CFAC and M2 Green Redevelopment LLC, Missoula site.
    (b) Definitions. Terms not defined below shall have the meaning 
given them in the Clean Air Act or EPA's regulations implementing the 
Clean Air Act. For purposes of this section:
    Boiler operating day means a 24-hour period between 12 midnight and 
the following midnight during which any fuel is combusted at any time 
in the EGU. It is not necessary for fuel to be combusted for the entire 
24-hour period.
    Continuous emission monitoring system or CEMS means the equipment 
required by this section to sample, analyze, measure, and provide, by 
means of readings recorded at least once every 15 minutes (using an 
automated data acquisition and handling system (DAHS)), a permanent 
record of SO2 or NOX emissions, other pollutant 
emissions, diluent, or stack gas volumetric flow rate.
    Kiln operating day means a 24-hour period between 12 midnight and 
the following midnight during which the kiln operates.
    NOX means nitrogen oxides.
    Owner/operator means any person who owns or who operates, controls, 
or supervises an EGU identified in paragraph (a) of this section.
    PM means filterable total particulate matter.
    SO2 means sulfur dioxide.
    Unit means any of the EGUs or cement kilns identified in paragraph 
(a) of this section.
    (c) Emissions limitations. (1) The owners/operators of EGUs subject 
to this section shall not emit or cause to be emitted PM, 
SO2 or NOX in excess of the following 
limitations, in pounds per million British thermal units (lb/MMBtu), 
averaged over a rolling 30-day period for SO2 and 
NOX:

----------------------------------------------------------------------------------------------------------------
                                                                    PM emission    SO2 emission    NOX emission
                           Source name                              limit  (lb/     limit  (lb/     limit  (lb/
                                                                      MMBtu)          MMBtu)          MMBtu)
----------------------------------------------------------------------------------------------------------------
Colstrip Unit 1.................................................            0.10            0.08            0.15
Colstrip Unit 2.................................................            0.10            0.08            0.15
JE Corette Unit 1...............................................            0.26            0.57            0.35
----------------------------------------------------------------------------------------------------------------


[[Page 57916]]

     (2) The owners/operators of cement kilns subject to this section 
shall not emit or cause to be emitted PM, SO2 or 
NOX in excess of the following limitations, in pounds per 
ton of clinker produced, averaged over a rolling 30-day period for 
SO2 and NOX:

----------------------------------------------------------------------------------------------------------------
                                                                                   SO2 emission    NOX emission
                Source name                           PM emission limit           limit  (lb/ton  limit  (lb/ton
                                                                                     clinker)        clinker)
----------------------------------------------------------------------------------------------------------------
Ash Grove Cement...........................  If the process weight rate of the              11.5             8.0
                                              kiln is less than or equal to 30
                                              tons per hour, then the emission
                                              limit shall be calculated using E
                                              = 4.10p \0.67\ where E = rate of
                                              emission in pounds per hour and p
                                              = process weight rate in tons per
                                              hour; however, if the process
                                              weight rate of the kiln is greater
                                              than 30 tons per hour, then the
                                              emission limit shall be calculated
                                              using E = 55.0p\0.11\ -40, where E
                                              = rate of emission in pounds per
                                              hour and P = process weight rate
                                              in tons per hour.
Holcim (US) Inc............................  0.77 lb/ton clinker................             1.3             6.5
----------------------------------------------------------------------------------------------------------------

     (3) The owners/operators of LP, Blaine County 1 
Compressor Station shall not emit or cause to be emitted from each 
5,500 horsepower Ingersoll Rand 616 natural gas-fired compressor engine 
installed at the facility total NOX in excess of 21.8 lbs/hr 
(average of three stack test runs).
    (4) These emission limitations shall apply at all times, including 
startups, shutdowns, emergencies, and malfunctions.
    (d) Compliance date. The owners and operators of Blaine County 
1 Compressor Station shall comply with the emissions 
limitation and other requirements of this section as expeditiously as 
practicable, but no later than July 31, 2018. The owners and operators 
of the BART sources subject to this section shall comply with the 
emissions limitations and other requirements of this section as 
follows, unless otherwise indicated in specific paragraphs: Compliance 
with PM limits is required within 30 days of the effective date of this 
rule. Compliance with SO2 and NOX limits is 
required within 180 days of the effective date of this rule, unless 
installation of additional emission controls is necessary to comply 
with emission limitations under this rule, in which case compliance is 
required within five years of the effective date of this rule.
    (e) Compliance determinations for SO2 and 
NOX. (1) CEMS for EGUs. At all times after the compliance 
date specified in paragraph (d) of this section, the owner/operator of 
each unit shall maintain, calibrate, and operate a CEMS, in full 
compliance with the requirements found at 40 CFR part 75, to accurately 
measure SO2, NOX, diluent, and stack gas 
volumetric flow rate from each unit. The CEMS shall be used by the 
owner/operator to determine compliance with the emission limitations in 
paragraph (c) of this section for each unit.
    (2) Method for EGUs. (i) For any hour in which fuel is combusted in 
a unit, the owner/operator of each unit shall calculate the hourly 
average SO2 and NOX concentration in lb/MMBtu at 
the CEMS in accordance with the requirements of 40 CFR part 75. At the 
end of each boiler operating day, the owner/operator shall calculate 
and record a new 30-day rolling average emission rate in lb/MMBtu from 
the arithmetic average of all valid hourly emission rates from the CEMS 
for the current boiler operating day and the previous 29 successive 
boiler operating days.
    (ii) An hourly average SO2 or NOX emission 
rate in lb/MMBtu is valid only if the minimum number of data points, as 
specified in 40 CFR part 75, is acquired by the owner/operator for both 
the pollutant concentration monitor (SO2 or NOX) 
and the diluent monitor (O2 or CO2).
    (iii) Data reported by the owner/operator to meet the requirements 
of this section shall not include data substituted using the missing 
data substitution procedures of subpart D of 40 CFR part 75, nor shall 
the data have been bias adjusted according to the procedures of 40 CFR 
part 75.
    (3) CEMS for cement kilns. At all times after the compliance date 
specified in paragraph (d) of this section, the owner/operator of each 
unit shall maintain, calibrate, and operate a CEMS, in full compliance 
with the requirements found at 40 CFR 60.63(f) and (g), to accurately 
measure concentration by volume of SO2 and NOX 
emissions into the atmosphere from each unit. The CEMS shall be used by 
the owner/operator to determine compliance with the emission 
limitations in paragraph (c) of this section for each unit, in 
combination with data on actual clinker production. The owner/operator 
must operate the monitoring system and collect data at all required 
intervals at all times the affected unit is operating, except for 
periods of monitoring system malfunctions, repairs associated with 
monitoring system malfunctions, and required monitoring system quality 
assurance or quality control activities (including, as applicable, 
calibration checks and required zero and span adjustments).
    (4) Method for cement kilns. (i) The owner/operator of each unit 
shall record the daily clinker production rates.
    (ii) The owner/operator of each unit shall calculate and record the 
30-operating day rolling emission rates of SO2 and 
NOX, in lb/ton of clinker produced, as the total of all 
hourly emissions data for the cement kiln in the preceding 30 days, 
divided by the total tons of clinker produced in that kiln during the 
same 30-day operating period, using the following equation:

[GRAPHIC] [TIFF OMITTED] TR18SE12.069


Where:

ED = 30 kiln operating day average emission rate of 
NOX or SO2, lb/ton of clinker;
Ci = Concentration of NOX or SO2 
for hour i, ppm;
Qi = volumetric flow rate of effluent gas for hour i, 
where
Ci and Qi are on the same basis (either wet or 
dry), scf/hr;
Pi = total kiln clinker produced during production hour 
i, ton/hr;
k = conversion factor, 1.194 x 10-7 for NOX 
and 1.660 x 10-7 for SO2; and.
n = number of kiln operating hours over 30 kiln operating days, n = 
1 to 720.

    For each kiln operating hour for which the owner/operator does not 
have at least one valid 15-minute CEMS data value, the owner/operator 
must use the average emissions rate (lb/hr) from the most recent 
previous hour for which valid data are available. Hourly clinker 
production shall be determined by the owner/operator in accordance with 
the requirements found at 40 CFR 60.63(b).

[[Page 57917]]

    (iii) At the end of each kiln operating day, the owner/operator of 
each unit shall calculate and record a new 30-day rolling average 
emission rate in lb/ton clinker from the arithmetic average of all 
valid hourly emission rates for the current kiln operating day and the 
previous 29 successive kiln operating days.
    (5) Method for compressor station. The owner/operator of Blaine 
County 1 Compressor Station shall install a temperature-
sensing device (i.e. thermocouple or resistance temperature detectors) 
before the catalyst in order to monitor the inlet temperatures of the 
catalyst for each engine. The owner/operator shall maintain the exhaust 
temperature at the inlet to the catalyst for each engine at a minimum 
of least 750 [deg]F and no more than 1250 [deg]F in accordance with the 
catalyst manufacturer's specifications. Also, the owner/operator shall 
install gauges before and after the catalyst for each engine in order 
to monitor pressure drop across the catalyst. During the initial 
performance test the owner/operator maintain the pressure drop within 
 2'' water at 100 percent load plus or minus 10 percent 
from the pressure drop across the catalyst measured. The owner/operator 
shall follow the manufacturer's recommended maintenance schedule and 
procedures for each engine and its respective catalyst. The owner/
operator shall only fire each engine with natural gas that is of 
pipeline-quality in all respects except that the CO2 
concentration in the gas shall not be required to be within pipeline-
quality.
    (f) Compliance determinations for particulate matter.
    (1) EGU particulate matter BART limits. Compliance with the 
particulate matter BART emission limits for each EGU BART unit shall be 
determined by the owner/operator from annual performance stack tests. 
Within 60 days of the compliance deadline specified in paragraph (d) of 
this section, and on at least an annual basis thereafter, the owner/
operator of each unit shall conduct a stack test on each unit to 
measure particulate emissions using EPA Method 5, 5B, 5D, or 17, as 
appropriate, in 40 CFR part 60, Appendix A. A test shall consist of 
three runs, with each run at least 120 minutes in duration and each run 
collecting a minimum sample of 60 dry standard cubic feet. Results 
shall be reported by the owner/operator in lb/MMBtu. The results from a 
stack test meeting the requirements of this paragraph that were 
completed within 120 days prior to the compliance date can be used by 
the owner/operator in lieu of the first stack test required. In 
addition to annual stack tests, owner/operator shall monitor 
particulate emissions for compliance with the BART emission limits in 
accordance with the applicable Compliance Assurance Monitoring (CAM) 
plan developed and approved in accordance with 40 CFR part 64.
    (2) Cement kiln particulate matter BART limits. Compliance with the 
particulate matter BART emission limits for each cement kiln shall be 
determined by the owner/operator from annual performance stack tests. 
Within 60 days of the compliance deadline specified in paragraph (d) of 
this section, and on at least an annual basis thereafter, the owner/
operator of each unit shall conduct a stack test on each unit to 
measure particulate matter emissions using EPA Method 5, 5B, 5D, or 17, 
as appropriate, in 40 CFR part 60, Appendix A. A test shall consist of 
three runs, with each run at least 120 minutes in duration and each run 
collecting a minimum sample of 60 dry standard cubic feet. The average 
of the results of three test runs shall be used by the owner/operator 
for demonstrating compliance.
    Clinker production shall be determined in accordance with the 
requirements found at 40 CFR 60.63(b). Results of each test shall be 
reported by the owner/operator as the average of three valid test runs. 
In addition to annual stack tests, owner/operator shall monitor 
particulate emissions for compliance with the BART emission limits in 
accordance with the applicable Compliance Assurance Monitoring (CAM) 
plan developed and approved in accordance with 40 CFR part 64.
    (i) For Ash Grove Cement, the emission rate of particulate matter 
shall be computed by the owner/operator for each run in pounds per hour 
(lb/hr).
    (ii) For Holcim, the emission rate (E) of particulate matter shall 
be computed by the owner/operator for each run in lb/ton clinker, using 
the following equation:

E = (CsQs)/PK

Where:

E = emission rate of PM, lb/ton of clinker produced;
Cs = concentration of PM in grains per standard cubic 
foot (gr/scf);
Qs = volumetric flow rate of effluent gas, where 
Cs and Qs are on the same basis (either wet or 
dry), scf/hr;
P = total kiln clinker production, tons/hr; and
K = conversion factor, 7000 gr/lb,

    (g) Recordkeeping for EGUs. The owner/operator shall maintain the 
following records for at least five years:
    (1) All CEMS data, including the date, place, and time of sampling 
or measurement; parameters sampled or measured; and results.
    (2) Records of quality assurance and quality control activities for 
emissions measuring systems including, but not limited to, any records 
required by 40 CFR Part 75.
    (3) Records of all major maintenance activities conducted on 
emission units, air pollution control equipment, and CEMS.
    (4) Any other records required by 40 CFR part 75.
    (5) All particulate matter stack test results.
    (h) Recordkeeping for cement kilns. The owner/operator shall 
maintain the following records for at least five years:
    (1) All CEMS data, including the date, place, and time of sampling 
or measurement; parameters sampled or measured; and results.
    (2) All particulate matter stack test results.
    (3) All records of clinker production.
    (4) Records of quality assurance and quality control activities for 
emissions measuring systems including, but not limited to, any records 
required by 40 CFR part 60, appendix F, Procedure 1.
    (5) Records of all major maintenance activities conducted on 
emission units, air pollution control equipment, CEMS and clinker 
production measurement devices.
    (6) Any other records required by 40 CFR part 60, Subpart F, or 40 
CFR part 60, Appendix F, Procedure 1.
    (i) Reporting. All reports under this section, with the exception 
of 40 CFR 52.1396(n) and (o), shall be submitted by the owner/operator 
to the Director, Office of Enforcement, Compliance and Environmental 
Justice, U.S. Environmental Protection Agency, Region 8, Mail Code 
8ENF-AT, 1595 Wynkoop Street, Denver, Colorado 80202-1129.
    (1) The owner/operator of each unit shall submit excess emissions 
reports for SO2 and NOX BART limits. Reports 
shall be submitted quarterly by the owner/operator for EGUs and 
semiannually for cement kilns, no later than the 30th day following the 
end of each calendar quarter or semiannual period, respectively. Excess 
emissions means emissions that exceed the emissions limits specified in 
paragraph (c) of this section. The reports shall include the magnitude, 
date(s), and duration of each period of excess emissions, specific 
identification of each period of excess emissions that occurs during 
startups, shutdowns, and malfunctions of the unit, the nature and cause 
of any malfunction (if known), and the corrective action taken or 
preventative measures adopted.

[[Page 57918]]

    (2) The owner/operator of each unit shall submit CEMS performance 
reports, to include dates and duration of each period during which the 
CEMS was inoperative (except for zero and span adjustments and 
calibration checks), reason(s) why the CEMS was inoperative and steps 
taken to prevent recurrence, and any CEMS repairs or adjustments. The 
owner/operator shall submit reports quarterly for EGUs and semiannually 
for cement kilns.
    (i) For EGUs: The owner/operator of each unit shall also submit 
results of any CEMS performance tests required by 40 CFR part 75 
(Relative Accuracy Test Audits, Relative Accuracy Audits, and Cylinder 
Gas Audits).
    (ii) For cement kilns: Owner/operator of each unit shall also 
submit results of any CEMS performance tests required by 40 CFR part 
60, appendix F, Procedure 1 (Relative Accuracy Test Audits, Relative 
Accuracy Audits, and Cylinder Gas Audits).
    (3) When no excess emissions have occurred or the CEMS has not been 
inoperative, repaired, or adjusted during the reporting period, the 
owner/operator shall state such information in the quarterly reports 
required by sections (h)(1) and (2) of this section.
    (4) The owner/operator of each unit shall submit results of any 
particulate matter stack tests conducted for demonstrating compliance 
with the particulate matter BART limits in paragraph (c) of this 
section within 60 days after the completion of the test.
    (5) The owner/operator of each unit shall submit semi-annual 
reports of any excursions under the approved CAM plan in accordance 
with the schedule specified in the source's title V permit.
    (j) Testing requirements for Blaine County 1 Compressor 
Station:
    (1) An initial performance test shall be conducted by the owner/
operator for each engine for measuring NOX emissions from 
the engines to demonstrate initial compliance with the emission limits. 
The initial performance test shall be conducted by the owner/operator 
as expeditiously as practicable, but no later than October July 31, 
2018.
    (2) Upon change out of the catalyst for each engine a performance 
test shall be conducted by the owner/operator for measuring 
NOX from the engines to demonstrate compliance with the 
emission limits and re-establish temperature and pressure correlations. 
The performance test shall be conducted by the owner/operator within 90 
calendar days of the date of the catalyst change out.
    (3) The performance tests for NOX shall be conducted by 
the owner/operator in accordance with the test methods specified in 40 
CFR Part 60, Appendix A. EPA Reference Method 7E shall be used to 
measure NOX emissions.
    (4) All tests conducted by the owner/operator for NOX 
emissions must meet the following requirements:
    (i) All tests shall be performed at a maximum operating rate (90 to 
110 percent of engine capacity at site elevation).
    (ii) During each test run, data shall be collected on all 
parameters necessary to document how NOX emissions in pounds 
per hour were measured or calculated (such as test run length, minimum 
sample volume, volumetric flow rate, moisture and oxygen corrections, 
etc.). The temperature at the inlet to the catalyst and the pressure 
drop across the catalyst shall also be measured and recorded during 
each test run for each engine.
    (iii) Each source test shall consist of at least three 1-hour or 
longer valid test runs. Emission results shall be reported as the 
arithmetic average of all valid test runs and shall be in terms of the 
emission limits (pounds per hour).
    (iv) A source test plan for NOX emissions shall be 
submitted to EPA at least 45 calendar days prior to the scheduled 
performance test.
    (v) The source test plan shall include and address the following 
elements:
    (A) Purpose of the test;
    (B) Engines and catalysts to be tested;
    (C) Expected engine operating rate(s) during test;
    (D) Schedule/date(s) for test;
    (E) Sampling and analysis procedures (sampling locations, test 
methods, laboratory identification);
    (F) Quality assurance plan (calibration procedures and frequency, 
sample recovery and field documentation, chain of custody procedures); 
and
    (G) Data processing and reporting (description of data handling and 
quality control procedures).
    (k) Monitoring, recordkeeping, and reporting requirements for 
Blaine County 1 Compressor Station:
    (1) The owner/operator shall measure NOX emissions from 
each engine at least semi-annually or once every six month period to 
demonstrate compliance with the emission limits. To meet this 
requirement, the owner/operator shall measure NOX emissions 
from each engine using a portable analyzer and a monitoring protocol 
approved by EPA.
    (2) The owner/operator shall submit the analyzer specifications and 
monitoring protocol to EPA for approval within 45 calendar days prior 
to installation of the NSCR unit.
    (3) Monitoring for NOX emissions shall commence during 
the first complete calendar quarter following the owner/operator's 
submittal of the initial performance test results for NOX to 
EPA.
    (4) The owner/operator shall measure the engine exhaust temperature 
at the inlet to the oxidation catalyst at least once per week and shall 
measure the pressure drop across the oxidation catalyst monthly.
    (5) The owner/operator shall ensure that each temperature-sensing 
device is accurate to within plus or minus 0.75% of span and that the 
pressure sensing devices be accurate to within plus or minus 0.1 inches 
of water.
    (6) The owner/operator shall keep records of all temperature and 
pressure measurements; vendor specifications for the thermocouples and 
pressure gauges; vendor specifications for the NSCR catalyst and the 
air-to-fuel ratio controller on each engine.
    (7) The owner/operator shall keep records sufficient to demonstrate 
that the fuel for the engines is pipeline-quality natural gas in all 
respects, with the exception of the CO2 concentration in the 
natural gas.
    (8) The owner/operator shall keep records of all required testing 
and monitoring that include: The date, place, and time of sampling or 
measurements; the date(s) analyses were performed; the company or 
entity that performed the analyses; the analytical techniques or 
methods used; the results of such analyses or measurements; and the 
operating conditions as existing at the time of sampling or 
measurement.
    (9) The owner/operator shall keep records of all deviations from 
the emission limit or operating requirements (e.g., catalyst inlet 
temperature, pressure drop across the catalyst) for each engine. The 
records shall include: The date and time of the deviation, the name and 
title of the observing employee and a brief description of the 
deviation and the measures taken to address the deviation and prevent 
future occurrences.
    (10) The owner/operator shall maintain records of all required 
monitoring data, support information (e.g., all calibration and 
maintenance records, all original strip-chart recordings for continuous 
monitoring instrumentation, and copies of all reports required) and 
deviations from operating requirements for a period of at least five 
years from the date of the monitoring sample, measurement, or report 
and that these records be made available upon request by EPA.
    (11) The owner/operator shall submit a written report of the 
results of the required performance tests to EPA within 90 calendar 
days of the date of testing completion.

[[Page 57919]]

    (l) Notifications. (1) The owner/operator shall submit notification 
of commencement of construction of any equipment which is being 
constructed to comply with the SO2 or NOX 
emission limits in paragraph (c) of this section.
    (2) The owner/operator shall submit semi-annual progress reports on 
construction of any such equipment.
    (3) The owner/operator shall submit notification of initial startup 
of any such equipment.
    (m) Equipment operation. At all times, the owner/operator shall 
maintain each unit, including associated air pollution control 
equipment, in a manner consistent with good air pollution control 
practices for minimizing emissions.
    (n) Credible evidence. Nothing in this section shall preclude the 
use, including the exclusive use, of any credible evidence or 
information, relevant to whether a source would have been in compliance 
with requirements of this section if the appropriate performance or 
compliance test procedures or method had been performed.
    (o) CFAC notification. CFAC shall notify EPA 60 days in advance of 
resuming operation. CFAC shall submit such notice to the Director, Air 
Program, U.S. Environmental Protection Agency, Region 8, Mail Code 8P-
AR, 1595 Wynkoop Street, Denver, Colorado 80202-1129. Once CFAC 
notifies EPA that it intends to resume operation, EPA will initiate and 
complete a BART determination after notification and revise the FIP as 
necessary in accordance with regional haze requirements, including the 
BART provisions in 40 CFR 51.308(e). CFAC will be required to install 
any controls that are required as soon as practicable, but in no case 
later than five years following the effective date of this rule.
    (p) M2Green Redevelopment LLC notification. M2Green Redevelopment 
LLC shall notify EPA 60 days in advance of resuming operation. M2Green 
Redevelopment LLC shall submit such notice to the Director, Air 
Program, U.S. Environmental Protection Agency, Region 8, Mail Code 8P-
AR, 1595 Wynkoop Street, Denver, Colorado 80202-1129. Once M2 Green 
Redevelopment LLC notifies EPA that it intends to resume operation, EPA 
will initiate and complete a four factor analysis after notification 
and revise the FIP as necessary in accordance with regional haze 
requirements including the ``reasonable progress'' provisions in 40 CFR 
51.308(d)(1). M2 Green Redevelopment LLC will be required to install 
any controls that are required as soon as practicable, but in no case 
later than July 31, 2018.

[FR Doc. 2012-20918 Filed 9-17-12; 8:45 am]
BILLING CODE 6560-50-P


