kc in 33 AM '09 - APSC

143
ARK pcy, ’:: LET’/ cci.:lA. -??RE iaav ‘1‘ :OMLln kc in 33 AM ‘09 K5 BEFOU THJ3 ARKANSAS PUBLIC SERVICE COMMISSION ‘,-it ED I I!.- IN THE MATTER OF ENTERGY ARKANSAS, ) ORDER APPROVING THE ADDITION OF THE ) EWIRONMJWTAI, CONTROLS PROJECT AT ) THE WHITE BLUFF STEAhl ELECTRIC 1 STATION MAR mDFELD, ARKANSAS 1 XNC.3 REQUEST FOR A DECLARATORY 1 DOCKET NO. 09-024-U RESPONSE IN SUPPORT OF STAFF’S MOTION TO IMMEDIATELY SUSPEND PROCEDURAL SCmDULE For the reasons set forth below, the Sierra Club and National Audubon Society (collectively “Sierra Club”) support Staff’s motion to immediately suspend the procedural schedule for this docket. I. As Staff correctly noted, the Arkansas Department of Environmentd Quality YADEQ”) must issue a PSD permit’ prior to EAI going forward with its environmental controls project (“Project”). Staff aIso correctly noted that both the U.S. Environmental Protection Agency and the U.S. Forest Service submitted detailed comments to ADEQ pointing out numerous technical and legal flaws in ADEQ’s proposed Draft Operating Air Permit for EAI’s Project. These comments will Iikely affect my permit ADEQ issues. 2. In addition to the comments Staff provided, the U.S. Fish and Wildlife Service, which manages an affected wiIdlife refuge, and the Sierra CIub also provided highly substantive comments based on expert opinion. Those comments are appended here as Attachments 1 and 2. The Sierra Club in particdar provided voluminous comments showing, among other things, that EAI’s proposed controls would not meet regional haze regulations and that certain plant a a [ej Prevention of Sisnificant Deterioration under the Clean Air Act. 1

Transcript of kc in 33 AM '09 - APSC

ARK p c y , ’:: LET’/ cci.:lA.

- ? ? R E iaav ‘1‘ :OMLln

kc in 33 AM ‘09 K5 BEFOU THJ3 ARKANSAS PUBLIC SERVICE COMMISSION

‘,-it ED I I ! . - IN THE MATTER OF ENTERGY ARKANSAS, )

ORDER APPROVING THE ADDITION OF THE ) EWIRONMJWTAI, CONTROLS PROJECT AT ) THE WHITE BLUFF STEAhl ELECTRIC 1 STATION MAR mDFELD, ARKANSAS 1

XNC.3 REQUEST FOR A DECLARATORY 1 DOCKET NO. 09-024-U

RESPONSE IN SUPPORT OF STAFF’S MOTION TO IMMEDIATELY SUSPEND PROCEDURAL SCmDULE

For the reasons set forth below, the Sierra Club and National Audubon Society

(collectively “Sierra Club”) support Staff’s motion to immediately suspend the procedural

schedule for this docket.

I. As Staff correctly noted, the Arkansas Department of Environmentd Quality

YADEQ”) must issue a PSD permit’ prior to EAI going forward with its environmental controls

project (“Project”). Staff aIso correctly noted that both the U.S. Environmental Protection

Agency and the U.S. Forest Service submitted detailed comments to ADEQ pointing out

numerous technical and legal flaws in ADEQ’s proposed Draft Operating Air Permit for EAI’s

Project. These comments will Iikely affect m y permit ADEQ issues.

2. In addition to the comments Staff provided, the U.S. Fish and Wildlife Service,

which manages an affected wiIdlife refuge, and the Sierra CIub also provided highly substantive

comments based on expert opinion. Those comments are appended here as Attachments 1 and 2.

The Sierra Club in particdar provided voluminous comments showing, among other things, that

EAI’s proposed controls would not meet regional haze regulations and that certain plant

a a [ej

’ Prevention of Sisnificant Deterioration under the Clean Air Act.

1

modifications actuaIIy would trigger more stringent control technology, i.e., best available

controI technology (“BACT”) versus best available retrofit technology (“BART’*).’

3. These comment letters to ADEQ raise serious questions concerning the Project’s

ability to IawfuIIy comply with air quality requirements. In the interest of administrative

economy, the Commission must allow these critical state and federal processes to rcsoIve so that

the Commission’s own process goes forward with a full and accurate record on which it can base

its own determination on the Project.

WHEREFORE, Sierra Club requests that the Commission immediately suspend the

procedural schedule in this docket.

Respectfully submitted,

Gloria D. Smith Sierra CIub, senior staff attorney

E M is proposing best available retrofit technology when in fact it should be implementing best available control

2

2

technoIogy.

CERTIFICATE OF SERVICE

I, Violet Lehrer, do hereby certify that a true and correct copy of the foregoing Response in Support of Staffs Motion to Immediately Suspend Procedural Schedule was served on all parties of record by electronic mail this 4* day of December, 2009,

violet Lehrer

United States Department of the Interior FISH AND WILDLIFE SERWCE

Na tianal W i Id1 i fe Refuge Sys tern Branch of AIr Quality

7333 W. Jcfferson Ave., 96k 375 Lakewd, CO 802352017

M REPLY = To F'NSIANWS-Ak AQ

November 19,2009

Mike Bates Chief, Air Division Arkansas Department of Environmental Quality 5301 Northshore Drive North Little Rock. Arkansas 72 1 18-53 17

SUBJ: Comments on Draft Air Permit # 0263-AOP-R7 AFlN: 35-00 1 10 for Entergy Arkansas, Inc. - White Huff Plant with additional reference to Best AvaiIable Retrofit Technology (BART)

Dear Mr. Bates:

The US. Fish and Wildlife Service (FWS) appreciates the opportunity to comment during the Public Notice comment period on the subject draft air permit for Entergy Arkansas, Inc. -White BIuff Plant, Units #1 and #2. The FWS recognizes the efforts of Entergy and the Arkansas Department of Environmental Quality (ADEQ) in proposing significant SO1 controls for Units $1 and $2. Flue Gas Desulfurization on these units will improve visibility at a11 nearby Class I areas. However, it is our conclusion that this proposed final permitting action for Prevention of Significant Deterioration does not address all of the requircnients outIined in the EPA BART Guidelines.

A significant change regarding this facility appears to have been made after the Arkansas RcgionaI Haze SIP was submitted to the U. S. Environmental Protection Agency without the Federal Land Managers, including the FWS, having adequate information to properly evaluate and comment on the proposed change. The primary change involved the seIection of a dry flue gas desulhrization system with a spray dry absorber for SO2 control rather than the previously- selected wet flue gas desulfurization system, Until the FWS receives and reviews the pertinent information which has not yet been supplied as described in this letter, definitive comments cannot be made. Future definitive comments made by the FWS after receipt of the information may adversely affect any actions taken in the near future by Entergy Arkansas, Inc. (Entergy), acting on any permit that might be currently issued as a resuIt of the upcoming pubIic hearing. Since the facility's permit is the enforcement mechanism for BART-related actions and since proper BART review is not possible at this time, we would ask that you not finalize the proposed permit as a BART decision. Nevertheless, we provide some substantive comments on the summary information that was available for review.

Attachment I

Mr. Bates 2

The FWS recentIy located a document entitled, "Revised BART Analysis for the White Bluff Steam Electric Station" dated August 2008 (2008 Entergy BART Determination) on the ADEQ website. The document appears to be a BART determination prepared by Entergy Arkansas, Inc. It seems to replace a document that we were earlier provided for review, "BART Analysis for the White Bluff S t e m Electric StationI'dated December 2006 (2006 Entergy BART Determination). The FWS was not provided with the required review period for the 2008 Entergy BART Determination. Further, the Arkansas Regional Haze SIP does not contain any ADEQ determinationsconfirming State adoption of Entergy conclusions from the 2008 Entergy BART Determination. It might be supposed that ADEQ concurs with the 2008 Entergy BART Determination, since it is proposing to issue an air permit based on that document's conclusions. The importance of the 2008 Entergy BART Determination in the draft Entergy Arkansas, Inc., permit application entitIed, "Application for Permit to Construct- Entergy White BIuff Units 1 & 2 Air Pollution ControI Project" dated January 2009 (Application) is that on page 2-7 of the document it states, "This action is being taken to reduce emissions of NO, and SO2 in response to Best Available Retrofit Technology (BART) State ImpIementation PIan (SIP) requirements."

The information that was available within the 2008 Entergy BART Determination provided a good summary on cost and visibility, but lacked both detailed information supporting deveIopment of those costs, and visibility modeling for Caney Creek and the other visibility- impacted Class I areas. The document indicated that Entergy considers this additional information to contain Confidential Business Information (CBT), and that information wiIl be provided at a future time under CBI safeguards. Likewise, the Application omits the missing information. Regarding cost estimates the EPA BART Guidelines state that, "The basis for equipment cost estimates also should be documented, cither with data supplied by an equipment vendor (Le., budget estimates or bids) or by a referenced source (such as the OAQPS Control Cost Manual). In order to maintain and improve consistency, cost estimates shouId be based on the OAQPS Control Cost ManuaI, where possible."' Regarding the fifth hctor of BART determinations; nameIy, degree of improvement in visibility, the EPA BART Guidelincs detaiI the procedure at Section 1V.D.S.STEP 5. Thus, the FLMs cannot yet make a definitive determinationon the condusions made by Entergy in the 2008 Entergy BART Determination or the draft air permit for Entergy Arkansas, Tnc. - White BIuff Plant, Units #1 and #2. We ask that we be allowed to make further review and comment on these documents when the full body of information becomes availabIe.

On the supposition that the currently unavdable background infomation properly justifies the conclusions arrived at by Entergy and presumably ADEQ, the FWS has substantive comments that shouId be considered at this time. The most significant issue is that Entergy and ADEQ imply that meeting the presumptive emission rate of 0.15 lblMMBtu for both SO2 and NO, meets the intent of BART and that permitted emission limits being set at 0.15 lb/MMBtu for each of these pollutants is BART. The EPA Guidelines define BART as, "... an emission limitation

See40 CFR €kt 51, Appendix Y, "Guidelines for BART DeterminationsUnder the Regional Haze Rule," section IV.D.4.STEP 4a.5.

Attachment 1

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 ... [a BART-eligible source] ... taking into consideration.. the costs of compliance.”2 The “best system” (as determined by the 5 factors) in terms of “cost” is regarded as a ’dominant control’ faIling on the ‘least-cost’ envelope.3 The PLM’s conclusion from this is that if a control alternative remains to be reasonabIe in cost and is a “best system” (dominant control), it is BART- even if that alternative happens to provide more control than the presumptive level.

Therefore, the proposed SO2 emission limits in the draft permit, which seem to have been based on a 0.15 IblMhlTBtu emission rate for dry flue gas desulfurization with spray dry absorbers (dry FGD with SDA), are too lenient. Let us use information provided in the Application for Unit ffl as an example. The 2008 Entergy BART Determination states that the proposed dry FGD with SDA has a controI efficiency of 92.5%. The Application states that the baseline cmission rate is 0.65 lb/MMBtu. A contrd efficiency of 92.5% applied to 0.65 lb S02/MMBtu wouId resuIt in a ilieureticd post-controI emission rate of 0.049 Ib S02IMMBtu. so consider the post-control emission rate to be 0.05 Ib SOtlMMBtu. This control efficiency, applied to a baseline emission of 17,733 tons per year results in a permitted emission limit of 1,364 tons SO2 per year ( I 7,733 x .051.65 = 1,364). This compares to the 5,880 1011s of SO1 per year proposed by Entergy in the Application. Converting the theoretically achievable 1,3G4 tons per year to an hourly rate results in an emission limit of 3 I2 lb SOhour ( 1,364 x 200018760), rather than the 1,342.5 lb SOfiour as proposed in the draft permit. In summary, the specifications for SO2 shown in the Statement ofBasis should be as foIlows:

Tons per Year Pounds per Hour Emission Rate

1,364 TPY SO2 3 I2 l b h SO2 0.05 IblMMBtu SO2

Using the same procedure for Unit $2 the specification for SO1 shown in the Statement of Basis should be as foIIows:

Tons per Year Pounds per Hour Emission Rate

1,329 TPY SO;! 303 Ibihr SO2

0.05 Ib/MMBtu SOz

A review of emission rates of other large Electric Generation Units with dry FGD systems using low sulfur coal proposed that permitted emission limits be as low as 0.065 lb SO$MMBtu, so regulatory emission limitations might be adjusted accordingly.

’ See40 CFR Part 51, Appendis Y, “Guidelines far BART DeterminationsUndcr tlic Regional Hau! Rule,” section 1V.A

Sce 40 CFR Part 51, Appendix Y, “GuideIines for BART Determinations Under thc Regional Haze RuIc,” section IV,D.4.c.2.

Attachment 1

Mr. Bates

The 2008 Entergy BART Determination concluded that dry FGD wlSDA should be installed as BART, even though the 2006 Entergy BART Determination concIuded that wet flue gas desulfurization (wet FGD) should be instalIed as BART. Without explanation, the annual cost for wet FGD went from approximately$l7,159,020 in the 200G document to $68,045,000 in the 2008 document ( a 297% increase) for each of the two units, while the annual cost for dry FGD wlSDA went from approxirnateIy$34,306,388 in 2006 to $65,155,000 in 2008 (a 90% increase) for each of the two units. Justificationof the wide disparity and uneven escalation in the cost figures is paramount in justifying which control alternative shouId be selected. It is interesting that due to additional H2S04 emissions under wet FGD, the analysis claims that dry FGD WlSDA will result in Icss visibility impairment at Caney Creek. It is reasonable to assume that the uhimate test under the regiond haze program is the effect on visibility. This position would favor the selected alternativeof dry FGD w/SDA. We would stiII Iike to see the visibility impact modeling and data for Caney Creek and for the other impacted Class I areas.

The baseline period values used for the 2008 Entergy BART Determination differ from those used for the Application. Since the values used in the 2008 Entergy BART Determination are primarily used for choosing between alternatives (and that one of the alternatives was in fact chosen as BART) and the Application values primarily determine permitted emission rates, it is not imperative that they be identical. However, some expIanation of the differences might be in order. For your convenience the values are shown below:

Baseline Emissions Application SO2 Annual Emissions Unit I 17,733 TPY SO2 AnnuaI Emissions Unit 2 NO, Annual Emissions Unit 1 NU, Annual Emissions Unit 2 SOZ Emission Rate Unit 1 Sa Emission Rate Unit 2 NO, Emission Rate Unit 1

18,077 TPY 6,792 TPY 7,206 TPY

0.65 IblMMBtu 0.68 IblMMBtu 0.25 1bMMBht

NO, Emission Rate Unit 2 0.27 lbMMBtu

2008 BART Det. 28,902.8 TPY 29,132.5 TPY 16,275.7 TPY 17,612.9 TPY 0.83 JblMMBtu 0.77 lblMMI3tu 0.468 IbMMBtu 0.463 IbhlMBtu

As discussed above for S02, implementation of NO, controls that simpIy meet the NO, presumptive emission rate of 0.15 Ib NOx/MMBtu (Le,, combustion controls) does not meet the intent of the EPA BART Guidelines. If post-combustion control equipment can be installed cost-effectively to achieve an emission rate beIow presumptive, then such equipment should be installed. Therefore, a cost analysis should have been considered for employing Selective Catalytic Reduction (SCR) and a judgment made as to its cost-effectiveness.

In our previous comments to ADEQ OR the draft Regional Haze SIP in 2008, it was noted that Section 9.3 should provide a summary of the BART determinations for the Subject-to-BART sources. If the emission limits proposed for the White Bluff plant as outlined in the Statement of Basis are the ADEQ BART conclusions for the plant, then Section 9.3 should be amended to

Attachment *I

Mr. Bates 5

include confirmation of the State’s decision and an explanation as to how a proper BART five- factor analysis led to those conclusions.

If you have any questions, or if you would like to discuss these comments in more detail, please contact Meredith Bond at (303) 9 143808.

Sincerely,

Sandra V. Silva Chief, Branch of Air Quality

Cc (via e-mail}: Thomas Rheawe, Permits Branch Manager, ADEQ Tony Davis, Planning & Air Quality Analysis Branch Manager, APEQ Joe Kordzi, EPA Region 6 Allen Chang, EPA Region 6 Judith Logan, USDAlFS Bruce Polkowsky, NPS Ben Mense, Refuge Manager, Mingo National Wildlife Refuge

Attachment I

The Law Office of William J. Moore, III

1648 Osceola Street Jacksonville, Florida 32204

Telcphonc (904) 685-2 172 Facsimilc (904) 685-2 175

via U.S* Mail and e-mail November 34,2009

Teresa Marks, Director Arkansas Department of Environmental Quality 530 1 Northshore Drive North Little Rock, AR 72 1 18-53 17 Attention: Air Permits Branch AirPsrmits~~irlndcci.statc.ar.us

Re : Sierra Club Cowmetits ott Drafr PSDflitZe V Permit and BART Permit for the Entergy Arkarisus White B lq f PZant

Dear Ms. Marks:

I am an attorney representing the Sierra Club. I am submitting the following comments from Sicm Club on the Draft Prevention of Significant Deterioration (PSD)n*itle V Permit and Best Available Retrofit Technology (BART) Permit for the Entergy Arkansas White Bluff Plant. In addition to these written comments, pursuant to express instructions from staff at ADEQ, Sierra Club has uploaded si number of the exhibits onto an ‘‘ftp” site which the Department has access to. Those exhibits are incorporated by reference into this document and support the comments made herein.

I. THE PUBLIC NOTICE FOR THE DRAPT PERMIT WAS INCOMPLETE AND DEFICIENT

A. The Public Notice and Federal Land Manager Notification for the Draft Permit and Finalization of BART Controls for White Bluff Failed to Meet Regional Ham SIP Requirements

ADEQ is proposing to issue the Entergy permit its part of its finalization of its BART determination for White Bluff. ADEQ has stated in its regional haze SIP that permit approvals such as this one for White Bluff essentially represented ADEQ’s finaI approval of the BART determination process for a particular BART eligible source.’

’ According to Section 9.2 of ADEQ’s Regional Haze SIP at 47, adopted in Scptembcr 2008, “it is dtimately the responsibility of ADEQ to either approve or reject the BART sources’ cngincering andysis during the permitting process.”

1

Attachment 2

However, Entergy’s proposed BART controls for sulfur dioxide (S02) are significantly different from what was Entergy proposed as BART in its analysis that was submitted to ADEQ in 2006. Specifically, Entergy previously concluded a wet scrubber was more cost effective than a dry scrubber with a baghouse. Entergy’s December 2006 BART Analysis for White BIuff at ES-1. Thus, Entergy’s proposed installation of a wet scrubber as BART and the modeling that was conducted to demonstrate the benefits of its BART determination on regiond haze in Class I areas was presumably based on the stack characteristics and emissions with a wet scrubber. Now, in its 2009 PSD permit application, Entergy has instead proposed to instal1 a dry scrubber {along with a baghouse) as BART for SO2. January 2009 Permit Application for White BIuff Air Pollution Control Project at 2-7. While Entergy bas proposed to meet the same SO2 emission limit with the dry scrubber as it did with the wet scrubber (0.15 IblMMBtu), it appears that there has becn no additiond modeIing to demonstrate the how the impacts of Entergy’s current BART analysis will affect regional haze in the affected Class I areas. Furthermore, Entergy’s PSD permit application failed to include an updated five factor BART anaIysis as required by 40 C.F.R. Part 5 I , Appendix Y, Section IV.

In reviewing the documents posted on ADEQ’s “Entergy White Bluff Draft Permit Information” website, Sierra Club has inadvertently discovered that Entergy did make a subsequent submittal of a revised BART anaIysis for White Bluff in August 2008 but it is buried in the “Entergy-ADEQ internal emaikpdf’ file. This does not constitute proper pubIic notice of Entergy’s revised BART analysis. Furthermore, based on conversations with staff at the U.S. Fish & Wildlife Service, the only Entcrgy BART analysis that the U.S. Fish and Wildlife Service has is Entergy’s 2006 BART submittal2 which reached an materially different conclusion as to what reflects BART for SO:! at White Bluff. Pursuant to 40 C.F.R. $5 1.308(i)(2), ADEQ is required to give the federal land managers (“FLMs”) the opportunity for consultation at least sixty (60) days prior to holding a public hearing on the regional haze State Implementation Plan (“SIP”) or SIP revision. The draft White Bluff permit, which reflects ADEQ’s final approval of BART controls, is essentially a revision to the regional haze SIP. That is because the regional haze SIP ADEQ submitted to EPA in 2008 included a BART analysis and presumably modeling showing that installation of a wet scrubber met BART for White Bluff. Now that Entergy has detcnnined that installation of a dry scrubber and baghouse meet BART for SO2, this is a major revision to the SIP, requiring ADEQ to undertake a new consultation process with the FLMs. That consultation process is required to include the opportunity for the FLMs to discuss impairment of visibility and recommendations on the development of regional haze implementation strategies inchding BART. The FLMs can not do that if they have not becn given a11 relevant documents at least sixty (60) days prior to the pubIic hearing on the ADEQ’s find approval of BART for White BIuff.

Not only have the FLMs not been provided with the company’s August 2008 revised BART analysis, but no additional modeling analysis was done regarding the company’s revised BART analysis. In addition, information in the “Entergy - ADEQ internal cmails.pdf’ file makes dear that this proposed PSD permit allows White Bluff to

* Wc wcrc unnble to reach the U.S. Forest Service to determine what BART mdysis it hnd rcceivcd for White Bluff.

Attachment 2

emit at higher NOx emission rates than what was previoudy modeled for the BART analysislregional haze SIP. See, cg., July 28,2009 e-mail Eom Thomas Rheaume to Siew Low and Anthony Davis. Further, it appears that ADEQ has done some new Class I area visibiIity modeling with the PSD permit’s proposed emission limits (see, e.g., August 1 1,2009 e-maiI from Mary PettyJolm to Thomas Rheaume), but that modeling has not been made available as part of this draft permit. And, if we are mistaken and no new visibility modeling has been done for Entergy’s new BART analysis and proposed PSD permit, clcady such modeling must be done because, as ADEQ staff stated in a July 28,2009 e-mail, the proposed NOx emission rates of the PSD permit are much higher than what was modeIed in the regional haze modeling?

These issues were raised in the comment letter on the draft White Bluff permit recently submitted by the Fish & WildIife Service to ADEQ. In fact, the Fish&WiIdlife Service asked that ADEQ not issue the permit at this time untiI ADEQ has provided the FLMs with the proper consultation period and all documents necessary to review the BART determination that is essentially being made with the issuance of the White Bluff permit.“

In addition, ADEQ is required to make its own determination of BART as part of its regional haze SIP pursuant to 40 C.F.R. $51.308(e)( I)(ii). 40 C.F.R. $5 1.308(c)(I)(ii)(A). The determination of BART for units the size of the White Bluff units (i.e., greater than 750 megawatts (MW)) must be made pursuant to the Guidelines for BART Determinations Under the Regional Haze Rule in 40 C.F.R. Part 5 I , Appendix Y including the foIlowing of the 5-step BART determination analysis. See 40 C.F.R. $SI.308(e)( I)(ii)(3) and Appendix Y of 40 C.F.R. Part 5 1. And the state’s determination is required to be made publically avaiIable as part of its regional haze SIP, as is required for any SIP or SIP revision, and a thirty (30) day comment period along with an opportunity to request a public hearing must be offercd to the public. See 40 C.F.R. $85 I. 102,5 1.308(e), Appendix V of40 C.F.R. Part 5 I. Eatergy’s August 2008 revised BART proposal clearly reflects a revision to the regional haze SIP submitted to EPA, and yet ADEQ has provided no independent analysis of BART for White BIuff in any of the draft permit documents or in the other documents posted with the draft White Bluff permit on ADEQ’s website (ft~~:/ilistserv.ades.stntc.ar.u~’AirPcnnitslEnter~v?~~20WhiteD/u10Bluft;3. For example, as previously mentioned, it appears ADEQ conducted revised visibility niodeIing based on the PSD permit limits, but this revised modeling has not been public noticed as part of this permit action. Presumably, no independent ADEQ analysis of BART for White Bluff has been transmitted to the FLMs either. For these reasons, ADEQ has failed to provide the FLMs with all relevant information regarding its finalization of the White Bluff BART determination within sixty (60) days of the public hearing that is currentIy scbcduIed for November 17, 2009. This directly contradicts the fedcral regional haze SIP

An August 12,2008 email from Thomas Rheaume to Mary Pettyjohn states the NOx rates proposcd arc 2.506 lbhr, and Entergy’s Deccmber 2OOG BART submittal indicates at page 4 4-1 that NOx emission rates of 1,353.1 to 1,464.2 lbihr were modeled as future peak emission rates for each unit.

Sec 1 IC0109 letter from Fish & WiIdlife Service to ADEQ on dmA White Bluff Permit, Ex. 1. 4

Attachment 2 3

requirements. ADEQ has also failed to make publicly avaiIable as part of the public notice on this draft permit its review and detcnnination of BART for White BIuff and its revised visibility impact modeling. And ADEQ failed to give adequate notice of Entergy’s 2008 revised BART submittal for White BIuff.

Furthermore, the 2008 revised BART analysis included on the ADEQ draft White Bluff permit website is incomplete. Entergy has withheld cost information for the White Bluff BART analysis as confidential business information which should have been made publicly available. Specifically, it appears that part of Appendix A (Cost and Emission Estimates for NOx and SO2 Control Options) and all of Appendix 3 (Costs and Emission Estimates for SO2 Control Options) has been withheId fmm the pubIic as confidential. Yet, this cost andysis and the assumptions used in that analysis are integral to the BART determination as it is one of the 5 factors that must be evaluated. There are certain requirements that must be met in estimating costs of controls, and the public has no way of evaluating whether Entergy properly evaluated costs for SO2 controls without this information. Given that Entergy’s 2006 BART submittal previously found that a wet scrubber was more cost effective and that now Entergy has chosen the most costly, less effective SO2 controls, the public and the FLMs have a right to evaluate Entergy’s cost analyses in detail.

For all the reasons discussed above, ADEQ cannot go forward with this finalization of BART for White Bluff until all SIP requirements arc met for FLM consuItation and for pubIic notice and untiI a11 relevant documents have been made available to the public for at least thirty (30) days prior to the public hearing.

B. Thc Public Notice Failed to Meet Arkansas Public Notice Requirements

1. The Public Notice for Proposed Permit Failed to Comply with AFTEC Reg. 8.207

APCEC Reg. 8.20(B)(3) requires that the public notice identify the “type of permit for which the permitting decision is proposed to be issued.” In the October 16, 2009 public notice for the subject permit, ADEQ only described the type of applications which were submittcd by Entergy, stating that it has submitted a “Prevention of Significant Deterioration and Best AvaiIable Retrofit Technology application. . . .” That was insufficient to identify what type of permit was being issued. Of particular significance, the public notice failed to state that this was also at Title V operating permit and failed to clarify whether the permit in question was a modification of an existing Title V permit or a renewal. For this reason, the pubIic notice violated APCEC Reg. 8.207(B)(3) and must be re-issued. The public comment period must be re-initiated as well.

Attachment 2 4

2. The Public Notice for Proposed Permit FaiIed to Comply with APCEC Reg. 26.602

APCEC Reg. 26.602 is part of Arkansas’ Title V rules. It requires that for “all initial permit issuances, significant modifications, minor modifrca~ons, and renewals,” a public notice be issued, which, infer alia, must identify “the activity or activities invdved in the permit action” and “the emissions change involved in any permit modification.” APCEC Reg. 26.602(A)(3) and (B). The public notice for this permit violated these requirements in several ways.

First, the proposed permit allows for a major turbine efficiency projcc? and for a 250 MMBtulhr increase in heat input to the White Bluff boiled, which were not identified in any manner in the public notice. Because the turbine project and the increase in heat input are clearly significant “activities” involving the permit action, they were required to be disclosed in the public notice.

Second, the proposed permit will resuIt in an increase in certain hazardous air pollutants (HAPS), including arsenic, benzene, cyanide and selenium. See, e.g., Draft Permit at 29. These increases were not identified in the public notice either. These increases constitute “emissions changes” and also represent an “activity” involved in the permit action which was required to be included in the public notice.

Third, the proposed permit includes ton per year limits on nitrogen oxides (NOx) and sulfur dioxide (502). Draft Permit at 42. These Iimits are stated as follows:

19. The permittee shaI1 not exceed 5880.0 tonslyear of SO2 emissions for any consecutive twelve month period from SN-0 1 and SN-02. [Regulation 19,s 19.50 1, and 40 CFR Part 52, Subpart E]

30. The permittee shall not exceed 5880.0 tondyear of NOx emissions for any consecutive twelve month period &om SN-0 1 and SN-02. [ReguIation 19,§ 19.50 1, and 40 CFR Part

Id. While the regulatory basis for imposing these limits is not entirely clear, they appear to be potential to emit limits relied upon in an attempt to avoid triggering PSD review for these pollutants. As such, these limits should have been disclosed in the public notice and the rcglatory basis for including these Iirnits in the permit should have been explained in ADEQ’s statement of basis or in the permit itself.

’ The turbine upgrade is not mentioned in thc dmfl permit, but ivas included by Entergy as part of the projects identified in its 2009 pennit application at 1-1. As i s described hrtlier below, it is also likely tied to the rcquested increase in heat input capncity.

Id. at 4-1: Section IV. of DraA Permit. 4

Attachment 2 5

For all the preceding reasons, the requirements of APCED Reg. APCEC Reg. 26.602(A)(3) and (B)has not been complied with. Therefore, the public notice should be re-issued and pubIic comment process should be re-initiated.

3. ADEQ Failed to CompIy with APCEC Reg. 8.208

APCEC Reg. 8.208 (E) provides in pertinent part that:

The Department shall make available the draft permitting decision and other material relevant to the draft permitting decision for inspection and copying at the Department during the public comment period and shall coriiply with fhe relevant provisions of the Arkansas Freedom of Information Act. The Depwtment shall provide copies fo any person itiakiitg CI request for copies, including my requesi by mail, fcieplione, elecfronic mail, or facsimile.

On October 19,2009, three (3) days of the issuance of the public notice for this permitting action, Sierm Club served ADEQ with a FOIA seeking relevant documents to the proposed permitting decision. Despite diligent efforts on Sierra Club’s part to gather the relevant information, Sierra Club has been unable to obtain a11 the relevant documents and data from ADEQ necessary to prepare and submit adequate comments on the subject permit. The difficulties and deficiencies associated with ADEQ’s documcnt production efforts are largely reflected in October 30,2000 and November 10,7009 letters that were submitted to ADEQ seeking extensions of the comment period deadline. Of particular significance, despite repeated requests by Sierra Club, ADEQ has failed to produce a notice letter from Entergy to ADEQ dated August 8,2007 that appears to contain critical baseline emission and post-project emission projections associated with the rephcement of an economizer at White Bluff Unit 2 in 2007, which, as explained itlfu, is highly relevant to this permitting action. See, e.g., Composite E-Mails from W. Moore, In, to ADEQ’s K. Robinson Regarding August 8,2007 Notice Letter and FOIA Issues, Ex. 64. Additional, as of November 22,2009, there is still an ADEQ compact disc in the process of being delivered by mail to Sierra Club containing additional ADEQ documents and data responsive to Sierra Club’s FOIA.

ADEQ has failed to comply with its obligations under Arkansas FOIA law and has consequently violated the requirements of APCEC Reg. 8.208 (E) to comply with FOIA’s and more informa1 inforniation requests when issued in regards to permitting actions. Moreover, in Iight of ADEQ’s failure to comply with FOIA in the face of Sierra Club’s persistent requests for reIevant information -- particularly with respect to the August 8,2007 notice letter from Entergy -- it clearly was an abuse of discretion for ADEQ to have denied Sierra Club’s request for an extension of the comment period.

Attachment 2 d

11. ENTERGY AND ADEQ HAVE FAILED TO ADEQUATELY ANALYZE BART FOR WHITE BLUFF

In 2007, ADEQ adopted Best Available Retrofit Technology (BART) limits for BART-eligible sources in Regulation 1 9, Chapter 15, Reg. 19.1505. SpecificalIy, ADEQ adopted the “presumptive BART” emission limits identified in EPA’s BART GuideIines of 40 C.F.R. Part 5 1, Appendix Y. Notwithstanding the presumptive BART limits, the EPA’s BART Guidelines require a five factor analysis to determine the BART controls and BART emission limits for each BART-eligible source to be done on a case-by-case basis. However, in the Arkansas Regional Haze SIP adopted by the state in 2008, ADEQ did not conduct a five factor analysis of BART for each source including White Bluff. Instead, ADEQ indicated that its final determination of BART for each BART-eligible source would be determined when the permit was issued for the source for the instalIation of the BART controls.

SpecificaIly, according to Section 9.2 of ADEQ’s RegionaI Haze SIP, adopted in September 2008, “it is ultimately the responsibility of ADEQ to either approve or reject the BART sources’ engineering analysis during the permitting process.” Thus, one of the purposes of the dmft PSDlTitIe V air pennit is to reflect ADEQ’s final approval of BART requirements for White Bluff. Entergy submitted a BART analysis for White Bluff in 2006 and then a revised BART analysis in 2008. These BART analyses essentially just provided Entergy’s review of how White Bluff would meet the presumptive BART limits adopted into Regulation 19, Chapter I5 and thus did not constitute a five factor BART analysis. ADEQ did not provide a review of Entergy’s 2006 analysis despite it being incorporated into the 2008 Regional Haze SIP, and ADEQ has aIso failed to provide its review of Entergy’s BART anaIysis for White Bluff as revised in August 2008 in any document made public with the dnfi PSDflitle V permit. The SO2 controls that ADEQ is “approving” as meeting BART at White Bluff through its proposed issuance of this permit are different than the SO2 controIs that Entergy proposed in its 2006 BART Analysis for White Bluff that ADEQ incorporated into its regional haze SIP that was submitted to EPA for SIP approvaI. See 2006 White Bluff BART Analysis in Appendix 9.3A of the Arkansas Regional Haze SIP submitted to EPA in July of 2008. ConsequentIy, this draft permit for White BIuff essentially reflects a revision to the Arkansas Regional Haze SIP and, accordingIy, there are several requirements that ADEQ must address.

A. ADEQ Failed to Adequately Addrcss All SIP Requircmcnts regarding BART in the Proposed Issuance of this White BIuff Permit

FederaI regulations mandate that states’ regionaI haze SIPS include emission Iimitations representing BART for each BART-eligible source. 40 C.F.R. $5 1.308(e). Best available retrofit technology or BART is defined as follows:

‘A copy of the entire Arkansas Regional Haze SIP including appendices is included as Ex. 2 to this letter.

7 Attachment 2

Best Available Retrofit Technology (BART) means an emission limitation based on the degree of reduction achievabk through the application of the best system of continuous emission reduction for each polIutant which is emitted by an existing stationary facility. The emission limitation must be estabIished, on a case-by-case basis, taking into consideration the technology available, the costs of compliance, the energy and nonair quality environmental impacts of compIiance, 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 resuIt from the use of such technology.

40 C.F.R. 3 51.301.

BART is to be determined based on a five factor analysis. 40 C.F.R. Part 5 1, Subpart Y, Section I.F. I. EPA's five factor analysis requirements stems fiom statutory and regulatory requirements regarding how BART is to be determined. Specifically:

The determination of BART must be based on an anaIysis of the best system of continuous emission control tcchnology available and associated emission reductions achievable for each BART-eligible source that is subject to BART within the State. In this analysis, the State must take into considemtion the technology availabIc, the costs of compliance, the energy and nonair qudity environmental impacts of compliance, any poIIution control equipment in use 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.

40 C.F.R. 5 51.308(e)( I)(ii)(A). See also Q 169A(g) of the Clean Air Act

For power plants that are over 750 megawatts in generating capacity, BART must be determined based on the Guidelines for BART Determinations under the Regional Haze Rule in 40 C.F.R. Part SI, Appendix Y. 40 C.F.R. $ S1.308(e)( l)(ii)(B).

40 C.F.R. $5 1.308(i)(2) provides:

The State must provide the Federal Land Manager with an opportunity for consultation, in person and at least 60 days prior to holding any public hearing on an implementation plan (or plan revision) for regionaI haze required by this subpart. This consuItation must include the opportunity for the affected Federal Land Managers to discuss their:

(i) Assessment of impairment of visibiIity in any mandatory Class I Federal area; and (ii) Recommendations on the development of the reasonable progress g o d and on the development and implementation of strategies to address visibiIity impairment.

Attachment 2 S

In adopting any SIP or SIP revision, there are certain requirements that States must meet. One such requirement is proper public notice and the availability of the SIP revision in at least one Iocation. 40 C.F.R. 4 5 1.102.

ADEQ has failed to meet these requirements for the issuance of the White Bluff pennit which essentially reflects a revision to the White Bluff BART detemination that was previously adopted into the Arkansas SLP and submitted to EPA for approval in 2008.

Specifically, ADEQ’s regional haze SIP submittal to EPA included Entergy’s 2006 BART Analysis for White BIuff in which the company found that a wet scrubber met BART for SO2 emissions from White Bluff. 2006 White Bluff BART Analysis at ES-1. In fact, Entergy found installation ofa wet scrubber to be more cost effective and to achieve a higher SO2 control efficiency than a dry scrubber. Entergy’s 2006 BART submittal and proposed BART controls are clearly part of the regional haze SIP that was submitted to EPA in July 2008 (in Appendix 9.3a).

However, in August 2008, Entergy submitted a revised BART andysis for White Bluff to ADEQ. Based on conversations Sierra Club has had with the Fish & Wildlife Service and the US. Forest Service, it appears that neither Entergy nor ADEQ provided the revised White Bluff BART anaIysis to the FLMs. Further, ADEQ did not give public notice of Entergy’s revised BART analysis for White Bluff as part of this current draR permit. Sierra Club found Entergy’s 2008 rwised BART analysis in the middIe of a file that is posted on ADEQ’s website titled “Entergy-ADEQ internal ernaiIs.pdf.”s That can hardly be considered adequate public notice.

Although Entergy’s rcviscd BART analysis did not propose any different SO2 or NOx emission limits than the presumptive BART emission limits adopted in Arkansas Rcgylation 19, Chapter 15, Entergy reached a much different concIusion in its 2008 BART analysis about the poIlution controls that the company believed satisfied BART for S02. Specifically, the company found that a dry scrubberhaghouse satisfied BART for S02, even though the company admitted that dry scrubbers achieved Iower levels of SO2 control than wet scrubbers. See August 2008 Revised BART Analysis for White Bluff at ES-I. Further the company revised its SO2 control cost analysis and now, the cost effectiveness of a dry scrubberhaghouse is shown to be very similar to a wet scrubber. I d at 3-5 and Appendix A. However, the detailed cost analysis that went into Entergy’s revised BART andysis and the conclusion that a dry scrubberlbrtghouse - instead ofthe previously proposed wet scrubber - satisfy BART has been withheld by Entergy fiom the pubIic as confidentiaI information.’

’ ftp://Iistserv.adeq.stnte.ar.us/AirPermitsrEntergyo/oZOWhitc%20Blu~.

’ Stc cover page to Appendix B of 2008 Revised BART Analysis which states “Notc the information will bc submitted at a h e r date with sworn affidavit for confidentiality ....” Nothing but the cover page to Appendix B was providcd in the version of the 2008 Revised BART Analysis that is on ADEQ’s wcbsitc, found within the file “Entcrgy-ADEQ internaI emails.pdf.”

9 Attachment 2

Thus, ADEQ has utterly failed to meet the regional haze and SIP requirements cited above in proposing to issue this permit which essentialIy provide ADEQ's final approval of BART controls at White Bluff. ADEQ has not provided the FLMs with the proper time period for FLM consultation, ADEQ has failed to provide the FLMs with Entcrgy's revised White Bluff BART analysis, ADEQ has failed to provide the public with proper notice of Entergy's revised BART analysis, and Entergy and possibly ADEQ have withheld an important section of Entergy's 2008 BART andysis that is necessary to conduct a proper review of Entergy's BART analysis.

Further, ADEQ has failed to conduct a proper five factor analysis of BART for White Huff as required by 40 C.F.R. 8 5 1.308(c)( I)(ii) and 40 C.F.R. Part 5 I , Appendix Y. Entergy's BART analyses also failed to include a proper five factor analysis of BART for White BIuff in accordance with 40 C.F.R. Part 5 1, Appendix Y. Both EPA's and the Fish & WildIife Service's comment letters to ADEQ on the draft regional haze SIP indicated that such an analysis was Iacking." The Fish & WiIdlife Service submitted a subsequent coinment letter on this draft permit that said the same.'' Below w e discuss why Entergy's BART andysis fails to meet those criteria and why neither the proposed BART controIs nor the BART emission limitations satisfy BART.

As previously stated, ADEQ has indicated that it would make its final dctcrmination of BART in the pennits issued for installation of the BART controls. Thus, this draft permit for White BIuff must reflect ADEQ's final determination of BART for White BIuff, and yet ADEQ provided no notice or a written review and determination that the proposed controls satisfy BART for White Bluff. Because of all of these flaws with proper FLM and pubIic notice and avaiIabiIity of information, ADEQ must re-notice the draft permit for White Bluff (after it makes its own proper determination of BART through a five factor analysis as outlined in 40 C.F.R. Part 5 I , Appendix Y and after it provides a proper consultation period to the FLMs), make all relevant infomation publicly availabk, and provide proper notice and opportunity for public hearing and to comment to the public.

B. Entergy's 2008 Revised BART Analysis for White Bluff Failed to Follow the EPA's BART Guidelines, and the Company's Proposed Pollution Controls and Emission Limits Do Not Reflect BART

1. Background of BART Requirements

As previously stated, the federal regional haze regulations require BART to be determined based on an analysis of the best availabIe system of continuous emission control technology and emissions reduction achievable for a BART eIigibIe source, taking into account the costs of control, any energy and non-air quality environmental benefits, pollution controls currently in use at the source, the remaining useful life of the

See EPA's July 2 1,2008 letter, Enclosurr nt 4-5. Ex. 3; see also June 2G, 2009 Icttcr from tho US.

November 20,2009 Letter from FWS to ADEQ, Ex. 1.

10

Department of Interior, Enclosure at I, Ex. 4. "

Attachment 2 10

source, and the improvement in visibility that will occur as a result of the controls bcing evaluated. 40 C.F.R. $51.308(e)( l)(ii)(A); Section 169A(g) of the Clean Air Act. In the EPA’s Guidelines for BART Determinations (BART Guidelines) in 40 C.F.R. Part 5 1, Appendix Y, EPA has spelled out a five step process to determine the level of control technology that represents BART for a particular source.

For BART-eligible power plants with generating capacity more than 750 MW like White Bluff, States are required to follow the BART guidelines in determining the controls and in adopting emission Iimits reflective of BART. 40 C.F.R. 8 51.308(e)( I)(ii)(B); Section I.F. 1. of Appendix Y, 40 C.F.R. Part 51. The five steps of determining BART are

STEP 1 -- Identify All [h 121 Available Retrofit Control TechnoIogies, STEP 2- Eliminate Technically Infeasible Options, STEP 3-- Evaluate ControI Effectiveness of Remaining Control Technologies, STEP 4-- Evaluate Impacts and Document the ResuIrs, and STEP 5 - Evaluate Visibility Impacts.

Fn 12: 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 Iist a11 permutations of available control levels that exist for a given technology - the list is compIete if it includes the maximum level of control each technology is capable of achieving.

Section N.D. of BART Guidelines at 40 C.F.R. Part 5 1, Appendix Y.

Clearly, EPA’s BART Guidelines require an evaluation of the top level of pollution reduction achievabIc with each control system evaluated in a BART analysis. EPA’s BART Guidelines provide that, if a control system can be operated at a wide range of control efficiencies, “the most stringent emissions control level that the technology is capable of achieving’’ must be evaluated. Section IV.D.3. of the BART GuideIines at 40 C.F.R. Part 5 1, Appendix Y. The BART GuideIines further require that “[YJou should consider recent regulatory decisions and performance data (e.g., manufacture+s data, engineering estimates and the experience of other sources) when identifying an emissions performance level or levels to evaluate.” Id.

The BART Guidelines aIso provide:

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 controI in the BART analysis. Without a showing of differences

Attachment 2 1 1

between the source and other sources that have achieved more stringent emissions limits, you should condude that the Ievel being achieved by those other sources is representative of the achievable level for the source being analyzed.

Id

Further, while one can consider varying levels of pollution control in evaluation of a particuIar controI device, one “must consider the most stringent level as one of the controI options.” Zd

In step 4, the BART Guidelines providc that, for a BART review, one is expected to quantify and report the costs of compliance, the energy and non-air quaIity impacts, and the remaining usefuI life for a11 of the available control options identified in Steps 1-3. Section IV.D.4. of the BART Guidelines. The State is “responsible for presenting an evaluation of each impact dong lvirh appropriate siipportiiig iilfoiitt~liioii.*’ id. (emphasis added.) Thc cost analysis is required to identify the emission units and their design parameters for emission controls and develop cost estimates based on those design parameters. Z d The cost estimates should be documented either with data from an equipment vendor or by a referenced source such as EPA’s Control Cost Manual. The Guidelines require that cost estimates be based on EPA’s ControI Cost Manual wherever possible.

EPA’s BART GuideIines include “presumptive BART” emission limits for units 750 MW or greater which were based on EPA’s broad review of the control tcchnoIogies and emission limits that could be met cost effectively at a wide range of coal-fired power plants. See Sections IV.E.4 and 5 of the BART Guidelines in 40 C.F.R. Part 5 1, Appendix Y. However, it must be stated that EPA’s presumptive BART limits do not negate the need for the state to determine BART for each BART-eIigible source on a case-by-case basis through a five factor analysis. In fact, EPA’s SART Guidelines make dear that a state may adopt alternative BART limits than presumptive BART based on a 5-factor analysis. Id. And the regdations and the Clean Air Act make clear that the determination of BART has to be source-specific. See 40 C.F.R. 5 1.30S(e)( l)(ii)(A); 9 I69A(g) of the CIean Air Act.

In spite of these regulatory and statutory requirements, ADEQ simply adopted the presumptive SO2 and NOX BART limits for Whitc Bluff in Reg. I9.1505(F) - (K), Chapter 15, of Arkansas PolIution and Ecology Commission Regulation No. 19, and faiIed to conduct a source-specific BART analysis for White Huff. While Entergy conducted a BART analysis in 2006 that was incorporated into the Arkansas regional haze plan and submitted to EPA for approval, Entergy later revised its BART determination, coming to a radically different conclusion on the most appropriate SO2 controls. In both submittals Entergy simply evaluated the controls to meet the presumptive BART limits adopted into Arkansas Regulation No. 19, Chapter 15, and did not conduct a proper five factor analysis. ADEQ cannot simply allow Entergy to meet the limits of Reg. 19.1505 of Regulation I9 with whichever poIIutant controls it wonts to

I2 Attachment 2

use and calI it BART. Furthermore, ADEQ is required to conduct its own five-factor analysis of BART. It cannot simply rely on Entergy’s 3ART analysis which was not a complete five factor BART analysis. As will be shown below, the proposed limits of Regulation 19 and Entergy’s revised BART analysis fail to reflect BART.

2. Entcrgy’s August 2008 Revised BART Analysis Does Not Reflect a Complete Five Factor Analysis of BART as Rcquircd by the Regional Haze Rules and EPA’s BART Guidelines

In 2008, Entergy submitted a revised BART analysis for White BIuff to ADEQ. The Executive Summary of Entergy’s August 200s revised BART analysis states that this analysis supersedes and replaces Entergy’s previously submitted 2006 BART Analysis for White Bluff. Entergy’s 2008 BART analysis failed to follow the procedures required by EPA’s BART Guidelines for determining BART controls and emission limits for White Bluff.

a. Entergy’s BART Analysis for the SO2 Emitted by White Bluff Is Flawed

Entergy evaluated two options for control of SO2 emissions - a lime spray dryer (dry scrubbing) and a forced limestone oxidation system (wet scrubber). For both types of SO2 controls, Entergy assumed the same emission level wouId be met with either - 0.15 Ib/MMBtu which just happens to be the presumptive BART limit that ADEQ adopted in Regulation No. I9 at APCEC Reg. 19.1505. Such an evaluation of BART controls simply to show the presumptive BART limit couId be met does not constitute a proper source-specific five factor BART andysis for SO2.

Regarding the presumptive BART limit, it must be noted that EPA said 95% SO2 control also reflected presumptive BART for SO2 The 0. I5 1blMMBtu presumptive BART SO2 emission limit simply reflected a level of control that EPA determined to be cost effective and achievable at all of the larger BART-eIigibIe sources. However, for a facility such as White BIuff, the 0.15 lblMMBtu SO2 h i t does not come anywhere near to reflecting the “degree of reduction achievable through the application of the best system of continuous emission reduction” of S02. According to Eniterg’s 2008 BART Analysis, White Bluffs primary fuel is subbituminous coal kom the Powder River Basin, and White Bluffs yearIy average SO2 emission rates has ranged from 0.645 -0.737 1bMMBtu from 2001 -1005. 2008 Revised BART Analysis at ES- I , 2-1. Because thcrc are no SO2 pollution controls currently instaIIed at either of the White Bluff units, these average emission rates reflect the uncontrolled SO2 emission rate at the inlet to a wet or dry scrubber. Thus, an SO2 emission limit of 0.15 lb/MMBtu reflects 76.7% to at best 79.6% removaI of SO2 at White Bluff based on the historical SO2 emission rates at each unit. A review of SO2 emission rates for more recent years (based on data from EPA’s Clean Air Markets Database) shows that the annuaI average SO2 emission rates at White Bluff have stayed within this range or lower. In 2009, averaged over the first three quarters of 2009 data, the SO2 emission rate at each unit averaged 0.62 1blMMBtu for which an SO2 limit of 0.15 lb/MMBtu only reflects 75.8% control. Whether considering

13

Attachment 2

a dry or a wet scrubber, there is no way such low levels of SO2 control could be considered to reflect the best system of continuous emission reduction. Indeed, as EPA indicated in its presumptive BART limitations for SO2, ninety-five percent control shouId be readily achievable with either a wet or a dry scrubber and low sulfur coal, and even higher Ievels of SO2 control should be achievable with a wet scrubber.

Further, as previously discussed, EPA's presumptive BART limits do not negate the need for a source-specific BART andysis pursuant to the five factor analysis. Thus, Entergy erred as a matter of law in sirnpIy evaluating SO2 controls to meet the 0.15 1blMMBtu presumptive BART limits. Moreover, ADEQ has arbitrarily and capriciously accepted Entergy's BART determination via its proposed issuance of the current draft PSD permit for White Bluff without any independent review of BART for White BIuff pursuant to a proper five factor analysis. As discussed previously, this is entirely inconsistent with what ADEQ said in its regional haze SIP that was submitted to EPA - Le., that ADEQ's approvaI of each company's BART anaIysis in accordance with the five factors would be finalized in its issuance of the PSD permit for the installation of BART controls."

i. Etitergy 3 Aiialysis of BART for SU2 Milst Be Based m i the Characteristics of the Cod it Cui-renrly Burns cri White Blrflti~tdtr its CurreJif Air Perntil.

Entcrgy had made clear that subbituminous Powder River Basin coal is the coal it has historically burned at White Bluff and its 2009 PSD permit application indicates it is the coal that will typically be burned at White Bluff. See August 2008 Revised BART Analysis at ES-I, and January 2009 PSD Application at 2- I . Yet, in its BART analysis, Entergy assumed an uncontrolled SO2 emission ratc of 2 lblMMBtu in its evaluation of a dry scrubber and an uncontrolled SO2 emission rate of 3 lb/MMBtu in its evaluation of a wet scrubber. The choice of these uncontrolkd SO2 emission rates is cIearIy arbitrary since the projected controlled emission rate with each type of control and Entergy's assumed 92.5% control efficiency with a dry scrubber and 95% controI with a wet scrubber equaIs 0.15 Ib/MMBtu (the presumptive BART limit for SO2). However, in projecting the uncontrolled SO2 emission rate for White Bluff that is different from the historical uncontrolled SO2 emissions, Entergy must provide coal analysis data of the other coals it plans to burn at White Bluff. Further, Enterby cannot assume uncontrolkd SO2 emission rates that arc not allowed under its current permit.

We arc aware that Entergy obtained a permit modification in 2006 to allow it to burn bituminous coal in addition to subbituminous coal.13 Section IV, Condition 26 of the current White Bluff permit, which was added with the 2006 permit modification

'' Scr: rg., ADEQ's rcsponsc to comments undcr'Talking Points - 4/17/08 Conference Call Between FLM Agcncy Staff and ADEQ Regarding AR Draft Regional Hnzc: SIP" in Appcndix 2.1 of the RegionaI Haze SIP at 25 in which ADEQ states "ADEQ hns not oficially approved the subjcct-to-BART SOUTCCS'

cngineering nnnlyses. These analyses will be reviewed and approved or rejected during the permit process."

Pcrmit Number %3-AOP-R4, Ex. 5. The previously-issued pcrmit for Whitc Bluff limited thc facility to burning coal from northeastern Wyoming (be.. Powder River Basin subbituminous coal).

I4 Attachment 2

allowing the burning of bituminous coal at White Bluff, limits the sulfur content of the coal burned to 0.72% unless a higher level is alIowed by the Equation in Condition 26. The intent of the equation is to allow higher sulfur coal but only if the coal has lower ash content.

To assess whether the 2 Ib/MM3tu and 3 lblMMBtu uncontrolIed 502 emission rates would be allowed by the existing White Bluff permit, we used AP-42 emission factors to back-calculate the sulfur content of the fuel that each of these uncontrolled emission rates would reflect at the inlet to the dry or wet scrubber“. Because these are very high levels of uncontrolled SO2 emission rates from Powder River 3asin subbituminous coal, we assumed that the uncontrolIed emission rates used by Entergy in its SO2 BART analysis were for bituminous coal. Thus, we used the “38s’ emission factor of Table 1.1-3 of AP-42, for a pulverized coaI tangentially fired boiler subject to NSPS burning bituminous coal and we assumed a range of heat vaIues of bituminous coal from 10,000 Btullb to 13,000 Btu/lb to convert the 385 lblton emission factor to units of lblMMBtu so that the percent sulfur content allowed by Entergy’s assumed uncontrolled SO2 emission rates of2 IblMMBtu and 3 IblMMBtu couId be estimated. Based on the assumed heat content range of bituminous coal, we determined that the 2 Ib/MMBtu uncontroIled SO2 emission rate assumed by Entergy to reflect emissions at the inlet of a dry scrubber reflected a sulfur content of 1.05% to 1.26%, which is much higher than the 0.72% allowed by thc current permit, Condition 26. The 3 lb/MMBtu uncontrolled SO2 emission rate assumed by Entergy to reflect emissions at the inlet of a wet scrubber reflects a coal sulfur content of 1.58% to 1.90%, we11 in excess of the permit’s 0.72% coal su lhr content limit.

We then andyzed the equation in current Permit Condition 26 to assess whether cod with this high of sulfur content wouId be alIowed under the equation. Permit Condition 26 would not alIow for the burning of coal of such high suIfur content unless the ash content was very low. For example, to burn cod with a sulfur content of 1.05% and n heat vaIue for bituminous coal of 10,000 Btu/Ib. the coal would be required to have an ash content of under 3%. That is an extremdy Iow level of ash content for either bituminous or subbituminous coal. Assuming coal with a sulfur content of 1.05% and a heat value of 12,000 Btullb, the coaI wouId be required to have an ash content of under 3.5%, which is stilI very low for bituminous or subbituminous coal. Thus, it would be virtually impossible for White Bluff to comply with Permit Condition 26 and burn coal with uncontrolled SO2 emissions at the inIet to the scrubber of 2 Ib/MMBtu, much less 3 IblMMBtu.” This shows it was entirely inappropriate for Entergy to assume such high levels of uncontrolled 502. emissions at the inlet to the SO2 controls in determining the achievable emission rate with a dry or a wet scrubber.

’‘ The AP-42 SO2 emission factors for bituminous and subbituminous coal combustion in Table 1.1-3 of Chaptcr 1.1 of EPA’s AP-42 emission factors take into account thc fact that 510% of the sulfur in coal drops out in the bottom ash. Scr footnote b of Table 1.1-3. available nt h t t p : / / ~ ~ . e p a . ~ o ~ f ~ c h i e f / a p 4 ~ c h O 1tindex.html. ’’ If a highcr sulfur contcnt is assumed such as thc 1.58% to 1.90% sulfur content that a 3 IblMMBtu uncontroiled SO:! emission rate reflects. that means the ash content of the coal has to be much lower than 3% in order to comply with existing permit Condition 26.

15

Attachment 2

The current draft permit excludes condition 26 and does not replace it with any comparabIe limitations. As is discussed further below in this comment letter, such a change in the permit cannot be alIowed without all PSD requirements being met. A change in coal to a higher sulfur andor higher ash coal is a change in the operation of White BIuff that must trigger a PSD analysis, including impacts on the NAAQS, PSD increments, and air quality related values of Class I areas. It is unlawfil for ADEQ to allow for this change without a proper PSD review.

While no coal specification data was provided in Entergy’s 2008 BART Analysis for White Bluff, Entergy did provide some coal specifications in Section 4.5.1 of Appendix C of its 2009 PSD Permit Application. Specifically, Appendix C of Entergy’s 2009 PSD permit application indicates that Powder River Basin coal is the design c o d for the dry scnibber. Appendix C to PSD Permit Application at 8. Appendix C identifies typicd coaI as having a sulfur content of 0.35% and design coal of having a sulfur content of 0.88%. Such a high sulfur content and uncontrolled SO2 emission rate of 0.88% is virtually unheard of for Powder River Basin coal. A review of coal data for power plants that bum Powder River Basin coal shows that the sulfur content of Powder River Basin subbituminous coals typically falls within a range of 0.7 - 0.4% sulfur and that the heating vaIues typicalIy range from 8,500 to 8,800 Btu/lb.Ih Entergy has not provided any citations to the design coal data in Appendix C of its permit application to show that these design characteristics are bascd on any Powder River Basin mine data. Thus, it is entirely unjustified to assume such a high uncontroIIed SO2 emission rate in the BART analysis for SO2 at White Bluff.”

Entergy’s 502. BART analysis and determination of achievable SO2 emission rates must be based on the coal the unit currentIy burns, is permitted to burn, and intends to burn in the future - i .c, Powder River Basin subbituminous coal. The purpose of BART is to reduce emissions from current levels to improve visibility, and that purpose will be ignored if BART limits are determined based on coals that White Bluff may potentially utilize in the fiturc. Instead, the SO2 BART limit should be based on the coal White Bluff is currently burning and has been utilizing for quite some time. In addition, as shown above, it is difficult to see how the current White Bluff Permit Condition 26 would alIow anything but Powder River Basin coal or low sulfur bituminous coal. Based on a review of average SO2 emission rates at the White Bluff units over the past 12 years, the highest the uncontrolIed SO2 rate has ever been is 0.74 IbMMBtu.’’ Thus, at the maximum, 0.74 lb/MMBtu should bc the uncontrolled SO2 cmission rate from which achievable SO2 emission limits with SO2 pollution controls are determined. AlternativeIy, if Entergy intends to switch coals andlor get a revision to its permit to

See 2000 Directory of Powcr Phnts Burning Wyoming Coal, Wyoming Statc GeologicaI Survey. Ex. 6.

In Appendix C of its permit application (at page 3), Entergy assumed M unconirollcd SO2 rate at thc inlet of thc scrubber to bc 2 IblMMBiu which is much higher than even tho characteristics of its “design coal” and fnils to account for the sulfur that falls out in the bottom nsh.

This is bascd on annual SO2 cmissions dividcd by annual heat input at each White Bluff unit from EPA’s Clean Air Markets Database.

I 6

17

IS

Attachment 2 16

allow it to burn higher s u h r coals at White Bluff in the hture, ADEQ should impose both an SO2 eniission limit and a minimum contrd efficiency requirement as BART.

ii. Entergy Failed to Evaluate [he Most Stringent Lwei of Corttrol Achievable with the SO2 Controls Evaluated as Om of the BART options.

in addition to basing its 3ART analysis for SO2 on incorrect and improper assumptions of uncontrolled SO2 emission rates, Entergy also failed to evaluate the most stringent level of SO2 removaI efficiency achievable with the wet and dry scrubbers evahated in its BART analysis for White Bluff. For its analysis of a wet scrubber, Entergy assumed 95% control was achievable. 2008 Revised BART Analysis at 3-5. However, higher leveIs of SO2 removal efficiency are readily achievable with both wet and dry scrubbers, even with low sulfur coals.’9

a) Enter= Failed to EvaIuate the Levels of SO2 Control Achievable with a Wet: Scrubber at White Bluff.

Wet scrubbers can achieve 99% removal efficiency. A prime exampIe is the Chiyoda CT- 12 1 FGD. Vendor information for this technology indicates that this scrubber has achieved 98-99% SO? removaI even with Iow sulfur coal.” For example, the Chiyoda’s bubbling jet reactor has consistently achieved >99% SO? removal during long-term operation at the Shinko-Kobe power plant in Japan. This facility consists of two 700-MW coal-fired utility boilers. The wet FGD was designed to achieve 0.0 14 Ib S02/MMBtu (9 ppmv at 3% oxygen) on an instantaneous basis and has consistently exceeded this Ievel of control while treating gases with inlet SO? concentrrltions of 1.78 1b/MMBtuv2’ This technology has been guaranteed by Chiyoda to achieve 99% SO:! removal on t h e e coal-fired boilers in Japan.= It also has been demonstrated in the U.S. at the University of Illinois’s Abbott power plant, Georgia Power’s Plant Yate?, Dayton Power & Light’s KiIlen Unit p4 and Plant Bowen Unit 3.3 It has aIso been licensed for installation on several additional units in the US, including the other three units at Plant Bowen in Georgia, the other units at Dayton Power & Light’s Killen plant, Dayton Power & Light’s Stuart plant, and AEP’s Big Sandy Unit 2, Conesville Unit 4,

See Sargent & Lundy. FIue Gas Desulfurization Technology Evaluation, Dry Lime vs Wct Limestone

See Black & Veatch vendor brochurc on CT-121. Ex 8.

I9

FGD, March 2007, Ex. 7..

” YasuIiiko Shimogama. Hiroknzu Yxsuda, Nnohiro Kaji, Fumiaki Tannka, and David IC Harris, Commercial Experience of the CT-121 FGD Plant for 700 MW Shinko-Kobe Electric Powcr Plant. Paper No. 37, presented at MEGA Symposium, Air & Waste Management Association. May 19-22.2003. W.9.

’* CT-I 21 FGD Proccss - Jet Bubbling Reactor. ht~n:, ‘ ~ ~ ~ ~ ~ v . b w c . d W f ~ ~ l l - c t l ~ 1 .litml,

XI Emission-controI Technologies Continue to CIcar the Air. Power, MaylJunc 2002.

” Sce BIack & Veatch, First Black&VentchlChiyoda Wet Flue Gas Desulfurization System in North America Successfully Goes Operational, Ex. 1 1.

http:llpepei.pennnet.cornldisplay~articIe/342997/6/ARTClJnonelnonel l/Go-Take-a-Batlll.

Ex. IO.

Blankinship, Steve, Go Take a Bath, Power Engineering, October 2008. available at 3

. ’

Attachment 2 17

Cardinal Units 1 and 2, and Kyger Creek, among others.” Black & Veatch and Southern Company are both US. licensees. Further, this tcchnology also has shown to be very effective in removing fine particulates, oxidized and elemental mercury, and acid gases, and the technology uses less energy compared to traditional wet scrubbers.

Further, Mitsubishi, a vendor of scrubber systems, reports it has uaranteed SO1 removal efficiencies up to 99.8 percent, including for coal-fired boilers: R . z s . x l

Finally, a recent Lake Michigan Air Directors Consortium (IZADCO”) and the Midwest Regional Planning Organization (“MRPO”) presentation indicated that advanced FGD technologies could achieve 99.5% control for S 1,240 to $2,875 per ton of SO? removed and wet FGD could achieve 99% SO2 control for $1438 1 to $3,440 per ton of SO2 removed. Ex. 17 These costs are well within the range that EPA normally considers to be cost effective in best available control technology (BACT) analyses.

Considering the Powder River Basin subbituminous coal that has been and will continue to be burned at White Bluff, Entcrgy’s proposed BART emission limit of 0.15 Ib/MMBtu only reflects, at best, eighty percent removal (assuming a worst case uncontrolled emission rate at the inkt to the scrubber of 0.74 Ib/MMBtu). Given the historical uncontrolled SO:! emissions at White Bluff, the most stringent BART emission limit that should have been evaluated in the BART analysis would be 98% control horn 0.74 lb/MMBtu - or an emission limit of approximateIy 0.0 15 IbMMBtu. As shown above and in the refcrenccd exhibits, this level of emissions has been achieved. Thus, the Entergy White Bluff BART AnaIysis shouId have evaluated this level of control.

The Ieast stringent leveI of SO2 control with a wet scrubber that should have been evaluated in the White Bluff BART analysis should be no less than 95% SO2 control. Such levels of control are standard today with a wet scrubber?’ Even Entergy claimed 95% removal could be achieved with a wet scrubber. 2008 Revised BART AnaIysis at 3- 5. Ninety five percent SO2 control eom 0.74 lb/MMBtu uncontrolled SO2 emissions at the inlet to the scrubber would equal an emission limit of 0.04 lb/MMBtu.

An annual average was compiIed of SO2 emission mres for 2008 using data submitted to the Clean Air Markets website by similar coal-fired electric generating units.

2b Chiyoda Liccnscs Its FIuc GRS Desulfurization Tcchnology in USA Newly for 5 Coal-Fired Generation Units. Press Release. May 2.2005, Ex. 12; Chiyoda Licenses its Flue Gas Desulfurization Process in USA for Georgia Power Owned 4 FGD Units, January 26.2005.E~. 13 . ’’ Jonas S. Klingspor, Kiyoshi Okazoe, Tetsu Ushiku. and George Munson, High Efficiency Double Contact Flow Scrubber for the US. FGD Market. Paper No. 135 presented at MEGA Symposium, Air & Waste Mnnngemcnt Association. May 19-22.2003. p.S. Table 4,Ex. 14.

Test Facility of MHI Single Tower FGD. Ex. 15.

19 Mitsubishi High SO2 RernovaI Experience, Ex. 16.

’’ See discussion of Acid Gad307 Control. Technologies on the Institutc of Clean Air Cornpanics website at http:flwww.icac.comli3~pageslindex.cfm?pa~eid=340 1 .

Yoshio Nnktlyama, Tetsu Ushih. and Takco Shinoda, Commercial Expcricncc and ActuaI-Plant-Scalc

IS

Attachment 2

The annual avera.ge SO2 emission rates were ranked from low to high to identi@ the best performing similar sources. This analysis is shown in Ex. IS. The best performing similar source in 200s was Pleasant Prairie Units 1 and 2 in Wisconsin. The 200s annual average achieved at Unit 1 was 0.021 IblMMBtu and at Unit 2,0.027 IblMMBtu. These units are equipped with a wet limestone scrubber and burn a low sulfur Powder River Basin Coal. Similar data for the first six months of 2009 indicate that other units are currently achieving even Iower SO? emissions, including Iatan Unit 1 at 0.005I IblMMBtu; Muscatine Unit 9 at 0.0 I3 lb/MMBtu; Harnmond Unit 2 at 0.0 1 6 Ib/MMBtu; Gorgas Unit 10 at 0.017 IblMMBtu; Prairie Creek Unit 4 at 0.019 lb/MMBtu; Hopewell Power Station Units 1 and 2 at 0.020 Ib/MMBtu; and Centralis Unit BW22 at 0.021 IbMMBtu. Florida Power and Light filed an application in March 2007 for the Glades facility with a SO2 BACT limit of 0.04 lb/MMBtu based on a 24-hour rolling average. The design fuel is 3.21 lb/MMBtu SO1 and 12,324 Bt~dlb.~' This facility was rejected by the Florida Public Service Commission for reasons unrelated to the BACT analysis, but the proposed SO2 BACT emission limitation nevertheless demonstrates that FP&L considered this limit to be achievable.

In summary, Entergy's evaluation of a 0.15 Ib/MMBtu SO2 emission limit faiIs to reflect the emission mtcs and levels of SO2 control achievable with a wet scrubber and the typical Powder River Basin coal. In accordance with EPA's BART Guidelines, Entergy should have evaluated the most stringent level of control achievable with a wet scrubber. And, in no event, should Entergy have assumed less than 95% removaI efficiency achievable with a wet scrubber, even with low suIfur Powder River Basin coal.

b) Entersy Failed to Evaluate the Levels of SO2 Control Achievable with a Dry Scrubber at White Bluff.

A dry scrubber can achieve 95% SO2 removal efficiency?' A new dry scrubber retrofit at White Bluff should be able to meet the same levels of control as dry scrubbers that are proposed for new coal-fired power plants. Thus, Entergy should have conducted a revicw of recent proposed and final BACT determinations as part of assessins the emission rate that could be achieved with a wet scrubber. There have been several proposed coal-fired power plants burning low suIfur Powder River Basin coal that have proposed to use dry scrubbers to meet PSD requirements and that are subject to much Iower SO2 BACT Iimits than 0.15 IblMMBtu. Those facilities include the Newmont Nevada TS power pIant, the proposed White Pine power plant, Toquop, and the Dry Fork power plant, The Newmont Nevada power plant is subject to a minimum 95% SO2 removal efficiency requirement when burning coal with a sulfur content equal to or greater than 0.45% and is subject to a minimum 9 I % SO2 removal efficiency when burning coal with sulfur content less than 0.45%?3 This facility is currently operating in compliance with its limits. The Newmont Nevada is aIso subject to an SO2 BACT limit

'I FP&L Glades Response to EPA Commcnts, March 26,2007, pp. 3-5. Ex. 19.

'' See discussion of Acid GaslSO:! ControI Technabgies on the Institute of Clean Air Companies website nt ht tpJlw~~.icac.co~i4~pagcs/indcx.cfm:id=340 1.

'' See Section V.A.2.a.8. of Newmont Nevada Permit. Ex. 20.

Attachment 2 19

of 0.065 lb/MMBtu when burning cod with less than 0.45% sulfur content. The proposed Toquop permit included an SO2 BACT limit of 0.06 lbPMMBtu on a 24-hr average basis.34 The Dry Fork power plant in Wyoming, which is also currently under construction, will burn Powder River Basin coal, will be equipped with a dry scrubber, and is subject to an SO2 BACT limit of 0.07 IblMMBtU? Other examples of low SO2 emission limitslhigh SO2 removal rates being required as BART can be found in the National Park Service's spreadsheet of BACT determinations for coal-fired electrical generating units, Ex. 23.

As stated above, Entergy's proposed BART emission limit of 0.15 lb/MMBtu only reflects, at best, SO% SO2 removal (assuming a worst case uncontrolled emission rate at the inlet to the scnibber of 0.74 IbMMBtu). Given the historical uncontrolIed SO2 emissions at White Bluff, the most stringent BART emission limit that should have been evaluated in the BART analysis for a dry scrubber would be 95% control from 0.74 lb/MMBtu - or an emission limit of approximately 0.04 lb/MMBtu.

Entergy identified a 92.5% control efficiency as achievable with a dry scrubber. See 2008 Revised BART AnaIysis for White Bluff at 3-5. An SO2 control eficiency of 92.5% from 0.74 lb/MMBtu uncontroIled SO2 emissions at the inIet to the scrubber would equal an emission limit of 0.06 1blMMBtu. The least stringent level of SO2 control with a dry scrubber that should have been evaluated in the White BIuff BART analysis should be no Icss than 9 1 % which was rcquired as BACT for the Newmont Nevada power plant when the facility burns coal with sulfur content less than 0.45%. A 9 I % control efficiency from the 0.74 IbMMBtu uncontrolled SO2 emissions at the inIet to the dry scrubber would equate to an emission limit of 0.07 lb/MMBtu

In summary, Entergy's evaluation of a 0.15 Ib/MMBtu SO2 emission limit utterIy fails to reflect the emission rates and levels of SO2 control achievabIe with a wet scrubber and the typical coal to be burned at White Bluff. Enterm shouId have evaluated 502 emission limits for a dry scrubber in the range of 0.04 to 0.07 Ib/MM3tu.

iii. Eittergy 's Cost Per Toit Analysis is Flawed Because it Fails to Reflect the Edssiorr Redirctioits Achievnhle with Either. u Wet or Diy Scrtrbber.

Entergy determined cost effectiveness of both a wet and a dry scrubber at White BIuff, but its assessment is flawed because it failed to take into accowt the achievable SO2 control efficiencies and the currently used low sulfur Powder River Basin coal in detcmiining cost effectiveness. h addition, in its cost per ton analysis provided in Appendix A of its 2008 BART analysis, Entergy incorrectly inflated baseline emissions. Specifically, it appears that Entergy calculated base case SO2 emissions for each unit based on the highest hourly emission rate and assuming 8,760 hours per year and 85 % capacity, rather than based on the actual annual emissions of each unit that has occurred

See Section V.ALa.(S) of draft Toquop permit. Ex. 21.

See Dry Fork PSD Permit. Ex. 27. 3s

Attachment 2 20

in the recent past. Table A- 1 of Entergy’s 2008 Revised BART Analysis identifies the base case SO2 emissions of White Muff Unit 1 and Unit 2 to be 28,902.8 tpy and 29,132.5 tpy, respectively. Yet, Entergy identified the actuaI annual SO2 emissions of each unit to range from 16,679.2 tpy to 21,653.4 tpy for Unit 1 and from 13,335.0 to 22,979.4 tpy for Unit 2 during 200 1 to 2005. See TabIe 2-1 of Entergy’s A u y s t 2008 Revised BART Analysis (at 2- 1).

It also appears that the costs of SO2 controls with both a wet and a dry scrubber at the White Bluff units are higher when compared to the costs of such controls at other coal-fired power plants. However, the details of Entergy’s SO2 cost analyses are not publicly available, as the details appear to be in Appendix B which Entergy has requested ADEQ withhold from the p u b k as confidential. Such data is very relevant to a proper review of the White BIuff BART analysis and cannot be withheId 6om the public. See Section IV.D.4. of the BART Guidelines in 40 C.F.R. Par t5 1, Appendix Y. Thus, ADEQ must not withhold Appendix B of Entergy’s 2008 BART ana1ysis from pubIic review. In any case, because it has thus far been withheld from the public, we cannot at this time offer a detailed critique of Entergy’s SO2 controls cost estimates.

In our first review of Entergy’s cost estimates, we compared Entergy’s annualized costs of a wet scrubber and a dry scrubber from Table A-I of Appendix A of Entergy’s 2008 Revised BART Analysis to estimates of similar SO2 controI installations at other coal fired power plants. Entergy projected an annualized cost of installing a wet scrubber to be $68,045,000 and an annualized cost for installing a dry scrubberhaghouse to be $65,155,000. These estimates are much higher than Entergy’s previous annualized cost estimates presented in Table A-I of its 2006 BART Analysis for White Bluff of $17,033,734 for a wet scrubber and $34,035,909 for a dry scrubber and baghouse. No explanation has been provided as to why thc costs two years later are 4 times higher now for a wet scrubber and almost two times higher now for a dry scrubber and baghouse. In comparison, the proposed BART controls for the Boardman power plant in Oregon, a G I7 MW unit, indicated a dry scrubber and baghouse would have an annualized cost of $36,600,000 (or $59,3 19/MW).36 The proposed BART control for Clay Boswcll Unit 3, a 375 MW unit, indicated a wet scrubber was estimated to have an annualized cost of S 17,933,022 (or $4732 llMw).37 The attached spreadsheet created by the National Park Service of SO2 BART data includes several more examples of projected annual costs of both wet scrubbers and dry scmbberslbagbouses which are much lower than Entergy’s assuming a $65 ($76,652.Mw) to $68 million ($80,053/MW) price tag However, without seeing the detaiIs of Appendix 13 of Entergy’s 2008 BART analysis, it is difficult to further pinpoint errors or flaws in Entergy’s cost estimates for SO2 controls at White Bluff,

To address the other deficiencies in Enter@ cost effectiveness (k, failure to evaluate the achievable emission limits and the improper inflation of base case

’’ National Park Service BART Evaluation Sprcadshcct cntitlcd “EGU Proposed SO2 BART ControI Spreadsheet,’’ Ex. 24.

37 id

Attachment 2

emissions), we recalculated cost effectiveness for the SO2 controls. First, we corrected the base case emissions to reflect actual annual emissions. Since we are using worst case actual SO2 emission rates in determining the achievable emissions limit with each SO2 controI, we decided to use the two worst case years of emissions provided in Table 2- 1 of the August 2008 Revised BART Analysis for White Bluff. Specifically, we used years 2003 and 2004. Second, we used the “Ieast stringent leveI of control that should have bcen evaluated as BART” that we discussed in the preceding section - Le., no less than 95% control from typical c o d with uncontrolled SO2 emission rate at the inlet of the scrubber of 0.74 IblMMBtu, which equates to an emission limit of 0.04 IblMMBtu as achievable with a wet scrubber, and no less dian 91 % control with coal with an uncontrolled SO2 emission rate of 0.74 IbpMMBtu, which equates to an emission limit of 0.07 IblMMBtu as achievable with a dry scrubberlbaghousc. We then projected future emissions, assuming these emission rates, the revised heat input capacity of each boiler that has been proposed with this permit action (k, S,950 MMBtulhr) and assumed each unit would operate at 85% capacity as Entergy assumed in its cost effectiveness caIculations. See TabIe A-1 of Appendix A of ZOOS BART Analysis. We used Entergy’s estimated annual costs of a wet scrubber and dry scrubberhaghouse unchanged (as provided in Table A-I of Appendix A). The foIlowing table shows the resulting cost effectiveness determinations.

Base Case Wet Scrubber Drv ScrubbedFF

‘50 Removal so2 Annualized %/ton of SO2 Emissions, tpy costs Removed

21,586 95% 1,333 $68,045,000 $3,360 91% 2.332 1 $65-155.000 $3.384

Base Case Wet Scrubber Dry ScrubberlFF

As the table shows, when the variations in control effectiveness are properly evaluated with the uncontrolled SO2 emission rate of the coal currently being utilized at White Bluff, the cost effectiveness of a wet scrubber is slightly less than the cost effectiveness of a dry scrubberhaghouse. Further, a comparison of these cost effectiveness numbers to other cost effectiveness estimates for other sources’ SO2 BART controls (as summarized in the attached National Park Service BART Evaluation Spreadsheet) shows that these costs are not outside of the range of the cost effectiveness of similar controls that have been proposed as BART for SO2 at coal-fired power plants.

20,3 15 95% 1,333 $68,045,000 $3.585 91yo 2,332 $65,155,000 $3,623

While we do not believe the 3ART limits for SO2 shouId be based on the worst case coal Entergy plans to burn at White Bluff in the f h r e , it does appear that in the future Entergy plans to burn higher sulfur content coal at White Huff. For example, a Fcbntary 18,2008 e-maiI fiom Thomas Rheaume to Mike Bates in the Entergy - ADEQ emakpdf file on the White BIuff permit websitc indicates that Entergy may pursue a permit change to allow it to burn or bIend with lignite cod. Further, Entergy’s 2008

22

Attachment 2

Revised BART Analysis seems to be based on thc possibility that White Bluff might burn higher sulfur coal in the future. Thus, we decided to determine cost effectivcncss for thc burning of a higher sulfur coal. SpecificalIy, we assumed a 2 lb/MMBtu uncontrolled emission ratc at the inlet to the SO2 contro1 device. We assumed the “least stringent level of control that should have been evaIuated as BART” for the wet scrubber of 95% control which equates to an emission limit of 0.10 IblMMBtu. For the dry scrubberhaghouse combination, we assumed the presumptive BART limit of 0.15 lbMMBtu which would equate to an SO2 removal efficiency of 92.5%, the same removal eficiency Entergy assumed in its evaluation of a dry scrubber. Everything else in this set of cost effectiveness calcdations remained the same as our previous calculations.

The results were that with a higher sulfur coal, the cost effectiveness of a wet scrubber in terms of $/ton of polIution removed is 56% Iower than the cost effectiveness of a dry scrubber. Specifically, the cost effectiveness of a wet scrubber with higher sulfur coal was $3,727/ton and $4,007/ton at White BIuff 1 and 2, respectively, while the cost effectiveness of a dry scrubberhaghouse was $3,92S/ton and $4,254/ton at White Bluff 1 and 2, respectively. What this shows is that, if Entergy plans to burn higher sulfur coal at White BIuff in the future, then installation of a wet scrubber will clearly be more cost effective than a dry scrubberhaghouse. Further, this analysis assumed what we contend are the least stingent levels ofS02 control achievable with a wet and a dry scrubber. Higher levels of control and lower emission limits could be achieved. For example, a wet scrubber designed to meet 98% removd efficiency, which was discussed further above is achievable, could meet an SO2 emission Iimit of0.04 lb/MMBtu with an uncontrolled SO2 emission rate at the inlet to the scrubber of 2 IbhfMBtu. Thus, a 0.04 IbMMBtu BART limit for SO2 based on current Powder River Basin coal couId still be met at White Bluff if it bums higher suIfbr coal in the future. However, it is not as clear that a dry scrubber could achieve the 96.5% removal eficiency when burning c o d with a scrubber inlet emission m e of 2 lb/MMBtu that wouId be needed to meet a limit of 0.07 lb/MMBtu BART.

Thus, use of a wet scrubber at White Bluff wouId result in Iower SO2 emissions than a dry scrubber and wouId aIIow for some flexibility in suIfur content of the coal while stilI achieving Iow SO2 mission rates, all at a reasonabIe cost.

iv, Etitergy Did Not Disclose the Eiiergy Inpacts of n Wet Scrubber versiis CI Dty Scrrtbbcr A its Pitblicly Avuilable BARTAnalysis.

Entergy stated that the energy impacts of a wet scmbber are higher than that of a dry scrubber due to the fan power required for the increased pressure drop across the scrubber. 2008 Revised BART Analysis at 3-5. However, the details of these power demands were not provided except in Appendix B to the revised BART analysis which has not been made pubIicIy available.

In its PSD Permit Application, Entergy did provide some information on the energy impacts of a dry scrubberhaghouse. Specifically, Entergy stated the new air

Attachment 2

pollution control systems would result in a 6% increase in parasitic loads. January 2009 White Bluff PSD Permit Application at 1 - 1. That is a very high assumption for the parasitic load of the planned pollution controls. In comparison, a recent BART analysis prepared for the Big Stone I power plant by Burns & McDonneIl indicated that the parasitic load of a wet scrubber for this 475 MW (gross) eIectricaI generating unit was 3.0% of nomina1 generation (or 7.5 MW) while the parasitic load of a dry scrubber wouId be only 0.7% of nominal generation (or 3.3 MW)?8 In addition, an expert for Constellation Power Source Generation Inc. indicated in testimony before the Maryland Public Service Commission that the addition of wet scrubbers and baghouses to the 1,370 MW (2 units) Brandon Shores coaI-fired power plant would add a parasitic load of approximately 35 MW in total for both units, which is only 2.5% of the plant’s nominal generati~n?~ These projections are consistent with other findings?0m4’ So, the difference in parasitic load of a wet scrubber versus a dry scrubber and a baghouse is probably closer 0.5-1 %.

Thus, since it appears Entergy overestimated the parasitic Ioad due to the SO2 pollution controls evaluated, Entergy likely overestimated the cost of the auxiliary power to operate these controls in its dctermination of costs for the SO2 COQWOI options.

v. Entergy Did Not Fiilly Evaluitte the Enviroiit~~ntal Ii~ipacts of a Wet Scrubber versus a Dgr Scrubber in its Pirblicly Avnilable BART Andysis.

In the 2008 Revised BART Analysis, Entergy listed several negative environmental impacts with wet scrubbing and none for dry scrubbing. 2008 Revised BART Analysis at 43. However, there are some major benefits of wet scrubbing versus dry scrubbing that go well beyond SO2 control.

Specifically, the higher SO2 removal efficiencies and very Iotv SO2 emission rates, on the order of single digit parts per rnilIion (ppm) concentrations, wiIl be needed for the effective removaI of carbon dioxide (CO2) from the gas stream. Many of the amine-based CO2 controI methods currently under development are very sensitive to sulfur and thus require very low SO2 inlet concentrations, on the order of 1 to 2 pprn.4’ This will require 98-99%+ SO2 removal or an outlet SO2 of 0.01 Ib/MMBtu. It will be more cost effective and operationalIy simpler to design and instal1 controls in one retrofit program.

38 See TabIe 3.4-5 of November 7009 Big Stone I BART Analysis.

the PubIic Service Commission of Maryland, Case No. 9075, October 23,2006. at 6, Ex. 25.

see ulso Lange, Inn et nl., P o k y Innovation Impacts on Scrubber EIectricity Usage. at 1, Ex. 27.

March 2007. Table 5.5-2. Ex. 7. ‘’ Chuck Dene, Lesley A. Baker, and Robert J. Kecth, FGD Performance CapabiIity. Mega 2003, Ex. 18.

See Direct Tcstimony of Dori J. Costa on behalf of ConstclIation Power Source Gcneration, Inc., bcfore

&e U.S. EPA, ControIling SO:! Emissions: A Revicw of Tcchnologies, EPA-6OO/R-OO-O93 at 35,Ex. 26;

Sargcnt & Lundy. FIue Gas Desulfurization Technology Evaluation, Dry Limc vs Wet Limestone FGD.

EJ

40

41

24

Attachment 2

It is well recognized that it is not a matter of if but when Congress and/or EPA wiIl mandate C02 reductions from industria1 sources such as White Bluff. Thus, if an SO2 control technoIogy will better prepare White BIuff to be abIe to effectively remove CO2 in the future, that must be taken into account in the BART analysis as another environmental benefit from a wet scrubber versus a dry scrubber. Indeed, as described above, there are wet scrubber technoIogies available that can remove 99+% of the 502. Dry scrubbers do not achieve as high IeveIs of SO2 emission reduction. Given that Entergy has expressed interest in burning higher sulfur coal in the future such as lignite, the differences between SO2 emission rates from a wet scrubber versus a dry scrubber become all the more relevant for C02 control.

Another environmenta1 benefit of wet scrubbers versus dry scrubbers is greatly improved removal of hydrogen chloride (HCI) and hydrogen fluoride (HF) as compared to a dry scrubber. Actual measurements have demonstrated that very high HCI and HF control efficiencies, 99.7% to 99.9% for HCI and 99.X% to 99.9% for HF, are being achieved at wet scrubbed Such high levels of HCI and HF removal have not been shown for coal fired boilers controlled with dry scrubbers. Tests at the recently constructed Wygcn I?, an electrica1 generating unit burning subbituminous coal from the Powder River Basin and equipped with a spray dryer absorber, showed only 49% removaI of HF and 58% removal of HCL4 According to the Institute of Clean Air Companies, “wet scrubbers also provide significant removal of arsenic, beryllium, cadmium, chromium, lead, manganese, and mercury 6om flue gas.’”

Further, lower emissions of SO2 that are achievable with a wet scmbber also equate to lower PM2.S concentrations since there will be less SO2 in the air to contribute to sulfate formation. And studies have demonstrated that sulfate addition to sulfate- limited water bodies or wetlands can increase the transformation of mercury to its neurotoxic form, methylmercury?6 Thus, with lower SO2 emissions from White Bluff via the use of a wet scrubber as compared to a dry scrubber, the resuIt should be less sulfate deposition which should decrease methyhation of mercury.

These environmental benefits must also be considered in evaluating the environmental benefits of the SO2 control options for BART at White BIuff.

There are also some negative impacts of dry scrubbing as compared to wet scrubbing that were not mentioned in Entergy’s BART analysis, including a solid waste

See t0/14/08 Letter from Alstom to Duke Energy, ks. 29A and B. See 3-12-08 Wygen I1 Performnncc Test Rcport submitted to the Wyoming Department of 44

Environmcntnl Quality by Black Hills Corporation. Ex. 30.

JJ See http:/lwvw.icac.comli4a/pages/index.cfm?p~geid=340 I .

Experimental Wetland, Environ. Sci. Technol., 2006,40.3800-3806. Ex. 3 I: Knbbenhoft. David P. et ul.. Unravelling the Complexities Mercury Methylation in the Everglades: The Use of Mesocosms to Test the Effecb of ‘‘New” Mercury, Sulfate, Phosphate. and Dissolved Organic Carbon, available at littp:llsofia.us~s.~oovlprojccts/mcrc~c~rbonlhgmeso_geer03abs.html.

See. e.g., Jereminson. Jeff D. et al.. SuIfatc Addition Increases Methylmercury Production in an 46

Attachment 2 25

product that cannot be recyckd, poorer removaI of hazardous air pollutants as compared to a wet scrubber, and very high short-term emissions during atomizer changeout.

vi. A Visibility Analysis with the Achievable SO2 Emission Rates with the Wet Scriibber and D y Scrubber Options Mtist be Cormdieted as Port of the BARTAiralysis for White BluffBccausc ADEQ is Not Requiring the Top Level of SO2 Control ns BART.

Entergy appears to have conducted a visibility analysis for the two SO2 control options it considered, but it assumed the same SO2 emission rate for each option. ZOOS Revised BART Analysis at 4-1, TabIe 4-1 As we have demonstrated above, the SO2 emission rates that Entergy assumed for each SO2 control option are much higher than what is achievable with these controls, and a wet scrubber can achieve much Iower SO2 emission rates than a dry scrubber. The modeling reported in Entergy’s 2008 revised BART analysis showed that, even with the same SO2 emission rate modeIed for the two SO2 control options, visibility would improve to a greater degree with a wet scrubber as compared to a dry scrubber at the White Bluff units4’. Had Entergy modeled the lower SO2 emission rates that are achievable at White Bluff with wet scrubbers as compared to the SO2 Iimits achievabIe with wet scrubbers, the differences in visibility improvement would have become even more pronounced. If ADEQ is not going to require the most stringent SO2 controIs and SO2 BART emission limits at White Bluff, such visibility modeling must be done (with the best achievable SO2 emission rates for each control option modeled) and considered in the BART determination for White Bluff.

b. Entergy’s BART AnaIysis for the NOX emitted by White Bluff is FIawed and Incomplete.

Entergy failed to evaluate any post combustion NOx controls in its NOx BART analysis for White Bluff. Lnstead, it only evaluated combustion controls including boiler tuning, overfire air, and low NOx burners. 2008 Revised BART Analysis at 3- 1. It appears Entergy did not evaluate post-combustion controls such as selective catalytic reduction (SCR) because “[tlhe EPA guidance states that for units without post- combustion NOx controls, the use of combustion controls should be able to meet the presumptive Iimits.” 2008 Revised BART Analysis at 2-3. As previously stated, nothing in the BART guidelines allows Entergy to adopt presumptive BART limits and skip the five-factor BART analysis. ADEQ must require or conduct a proper five factor analysis of NOx BART for White Bluff including consideration of SCR along with newhpgraded combustion controls as a NOx 3ART option. Not only did Entergy not evaluate SCR, it did not evaluate the NOx emission rates that are achievabIe with combustion controls at White Bluff. Instead, it only evaluated the presumptive NOx BART Iimits as achievable with combustion controIs. As we will show below, lower NOx rates can be achieved at While Bluff with cornbustion controls.

See Reviscd BART Analysis nt 5-1 (Table 5-1). 43

Attachment 2

I. Ewer-gy 's Analysis of BART for. NOX Mmt Be Bnsed on the Coal it Ciirreirtly Burm at White Bhcfrimier its Cttrrettt Air Perittit.

As discussed above in our SO2 BART coments, Entergy's evaluation of BART for NOx should be based on the coal it currently burns at White Bluff. Entergy has made clear that subbituminous Powder River Basin coal is the coal it has historically burned at White Bluff and its 2009 PSD permit application indicates it is the coal that will typically be burncd at White Bluff. See August 2008 Revised BART Analysis at ES- 1, and January 2009 PSD Application at 2-1. The purpose of 3ART is to reduce emissions from current lev& to improve visibility, and that purpose is not served where BART Iimits are determined based on coals that White Bluff mypoientially uiilizc in the future. Therefore, the evaluation of NOx BART and the setting of a NOx BART limit should be based on the burning of Powder River Basin coal. That does not mean that ADEQ shouId not consider the fact that Entergy plans to burn different coals at White Bluff in the future, but that information should be taken into account in the evaluation of the suite of BART controls for NOx avaiIable at White Blnff.

ii. EvtergV Failed to Evaittate the Most Striirgem Level of Con froi Achievable with die Combtrstioii Coil frois Evdiimd BA R T for NOX.

In evaluating BART for the NOx emissions from White BIuff, Entergy assumed an achicvablc NOx emission rate of 0.145 lb/MMBtu for Unit I and 0.143 IblMMBtu for Unit 2 for the only NOx control option Entergy evaluated - boiler tuning, overfire air, and low NOx burners. 2008 Revised BART Analysis at 3-2. A study conducted by Babcock & Wilcox at tangentially-fired units burning subbituminous Powder River Basin coal showed NOx emission rates with uItra Iow NOx burners and overfire air that were generally less than 0.13 Ib/MMBtu:* The NOx BART evaluation must include the lowcst N O x rate that can be achieved with the BART option being evaluated. Entergy did not provide any details as to how it determined the 0.143-0.145 lb/MMBtu limits were the best NOx rates achievable with boiler tuning, overfire air, and low NOx burners at the White Bluff units.

iii. Eutergy 's Cosr Arialysis foi- the Coiitbiistioir Controls It Evaltiutcd for N0-v BART Is High Compared to Other N0.r BART Cos1 hiahsex

In Appendix A of its revised BART analysis, Entergy provided annuaI cost estimates of the only NOx BART option it evaluated - boiler tuning, overfire air and low NOx burners. Specifically, Appendix A shows the annualized cost of this set of NOx controls to be $5.2 million for White Bluff Unit I and $5.3 million for White Bluff Unit 2. 2008 Revised BART Analysis, Appendix A at A-3 (Table A-2). These cost estimates are very high in comparison to the cost estimates for similar controls at other

'' See Whitfield, T. et nl., Comparison of NOx Emission Reductions with PRB and Bituminous Coals in 900 MW Tmgcntinlly-Fired Boilers, presented to EPRI-DOE-EPA-AWMA Combined Power Plant Air PolluL.ant Control Mega Symposium, May 19-22.2003, Washington, D.C. at 8, Ex. 32.

27

Attachment 2

coal-fired electrical genemting units. For example, the BART analysis for the Boardman power plant in Oregon, a 6 I7 MW Powder River Basin coal burning unit, projected an annualized cost for overfire air and low NOX burners to be only $3.7 million?g The BART proposd for the Four Corners Power Plant units 4 and 5, which are both 790 MW units that burn western bituminous coal, estimated the cost of overfire air and low NOx burners to be $3.0 million for each of those units?’ The BART analysis for the Sherburne County Unit 1 Power Plant, a 690 MW unit that burns Powder River Basin coal, assumed an annualized cost of $2.2 million.” Thus, Entergy’s cost estimates for these controls are much higher than the costs that other coal-fired units are claiming, and it is not clear why because Sierra Club does not have access to the details that went into these cost estimates. As previously stated, such details are required to be part of a BART analysis and must be made available to the public. It is imperative that the cost assumptions of BART controIs being evaluated be justified, in order to have a fair comparison to BART controls that achieve higher lev& of NOx emission reduction.

In addition, in Entergy’s cost per ton analysis provided in Appendix A of its 2008 BART anaIysis, Entergy incorrectly inflated baseline emissions. Specifically, it appears that Entcrgy cahla ted base case NOx emissions for each unit based on the highest hourly emission rate and assuming 8,760 hours per year and 85% capacity, rather than based on the actual annual emissions of each unit that has occurred in the recent past. Table A-2 of Entergy’s 2008 Revised BART Analysis idcntifies the base case NOx emissions of White BIuff Unit 1 and Unit 2 to be 16,275.7 tpy and 17,612.9 tpy, respectively. Yet, Entergy identified the actual annual NOx emissions of each unit to range from 8,268.0 tpy to 10,854.9 tpy for Unit 1 and fTom 7,046.2 to 10,550.9 tpy for Unit 2 during 200 1 to 2005. See Table 2- 1 of Entergy’s August 2008 Revised BART Analysis (at 2- 1).

iv. Eiitts‘gy Failed to Evdiiole Post Conibirs!iorr Conlrols as An Option for N& BART.

As previously stated, Entergy failed to evaluate any post-combustion controls in addition to the proposed combustion controls to further reduce NOx emissions. Such post-combustion controls include selective non-catalytic reduction (SNCR) and selective catalytic reduction (SCR). This is a major omission in the White Bluff NOx BART analysis and ADEQ must not make a final NOx BART determination until a complete five factor BART analysis is made with consideration of these post-combustion NOx controls including consideration of these controls in combination with the proposed combustion controls of boiicr tuning, overfire air, and low NOx burners.

SCR in combination with cornbustion controls results in lowest NOx rates from coal burning power plants. The technoIogy is routinely required as BACT today dong

National Park Scwicc BART Evaluation Sprcadsheet entitIed “EGU Proposed NOx BART Control

I d

49

Spreadsheet,” Ex. 33.

5’ Id.

Attachment 2

with combustion controls for proposed new coal-fired power plants. Thus, it must be evaluated as one of the BART options for NOx control at White Bluff.

A review of recent SCR retrofits definitively shows that very high levels ofNOx removal are being achieved by recent SCR retrofit installations. NOx emission rates Iess than 0.05 I b M B t u are routinely achieved, and NOx removal efficiencies are typically around 90%.52 Permitting agencies have required lower NOx limits in recent BACT determinations, with many proposed and required BACT Iimits of 0.05-0.06 Ib/MMBtu and at Ieast one NOx BACT limit as low as 0.035 IblMMBtu?

SCR dong with combustion controIs have also been proposed or required as BART or to meet regional haze progress goals at sevcml coal-fired power pIants including the Boardman power plant in Oregons4, Boswell Energy Center Unit 3 and the Alan S. King Unit 1 facility in Minnesota55, and Naughton Unit #3 and Jim Bridger Units 3 and 4 in Wyoming". Thus, SCR along with instaIlation of combustion controls including low NOx burners has been found to be cost effective at these units, and according to data compiled by the National Park Service, the cost effectiveness of these controls ranged from $2,258 to $4,262/ton?'

Even higher levels of costs projected for SCR instalIation have been determined by permitting agencies to be cost-effective. One ofthe earliest SCRs on a coal-fired power plant in the United States was permitted in 1991 in New Jersey at the Logan Generating Plant?' The facility was buiIt with an SCR, started up in September 1994, and continues to operate with SCR.59 T i e SCR cost effectiveness value from I99 1 escalates to $2 1 ,O 1 6/ton in 2008 dolIars. In 1992, an SCR was permitted at the Stanton coal-fired power plant in Florida. The cost effectiveness was S4,262/ton for 70% NOx

See Erickson, Clayton A. et al., Sclcctivc Cntdytic Reduction System Performance and ReIinbility

TIE Desert Rock Energy Facility pennit requires the facility to achicvc. aner a NOx optimization period, Review, The 2006 MEGA Symposium. Paper # 121,Ex. 34.

a NOx emission m e of 0.035 IblMMBtu on a 365 day rolIing avcmgc 3nd an miss ion rntc of 0.05 Ib/MMBtu on n 30-day rolling average. Ex. 35. See also the Nationnl Park Servicc sprendshect on BACT Iimits for New PC Power Plants, Ex. 23.

Regional Hazc Rule at 152-153. Ex. 36.

Control Agency. at 71, awiIable at http:!!www.p~~.sta~e.inn.u~uir~re~ionull~~~c.l~tinl#sjp, Ex. 37.

'' Sue August 25.2009 Dmft Wyoming Stntc Implementation PIan Regional Haze at 92, 1G4. nvaifnble a1 h t l n : l ~ d ~ u . s i n r r . \ ~ v v . t i s ~ : ~ ~ i ~ ~ r ~ ~ i ~ n ~ ~ l h ~ ~ ~ . ~ ~ , Ex. 38.

'' National Park Service BART Evaluation Spreadsheet entitled "EGU Proposed NOx BART Control Spreadsheet." Ex. 33.

Scptembcr 30, 1999. http:l/www.allbusiness.comlenergy-utilitiesluti~itics-industry-clcctric- powcd6710358-1 .html,Ex. 39); D.W. Bullock, Long-term SCR Operating Experience at PG&E Generating's Coal-Fueled Plants, ICAC Forum 2000, March 2000, Ex. 40.

'' Ii?tp//cfpubI .epa.gov/rbIc/htmlblOZcfm, Keystonc Cogmcmtion Systems, Inc., Permit No. 0 1-89-3983.

June 19,2009 Oregon Regional Hnze Plan for Implementing Section 305 (40 CFR 51.305) of the

See Revised Draft Regional Haze State ImpIernentation Plan, July 2009, Minnesota Air PolIution

Award-winning Logan Gcnenting Plant Cctcbmtes Five Years of Opcrating Exccllcncc. Business Wire, 5s

29 Attachment 2

removal and a 2 year catalyst life."' This is $6,846/ton in 2008 dollars,6' An application is currently pending before the Wisconsin Public Services Commission to retrofit an SCR on the 430 MW Edgewater Unit 5, estimated to cost $15,40O/ton.c"" The state of Arizona determined that a cost effectiveness value of more than $4,4891ton of NOX reduced is cost eff~ctive."~ Thus, it is extremely IikeIy that the cost effectiveness of installing SCR systems at White Bluff Units 1 and 2 would also be within the range of costs considering by other state and local envirbnmental agencies as being reasonable.

If White Bluff were to install SCRs at each unit as BART, it could greatly reduce NOx emissions beyond the combustion controls it has proposed as 3ART. SpecificaIIy, based on actual annual emissions, the combustion controls proposed as BART wiIl only remove approximately 42-50% at each unit, but with SCR added to the combustion controls, 94% of the NOX emissions could be removed at each White Bluff unit.

Entergy aIso failed to evaIuate sekctive noncatalytic reduction (SNCR) along with combustion controls. This post-combustion controI technology is not as effective at removing NOx as an SCR, but it less costly. It has been required as BART for NOx along with combustion controIs for six lignite-fired power plant units in North Dakota at costs ranging from $1,268 to $3,778 per ton of NOx removed.w

Entergy's NOx 3ART analysis for White Bluff cannot be considered complete without an evahation of the top NOx control option - combustion controls plus SCR, as well its an evaluation of combustion controls plus SNCR. Based on testimony of Entergy's Anthony P. Waltz before the Arkansas Public Service Commission, Entergy apparently is planning to have to install SCR at both units at some paint in the near hture and has determined the costs of those pollution controls. See March 27,2009 Direct Testimony of Anthony P. Walz, on behaIf of Entergy Arkansas before the Arkansas Public Service Commission (Docket No. 09-024-U), at 9-10, Ex. 44. Therefore, ADEQ should require Entergy to provide that SCR cost data and should include the option of combustion controls plus SCR in its BART analysis for White Bluff.

Meniomndum from Greg Worley to Brian Beals, U.S. EPA, Re: Rcvicw of the Cypress Energy Projcct MI

PSD Application. September 25, 1392, EPA4PER054453 at 54454, Ex. 41.

'' 2008 doIIars = (%4,26~~on)(S75.4/355.2) = S6,846.21ton.

a Wisconsin Power & Light Co.. Certificate of Authority AppIicntion. Edgewater Generating Station Unit 5 NOx Reduction Project, November 2008, p. 12. Ex. 42.

61 Letter from Arizona Department of Environmental Quality to Steve Fry, EPA Region IX. Re: Consultation Regarding Best Available Retrofit Technology Analyses for the Four Corners Power Plant and Navajo Generating Station. May 12,2005. Ex. 43.

Spreadsheet," Ex. 33. Sw National Park Service BART Evaluation Spreadsheet entitIed "EGU Proposed NOx BART Control

30 Attachment 2

As stated above, if Entergy were to install SCR systems along with combustion controls at the White Bluff units, it couId reduce NOx emissions by 90% from base case emissions. Not only would that improve visibility in the region's Class I areas, it would also likely improve the region's ozone pollution, since NOx is a precursor to ozone. PuIaski County to the north of White Bluff has been recommended as nonattainment for ozone. Since vegetation damage has been found to occur even where ozone concentrations are below the ozone NAAQS, greater reductions in NOx emissions from White BIuff as wouId bc obtained via combustion controls plus SCR could also benefit the vegetation of the region's CIass I areas.

In addition, nitrogen deposition is aIso an issue in the region. The significant adverse impacts of nitrogen deposition, which is caused by wet and dry deposition of nitrates derived from NOx emissions, on ecological systems is well known. Critical loads &e., the amount of nitrogen deposition that an ecosystem can tolerate, above which the system become adversely impacted) are likely to SOOR be used to formally inform poIicy developments (e-g., NADP-CLAD Meeting, Pensacola, FL Spring 2009). And the National Park Service trends data shows an increasing trend in nitrate deposition in Arkansas". This presents another compelling reason for installation of the top NOx control measures in meeting 3ART at White Bluff. In addition, greater reduction of NOx emissions at White Bluff would aIso equate with reductions in PM2.5 formation.

While some utility companies have raised SO3 formation as an adverse environmental impact associated with use of an SCR system, there are options available to address that issue. First, low SO1 to SO3 conversion catalysts are available for the SCRs. For example, the CX Series type catalyst developed by Babcock- Hitachi has achieved SO? to so3 conversion rates its low its 0.1 %:'

In addition, the injection of alkaIi into the furnace can significantly reduce sulfate formation!' The addition of limestone to thc coal before pulverizing has shown a 50% reduction in sulfate forrnation.bs The injection of alkaIine materials such as hydrated lime, limestone, magnesium oxide and sodium carbonate after the furnace has also been shown to reduce sulfate formation by 40%-90%.69 In its recent BACT determination of

'' Scc Appendix B of the National Park Service's Air Quality Conditions and Trends Report of2008, availnbIe at http~/~.nature.nps.~oovlair/wfiolnpsPerfMeasures.cfm.

Anthony C. Favale et ai., Application and Operating Results of Low SO2 to SO3 Conversion Rate Catalyst for DeNOx AppIication at AEP Gavin Unit I , Proceeding of the 2006 Environmental ControIs Conference, U.S. Department of Energy. National Energy Technology Laboratory, Ex. 45. See also, Keiichuro Kai et al., SCK Catalyst with High Mercury Oxidation and Low SO2 to SO3 Conversion, Paper #S6, Ex. 46.

Mrrrrage. Assuc.. 54: 750-762 at 755-759. June 2004, Ex. 47.

hb

R.K.Srivnstavn et d., Emissions of Sulfur-Trioxidc from CoaCFired Power Plants, J. Air & Warre 67

48 Id

I d m

Attachment 2 31

the Desert Rock power plant, EPA Region 9 required hydrated lime injection to reduce SAM and sulfate^.^'

Ammonia injection before the elcctrostatic precipitator (“ESP”) has also been shown to reduce sulfate formation by high levels?’ And another option that could be used to reduce suIfate and SAM emissions would be the addition of a wet ESP. Wet ESPs have been shown to reduce sulfates by 90% and resuIt in very low opacity of

Furthermore, wet ESPs can capture other pollutants such as mercury.73

In summary, there are some significant environmental benefits that could be obtained if White BIuff was subject to the best control technology for NOx reductions, ix., combustion controls plus SCR, as coinparcd to installing only combustion controls to reduce NOx at White Bluff, due to the significant additional reductions in NOx emissions that could be obtained with combustion controls pIus SCR. Those benefits must be considered by ADEQ in determining BART for NOx at White Bluff.

vi. A Visibility Andysis with rlte Achievable NOx Entissioti Rates with Conibirsfioir Controls nitd ait SCR Must be Cotidircted os Part of the BARTAiialysis for. White BhflBecaiise ADEQ is Not Reqtiiririg the Top Level of NOx Control m BART.

Entergy appears to have conducted a visibiIity analysis for the NOx control option it considered, but only dong with the SO2 controI options it evaluated. 2OOX Revised BART Analysis at 4-1, Table 4-1. As we have demonstrated above, the NOx emission rates that Entergy assumed for the combustion controls evaluated are higher than what is achievabIe with these controls. Further, Entergy failcd to evaluate the top NOx controI option of combustion controls plus SCR. If Entergy instaIIed SCRs at White Bluff along with updatedlnew overfire air and Iow NOx burners, it could achieve much greater emission reductions than with just combustion controls alone. SpecificalIy, with SCR, White Bluffs NOx emissions could be significantly reduced beyond the NOx emissions that would be achieved with the combustion controls alone - by at least 2,666 tpy more per unit (or a total of 5,33 I ) to 3,832 tpy per unit (or a total of 7,664 t p~ ) . ’~ Clearly, such significant reductions in White Bluffs NOx cmissions achievable with combustion

’’ JuIy 31,2008 Prevention of Significant Deterioration Permit issued to Desert Rock Energy Compnny, Condition IX.B.2.b. Ex. 35.

R.K.Srivastava et aI., Emissions of SulfiwTrioxide from Coal-Fired Power Plnnts,A Air d IVrrsrc Mawage. /ISSDC., 54: 750-762 at 758-759, June 2004, Ex. 47.

Id at 760; see also, e.g.. K.S. Kumar et al.. Wet ESP for controlIing sulfuric acid plume foIlowing an SCR system. presented at the 1002 ICAC Forum, Ex. 48.

71

73 id ’‘ This was determincd using the annual avvcragc of 2003-2004 as base case cmissions and assuming a worst case NOx emission rate with combustion conids plus SCR of 0.07 IbfMMBtu and a Iawest NOx emission rate with combustion controIs plus SCR of 0.035 IbfMMBtu. In projecting future controlled emissions, it was assumed hat mch unit operated at 85% capacity at the proposed heat input capmity (8,950 MMBtu/hr).

Attachment 2 32

controIs plus SCR would hrther improve visibility in the Region's Class I areas, If ADEQ is not going to require the most stringent NOx controls and NOx BART emission Iimit at White Bluff, such visibility modeling must be done and considered in the BART determination for White Bluff.

c. Entergy Failed to Conduct Any BART Analysis for PMIO.

Entergy did not conduct any BART analysis for PM 10, claiming that use of electrostatic precipitators effectively rcmoves PM IO. 2008 Revised BART Analysis at 3- 5. However, an ESP is not the top control technology for PM10. Instead, use of a fabric filter baghouse is. AIthough Entergy is proposing to install a baghouse with the dry scrubber, Entergy has not conducted a BART analysis for PM10 or proposed a BART emission limit for PM 10. Thus, Entergy should have conducted a BART analysis to justify continued use of an ESP as BART for PMIO.

3. ADEQ Failed to Provide its AnaIysis of BART for White Bluff in Proposing Issuance of this Permit, and the Permitted Emission Limits Do Not Satisfy 3ART Requirements.

As we previously stated, ADEQ has failed to conduct its own analysis and determination of BART for White Bluff. This is required as part of the Regional Haze SIP and ADEQ previousIy indicated that it made such a determination in its issuance of the permit authorizing installation of the BART controls. However, ADEQ has not provided any such analysis with this draft permit.

a. ADEQ Must Find a Wet Scrubber to Reflect BART for SO2 at White Bluff.

Based on the information we have provided above, use of a wet scrubber should be considered BART for SO2 control at White BIuff for numerous reasons including:

Wet scrubbers are more efficient than dry scrubbers in removing SO2 and can achieve lower SO2 emission rates. When lower SO2 emission rates that are readily achievabk at White Bluff are considered, use of a wet scrubber is more cost effective than a dry scrubber on a dollar per ton basis - even using Entergy's assumptions for costs of each contro1 option. And the dollar-per-ton costs of a wet scrubber are in the range of what has been considered reasonable in other BART decisions. Wet scrubbers have other important environmental benefits in that lowered SO2 emissions will more readily enable White Bluff to effectively achieve C02 reductions. Further, lowered SO2 emissions wilI mean less downwind sulfate formation (and thus less downwind particulate formation) and potentially less methylization of mercury in wetlands and water bodies. Last, use of a wet scrubber will provide for greater visibility improvements than use of a dry scrubber in the affected Class I areas.

33 Attachment 2

For all of the above reasons, use of a wet scrubber clearly reflects BART for White Bluff. This is especially true because Entergy appears to be planning to burn different coals including lignite at White Bluff in the future. Installation of the best SO2 controls now, accompanied by a stringent SO2 BART emission limit, will provide Entergy with the fueI flexibility it needs for White Bluff and at the same time, ensure the most benefit from BART controls for Arkansas' wilderness areas.

b. ADEQ Must Find SCR Plus Combustion ControIs to Reflect BART for NOx at White BIuff.

Based on the information we have provided above, use of SCR pIus the combustion controls of boiIer tuning, overfin: air and low NOx burners should be considered BART for NOx control at White BIuff for numerous reasons including:

SCR plus the proposed combustion controls wiI1 achieve much lower NOx emission rates than just the combustion controIs by themselves. SCR will cut NOx emissions by more than half again more than what combustion controls done will result in. SCR plus combustion controls has been proposed or required as BART for several coal-fired power plant units and has been found to be cost effective at a wide range of costs. Thus, it is very likely that the costs of SCR plus combustion controls are in the range of what has been considered reasonable in other BART decisions. An SCR plus combustion controls will provide much greater reductions in NOx than just combustion controls which will not onIy improve visibility, but could also improve ozone andoc PM2.5 concentntions due to Iess NOx precursors and improve downwind formation of nitrates and nitrogen deposition due to less NOx emitted. Last, use of an SCR pIus combustion controls will provide for greater visibiIity improvements than just use of combustion controls alone in the affected Class I areas.

For all of the above reasons, an SCR plus the combustion controls of boiler tuning, overfire air, and low NOX burners reflects BART for NOx control at White Bluff. This is especiaIly true because Entergy appears to be planning to burn different coals including bituminous coal and possibly lignite at White Bluff in the future. The benefit of requiring an SCR system along with combustion controls to meet BART at White Bluffis that, even if White Bluff switches to c o d that resuIrs in higher NOx emission rates at the outlet of the boiler than would be obtained with subbituminous coal (h., considering combustion controls), the same low levels of NOx can be achieved because SCRs can reduce NOx emissions by 90% or more.

Attachment 2 34

C. The Draft Permit Fails to Set an SO2 or a NOx BART Limit Applicable to the White Bluff Units When They Burn or Blend with Coal Other Than Subbituminous Coal.

ADEQ did not specify an SO2 or a NOx BART limit in the draft permit for when White Bluff burns bituminous or other coal or when White Bluff burns a blend of coals. Instead, the permit just specifies an SO2 BART limit and a NOx BART limit of 0.15 lb/MMBtu, respectively, applicable when the units burn subbituminous coal. Draft Permit, Post -BART Scenario, Section IV. Conditions 3.a. 1 and 2). This is a major flaw in the draft White Bluff permit given that one of the purposes of this permit is to implement BART at White Bluff.

It must first be noted that the permit lacks a clear description of how compliance with a limit that only applies when the units burn subbituminous c o d will be determined. While there is a requirement in the permit that Entergy keep records of how much coal is burned to show compliance with the total tonnage limit on c o d burned in a year in Section VI, Condition 16 of the draft permit, there appears to be no recordkeeping or reporting requirement governing the information necessary to determine compliance with the 0.15 IblMMBtu SO2 and NQx limits that apply on a 30 day rolling average basis when the units burn only subbituminous coal. Thus, the proposed SO2 and NOx limits of 0. I5 lb/MMBtu that apply only when subbituminous coal is burned are unenforceable without proper recordkeeping and without a proper description in the permit detailing how compIiance is to be determined.

The draft opemting permit for White 31uff is required to include all applicable requirements, and so the permit is deficient because it faiIs to include any SO2 or NOx BART emission limits for when the White Bluff units burn or blend with bituminous coal. APEQ must correct this error and specify BART limits that appIy during a11 periods of operation of the White Bluff units, not just when the units burn subbituminous coal, Further, the permit must include adequate conditions describing how compliance with the BART emission limits are to be determined.

In the absence of a different NOx BART Iimit specified in the permit for when the White Bluff units burn or bIend with coal that is not subbiuniinous coal, it appears the NOx BART limit of APCEC Reg. 19 wouId apply. APCEC Reg. No. 19 allows for a NOx BART Iimit higher than 0.15 Ib/MMBtu when White Bluff burns bituminous coal or when it burns a blend of bituminous and subbituminous coal. See APCEC Reg. 19.1505(F)(2), (H), (1)(2), and (K). Spccifically, the Arkansas BART regulation allows the White Bluff units to emit NOx up to a rate of 0.28 lblMMBtu when burning bituminous coal, or a prorated emission limit when burning both of bituminous and subbituminous coals. There are numerous reasons why the 0.28 lb/MMBtu does not reflect BART and also why ADEQ cannot allow the White Bluff units to emit NOx at any higher rates than 0. I5 lbMh4Btu.

The purpose of BART is to reduce emissions from current levels to improve visibility. The White Bluff units have actuaIIy been emitting at a lower NOx rate than

35 Attachment 2

0.28 1blMMBtu. White Bluff 1 has had an annual average NOx emission rate of 0.25 IbMMBtu in recent ~ e a r s . 7 ~ White BIuff 2 has been emitting NOx in 2009 of 0.274 I b M ~ l B t u . ~ ~ Thus, if ADEQ allowed the 0.28 lb/MMBtu limit to be met, it would actually allow the White Bluff units to increase emissions.

Further, allowing the White Bluff units to emit NOx at any higher of a rate than 0. I5 IblMMBtu would be inconsistent with the modeling that was conducted for the Arkansas regional haze SIP to show the benefits of BART at White Bluff. Information in the “Entergy - ADEQ internaI emails.pdf‘ file on the ADEQ White Bluff permit website makes clear that this PSD permit allows White Bluff to emit at higher emissions than what was previousIy modekd for the BART analysis/regional haze SIP. SpecificalIy, a July 28,2009 e-mail from Thomas Rheaume to Siew Low and Anthony Davis states as folIows:

It appears that the NOx limits in our BART regulations for White Bluff, burning bituminous coal, were never part of any modeling used in the development of the rule. Entergy never submitted any numbers on that order. In fact, even the subbituminous rates in the regdation are higher than that used in the modeled.

Because only the 0.15 1blMMBtu NOx emission rate was modeled for the RegionaI Haze SIP, ADEQ cannot allow the White Bluff units to emit NOx at any rates higher than 0.15 IblMMBtu.

In addition, ADEQ also cannot allow Entcrgy to emit at the higher NOx rates allowed in the Arkansas BART regulation because it determined that the modifications being authorized in this permit action would trigger PSD review for NOx if White Bluff was allowed to emit at the higher NOx emission rates. See, e.g., May 26,2009 email from Siew Low to Thomas Rheaume in the file Entergy-ADEQ emaikpdf. ADEQ has proposed a 5,880 tons per year NOx limit in Section IV., Condition 20 of the draft permit which appears to have been imposed to limit potential to emit NOx of each unit, in an attempt to alIow the modifications at White Bluff being authorized in this permitting action to avoid PSD review for NOx. This NOx limit is based on a NOx emission rate of 0.15 Ib/MMBtu, a maximum permitted heat input capacity of 8950 MMBtu/hr, and continua1 operation throughout the year. The proposed imposition of this apparent limit on potential to emit reflects what is likely ADEQ’s determination that Entergy cannot be allowed to emit NOx at a rate any higher than 0.15 IbMvfBtu and still avoid PSD review for the modifications being authorized in this permit?’

’’ Baed on CEMS dxta obtained for White Bluff Unit 1 from EPA’s Clmn Air Markets Database for 2007 through the first 3 quarters of 2009.

’* Based on CEMS data obtained for White Bluff Unit 2 fiom EPA’s Clean Air Markets Database for the first 3 quarters of 2009.

’I’ As is discusscd furthcr beIow, we believe the projects being authorized in this permit must trigger PSD review for SO2 and NOx.

36 Attachment 2

For all of the above reasons, the draft permit is fatally flawed because it fails to set SO2 and NOx BART limits applicable when White BIuff burns or blends other coals beside subbituminous coaI. Further, ADEQ must proposed a NOx BART limit in the draft White Bluff BART permit that is consistent with what was modeled for the Regional Haze SIP - which means at the very minimum, the NOx BART h i t for the White Bluff units must be no higher than 0.15 Ib/MMBtu regardless of coal type. In addition, as shown above, ADEQ must also evduate post-combustion controls in determining the appropriate NOx BART limits for White Bluff. Last, the permit must include adequate conditions detailing how compliance with BART emission limitations wiIl be determined, especidly if ADEQ adopts emission Iimitations that vary depending on the coal burned at the White Bluff units.

111. ADEQ IMPROPERLY DETERMINED PSD APPLICABILITY FOR TEIE PROJECTS AT TIfE WHITE BLUFF BOILERS THAT ARE BEING AUTHORIZED IN THIS DRAFT PERMIT.

Along with addressing the installation of pollution controIs to meet best available retrofit technology (BART) requirements, Entergy has requested that the heat input capacity of the White H u f f boiIers be increased in this PSD/TitIe V permit action from 8,700 MMBtulhr to 8,950 MMBtuhr. In its January 2009 PSD permit application (submitted to ADEQ by Entergy via a February 4,2009 letter), Entergy indicated that to recover the Iost generating capacity fiom the parasitic load of the planncd dry scrubber and baghouse, Entergy would be undertaking “eficiency upgrades to the existing steam turbine generators ... which wiII result in an estimated net 2.9% increase in coal feed rate (MMBtulhr) to the Units after implementation of the proposed steam turbine efficiency improvements.” January 2009 White Bluff Permit Application at 2-7. Entergy also stated “[elxisting coal handling and storage facilities will continue to operate as at present, except that their actuaI and potentia1 actual throughput may increase by up to 3%.” Id. These and other reIated statements were induded in revisions to Entargy’s permit application that were submitted to ADEQ in July 2009 and August 2009.

Subsequently, in October 2009, Entergy submitted a revised application that removed all references to any increases in coal feed rate to the boilers as a result of the turbine efficiency project. Specifically, Entergy’s October X, 2009 letter to ADEQ states:

Per our discussion, Entergy is submitting revised pages necessary to more accurately describe the beneficial impact of the White Bluff Turbine Upgrade Project. . , .The current application refcrences “parasitic losses” in MWs that would be incurred as a result of this project and that would necessitate additional fuel use. Based on further discussions with staff and review of engineering studies, this loss will be fully recovered by the planned Turbine Upgrade Project without the utilization of additional fuel. These clarifications do not impact emissions cdculations contained in the application.

Attachment 2 37

The revised application pages include a strikeouthedline version in which Entergy appears to have removed most of the discussion regarding increase in fueI use, and Entergy also deleted information about how much of a amsitic load the pollution controIs would have on the units' generating capacity. 7 f

Interestingly, Entergy stated in its October 2009 revised PSD application that its revisions to the permit appIication (to no longer discuss the need to increase the amount of coal burned) do not impact the emissions calculations previously submitted. This does not make sense because Entergy claimed a 250 M M B t u h heat input increase to each boiler and calculated future potential emissions based on this higher heat input capacity for each boiler in its January 2009 permit application and in subsequent application revisions submitted in July and August 2009. See Appendix A of January 2009 PSD permit application which identifies the current heat input of each boiler as 8,700 MMBtu/hr and the future heat input of each boiler as 8,950 MMBtulhr. In its October 2009 PSI) permit application changes, Entergy addresses this discrepancy by stating "[t]he assumed maximum heat input will be increased from 8700 MMBtulhr to 8950 MMBtdhr, based on past historical data."79

Further, tve note that the draft permit allows for hourly and annual emission rate increases for PM 10 from the rail car rotary dumper from previously permitted limits of 0.1 Ibhr and 0.1 todyear to 16.0 Ibhr and 70.1 todyear, respectivdy. Draft Permit Conditions 43 and 15 1, as compared to Condition 37 of the currently effective White Bluff permit No. 0263-AOP-R6. Also, the draft permit allows for an increase in the annual amount of coal burned. Condition I G of the Draft Permit as compared to Condition 14 of the current Permit No. 0263-AOP-RG. If White Bluff had historically been operating each boiler at this higher heat input, then why do these current projects require such significant increases in the tom1 amount of cod burned or in the emissions of the rail car rotary dumper?

Regardless of whether this increase in heat input capacity is due to needed fuel to make up for the parasitic Ioad of the pollution controls being installed, is otherwise needed for the turbine efficiency project, or is truIy based only on "past historical data," this permit change to provide for an increase in allowable heat input capacity of the boilers constitutes a project" that must be reviewed to determine PSD applicability. And this project must be reviewed independent from the polIution controls project. There is no justification for considering an increase in permitted heat input capacity as part of the same project as the pollution controIs project as these are distinct physical changes or clianges in the method of operation at the White Bluff facility.

'' SPLJ file entitled "Mark-up of change pages for app1ication.pdf' at Ite:' :I ktscw.adcu .stntr.;ir.udAirPernlits/ Enter~~~"2OWIiitc~~n70U 111 fV.

See file entitIcd "Mark-up of change pagcs for appIication.pdf' at 3-1.

A "project" is defined as a "physical change. or change in the mcthod of operation of nn existing major

79

80

stationary source." 40 C.F.R @52.21(b1(52), Arkansas Regulation 19, Chnptcr 9, Rcg. 19.904(A).

3s Attachment 2

Neither Entergy nor ADEQ properly determined whether the project of increasing heat input capacity triggered PSD for all regdated NSR pollutants. As we will show below, w e believe this project will result in a major modification ofS02, NOx, CO2, PM 10 and PW.5, along with the polIutants for which ADEQ has already determined the White Bluff projects are major modifications @e., PM, CO, H2S04, VOCs, and lead). Thus, ADEQ’s proposed issuance of this drat? permit is improper without determining and requiring best available control technology (BACT) for a11 of these pollutants and without requiring modeling to ensure protection of the national ambient air quality standards (NAAQS), PSD increments, and visibility and other air quality related vaIues in affected Class I areas in the region.

A. ADEQ Must Obtain More Details from Entergy to Determine the Increases in Capacity andlor Emissions that WilI Be Allowed with the Turbine Efficiency Project.

As stated above, Entergy’s permit application indicated that it would be undertaking steam turbine efficiency upgrades at the White Bluff units. In its original PSD permit application and its July and August revisions to its permit application, Entergy further stated that it would be increasing capacity of the generating units by 6% and that about half of this increased generation wouId come from turbine efficiency upgrades. See, cg., January 2003 PSD permit application at 1-1. Entergy also has indicated that the other half of this 6% increase would come from an increase in heat input to the boilers. Id. As statcd above, Entergy Iatec revised its permit application to take out references to the percent increase in capacity and the increase in heat input, and instead indicated heat input capacity was being increased based on historical data. However, it is very likely that the increase in permitted heat input capacity and the turbine efficiency projects are related, as discussed below.

Based on the publicIy availabIe information regarding this permit, it does not appear that ADEQ has requested any additional information on the turbine upgrades. However, turbine cfficiency projects can result in an increase in annual emissions because the projects make the unit more efficient which ultimately results in the unit being dispatched more often. Further, if the units had more down time for maintenance andlor partial or forced outages before the turbine efficiency upgrade, the turbine efficiency project would allow for greater hours of operation andor operation at higher capacities post-project?’ On the EPA’s New Source Review Policy and Guidance database, there are numerous EPA letters regarding turbine efficiency projects and applicabdity to new source review?’ In fact, a review of EPA’s comment letters on turbine eficiency projects shows that EPA has typically requested significant detail on the projects to determine if the projects could result in increased emissions due to being more efficient (less costly to operate) and thus being dispatched more fiequently andor

See, e.g., Lesiuk, J.F.. Stcorn Turbine Upmtcs, GE Power Systems, Atlanta, GA, at 2, Ex. 49. dl

’* See EPA‘s Ncw Source Review Policy & Guidance Database at 11 t t p : l ~ w ~ ~ ~ ~ ~ . c p n . ~ ~ ~ ~ ~ ~ ~ I ~ n O 7 / p r ~ ~ ~ ~ ~ 1 i ~ ~ ’ ~ ~ d ~ ~ i r ~ p o l ~ ~ y ~ s ~ ~ h . l ~ t i n .

39

Attachment 2

due to increased operating hours because of increased a~ailability.8~ One thing is clear: EPA has not found turbine upgrade projects to be routine rnaintenance.&l

In addition, available documentation indicates there may be other rcasons that a turbine efficiency project would result in increased heat input to the boiler which wouId also result in an emissions increase. Testimony provided to the Maryland Public Service Commission by an expert for the owner of the coal-fired Brandon Shores power plant stated that heat input to the boilers would increase as a result of turbine efficiency projects. The Brandon Shores power plant consists of 2 coatfired units with a total generating capacity of 1,370 MW. This proceeding pertained to a request by the company, Constellation Power Source Generation, Inc., for a certificate of public convenience and necessity to retrofit pollution controls and conduct other enhancements. A turbine efficiency project was to be conducted concurrent with the installation of air pollution control equipment. According to the testimony of Dori J. Costa, who is empIoycd by and testified on behalf of Constellation Power, the turbine efficiency project, which included an upgrade to the high pressure steam turbine, would require more heat input to the boiler. See the folIowiog excerpt from Ms. Costa’s testimony:

Power block enhancements will include an upgrade of the high pressure steam turbine steam path components to increase turbine efficiency. The rcsults of this upgrade will improve heat rate and increase generator output at current steam flow.’ The increased turbine efficiency will result in a reduced high pressure steam turbine exhaust temperature. In order to compensate for the lower temperature, additional enhancements to the boilers will be needed, which include upgrades to the economizers, superheaters, upgrades to related process equipment, as we11 sis requiring an increase in fuel-derived heat input to the boilers.

October 23,2006 Testimony of Dori J. Costa, on behalf of Constellation Power Source Generation, Inc., Before the Public Service Commission of Maryland (Case No. 9075), at pages 6-7 (Ex. 25).

Entergy’s planned efficiency upgrades to the high pressure steam turbines at White Bluff thus cauId very well require additional heat input {ie-, more coal burned) to the boilers as well as boiler changes to increase the lower temperature high pressure steam turbine exhaust tempemure.

In the case of Brandon Shores, the company planned collective “power block enhancements” to not only gain back the parasitic load from the planned pollution controI equipment but to also provide for an additional 25 MW of additional power. Id. at 7. This would also typically require additional heat input to the boiler to generate higher steam flows,

See. e.g.. May 23,2000 letter from EPA to Ileniy NickeI regarding a turbinc upgrade at Detroit Edison’s

Id.

Monroe power plant, Ex. 50.

Attachment 2 40

Turbine vendor literature also indicate that turbine efficiency upgrades can accommodate such increases in steam flow which would generate more eIectricity but would also require additional he1 derived heat input to the boiler?’ Alstom states one of the benefits of steam turbine retrofits is a capacity increase:

The improved eEciency of a [turbine] retrofit produces additional capacity. It can be further optimized to match the increased steam flow from an uprated boiler. . . .

Alstom Power Brochure, “Steam Turbine Retrofit, Add Life, Add power,” at 3. Ex. 52.86

Entergy origindly stated it wouId recover about half of the 6% Ioss in generating capacity due to the parasitic Iosses of the pollution controIs, and that to “maintain net generating capacity” (which means to recover the fuIl6% loss in generating capacity), a 2.9% increase in fueI-derived heat input to the boilers would be needed. See January 2009 White Muff PSD Permit Application at 1 - 1. A 2.9% increase in the permitted heat input capacity of 8,700 MMBtu raises the heat input capacity of each unit to 8,950 MMBtulhr, which is the level that Entergy stated in a subsequent revision to its permit application it needed to raise permitted heat input capacity. October 8,2009 revised PSD permit appIication at 4- 1. Entergy’s October 2009 cover letter for its revised permit application states that the White Bluff turbine upgade project would hlly recover the parasitic losses from the pohtion controls without additional heat input. This may be true as our research shows that Entergy greatly overstated the parasitic losses &om the planned pollution controls as discussed in our SO2 BART comments below.87 However, since Entergy is still requesting an increase in the permitted heat input capacity of the boilers, its seems quite likely that Entergy is planning on hrther capacity increases with the turbine upgrade project and/or that Entergy is aware it wiIl need to increase heat input to increase the high pressure turbine exhaust steam temperature.

Entergy’s January 2009 permit application aIso stated that Entergy would be conducting “simuItaneous maintenance activities” with the poIIution control project including “tubing repairs, refractory repairs, maintenance, [and] repairs to ancillary systems (such as fans, pumps, piping, etc.) . . . .” Section 2.2.1.6. of January 2009 PSD permit application at 2-9. It seems quite possible that some of these activities may be designed to increase the steam flow of the boiIer (which would account for their projected

%%e, c.g. Drcier, Jr., D.W.. UpgndnbIe Opportunitics for Steam Turbincs, GE Power Systems. Schenectady, NY, GER 3693D. at 13. (Ex. 5 1). Downloaded from http:/lwww.gepower.cod prod-scwlproductdtcch-docdedsteam_tuines.him.

http~/~.powvcr.alstoni.comlliome/media_ccntrc/brochurc_libn~/index.EN.php?isSearched= I &si tcId=? ~lnngua~cId=EN&dir=%2~ome%2Fmedia~ccntre%2Fbrocfiure~Iib~ryD/o:!FWPRODUff= 1 1370GX4529 9 1 &PRESENTATIONSUBJECT=&theKeyword=&Search=SEARCH.

losses from the dry scrubber and baghouse might be at thc maximum 2.5% and is probably much less than that, as compared to the 6% increase in pamitic load claimed by Entergy.

Also available at

As is discussed in our commcnts on the SO2 BART anaIysis above, our research shows thc pamitic X I

41

Attachment 2

increase in heat input capacity of the units) to take fuII advantage of the turbine upgrade project?’ ADEQ should have requested hrther detail from Enrergy on these simultaneous boiler changes. It is aIso possible that Entergy has aIready begun making boiler changes that would allow for increased steam flow with the turbine upgrade project. For example, the economizer was replaced at each White Bluff unit in the past few years, as were the circulating water p~rnps.9~

Thus, for all of the above reasons, ADEQ shouId have obtained further information from Entergy on the turbine upgrade project and the requested increase in heat input capacity of the boilers?’ Based on our research, summarized above, we beIieve it is very Iikely that Entergy’s request to increase heat input capacity of the boilers is tied to the turbine efficiency project. Thus, these activities are a project that needs to be reviewed for applicability under PSD.

B. The Past Operatima1 History of the White Bluff Units is Not Relevant to the Fact that Entergy is Currently Requesting an Increase in the Permittcd Design Heat Input Capacity of the White Bluff Boilers Because a Permit Modification to Authorize this Change is Still Required.

As previously stated, Entergy’s most recent rcvision to its White BIuff PSD Permit application claims the increase in permitted heat input capacity of the boilers is based on “past historical data” rather than being associated with the turbine efficiency project. See October 2009 Revisions to the PSD Permit Application, Section 4.1 at 4- 1. Entergy’s Title V renewaI application makes the same claim and states that annual emissions will not change due to the Iantwide permit limit of 9.2 million tons of coal per year per Condition 14 of the permit. It should first be noted that the plantwide permit limit of 9.2 million tons of coal per year does not ensure there wilI be no increase in actual emissions because it is highly unlikely that the plant would burn 9.2 million tons of coal per year in

P P

KsSuc Mokansi, Jason, OPPD Unleashes SteaIth Low-Cost Capacity, Power, May/June 1998. This article expIains changes to thc boiler that had to be made to accommodate the increased generating capacity of a modernized stem furbine and the use of cxcess capacity of the boiler.

See. e.g., July 3 I , 2006 h e r to ADEQ from Entergy regarding the economizer replaccment at White Bluff Unit I (Ex. 531, tlie August S, 2007 lcttcr from Entergy to ADEQ regarding the cconomizcr replacement at White Bluff Unit 2 (Sierra CIub has requested this document from ADEQ specifically and through a Freedom of Information Act request IO ADEQ. but it has not yet been produced).

In addition to obtaining documentation from Entergy on the turbine upgrades. ADEQ should obtain Generating Availability Data Systcm (GADS) data to determine if these units were having reliabiIity issues and forced outages due lo the turbines. I t appears that Entergy has provided summary rcports of at least part of the GADS data to the Arkansas Public Service Commission, which nrc publicly avaiIabIe on the Public Service Commission wcbsitc. We have attached the Inst 6 years of Encrgy’s ”Arhnsas Reports” which we downloaded from the PSC as Exs. 54A-F.

” See October 20.1009 cover letter Entergy’s TitIe V renewal pennit application for Whitc Bluff, Ex. 55.

’’ Based on a typical heat value of Powder River Basin coal being 8,500 Btullh the 9.2 million ton of coal per year plantwide Iimit would allow each unit to operate a maximum pcrmittcd heat input capacity of 8700 MMBtulhr for every hour of the ycar.

89

Kl

Attachment 2 42

More importantly, if the White BIufTunits have been operating in excess of permitted heat input capacities of 8,700 MMBtu/hr, that means Entergy has been operating White Bluff in violation of its permit. The heat input capacity of the Whitc Bluff units bas been identified as part of the air permits for White Bluff since at least 1991 as 8700 MMBtulhr. See Section IV of current January 2009 White Bluff Title V permit, Ex. 56. There have been several permit actions and permit modifications in recent years, and not once did Entergy ask for an increase in permitted heat input ~apacity.9~ Further, it is clear that ADEQ has relied on the heat input capacity of the boilers in setting allowable emission limits. For example, the 6,090 lbhr NOx limit of Condition I of the current Title V permit is based on the NSPS NOx emission limit of 0.70 Ib/MMBtu (inchded in Condition 3.d. of the permit) multiplied by the 8,700 MMBtu/hr heat input capacity of the boiler. The 10,440 l b h SO2 limit is similarly based on the NSPS SO2 limit of 1.2 lb/MMBtu (included in Condition 3.c. of the permit) multiplied by the 8,700 MMBtulhr heat input capacity of the boiler.

The fact that Entergy has been operating the Wiite Bluff units at higher heat input than ailowed by the permit is an enforcement issue. ADEQ must investigate this past non-compliance and instigate a compliance schedule to bring the White BIuff units into compIiance with the Title V permit. ADEQ must also determine the physical or operationd modifications at the White Bluff faciIity that allowed for the increased heat input at each boiIer, and determine if Clean Air Act violations have occurred (including whether activities at the White BIuff units should have triggered PSD or NSPS applicability).

In addition, because Entergy has been operating the White Bluff units at higher heat inputs than permitted, that means the baseline actud emissions are inflated above allowable emission Iimitations. As provided in the definition of “baseline actual emissions,” in such cases, the baseline actuaI emissions must be adjusted downward to exclude the noncompliant emissions. 40 C.F.R. 0 52.2 1 (b)(48)( i)(b). This issue affects all of the determinations of PSD applicability conducted for this permit action.

Thus, ADEQ cannot discount the requested increase in permitted heat input capacity based on Entergy’s claims that they have historically been operating the White Bluff units in this manner. The fact is, an increase in the permitted hourly heat input capacity is a change in the method of operation andlor is due to a physical change that wilI increase actual emissions and which will necessitate a permit modification and review to determine applicability to PSD permitting requirements. Since no change in permitted heat input capacity has been authorized by ADEQ to date, this permit action must address the requested change in permitted heat input capacity as a change in the method of operation {if not also due to a physical change at each unit). And as stated

”ADEQ’s website shows that there were eleven (1 I ) operating permits or permit modifications issued for White Bluff since 1998, and all ofthese included the permitted heat input capacity of the boilers as 8700 MMBtufhr. See http://~~.adeq.state.ar.uslAIWbnnchgermitslpermi tting/p_facil_details~~p?AFIN=3500 1 10.

Attachment 2 43

above, it is quite likely that this increase in permitted heat input capacity is related to the turbine upgrade project.

C. Entergy and ADEQ Have Improperly Allowed Entergy to Take Into Account the SO2 and NOx BART Requircments In Determining Whether a Significant Emissions Increase of SO2 or NOx Would Occur as a Rcsult of the Proposed Projects at White Bluff.

Enrergy treated the pollution controls project and the increase in the permitted heat input capacity and turbine upgrades as one “project” in determining applicability to PSD. However, a “project” is a distinct activity under the PSD regulations. A “project” is defined as a “physical change in, or change in the method of operation of, a major stationary source.” 40 C.F.R. $53.3 1 (b)(52); Arkansas Regulation 19, Chapter 9, Reg. 19.904(A). Thus, the installation of the SO2 and NOx BART controls are one project and the increase in permitted heat input capacity is another separate and distinct project which, as stated above, we believe is related to the turbine upgrade project. Entergy considered these projects together in determining whether a significant emissions increase would occur and therefore took into account the SO2 and NOx BART emission reductions in projecting future potential emissions. However, in the first step of determining applicability to PSD (k, determining whether a project will resuIt in a significant emissions increase), the regulations onIy provide for the evduation of emission iiict-mses from the project. If a significant emissions increase is projected to occur as a resuIt of a project, then the net emissions increase from the project must be determined and, in that next step of PSD applicability, one can take into account contemporaneous and creditable emission decreases. The definition of “nct emissions increase’’ defines the emission decreases that are considered contemporaneous and creditable. Entergy’s approach of combining the projects and taking into account SO2 and NOx BART requirements in determining whether the project wiIl result in a significant emissions increase of SO2 and NOx csscntialIy circumvented the restrictions on creditable emission increases and decreases of the definition of “net emissions increase” and therefore was legally improper.

ADEQ did not clearly indicate how it determined PSD applicabiIity in either the draft permit or the statement of basis, so we assume ADEQ is dying an Entergy’s PSD applicability analysis.

I. LcgaI Background.

Under the PSD regdations, a “project” is a major modification if it causes two types of emissions increases, ( I ) a significant missions increase and (2) a significant net emissions increase. 40 C.F.R. 8 52.2 I (a)(2)(i~)(a).~

The PSD regulations define a “significant emissions increase” as an increase in emissions that is considered to be significant. 40 C.F.R. $52.2I(b)(40). For an emission

’‘ Arkansas has incorporated by reference the fcdcml PSD regulations of 40 C.F.R. 952.2 I at Arkansas Regulation 19. Chapter 9, Reg. 19.903(A).

44 Attachment 2

increase to be significant, it must exceed a particular “ton per year” threshold. Significant is defined as an emission increase that cquaIs or exceeds 40 tpy for SO2 and NOx, among other pollutants. 40 C.F.R. 5 52.2 1 (b)(23). A ‘‘significant net emissions increase” is simply a “net emissions increase” that is “significant.” Id. Again, “significant” for these polIutants is more than 40 tons per year. Id. A “net emissions increase” involves an arithmetic determination of whether a project will resuIt in an emissions increase by adding all the emissions increases that will result f?om a project and then adding and/or subtracting a11 contemporaneous, creditable emission increases and emission decreases. The definition of “net emissions increase” includes limitations on the emission reductions can be credited. 40 C.F.R. 9 52.2 1 (b)(3).

The regulations specify a procedure for determining whether a project will result in a “significant emissions increase” and a “significant net emissions increase.” 40 C.F.R. Q 52.2 1 (a)(2)(iv)(b). To determine whether a “significant emissions increase” from a project will occur, one must use a specific methodology depending upon the type of modification that will occur. Id. If the project involves only existing emissions units as is the c u e with this White Bluff permit, then one nccds to use an actual-to-projected- actual applicabiIity test. 40 C.F.R. § 52.21(a)(?)(iv)(c). The rules also provide for using an actuaI to potential test to determine whether a significant emissions increase will occur as a result of a projcct. 40 C.F.R. 9 52.21@)(41)(d) (definition of “projected actual emissions” allows for the use of potential to emit of the emission unit in lieu of the methodology for determining projected actual emissions). Ln the current permitting action for White BIuff, Entergy chose to determine applicability based on the actual to potential test. See Section 4.2 of January 2009 PSD Permit Application at 4-1. However, in doing so, Entergy took credit for BART reductions in determining potential to emit of SO2 and NOx at the White Bluff units. Such an approach is iIlegal as discussed below.

2. Thc Basdine Actual Emissions For White Bluff Units 1 and 2 Are Improperly Inflated Because Each Unit Has Been Operating Above the Permitted Heat Input Capacity.

Entergy’s October 2008 revised PSD permit appIication states that they are requesting an increase in the permitted design heat input capacity based on past historicaI data. October 2008 Revised PSD Permit AppIication at 4- 1. Subsequent to learning this, we examined the hourly heat input and cmissions data reported by Entergy to EPA’s Clean Air Markets Database (CAMD) and confirmed that the hourly heat input reported to CAMD was higher than the permitted 8,700 MMBtulhr, as hiFh as 1 1,827 MMBtu for White Bluff Unit 1 and 14,476 MMBtu for White Bluff Unit 2.

Because these units were operating at heat input capacities greater than that allowed in the White Bluff operating permit, their basdine emissions were improperly inflated, “Baseline actual emissions” is defined in the PSD regulations in pertinent part as

~ ~ ~ ~~~

’’ CAMD data downIoaded from EPA’s CAMD website at Iitlp:llcnmddatnnndmaps.epn.~ovl~dmlindcx.cfm?fuscact ion=cmissions.wizard.

Attachment 2 45

. . . the rate of emissions, in tons per year, of a regulated NSR poIlutant, as determined in accordance with paragmphs (b)(48)(i) through (iv) of this section.

(i) For any existing electric utility steam generating unit, baseline actuaI emissions means the average rate, in tons per year, at which the unit actuaIIy emitted the pollutant during any consecutive 24-month period selected by the owner or operator within the 5-year period immediatdy preceding when the owner or operator begins actual construction of the project. The Administrator shall allow the use of a different time period upon a determination that it is more representative of norma1 source operation.

... (b) The average rate shall be adjusted downward to exclude any non- compliant emissions that occurred while the source was operating above any emission limitation that was IegalIy enforceable during the consecutive 24-month period.

40 C.F.R. 0 52.2 1 (b)(48).

Because the White BIuff units were operating out of campIiance with their permitted heat input capacity requirements identified in Section IV of the current White Bluff operating permit (as well as in all previous White Bluff operating permits), the baseline emissions for these units must be adjusted downward. The heat input capacity identified in the permit is cIearIy an “emission Iirnitation” as that tern is defined under the CIeaa Air Act.

Specifically, the term “emission limitation” is defined in the federal SIP regulations as

a requirement estabIished by a State, local government, or the Administrator which limits the quantity, rate, or concentration of cniissions of air pdlutants on a continuous basis, iircitrdirig any rcgtiirenreitts which Iimit the Ievel of opacity, prescribe eqitipment, set fuel specifications, or prescribe operation or maintenance procedures for a source to assure continuous emission reduction.

40 C.F.R.$SL.l OO(z) (emphasis added)

Further, Section 302(k) of the Clean Air Act provides:

The terms “emission limitation” and “emission standard” mean a requirement established by the State or the Administrator which limits the quantity, rate, or concentration of emissions of air pollutants on a continuous basis, irrchidiug my reqitirernc~ir relnting to the operntion or

Attachment 2 46

ttmitttetiarice of a sozirce to assure continuous emission reduction, and miy dcsigt, equipment, work practice or operational standard promulgated under this chapter.

(Emphasis added; see also Sierra Club Y. Teiiiiessee Valley Airthorig, 430 F.3d 1337, 1348 (1 lLh Cir. 2005) (quoting 42 U.S.C. 9 7602(k)).

The permitted heat input capacity of the White Huff boilers falls into the category of “emission limitation” as defined in the regulations and statute. Thus, the baseline emissions need to be adjusted downward to reflect the permitted heat input capacity of 8700 MMBtulllr per each White Bluff unit. Entergy’s and ADEQ’s emission increase analysis relies on an improperIy inflated level of basdine actuaI emissions and thus their emissions increase calculations in Section 4 and Appendix A of the Mite Bluff PSD permit appIication are flawed.

3. The Baseline ActuaI Emissions For White Bluff Units 1 and 2 Are Improperly Inflated Because Each Unit Has Undertaken a Major Modification that Should Have Triggered ApplicabiIity of BACT for SO2 and NOx Among Other PolIutants.

As explained in Section IV of these comments, the economizer replacements which took pIace in 2006 and 2007 at Units 1 and 2 respecthlly likely triggered the application of BACT for S02, NOx, CO2 and potentially a host of other PSD regulated pollutants. Assuming PSD was triggered, Unit 1 would have been required to have continuously complied with BACT limitations for these poIlutants from 2006 and going fonvard. Unit 2 would have had to do the same from 2007 onward. As a consequence, all the emission caIculations and projections which provide the basis for the present permitting action are based on unlawful emissions. And a source cannot take credit from any actual or projected reductions in such unlawful emissions. Instead, all aspects of the evaluation of White Bluffs emissions including, without limitation, the establishment of 3ART and PSD limitations, shouId have presumed that Units 1’s emissions werc reduced to BACT levels for S02, NOx, C02 and all other PSD pollutants that were triggered in 2006. And ADEQ shouId have presumed Unit 2’s emissions were reduced to BACT levels for 502, NOx, C02 and all other PSD pohtants that were triggered for PSD from 2007 on. Because of this failure to properly account for the unlawful emissions from Unit 1 and 2 and the fact that basdine emissions have been inflated see srpra, ADEQ, at a minimum, must re-evaluate all the BART and PSD emission limits and determinations made in the context of the permitting action in order to avoid the issuance of legally invalid permit.

Attachment 2 47

4. Entergy and ADEQ ImproperIy Combined the SO2 and NOx BART Projects with the Beat Input Capacity IncreaselTurbine Upgrade Project in Determining Whether a Significant Emissions Increase of SO2 and NOx Would Occur.

As previously stated, Entergy improperly combined the projects that are part of the White Bluff permit application in determining whether a significant emissions increase of SO2 or NOx would occur, which was illegal because the PSD rules do not allow one to take credit for emission reductions in the first step of PSD applicability - that is, in determining whether a project wiIl result in a “significant emissions increase.”

PSD applicability for a pollutant to be emitted by a project requires both a “significant emissions increase and a “significant net emissions increase.” To determine whether the first type of increase will occur (Le,, a significant emission increase), one must first determine the emissions increases that will occur as a result of the project as provided in 40 C.F.R. 8 52.21(a)(2)(iv)(c). When EPA promulgated the revisions to the PSD regulations that specified the two step applicabiIity process, EPA stated “[w]e have revised the definition of major modification to clarify what has always been our policy- that determining whether a major modification has occurred is a two-step process.” 67 Fed.Reg. $0190 (December 3 1,2002). EPA’s policy on this issue, first issued in 1983 and then subsequently in 1988 and I989”, stated that a modification must first result in a significant emissions increase before one takes into account a11 contemporaneous emission increases and decreases in determining net emissions increase. EPA’s October 1990 New Source Review Workshop Manual also incorporates this policy in determining if a modification is major. SpecificaIIy, Table A-5 of the New Source Review Workshop Manual states as the first step in determining New Source Review applicabiIity:

Determine the emissions increase @ut not any decreases) from the proposed project. If increases are significant, proceed; If not, the source [sic] is not subject to review.

October 1990 New Source Review Workshop Manual at A.45 (emphasis added)?’

Contrary to this approach, , Entergy and ADEQ are attempting to take into account the SO2 and N O x BART decreases and emission limits concurrentIy with the emission increases that will result from the increase in permitted heat input capacity of the boilers. By doing so, Entergy and ADEQ unlawfully and improperly avoid following

. -

I)6sLT 19x3 memo with subject “Nct Emissions Incrcnses under PSD,” and September 18.1989 EPA Memo with Subject “Request for CIarificatbn of Policy Regarding “Nct Emissions Incrcase.” availnblc at http:ll~~.epa.govlregi on07/p rogramslxtdhi r/ pol ic ylsenrch. htm.

The next page of the New Sourcc Review Workshop Manual providcs an eaampie of how to determine applicabiIity and states, with respect to thc first step of determining applicability, that “only emissions increnscs expected to rcsult from the proposed project are examincd ... Emission decreases associated with a proposed project ... are contempanneous and may be considered along with other coritemporaneous cmissions changes at the source. However, tliey are not considered at this point in the analysis ....” New Source Review Workshop Manual tit A.4d (emphasis in original), Ex. 57.

93

4s Attachment 2

the rules of determining a net emissions increase. This circumvention of the PSD regulations is incorrect as a matter of law and cannot be allowed.

Instead, Eritergy and ADEQ must first, in step 1, determine whether the projects will result in a significant emission increase. Moreover, the projects need to be cvaluated separateIy, with the BART controls being one project and the increase in permitted heat input capacityhrbine upgrade being another project as there is no legitimate nexus between them

Thus, in determining applicability to PSD of the requested increase in permitted heat input capacity of the White Bluff boilers from 8700 MMBtulhr to 8950 MMBtulhr, Entergy and ADEQ should have first just evaluated the emission increases from this project. Such a review should have included the gathering of more information from Entergy to determine if emissions might increase from the turbine upgrade project, such as if the units were projected to be dispatched more often or otherwise will be operating more hours or at higher capacity more often.

The PSD rules allow for this determination to be based on an actual-to-projected future actual analysis or on an actual-to-potential to emit andysis. See 40 C.F.R. $52.2 1 (a)(2)(iv)(c); (b)(4l)(ii)(d). However, the potential to emit based on those requested BART emission limits cannot be taken into account in this first step of the process because, as stated above, one cannot take credit for emission reductions in the first step of determining PSD applicability. Emission reductions can only be considered in the second step of determining applicability - k, in determining net emissions increase at the source - if the reductions arc contemporaneous and otherwise creditable. See 40 C.F.R. 5 52,2l(b)(3).

Sierra CIub made a determination that the heat input capacity/turbine upgrade project would result in a significant emissions increase of both SO2 and NOx at each White Bluff unit. For the purpose of our analysis, we used an actual-to-projected actual emission comparison” Although we belicvc baseline actual emissions are improperly inflated for the reasons discussed in the above two sections (i-e., the economizer modifications and the fact that the White Bluff units were operating out of compliance with their pennitted heat input capacity), we relied on Entergy’s determination of baseline actual emissions from its August 5,2009 submittaI of revised baseline emissions and emissions caIculations (in Appendix A of the permit application), We also used Entergy’s data presented in revised Appendix A of its August 2008 submittal on current actual emission factor (IbMMBtu) and current capacity factor for each unit. This methodology underestimates the true projected actual emissions of each White Bluff unit, which is supposed to include among other things “the maximum annual rate, in tons per year, at which an existing emissions unit is projected to emit a regulated NSR pollutant” in any one of the 10 years foIlowing the date the unit resumes normal operations if the

’’ An aclual-to-projcctcd actual analysis was done because Entergy connot Lakc credit for thc SO2 and NOx BART cmission limits in this step of determining applicability and also because thc increme in pcrmitted heat input must be treated as a separate project from the BART controls.

49 Attachment 2

project involves increasing the emissions unit’s design capacity - which this permit change allows for. 40 C.F.R. §5221@)(41)(i).

In projecting actual emissions, such projections must be based on (among other things

. . .a11 rdevant information, including but not Iimited to, historical operational data, the company’s own representations, the company’s expected business activity and the company’s highest projections of business activity, the company’s fiIings with the State or Federal regulatory authorities, and compliance plans under the approved State Implementation Plan. . .

and [slhall exciude, in calcdating any increase in emissions that results from the particular project, that portion of the unit’s emissions following the project that an existing unit could have accommodated during the consecutive 24-month period used to establish the baseline actual emissions under paragraph (b)(48) of this section and that are also unrelated to the particular project, inchding any increased utilization due to product demand growth. . . .

40 C.F.R. $52.21@)(41).

In this case, the record is greatly lacking in information regarding the turbine efficiency project and in the other information that would be necessary to more appropriately project actual emissions for the White Bluff units after the increase heat input capacity and turbine upgrade project, so not all data needed to properly project future emissions is available. Thus, our analysis of projected actua1 emissions for the White Bluff units is likely to underestimate emissions as it does not address any increase in emissions from the units being dispatched more often and/or operating more hours each year as a resuIt of the turbine upgrade project. It also does not take into account projected increases in sulfur content of coal, when it is dear that Entergy is planning on burning higher sulfur coal and it appears ADEQ might be allowing use of such higher sulfur coal in this permit action. The table beIow shows that, even when projected actual emissions for the White Bluff units are based on the same current SO2 and NOx emission factors and capacity factors for each unit, the increased heat input capacity of each boiler being authorized in this draR permit will result in a significant emission increase of both SO2 and NOx.

Attachment 2 50

ActuaI Emissions (Bnscd on

tons (a) ‘07-’08),

Actuat Emission Factor,

IbliMMBtu {a)

Bascline 1 Baseline I Future I Future ActuaI Permittcd Projcctcd

Capacity Hourly Actual Factor (a) Heat Emissions,

MMBtuhr Input, tons

so2 16.282 I 0.G5 I 0.63 I 8.950 I 16,562 I 280

Projcctcd ActuaI

Emissions - Baseline Actual

Emissions, tons 1

the permitted heat input capacity of each unit will result in a significant emissions increase of SO2 and NOx based on an actual to projected actual emission comparison that likely underestimates projected actual emissions (and that also reflects an improperly inflated baseline as discussed in the above two sections). Clearly, this project will result in a significant emissions increase of both SO2 and NOx at each unit and at the plant, and thus a proper analysis of net emissions increase of SO2 and NOx should have been done but Entergy and ADEQ failed to do so.

5. Entergy and ADEQ Failed to Conduct a Proper Analysis of Whether a Significant Net Emissions Increase of SO2 or NOx Would Occur as A Result of Entcrgy’s Projects.

Because the increase in permitted heat input capacity wouId result in a significant emissions increase at each White Bluff boiler of SO2 and NOx, a determination of net emissions increase must be done to determine if the increase in heat input capacity/turbinc upgrade project is a major modification for SO2 and NOx. But neither Entergy or ADEQ performed such an analysis. While Entergy did provide a “White Bluff Emission Increase and Decrease Analysis Summary” in Appendix A and Section 4.0 of its PSD permit application, this analysis did not constitute a proper evaIuation of net emissions increase. Neither did the emission cahlntions provided in the tables entitled “Net emissions changes - Data and calculations” provided for each White BIuff unit in Appendix A of the permit application.

After the increase in actual emissions from a physical change or change in the method of operation is determined to be significant, the next step in determining net emissions increase is to evaluate all othcr contemporaneous emissions increases and decreases at the source that are contemporaneous with the change. The contemporaneous period is defined in the regulations as beginning on the date five years before construction commences on a change and cnding on the date the increase fiom the change occurs. 40 C.F.R. $ 52.2 I (b)(3)(ii).

Attachment 2 51

According to the White Huff permit application, construction is expected to commence on the projects right after the permit approval is issued and will be completed by 201 3. So, assuming the permit is issued in January 20 10 that means the contemporaneous period starts in January 2005 and ends in 20 13.

The definition of “net emissions increase” is as foIIows:

(3)(i) Net ennrissiotis iticreasc means the amount by which the sum of the following exceeds zero:

(a ) Any increase in actual emissions fiom a particular physical change or change in method of operation at a stationary source: and (b) Any other increases and decreases in actual emissions at the source

that are contemporaneous with the particular change and are otherwise creditable.

(ii) An increase or decrease in actual emissions is contemporaneous with the increase from the particular change only if it occurs between:

(a) The date five years before construction on the particular change commenccs; and (b) The date that the increase from the particular change occurs.

(iii) An increase or decrease in actual emissions is creditable onIy if the Administrator has not relied on it in issuing a permit for the source under this section, which permit is in effect when the increase in actual emissions fiom the particular change occurs. (iv) An increase or decrease in actual emissions of sulfur dioxide, particulate

matter, or nitrogen oxide, which occurs before the applicable minor source baseline date is creditable only if it is required to be considered in calculating the amount of maximum dlowable increases remaining available. With respect to particulate matter, only PM-10 emissions can be used to evaluate the net emissions increase for PM- 1 0. (v) An increase in actual emissions is creditable only to the extent that the ncw level of actual emissions exceeds the old level. (vi) A decrease in actuaI emissions is creditable only to the extent that:

(a) The old level of actual emissions or the old IeveI of allowable emissions, whichever is lower, exceeds the new level of actual emissions; (b) It is federally enforceable at and after the time that actual construction on the particular change begins; and (c) It has approximately the same qualitative significance for public health and welfare as that attributed to the increase Eom the particular change.

40 C.F.R. 8 52.2I(b)(3).

Because we have shown that the increase from the 250 MMBtu/hr increase in permitted heat input capacity of each unit is itself a significant emission increase, ADEQ should have conducted an evaluation of and decreases at the White Bluff facility to deternine whether a significant net emissions

other contemporaneous emission increases

52

Attachment 2

increase will occur and for which polhtants. Such a determination of net emissions increase must at least include the following:

1) The increase in actual emissions from the turbine upgrade project. As w e have previoudy discussed, the turbine upgrade project could increase emissions due to the units being dispatched more frequently andlor due to greater reliability both of which would result in increased operating hours. Further, the increase in permitted heat input capacity could very we11 be tied to the turbine upgrade project as either necessary to reheat the steam coming off the high pressure steam turbine and/or to provide for a capacity increase that the turbine efficiency project allows for.

2) The increase in actual emissions from the change in coal that appears to be allowed in this draft permit. This issue is discussed in more detail elsewhere in this comment letter, but it appears that ADEQ is removing Iimitations on sulfur and ash content of the coa1 that exist in Section IV., Condition 26, of the currently effective White Bluff permit. Such increases in sulfur and ash content that were previously prohibited by this permit condition constitutes a change in the method of operation at White Bluff that must be evahated for PSD applicability.

3) All other emission increases and decreases at the White Bluff facility that are contemporaneous and are otherwise creditable.

With respect to the SO2 and NOx emission reductions that will occur as a result of the pollution controk installation to meet BART, none of those emission reductions can be credited in the netting analysis. That is because emission reductions that are relied on for the state implementation plan (SIP) cannot be double counted and credited for netting. This is dear in the PSD regulations and past policy.

Specifically, the definition of ‘”net emissions increase” provides that a decrease in actual emissions is creditable onIy to the extent that “the old level of actuaI emissions or the old IeveI of allowable emissions, whichever is lower, exceeds the new level of actuaI emissions.” 40 C.F.R. $ 52.21 (b)(3)(vi)(a). “Allowable emissions” is, in turn, defined as follows:

the emission rate of a stationary source caIculated using the maximum rated capacity of the source . . , and the most stringent of the foIlowing: (i) the applicable standards as set forth in 40 CFR parts 60 and 61; (ii) the applicabIc State Implementation Plan emissions limitation, including those with a future compliance date: or (iii) The emissions rate specified as a federally enforceable permit condition, including those with a hture compliance date.

40 C.F.R. 6 52.2 1 (b)( I 6).99

In addition, EPA’s Emission Trading Policy Statement. which applies to netting in addition to other emissions trading, statcs that one requiremcnt for an emission reduction to be creditable is that it has to bc 99

53 Attachment 2

The BART emission Iimitations applicable to White Bluff were adopted by the state as part of the state’s regionaI haze SIP in 2008’00, which was submitted to EPA in July of 2008. APCEC Regulation No. 19 includes the BART emission limitations adopted by the state for each BART-eligible facility including White Bluff, and this regulation is part of the state’s regionaI haze SIP at Appendix 9.3C. SpecificaIly, APCEC Regulation No. 19, Chapter 15, Reg. 19.1505(F)-(K) set the BART limits for the White Bluff boilers, which includes a 0. I5 lb/MMBtu SO2 limit for each unit regardless of coal burned, a 0. I5 lbMMBtu NOx limit for each unit when burning subbituminous coal, a 0.28 lblMMBtu NOx limit for each unit when burning bituminous coal, and a prorated limit for each unit when the unit is burning a blend of subbituminous and bituminous coal.

Entergy has indicated the White Bluff units bum primarily subbituminous coal, so the subbituminous coal NOx is the limit that will prirnady apply. January 2009 PSD Permit Application at 2-1. Thus, White Bluffs allowable SO2 and NOx emissions based on the SIP limits applicable to Units 1 and 2 arc as given below.

White Bluff Units 1 and 2 Allowable Emissions Under the 2008 AR Regional Haze SIP: 502: 0.15 IblMMBtu x 8,700 MMBtulhr x 2 boilers x 8,760 hrs per year x 1 tod2000 Ib = 1 1,43 1.8 tons per year.

NOx: 0.15 lb/MM3tu x 8,700 MMBtulhr x 2 boilers x 8,760 hrs per year x Itod2000 lb = I I ,43 I -8 tons per year.

The “old level” of actuaI emissions from the two White Bluff boilers before this permit change (based on 2007-2008 emissions as used by Entergy in its August 2009 revised permit application submittal, Appendix A) are provided below:

Actual Emissions Based on 2007-2008 Averace White Bluff Unit 1 : SO2 19,437 tpy

NOx: 7,812 tpy

White Bluff Unit 2: 502: 16,282 tpy NOx: 7,269 tpy

Thus, the old allowable emissions based on the BART emission limits of APCEC Regulation No. 19 are lower than the old actual emissions at the White Bluff units. Consequently, only reductions in SO2 and NOx emissions that bring emissions lower

“surplus.” EPA’s policy states “At a minimum. onIy emission reductions not required by current regulations in the SIP, not already r c k d on for SIP planning purposes, and not used by the source to m e t any other rcgulatory rcquirement, can be considered surplus.” See II.A.1 of EPA’s Emission Trading Policy Stntcmcnt published at 51 Fed. Reg. 43814 (Decembcr4. 1986).

‘O0A copy of the complete Arkansas RegionaI Haze SIP is included as Exs. 2A and I3 to this Ietter. The SIP was adopted by ?hc state sometime between JuIy and September 2008.

54

Attachment 2

than 1 1.43 1.8 tpy of SO2 and 1 1,43 1.8 tpy of NOx, Le., that go beyond BART, are potentially creditable. lo'

Another requirement for emission reductions to be creditable is that the reductions have to be enforceable as a practical matter at and after the time construction commences on the project. 40 C.F.R. $52.2 l(b)(3)(vi)(b). There are no requirements in the draft permit that would require enforceable reductions lower than the old leveI of allowable emissions defined by the state's regional haze SIP (Le., lower than 1 1,43 1.8 tpy of SO2 and 1 I .43 I -8 tpy of NOx).

Thus, because there are no SO2 or NOx emission reductions that are creditable for netting, this permit action allowing an increase in permitted heat input capacity wilI resuIt in a significant net emissions increase of SO2 and NOx. Specifically, the emissions increase and the net emissions increase from the increase in permitted heat input capacitylturbine upgrade project wiIl be at least as high as shown in the table above, we11 above PSD significance Ievels. A major modification of these pollutants wiII result from the permitted heat input capacityhrbine upgrade project and, therefore, all PSD requirements including applicable of best available control technology (BACT) must be met for these poIlutants. The d n R permit fails to address these PSD requirements for SO2 and NOx and, consequently, is legally deficient as it constitutes a violation of the applicable PS D regulations.

'"ADEQ has also acknowledged that Entergy can not use the BART reductions for netting. See. rg.. August 20,2009 email from Thomas Rheaume to Mike Bates and others, and also September 1.2009 email fmm Thomas Rhcaume to Siew Low, in the file "Entcrgy-ADEQ inkma1 emails.pdf' downbaded from the ADEQ websitc.

55 Attach men t 2

IV. ADEQ FAILED TO EVALUATE PSD APPLICABILITY FOR PRIOR ECONOMIZER REPLACEMENTS AT WHITE BLUFF WITS 1 AND 2 WHICH APPEAR TO HAVE TRIGGERED PSD RF,VIEW AND FAILED TO IDENTIFY ALL APPLICABLE REQUIREMENTS AND AN APPROPRIATE COMPLIANCE SCHEDULE IN THE PROPOSED

COMPLIANCE ISSUES OPERATING PERMIT COVERING THOSE PROBABLE NON-

A. Title V Requires ADEQ to Have Included a Compliance Schedule in the Proposed Operating Permit for Prior PSD Violations at White Bluff

All sources subject to Tide V must have a permit to operate that “assurcs compliance by the source with all applicable requirements.” See 40 C.F.R. § 70.1 (b); Clean Air Act (CAA) 504(a), 42 U.S.C. 8 7661~. To meet this requirement, every TitIe V permit application must provide “a description of all applicable requirements” and must disclose any violations at the facility. Arkansas PolIution Control and Ecology Commission Regulation (hereinafter “APCEC Reg.”) 26.402(4)(a) and @>(a>, (b)(iii) and (c) (iii); 42 U.S.C. 0 766Ib(b); 40 C.F.R. $0 70.5(~)(4)(1), (5), (8).

Arkansas and federal law define “applicable requirements” to include:

“Any standard or other requirement provided for in the applicable implementation plan approved or promulgated by EPA through rulemaking under title I of thc Act that implements the relevant requirements of the Act, including any revisions to that plan promulgated in 40 CFR part 52.”

APCEC Reg. 26, Chapter 2 (definition of “‘applicable requirement”); 40 C.F.R § 70.2. This definition encompasses the requirement for new and modified major stationary sources to obtain PSD permits that fulIy comply with a11 applicable PSD requirements under the Act and the Arkansas SIP, inchding the requirements to apply best available control technology (BACT) and to perform air quaIity demonstrations. See generally CAA I IO(a)(2){C), 1GO-69, 173; 40 C.F.R. $5 52.21 el seq.; APCEC Reg.19.901 el scq. I02

For any appIicable requirements, inchding PSD requirements and other preconstmction requirements, for which the source is not in compliance at the time of

Vie PSD rulcs of the Arkaoss SIP are set forth at APCEC R e g u l a h 19.901 cf scq. and they adopt the federal PSD regulations at 40 C.F.R. 52.21 et sq. with limited exceptions. The most recent vcrsion of APCEC Rcgulation 19.801 ct seq. was approved by EPA as part of the Arkansas SIP on April 11.2007,72 Fed. Reg. 18394 (April 13,2007). and was effective as part ofthe SIP on May 14,2007. It adopted die version of the fedenl PSD regulations at 40 C.F.R. § 52.1 I et scq. in effect as of July 73,2004. fd: APCEC Regulation 19.904. Prior to that effective date, an carlicr version of APCEC Rcgulntion 19.901 ct scq, of the Arkansas SIP appIicd. which adopted the version of 40 C.F.R. 52.21 e/ scq. in effect as of June 3, 1994. APCEC Regulnlion 19.904. This version of ReyIation 19.901 cf scq. was approved by EPA as part of the Arkansas SIP on October 16,2000.65 Fed. Reg. 61 103 (October 16, ZOOO), and was effective on Novembcr 15,2000.

56

Attachment 2

permit issuance, the source’s application must provide a narrative description of how the source intends to come into compliance with the requirements. APCEC Reg. 26.402(S)(b)(iii); 42 U.S.C. $766 I b(b); 40 C.F.R. 0 70.5(~)(8)-(9). The application must fbrther propose a compliance schedule for any applicable requirements for which the source is not in compliance. APCEC Reg. 26.402(8) (c)(iii); 40 C.F.R. 4 70.5(~)(S)(iii).’~~ If any statements in the appIication were incorrect, or if the application omits relevant facts, the applicant has an ongoing duty to supplement and correct the application. APCEC Reg. 26.409; 40 C.F.R. 8 70.5(b).

As explained below, the record demonstrates that ADEQ failed to perform a PSD applicability determination regarding the recent replacement of economizers at White Bluff Units I and 2 in 2006 and 2007 respectively or to otherwise adequately evaluate PSD application prior to the commencement of those projects. These modifications appear Iikely to have triggered PSD for NOx, S02, C02 and other regulated PSD pollutants, including PM-10 and PM-2.5. If so, Entergy was required to obtain a PSD permit imposing BACT limits for emissions and to comply with all other PSD-related preconstruction requirements. Because ADEQ has failed to adequately evaluate the White Bluff Plant’s compliance with the PSD requirements of the CIean Air Act and the Arkansas SIP and it is probable that PSD violations are ongoing, the proposed Title V permit cannot be issued because a compliance schedule to address these ongoing PSD violations has not been included in the permit.

B. The Economizer Replacements at White Bluff Appear to Have Triggered PSD

1. The Economizer Projects at White Bluff Units 1 and 2

On July 3 1,2006, Entergy notified ADEQ that it intended to perform what it characterized as “a rnaintmnnccprujecf[ section of the Unit 1 boiler.” JuIy 3 1,2006 Letter from Entergy’s M. Bowles to ADEQ’s T. Rhcaurne at 1 , Ex. 53 (cmphasis added). According to Entergy’s response to a Sierra Club data request in a current proceeding before the Arkansas Public Service

invoIving replacement of the economizer

~~

‘03 The EPA Administrator has stated ns follows: 40 C.F.R. 4 70J(c)(X)(iii)(C) and 70.6(c)(3) requirc that. if a facility is in violation of an applicable requirement and it wiIl not be in compliance ai the time of pcrmit issuance, its permit must include a compliance schedulc that meets certain criteria. For sources that ace not in compliance with appIicnble requircmcnts at the time of permit issuance, compliance schcdules must include “0 schedule of remedial measures, including an enforceable sequence of actions with milestones, leading to compliance.” 40 C.F.R. 9 705( c)@)[iii)[C). III /lie Marrcr oJOg!r E~iviror~ina~md Scrviccs, Petition No. V-2005-1, Ordcr at pp. 6-7 (Adm’r Feb. 1,2006).

I M Except for asserting that this was a ‘maintenance project’’, Entergy failed to provide any information to ADEQ from which n four-factor routine maintenance. repair and replacement analysis could be performed. Based on the records from ADEQ that Sierra Club has so far obtained. it appears thot ADEQ failed IO rcquest any such information from Entergy and failed to perform any meaningful nndysis ofroutine maintenance issue. Entergy did not provide ADEQ any cost data on this project or any meaningful information on the nature or extent of this project, other than tbnt it involved an economizer replaccment. id Entergy failed to provide ADEQ with any information on purpose of the project or frcquency with which such projects had been performed nt White Bluff or anywhere else. id.

57 Attachment 2

Commission (APSC), the capital cost of the economizeriductwork replacements was $16.3 million.’05 In this letter, Entergy submitted calcuIations which it claimed demonstrated “this project [would] not resuh in a significant emission increase.” Id. While Entergy acknowledged that “the actual to projected actual test shows a significant increase in emissions,” it claims that “further analysis demonstrate the emission increase in not attributable to this project.” Id Instead, Entergy argued that “the project [would ] result[] in an actual decrease in emissions relative to future projected emission without undertaking the project.” Id. Entcrgy attributed its projected actual emissions to “normal economic activity including projected fuel prices and system electricity demand.” Id.; see also February 25,2008 Letter from Entergy’s M. Bawles to ADEQ’s T. Rheaume Regarding Unit 1 Economizer at I , Ex. 60 (correspondence after project was completed recognizing increased emissions above baseline which Entergy attributed to factors unrelated to the project” such as increased demand.). Entergy also committed to providing ADEQ with annual emission reports for five ( 5 ) years after the project was completed. Id.

Based on what records Sierra CIub has been able to obtain from ADEQ, it does not appear that AJXQ objected to any of Entergy’s contentions regarding this project and that nothing was done by ADEQ to determine whether the claims Entergy made were factually and technically supported or consistent with applicable law. Entergy commenced construction on the economizer replacement project for White Bluff Unit I on approximatdy September 15,2006.106 By approximately November 19,2006, that construction work was completed and Unit 1 had commenced operations.

After White BIuff Unit I was returned to service, Entergy submitted onIy one (1) of five (5) annual reports of emissions for Unit 1 that it had committed to provide to ADEQ. On February 25,2008, Entergy submitted its first annual report covering 2007. February 25,2008 Letter from Entergy’s M. Bowles to ADEQ’s T. Rheaume Regarding Unit 1 Economizer at I , Ex. 60. In the cover letter enclosing that rcport, Entergy stated:

Since the original projected actual emission baseline average emission evaluation for this project indicted that there was not a “reasonable possibility” of a significant increase in emissions, Entergy is proposing to suspend submittd of these reports.

Zd. at 1 (recognizing “8% increase in the unit’s capacity factor” and increased emissions above baseline which Entergy attributed to “factors unrelated to the project” such as “increased demand for power.”); see dso February 5,2008 Letter from Entergy’s M. Bowles to ADEQ’s T. Rheaume Regarding Annual Emission Reports for Four White BIuff Projects at 1, Ex. 6 I. There is no record of ADEQ either expressly agreeing with Entergy’s position or any objecting to the suspension of reporting. However, no further

~

’o’Sw Response of Entergy-Arkansas, Inc., to Sierra Club’s Fourth Sct of Data Requcsts, Responsc to Rccqucst 4-Le. (Dockct No, 09-024-U. White Bluff Dcclaratory Order), Ex. 59.

refcrcncc the version of 40 C.F.R. 5 52.21 el scq. effective as of June 3. 1994. Ai this point in timc. the older version of APCEC Reg. 19.901 etscq. ivas applicable. which adopted by IM

5s Attach men t 2

annual emission reports were submitted. Because at the time of commencement of construction, the appIicable Arkansas SIP ruIes required five (5 ) years of reporting in order to avail itself to the achlal-to-representative actual annual emissions, 40 C.F.R. Q 53.2 I (b)(2 l)(v) (1 994), Entergy’s failure to continue providing annual emission reports meant that this modification, when evaluated under the applicable Arkansas SIP rules, had to be assessed using the actual-to-potential emission test instead of the actual-to- projected future actual emission test. If this emissions test is applied to the White Bluff Unit 1 economizer project, the project clearly triggered PSD as it constitutes a “major modification” which results in a “significant net emissions increase” for S02, NOx, COZ,’07 see 40 C.F.R. 6 52.21(b)(2)(1), and potentially a host of other PSD regdated pollutants including PM- 1 0 and PM-2.5.

On August 8,2007, Enter notified ADEQ that it intended to replace the economizer at White Bluff Unit 2?’See December 7,2007 Letter from Entergy’s M. Bowles to ADEQ’s T. Rheaunie at 1, Ex. 62; see also November 15,2007 Letter from Entergy’s M, Bowles to ADEQ’s T. Rheaume at 1, Ex. 63; February 5,2008 Letter from Entergy’s M. Bowles to ADEQ’s T. Rheaume Regarding Annual Emission Reports for Four White Bluff Projects at 1, Ex. 6 I . Subsequent correspondence suggests that a notice submitted on this date by Entergy to ADEQ provided emission calculations similar to those provided from the economizer project at Unit 1 on JuIy 3 1,2006. Id. Despite sewing a FOIA request on ADEQ on October 19,2009 and making a diligent efforts to locate this document, Sierra Club has stiIl not obtained this critical information from ADEQ. See Composite E-Mails from W. Moore, LII, to ADEQ’s K. Robinson Regarding August S, 2007 Notice Letter, Ex. 64.”’ The available record indicates that Entergy again cIaimed this was a “maintenance project,”see. eg., December 7,2007 Letter from Entergy’s M. Bowles to ADEQ’s T. Rheaume at 1, Ex. 62, which did not trigger PSD but committed to providing ADEQ with annual emission reporls for five ( 5 ) years after the project was completed. February 5,2008 Letter from Entergy’s M. Bowles to ADEQ’s T. Rheaurne Regarding Annual Emission Reports for Four White Bluff Projects at I ,

Regardless of whether ADEQ agrees that C02 was “subject to regulation” under the Act and thus a PSD 107

regulated pollutant us of 2006, it is clearly “subject to regulation” under the Act because i t is subject Io ‘‘actual controI of emissions” at this juncture. See iifru at Section V (discussion of C02 LIS PSD regulated polIutant). Because any rcmcdy for the PSD violations associated with the 2006 and 2007 economizer replacements discusscd in this section wouId require the application of BACT as it is determined at the time the remedy is imposcd as opposed IO the time of the violations, Entergy would have to comply with BACT for CO? for the two economizer repIacemcnts.

Sicrra Club has recently learned thnt during 2007. and most likely during the same fall outage during which the economizer work was performed, Entergy replaced associnted duct work, bowl mi11 piping at a costs of $1 million. and an induced dmR fan blade which cost $1 million. See Response of Entergy- Arkmsas. Inc.. to Sierra Club’s Fourth Set of Data Rcqucsts, Response io Request 4-1 .e. (Dockct No. 09- 024-U, White Bluff Dmlantory Order). Ex. 59. ‘OPADEQ’s failure IO indude this and other documents in the permining file and failure to producc this documcnt in response to Sierra Club’s FOIA request before the comment period has expired has significantly impaired Sierra Club’s ability to evnluatc and comment on the economizer project at White Bluff Unit 2. Bccausc of ADEQ‘s failure to make critical information related to this matter publicly available, Sierra. Club will seek to supplcment these comments shouId it be able to obtnin additional relevant information related to this economizer project and other similar issues.

Io8

Attachment 2 59

Ex. 61 .'lo According to Entergy's response to a Sierra Club data request in a current proceeding before the Arkansas Public Service Commission (APSC), the capital cost of the economizedductwork replacements was $16.5 milIion."'

It does not appear that ADEQ asserted any objections to this project. On approximately September 14,2007, physical work on the Unit 2 economizer project began."' On November 15,2007, the economizer replacement project and the replacement of associated ductwork at White Bluff Unit 2 was completed at a cost of S 16.5 rn i l l i~n ,"~ and on Noveniber 17,2007, the unit was placed back in service. December 7,2007 Letter from Entergy's M. Bowles to ADEQ's T. Rheaume at 1, Ex. 62. Before the first annual emissions report for White Bluff 2 was due to ADEQ, Entcrgy submitted a letter to ADEQ claiming that because the August 8,2007 project notification for the White Bluff Unit 2 economizer project "demonstrate[d] there is not a "reasonable possibility" that a significant emission increase will occur as a result of this project, ... [Entergy] . . . determined that the provisions of 40 CFR 5 5221(r)(6) do no apply," inchding the requirement to provide five ( 5 ) years of annual emissions reports. Zd Based on this reasoning and Entergy's "emission evaluations" submitted to the ADEQ which are have not been made available to the public, Entergy sought to suspend the submission of its annual reports associated with the White Bluff Unit W S economizer project, among others. Although no response from ADEQ to this correspondence has been located, Entergy never submitted any annual emission reports of White BIuff Unit 2.

C. The Economixer Projects Appear Likely to Have Triggered PSD AppIicabiIity And As A Consequence the Title V Permit is InvaIid Due to a Failure to Identify A11 Applicabk Requirements and a Lack of a CompIiancc Schcdule Covering Thew Non-Compliance Issues

1. The PSD Program

The Clean Air Act was passed to protect and enhance the quality of the nation's air so as to promote the pubIic health and welfare and the productive capacity of the United States' population. 42 U.S.C. 8 740 1 (b)( 1). Congress intended to "speed up, expand, and intensify the war against air pollution in the Untied States with a view to assuring that the air we breathe throughout the Nation is wholesome once again." Wis.

' lo Again. there is no evidence in the permitting record or clsewhere suggesting tlint Entcrg provided any information to ADEQ from which ii routine maintennnce analysis could be performed or that ADEQ asked for such information or performed any such analysis or applicability determination.

' I ' Sue Responsc of Entergy-Arkansns, Inc., to Sierra Club's Fourth Set of Data Rcquests, Response to Request 4-1.e. (Docket No. 09-0244. White BIuff Dcdaratory Order), Ex. 59.

'I2 Bnscd on the date that II physical progtam of on-site construclion appmrs to have begun for the Unit 2 economizer rcplacemcnt. the version of the Arkansas SIP that wouId have been applicable to this project would have been thc most rccent version of APCEC Reg. 19.90 1 et .wq., which was effective as of May 14, 2007 and which adopted by reference the version of 40 C.F.R. S 52.2 I et scy. effective on and after July 23. 2004.

Request 4-1.e. (Docket No. 09-0244, White H u f f DecIaratory Order), Ex. 59. Rcsponse of Enkrgy-Arkansas. Inc.. to Sierra Club's Foourlh Set of Data Requesls. Response to 113

Attachment 2 60

Elec. Power Co. v. Rei& 893 F.2d 90 1,909 (7th Cir. 1990) (quotiiig H.R. Rep. No. 9 1- 1 146, at 1 (197O), as reprinted in I970 U.S.C.CA,N. 5356,5354)). As its name implies, the Prevention of Significant Deterioration program in Part C of the Clean Air Act, 42 U.S.C. $3 7470-7492, creates a program to prevent those areas currentIy attaining the minimum national air quality standards fiom deteriorating. The PSD provisions prohibit a major emitting facility from being constructed or modified unless, among other requirements, it: obtains a PSD permit, 42 U.S.C. 9 7475(a)( 1); by a permitting agency and through a pubIic hearing, 42 U.S.C. 6 7475(a)(2), has demonstrated that it will not cause or contribute to a violation of NAAQS or a “maximum allowable increase” over existing pollution levels (“increment”), 42 U.S.C. 8 7475(a)(3); and meets pollution limits based on “best available control technology” (BACT), 42 U.S.C. 6 7475(a)(4).

Although Congress intended the Clean Air Act to clean up old, polluting facilities, it recognized that it was not economically feasible to retrofit polIution controls on all existing sources. Therefore, Congress “grandfathered” existing facilities, effectively exempting them from compliance with new regulations until the facilities were modified. Alnbama Power v. Costle, 636 F.2d 323,400 (D.C. Cir. 1979); Utrited States Y. Miirphy Oil USA. hc., 155 PSupp3d 1 1 17, 1 137 (W.D. Wis. 200 1) (citing Wiscomsiir Ekctric Power Co. v. Reilk (WEPCO), 893 F.2d 901,909 (7th Cir. 1990)). This “grandfathering” was intended to be temporary - not “to constitute perpetuaI immunity” from all standards under the PSD program. Alclbrrnra Power; 636 F.2d at 400; WEPCO, 893 F.2d at 909 (“But Congress did not permanentIy exempt existing plants from these (PSD] requirements; section 741 I (a)&) provides that existing plants that have been modified are subject to the Clean Air Act programs at issue here.”); U.S. v. Ohio Edison Co., 276 F. Supp.2d 829,850 (S.D. Ohio 2003) (Congress did not intend that existing sources be granted perpetual immunity from installing modern pollution controls).

The White Bluff Unit I economizer project must be evaluated with reference to the prior version of the Arkansas SIP’S Reg. 19.90 1 et seq.; 40 C.F.R. 8 52.2 1 er seq. (1994), while the Unit 2 economizer project should be evaluated under the most recent SIP-approved version of Reg. 19.90 1 et seq.; 40 C.F.R 8 53.2 1 er seq. Although there are some significant differences in the two versions of the Arkansas SIP’S PSD rules, there are two fundamental components to determining applicabiIity of these modifications at the White Bluff units to PSD. To constitute a “major modification” which triggers PSD applicability: ( I ) there must “physical change or change in the method of operation”; and (2) there must be a significant emission increase. More specifically, under the prior version of the Arkansas SP’s Reg. 19.901, there must be a “significant net emissions increase,” APCEC Reg. 19.904; 40 C.F.R. Q 53.2 1 (b)(2)(1)( 1994), and under the most recent version of the Arkansas SIP, them must be both a “significant emissions increase” and a “significant net emission increase.” APCEC Reg. 19.904; 40 C.F.R. 5 52.2 I(a)(2)(iv)(a) (2004).

The term “physical change” is very broad. Congress intended that “any physical change” trigger the PSD program requirement, and intended “any physical change” to have an expansive meaning. New York v. EPA, 443 F.3d 880,855-87 (D.C. Cir. 200G)(holding that Congress’ use of the phrase “any physical change” was intended to

61

Attachment 2

apply to the broadest possibIe category of changes); New York v. EPA, 4 13 F.3dd 3,40- 42 (D.C. Cir. 2005). As stated recently in UitiredStafes v. Cirtergy Curp.:

The CAA defines the term ‘modification’ broadIy as ‘any physical change . . . which increases the amount of any air pollutant emitted . . . .* 42 U.S.C. 5 741 l(a)(4). As the Seventh Circuit has noted, the potential reach of this definition is broad and encompasses even the most trivial of activities.

495 F. Supp. 2d 892,901-02 (S.D. Ind. 2007) (citing WEPCO, 893 F.2d at 905; AIdmna Power, 636 F.2d at 400 (modification “is nowhere limited to physical changes exceeding a certain magnitude.”)).

An exemption does exist in the definition of “major modification” for “routine maintenance, repair and replacement.” APCEC Reg. 19.904; 40 C.F.R. $52.2 1 (2) (iii)(a). However, this exemption is exceedingly narrow. Urtited Sfatcs v. So. Itid. Gas & Elec, Co., 245 F. Supp. 2d 994, 1009 (S.D. Ind. 2003) (“Giving the routine maintenance exemption a broad reading couId postpone the application ofNSR to many faciIities, and would flout the Congressional intent evinced by the broad definition of medication.”).’ l4 To fa11 within this exception, the burdenii5 is on the source to demonstmte that the project in question satisfies a rigorous four-factor test which assesses the nature and extent, purpose, frequency and cost of the work. WEPCO, 893 F.2d at 9 IO (quoting September 9, 19SX Memorandum from Don R. Clay, USEPA, to David A. Kee, “Applicability of Prevention of Significant Deterioration (PSD) and New Source Performance Standards (PEPS) Requirements to the WEPCO Power Company Port Washington Life Extension Project.”)( I988 Clay Memo), Ex. 65; Unired Sfotes v. Cineray, 2006 WL 372726, “4 (S.D. Ind, Feb. 16,2006) (“The party claiming the benefit of an exemption to compliance with a statute bears the burden of proof as to the exemption.”) (citing Utiited States IS

First City Nat ’I Bonk of Hoirstwn, 386 US. 36 1,366 (1 967)); Ohio Edisoti, 276 F. Supp. 2d at 856; Sierra Club v. Morgm, No. 07-C-25 1-S 2007 US. Dist. LEXIS 82760 at *34 (W.D. Wis. 2007); Nat’l Parks Comemation Assh v. TVA, 6 I 8 F. Supp. 2d 8 I 5,824 (E.D. Tenn. 2009)c‘Defendant TVA bears the burden of proof as to the applicability of the RMRR exception in this case.”); United 3 d e s v. E. KJ~. Power Coup.. Inc., 498 F. Supp. 2d 976,995 (E.D. Ky. 2007). As stated in Ohio Edison, 276 F. Supp. 2d at 834:

Routine maintenance, repair, and replacement occurs regularly, involves no permanent improvements, is typically limited in expense, is usudly

“%PA’S 198s Clay Memo at 3 reinforces the narrow scope of the routine mnintenancc cxccption, stating: ”[tlhe clear intent of the PSD regulations is to construe the term “physical change” very broad@, to cover virfvally any signifcuid afterutioa to an existing plant. This wide reach is demonstrated by the very izarrow cxclusion provided in the regulations.” (emphasis added).

I” SimilarIy, through the permitting application process, Enlcrgy had the burden of proving that the routine maintenance exemption applied to both economizer projects and of providing rhe supporting bnsis for such an exemption. APCEC Rcg. 26.402 (B)(6): 40 C.F.R. 9 70.5(~)(6). It has failed to mcct its burden of proof and has not included any supporting documentation for such an exemption in the prcscnt permit application for either the Unit 1 or 2 economizer replacement projccts.

62 Attachment 2

performed in large plants by in-house employees, and is treated for accounting purposes as an expense. In contrast to routine maintenance stand capital improvements which generally involve more expense, are large in scope, often involve outside contractors, involve an increase of value to the unit, are usuaIly not undertaken with regular frequency, and are treated for accounting purposes as capital expenditures on the balance sheet.

Ohio Edison, 276 F. Supp. 2d at 834 (citations omitted).

The second part of a PSD applicability analysis involves an assessment of emissions increases under the applicable rules. As stated previously, under the prior version of APCEC Reg. 19.904, there must be a “significant net emissions increase,” APCEC Reg. 19.904; 40 C.F.R. 0 52.2 I (b)(2)(I)( 1994), and under the most recent version of the Arkansas SIP, there must be both a “significant emissions increase” and a “significant net emission increase.” APCEC Reg. 19.904; 40 C.F.R. 3 522I(a)(2)(iv)(a) (2004).

2. Application of PSD to the Replacement of the Economizer at White Bluff Unit 1

There is no question that the 2006 economizer replacement project at White Bluff Unit 1 constituted a “physical change.” And this work was clearly not exempt from PSD applicability as “routine maintenance” work. Other than calling that project “maintenance,” Entergy submitted no documentation to ADEQ, at the time the Unit 1 economizer replacement project was noticed or commenced or in the context of its permit application, that would support a claim that this project constituted routine maintcnance according to the applicable four factor test set forth in the CIay Memo. Based on a review of the pertinent case law, which has never resulted in a ruling that an economizer replacement project at a coal-fired power plant was routine, and the fact that Entergy failed to provide any documentation to ADEQ sufficient to support such a finding, it is beyond question that this project could not legitimately satisfy the routine maintenance exemption. See, e.g.. Morgan, 2007 US. Dist. LEXIS 82760 at “4 1-42 (replacement of economizers every 24 years “can hardly be considered ‘routine.’”); see gemrally WEPCO, 893 F.2d at 909-1 I , CinerBs 495 F. Supp. 2d at 933-948; Uttited Stales v. S. Idiatra Gas arid Ekc. Co. (SIGECO), 245 F. Supp. 2d 994, 1008 (S.D. Ind. 2003); Ohio Edism, 276 F. Supp. 2d at 834.

Nonetheless, the lack of any evidence that ADEQ evaluated the projected emissions increase calculations that were submitted for White Bluff Unit I project suggests that ADEQ may have made a Iegally erroneous routine maintenance decision regarding the White Bluff Unit 1 economizer replacement project which, once made, would have obviated the need for ADEQ to go further in the PSD analysis and evaluate emission increases. ADEQ has improperly dctemined a modification at another facility to be routine.

Attachment 2 63

Specifically, in response to an April 2,2001 request for an applicability determination from Green Bay Packaging, Inc., April 2,200 1 Letter from Green Bay Packaging, Lnc.’s M. Swindell to ADEQ’s T. Rheaume at 1-3, Ex. 66, ADEQ concluded that a boiler tube replacement project in question at that facility was routine maintenance due to the fact that the project was being performed “for safety concerns, is a maintenance issue, does not recover lost capacity, and does not have the potential to increase production in any downstream equipment . . . .” April 17,200 I Letter from ADEQ’s T. Rhertume to Green Bay Packaging, Inc.’s M. Swindell at 1, Ex. 67. In support of its position, ADEQ explicitly relied on a April 20,2000 letter from the Tennessee Department of Environment and Conservation (TDEC) to EPA Region 4 regarding Recovery BoiIer No. 1 at a Packaging Corporation of America facility, which ADEQ recognized had not been objected to by EPA. April 20,2000 Letter from TDEC to EPA Region 4, Ex. 68. If ADEQ reIied on the April 20,2000 TDEC Tetter andlor its previously expressed position on the contours of the routine maintenance exemption to determine that the White Bluff Unit I project was exempt from PSD, then ADEQ has made an error o f law in its PSD applicability determination for the White BIuff Unit 1 economizer replacement project for two reasons. First, the April 20,2000 letter from TDEC to EPA Region 4 was subsequently objected to by EPA Region 4. On September 14,2001, EPA Region 4 rejected TDEC’s position on routine maintenance set forth in its letter and informed TDEC that the replacement of tubes on Packaging Corporation of America’s Boiler No. 1 was not routine maintenance. September 14,200 1 Letter &om EPA Region 4’s G. Worley to TDEC’s B. Stephens at 2, Ex. 69. Second, ADEQ’s interpretation of routine maintenance set forth in its applicability determination concerning Green Bay Packaging, Inc. is inconsistent with EPA’s longstanding interpretation of routine maintenance and unlawhlly fails to folIow the four factor analysis set out in the 1988 Clay Memo and a series of other EPA applicability determinations and federal decisions interpreting and applying the routine maintenance exemption. Therefore, if ADEQ determined that the repIacemcnt of the economizer at White Bluff No. 1 was exempt from PSD as routine maintenance based on either the prior ADEQ position on routine maintenance and/or the April 20,2000 TDEC letter, it must re-evaluate and reverse that dctermination on PSD applicability.

It is highly unlikeIy that a project as expansive, expensive and as infrequently performed as the economizer replacement at White Bluff Unit 1 would be considered routine maintenance under the applicable four factor test. Moreover, when properly evaluated, it appears very likely that the 2006 White Bluff Unit 1 economizer replacement resdted in a “significant net emissions increase” under the applicable Arkansas SIP PSD rules, APCEC Reg. 19.904; 40 C.F.R. 8 52.2 1 @)(2)(I)( 1994). If this is correct, the White Bluff Unit 1 economizer project would have triggered PSD review for SO2, NOx, C02 and potentially a host of other PSD regulated pollutants, including PM-I 0 and PM-2.5 in 2006, and Entergy would now be under an obligation to obtain a PSD permit, comply with BACT limits and perform all necessary pre-construction requirements, including all requirement air quality demonstrations, for these pollutants. Furtliermore, the present operating permit would be invalid because it does not identify all applicable requirements and lacks a compliance schedule for these PSD violations.

Attachment 2 64

With regard to the emissions increase component of the PSD analysis for the White Bluff Unit 1 economizer project, there a number of errors in Entergy's submission to ADEQ on July 3 1,2006 which, when corrected and filly evaluated, are likely to show that this project resulted in a "significant net emissions increase" which triggered PSD. See July 3 1,2006 Letter from Entergy's M. BowIes to ADEQ's T. Rheaume at 2-5.

First, the two year baseline of actual emissions which Entergy relied on are improperly inflated for several reasons. The basehe period selected by Entergy covered 2003 and 2004. During the five ( 5 ) years before the project, White Bluff Unit 1 was being operated above its federalIy enforceable permit limit on maximum design heat input capacity of 8700 MMBtulhr.' significantly higher than they wouId have been if White Bluff Unit f had been operated in compliance with its heat input capacity limit. Entergy cannot Izlwfully rely on an inflated baseline emissions in a PSD applicability anaIysis. See October 1990 Draft New Source Review Workshop Manual at A.41-42, A.48. Instead, Entergy's baseline emissions for this project should have been adjusted downward to correspond to White BIuffs allowable heat input capacity permit limit of 8700 MMBtdhr before being compared to projected representative actuaI annual emissions.

Thus, emissions during the baseline period are

Second, Entergy's two year baseline period of actual SO2 and NOx emissions is also improperly inflated because the emissions used were based on CEMS data collected which included significant periods where Acid Rain Program 95% data capmrc requirements were not complied with.' l7 When a significant amount of data from an Acid Rain CEMS is missing, the methodology used for data substitution requires a source to conservatively estimate higher emissions for SO2 and NOx"? We reviewed the data submitted by Energy to EPA's Clean Air Markets Database (CAMD) and found that in 2004, there were 1,128 hours of substituted hourly SO2 emissions data for White Bluff 1. Since the actual emissions used in Entergy's baseline period to assess PSD applicability for this project included a significant portion of substituted data"', the baseline emissions overstate actual emission levels. Thus, the 2003-2004 baseline emissions data based on CEMs cannot be considered representative of uormal source operation for White BIuff Unit 1. 40 C.F.R. 8 52,21(b)(2I)(ii).

Third, in Entergy's July 3 I , 2006 letter to ADEQ, Entcrgy recognizes that its own PSD andysis for the Unit 1 economizer replacement was projected to result in a significant increase in emissions. See July 3 1,2006 Letter from Entergy's M. Bowles to

"'Section IV of the White Bluff operating permit spccifies the heat input capacity ofthc White Bluff boilers as 8700 million BTU per hour. A copy of the operating permit that applied prior to the baseline period used by Entergy (Permit # 263-AOP-R2) is attached as Ex. 70. This limitation is in the currently effective operating permit for White Bkff ns well. White Bluff Openting Permit at 16, Ex. 71.

"'See, eg., August 5,2005 lcttcr from ADEQ to Entergy re March 17.2005 compliance inspection and relatcd crnniIs. Ex. 72.

"'40 CFR 9 75.33 and 75.34; sec nko U.S. EPA, Clean Air Markets Division, Plain English Guide to the Part 75 Rule. Septembcr 2005.

"'See CAMD Data for White Bluff Unit 1 (2003 - 3rd Quarter 2009). Ex. 73.

Attachment 2 65

ADEQ’s T. Rheaume at 1, Ex. 53. Specifically, a comparison of the “Future Projected Actual with Project” emissions to the “Actual” emissions before the project in “White Bluff Unit 1 Economizer - NSR Actual to Future Projected Actual Calculation” enclosed with Entergy’s July 31,2006 submittal to ADEQ shows that Entergy predicted a significant emissions increase of S02, NOx, PM, and C02. For this determination, we first averaged Entergy’s hture projected actuaI emissions with the project for the two years after the project (2007-2008), consistent with the definition of “representative actual annual emissions” at 40 C.F.R. 8 52.2 1 (b)(33) ( I 994). See also 40 C.F.R. 9 52.2. I (b)(2 I)(v)( 1994). The difference between the 2007-2008 average of future projected actual emissions with the project and the baseline emissions is 1,288 tpy of S02,598 tpy of NOx, 47 tpy of PM, and 370,389 tpy of CO2, all in excess of the PSD significance Ievels for these pollutants. 40 C.F.R. § 52.2 1 (b)(23)(I) (1 994).

However, without adequately quantifying the impacts from the project which could have resulted in emission increases, Entergy claimed virtually all of this projected increase was a reflection of demand growth. The definition of “representative actual annuaI cmissions’’ includes Iimitations on such exclusions. SpecificalIy, a source can only claim such exclusion of emission increases if such increases in emissions couId have been accommodated during the representative baseline period and if the increase in emissions is unrelated to the particular change. 40 C.F.R. #52.12(b)(33)(ii) (1 994).

Our review of available information indicates that Entergy’s projected emission increases and increases in capacity utiIization of White Bluff Unit 1 ore related to the installation of the new economizer and therefore cannot be exduded in determining representative actual annual emissions. We base this conclusion on several factors. First, we have reviewed Entcrgy’s ‘‘Arkansas reports” to the Arkansas Public Service Commission (APSC) that are publicly available on the APSC website.”’ Based on our review of these reports, it is dear that White Huff Unit 1 was having issues with the economizer plugging and abo with the induced draft fan that are likely related to the economizer pluggage. We also reviewed the gross MW generation data for White Bluff Unit 1 from 2003 through third quarter of 2009, Ex. 73, and determined that the average of the highest 100 hours of gross megawatts generated increased significantly after the economizer was replaced - by 10-15 MW.

Entergy claimed a 2.1% increase in the White Bluff Unit 1 capacity factor as compared to the capacity factor during the basdine period, but then appears to have claimed that all but 0.1 % of that increase in Unit 1’s capacity factor was due to demand growth.’” However, based on our review of Entergy’s Arkansas Reports and, especially, the peak 100 hours of megawatts generated before and after the economizer replacement,

The cover page to thcsc reports is cntitled “FERC Form I Supplement, AnnuaI Report of the Entergy- Arkansas, Inc.” The 2003 to 2009 Entergy “Arkansas Reports“ are nttached ILS Exs.54A - 54F. ‘’I We took the avcragc of Entergy’s projected capacity factor for 2007 and 2008 with the projected, which equaied 78.1%, and subtracted the avenge capacity factor of the 2003-2004 baseline period of 76%. July 3 I , 2006 Letter from Entergy’s M. BowIcs to ADEQ’s T. Rheaume, Ex.53. It is clear from a comparison of the “projected future actunl with project” to the ‘‘projected future actual without project” tables that Entcrgy only assumed a 0.1% increase in capacity Factor was due io the economizcr replacemcnt.

Attachment 2 66

it appears that White Bluff Unit 1’s capacity factor was reduced in the years before the economizer replacement because pluggage in the economizer was acting as a bottleneck to operation at full generating capacity. Assuming this is We, that means Entergy could not exclude the cmissions increases due to operating White Bluff Unit I at increased capacity after the economizer replacement because the unit was not capable of accommodating that capacity before the economizer replacement. Furthermore, the increase in capacity factor would clearIy be reIated to the replacement of the economizer. We determined that the average increase in the top IO0 hours of electricity produced from the baseline period to 2007-2008 was 13.8 MW or about 1 % increase over the average top 100 hours during 2003-2004.

Another factor that was e n t i d y ignored in Entergy’s analysis was the impact that improvements in heat rate derived from the economizer project could have in moving White Bluff Unit I up Entergy’s dispatch order. Entergy indicated in its July 3 1,2006 submittal to ADEQ that there would be a 59 Btu/kWhr decrease in heat rate of Unit 1 after the economizer replacement. Ex. 53. A review of the average annual heat rates reported in the Arkansas Reports for Entergy shows that heat rate of Unit 1 decreased even more significantly - annual average heat rate of White Bluff unit 1 was reported to be 10,491 B m W h in 2003 and 1 I,9$1 Btu/kWh in 2004, while in 2007 the annual heat mte decreased to 10,383 BWkWh. Exs. 54A, 54B, 54E. That change was sufficient to provide Unit 1 with a better heat rate that Unit 2, which should have moved Unit I above Unit 2 on the dispatch order. This should have directly led to increased hours of dispatching of Unit I and, consequently, higher emissions of SO2 and NOx, among other pollutants.

In addition, Entergy’s emission calculations also fail to take account, much less to quantify, the improvement in availabiIity for Unit 1 which resulted from the replacement of the economizer. As discussed above, the Arkansas Reports seems to indicate that, on top of having tube leaks in Unit 1’s economizer, the ccanomizer was experiencing problems with pluggage and air flow which caused operational and other problems with ID fans associated with Unit 1. These problems and any others that the economizer was having pre-project should have been projected to be eliminated post-post project, resulting in a projection of increased availability for Unit 1.

Last, Entergy has used an inappropriate and unrepresentative baseline period for its emissions increase analysis. Pursuant to the version of APCEC Reg. 19.90 1 which was applicable to this economizer project, 40 C.F.R. 3 52.21 (b)(ii)(2004) governed the selection of a baseline period. That provision required that the baseline selected be “during a hvo-year period which precedes the particular date and which is representative of normal source operation.” ld Entergy’s baseline period failed to comply with that requirement. Instead of using 2004 and 2005, Entergy used 2003 and 2004. According to the applicable rule, if Entergy had obtained a determination from the Administrator that its baseline period was more representative, it could have properly used 2003 and 2004, id., but no such determination was made. Moreover, the use of the 2003 and 2004 baseIiac period could not have been approved because it was not more representative that the preceding two years period before the project. The 2003 and 2004 baseline was not

67

Attachment 2

representative because the emissions data availabk during those two years included pcriods where CEM data substitution procedures led to over-reporting of SO2 and NOx emissions. It was also likely not representative because that period of time failed to encompass the more significant unplanned outages and partial unplanned outages or derates associated with the Unit 1 economizer which increased during the 2004 and 2005 time frame and which would appear to have played a substantial roIe in the decision to replace the 2006 economizer.

The record demonstrates that ADEQ has failed to adequateIy evaIuate whether the White BIuff Unit 1 economizer replacement project triggered PSD in 2006. ADEQ has not fuIly evaluated each of the specific issues relating to PSD applicability discussed above, and it has not gathered and evaluated thc types of data that would be necessary to perform a credibk PSD applicabiIity determination for this project. Moreover, when a11 the flaws in Entergy’s PSD applicability anaIysis are considered cumuIatiuely, it appears more probabk than not that the 2006 Unit 1 economizer project triggered PSD applicnbiIity for 502, NOx, PM, and C02 as well as potentially other PSD regulated pollutants. In this context, the failure of ADEQ to perform a proper PSD applicabiIity dctermination results in the proposed operating permit being invalid since, among other things, there is no assurance that all applicable requirements have been identified in the permit and there is a substantial probability that a IegaIly required compIiance schedule to address ongoing PSD violations has been omitted from the proposed permit.

3. Application of PSD to the Replacement of the Economizer at White Bluff Unit 2

As expIained szpru, Entergy submitted a document to ADEQ on August 8,2007 which appears to have included critical information similar to Entergy’s July 3 1,2006 notification letter concerning the Unit 1 economizer replacement project. Without this critical infomation relating to the Unit 2 project completed in 2007, Sierra Club cannot specifically critique any of Entergy’s analysis d a t i n g to PSD applicabihty for this project.

However, what is cIear is that, like the Unit I economizer project, the Unit 2 project constituted a “physical change” to Unit 1 and this work was clearly not exempt from PSD applicability as “routine maintenance.” Beyond that, Sierra Club can onIy presume that the August 8,2007 notice contained a comparable PSD analysis to what was presented with regard to Unit I . Assuming that is the case, Sierra Club hereby adopts the same basic comments and critiques set forth above in 7IV.C.2 with regard to PSD applicability to the White BIuff Unit 2 economizcc project.

Based on a review of the Arkansas Reports to the ARPSC, it is apparent that White Huff Unit 2 was likely having the same issues as White Bluff I had with economizer pluggage and also with the induced draft fan that are likely related to the economizer phggage. Exs. 54A-F.

Attachment 2 63

Furthermore, as explained with regard to the Unit 1 economizer replacement project, the record demonstrates that ADEQ has failed to adequately evaluate whether the White Bluff Unit 2 economizer replacement project triggered PSD in 2007. ADEQ has not fully evaluated each of the specific issues relating to PSD applicability discussed above in 7IV.C.2, and it has not gathered and evaluated the types of data thnt wou?d be necessary to perform a credible PSD appIicability determination for the 2007 Unit 2. project. Based on a review of the available data, It is probable that the 2007 Unit 2 economizer project triggered PSD applicability for S02, NOx, PM, and C02, as well as potentially other PSD regulated poIlutants. In this context, the failure of ADEQ to perform a proper PSD appIicability determination for the Unit 2 economizer replacement project results in the proposed operating permit being invalid since, among other things, there is no assurance that all applicable requirements have been identified in the permit and there is a substantial probability that a legally required compliance schedule to address ongoing PSD violations has been omitted from the proposed permit.

4. Other Potential Modifications at the White BIuff PIant

There are a number of other projects which were performed at the White Bluff Plants that constitute physicaI changes or changes in the operation and which, when evaluated individualIy or collectively with other work, couId have constituted “major modifications” in the context of PSD or “modifications” in thc context of the NSPS standards. Specifically, Sierra CIub has recently become aware that the following work has been perfornied at the White Bluff Plant:

2008 $30 million expenditure on coal handling equipment;

2008 White Bluff Unit 1 $1.5 million expenditure to replace Reheater OutIet Header Terminal;

200s White Bluff Unit 1 $1.3 million expenditure to replace Low Temp Superheater Boiler;

2008 White Bluff Unit 2 0 1.5 expenditure to replace Reheater Outlet Header Terminal;

2007 White BIuff Unit 2 complete Generator Rewind;

0 2007 White Bluff Unit 2. $1 million expenditure on Bowl Mill Piping and a $1 milIion expenditure for an induced draft fan blade; and

2004 White BIuff Unit I Refractory Replacement requiring 100 hours of curing and downtime;

Response of Entcrgy-Arkansas, Inc., to Sierra Club’s Fourth Set of Data Requests, Response to Request 4-1 .e. (Docket No. 09-024-U, White BIuff Declaratory Order), Ex.

69

Attachment 2

59; see also April 6,2004 E-MaiI from Entergy’s G. Johnson to ADEQ’s D. Stowers at 1 (regarding refractory work on White Bluff l), Ex. 74. Because these projects, and potentially others that Sierra Chb has not yet been able to identify, have not been evaluated by ADEQ to determine PSD and NSPS applicability, the proposed Title V permit is potentially incomplete due to a failure to identify all applicable requirements and invalid for a lack of a compliance plan. ADEQ should have closely examined each of these projects to assess whether any of this work Gggered PSD or NSPS for any pollutant.

70

Attachment 2

V. SINCE C02 IS NOW SUBJECT TO "ACTUAL CONTROL OF EMISSIONS," THE PROPOSED PROJECT WILL RESULT IN AN INCREASE IN C02 EMISSIONS WHICH WILL TRIGGER PSD AND THE APPLICATION OF BACT TO C02

On June 30,2009, EPA authorized the state of California to implement its motor vehicle greenhouse gas emission standards pursuant to Section 209(b) of the Clean Air Act, 42 U.S.C. 6 7609(b). 74 Fed. Reg. 32744 (July 8,2009). As a result, carbon dioxide ("C02") was immediatdy subject to emission limits not onIy in California, but also in ten ( IO) of the fourteen (14) other states that have imposed these same standards pursuant to their independent authority under Section 177 of the Act, 42 U.S.C. 5 7507. Therefore, even under EPA's unduly narrow interprctation of the phrase ''subject to regulation" in CAA Section 165(a)(4) as meaning ''subject to actual control of emissions", carbon dioxide is now ''subject to regulation." Accordingly, C02 emissions from major emitting facilities are now unambiguously subject to "best available control technology'' YBACT'') emission limits.'22 See 42 U.S.C. $9 I65(a)(4) & 169(3) (requiring BACT for all pollutants "subject to regulation" under the Act).

According to EPA, a pollutant is @'subject to regulation" only if the pollutant is subject to "actual control of emissions." Memorandum from Stephen L. Johnson, EPA's Imt-prerufion of Regtrlnfions thnt Deteimirre Polliifanfs Covered By Federal Pt-eveirfion of Sigttifjcnrrf Deterioration (PSD) PertM Progrnm at 1 (December 18,2008); see 73 Fed. Reg. 80300, 80301 (Dcc. 3 1,2008). While Sierra CIub continues to believe that the extensive CIcan Air Act regulations requiring monitoring and reporting of C02 emissions satisfy the "subject to regulation" criterion, these new motor vehicle standards unequivocally require ''actual control" of CO2 emissions:

California's greenhouse gas missions standards establish allowable grams per mile ("gpm'l) levels for greenhouse gas emissions, including tailpipe emissions of carbon dioxide (CO2), nitrous oxide (N20), and methane (CH4), as well as emissions of C02 and hydrofluorocarbons (HFCs)) related to operation of the air conditioning system.

74 Fed. Reg. at 32752.

California's grams-per-mile standards (the 'TO2 Emission Limits") are effective for model years 2009 through 201 6:

[CaIifomia's] regulation covers large-volumc motor vehicle manufacturers beginning in the 2009 mode1 year, and intermediate and small manufacturers beginning in the 201 6 model year and controls greenhouse gas emissions from two categories of new motor vehicles -- passenger cars and the lightest trucks (PC

'= Sierra Club also maintains that even under EPA's interpretation. C02 first became "subject to aciual control: of emissions" on April 29,7008, when EPA approved a Delaware SIP revision that established limits on C02 emissions from new and existing distributcd gcnentors. Error! Main Documcnt Only. 73 Fed. Reg. 73,101 (April 29,2008).

71

Attachment 2

and LDT 1) and heavier light-duty trucks and medium-duty passenger vchicles (LDT2 and MDPV).

Id. at 32746. Because Model Year 20 10 began on January 2,2009 (and Model Year 2009 began on January 2,2008, see 40 CFR 85.2304), the “C02 Emission Limits“ are currently in effect and govern C02 emissions 6om a11 new motor vehicle saIes and registrations.

The C02 Emission Limits are in effect not only in California, but aIso in 10 other states that have also promulgated these standards for Model Years 2009 or 20 IO: Connecticut, Maine, Massachusetts, New Jersey, New York, Oregon, Pennsylvania, Rhode Island, Vermont, and Wa~hington.”~

Each of these states adopted the C02 Emission Limits pursuant to Section 177 of the Clean Air Act, 42 U.S.C. 8 7507. Section 177 expressly grants other states the authority to adopt California’s vchicle emission standards:

Section I77 of the Act contains an “opt-in” provision that allows any other state to “adopt and enforce for any mode1 year standards relating to control of emissions from new motor vehicles” if “such standards are identical to the CaIifornia standards for which a waiver has been granted for such model year” and are adopted “at least two years before commencement of such mode1 year.” 42 U.S.C. 8 7507.

Ainerican Autotitobile Mamlfachirers Association v. Cahill, 152 F.3d 196, 198 (2d Cir. 1998).

States have been exercising their Section 177 authority for aImost two decades; the first to do so was New York, adopting California’s original Low Emission Vehicle standards in 1 992. Motor Vcliicle Mairt[acttirers Associatioii v. New York State Deparment of Eiwironntetitd Cotisetvnfioii, I7 F.3d 52 1, 529 (2d. Cir. 1994). Not only have states adopted these emission standards under their Section 177 authority, but each state has historically incorporated these standards into their State Implementation Plans (“SIP”) under Section 1 10 of the Act, 42 U.S.C. 8 7410. See, cg., 40 C.F.R. 8 52.370(~)(79) (EPA approval of 8 177-adopted standards as part of Connecticut’s SIP); 40 C.F.R. 8 52.102O(c)(58) (Maine); 40 C.F.R. 9 52. I 120(c)( 132) (Massachusetts); 40 C.F.R. $52.1570(~)(84)(i)(A) (New Jersey); 40 C.F.R. 9 52.2063(c)( 141)(i)(C) (Pennsylvania). Once incorporated into a SIP, these requirements become CAA

Cal. Code Regs. tit. 13, § fBbI.l(a); Conn. Agencies Regs. g 22a-174-36b(b)(3); 06-086-127 Me. Code R. fi 1(B)[4); 310 Mass. Code Regs. 7.40(2)(n)(6); N.J. Admin. Code 6 7:27-29.13; N.Y. Comp. Codes R. & Regs tit. 6.8 215-5.2; Or. Admin. R 340-257-0050(2)(e); 25 Pa. Code 124.413; .~ccnlso 36 Pa. Bull. 7424; 12-031 R.I. Code R. $37.2.3; 12-031-001 Vt. Code R. § 5-1 lOG(aI(5); Wash. Admin. Code 17343- 090(2). In three more sta?es and the District of Columbia, these standards will come into effect in subscquent mode1 years. A r k Admin.Code 5 R1S-7-1501; Md. Code Regs. 26. I I .34.03; N.M. Code R. 5 20.2.38.101; D.C. Law 17-0151.

Attachment 2 72

standards, and numerous provisions authorize both EPA and citizens to enforce such SIP requirements, cg., 42 U.S.C. 8 7413; 42 U.S.C. §7604(a)(l), (f)(3).'24

Because the C02 Emission Limits are no different than any other vehide emission standards that states have been adopting and enforcing under the Clean Air Act for decades, it is clear that C02 is now "subject to regulation" under the Act. In fact, two federal courts have found that these very CO2 Emission Limits are indeed federa1 Clean Air Act standards. In Central Vdky Chtysler-Jeep. Im. Y. Goldstem, 529. F.Supp.2d I 15 1, 1 165 (ED. Cal. 2007), the court rejected the notion that even when approved under Section 209 of the Act, the C02 Emission Limits "are and remain state regulations and therefore subject to preemption'' by the federa1 Energy Policy and Conservation Act ("EPCA"): "The court can discern no Iegal basis for the proposition that an EPA- promulgated regdation or standard functions any differently than a California- promulgated and EPA-approved standard or regu1ation.l' Id. at 1 173.

Faced with the identical argument, the court in Green Moiirrlaiti Chtysler v. Crombie, 508 F.Supp.2d 295,350 (0. Vt. 2007) (emphasis added), also rejected the idea that the C02 Emission Limits were not federal standards, concluding "that the preemption doctrine does not appIy to the interplay between Section 209(b) of the CAA and EPCA, in essence CI claim of conflict betweeir fivo federal regulatory sc1tentR.r."

Because carbon dioxide became "subject to reguIatiod' under the CIeaa Air Act no later than June 30,2009, Section I65(a)(4) requires that BACT emission limits be required for C02 emissions from the White Bluff facility.

In the context of the present permitting action involving the White BIuff Plant, the 250 MMBtulhr increase in permitted heat input capacity wiIl result in a large increase in C02 emissions at each unit. SpecificaIly, assuming each unit burns subbituminous coal, the potential C02 emission increase from just the heat input capacity change is 228,855 tons per year from each unit and a total of 457,7 IO tons per year increase in C02 at the entire Wiite Bluff facility.'= As we have previously discussed, there could Iikely be additional increased emissions with to the turbine upgrade project due to the units being dispatched more often and/or due to increased reliability and those emissions also need to be taken into account. Assuming 85% capacity, the C02 increase is still massive, at 389,054 tons per year at the White Bluff facility. See 40 C.F.R. 752.2I(b)(23)(ii). ConsequentIy, PSD will be triggered for C02 and Entergy is required to comply with a11 applicable PSD requirements, including the obligation to comply with BACT for C02, the requirements of which are discussed fbrther below.

'" Because thc C02 Emission Limits also provide significant criteria pollutant bcnefits (74 FR 32758) CaIifornia has already included thcse emissions reductions into its 2007 ozone and PM SIP submittals io EPA. h t ~ : / l w \ ~ ~ . a r b . c a . ~ o v / n m n n i n ~ s i ~ / 2 0 0 7 ~ i ~ / 2007sin.htm. Other states will presumably now begin doing so as well. I*' This determination of potentisl CO2 emissions increase from cach boiler ivns based on Enargy's request to increase heat input capacity by 250 MMBhllhr and a CO2 emission factor for subcriticai boilers burning subbituminous coal of 209 Ib/MMBtu from EPA's Environmental Footprints and Costs of Coal- based Integmtcd Gasification Combined Cycle 3nd Pulverizcd Coal Technologics. EPA-4301R-061006. July 2006, at 3-2 1 (EX. 82).

73 Attachment 2

VI. THE WHITE BLUFF MODIFICATIONS TRIGGER PSD FOR PMlO AND PM2.5, BUT THE PERMIT FAILS TO REQUIRE BACT OR ADDRESS OTHER PSD REQUIREMENTS FOR THESE POLLUTANTS

According to the draft permit, the projects at White Bluff have been determined to result in large increases of PM and sulfuric acid mist (H2S04 or SAM). See the Draft White Bluff Permit at 9. Specifically, ADEQ has indicated that, based on a basdine of 2006-2007, the projects wilI resuIt in a 394 tpy increase in PM and a 295 tpy increase in SAM. Revicw of this increase of PM emissions shows that it consists of 234 tondyear of PM emissions from the boilers plus 160 tonslyear fugitive PM emissions associated with new sources including FGD waste handling and lime delivery and handIing as well as increased fugitive emissions from existing sources including the ash silos, rail car rotary dumper, storage piles, cooling towers, and coal barging activities.IZ6 According to Entergy’s August 2009 submittal of revisions to its permit appIication in which basdine actual emissions were determined based on 2007-2008 data, PM emissions will increase by 430 tpy and S A M emissions will increasc by 295 tpy. August 2009 Revised PSD Application at 3-2. According to Entergy, its calculations of PM are total filterable particulate matter. Id

However, the draft permit shows a decrease in PM 10 and PM2.5 emissions. Draft Permit at I I , 2 1. This is inexplicable. PM 10 is a major component of PM emissions - AP-42 indicates that 92% of the particulate matter exiting a baghouse is 10 microns or smaller and 53% is 2.5 microns or Further, S A M emissions are condensable PMlO and, in fact, are in the size fraction of less than 2.5 microns. Since this permit allows for large increases in both PM and SAM, it is not technically possible for the permit to not also allow for significant increases in PMI 0 and PM2.5.

The record for this permit lacks an anaIysis and adequate description of the basis for this inexplicable determination of PM, PM IO, and PM2.5 emissions changes being authorized in this draft permit for the White Bluff projects. ADEQ must provide further detail and information on these emissions calculations and Eiitergy must include documentation to support these emission calculations.

A. The White Bluff Projects Arc a Major Modification for PMIO.

Entergy’s calculations do not show an increase in PM 10. According to Entergy “The collateral decrease in total PM 10 is somewhat counter-intuitive, but is the result of hisher measured baseline emissions from stack testing compared with the future emissions guarantee.” Id It is indeed counter-intuitive. According to AP-42,92% of the

Thc calculations in workshect “Summary Actuals - PTE“ in Excel workbook “Appendix A “White Bluff Net Emission Changes( 1) 2-04-09.~1~” incorrectly add 160 tonslycar for fugitive PM emissions instead of I G3 tondycar as calculnted in the workshcct “Fugitive Summary.”

See AM?. Table 1 - 1 4

74 Attachment 2

particulate matter from a boiler burning subituminous or bituminous c o d controlled with a baghouse is 1 0 microns or smaIler diameter. AP-42, Table 1.1-6. So that means the 234 tpy increase in filterable PM projected by ADEQ would equal a 2 15 tpy increase in filterable PM 10.

Furthermore, ADEQ's draft permit shows an increase in S A M of 295 tpy as stated above. In the Method 202 used to measure condensable PM 10, this pollutant condenses and is measured as condensabk particulate matter, typically in the size fraction of Iess than 2.5 microns. In determining whether the projects at White Bluff will have a significant PM 10 emissions increase, condensable PMIO must be included.'2g

The increase in total PMIO, consisting of the increase in filterable (2 15 tpy) plus the increase in condensable (295 tpy + 159.2 tpy) greatly exceeds the significance threshold of I5 tpy, thus triggering BACT requirements for total PMI 0 pursuant to 40 C.F.R.fi52.2 I (j)(3). The draft permit contains no limit for total PM 10, which should be set at or below 0.01 8 lb/MMBtu, as discussed below.

B. The White BIuff Projects Are a Major Modification for PM2.5.

Tfiere are numerous reasons why the White Bluff projects are a major modification for PM2.5. First, as shown above, the White Bluff increase in permitted heat input capacityhbine project is a major modification for SO2 and NOx. These poIlutants are considered precursors to PM2.5, and a project that results io a 40 tpy increase (Le., a significant emissions increase and a significant net emissions increase) of either SO2 or NOx is considered significant for PM2.5. See 40 C.F.R. §522l(b)(23)(i).

Second, for similar reasons as discussed above, the increase in filterable PM2.5 from these projects is significant. According to AP-42,53% of the particulate matter from a boiler burning subbituminous or bituminous c o d controlIed with a baghouse is 2.5 microns or Iess. AP-42, Table I. 1-6. So that means the 234 tpy increase in filterable PM projected by ADEQ would equal 124 tpy of PM2.5. Since that is we11 over the PM2.5 significance level of 10 tpy, the projects at White Bluff are a major modification for PM2.5 just based on filterable PM2.5 alone.

Third, the projects will result in a lwze increase in SAM which, as discussed above, condenses to form particulate matter in the size range Iess than 2.5 microns. The projected emission increases of this pollutant is an order of magnitude greater than the 10 tpy PM2.5 significance threshold.

Therefore, the projects at White Bluff being authorized in this action are a major modification of PM2.5 and the boilers will have a net emissions increase of PM2.5, triggering BACT requirements pursuant to 40 C.F.R.952.2 1 Cj)(3). The proposed PM2.5 limit of 0.024 lblMMBtu does not satisfy BACT, as discussed below.

See 40 C.F.R 51.1OO(qq) and (IT); see aLo March 3 1, 1994 EPA memo to the Iowa Department of Natural Resources. Ex. 83.

Attachment 2 75

I . ADEQ Cannot Rely on the EPA’s PM2.5 Surrogate Poky.

In its draft permit, ADEQ states that it is not required to consider PM2.5 emissions in PSD permitting because it has not yet adopted the PSD provisions of the PM2.5 implementation rule into its SIP and because the PM2.5 implementation rule allows for use of the surrogate policy of impIementing PSI) for PM IO during the transition period until the SIP is revised129. Draft Permit at 2 1. The fact that Arkansas has not yet updated its SIP to adopt the provisions of EPA’s May 2008 PM2.5 implementation rule is irrelevant. PM2.5 is a regulated NSR poIlutant as that term is defined in 40 C.F.R. $52.2 f (b)(SO)(a) because a NAAQS has been promulgated for PM2.5. And the PSD regulations incorporated into the Arkansas regulations and SIP require the PSD permitting requirements to apply to all major modifications. Arkansas Regulation 19, Chapter 9, Section 19.904(a); 40 C.F.R. 52.2 1 (a)(2). Under 40 C.F.R. 5 52.2 1 (b)(2), a major modification is any physical change in or change in the method of operation of a major stationary source that would result in: a significant emissions increase and a significant net emissions increase of any regdated NSR pollitianf. Thus, under state and federa1 law, ADEQ is required to implement the PSD requirements for major modifications of PM2.5.

In 1997, EPA issued guidance providing that sources wouId be allowed to use implementation of a PMlo program as a surrogate for meeting PM2.s NSR requirements. John Seitz, “Interim Implementation for the New Source Review Requirements For PM[2.5],” (October 23,1997). The purpose of that guidance was to provide time for the development of necessary tools to cdculate the emissions of PMls and related precursors, adequate modeling techniques to project ambient impacts, and PM2.s monitoring sites. 70 Fed. Reg. 65984,66043 (Nov. I , 2005). EPA stated in the May 2008 PM2.5 NSR Implementation RuIe that “these difficulties have largely been resolved.” 73 Fed. Reg. at 28,321,28,340-3 (May 16,2008).

Arkansas’s PSD permitting regulations as approved into the SIP provide “[cjxcept where manifestly inconsistent with the Clean Air Act,” the ADEQ shaIl have the responsibility and authority granted to EPA under 40 C.F.R. $52.2 1 as in effect on July 23,2004. Arkansas ReguIation 19, Reg. 19.904(A). The PSD regulations in effect in 2004 did not authorize the use of the surrogate policy. For the reasons set out above, the PSI) regulations in effect in 2004 mandated that PSD reguIations apply to a11 regulated NSR pollutants, and PM2.5 is cIearIy a regulated NSR poIIutant. The EPA’s I997 surrogate policy contravenes the regulations and the Clean Air Act. Thus, under state law and under the Arkansas SIP, ADEQ has no authority to follow it.

Notwithstanding the above argument, EPA’s PM2.5 impIcrnentation rule only alIows a state to rely on the I997 surrogate policy “if a SIP-approved State is unable to implerncnt a PSD program for the PM2.5 NAAQS based on these final rules.” 73 Fed.Reg. 28341 (May 1 6,2008). Arkansas’s Regulation 19, Regulation 19.904(A) does

‘29 Presumably. ADEQ is referring to John Seitz, “Interim Implementation for the New Source Review Requirements for PM[2.5].” (October 23. 1997).

Attachment 2 76

not prevent ADEQ from implementing a PSD program for the PM2.5 NAAQS. In fact, Reg. 19.904(A) states:

In the absence of a specific imposition of responsibility or grant of authority, the Department shall be deemed to have that responsibility and authority necessary to attain the purposes of the Plan, these PSD Regulations, and the appIicable federa1 regulations, as incorporated herein by reference.

Arkansas has adopted the PM2.5 NAAQS as state regulation in Regulation 19, Chapter 3, Reg. 19.301. Arkansas Reg. 19.304 provides ADEQ with the authority to implement the PSD program for the PW.5 NAAQS, and mandates faciIities such as White Bluff to comply with the PM2.5 NAAQS under the PSD program. These provisions in Chapter 3 are part of the state implementation plan and has been approved by EPA. 72 Fed.Reg 18396 (April 12,2007). Thus, the ADEQ has the responsibility and authority under the Plan, the state’s PSD regulations, and the federal reguIations to implement the PM2.5 NAAQS in the PSD program. Consequently, under EPA’s 2008 PM2.5 NSR Implementation rule, ADEQ does not have authority or a legitimate justification for using the PM I O surrogacy policy.

The EPA’s recent objection to a Title V permit for the Louisville Gas & Electric’s TrimbIe power plant also substantially limits a permitting authority’s ability to rely on the surrogate policy in PSD permitting. The Trimble decision elaborates on court decisions that have fouiid it is only reasonable to rely on the surrogate policy when it has been shown to be reasonable to do so. I3O An adequate technical rationale has to be provided in the permit record to justify use of PMlO as a surrogate for PM2.5. Entergy attempted to provide such a rationale in its September 9,2009 submittal to ADEQ, but its rationale is far from adequate. Neither ADEQ’s draft permit or i ts Statement of Basis provide an adequate technical rationale either.

First it must be noted that this permit differs from the Trimble permit because Entergy is claiming the White Bluff units are not subject to BACT requirements for PM 10. In the Trirnble decision, EPA stated as one of the possible ways of making such a demonstration of an adequate technicaI rationale to use PMI 0 as a surrogate is to show that “the degree of control of PM2.5 by the control techndogy selected in the PMlO BACT anaIysis wiIl be at least as effective as the technology that would have been seIected if a BACT analysis specific to PM2.5 emissions had been conducted.”131 In this permit for White Bluff, no BACT analysis has been done for PM 10, so no such demonstmtion can be made.

130Sue EPA Order In the Matter Of: Louisville Gas and EIectric. TrimbIc County, Kentucky, TitIe WPSD Air Quality Pennit, 8V-02-043, Rcvisions 2 and 3, Petition No, IV-2003-03, nt 4346, availabk at h t t p ~ / ~ v w w . e p a . g o v / r ~ r e g i o n 0 7 / p r o g ~ i ~ a ~ ~ a i r / ~ i ~ l ~ / p c t i t i o n d ~ / p c t i ~ i o n ~ ~ ~ e ~ ~ ~ m b l ~ ~ 2 n d p c t i t i o ~ O O 6 . p d f .

”‘ Id. et 45.

Attachment 2 77

Second, both Entergy’s September 2009 submittal and the draft permit faiI to acknowledge that this permit will allow for the increase of SAM that condenses into PM2.5. As stated above, the permit allows a 295 tpy increase of SAM. Statement of Basis, Appendix A. Both Entergy and ADEQ completely ignore this significant increase in condensabIe PM2.5 in shting that condensable PM2.5 emissions will decrease with this permit a~tion.’~’

Thus, for all of the above reasons, APEQ cannot rely on EPA’s 1997 surrogate policy for PM2.5.

2. ADEQ Must Include AH Condensable PM2.5 Emissions In Determining AppHcabiIity of the Projects at White Bluff to PSD Permitting Requirements.

WhiIe both Entergy and ADEQ cIaim that the EPA’s PM2.5 implementation rule does not require SIP-approved states to regulate condensable PM2.5 until test methods are developed, that is not a justified basis for ADEQ to not include all ofthe condensable PM2.5 emissions including in determining whether the projects at White Bluff are a major modification for PM2.5. First, the PM2.5 implementation rule is not part of the Arkansas SIP as discussed above. Under the Arkansas SIP and PSD rules, ADEQ is required to impIemcnt the PSD program for PM2.5, and PM2.5 necessarily includes condensable PM2.5.

Second, the PM2.5 implementation rule makes clear that PM2.5 as we11 as PM and PM 10 “shall indude gaseous emissions from a source or activity which condense to form particulate matter at ambient temperatures.” 40 C.F.R. 52.2 f (b)(SO)(vi) (as promdgated at 73 Fed.Reg.28349 (May 16,2005)).

Third, the PM2.5 implementation rule docs not preclude states from including condensable PM2.5 in applicability determinations; it just does not require them to do so until January I , 20 1 1 or any earlier date in upcoming rulemaking codifying test methods. EPA has proposed revised test methods for condensabk PM 10 and PM2.5 on March 25, 2009. (74 Fed.Reg.12970 (March 25,2009)). Thus, it is reasonable to assume these test methods will be promulgated soon, and it is almost certain that the test methods will be promulgated before Entergy completes construction of the projects at White Bluff being permitted in this current draft permit. By the time White Bluff has begun operations of its modified facility and has to demonstrate compliance with PM2.5 emission limits, the revised EPA test methods for condensable PM2.5 will bc final.

Moreover, given that the only stated reason EPA had for deferring condensable PM2.5 from being accounted for in PSD applicability determinations has to do with (according to EPA) the variability in condensabIe PM2.5 test methods, it makes no sense that ADEQ did not consider condensable PM2.5 in determining applicability of the White Bluff projects to PSD because the determination of PSD applicability for PM2.5 does not

‘”See Entergy’s September 2009 submittal at 3; Drat? Permit at 22.

7s

AtEachment 2

require the use or reliance on any condensable PM2.5 test methods. Just based on the permit’s allowable increase of S A M which condenses into PM2.5, ADEQ can and must make a determination that the modifications being permitted here are major for PM2.5.

As EPA has stated, most direct PM2.5 emissions are condensables. 70 Fed.Reg. at G5,984,6605I (November 1,2005). Given that, there is no adequate justification for ADEQ ignoring the significant increase in SAM, which is known to condense into PM2.5 before exiting the stack of a coal-fired power plant Iike White Nuff, in finding that these projects at White Bluff are not a major modification of PM2.5. ADEQ’s actions in ignoring this significant increase in condensable PM2.5 emissions is inconsistent with state and federa1 law.

In summary, for all of the reasons discussed above, the projects at White Bluff being authorized in this permit are a major modification of both PM 10 and PM2.5. Accordingly, the permit is deficient because it fails to require BACT for PM 10 and PM2.5 and because it failed to require that all other PSD requirements be met for these pollutants.

VII. THE DRAFT PERMIT FAILS TO REQUIRE BEST AVAILABLE CONTROL TECJ3NOLOGY FOR THE MAJOR MODIFICATIONS AT WHITE BLUFF,

Entergy’s determination of PSD applicability concluded that PSD review was required for some poIlutants, but not others. Our comments on BACT are organized within this framework. First, for those pollutants that triggered PSD review, we comment on Entergy’s BACT analysis. Second, for those pollutants that did not trigger PSD review under Entergy’s analysis, we explain why BACT is required and discuss what it would be if a proper PSD applicability anaIysis were done.

A. BACT Was Not Required For Pollutants That Entergy Assumed Would Trigger PSD Review

The Application performed a PSD applicabiIity analysis and concluded that PSD review is triggered for five pollutants: sulfuric acid mist (“SAM”); filterable particulate matter (“PM’); lead (“Pb”); volatile organic compounds (“‘VOCs”); and carbon monoxide (“CO‘), requiring Best Available Control Technolorn (“BACT’). 119 Ap., Sec. 4.0 and Draft Permit at 9. The Application elected to follow the five-step top down BACT analysis set out in the New Source Review Manual (“NSR Manual”) to determine BACT. 1/09 Ap. at 5-1. Thus, it must folIow it. The five steps are:

Step 1 : Identify all potential controI technologies Step 2: Eliminate technically infeasible options Step 3: Rank remaining controls by effectiveness Step 4: Evaluate most effective controls and document results

79 Attachment 2

Step 5: Select BACT, the best control option that does not have unacceptable energy, environmental or economic impacts.

The BACT analyses in the Application and Dnf t Permit fail to follow this top- down approach and faiI to identify BACT. As explained below for each pollutant, the Draft Permit contains a general discussion of some of the applicable control options that appears to sewe as Steps I and 2. The Step 1 analyses are always incomplete, most notably consistently failing to include dean fuels. The BACT analyses then skip over critical Steps 3 and 4 and propose BACT limits without any support or explanation whatsoever. A review of other permits and stack test data indicates that none of the proffered limits satisfjl BACT.

As the Environmental Appeals Board recentIy explained in the Northern Michigan opinion: BACT is “[nlot merdy an option-gathering exercise with casuaIly considered choices, the NSR Manual or any BACT analysis calls for a searching review of industry practices and control options, a careful ranking of alternatives, and a final choice able to stand as first and best. If reviewing authorities let slip their rigorous look at ‘dl’ appropriate technologies, if the target ever eases from the ‘maximum degree of reduction’ available to something less or more convenient, the result may be somewhat protective, may be superior to some pollution control elsewhere, but will not be BACT.” In re Northern Michigan University Ripley Heating Plant, Slip. Qp. at 15-16 (EAB 2/18/09). As explained below, the Application and Draft Permit fail to determine the maximum degree of reduction for each subject polrutant, fail to consult other permitted limits and stack tests, and fail to review ail appropriate technologies.

1. BACT Is Not Required For Sulfuric Acid Mist Emissions

The Draft Permit condudcd that BACT for sulhric acid mist is an emission limit of 0.0042 IbMMBtu, achieved using a combination of low sulfur coaIs and dry FGD. Draft Permit at 15. This determination does not satisfy BACT for three reasons. First, the BACT andysis itself is flawed. Second, the mission rate is much higher than the many lower limits that have been permitted and achieved and thus does not represent the maximum degree of reduction that is achievable. Third, the proposed pollution control train does not represent BACT-level controls.

a. The SAM EACT AnaIysis Is Flawed

The Draft Permit contains a discussion of various control technologies which appears to be an attempt at Steps 1 and 2. At the end of this discussion, the Draft Permit concludes that BACT for SAM is 0.0042 IblMMBtu, without going through Steps 3 and 4 of a BACT analysis. Instead, the Draft Permit’s only support is stated: “based on an initial Appendix A stack test.” Draft Permit at 15. The Draft Permit fails to provide any information on this “initial Appendix A stack test” or any other basis for the proposed SAM BACT limit. In the PSC proceeding, Entergy’s response to S i e m Club Data Request 5-6 indicates that this reference to Appendix A stack testing is an error and that Entergy is drafting comments to correctly characterize these values. Absent this

Attach me nt 2 so

correction, we are unable to filly comment on the proposed SAM Iimit. The DraR Permit should be recirculated for public review after the two instances of this cite to Appendix A (SAM and VOCs) are cured.

The control technologies that are found to be feasible in Steps I and 2 shouId be ranked according to their control efficiency in Step 3 as BACT is “an emission limitation based on the maximum degree of reduction of each pollutant.. .” 40 CFR 52.2 1 (b)( 12). The degree of reduction means the amount by which a pollutant concentration is reduced, relative to the uncontrolled level. The degree of reduction information is used in Step 3 to rank control options based on emissions from the lowest to the highest. NSR Manual, p. B.25 and Tables B-2 and B-3.

The control efficiency must be determined first so that the controI options can be ranked and the top option selected. One cannot determine whether a given emission limit corresponds to the maximum degree of reduction without first determining what that reduction is and how it compares with reductions achievable by other methods and combinations of methods. The documents posted on ADEQ’s website do not disclose the degree of reduction or any of the information required to estimate the degree of reduction and prepare Step 3 rankings.

Instead, it contains what it calls “typicaI I-IzSO4 emission levels,” a compilation of generic emission rates that are cited to no source. The Application indicates that these typical levels are a compilation of limits from other permits, including several that are lower than proposed for White Bluff: Newmont at 0.001 IblMMBtu, Santee Cooper Cross Units 3 & 4 at 0.00 14 Ib/MM8tu (incorrectly listed at 0.0 14 lb/MMBtu in the Application), and Sandy Creek and Spruce 2 at 0.0037 Ib/MMBtn. I/9 Ap. at 5-10. The ranking required in Stcp 3 must be specific to the project under review. Regardless, the Draft Permit rankings include a range of 0.001 - 0.04 1blMMBtu for a spray dryer absorber, the technology selected as satisfying SAM BACT. Draft Permit at 13. The record is silent on why the lower end of the reported range, 0.00 1 Ib/MMBtu, or h i t s permitted at other facilities do not satisfy BACT for SAM at White Bluff. In addition to skipping Steps 3 and 4, the Steps 1 and 2 andyses are incomplete.

First, Step I must consider all potential control options. It omits the consideration of clean fuels, including conversion to natural gas, which wouId essentially eliminate SAM emissions. It also omits ammonia injection ahead of the ESP, which can achieve up to 90% S A M reduction and modifications to the air heater. SC WB Ex. 100, pp. 759- 760. Further, one of the identified control technologies is sorbent injection. The only chemical considered is Iimestone, which is reported as having an “undetermined” SAM emission level. Dmfi Permit at 13. It is well known that sorbent injection removes 40% to 90% of the SAM. Further, many other sorbents can be used, including trona, magnesium oxide, and hydrated lime. SC WB Ex. 100, p. 759.

Second, the Step 2 analysis concludes that “EwJhile WESP is in theory, an effective control device for collection of aerosols such as sulfuric acid mist, it is a tcchnoIogy that has not been demonstrated in practice on a utiIity scale coal-fired boiler

81

Attachment 2

similar to the White BIuff boiler.” Draft Permit at 14. This is incorrect. WESPs are in operation at many similar boilers in the United States and elsewhere, as discussed below.

b. The SAM BACT AnaIysis Failed To Consider Other Lower Permitted Limits

There are many permits for similar sources that contain lower SAM limits than proposed for White Bluff. Some of these are summarized in Table 1.

Table 1 Summary of Lower Sulfuric Acid Mist Permit Limits

Status: C = under construction; DP = draft permit; P = final permit, construction status unknown; 0 = operating Fuel: SB = sub-bituminous, 3 =bituminous * Sargent & Lundy design Iimit post dry FGD installation from “Emission Summary (drafi-08250X)” spreadsheet ** Draft Permit proposed limit

This table compares the proposed SAM BACT Iimit of 0.0042 lblMMBtu for White Bluff (in bold in the last line) with limits that have been permitted at other similar faciIities and with Iimits estimated by Sargent & Lundy specifically for White 3luff (bold italics). This comparison shows that the proposed S A M BACT limit is very high and thus docs not satisfy BACT.

Attachment 2

These units arc all similar because they burn coal in boilers to generate electricity. They are all pulverized cod-fired boilers. The resulting exhaust gases are similar in that they can and are routinely controlled using the same contro1 equipment. The specific control equipment used at each unit is set out in the permits and typically includes SCR to controI NOx, fabric filter baghouses to control PM, and dry FGD (Springfield, Whelm, Newmont, White Pine, Sandy Creek) or wet FGD to control sulfur dioxide ("SO;') Under ADEQ's theory that dry scrubbers remove more S A M than wet scrubbers, the units in Table 1 that are controIled by wet FGDs (Parish, Iatan, Trimble, Santee Cooper Cross) represent a worst case for White Bluff as wet scrubbers allegedIy emit more sulfuric acid mist than a dry FGD. All of these units include an SCR to control NOx, which would increase uncontrolled S A M relative to White Bluff, which does not yet have an SCR. Further, the units that bum high sulfur coaI (Cross 1-4 and Trimble 2) also represent a worst case as SAM emissions ultimately depend on the sulfur content of the coal. Thus, White BIuffs dry FGD should be able to do much better than, say Springfield 2 or the Cross units or Man.

c. The BACT Analysis Failed To Consider Test Data

We collected stack tests for similar sources that have been tested in the last five years. These data are summarized in Table 2 and where available, are compared to permitted lcvels. Some of the units that have been tested recently for SAM do not have a SAM permit limit as they were constructed before the PSD program was implemented.

Table 2 Summary of SAM Test Resuits

Plant

Marshall 4 I White Bhff

Coal Startup Limit Test sc (IblMMBtu) (lb/MMBtu) W B

EX. SB 1983 - 4.000 I29 1 1 1

0.00036 117 0.00064 118

SBlB I 1 0.0042 1 I 1

83

Attachment 2

This data shows that where permit limits exit, measured levels are factors of 1.5 to 18 times lower than the limit. Further, these data demonstrate that much lower SAM emissions have been achieved than the BACT Iirnit proposed for White Bluff.

d. Engineering Calculations Indicate SAM BACT Limit Is Overestimated

The Applicant’s engineering consultant, Sargent & Lundy, prepared an Excel workbook calculating future potential emission rates from the boilers after installation of the dry flue gas desulfurization (“FGD”) unit.’33 The emission calculations for the uncontrolled and controlled design target and estimated 30-day rolling average emission rates assume 0.87% sulfur content in Jacobs Ranch Powder River Basin (“PR3”) coal with a higher heating value (,,HHV’) of 8,725 BTU/lb. Assuming complete conversion of coal sulfur to sulfbr dioxide (“SO;’), this corresponds to an uncontrokd emission rate of 2.0 Ib/MMBtu of SO?.

Based on this uncontrolled emission rate of 2.0 Ib/MMBtu of SOz, Sargent & Lundy calculated the uncontrolled SAM emission rate from the boilers as 0.0245 lb/MMBtu, assuming an 0.8% conversion of SO1 to SO3 in the boiler and 100% conversion of sulfur trioxide (“so3“) to SAM. Tlie controlled design target S A M emission rate was calcuIatcd as 0.00170 Ib/MMBtu, based on 30% rcmovaI efficiency across the air heater and 90% removal post air Heater. The estimated 30-day rolling average emission rate was calculated as 0.0029 IblMMBtu based on 20% removal efficiency across the air heater and 85% removal post air heater. These arc well bdow the limit that was ultimately chosen as BACT. No expIanation is offered for the fact that White Bluffs engineer estimated much lower S A M emissions than claimed as BACT.

The Sargent & Lundy calculated values are higher than levels that have been permiff ed and achieved because they make conservative assumptions that overestimate emissions at the stack. First, the calculations assume that 0.8% of the SO2 is converted to SAM. This is the value typically used for a d bottom boiler firing a low sulfur eastcrn bituminous coal. SC WB Ex. 125, TabIe 41F4 The subject boilers are wet bottom and low sulfur eastern coals are unlikely to be a dominant fuel trpe due to their cost. The conversion rate for low sulfhr western bituminous and subbituminous coals that would allow the SO2 to be met range from 0.05% to 0.1%. SC WB Ex. 100, p. 754; SC WB Ex. 125, Fig. 4-1.13’ As coal quality must be considered in a BACT determination and the choice of fuels is limited by the design basis of the dry scrubber, which limits fuel choice to western low sulfur bituminous and subbituminous coals, a much lower conversion rate should have been used.

’” Excel workbook “Emission Sumrnnry~D~A_O33508).xls” including worksheets “INPUTS” and “EM IS S I ONS . ”

March ZOOS.

fired Power Plants, J. Air & Waste Manage. ASSOC., v. 54.2003. p. 754.

EPRI. Estimating Total SuIfuric Acid Emissions from Stationary Power PIamtS, Technical Update,

R.K. Srivastava, C. A. Miller, C. Erickson, and R. Jambhekar. Emissions of Sulfur Trioxide from Coal-

IM

135

84

Attachment 2

Second, the Sargent & Lundy calculations assume 20% to 30% removal of S A M across the air heater. This is low based on available measurements, which range from 56% for PRB coaIs to 85% for medium to high sulfur eastern bituminous coals. SC WB EX. 125, Table 3-1.

Third, the Sargent & Lundy calculations assume 85% to 90% removaI of S A M post air heater. The controls that are downstream of the air heater include an ESP and a dry FGD. The EPRI study in SC WB Ex. 125 indicates that the ESP removes 27% of the S A M (SC WB Ex. 125, Table 3-2) and the dry FGD plus its baghouse remove 93% of the SAM (SC WB Ex. 125, Table 3-3). The combined removal for the ESP and dry FGD is 99.3%.

Using these more reasonabIe estimates of removal efficiencies, the SAM emissions at the stack would be 0.000075 I ~ / M M B ~ - U . ’ ~ ~ This is consistent with the measurements in Table 2 at Holcomb but lower than measured at other units. The likely reason is that the method used to measure S A M in these stack tests, EPA Method 8, overestimates SAM emissions. Sec Section VIn.

Neither the SOB nor the Draft Permit contains any discussion of why the proposed dry FGD cannot achieve the emission rates calculated by the Applicant’s engineering consultant. These h i t s contain substantia1 margin as they overestimate boiler conversion and underestimate S A M removal downstream of the boiler.

The S A M BACT limit of 0.0042 IbMMBtu corresponds to a coal sulfur content of 4.9 lb SOzIMMBtu, based on the same assumptions made by Sargent & Lundy in the above-referenced spreadsheet for thc design casei3’ and 2.9 lb/MMBtu for the 30-day rolling average case. These values are much higher than the design basis coal at 2.0 IbhlMBtu and would result in exceedanccs of the proposed SO2 BART limits given the dry scrubber. If SAM emissions at this level are actually anticipated at the facility, additional controls must be required to reduce S A M emissions to levels caIculated by Sargent & Lundy and permitted and achieved at other facilities. These levels can be achieved by installing a sorbent injection system designed to remove 90% of S A M or a wet electrostatic precipitator following the baghouse.

Step 1 of the top-down process requires that “all” control options with potentia1 application be identified and evaluated. Applicable control technologies include combinations of inherently lower emitting processes and add-on controls. For example, the application of combustion and post-combustion controls to reduce NOx emissions at a gas-fired turbine.” NSR Manual, p. B.1 (emphasis in original.) Elsewhere, the NSR Manual notes:

The SAM emissions at thc stack. assuming cod with uncontroIled SO? emissions of 2.0 IblMMBtu, conversion of 0.8% of SO2 to SAM, 56% removal across the air heaters, and 99.3% removal ncross the ESP and dry FGD: (2.0 IblMMBtu of SOz) x (0.00s) x (98164) x (1-0.56) x (1-0.993) = 0.000075 WMMBtu oFSAM.

136

Coal sulfur contcnt corresponding to 0.0042 Ib/MMBtu of SAM: (0.0042 Ib/MMBtu of SAM) / (0.00s) 137

x (98/64) x (0.7) x (0.10)=4.9 lblMMBIu of SOr.

85

Attachment 2

“Combinations of inherently Iower-polluting processeslpractices (or a process made to be inherently less polluting) and add-on controls are IikeIy to yield more effective means of cmissions control than either approach alone.. ..These combinations should be identified in step I of the top down process for evaluation in subsequent steps.” NSR Manual, p. 14.

Combinations of techniques should be considered to the extent they result in more effective means of achieving stringent emissions levels represented by the “top” alternative, particularly if the “top” alternative is eliminated.” NSR Manual, p. B. 1 6. The ADEQ failed to evaluate the obvious combination of the dry scrubber, sekcted as BART and additional control for SAM, such as sorbent injection or a WESP, to satisfy BACT.

Based on this information, we conclude that BACT for SAM at White Bluff is an emission Iimit of 0.00 I Ib/MMBtu, based on a 3-hour average. Compliance should be determined with an SO3 CEMS or quarterly using the modified controIIed condensation method. See Section VIII.

2. BACT Is Not Required For Filterable Particulate Matter

The DraR Permit concluded that BACT for filterable particulate matter (“PW) is an emission limit of 0.0 12 Ib/MMBtu, achieved using a fabric filter baghouse. Draft Permit at I9 and Condition 4. As discussed for S A M , Step 1 fails to consider clean fuels, including conversion to natural gas, which would substantialIy reduce filterable PM emissions. The BACT analysis skips from Steps 1 and 2 to Step 5, omitting critical Steps 3 and 4. The BACT limit is simply stated, with no support. Correspondence on ADEQ’s websitc suggests this h i t was based on discussions with ADEQ, vendors, and Sar cnt &

The proffered limit is higher than other recently issued permits and much higher than actual measurements. As such, it does not comply with the definition of BACT, which is an emission Iimit based on the maximum degree of reduction.

Lundy. However, na actual support is provided beyond Enterm’s representations. 6 8

Other permits contain filterable PM limits that are lower than 0.0 I2 lb/MMBtu, including: Desert Rock, NM (SC WB Ex. 126: 0.01 Ib/MMBtu, 24-hr average); Tocquop, NV (SC WB Ex. 127: 0.01 IblMMBtu, 3-hr average); Two Elk, WY (SC WB Ex, 128: 0.0 10 lb/MMBtu); Las Brisas, TX (SC WB Ex. 129: 0.01 1 IblMMBtu, 3-hr average); and White StalIion, TX (SC WB Ex. 130: 0.01 1 IblMMBtu, 3-hr average). Compliance is determined for Desert Rock using a PM CEMS.

Source test data have shown that even lower emission levels have been achieved. Some of the more recent tests are summarized in Table 3. These data indicate that filtemble particulate matter permit limits are much higher than emissions that are achieved in practice, justifying lower BACT limits than required in existing permits.

Mark BowIes. Entergy, tetter to Siew Low, Arkansas Department of Environmental Quality, Re: Request for Additional Informntion, White BlufFPollution Control Project. April 29,2009, p. 1.

56

Attachment 2

Thus, a filterable PM BACT limit of 0.0 12 Ib/MMBtu is too high and should be lowered to no more than 0.01 0 lb/MMBtu, 3-hour average.

Table 3 Comparison of Permitted Filterable PMlPMlO Emission Limits

with Measured Firterable PiM/PMlO Emissions

These recent tests are consistent with historica1 filterable PM data for coal plants in the State of Florida. The Florida database contains 17 1 performance tests at coal-fired p h i s , conducted from 1990 to 2003, that measured filterable PM/PMlo at 0.010 Ib/MMBtu or lower, and 82 tests that measured filterable PM/PMlo emissions at less than 0.005 lb/MMBtu. The lowest reported PMlPMlo emission rate was 0.0004 IblMMBtu. SC WB Ex. 142.’39 Thus, clearly, a much lower fiIterablc PM limit has been achieved than proposed as BACT for White Bluff.

The ability of a baghouse to reliabIy achieve lower filterable PM emissions than proposed here as BACT depends on the design ofthe baghouse. Particulate is captured on the fabric filter as we11 as the filter cake that develops OR the fabric as the device is run over time. The control or removal efficiency achieved depends not only on thc inlet emissions levels but aIso on all of the variables that affect the development, maintenance, morphology, and other characteristics of the filter cake, and the variables associated with the cleaning cycle of the baghouse. In short, the outlet concentration depends on the

’’’ FIoridn Source Tests compilation.

87

Attachment 2

design of the fabric filter, the choice of filter materials, and the manner of operation and maintenance of the fiIters. However, the record in this case does not contain any detaiIed technical discussion of any of these aspects of baghouse performance nor does it disclose the baghouse design or type of filtration media. Much lower filterabIe PM limits can be achieved through baghouse design.

Thus, the filterable PM limit should be reassessed and based on the actual capabilities of a baghouse designed to achieve the maximum degree of reduction, using, for exampIe, filtration media capable of capturing over 99.99%+ of the inlet particulate loading. At a minimum, the revised limit should take into account the numerous low test results provided above and other similar data that are available from other states.

Based on this information, we conclude that BACT for filterable PM for White Bluff is an emission Iimit of 0.010 IblMMBtu based on a 24-hour average, with compliance determined by a PM CEMS. See Section VIII.

3. BACT Is Not Required For Lead

The Application and Draft Permit do not contain a BACT analysis for Iead. Instead, the documents argue that lead is emitted as solid particulate and is incIuded in the fiIterabIc PM emission limit that was selected as BACT. Thus, the documents reason, emissions of lead will be controIled by the baghouse. 1/9 Ap. at 5- 13 and Draft Permit at 20. The lead BACT limit is then arb i tndy set at the level in the prior permit, at 2.62 x I V5 IblMMBtu. DraA Permit at 20 and Condition 4. This limit can also be expressed as 26.2 pounds per trillion Btus (“IbnBtu”). This limit does not satisfy BACT as it ignores the requirement to consider clean fuels, it fundamentaIIy misunderstands the nature of lead emissions from coal-fired boilers, it fails to consider other permitted and achicved lead emission rates, and it purports to be based on the filterable PM limit, which does not itself satisfy BACT.

First, this anaIysis completeIy ignores the obIigation to consider clean fuels, including 100% conversion to natural gas, in Steps 1 and 2 of a BACT analysis. There is no lead in natura1 gas. Thus, firing natura1 gas would eliminate lead emissions and thus represents the maximum degree of reduction that is achievable. This option cannot be ignored by arguing that BACT for PM satisfies BACT for Iead as natural gas was not considered for filterable PM either. Further, some coals contains more lead than others. The BACT anaIysis failed to consider the lead content of the coal and failed to evduate dean fuels.

Second, the assumption that BACT for PM satisfies BACT for lead is not correct. Lead is voIatilized in the boiler and condenses as very fine particdate matter or nanoparticles ( a . 5 microns) in the pollution control train. SC WB Ex. 143.14’ The highest concentrations of Iead are consistently found in the smallest particles. SC WB

“* R.C. Flngan and S.K. Fricdlnnder, Particle Formation in PuIverizcd Coal Combustion - A Review, In: Recerrt Developnimls irr Aerosol Sckticc, D.T. SIinw (Ed.), 1975, Chapter 2.

gg

Attach men t 2

Exs. 144, 145,''' and 146.14? The particulate collection efticicncy of a baghouse designed to collect PM is generally lower for these nanoparticles that contain most of the lead than for larger particles. AP-42, Table 1.1-5, SC WB Ex. 147Iq3 p. 1582, SC WB Ex. 146, p. 1538. Thus, a fabric filter system designed to meet BACT for PM does not necessarily meet BACT for the submicron (0.3 micron) particks wherc most of the lead is found. These smaller particles also cause proportionately more of the adverse health impacts because they can penetrate deep into the lung. SC WB Ex. 147.

A BACT analysis for lead must consider methods to enhance the removal of these finer particles. Methods to enhance the control of fine lead particles include: ( I ) use of a filtration media with a higher removal efficiency for nanoparticles; (2) use of a wet electrostatic precipitator in addition to the baghouse; and (3) use of an agglomerator upstream of the baghouse. An agglomerator uses electrical charges to attach nanoparticles to larger particles, which are then more efficiently removed by the baghouse.lJJ Agglomerators have been used to reduce opacit (caused by nanoparticles) and PM at several coal fired power plants. SC WB Ex. I48 14; and f49.'46

Third, many permits have been issued with Iead Iimits that are lower than the 26.2 IbKBtu proposed as BACT for White Bluff. These include: Thoroughbred, KY (SC WB Ex. 150: 3.86 IbflBtu), Man, MO (SC WE3 Ex. 108: 5.93 IbTTBtu), Spurlock Gilbert Unit 3, KY (SC W Ex. 151: 6.3 lb/TBtu), AMP Ohio (SC WB Ex. 152: 9.8 IbffBtu), NEVCO Sevier, UT (SC WB Ex. 153: 11.3 Ib/TBtu), Longleaf, GA (SC WB Ex. 154: IS lb/TBtu), SWEPCO Turk, AR (SC WB Ex. 155 16 IbRBtu), Springerville Unit 3, AZ (SC WB Ex. 156: 16 IblTBtu); Longview, WV (SC WB Ex. 157: 18 IbffBtu), and Trimble Unit 2, KY (SC WB Ex. I IO: IS 1bTTBtu).

Third, stack tests demonstrate that lower limits have been achieved. The Springerville Unit 3 2006 stack test did not detect any lead and concluded emissions were less than 0.315 I b f l B t ~ , ' ~ ~ which is 75 times lower than proposed for White BIuK

'" Richnrd L. Davidson and others, Trace Elements in Fly Ash, Envir.otimartafSc~ctic~ & Tecftnolog: v. 8. no. 13. Dccember 1974, pp. 1107-1 I 13; E.S. Glodney and others, Composition and Size Distribulion of fn- State Particulak Materid at a Coal-Fired Power Plant, Ahriosplreric Eiivironmerrt. v. 10, 1976,

'" W.P. Linnk and others. Coniparison of Particle Size Distributions and EIcmental Partitioning from Combustion of PuIvcrized Coal and Residual PueI Oil, J. Air & Waste Marrag.. Assor.. v. 50,2000,

pp. 1071-1077.

pp. 1532-1544.

"' J. S. Lighhty. J.M. Vemnth, and A.F. Sarofirn, Combustion Aerosols: Factors Governing Their Size and Composition and Implicatians to Human I-Ienlth. J. Air & Waste Manrrgc. ASSOC., Y. 50,2000, pp. 1565- 1618.

IJ.I McIlvninc Hot Topic Hour, Impact of PM2.5 on Power Plant Choices, November 2,2006. Voice recording availabIc online to subscribers of McIkaine Power Plant Knowledge System.

I" R. Truce et al.. Reducing PM2S Emissions Using the Indigo Agglomcmtor, Mega 3006.

PRECIP NcwsIetter, indigo Agglomerntor Reduces Emissions in China, no. 405. October 2009.

SpringcrvilIe Unit 3 Stack Tat, AuSust 2006.

IJb

I47

89

Attachment 2

Weston Unit 4, permitted at 25 IbKBtu, w% tested at 2 Ib/TBtu (SC WB Ex. 1 3614’) or 13 times lower than proposed for White Bluff. Wygen 11, which has no lead permit limit, was tested at 6.6 lb/TBtu, or four times lower than proposed for White Bluff. SC WB Ex. lZO.’” MidAmerican Council Bluffs 4, permitted at 26 IbA‘Btu, was tested at 3 lbnBtu in May 2007 and at 0.2 lb/T‘Btu in August 2007. SC WB Ex. 1 I6 - 1 18. Holcomb Unit 1, also with no lead limit, tested at 5.65 IbflBtu. SC WB Ex.1 I I. These wits all burn coal similar to that proposed for White Bluff and use the same particulate controI train.

Based on this information, we conclude that BACT for lead for White Bluff is an emission Iimit of 0.6 IbRBtu based on a monthly average. Compliance should be determined based on daiIy coaI samples, cornposited and andyzed weekIy, and an annual stack test. See Section VIII.

4. BACT Is Not Required for CO

The Draft Permit concluded that BACT for carbon monoxide (TO”) is an emissions limit of 0.15 IblMMBtu, based on a 30-day rolling average and achieved using good combustion. Draft Permit at 16 and Condition 4. CompIiance would be determined by a continuous emission monitoring system. Draft Permit, Condition 4. This limit does not satisfy BACT. The Application performed a top-down BACT anaIysis for CO, purportedly following the procedures set out in the NSR Workshop Manual. 119 Ap. Sec. 4.0. The ADEQ adopted this analysis in the Draft Permit at I 1 - 13. The CO top-down BACT analysis in the Application does not follow the top-down process and as a result selected the wrong CO emission IeveI.

a. Step 1 Incorrectly Evaluates Combustion Controls

The first step of the top-down BACT process is to prepare a “comprehensive” list of all available controI options with a practical potential for application. NSR Manual, pp. B.5-B.7. The Application did not identify all available CO control technologies, and, in fact, failed to even mention some of the most attractive technologies.

The application identified two genera1 classes of controls, combustion controls, comprising boiler design and operation or good combustion practices, and post- combustion controls, comprising catalytic oxidation and EMx (“SCONOx”). Ap., Sec. 5.3.2.2. There are several problems with this list of technologies.

The general class “combustion control” defined as CO conk01 through design and operation of the boiler to Iimit CO is far too broad for a top-down BACT andysis. All available control options must be listed (NSR Manual, p. 3.51, not just general classes of options. Tbe combustion control class includes many control options that were not considered. These includc clean fuels, including firing 100% natural gas, different types

METCO, Source Emissions Survey of Tucson Electric Power Company SpringerviIle Gcnemting

TRC. Performance Test o f the Wygcn I1 Unit 4 at BIack Hi lb Generation, Inc., January 2008.

14s

Station Unit Number 3 Stack, Springewille. AZ for Bechtcl Power. August 2006.

90

Attachment 2

of low N O x burners (leNBs’’), different types of overfire and underfire air systems, flue gas recirculation, gas reburn systems, and various combinations of these tcchnologies.I5’ Instead, the BACT analysis lumps them a11 together and argues that the burners it plans to install to satisfy BART for NOx satisfy BACT for CO. This is wrong.

First, as discussed above, the proposed low NOx burners and modifications to the overfire air system do not satisfy BART for N O x and require the use of a post- combustion control, SCR. A proper BART determination would relax any hypothetical constraint associated with the low NOx burners as boiler outkt N O x could be increased to minimize CO. Thus, SCR would be evaluated as a CO and VOC control option, as it would allow any hypothetical existing boiler constraint to be addressed by selecting burners that yield a higher boiler outlet NOx, which high NOx could be reduced by the downstream SCR.

Second, the source of the CO that triggers PSD review is the LNBs used to control NOx. Entergy argues that NOx emissions increases in the presence of excess oxygen, while emissions of VOClCO decrease with more excess oxygen and vice versa. 4/29/08 Bowles Letter at 1. Entergy claims that the LNBs seIected to meet N O x BART limit the CO (and VOC) emissions that can be achieved. I19 Ap. at 5-4 and Draft Permit at 12-13.

The Application and supporting materials posted on ADEQ’s website do not contain the design basis for the proposed LNB/OFA system. The burner vendor and mode1 are not identified nor the design of the OFA system. These are critical to truth the claims that 0.15 IblMMBtu is the best that can be achieved. The burners and overfire air system selected to satisfy BART for NOx does not necessariIy satis@ BACT for CO and vocs.

Entergy claims the particular type of LNB selected for the boiler upgrade simultaneously increases CO and reduces NOx. They argue their abihty to compensate for the increase in CO is constrained by the furnace design, essentially arging that their hands are tied and they can do nothing to control CO beyond “good combustion control,” which is never defined. Draft Permit at 12-13. This line ofargument is whoIly inconsistent with the definition of BACT and the top-down BACT process.

The Project would replace existing burners with new LNBs and upgrade the overfire air systcm. Thus, the BACT analysis assumes a certain type of burner as the starting point for its analysis and does not evaluate other types of burners that would reduce CO while simultaneously achieving its NOx reduction goals. A different type of burner could be selected with lower NOx and CO emissions.

All LNBs are not created equal. There are LNBs that simultaneously reduce both NOx and CO and others that either do not increase CO or increase it by much smaller amounts than the particular burners apparently chosen (which are unidentified) for this

Is’ GE Environmentnl Energy, Reburn & Advanced NOx Control Technologies. US-China NOx and SO2 Control Workshop, November 2003.

91

Attachment 2

Project. These include Babcock & Wilcox’s AireJet burners (SC WB Ex. 158)‘’’ and low-NOx CCV burner (SC WB Ex. 159);’52 Alstom’s TFS 2000TMR burner (SC WB Ex, 1 60p3 and Babcock Power’s Combustion ControIIed Venturi (CCV) burners with Dual Air Zone (“DAZ”). SC WB Ex. 165.IM See also SC WB Exs. 161 burners are designed to mitigate increases in CO and VOCs.

and 162. These

The BACT analysis does not identify any alternative burner systems, stating with no support that “Iower IeveIs are not technically feasible due to physicaI constraints o f the existing boiler geometry and the aggressive NOx control requirements needed to satisfy Best Available Retrofit Technology for NOx.” 119 Ap. at 54. The ADEQ buys this line of argument, providing no support beyond a conchsory restatement of the applicant’s argument. Draft Permit at 12-13. Thus, the Application rejects the entire universe of LNBs without identifying a single alternative to the chosen (and unidentified) burners proposed for BART. BART does not satisfy BACT. This is not only incorrect, as noted above, it violates the top-down procedure by eliminating control options in Step 1 by grouping them into a class and claiming the class cannot do any better without distinguishing class members.

Technologies are eliminated in Step 2. “A demonstration of technical infeasibility should be clearly documented and should show, based on physical, chemical, and engineering principles, that technical dificdties would preclude the successful use of the control option on the emissions unit under review..” NSR Manual, p. B.7. No such showing is made in the Application.

Second, overfire air systems are commonly installed with low NOx burners to further reduce NOx and to minimize CO emissions. The Project involves modification to an existing overfire air system. Overfire air systems can be designed to reduce or eliminate the increase in CO created in the LNB. In an overfire air system, 20-30% of

~ ~ ~ ~

A. LnRue and others, B&W’s AireJetTM Burner for Low NOx Emissions. 2006 PowerGcn. 151

‘’I Craig A. Penterson and Kenneth R Hules, Reducing NOx Emissions to Below 0. I5 Ib/l06Btu on a 600 MW Utility Boiler with Combustion Control Only. 2003.

Is’ R. Lewis and others, Summary of Recent NOx AcIiievcment with Low NOx Firing systems and I.Iigbly Reactive PRB and Lignite Cod: As Low As 0.10 Ib NOx/MMBtu; Alstom Power. Ultra Low NOx Jntegratcd System for NOS Emission Control from Coal-Fired Boikrs, Final Report. U.S. DOE, December 30,2002.

Company Coal Fircd Boibrs with Low NOx Burners 3nd CFB Annlysis, Electric Powcr 2003.

“B&W’s Experience Reducing NOx Emissions in Tangentially-Fired Boilers.“ Presented to the U.S. EPMDOWEPR Combined Power Plant Air Pollutant Control Symposium: ‘The Mego Symposium.’ August 20-23,2001; Kokkinos, A., D Wasyluk. D, Adams, R. Yavorsky, M. Brower, The Babcock and Wilcox Company. “B&W’s Experience Reducing NOx Emissions in Tangentially-Fired Boilers - 2001 Update.” Presented io Power-Gcn International 2001. December 11-1 3.2001. Whitfield, T., Chip Witon, Dan Pitsko. A. Kokkinos, D. WnsyIuk, M. Boris, and J. Hicks. “ Comparison of NOx Emissions Reduction with PRB and Bituminous Coals in 900 MW Tnngentially Fired Boilers.” Presented to EPRI-DOE-AWMA Combined Power Plant Air Pollutant Control Mega Symposium, May 1922,2003.

B. Courternanche and others. Reducing NOx Emissions and Commissioning Time on Southcm

Kokkinos. A., D. Wasyluk, D. Adams, R. Yavorsky, M, Brower, The Babcok and Wilcox Company.

I .u

Attachment 2

the combustion air is diverted to an elevation above (or in some types of boilers, below) the main burners. The Mobotec ROFA (rotating opposed-fired air) is an overfire air system that turbdcntly mixes the flue gas with high velocity secondary air, thus enhancing combustion. This increases NOx reduction beyond that achieved by an LNB or an LNB plus conventional OFA system, and routinely keeps CO concentrations below 20 ppm (0.0 14 1blMMBtu). SC WB Ex. 1 63.”‘ These systems are commercialIy available, applicable to all types of boilers including turbo-fired boilers, and are in operation on many coal-fired boilers in both the US and Europe. SC WB Ex. 1 64.15’

b. Step 1 Fails To IncIude All CO Control Technologies

In addition to failing to consider clean fuels and properly evaluate various low NOx burners and overfire air options, Enterm also failed to identify and evaluate other types of CO control options. The BACT analysis is therefore incomplete.

First, modern boilers employ sophisticated burner and combustion management systems that sene to optimize overall combustion conditions and often result in 1520% NOx reduction in the boiler itself. reference to these technoIogies and their impkmentution as part of BACT for CO and vocs.

SC WE3 Ex. 166. Yct, the record makes no

Second, combinations of controls were not considered in the CO BACT analysis. The NSR ManuaI indicates that in Step 1 “combinations of inherently lower emitting processes and add-on controls” should be considered. NSR Manual, p. B. 10. The increase in CO is created by reducing NOx with low NOx burners. Additional NOx is removed downstream of the boiler using a selective catalytic reduction (“SCR”) system. The increased CO emissions could be reduced by reducing the NOx removat efficiency of the burners (and thus reducing CO) and increasing the NOx reduction efficiency of the SCR.

Third, the Application failed to consider a number of commercial muhipollutant removal processes that simultaneously reduce NOx and SO2 without causing any increase in CO. These include processes such as the CansolvT” and Powerspan Etectro Catalytic Oxidation (‘tECO’rh’’y) multi-pollutant processes. The Cansolv process was created for the removal of SO? from industrial gas streams, and has been further developed to achieve simultaneous removal of NOx and SOz. The process has achieved 99%+ removal of SO2 in a number of applications and 90% to 95% NOx removaI at the pilot scale.

E.E. Haddad, J.S. CriIley. and Brian S. Higgins. The Viability and Economics of Adding a 156

ROFNRotamix MobotecSystem to a SeIective Catalytic Reduction (SCR) Inskdlation. NETYDOE 2003 Confcrence on SCR and SNCR, Oclober 2003; L A . Coombs and others. SCR Levels of NOx Reduction with ROFA and Rotnmix (SNCR) nt Dynergy’s Vcrmilion Power Station. 2004 Stack Emissions Symposium. JuIy 2004

‘’’ See, for example, http:llwww.neuca.nctnibrarylcase-studiesldefaulr.cfm. Mobotec Projects - USA. I57

93 Attachment 2

The ECO pro~ess’’~ can simultaneously achieve high removal efficiencies for NO, (90%), SO1 (99%), fine particulate matter (95%), and mercury (8540%) as well as acid gases, and, thus, from a systems standpoint, offers advantages over conventional pollution control trains. It requires no reagent preparation or landfill, it creates a marketabIe fertilizer, it can be easily retrofitted for CO? capture, and it reduces outage time for installation as it Iocated at the backend of the flue gas path.’” The process has been successfuIIy demonstrated at the First Energy Burger coal-fired station and a 2 1 5 MW unit is currently being installed at the FirstEnergy’s Bay Shore Plant.’“ It has been selected by American Municipal Power-Ohio (“AMP-Ohio”) for use at its proposed 1000-MW American Municipal Power Generating Station, which wiIl bum high sulfur Ohio coal, and has been permitted.IQ

Fourth, the CO BACT anaIysis failed to consider modifications to the pulverizers that would improve combustion and thus lower CO. These include technologies such as the MasMilI, which separates the hard abrasive minerals while recovering their carbon content. This reduces boiler fouling and increascs combustion efficiency, thus reducing COY

Fiflh, the Application did not consider thermal oxidation, which routinely removes 90%+ of the CO (98% of the VOC) from gas streams simiIar to those from White Bluff. Therefore, thcrmaI oxidation will achieve greater reductions of CO (and VOC) from and must be evaluated to establish BACT limits unless Entergy adequately documents adverse energy, environmental, and economic impacts. NSR Mama1 at 8.6. This has not been done.

ThermaI oxidation is an available pollution control technology. At least one Portland cement kiln, in Midlothian, Texas, uses thermal oxidation to control CO emissions. Tlicmal oxidation is widely used in ethanol plants, refineries, and other sources to control VOC and CO emissions. Therefore, thermal oxidation is an available control technology that must be considered in a top-down BACT analysis. NSR Mmrzial 3.1 1 (“AppIicants are expected to identify all demonstrated and potentially applicabIe control technology alternatives. -. Opportunities for technology transfer lie where a control technology has been applied at source categories other than the source under consideration.”)

Is’ Powerspan Clean Energy Technology, ECO Commercially Demonstrated; FulI-ScaIe InstalIntion Proceeding. See also ivww.powerspan.com.

Boylc, D. Steen, and A.J. Dovale. Jr., Commercial Demonstration of ECO Multi-Pollutant Control Technology, 1CAC Forum 2003.

Ih’ PoiverSpan. ECO CommcrcinIIy Demonstmted: FuIl-ScaIe Ins!allation Proceeding.

American Municipal Power Gcnerating Station Project, November 17,2006; http:llwww.powcrspan.co~ne~lrs/reIcase_28.pdE

FirstEnergy. FirstEnergy Experience, SIides from Powerspaii Open House. Scptembcr 28,2005; P.D.

Ncws Release, AMP-Ohio Dedarcs Intent 10 Pursue Powerspan Air Emissions Control Tccbnology for

www.nin~rniI 1-1 Ic.com.

94

Attachment 2

As the NSR Mama/ provides: “[tJechnology transfer must be considered in identifying controI options. The fact that a controI option has never been applied to process emission units similar or identical to that proposed does not mean it can be ignored in the BACT analysis as here if the potential for its application exists.” e Mama1 at B. 16. However, Enterm and ADEQ did not even mention this technology. This technology should be listed in Step 1 and should not be eliminated in Step 2 because it has never been used on a coal plant.

c. No Step 3 Ranking

Step 3 of the top down process requires that the remaining technologies, not othenvise eliminated in Step 2, be ranked based on control effectiveness. NSR Manual, Sec. UI.C, p. B.7. The AppIication and Draft Permit do not contain any ranking for White Bluff, but rather a chart with three entries -combustion controIs as a class, oxidation catalyst, and EMx - with emission levels for only combustion controIs gleaned from the U.S. EPA’s RACTlBACTlLAER Clearinghouse (“Rl3LC”). This is not an adequate Step 3 ranking.

First, as discussed above, combustion control is a broad category that encompasses a large number of individud controI options. The Application and Draft Permit are silent as to which specific combustion controls will be used in this case, beyond low-NOx burners that increase CO emissions and the control efficiency that can be achieved by each, when appIied to the White Bluff boilers. Thus, it is not possible to determine the emission Iimit that corresponds to the maximum degree of reduction.

The Draft Permit states only that “good combustion” will be used. Draft Permit at 12. The Application and Draft Permit faiI to disclose what constitutes “good combustion.” Because this can encompass a number of separate control options, a full top-down analysis must be done that evaluates each of these options, including Step 3 ranking and Step 4 evaluation of energy, environmental, and economic impacts if the top option is not selected.

Second, the rankings must be based on emission reductions that can be achieved by each control option when applied to the subject Project, not rankings based on unidentified projects cberry picked from the RBLC. BACT is ‘an emission limitation based on the maximum degrec of reduction of each pollutant.. .’* 40 CFR 52.2 I (b)( 12). Thus, in Step 3, control options are ranked by their effectiveness for the project at hand. This requires that the degree of reduction be determined and used to estimate emissions from the Project. NSR Manual, p. B.25 and TabIes B-2 and 3-3. The Application and Draft Permit do not contain a Step 3 ranking. Thus, it is not possible to determine BACT.

d. Ignores Lower Permit Limits And Test Data

Lower CO limits than 0. I5 IblMMBtu, 30-day average, have been permitted at many facilities including: Trimble Unit 2, KY (SC WB Ex. I IO: 0.10 IblMMBtu);

Attachment 2

Thoroughbred, KY (SC W3 Ex. 150: 0.1 Ib/MMBtu); Longview Power, WV (SC WB Ex. 157: 0.1 1 IblMMBtu); Prairie State Energy Center, IL (SC WB E x . 167: 0.12. IbMMBtu); Dallman Unit 4, IL (SC WB Ex. 168: 0.12 IblMMBtu); CoIeto Creek, TX (SC WB Ex. 169: 0.12 IblMMBtu); WE Energies-Elm Road Station, WI (SC WB Ex, 170: 0.12 lb/MMBtu); Comanche Unit 3, CO (0.13 Ib/MMBtu); Two Elk Power Plant (0.135 IbMMBtu); and Big Cajun 11, LA (SC WB Ex. 171: 0.135 lb/MMBtu). The Review Summary does not contain any justification for rejecting these lower CO limits or explaining why they do not constitute BACT in this case. These lower limits can be met whiIe achieving a low NOx emission rate, as explained above.lM Dallman Unit 4 is operating and is in compliance with its CO limit.

Lower CO limits have also been achieved in practice. SpringerviIlc Unit 3, which bums a simiIar coal and has the same pollution control train as White Bluff, has been stack tested annually from 2006 to 2009. The measured CO emissions are 0.062 IbMMBtu, 0.016 IblMMBtu, 0.005 lb/MMBtu, and 0.047 1blMMBtu. SC WB EXS. 136 - 139.

Based on the forgoing, we conclude that BACT for CO for the White Huff boilers should be an emission limit no higher than 0.10 lb/MMBtu, based on a 30-day rolling average.

4. BACT Is Not Required For VOCs

The Draft Permit concludcd that BACT for Volatile Organic Compounds (“VOCs”) is an emissions limit of 0.004 IblMMBtu, achieved using good Combustion. Draft Permit at 16 and Condition 4. This Iimit does not satisfy BACT for the same reasons explained above the CO. The above comments on the flawed COBACT analysis are incorporated here by reference for VOCs. The VOC BACT analysis has additional problems, discussed below.

First, the Draft Pernit claims the limit is based on an Entergy proposal, which is based on “performance of an initid Part 60 Appendix A Reference Method stack test.” Draft Permit at 16. As discussed above for SAM, the record does not contain the subject stack test or any other basis for the proposed VOC BACT limit. In the PSC proceeding, Entergy’s response to Sierra Club Data Request 5-6 indicates that this reference to Appendix A stack testing is an error and that Entergy is drafting comments to correctly characterize these values. Absent this correction, we are unable to fully comment to the proposed VOC limit. The Draft Permit should be recirculated for public review after the two instances of this cite to Appendix A are cured.

The Draft Permit and Application contain no further support for the proffered Iimit. It is simply stated as fact without consideration of lower VOC limits permitted at other facilities or measurements made at other facilities.

IM Ex. 143; EX. I* Ex. 145

96 Attachment 2

However, a letter from Entergy to ADEQ suggests the VOC limit is based on the following discussion:

"BACT is a case-by-case evaluation. In the case of the White BIuff Air Pollution ControI Retrofit Project, the critical factor is that the White Bluff boiIers are existing emissions units, and are fundaInentally constrained by their existing design, geometry and fuel combustion characteristics. As stated in the BACT Analysis, Section 5.6.5, there are no technically feasible add-on VOC controls applicable to these units. Therefore, the lowest emission limit that can be continuously achieved for these units through good combustion controI represents BACT for VOC for these units. This emissions level was determined based on an engineering determination supported by the boilers original equipment manufacturer, in particdm in combination with other combustion modifications being applied to further reduce emissions of NOx from these existing units. The existing White Bluff boiIers can not physically achieve the Ievels of controI being permitted today for the latest generation of PC boilers. The White Bluff boilers must optimize the tradeoff between NOx and VOClCO emissions to sirnultaneoudy achieve BACT limits for each. As stated in the application, N O x emissions increase in the presence of excess oxygen, while emissions of VOClCO decrease with more excess oxygen, and vice versa. Given the modifications being proposed to reduce NOx from the White Bluff boilers, and given the limitations inherent to retrofit of existing equipment, 0.004 IbMMBtu VOC is the lowest BACT limit that can be reliably achieved for these particular boilers, and this emissions level has been determined to represent the top Ievel ofcontrol for this particular air poIlution controJ retrofit project.**165

This does not satisfy BACT. The discussion offers no physicaI evidence that the boilers are in any way constrained by geometry, e.g., engineering calculations, original equipment design, drawings, vendor quotes. As the Project involves retrofitting new low NOx burners and modifications to the existing overfire air system, absent a convincing engineering analysis to the contrary, these changes can be made to achieve Iower VOC (and CO) Iimits than proposed using different burners and other options discussed above for CO.

As discussed for CO, there is not necessady a trade-off between lower NOx and lower VOClCO emissions. Newer low NOx burners can achieve low NOx as well as low VOC and CO emissions. For example, the DRBlCZ NOx burners developed by Babcock and WiIcox have demonstrated via testing at Wygen Unit 1 that NOx values as low as 0.13 IblMMBtu were achieved, while simultaneously providing CO values as Iow as 100 ppm (0.072 lb/MMBtu) and very low Loss on Ignition (LOI), which is indicative of Iow volatile organic compounds (VOC). Wygen Unit I burns PRB coal. SC WB Ex. 158.166

Bowles. Entergy Arkansas, Inc., Letter to Arkansas Department of Environmental Quality, Re: Request IM

for Additional Informntion, White BIuff Pollution ControI Project, ApriI 29,2009

"* E& IV'S A i r d d A ' Bttwer for Low N0.s Emissiots, 2006 Power-Gcn Intcrnationaf, November 28-30, 2006, Orlando. Floridn. U.S.A.

Attachment 2 97

Wygen is not the only example showing that lower NOx can be achieved while achieving low CO and VOC. In addition to the examples provided in the CO BACT analysis, other vcndors have provided examples of low NOx and low CO for non-PRB fueIs. A different vendor, Foster Wheeler, in a presentation at Power-Gen Asia in September 2006, SC WE Ex. I7 1 ,lb7 also provides examples of testing confirming this fact. Using proper air-fuel biasing technologies, Foster Wheeler was abIe to achieve very low NOx and CO emissions for different fuels. In one case study discussed in the paper, on a unit burning PRB coal, NOx emissions of 0. I I IblMMBtu were achieved while keeping CO levels to 5 ppm (0.0036 lb/MMBtu). IncidentalIy, unburned carbon levels Ieaving the boiler were also low. This was demonstrated in 2002. Similarly, burner retrofits at Georgia Power's Scherer Units 3 and 4 demonstrate NOx levels of 0. I3 lb/MMBtu after the retrofit, with CO ranging fiom 1 14 to 121 ppm (0.08 - 0.087 IblMMBtu at 3% 02). As such, this further rcfutes the contention that low NOx levels can only be achieved with .corresponding higher levels of VOC (and CO) emissions. SC WB Ex. 172."'

required to satisfy both BART and BACT for NOx. Even assuming a boiler constraint related to optimizing NOx and VOCICO, the use of SCR will allow burner choice to achieve Iower VOC (and CO) emissions by allowing an increase in boiler outlet NOx.

Further, as discussed elsewhere, post-combustion control for NOx, SCR, will be

Regardless, Entergy 's engineer, Szlrgent & Lundy, was not persuaded by this argument as they calculated a design target emission rate and an estimated 30-day rolling average emission rate of 0.0034 lb/MMBtu based on a boiler VOC emission rate of 0.06 lb/ton from AP-42, Section 1.1 - 1 9 (PC-fired, dry bottom or wall-fired boiler firing bituminous or sub-bituminous) and a gross higher heating value of 8,725 BTUllb.

Second, the Draft Permit's ranking of control technologies indicates that combustion controI can achieve VOC emission levels of 0.0024 to 0.004 Ib/MMBtu. DraR Permit at 16. The DraA Permit does not explain why the lower end of this range, 0.0024 IblMMBtu, is not BACT for the White Bluff boilers, especiaIly given that new burners will be retrofit.

Third, the DraR Permit does not consider VOC limits that have been required at other facilities. Many permits have been issued with Iower VOC BACT limits for similar coal-fired boilers as summarized in Table 4

'" Fuel: Injection for Pulvcrized Coal Fired Power Boilcrs - Automatic Air IO Coal Binsing For Lower Ovenll Emissions. presentcd at 2006 Powcr-Gen Asia, Hotig Kong. Hong Kong, Septcmber 57,2006.

'M A. Kokkinos, et al., "Which i s Easier: Reducing NOx from PRB or Bituminous Coal. Powcr 2003.

Attachment 2

Table 4 VOC BACT Permit Limits

Ex. Santee Cooper Cross 3,4 0.0024 105

, Intermountain 3 0.0027 I73

Plant I 0.0030 I74

0.0034 121 0.0034 169

I Holcomb2 0.0035

' '{Creek 0.003'

Lon leaf

0.0037 White Bluff 0.004

The AppIication identified some of these limits, but argues for a higher limit as there are many VOC BACT limits that are higher than 0.004 Ib/MMBtu. 4/29/09 Bowles Letter at 2 and 119 Ap. at 5-13. This badly mangles BACT which is the emission limit based on the maximum degree of reduction. The Application also inconectly rejects the lower VOC limit permitted in Arkansas for the Turk Plant (0.0036 Ib/MM3tu), claiming White 3luffis not subject to 112(g) MACT. 119 Ap. at 5- 13. Any permit limit, regardless off its regulatory origin, is fair game for a BACT determination.

Fourth, Hawthorne Unit 5, a Missouri plant with the identical control train plus SCR, has been source tested for VOCs six times and has consistently achieved a lower VOC limit than proposed for White BIuff (see Table 5) . If SCR alIows superior performance at Hawthorn, then SCR must be evaluated in the VOC (and CO) BACT analyses.

Attachment 2 99

TabIe 5 VOCs Measured at Hawthorn Station Unit 5

Average of SC WB

2003 2004 2005 2006 2007 2008

0.002 182 0.0003 I83 0.0004 184 0.0005 IS5 0.0010 186 0.0038 187

Stack tests are a reasonable basis for establishing BACT when compliance is determined by annual stack tests.

3. BACT Is Not Required For PoIIutants That Entergy IncorrectIy Assumed Would Not Trigger PSD Review

Entergy performed a PSD applicability analysis to determine if PSD review was triggered. Based on this analysis, Entergy concluded that PSD review was not required for NOx, SO:, total PM10, and PM2.5. 1/09 Ap., Sec. 4.0. However, Entergy's applicability analysis was flawed for several reasons as discussed in Section I11 of this letter, and our analysis shows that project being authorized in this permit wouId result in a major modification of SO2, NOx, and C02. Further, Entergy impropedy concluded that PSD review was not required for PM 10 and PM2.5. The emission limits for these pollutants included in the Draft Permit do not satisfy BACT. BACT limits for NOx and SO2 are discussed above. BACT limits for PM 10 and PM2.5 are discussed below. And finaIly, Entergy totalIy ignored CO2. The Draft Permit does not contain a C 0 2 BACT analysis or any C02 limit, which is discussed below.

1. The SO2 BART Emission Limit of the Draft Permit Docs Not Reflect BACT for SO2.

As shown in Section 111. above, the increase in permitted heat input capacity of the boilers and turbine upgrade project will result in a major modification of SO2 at the White BIuff plant and a significant net emissions increase of SO2 at each White Bluff unit. Further, it is likeIy that each White Bluff unit triggered PSD applicability for SO2 with the economizer replacements and possibly other changes at the White BIuff facility. Therefore, ADEQ must require the White Bluff units to meet BACT for 502, which ADEQ has failed to do so. Entergy and ADEQ improperly found that the projects proposed in the White Bluff permit application were not subject to PSD for 502. Thus, ADEQ has erred in issuing this draft permit White Bluff without requiring BACT for SOL

Attachment 2 100

The dry scrubber and the proposed SO2 emission limit of 0.15 lb/MMBtu that ADEQ has proposed in the draft permit to meet BART requirements does not satisfy BACT for S02. BACT is required to be based on the *‘maximum degree of emission reduction” that is determined to be achievable for the White Bluff units, taking into account environmental, energy and economic impacts. As we have shown in our comments above on Entergy’s proposed 3ART controls, a wet scrubber achieves a much greater degree of SO2 emissions reductions than a dry scrubber. Even with Iow sulfur coal, a wet scrubber can achieve 98-99% SO2 removal, whereas a dry scrubber may onIy achieve 9 2 5 9 5 % SO2 removal.

There have been numerous proposed and final SO2 BACT determinations for new coal-fired power plants burning subbituminous coal or other low sulfur coal that have required installation of a wet scrubber as BACT. Those include, but are not limited to, the proposed EIy Energy Center, the proposed Toquop power plant, and the Desert Rock

Other facilities that plan to burn subbituminous coal or si blend of subbituminous coal and bituminous coal have proposed andor have been required to install a wet scrubber to meet SO2 BACT, such as Plant Washington,l7’ The proposed permit for this facility requires as part of the facility’s BACT limits a 97.5% SO2 removal efficiency re uirement that appIies on a 30 day rolling average regardless of the type of coal burned.’” Such high levels of SO2 control cannot be achieved with the dry scrubber currently planned to meet SO2 BART requirements at White Bluff.

The fact that wet scrubbers have been either proposed by the owners as BACT for these facilities or proposecUfinalized as BACT by a permitting agencies means that the cost of a wet scrubber at these subbituminous coal-fired power plants has been considered reasonable to meet BACT requirements. Absent detailed and supported contro1 cost data that show othenvise, what has been determined to be economically reasonable for BACT at a similar coal-fired power lant unit should also be considered as economicalIy reasonable for the White Bluff units. 8 2

The top level of SO2 control with a wet scrubber that should be evaluated in a BACT analysis for White Bluff should be no less than 9849% SO2 control. Our comments and cited documents in the above SO2 BART comments provide support that such levels are achievable and have been achieved in practice. At the minimum, BACT for the White Bluff units should be no less than 98% SO2 removal with a wet scrubber which, based on current uncontrolled SO2 emission rates exiting the boiler would equal an emission limit of 0.02 Ib/MMBtu. Such IeveIs of control are standard today with a wet ~crubber.”~ ADEQ should refer to our discussion of SO2 controls in our BART

‘M The proposed Ely Energy Center permit i s airached as Ex. 75, and the proposed Toquop permit is attached as Ex. 11. The Desert Rock permit is at Ex. 35. ”O Proposed Plant Washington pcrinit is attachcd as Ex. 76. I” Id at Condition 3.14.

I f ‘ Sea ng., EPA’s New Source Rcvicw Workshop Manual at B.44

at htr n : ~ ~ w i v w . i c x .cuin/i4dpn~t.&i ndex.c fm’?narreid-3?O 1, Sargent & Lundy, Flue Gas DesuIfurization See discussion of Acid Gas/S02 Control Tcchnologies on the Institute of Clcan Air Cornpanics wcbsite 173

101

Attachment 2

commcnts above for more information and documentation that shows wet scrubbers achieve the maximum degree of SO2 control, the lowest SO2 emission rates, and have other environmental benefits inchding achieving low levels of SO2 concentration which is necessary for cost effective C02 removal.

Power plant units burning low suIhr subbituminous coal similar to White Bluff have been required to meet SO2 3ACT limits much Iower than 0.15 IblMMBtu, whether being controlled with a wet or a dry scrubber. The attached NationaI Park Service spreadsheet of BACT determinations for coal-fired power plants shows that BACT limits as high as the 0.15 Ib/MM3tu SO2 BART Iimit applicable to White Bluff are unheard of. Ex. 23. Low sulfur subbituminous coaI plants with dry scrubbers have recentIy been permitted with SO2 emission rates of 0.065 lblMMBtu or lower. For example, the proposed Toquop permit included an SO2 BACT limit of 0.06 1blMMBtu on a 24-hr average basis (Ex. 22). The Newmont Nevada plant is subject to an SO2 BACT limit of 0.065 1bMMBtu on a %hour averaging period (Ex. 20) and is currently operating in compliance with its BACT limits.

Coal-fired power pIants controlled with wet scrubbers have also proposed or been subject to BACT limits much lower than the 0. I5 lblMMBtu BART limit applicable to White Bluff. For example, Florida Power and Light proposed an SO2 BACT limit of 0.04 IbMMBtu based on a 24-hour rolling average for the proposed Glades facility. See Ex. 19. The Desert Rock facility, which will burn low sulfur western coal, is subject to an SO2 BACT limit of 0.060 lb/MMBtu on a 24 hour basis, and the proposed Toquop (Ex. 22) and EIy Energy Center (Ex. 75) permits, both of which would burn Powder River Basin coal, aIso included BACT Iimits of 0.06 lb/MMBtu for S02.

The determinatiori of BACT and the appropriate BACT emission Iimit is not just limited to prior BACT determinations. BACT is to be based on the maximum degree of emission reduction that is achievable. As Ex. 18 shows, there are numerous coal-fired units that are achieving SO2 cmission rates much Iower than the 0.1 5 IbMMBtu SO2 BART limit appIicable to White Bluff. The best performing similar source in 2008 was Pleasant Prairie Units 1 and 2 in Wisconsin. The 2008 annual average achieved at Unit 1 was 0.02 1 Ib/MMBtu and at Unit 2,0.027 IblMMBtu. Ex. 330. These units are equipped with a wet limestone scrubber and burn a low sulfur Powder River Basin CoaI. SimiIar data for the first six months of 2009 indicate that other units are currently achieving even lower SO? emissions, including Iatan Unit I at 0.005 I Ib/MMBtu; Muscatine Unit 9 at 0.01 3 IblMMBtu; Hammond Unit 2 at 0.0 16 IblMMBtuu; Gorgas Unit 10 at 0.0 17 IblMMBtu; Prairie Creek Unit 4 at 0.0 19 1blMMBtu; Hopewell Power Station Units 1 and 2 at 0.020 lb/MMBtu; and Centralia Unit BW22 at 0.021 lb/MMBtu.

In summary, for the reasons we have discussed above, there is no way the proposed 0.15 1blMMBtu SO2 BART Iimit for the White Bluff units could be considered as satisfying SO2 BACT requirements for White Bluff. Further, use of a wet scrubber instead of the proposed dry scrubber currently reflects BACT for White Bluff based on

Techology Evaluation. Dry Limc vs. Wet Limestone FGD, March 2007.

I 02 Attachment 2

recent proposed and final BACT determinations and PSD permits for coal-fired power plant units burning low sulfur coal as White Bluff burns. Thus, the SO2 emission limit of the draft White Bluff permit cIearIy do not reflect BACT. Because this permit aIlows for a major modification of SO2 at White Bluff, the draft permit is deficient for failing to require BACT for SO2 from the White Bluff units. Further, because it appears that each White Bluff unit triggered PSD applicability for SO2 with the economizer replacements and possibly other changes at the White Bluff facility, the dmft permit is also deficient for not including BACT requirements for SO2 at the White Bluff units.

2. The NOs BART Emission Limit of the Draft Permit Does Not Reflect BACT for NOx.

As shown above, the increase in permitted heat input capacity of the boilers and turbine upgrade project will result in a major modification of NOx at the White Bluff plant and a significant net emissions increase of NOx at each White Bluff unit. Further, it is likeIy that each White Bluff unit triggered PSD applicability for NOx with the economizer replacements and possibly other changes at the White H u f f facility. Therefore, ADEQ must require the White Huff units to meet BACT for NOx, which ADEQ has failed to do so. Enterg and ADEQ improperly found that the projects proposed in the White Bluff permit appIication were not subject to PSD for NOx. Thus, ADEQ has erred in issuing this draft permit White Bluff without requiring BACT for NOx.

The low NOx burners and overfrre air that Entergy plans to install to meet BART in no way satisfy BACT for NOx. BACT is required to be based on the “maximum degree of emission reduction” that is determined to be achievable for the White Bluff units, taking into account environmental, energy and economic impacts. BACT determinations for pulverized coal-fired boilers such as the White Bluff units have consistently rc uired SCR in addition to Iow NOx burners and overfire air to meet BACT require~nents.~’~ This combination of controls is the top level of controI for NOx at coal- fired power plant units.

An SCR system can achieve 90+% NOx removal. According to Babcock & Wilcox, commercial SCR installations have shown that 90% NOx reductions can be achieved with low ammonia sIip.”’ Indeed, Babcock & Wilcox states that up to 95% NOx control can be achieved with SCR.176 A review of recent SCR retrofits most definitively shows that very high levels of NOx removal are being achieved by rcccnt

”‘ See spreadsheet summarizing proposcd and final BACT deleminations prepared by the National Park Service, Ex. 23.

See Bielawski, G.T., J.B. Rognn, and D.K. McDonald, How Low Can We Go? Controlling Emissions in New Coal-Fired Power Plants, Prcscnted to the U.S. EPMDOIYEPRI Combined Power Plant Air Pollutnnt Control Symposium: “The Mega Symposium,” August 2001, Ex. 75. ”’ Id.

Attachment 2 103

SCR retrofit installations. NOx emission rates less than 0.05 lblMMBtu are routinely achieved, and NOx removal efficiencies are typicdly around

Permitting agencies have routinely required lower NOx limits in recent BACT determinations for coal-fired power pIant units, with many proposed and required BACT limits of 0.05-0.06 IblMMBtu and at least one NOx BACT limit as Iow as 0.035 IblMMBtu. The Desert Rock Energy FaciIity permit requires the facility to achieve, after a NOx optimization period, a N O x emission rate of 0.035 IblMMBtu on a 365 day rolling average and an emission rate of 0.05 Ib/MMBtu on a 30-day roIling average. Ex. 35. The proposed Plant Washing permit includes a NOx BACT limit of 0.05 IblMMstu (30 day average) (Ex. 77). There are numerous other examples of proposed or final emits for coal fired power pIants with NOx BACT limits of 0.06 IblMMBtu or lower. 171:

SCR has been required to be retrofitted at cxisting coal-fired power plant units as well. SCR systems have been retrofitted to numerous existing coaI-fired power plant units as a resuIt of the NOx SIP call andor the Clean Air Interstate Rule. SCR systems have also been required to be instalkd at existing power pImt units pursuant to federa? claims of PSD applicabiIity. For example, the Coronado Consent Decree requires installation of SCR along with a low NOx combustion system at each unit and a NOx emission limit for each unit of 0.08 Ib/MMBtu (30 day average) be met.179

The determination of BACT and the appropriate BACT emission limit is not just limited to prior BACT determinations. BACT is to be based on the maximum degree of emission reduction that is achievable. High IeveIs of NOx removal and low NOx emission rates have been proven. Ninety percent NO, removaI was achieved on 10,000 MW of coal-fired gencration in 2004. Ex. 79.”’ The McIIvaine reports indicate three of I-Ialdor Topsoe’s SCR instdlations averaged over 95% NO, reduction during the 2005 ozone season. Ex. SO.’8’

I n See Erickson, Clayton A. et d. SeIective CataIytic Reduction System Performance and ReliabiIity Revicw, Thc 2006 MEGA Symposium, Paper # 121. Ex. 34.

Scc also thc National Park Service spreadsheet on BACT limits for New PC Power Plants, Ex. 23.

See United States of Amcrica v. SaIt River Project AgricuItunl Improvemcnt and Power District, CiviI Action No. CV OS- 1479-PHX-JAT Consent Decree (Ex. 78). This senlement rcsolvcd an EPA complaint alleging that thc Coronndo pImt had been modified without obtaining prior PSD permits. Coronndo Conscnt Dccrcc at 4. The Consent Dccrcc resolvcs all civil claims brought by the EPA, thus one can presume that the costs of instnllntion of SCR and combustion contds were dctcrniincd by EPA to be reasonable even as II retrofit to nn existing facility.

Competitive Power CoIlcge. PoiverGen 2005. Selective Catalytic Reduction - From Planning to Operation. 77.

Is‘ Mcllvaine Utility e-Alert, No. 798. Novcmber 3,2006. Mr. Nate White of Haldor Topsoe provided the following information: “Topsoe has over 100,000 hours of opemting experience on PRB coal. In fact, three Topsoc supplied SCRs achieved the highcst NO, cfficicncy for a11 U.S. cod-fired high dust SCRs, avcraging ovcr 95% NO, reduction over the 2005 Ozonc scuon.”

17R

Attachment 2 104

A review of NO, emissions data for coal-fired power plants equipped with SCR shows that NO, emission rates of 0.05 lblMMBtu and Iess are being achieved in practice on a 24-hour avenge basis. See Ex. 8 1 which includes a tabIe of coal-fired power plant units achieving daily NO, emission rates less than 0.06 IblMMBtu during the 2006 ozone season.'82 For example, the maximum daily NO, emission rate during the 5 month ozone season at W.A. Parish Unit 8 was 0.044 IblMMBtu and the lowest was 0.021 1blMMBtu. The maximum daily NO, emission rates at W.A. Parish Units 6 and 7 during the 2006 ozone season was 0.05 and 0.049 lb/MMBtu, respectively, and the lowest daily NO, emission rate was 0.033 and 0.03 lb/MMBtu, respectively. The lowest NO, emission rate being achieved at the W.A. Parish Unit 8 is consistent with a NO, removal efficiency of the SCR system ranging from 86% to 90% depending on the boiIer-out NO, emission rate.'83 AI1 of these facilities burn Powder River Basin subbituminous coaI which is the typical coaI burned at White Bluff. This data shows that not only are recently required NO, BACT emission limits of 0.06 Ib/MMBtu and lower achievable, but much lower NO, emission rates are achievable with SCR systems. As shown in the attached table of NO, emission rates at plants with SCR systems, NO, emission rates as low as 0.009 1blMMBtu have been achieved on a daily average basis.

In summary, for the reasons we have discusscd above, there is no way the proposed 0.15 lb/MMBtu NOx BART limit for the White Bluff units could be considered as satisfying NOx BACT requirements for White Bluff. Clearly, instalIation of an SCR along with combustion controls currently reflects BACT for White Bluff based on proposed and final BACT determinations for the past decade. Thus, the NOx BART emission limit of the draft White Bluff permit does not reflect BACT. Because this permit allows for n major modification of NOx at White Bluff, the dmfi permit is deficient for failing to require BACT for NOx &om the White BIuff units. Additionally, since it appears that each White Bluff unit triggered PSD applicability for NOx with the cconomizer replacements and possibly other changes at the White Bluff facility, the draft permit is deficient for not including BACT requirements for NOx as a consequence of those major modifications as well.

3. BACT Is Not Required for Total PMlO

The Dmfi Permit concluded that PSD review was not required for totaI PM IO as the Project would result in a reduction of 45 tpy. DraR Permit at 10. However, as expIained above, it is technicaIIy impossible for the Project to cause a Iarge increase in SAM, the major component of the condensable h c t i o n of PMlO and a large increase in filterable PM, the major component of the filterable fraction of PM 10, while

"' This table was excerpted from Ranajit Sahu's Response to Executive Secretary's Written Deposition Questions for S i e m Club Expert Witness Ranajit Sahu and Related SuppIenicntaI Request for Production of Documents, In the Matter of Sevier Powcr Company Powcr Plant, appeal bcforc the Utah Air Quality Board. This document indicates that this data reflects actual daily NO, emission rates at coal-fircd units with SCR from the acid rain database (i.e., data available at http:i:l:nniddataandmaps.wn.~i~~,cdnii).

IbMMBtu to 0.77 Ib/MMBtu. This nngc of NO, rcmoval eficicncy is based on n boiler-out NOz emission rate ranging from 0.15 IS3

Attachment 2 105

simultaneously reducing emissions of PMIO. I n fact, as explained above, the Project will resuIt in a significant increase of total PM 10, triggering PSD review.

The Draft Permit and underlying Application do not contain a BACT andysis for total PM IO, based on the erroneous assumption that emissions were less than the PSD significance level of I5 tpy. However, the Draft Permit does contain a limit on total PM I O of2 14.8 lbflir per boiler. This corresponds to 0.024 Ib/MMBtu. Draft Permit at 34, Condition I . This is not BACT for PMIO. The ADEQ must prepare a top-down BACT adysis for total PM 10 and circuIate it for public review. As expIained below, this analysis should conclude that BACT is an emission limit no higher than 0.0 18 1blMMBtu based on n %hour average.

Permits for similar sources that contain Iower total PMl 0 limits than proposed for White Bluff are summarized in Table 6.

Table 6 Permitted TotaI PMlO Emission Limits

(IblMMBtu)

Status: A = application; C = under construction (startup date); Can = cancelled; DP = draft permit; 0 = operating; P = permitted Fuel: B = bituminous; SB = subbituminous; V = various, including bituminous, GOB, wood, distillate oil, and diesel fuel

Attachment 2 106

This table demonstrates that many similar sources firing a wide range of fuels have been permitted with totaI PM I O limits that are lower than proposed for White Bluff. These permitting decisions represent the judgment of nine separate permitting authorities that a tota? PMlO emission Iimit of 0.0 IS 1blMMBtu or lower, based on a 3-hour average, is BACT for 17 simiIar sources. In fact, ADEQ itself permitted a very similar facility, Plum Point, with a BACT total PMlO emission limit of 0.0 18 IblMMBtu. SC WB Ex. 194.

Stack tests for similar sources summarized in Table 7 demonsirate that these limits have been achieved.

Table 7 Total PMlO Stack Test Results

(IbliMMBtu)

The data in Table 7 show that a total PM 10 emission limit of 0.0 1 8 lbMMBtu is routineIy met at a range of coal-fired facilities. One of these units, Hawthorn, has been tested for eight consecutive years and has consistently met its permitted total PM 10 Iimit of 0.0 1 8 1bMMBtu. These data demonstrated that 0.0 18 IblMMBtu is achieved in practice and is achievable for White Bluff as configured.

Attachment 2 107

4. BACT Is Not Required For PM2.5

The Draft Permit concluded that PSD review was not required for PM2.5 as the Project would result in a decrease of PM2.5 of at Ieast 47 tpy. Draft Permit at 2 1. However, as explained above, PSD review is triggered for PM2.5. The Draft Permit and underIying Application do not contain a BACT andysis for PM2.5. Instead, the Draft Permit proffers an enforceable numerical limit of 0.024 Ib/MMBtu, reportedly based on the vendor’s guarantee. Draft Permit at 22 and Condition 26. This limit does not satisfy BACT for PM2.5 for the reasons explained below. The ADEQ must prepare a top-down BACT analysis for total PM2.S and circulate it for public review.

First, a BACT PM2.5 limit has been established at 0.0123 1WMMBtu based on a 3-hr average for PIant Washington, Georgia. SC WB Ex. 201, p. 8 and SC WB Ex. 175. This 850 MW unit will burn similar coal and use a wet scrubber, baghouse, and sorbent injection to control particdate matter. A proper BACT analysis must adopt this h i t as BACT for PM2.5 or expIain why White Bluff cannot meet it.

Second, the Draft Permit aIso sets a PM IO limit of 0.024 IbMMBtu. As PM2.5 is a subset of PM 10, a PM2.5 limit must be smaIler than the PM 10 limit.

Third, the Draft Permits establishes a filterable PM BACT limit of 0.01 2 IbMMBtu. According to AP-42,53% of the particulate matter from a boiIec burning subbituminous or bituminous coal controlIed with a bashouse is 2.5 microns or less. AP-42, Table 1.1-6. Thus, filterable PM2.5 is 0.01 2 x 0.53 or 0.0064 1blMMBtu. All of the condensable fraction of PM I O is PM2.5. As expIained above, BACT for total PM 10 for White Bluff is a limit no higher than 0.01 8 lblMh4Btu. Thus, the condensable fraction is 0.0 1 8 - 0.0 12 = 0.006 IblMMBtu. The PM2.5 BACT limit then would be 0.0064 + 0.0060 or 0.0124 lb/MMBtu.

Fourth, Sargent & Lundy estimated primary PM2.5 emissions based on stack rests as 0.0069 Ib/MMBtu.’W The subject stack tests were not produced. This is presumably the filterable fraction of PM2.5. Based on the Sargent & Lundy total PM I O estimate of 0.030 Ib/MMBtu and filterable PM 10 of 0.0 12 IblMMBtu, the condensable PM2.5 fraction would be 0.020 - 0.012 = 0.008 IblMMBtu. Thus, according to Sargent & Lundy’s calcuIation, the tom1 PM2.5 is 0.0069 + 0.OOX = 0.0149 Ib/MMBtu.

Fifth, controls are available, above and bcyond those required to satisfy filterable PM BACT, that would allow much lower PM2.5 emission rates to be achieved than caIcuIatcd assuming the proposed control train. These include coal selection, coal cleaning, modified baghouse, electrostatic precipitator (“ESP”), (“Wet ESP” or “WESP”), , membrane WESP, Compact Hybrid Particulate Collectors, venturi scrubber, cyclones, advanced hybrid particdate collectors, and an agglomerator.

Sprcadshcct S&L EmissionSummary (DraA_OS2508), tab: Emissions. IS4

Attachment 2

A wet cIectrostatic precipitator, rejected as BACT for S A M , could be placed afier a fabric filter and would eliminate significanf amounts of PM2.5 emissions. SC WB Ex. 2 I 8.lSs The EPA and others have recognized that WESPs reduce PM2.5 emissions.186 Indeed, “the WESP is the ultimate device capable of. . . removing uItrafine particles. Many industries are considering the WESP as the maximum achievable control technology (MACT).” SC WB Ex. 2 18, pp. 6-7. Examples of facilities using WESP technology include: (1) Xcel Energy, Sherburne County, Units 1 and 2; (2) First Energy, Mansfield, Unit 2 (pilot test); (3) Duke Power, Cliffside, Units 6 and 7; (4) AES, Deepwater (operating since 1986); (5) New Brunswick Power, Coleson Cove; and (6) DalIman Unit 4.

In addition to the WESP, other options are available to reduce PM2.5 emissions. For example, the EPA’s Environmental Test Verification (ILETV”) program has verified the performance of the “Advanced Hybrid Particulate CoIlector” (ILAHPC“) system “as providing the lowest filter outlet concentrations for both PM2.5 and total mass on cent ration.''^^^ The AHPC system is instaIled at Otter TaiI Power’s Big Stone plant in South Dakota. AnaIyzing the performance of the system at that plant, the US Department of Energy explained that (SC WB Ex. 2 19”’):

The Advanced Hybridr’*’ consists of alternating electrostatic precipitation and fabric filtration elements in a single casing to achieve exceptional removal of particulate matter (PM) in a compact unit. Very high removal is achieved by removing at least 90 percent ofthe PM before it reaches the fabric filter and using a membrane fabric to collect the partides that reach the frIter surface. . . . Combining precollection with the ESP elements and membrane filter bags results

Albcrt L. Moretti and Ronald J. Triscori. Application of Wet Electrostatic Precipitators to Address Fine I u

Particulate Emission Requirements from Fossil-FueIed Combustors, ICAC Clean Air TechnoIogies and Smtegies Conference. Mmh 7 -10,2005.

Candidate Stationary and Area Control Measures, Chicago PM2.5 Workshop. June 21,2007, Tim Smith, USEPA nt slidc IS (recognizing Wet ESP’s as “innovative PM2.5 controls. available ot Iittp:iienrtli 1 .ens. ~ 0 % ittn/ti:ina~~rn!presenfs/conrrnl I ~ I C ~ S ~ I ~ C S stntionnw and imn-tim sm ith.npt.

EvaIuation of Potential PM2.5 Reductions by Improving Performance of Control Devices: ConcIusions and Recommendntions, Prepared for: U.S. Environmental Protection Agency by E.H. Pechan & Associates. Inc., uvdabie f i r Iitm:l,wiMv.epn.~~nv/17mllneasurrs/oin7f mommend 2007.nd I+ (describing Wct ESP as an “innovntive control systcm” that ‘lield[s] higher PM2.S emissions reductions tlinn the methods identified to improve existing control device perf‘onnancc’’); CIS0 Industrial Emissions Control TechnoIogy 11 Conference. August 2 - 4,2004 Portland, Maim at 6 (explaining that Wet ESP’s nrc an effective control technology for PM2.5: “There are no moving parts in a wct ESP. The [fine] particIcs never rcalIy reach the electrode and arc constantly washed away by the water flow)” niwilable at ~iu~i:!::wwrv.cibo.o~cniissiani~00.1isummarv.ndf.

1%

EPA Tat Program Verifies Performance of GORE@ Filter Laminate (October 2005) nvailulde at httn:/.:’\~vwrv.~ore.c~[~/et~ xx !ne\vd;/mn test nromrn ehJ.htnil.

Demonstration of a Full-Scale Retrofit of lhe Advanced Hybrid Particulzte Collector Technology, U.S. Dcparhncnt of Enerky (February 2007) available at http: ~203.154.137.1 cl/t~hnoloeicslcclnlpower/~ctc/PPI I~~i~t ic~c~.nhvldcmonstnt iunlc~i~ ironmrntal!ntter/P P.4 OtterSon20Tail PPA FinnI’ ,blOToPirlOPu~t~~~.~~f .

18’1

188

109

Attachment 2

in a small, economical unit that can achieve very high coIlection of all particle sizes.

The proposed baghouse itself should be redesigned to improve its capture efficiency for the PM2.5 size fi-action. Fabric filter baghouses are only as efficient as the bags they use. The filtration media determines the control effxiency of a baghouse for very small particles. There is a wide range of media that can be used, most of which are more efficient for larger particles. The filtration media proposed for the White Bluff baghouse is unknown, but would typically be Ryton. The exhaust gas conditions should allow wide Iatitude in selection of filtration media to achieve high removal of flie PM2.5 particles. Media have been developed over the Iast decade that remove over 99.99%+ of the 2.5 micron particles. These indude Daikin's AMIREXT", PTFE membrane filters (SC WB Ex. 202'") and W.L. Gore's L3650 (SC WB Ex. 203).'90 See summary of U.S. EPA's ETV test results in SC WB Ex. 204.

In SUM, the Draft Permit fails to address PM2.5 as a PSD pollutant. The ADEQ cannot issue a final PSD permit for the Project without first conducting PM2.5 modeling to ensure that the plant's PM2.5 emissions will not violate the PM2.5 NAAQS. In addition, ADEQ may not issue the permit without first conducting a PM2.5 BACT analysis and setting BACT-based limits on PM2.5 emissions. We recommend a PM2.5 BACT limit of 0.01 33 lb/MMBtu, based on a filterable PM 10 Iimit of 0.01 0 lblMMBtu and a total PMlO h i t of 0.01 8 lb/MMBtu (0.01 x 0.53 + 0.01 8-0.01).

5. The Permit Must IncIude BACT Limits for COz

As expIained in Section V. above,-PSD will be triggered for COZ and Entergy is required to comply with all applicable PSD requirements, including the obligation to comply with BACT for COz. Thus, a BACT analysis must be conducted For COz and COz BACT limits included in the permit. This analysis should consider at least four types of COr control options which are currently technically feasible: (1) conversion of the facility to an Integrated Gasification Combined Cycle ("IGCC") plant; (2) conversion of the facility to fire 100% natural gas; (3) carbon capture and sequesmtion; and (4) efficiency improverncnts. One or more of these options can be implemented to reduce CO2 emissions.

The facility could be converted to use IGCC technology, which would reduce emissions of a11 criteria polhtants. In addition, it would facilitate the capture of CO?. The syngas produced in coal gasifiers at IGCC plants is at high pressure and has a high C02 concentration, which allows high C02 capture, about 90% using currently availabIe carbon capture technologies. IGCC is a t e c h n i d y feasibk and available control technology now. Currently, there are around 130 gasification plants worldwide -

I** Mcllvainc Hat Topic Hour, Filter Mcdia SeIection for CoatFircd Boilers, September 13.2007, Presentation by Todd Brown, Dnikin America. Inc. Voice recording availabIc online to subscribers of McIIvainc Powcr Plant Knowledge System and available for purchase.

U.S. Environmentnl Protection Agency. ETV Joint Vcrificrltion Statement. Baghouse Filtrtltion Products, W ,L. Gore & Associates, L3650 .at htr~ , : i !cp~.~o~~r~v!~uhsl6 l ) l )e t~~116M~s.~df .

Attachment 2 110

fourteen are IGCC plants, with a capacity of 3,632 megawatts (w of electricity, worth nearIy $8 billion, and using a variety of fuels such as oiI residues, petroleum coke and coaI. Currently, there are over thirty proposed cod-fired power plants in the U.S. using gasification technology.'9' These proposed plants include:

American Electric Power Company's 629 MW Great 3end IGCC plant, Ohio; American Electric Power Company's 629 MW Mountaineer IGCC plant, West Virginia; Duke Energy's 630 MW Edwardsport IGCC pIant, Indiana; Buffalo Energy's 1 100 MW Glenrock IGCC plant, Wyoming; ERORA Group's 630 MW Taylowille Energy Center IGCC plant, Illinois; ERORA Group's 773 Mw Cash Creek IGCC plant, Kentucky; Excelsior Energy's 1200 MW (two GOOMW plants) Mesaba I & II IGCC plants, Minnesota; and Mississippi Power's 600MW Kemper County IGCC plant, Mississippi.

The range of U.S. IGCC proposals includes those using petroIeum coke, bituminous coal, subbituminous coal, and 1ig11ite.I~~

The facility could also be converted to fire up to 100% natural gas, which would reduce C02 emissions by up to about 45%. Data reported to the EIA on Forms EM-860, -767, and -923 indicate that there are over 150 units that can switch between coal and natura1 gas. At White Bluff, this can be accomplished by just converting the boilers to fire 100% natural gas, or by replacing the coal boilers with a gas turbine. Converting just the boiler to 100% gas firing is a mature t echn~ logy . '~~ It requires replacing the pulverized coatfired burners with low NOx natural gas-fired burners; upgrading oil-fired ignitors to natural gas service; revised burner management controls; gas piping; control valve stations; and a pipeline to the nearest source of natural gas, about 20 miles from the faci 1 i ty .

An alternative to retrofitting the boilers to fire natura1 gas is to replace the coal- fired boilers with gas turbines, reusing as much of the existing plant as feasible, e.g., the steam turbine. These types of conversions are sometimes referred to as rcpowering. Several coal-fired boilers have been repowered, including Noblesville in Indiana; the Bayside Power Station in Florida; and several in Minnesota inchding King, Riverside, and High Bridge.

Finally, increasing the efficiency of a coal-fired generating unit by 1% will reduce CO2 eniissions from that unit by 1 %. The total reduction in C02 emissions will be

W.S. Department of Energy, Ti-mkiiig N m Coal-Fired Power Piunts. October 10.3007. I91

19' US. Department of Energy. FossiI Energy Techlinc, T Q ~ Credit Pmgrutiis Promno& Conl-BasedPoaw

Gcncration Technologies. August 14,20015. 193 Coal Pfant Conversion Projects. http:l/~v.source~~~tch.ordindex.php?title=Coalglant_conversionqrojects.

Attachment 2 1 1 1

somewhat larger, since the more efficient unit wiIl be run more, reducing usage of less efficient units. The proposed average heat rate is reported as 1 1, IS9 MBtu/MWhr (net). 119 Ap., Appx. A. This is a relatively high heat rate and indicates good opportunity for efficiency improvements in a COz BACT analysis

The technology exists today to increase the efficiency of an existing coal unit by about 20%. The cost of doing this wouId be high, but wouId be less than the cost of buildin2 a new unit, and would be partly offset by increased capability of the unit and decreased fuel costs. Eficiency increases of 5 to 10% are achievable at considecably lower costs. Efficiency can be improved by upgrading the steam turbine, boiler, and condensers, doing more maintenance than economically optimal, reducing boiler in- leakage, reducing air prcheater leakage, reducing condenser pressure, and enhancing cooling tower performance, among others. A proper BACT analysis would include a thorough review of the White Bluff units to identify all opportunities for eff~ciency improvement.

VIII. TIIE DRAFT PERMIT APPEARS TO IMPROPERLY ALLOW FOR A SIGNIFICANT CHANGE IN THE TYPE OF COAL BURNED AT WHITE BLUFF WTHOUT A PSD REVIEW

The existing White Bluff operating permit has limitations on the sulfur content and ash content of the coal burned. White Bluff Perniit No. 0263-AOP-R6, Section IV, Condition 26, (Ex, 71). SpecificalIy, Condition 26 of the currently effective permit limits sulfur content to 0.72% and ash content to no more than 15-96 Ib/MMBtu unless the terms of the equation given in Condition 26 can be met. The equation essentially allows a higher sulfur coal to be burned at White Bluff if the ash content is lower than 15.96 IbMMBtu and vice versa. With no discussion or indication why, ADEQ has deIeted this condition in the current dmft permit action in the post-BART scenario.

This permit condition was added to the White Bluff Permit in a permit action finalized in 2005. White BIuff Permit $0263-AOP-R4 (Ex. 5) . The 2005 permit action allowed White Bluff to burn subbituminous and bituminous coal and to receive coal via barge. Condition 26 was added to limit sulfur and ash content of the coal, and this was apparently included to keep the White Bluff facility from exceeding a PSD limit.'94 Given that White Bluff does not currently have an existing PSD permit to our knowledge, we interpret the reference to "PSD limit" in the cited email to mean a limit that was imposed to alIow White Bluffto avoid PSD review. 40 C.F.R. $52.21(r)(4) prohibits the relaxation of any limit taken to avoid an otherwise applicable PSD requirement without the source having to meet aI1 PSD permitting requirements as though construction has not yct commenced. Thus, ADEQ cannot remove a limit that was intended to allow White

See February 2 1,1006 Emnil from Ann Sudmeycr to Georgc Johnson, which stated "I intcndcd for the 1%

ash to be less than 15.96 regardless of the outcome ofthc equation. if this ash content i s excecdcd. you would be exceeding other limits like the PSI) limit of 0.04 IblMMBtu. The equation was only mcant to alIow for h i g h sulfur contents if you have a lower ash content." Ex. 22 I . See also EPA letter to ADEQ regarding White Bluff fuel witch dated October 4,2006 (Ex. 841.

112 Attachment 2

Bluff to avoid PSD review in this permit action without requiring White Huff to meet a11 PSD requirements.

In 2006, Entergy apparently wrote to EPA to determine if burning lignite coal at White Bluff would be subject to PSD andlor NSPS requirements. EPA’s response, provided in an October 4,2006 letter to ADEQ (Ex. 84), discussed the limitations on fuel switches being exempted from the definition of major modification, inchding that the source had to be capable of accommodating the h e 1 before I975 and that the use of the alternative fuel must not be prohibited by any federaIly enforceable permit condition. EPA elaborated that such permit conditions could include a change in emissions that could “distort a prior assessment of a source’s air quality assessment,” and EPA informed ADEQ that it needed to review Entergy’s existing permits to see if any term or condition could exclude lignite as a possible fuel. Id. at 2. EPA also said ADEQ needed to review the SIP to see if it included any requirements that could limit Entergy from burning lignite.

Given that ADEQ added Condition 26 during the context of a permit change to alIow for a change in coal (from subbituminous coal from northeast Wyoming to subbituminous and bituminous coal), it seems very likely that it was imposed to ensure White Bluffs fuel switch did not distort a prior air modeling anaIysis, did not violate another applicable requirement, and/or did not result in a significant actual emissions increase of SO2 or PMIPMZO. ADEQ must disclose the basis of the existing Condition 26 and must justify how the condition can be legitimately removed under the requirements of the Arkansas SIP and the Clean Air Act.

It appears to Sierra CIub that ADEQ is removing this Condition 26 to allow Entergy the fIexibiIity burn higher sulfur and/or higher ash coal without triggering a PSD review. We are aware that Entergy has again approached ADEQ about burning lignite at White Bluff, but ADEQ informed Entergy to delay that action untiI this permit is processed. February 18,2009 email from Thomas Rheaume to Mike Bates in the Entergy - ADEQ internal emails file. ADEQ cannot remove this h i t to pave the way far another permit change to burn dirtier fuel at White Bluff.

As discussed above in the BART comments, Entergy relied on assumptions for worst case coal than were not alIowed in Condition 26 of its existing permit. So it is clear the company is intending to burn higher sulfur coal. If so, such change is a change in the method of operation that cannot be permitted in this action without an evaluation of whether the change triggers applicability of PSD. This change in fuel could not have been accommodated prior to the Project as the existing ESPs Iimit the type of cod that can be burned due to the impact of changes in ash resistivity on ESP performance. The new baghouse are apparently the lynchpin of Entergy’s quest for fueI flexibility.

And, as we described in our comments in Section IT., Entergy cannot take credit for any emission reductions occurring in Step 1 of the applicability analysis. Only emission increases are taken into account in determining whether a modification such as n switch to a higher sulfur andlor higher ash content coal than currently allowed in the

113

Attachment 2

permit would have a significant emissions increase. If so, then all contemporaneous and creditable emission increases and decreases can be taken into account in determining whether a net emissions increase will occur. However, as we stated above in Section II., Entergy cannot take credit for the BART SLP emission limitations in a netting analysis.

ADEQ cannot delete the limitations of Condition 26 without 1) requiring White Bluff to undergo PSD review as though construction had not yet commenced, if Condition 26 was included to avoid an othenvise appIicable PSD requirement; and 2) requiring a PSD appIicability analysis for the change in fuel to a higher sulfur andlor higher ash content coal.

IX. ADEQ FAILED TO MQUIRE ENTERGY TO MEET OTHER PSD MQUIREMENTS FOR S02, NOX, PM2.5, AND PM10.

As discussed above, the projected being authorized in this permit would result in a major modification of S02, NOx, PM2.5, PM 10, and CO2 among other pollutants. Yet, ADEQ failed to require Entergy to address all other PSD requirements for these polhtants. Such requirements include preconstruction monitoring, air quality analyses to determine whether White Bluff wouId cause or contribute to a violation of the NAAQS or PSD increments for these pollutants, and analyses of impacts on air quality related values of nearby Class I areas, among other requirements of 40 C.F.R. 52.2 I . Therefore, the draft permit is invalid because Entergy has not demonstrated compliance with all applicable PSD permitting requirements.

X. THE PERMIT LIMITS ARE NOT ENFORCEABLE

A. The Draft Permit Does Not Require Continuous Compliance with Permit Limits

Emission limits must be met on a continual basis at all levels of operation. Section 302(k) of the CIean Air Act expressly defines the term “emission limitation” as a limitation on emissions of air pollutants “on a continuous basis.” Section 169(3) of the Clean Air Act, in turn, defines BACT as an “emission limitation.” Accordingly, the Clean Air Act mandates that BACT continuously limit emissions of air pollutants. Thus, the permit must require sufficient monitoring and recordkeeping to assure that the BACT limits are met continuoudy. It does not.

Compliance with the BACT limits for filterable PM, lead, VOCs, and S A M emissions is determined by an annual stack test. Draft Permit, Condition 5. Compliance with certain other emission limits -- PM2.5, total PM 1 0, HC1, and IIF -- is determined by stack testing every five years. Draft Permit, Conditions 27-29. Finally, no testing or recordkeeping at all is required to determine compliance with the emission limits for 40 hazardous air pollutants (“HAPS”). Draft Permit, Conditions 2 and 13. These HAPS emission limits are not enforceable.

Attachment 2 114

A stack test conducted annually or once every five years is not suficient to show that the BACT and other emission limits are being met day-in and day-out on the coal- fired boilers as emissions are highIy variable from hour to Stack tests do not capture the variability in operating conditions, such as fluctuations in load, flue gas flow rates, temperatures, pressures, burners in and out of service, fuel characteristics, soot blowing, startups, shutdowns, ctc.

Further, the U.S. EPA, for example, stated in a Federal Register preamble: “[mJanual stack tests are generalIy performed under optimum operating conditions, and as such, do not reflect the fuI1-time emission conditions fiom a source.”196 A well-known textbook confirms, stating: “Manual source tests also require a degree of preparation, and the coordination and the prior schedule of a test may result in source operations being highIy tuned before such testing takes place. The manual test results, therefore, may not necessarily be representative of day-to-day erni~sions.~~.’~’ Further, some consent decrees prohibit pre-test maintenance due to the prevalence of this practice. Thus, periodic monitoring by stack testing does not nccessady represent emissions that occur day-in and day-out, under all operating conditions, as required to show compliance.

The monitoring and testing provisions of a permit must ensure the underlying performance standards are being met. If they do not do so, then they cssentially weaken the performance standards. Generally, the hierarchy for specifying monitoring to determine compliance is: ( I ) continuous direct measurement where feasible; (2) initiaI and periodic direct measurement where continuous monitoring is not feasible; (3) use of indirect monitoring, e.g., surrogate monitoring, where direct monitoring is not feasible; and (4) equipment and work practice standards where direct and indirect monitoring are not feasibIe. NSR Manual, p. i.3. Thus, we recommend that CEMS be required where feasible and where not, that a valid surrogacy approach be employcd.

B. Startups, Shutdowns, Maintenance, and Malfunctions Arc Excluded From the Monitoring and Recordkeeping Provisions

Stack tests are nomaIly conducted during full load operation. In fact, the Draft Permit requires testing for PM2.5, filterable PM, totaI PM IO, hydrogen chloride, and hydrogen fluoride to be conducted while operating at 90% or greater capacity. Draft Permit, Conditions 27-29. Emissions can be substantially higher during other operating conditions, such as startups, sliutdowns, atomizer change-out, soot-blowing and malfunctions. During startups, for example, the baghouse may not be operational, resulting in very high total PM I 0, PM2.5 and filterable PM emissions, or combustion would be incomplete, resulting in very high VOC emissions. These emissions wouId not be captured by an annual stack tests during full load operation. Thus, annual stack testing does not assure compliance under all operating conditions. The Draft Permit should be

In re: Ra~~e~~swoorlS~eut~r Ptmt. 11-7001 -08 (EFA Admin. Sept. 30,2003) at 20 (stack test once every 195

pennit term is not sufficient to ensure compliance with PM emission limit for coal fircd boiler). I q b 40 Fed. Rcg. 46,240.46J4 I , (Oct. 6, 1973,

”’ James A. Jahnkc, Continuous Emission Monitorin?, John Wiley & Sons, 7007. p. 1.

Attachment 2 115

modified to require continuous testing during all modes of operation, where feasible and othenvise, surrogates.

C. A ParticuIate Matter CEMS Should Be Uscd To Determine Compliance with the FiltcrabIe PM BACT Limit

Continuous emissions monitoring systems are the preferred method for determining compliance with PM limits. 40 C.F.R 85 60.42, et seq. The US. EPA has in practicality required PM CEMs, based on its determination that PM CEMS are reIiabIe and accurate. In comments submitted on the Robinson Power Company PSD Application and Draft Plan Approval, for a proposed 270 megawatt waste coal-fired, circulating fluidized bed (I‘CFB”) boiler facility at Robinson Township, Pennsylvania, about 12 miIes west of Pittsburg, the U.S. EPA commented that:

“The proposed plan approval requires annual stack testing to assure compliance with the particulate matter emission limits from the CFB and its associated fabric-filter baghouse. In light of the evolution of CEMS systems for particulate matter, EPA is strongly urging the requirement to install and operate a particulate matter CEMS at the proposed facility. Currently, there are several facilities that operate PM CEMS and have demonstrated that the systems are reliable and accurate. These are Tampa Electric power plant (Florida), Eli LilIy Corporation (Indiana), and the U.S. Department of Energy (Tennessee). EPA has also secured commitments from up to 30 existing coal-fired utility installations to install PM CEMS over the next couple of years. It is fair to assume that the state of technology for PM CEMS will be even further evolved by the time the proposed Robinson Power facility begins operation. Further, the facility will be required to establish a compliance assurance monitoring plan (CAM) as part of its title V operating permit and the federal CAM regulations strongly encourage reliance on continuous monitoring systems as a means for assuring compIiance.”

Filterable particulate matter CEMS have a Iong and successful history. Continuous monitoring of PM started in Germany in the 1960s and was a requirement by the 1970s. Other European countries and Canada have been using PM CEMS for many different sources including power plants, incinerators, pulp and paper miIls, and cement kilns. PM CEMS are in use in the United States, where they have been required by the US. EPA and state agencies as part of over 30 consent decrees, including with Tampa Electric Company, Virginia Electric Power Company, Wisconsin EIectric Power Company, and many others.’9s SC WB Ex. 205.

They have been specified to determine compliance in permits for many new coal-fired power plants, e.g., Desert Rock (SC WB Ex. I2G), Iatan (SC WB Ex. 108),

Robynn Andncsek and others, PM CEMS: The Current Reality of Monitoring Particulate Matter, I V#

Powefien 2006.

Attachment 2

City of Springfield UtiIities (SC W 3 Ex. 1 Ol), and Longview. PM CEMS are now routinely recommended by the US. EPA for all new fossil-fuel-fired power plants, as evidenced by the U.S. EPA's recent comments on draft air permits. See, e.g., SC WB Ex. 206 and 207199 Further, they have been vohntarily selected by utilities that operate wet scrubbers, e-g., Harding Street and Bruce Mansfield. Published studies indicate PM CEMS are accurate and compIy with performance standards. SC WB Ex. 208.'00 Thus, the White Bluff permit shouId be modified to require the use ofa PM CEMS to determine continuous compliance with the White Bluff filterable PM BACT limit.

D. A CEMS Should Be Used To Determine Compliance with The VOC BACT Limit

CompIiancc with the VOC BACT limit is determined annually using EPA Method 25 or 25A. Draft Permit, Condition 5. Performance standard and conforming CEMS are available for volatile organic compounds are avnilabIe.'O' They are required, for example, in the draft Portland cement manufacturing rulemaking to comply with MACT standards"' and for compliance in other industries. SC WB Ex. 209.2"3 We recommend that a VOC CEMS be used to determine compliance with the White Bluff VOC BACT limit.

E. A CEMS ShouId Be Used To Assure That Emissions of HCI and HF Do Not Trigger MACT

Entcrgy claims that the Project is not subject to I 12(g) MACT requirements, but does not explain why. The emission calculations for HCl and HF, the major HAPS that wouId be emitted by White Bluff, suggest that current actual to expected hture actua1 emissions would decrease. Spreadsheet "Appendix A - White Bluff Net Emission Changes. However, these calculations are wholly unsupported in the record. They are based on the 200612007 Point Source Inventory, for which there is no support, and expected future potential emissions based on the Applicant's spreadsheet, again for which there is no support. It is not believable that a Project that increases the firing rate and coal throughput and allows it switch from Iow chlorine and fluorine PRB coal to higher chlorine and fluorine bituminous coals would not increase actual emissions of HCl and HF. As the Permit requires only one stack test every five years for KC1 and HF (Draft Permit, Condition 29), these claims are unenforceable. Absent a federally enforceabIe

U.S. Environmcntal Protection Agency, Comments on Longleaf. GA (4/20/06); U S . Environmental Protection Agency, Comments on White Slallion (4/14/09 Icttcr from Jeff Robinson to TCEQ). 2w Donald Picning, Jason Wilkersan. and Manfred Strombcrg, Compliance with EPA's PS-11 Correlation Rcquircmcnts Using the Sick-FWE200 PM Monitor, no date.

lo' Performance Specificaiion 8. Performance Spccificntions for VolatiIe Organic Compound Continuous Emission Monitoring Systems in Stationary Sources.

'02 74 Fed. Reg. 21.136.21.140/41,21,150/51,21,157 (May 6,1009).

'03 Dale Evarts. A History and Status ofCEMS Applications in USEPA Regulntions, Slides. Dcccnibcr 16. 2002.

IW

117

Attachment 2

limit that restricts emissions of HCI artd HF to less than a de minimis increase, Section 1 12(g) MACT is required for the Project.

CEMS are available for both HCI and HF and are wideIy used in other industries. See, e.g. SC WB Exs. 210 and 21 I?M An HCI CEMS is required, for example, in the US. EPA’s recent draft ruIemaking for cement kilns if a source does not use a limestone wet scrubber. ’05 White Bluff does not use a limestone wet scrubber. While these CEMS have not yet been used on coal-fired power plants in the United States, as case-by-case MACT limits have only recently been imposed on coal-fired power plants, HCl and EIF CEMS are entirely capable of being used on such power plants.

The US. EPA has recently recognized that CEMS are the proper means of measuring compliance with HCI limits. On April 22,2009, EPA recommended that North Carolina require an HCI CEMS to assure that IlCl emissions at Cliffside Unit 6 remain below the MACT applicability threshold. The US. EPA wrote: ”While there are monitoring alternatives to an BCl CEMS, a HCI CEMS is expected to provide the most reliable assurance of compIiance.” SC WB Ex. 2 12.’06 In addition, Florida recently issued a revised draft permit for the Seminole plant that requires the use of HCI and HF CEMS to demonstrate that emissions remain below 9.75 todyr for HCI plus HF combined, which works out to the lowest WCI and HF emissions in any coal pIant permit. SC WB Exs. 213 and 214.

F. The Methods Specified To Determine CompIiance with SAM BACT Limit Are Inaccurate

Compliance with the SAM BACT limit is determined annually using U.S. EPA Method 8 or an undefined ‘%ontrolIed condensate method.” Draft Permit, Condition 5. These stack test methods arc known to yield inaccurate results when used to measure SAM from coal-fired boilers.

Method 8 was designed for use at sulfuric acid plants. SC WB Ex. 215.’07 The test method does not work equally well for a11 sources that emit SAM or sulfur trioxide.”’ The U.S. EPA Emission Measurement Center suggests that Method 8 is not an appropriate test method for certain sources: “It [Method S] should not be used to measure suIhric acid/suIhr trioxide from the following kinds of sources: (1 ) Those sources that have significant emissions of solid sulfates that are water soluble. Solid sulfates are compounds like sodium sulfate. (2) Those soiirces that have sigiii3cmt

Offcrings by Gasmet and Thermo Scientific. 1al

m 73 Fed. Reg. 21,13G, 21.14I-21,157 (May 6,2009).

’Oa Letter from A. Stanley Meiburg, Acting Regional Administmtor, EPA Region 4. to Dee Frecman, Secretary, North Carolina Department of Environment and Natural Resources, ApriI 30,2009. lo’ Paul S. Farber and DnnieI L. Mnrmcr. Condcnsiblc Particulate Matter Emission Sources and Control in Coal-Fired Power PIants. PowcrGcn 2005.

Wcbsite for EPA Tcchnology Transfer Network Emission Measurement Center - Method S Sulfuric Acid Mist (tlt~p:i ~wvw.epn. coviitnienic!nirt had?imetIind): .html) Website last accessed November 17,2007.

1 IS Attachment 2

airissiorrs of srr&w dioxide mid nnmronia. r30’ (etiiplmsis added) Because coal fired power plants have significant emissions of sulfur dioxide, Method 8 is not an appropriate test method for the proposed facility. Other studies have identified many sources of error when Method 8 is used to measure S A M from coal plants. SC WB Ex. 216.”’

The other SAM test method, the controlled condensation method, was investigated by the EIectric Power Research Institute (I‘EPRI”) by comparing laboratory data with field data for a variety of coals with a range of sulfur and fly ash compositions. This study demonstrated a significant underestimate of 40% for PRB coals and 20-25% for higher s u h r coals due to reactions between the ash collected on the quartz filter and SO1 and SO3. SC WB Ex. 100, pp. 757-758.

As a result, an alternative method has been devcloped and is widely used within the utility industry to measure SAM emissions from coal-fired boilers. This method, referred to as the modified controkd condensate method, uses a tubular electrostatic prccipifator to remove ash particles to the side wall away from the gas stream. This method has been demonstrated to produce accurate and reproducible SAM measurements at coal-fired boilers. SC WB Ex. 217.”’ We recommend that this method be used on at least a quarterly basis to determine compliance with the SAM BACT limit. Alternatively, continuous emission monitoring systems for SAM have been devdoped by EPRI and proven in fulI scale tests.”’ A CEMS is always prcferabIe to a periodic stack test.

G. Coal Sampling Should Bc Required To Assure CompIiance with The Lead BACT Emission Limit

The Draft Permit includes a BACT limit for lead emissions. Draft Permit, Condition 4. Compliance is determined annually by measuring lead in the stack gases. Draft Permit Condition 5. However, lead originates in the coal and can be highly variable, especially given that Entergy is increasing the universe of coals that it will be a b b to draw from, with installation of a baghouse. Annual testing is not adequate to assure continuous conipIiancc with the lead BACT emission limit.

A permit Iimit should be established for lead in coal, calculated from the maximum amount of lead that can be present and stilI compIy wit lead BACT at the stack. Compliance with the lead BACT emission limit should be determined by analyzing coal samples. The samples should be measured continuousIy using the multi-

id m

m Scott Evans, Demonstrating Compliance with Sub-ppm Acid Mist Limils: Can Method 8 Handle the Chalbng? Electric Power Conference & Exhibition, May 2,2007.

”’ EON, SOJ Testing at Ghcnt Units 1.3 & 4, November 15,2009.

”* EPRI and IMACC. FUR Monitoring of NOS, SOX. SO3, & H2S04. Environmental Controls Conference. May 16-18.2006; Robert Spellicy. Richard I-Iimes, and John Pisano. Real-time Monitoring of SOJH$~OJNHJ in SCR Outputs. Proceedings of the 2006 Environmental Controls Conference.

Attachment 2 119

metal CEMS’I3 or coIlected daily, composited and analyzed weekly. Coal samples should be collected from feeder belts prior to the pulverizers. An annual stack test using Method 29 should also be conducted to confirm that coal testing adequatdy Iimits lead emissions.

H. The NOx and SO2 BART Limits Are Unenforceabk

The permit lacks a clear description of how compliance with the SO2 and NOx BART Iimits that only apply when the units burn subbituminous coal will be determined. While there is a requirement in the permit that Entergy keep records of how much coal is burned to show compIiance with the total tonnage limit on coal burned in a year in Section VI, Condition 16 of the draft permit, there appears to be no recordkeeping or reporting requirement governing the information necessary to determine compIiance with the 0.15 IblMMBtu NOx and SO2 limits that apply on a 30 day rolling average basis when the units burn only subbituminous coal. Thus, the proposed SO2 and NOx limits of 0.15 lb/MMBtu each that apply only when subbituminous coal is burned are unenforceable without proper recordkeeping and without a proper description in the permit detailing how compliance is to be determined.

XI. ENTERGY’S WHITE BLUFF PLANT HAS A LONG HISTORY OF NON- COMPLIANCE WITH NSPS OPACITY STANDARDS WHICH MANDATE THE INCLUSION OF A COMPLIANCE SCHEDULE IN THE TITLE V PERMIT

Because Enterg’s White Bluff Plant commenced construction after August 17, I 97 1 , it is currently subject to NSPS Subpart D, 40 C.F.R. 8 60:14 and the 20% opacity limitation set forth at 40 C.F.R. 0 60.42(a)(2). Entergy has a long and extensive history of violating these standards. It appears that White Bluffs ekctrostatic precipitators (ESPs) are incapable of maintaining opacity levels below the applicabIe 20% opacity limit. This is likely due to the ESPs being undersized, other design flaws or deterioration of these controls over time, which are likely at the end of their useful life. Because of White Bluffs history of opacity violations, the Title V permit was required to include a compIiance schedule designed to bring the facility into compliance with 40 C.F.R. 60.42(a)(2) and is invalid without it. Moreover, ADEQ should explain why White Bluffs ESP’s, which have repeatedly failed to maintain compliance with the applicable 20% opacity are expected to do so in the future without substantial modifications to them and an expansion in their size.

It’ Cooper Environmental Services, LLC, Xactn‘ 640 Multi-Metal Continuous Emission Monitoring System (MM-CEMS).

’’‘ This conclusion assumes that White Bluff has not undergone an NSPS “modification” that would have triggered a more rccentIy promuIgated NSPS standnd. It is possible that one or both of l t e White Bluff units have undergone NSPS modifications as a conscqucncc of the 2006 or 2007 economizer replacements (or other modifications which may not yet have been idcntified and thoroughly evaluated to determine NSPS applicability) and thcreby triggered the application of n more stringent NSPS standard.

120

Attachment 2

M I . THE AGENCY SHOULD CONSIDER RETROFITTING OR REPOWERING THE BOLERS TO FIW NATURAL GAS AS AN ALTERNATIVE TO LOWER CRITERIA POLLUTANT AND HAZARDOUS AIR POLLUTANT EMISSIONS AND REDUCE OTHER ENVIRONMENTAL IMPACTS

Pursuant the Clean Air Act Section 165(a)(2), 42 U.S.C. 8 7475(a)(2), Sierra Club requests ADEQ and Entergy to fully evahate the alternative of retrofitting or repowering the boilers at the White Bluff facility to fire natural gas as a means of reducing criteria pollutant and hazardous air poIlutant emissions and to reduce other negative environmental impacts. Retrofitting or repowering the existing coal-fired boilers for firing natural gas are available demonstrated technologies with significant environmental benefits including considembIy fewer emissions of criteria and hazardous air pollutants and fewer by-products. However, the Draft Permit does not contain an analysis of retrofitting or repowering the existing cod-fired boilers to fire natural gas.

A. Retrofit and Repowering for Natural Gas Are Available and Demonstrated TcclinoIogies That WouId Reduce Criteria Pollutant and Hazardous Air Pollutant Emissions

The White Bluff facility could be converted to fire 100% natural gas instead of coal. This can be accomplished by converting the boilers to fire 100% natura1 gas, or by replacing the coal boilers with natural gas-fired turbines. In comparison to coal, firing natural gas would result in considerably lower emission levels of NOx and particuIate matter and produces virtually no SO2 or mercury emissions. Converting coal-fired boilers to firing 100% natural gas is a mature technology and has been demonstrated in pra~tice.”~ There are currently over 150 coatfired boiler with the capability to fire 100% natural gas. EIA Forms EIA-860,7EiA-767, and EIA-923. Converting the boiler to gas firing would require replacing the pulverized coal-fired burners with Iow NOx natura1 gas-fired burners; upgrading oil-fired ignitors to natural gas sentice; revised burner management controls; gas piping; control valve stations; and a pipeline to the nearest source of natural gas (here, the nearest pipeline is located about 20 miles from the White Bluff facility). The cost to convert the boilers to natural gas firing and build supporting infrastructure is substantially less than the cost to retrofit the proposed BART controls on the existing cod-fired units. Maintenance costs would also be substantially lower as gas is clean burning, substantiaIly reducing fouling and corrosion of the boiler and eliminating costs of operating the scrubber and for material handling.

An alternative to retrofitting the existing coal-fired boilers to fire natural gas is replace the boilers with gas-fired turbines, reusing as much of thc existing pIant as feasible, cg., the steam turbine. These types of conversions, sometimes referred to as repowering, generally involve combined cycles in which heat from the turbine is recovered in a heat recovery steam generator. Thus, they have a much lower heat rate than an equivalent coal-fired or naturaI-gas-fired boiler. Several previoudy coal-fired

to

Attachment 2 I21

boilers have been repowered, inchding Noblesville in Indiana; the Bayside Power Station in Florida; and seved facilities in Minnesota inchding King, Riverside, and High Bridge.

Firing natural gas, either retrofitting or repowering the boilers, would aIso eliminate the fugitive dust particulate emissions associated with material handling of coal and the associated by-products such as bottom ash, fly ash and FGD waste. Fugitive emissions from material handling have been estimated by the AppIicant at 398 tpy of PM and 18Xtpy0fPMlO.”’~

€3. Additional Environmental Benefits of Retrofitting or Repowering for Natural Gas-Firing Compared to Cod-Firing

Natural gas-fired facilities create considerabIy fewer waste products that need to be disposed. Coal-fired electric generating units generate solid wastes - gypsum, bottom ash, and fly ash - that are typically disposed in landfills, posing a threat to water resources. In addition, because natural gas-fired boilers do not need the scrubbers required by coal-fired power plants to reduce SO2 emissions, they do not create scrubber by-products that need to be landfilled. Here, the Applicant proposes to expand the existing Iandfill at the White Bluff facility to accommodate the by-products associated with the proposed dry FGD that wouId be installed as part of the EnvironmentaI Controls Project. Entergy estimatcs that approximately 17,250 l b s h of FGD waste and 1,400 Ibslhr of fly ash per unit for a totd of 53,300 I b h of waste products for both units wiIl be disposed of at the expanded landfill after modification of the fa~ility.2’~ This landfdl expansion and the associated poIIutant emissions and other environmentd impacts would be unnecessary if the facility were converted to fire natural gas instead of coal.

Because of the negative environmental impacts associated with the proposed project and the significant environmental benefits associated with convcrting the White Huff facility to natura1 gas, this option is clearly preferable to the proposed permitting action.

XIn. ENTERGY’S WHITE BLUFF PLANT IS SUBJECT TO CLEAN AIR ACT SECTION 112(G) MACT REQUIREMENTS AND THE DRAFT OPERATING PERMIT IS INVALID BECAUSE MACT LIMITS HAVE NOT BEEN INCLUDED IN PERiiIT

CIean Air Act Section I I?(&, 42 U.S.C. 8 7412(g)(Z)(A), provides that:

(2) Construction, reconstruction and modifications

See Exccl workbook ‘‘Appendix A - Whitc Bluff Net Emission Incrcascs.”

”’ Entergy, Response to Sierra CIub Data Request # SC 5-7. November 1 1,2009, THll43. (Estimate of tatat wil~fe product to be disposed of includes water.)

122

Attachment 2

(A) After the effective date of a permit program under subchapter V of this chapter in any State, no person may niodifjl a mujor soiirce of hazardoiis airpollutants in such State, unIess the Administcator (or the State) determines that the tttadnzunt nclzicvablc control technology emission hi ta t ion under t1zi.r section for existing sources will be met. Such determination shall be made on a case-by- case basis where no appIicable emissions limitations have been established by the Administrator.

(emphasis added).

On December 14,2000, the EPA announced that it was adding coal- and oil-fired power plants to Clean Air Act Section f I2(c)’s list of sources. 65 Fed. Reg. 79825 (Dec. 20,2000). ThereaRer, each electric utility steam generating unit was subject to the case-by-case provisions of the Act until EPA promulgated a nationally applicabIe MACT standard to address hazardous air pollutants for this source category.

In 2005, EPA issued two reguIations relating to hazardous air poIlutant (HAPS) emissions from coal-fired power plants: (1) a rule that removed such power plants from the list of industries requiring the Clean Air Act’s rigorous “Maximum Achievable Control Technology” (MACT) standards for each electric generation unit in the country to sharply reduce its toxic air pollution; and (2) a regulation that substituted a mcrcury pollution trading regime, which weakened required mercury cuts from power plants, dispensed with the need to reduce mercury from each electric generation unit in the country, and walked away from reglating all other forms of toxic air pollution from power plants. However, on February 8,2008, the D.C. Circuit vacated EPA’s rule removing power pIaIits from the Clean Air Act list of sources of hazardous air pollutants and vacated the Clean Air Mercury RuIe. New Jersej? v. EPA, 517 F.3d 574,578 (D.C. Cir. 2008). This step reinstated the obligation for all new coal-fired power plants to include MACT limits covering a11 of their hazardous air pollution because the court’s ruling returned power plants to the regulatory list requiring adoption of MACT standards. 40 C.F.R. 63.42(~)(2). Therefore, for the reasoils set out below, Entergy’s White Bluff Plant is subject Section 1 12(g) as a consequence of the proposed “modification” contemplatcd by this permitting action and must undergo a case-by-case MACT analysis which establish MACT emission limits for all applicable H A P S to be emitted by the plant.

Entergy’s White Btuff Plant and each White Bluff electric utility steam generating unit is a “major source” of HAPS because each “emits or has the potential to emit considering controls, in the aggregate, I O tons per year or more of any combination of Iiazardous air pollutant or 25 tons per year or more of any combination of hazardous air pollutants.” 42 U.S.C. 5 74 12 (a)( 1 ). Specifically, White Bluff Unit 1 ’s emissions of hydrogen chloride (HCI) and hydrogen fluoride (I-IF) are 2,125 tpy and 265 tpy, respectively, and White Bluff Unit 2’s emissions of HCl and H F are 1,773 tpy and 222

Attachment 2 123

tpy, respectively.”’ These pollutants are both listed as HAPs at 42 U.S.C. $7412(b)(l). 42 U.S.C. 7412(a)(6). Since each unit emits more than IO tpy of each of these HAPs, each unit is a major source of HAPs.

The proposed permitting action, which involves, inter alia, an increase in heat input of 250 MMBWr, constitutes a “modification,” because this project covered by the proposed permit constitutes a “physical change in, or change in the method of operation of, a major source which increases the actual emissions” of HAPS above de minimis levels. 42 U.S.C. 8 7412(a)(5).

While Entergy provided an analysis of future actuaI HCI and HF emissions that projected a decrease in actuaI emissions of these pollutants (in Appendix A of the August 2009 White Bluff Permit Application Revisions), Sierra Club questions the validity of those emission projections. As discussed in the above section OR enforceabihty, Entergy’s calculations are wholly unsupported in the record. The current emissions are based OR the 2006/2007 Point Source Inventory, for which there is no underlying documentation provided in the record. And expected future potential emissions are based on the Applicant’s spreadsheet, which again lacks any support in the permit record.

Moreover, despite Entergy’s projections of emission decreases in HCI, ADEQ’s draft permit allows for a significant increase in HCl emissions at each White Bluff unit. Specifically, the currently effective White Bluff permit Iimits HCI emissions at each White Bluff boiler to 2,759.4 tpy. Section IVY Condition 2 of White Bluff Permit No. 0263-AOP-R6 (Ex. 71). The draft revised permit alhws emissions of HCI to increase to 2,839.0 tpy. Section Tv, Condition 2 of the post-BART scenario. This is a 79.6 tpy increase of HC1 at each White Bluff unit, over seven times the major source emissions threshold of 10 tpy for a single HAP.

In addition, the draft permit post-BART sccnario allows for increases in numerous other HAPS compared to the allowabIc emissions ofthese HAPs in the currently effective permit, including arsenic, benzene, cyanide, and selenium, among others.

For the reasons discussed above, the modifications being permitted at each White BIuff unit in this draR permit arc modifications to an existing major source of HAPs, which triggers the application of Clean Air Act’s Section I 12(g) MACT standards to each White Bluff unit. And as a consequence, Entergy’s White Bluff units must undergo case-by-case MACT determinations which establish MACT limitations for a11 applicable HAPS emitted by those modified units. Until ADEQ evaluates and proposes MACT emission limitations appIicable to the HAPs emitted from the modified White Bluff units, the proposed White Bluff operating permit is deficient for failing to include all appIicabIe requirements incIuding emission limitations for each HAP reflective of MACT.

‘Is August 5.2009 Entergy Supplement to White BIuff Permit Application, Appcndix A, Table Entitled “Net Emissions Changes Summary - Design Target.”

Attachment 2 I 24

XIV. The SO2 W e t " Mudulirg Is Tccl~aia~lly and Legally Unjusrified

In its niiibicnI nir nrralysis fbr rhc Wliiie Blull'pemiit. Eiirergy inodcted SO2 bascd on "fitturr maximum potential emissions." Jaiiunry 2009 Wtiitc Blu1'1'Permit Application at 6-39. A review of the 1blhr emission rites modeled shows that they were indeed based on fiiiure potenritil etnissions - ix., based on the 0. I 5 IblMMBtu SO2 and NQx limits of the pcnnii intilliplied by the requested permitted heat input capacity 01'8950 MMBtdhr. And the inodeling or these requested allowbte emission mtes showed SO2 impucts above the modeling signiticance levels. Id.

According to Sccliori 6.6 o f its permit applicalion, Entcrgy then performcd a "nct'' modeling anntysis to take into account thc eiiiissioii reductions xsocintcd with the project. This makes absolutely 110 teclinical seiise siiicc Entcrgy iiiodcled tlic ncw allowable emission rates that reflect the reduced SO:! and NOS emission rates that the cornpniiy has indicated meet BART rrquircmen~s. By modeling the reduced SO2 emission rates, it Ira already essentially token credit for the SO2 emission reductions. Such net inodeling is legally inappropriate. 40 C.F.R. 52.21(1) requires all estimates of air quality impacts be based on the EPA's Guidelines on Air Quality Models in 40 C.F.R. Part j I , Appendis W. EPA's tnodcling guidelines spcciQ that, for NAAQS compliance modcling. pcnnit applicants ~iiust modc1 tlic itinsirnuin allownble emission rate iiiuItipticd by the dcsign capacity or die unit assuming continual operation ttirotiyhour the year. See 40 C.F.R. Pnrf 5 I. Appciidis W, Tablc 8-2. When Eiitergy did such modcling. it showcd that tlic White B Iu ff'ficility's itripacls would g~nrJy escced the 3-hour nnd 24-hour nvcrnge SO2 significance levefs. See Jnnuary 2009 Whitc Bluff Permit Application. Table 6-15 at 6-30. Conscqticutly, Entcrgy mtisl bc requircd to do R cumulntive SO2 NAAQS madding aiinlysis to dctenii ine wltcthcr Whitc I3 lir fT wotrld caiise or contribute to an SO2 NAAQS violation.

'I'he drali permit must not be issued until there is a11 adcqiintc demonstration in coinpliunce with EPA's rriodclirig guidelines ilia1 demonstrates While BIulTwon't cause or contributr to EI violation d t l w SO2 NAAQS.

TIiunk yoti h r considering tltesu comments.

Sincerely. ,,y

Wi in111 1. Moorc. I l l

Esliibits Enclosed:

Attachment 2

& Waste Management Association,

12

I3

14

America Successfully Goes Operational Chiyoda Licenses Its FIue Gas Desulfurization TechnoIogy in USA Newly for 5 Cod-Fired Generation Units, Press Release, May 2,2005 Chiyoda Licenses its Flue Gas Desdfurization Process in USA for Georgia Power Owiied 4 FGD Units, January 26,2005 Jonas S. Klingspor, Kiyoshi Okazoe, Tetsu Ushiku, and George Munson, High Efficiency

15

DoubIe Contact Ffow Scrubber for the US. FGD Market, Paper No. 135 presented at MEGA

Yoshio Nakayama, Tetsu Ushiku, and Takeo Shinoda, Comiiiercid Experience and Acttial- j

- - I Organization r c ~ ~ ~ ’ ) presentation

16 17

1

Attachment 2

Plant-Scale Test Facilily of MHI Single Tower FGD Mitsubishi High SO2 Removal Experience Lake Michigan Air Directors Consortium (‘ZADCO”) and the Midwest Regional Planning

2

Attachment 2

37

35

Revised Draft Regional Haze State Implementation Plan, July 2009, Minnesota Air Pollution ControI Agency j

40

3

Business Wire, September 30, 1999. D.W. Bullock, Long-term SCR Operating Experience at PG&E Generating’s Coal-Fueled

Attach men t 2

41

42

43

44

45

46

47

48

49 50

Plants, ICAC Forurn 2000, March 2000 . Memorandum from Greg Worley to Brian Beals, U.S. EPA, Re: Review of the Cypress Energy Project PSD Application, September 25, 1992, EPA4PERO54453 Wisconsin Power & Light Co., Certificate of Authority Application, Edgewater Generating Station Unit 5 NOX Reduction Project, November 200s Letter from Arizom Department of Environmental Quality to Steve Fry, EPA Region IX, Re: Consultation Regarding Best Available Retrofit Technology AnaIyses for the Four Corners Power Plant and Navajo Genemting Station, May 12,200s March 27, 2009 Direct Testimoiiy of Anthony P. Walz, on behalf of Entergy Arkansas before the Arkansas Public Service Commission (Dockct No. 09-024-U). Anthony C. Favde et al., Applicalioii and Operating Results of Low SO2 to SO3 Conversion Rate Catalyst for DeNOx AppIication at AEP Gavin Unit 1, Proceedings of the 2006 Environmental Controls Conference, U.S. Deparhient OF Energy, National Energy Technology Laboratory Keiichuro Kai et al., SCR Catalyst with High Mercury Oxidation and Low SO2 to SO3 Conversion, Paper #E6 R.K.Scivastava et nl., Emissions of SuIfur-Trioxide from Coal-Fired Power Plants, J. Air d Wmte MnnnEe, Assoc., 54: 750-762 K.S. Kumar et al., Wet ESP for controlhg sulhric acid phme following an SCR system, presented at the 2002 ICAC Forum Lesiuk, J.F., Steam Turbine Uprates, GE Power Systems, Atlanta, GA May 23, 2000 letter from EPA to Henry NickeI regarding a turbine upgrade at Detroit Edison’s Monroe power plant

I 51 I Dreier, Jr., D.W., UpgradabIc Opportunities for Steam Turbines, GE Power Systems, 1 52 53

Alstom Power Brochure, “Steam Turbine Retrofit, Add Life, Add power July 3 1,2006 letter from Entergy to ADEQ regarding the economizer replacement at White

New Source Review Workshop Manual 5s Intentionally Omitted

4

Attachment 2

59 I Respoiise of Entergy-Arkansas, Inc., to Sierra Club’s Fourth Set of Data Requests, Response to

60 Rcqucst 4-1.e. (Docket No. 09-024-U, White Bluff Declaratory Order) February 25,2008 Letter from Entergy’s M. BowIes to ADEQ’s T. Rheaume Regarding Unit I

Requirements to the WEPCO Powcr Company Port Washington Life Extension

5

Attachment 2

77

78

79

80

I Bielawski, G.T., J.B. Rogan, and D.K. McDonald, How Low Can We Go? Controlling Emissions in New Coal-Fired Power Plants, Presented to the U.S. EPADOEEPRI Combined

United States of America v. Salt River Project Agricultural Improvement and Power District, Civil Action No. CV 08- 1479-PI-UC-JAT Consent Decree Competitive Power College, PowerGen 2005. Selective Catalytic Reduction - From Pfanning to Operation McIIvaine UtiIitv e-AIert. No. 798.

103 104 105 I06 107 108

31

In ten tionally Omitted Parish 8 Permit Santee Cooper Cross Permit Intentionally Omitted White Pine Permit Iatan Permit

Table was excerpted from Raiiajit Sahu's Respoiise to Executive Secretary's Written Deposition Questions for Sierra CIub Expert Witness Ranajit Saliu mid Related Supplemental Request for Production of Documents, In the Matter of Sevier Power Company Power PIant, ameal before the Utah Air Oualitv Board

109 Saiidy Creek Permit 110 TrirnbIe Unit 2 Permit 111 Holconib Test 10104 112 Holcomb Test 7/04

6

Attachment 2

113 114 115 116 1 I7 1 IS I19 120 I21

Newmont Test Cross 3 Test Cross 1 &2Test CounciI Bluffs 4 Test 5/07 Council Bluffs 4 Test 8/07 Council Bluffs 4 Test 5/07 Weslon 4 Test Wygen I1 Test Hardin Test

122 123 1 24 125 126 127

128 I29 130 131 I32 I33

7

Attachment 2

PIeasant Prairie 1 Test Pleasant Prairie 2 Test Marshall 4 Test EPRI SAM Report Desert Rock Permit Tocquop Permit

Two Elk Permit Las Brisas Permit White Stallion Permit Gilbert 3 2005 Test Gilbert 3 2007 Test Northamaton Test 200 1

L

134 Northanipton Test 1995 135 IntentionalIy Omitted 136 Springerville 3 2006 Test

I54 155 156 I57 158 I59 I GO IGI f 62 163 164 165

S

Attachment 2

Longleaf Permit Turk Permit Springerville Permit Longvicw Permit La Rue B&W AirJet Burners B&W CCV Burners Als tom TFS Firing Sys tern Other B&W Burnes Zink Halo Burners Haddad ROFA Rotamix ROFA Experience List 2006 Countenianche 2003

166 167 168 169

Pegasus Slides Prairie Stare Permit Dallman Unit 4 Permit CoIeto Creek Permit

I 176 I Nebraska City 2 Permit

172 173 1 74 175

Scherer Burner Retrofit Intermountain 3 Permit Roundup Permit Plant Washington Permit

I82 IS3 IS4 185

I 195 I Cliffside 6 Permit I

Spruce 2 Permit Hawthorn 5 2003 Test Hawthorn 5 2004 Test Hawthorn 5 2005 Test

9

Attachment 2

189 190 191 192 193 194

Northampton Permit RiverHill Permit Dominion Permit Hawthorn Unit 5 Permit Pee Dee Permit Plum Point Permit