November 16, 2000
John Hornback, Director
Division for Air Quality
803 Schenkel Lane
Frankfort KY 40601
Re: Kentucky Mountain Power
Draft Air Quality Permit
Title V/PSD Circulating Fluidized Bed Coal-Fired Power Plant
These preliminary comments are submitted on behalf of the Kentucky ResourcesCouncil, Inc., a non-profit environmental advocacy organization whose membership is dedicated to prudent use and conservation of the natural resources of the Commonwealth.
At the request of local residents and at the invitation of the applicant Kentucky Mountain Power Company, the Council has been involved in the review of this project since its' inception, and has had a consulting firm review the draft permit for compliance with applicable regulations. The Council has had two discussions with the proponents of the power plant project.
After a review of the draft permit and permit application, and afterconsideration of the air emissions associated with the facility, the Councilhas these preliminary concerns regarding the draft air permit:
1. Fine particulates, expressed as PM 2.5, are a significant healthconcern, being responsible for as many as 60,000 premature deaths annually,and significant adverse health consequences to children and adults frominhalation of fine respirable particulates.
It is not apparent from the application what tonnage of fine particulatewill be emitted by this proposed plant into the airshed of this operation,and hence it is difficult to assess the extent of any health impact todownwind communities.
The company was asked during discussions whether it would consider applyingavailable technologies to address pollutants of concern beyond thosewhich might currently be required by regulation, and the company declinedto do so, but stated that they would comply with all applicable standards.
Unfortunately, given the overwhelming health science indicating thatPM 2.5 should be better controlled, and the impending adoption of standardsfor controlling PM 2.5 from one of the major sources of the fine particles,coal-fired power plants, simply following existing regulations which controltotal particulate matter or PM 10, is not necessarily sufficient to assurepublic protection.
The draft permit and application confine the discussion of ParticulateMatter to Total Suspended Particles and PM10, and for the purposesof this permit PM10 is considered identical to PM. The PM2.5subset of PM10 is not addressed anywhere in the documentation.
If the EPA PM2.5 standards become enforceable, will thisplant as designed be in compliance with these new health-based standardsor will retrofitting be necessary? Is it not better to design to PM2.5standards now and be ahead of the curve, since the health science overwhelminglysupports better controls in order to protect the public?
Will current baghouse technology capture particulates less than 2.5microns at the same efficiencies as particulates of less than 10 micronsare captured?
2. Additionally, the Council is concerned with the emissions of otherpollutants of concern. From the information provided, it does not appearthat this emission source is intended to replace power from existing olderand less-efficient base load coal-fired power plants, but instead representsnew emissions from power generation that will possibly be dispatched aheadof gas-fired peaking plants and may replace that cleaner-burning fuel sourcerather than offsets of older, less efficient power plants.
The Council is concerned with the proposal to emit into the airshed, an additional 5,585 tons per year of sulfur oxides, and between 1,564 and2,792 tons per year of nitrogen oxides. The Council endorses the commentsof the National Parks and Conservation Association regarding the need formore thorough analysis of the impacts of the emissions of these ozone precursorson the Great Smoky Mountains, including the use of the more appropriateCALPUFF air quality modeling. The Council also supports the proposal forwithholding final action on the permit pending that re-assessment and othermodifications as recommended by the federal land manager for that ClassI area.
3. The Council is also concerned with the lack of quantitative dataregarding air toxic emissions from the combustion of coal and waste coal,and believes that a more thorough assessment of whether the plant willemit hazardous compounds in harmful quantities is mandated by state regulationand is appropriate. Hazardous Air Pollutants that would be regulated underthe "maximum achievable control technology" or MACT standards if emittedby a different industrial source, are not currently regulated under MACTstandards in the case of steam electric generating stations. Has the applicantand the state speciated the Hazardous Air Pollutants that would be emittedfrom this plant and estimated total emissions? What are the numbers andpotential consequences to public health and the environment?
4. The emission of up to 1.81 tons per year of mercury into the atmosphereis also of significant concern. Coal-burning utilities are responsiblefor about 1/3 of the total mercury emissions from human activities in theUnited States, and power plants are, according to the NESCAUM, the leadingsource of unregulated mercury emissions in the United States.
There are two primary mechanisms by which mercury emissions can be controlled.The first mechanism is the choice of fuel source and corresponding mercurycontent. The second mechanism is the fraction of mercury captured duringthe control of particulate matter (PM10) emissions. Every quarterthe facility is required to take a grab sample of the fuel "as fired" tothe CFB (Circulating Fluidized Bed) to determine Lead, Beryllium, Mercuryand Fluoride content.
It is unclear, however, how will this information be used to effectquality control on the nature of coal refuse supplied to the power plant?
By nature, the quality and properties of coal refuse is highly variable.If several months of coal refuse supplied during a period between testshappens to contain levels of these elements higher than what is establishedduring baseline testing, will the fuel be burned anyway before the grabsampling indicates a potential problem? Will the state require qualitystandards on the coal refuse supply to ensure this event dies not occur?Will quality control on-site include stockpiling and blending of coal refusedelivered to achieve a consistent feed quality for the constituents ofconcern? Is a grab sample truly representative of the "as fired" fuel?Since emissions of these elements are not actually being monitored (otherthan initial performance testing), what percentage of the lead, beryllium,mercury and fluoride in the coal is emitted? What percentage is assumedto be captured within the PM10 fraction of particulates? Oris there a percentage recovery established between fuel "as fired" contentand emissions measured during performance testing? No control efficiencyis assigned to mercury emissions in the permit application, and no meaningfulcontrols on mercury content and on on-going fuel quality and compositionappear to be contemplated by the state.
Meaningful measures to control mercury emissions must be included, orappropriate calculations based on actual performance testing, and thenmodeling and fate and transport studies must be conducted to determinewhether the emissions would constitute a hazard to public health or theenvironment in violation of state regulation.
5.. The cumulative impact of emissions of lead, beryllium, fluoride,mercury and other unspeciated air toxics, should also be evaluated, quantified,and assessed for human health and ecological impact under state regulation.
6. When modeling was done again to reflect the new location, increasedboiler capacity, and revised emission rate limits, it is unclear whetherfluoride was modeled correctly at 0.0053 lbs/MMBTU to obtain the resultof 0.238 micrograms per cubic meter (g/m3), or whether the previouslimit rate of 0.0012 lbs/MMBTU was used as indicated on the KMP MODELINGIMPACT COMPARISON dated 8/2/00? The entry of 0.0012 lbs/MMBTU on the sheetmay simply be a typo, but if not, and the revised limit was not input intothe model, there could very well be pre-construction monitoring implications.
7. On page 22 of the KMP PSD/Title V/Phase II Application, the moststringent RBLC (RACT/BACT/LAER Clearinghouse) Emission Limit for PM/PM10applied to CFB Boilers is 0.011 lb/MMBTU. Yet the permit is based on alimit of 0.015 lb/MMBTU. The explanation for this discrepancy is the statementthat the ash content of coal refuse is inherently higher than expectedfrom burning coal.
The Council is concerned that this emission limit should not be increased,and that the applicant should be required to demonstrate the inabilityto meet the more stringent limit, and to speciate and discuss the healthimpacts of the PM 2.5 fraction from the proposed higher limit, and whatthe efficiency of capture of the proposed baghouse is for PM 2.5.
8. Concerning Nitrogen Oxide emissions, there appears to be a discrepancybetween the draft permit emissions limitations (page 5, part e) and theStatement of Basis (page 6) regarding allowable NOX emissions.The draft permit emissions limitations specify a maximum concentrationof 0.07 lbs/hour, or 1,564 tons/year. The Statement of Basis indicatesthat the facility, when burning more than 25% coal refuse, is permittedto emit up to 0.1 lbs/hour, or 2,234 tons per year. In the case of KentuckyMountain Power it appears the CFB fuel will consist of at least 25% coalrefuse by weight. Coal refuse is defined as waste products of coal mining,physical coal cleaning, and coal preparation operations containing coal,matrix material, clay and other organic and inorganic material. Furthermore,the Statement of Basis allows an annual limit of 2,798 tons per year whilefiring coal refuse, implying an emissions rate of about 0.125 lbs/hour.In Section D of the draft permit, pages 27 & 28, there is a discussionregarding optimization of the SNCR (Selective Non-Catalytic Reduction)for NOX emissions reductions. During the optimization process(up to one year) the NOX emission limit is waived. After theoptimization process is complete (within 18 months after the plant's initialstartup date), the Kentucky Division for Air Quality may choose to re-evaluatethe NOX emission limit set forth in the permit. Thus it appearsthe long-term NOX emission limit is undefined at thispoint in time, other than the "provisionally guaranteed" 0.1 lbs/hour.Emissions of NOX are permitted to start at 0.125 lbs/hour withthe ultimate goal of reducing this figure to 0.07 lbs/hour. A brief reviewof SNCR technologies and case histories indicates the NOX reductionefficiency varies from process to process.
Given the potential for adverse air quality impacts on a Class I area,the Council believes that the permit should include an enforceable upperlimit for NOx emissions when the plant is in coal-only, or co-firing refusewith waste coal. We concur with the National Park Service recommendationthat the target of 0.07 pounds per million Btu should be coupled with anupper bound enforceable limit of 0.10 pound.
9. The air quality modeling leaves some questions unanswered.A. This proposal used the ISCST3 air dispersion modeling. Wouldit be more appropriate to model pollutant dispersion using the newmodels AERMOD or ISC-PRIME?
B. Air dispersion modeling seems to be limited to activities on-site.Where is the coal refuse loading and transport Particulate Matter accountedfor?
C. Previous modeling accomplished for the Class 1 areas of LinvilleGorge and Great Smokey Mountains also used the ISCST3 model. The IndustrialSource Complex Short Term (ISCST3) model is the US EPA's current regulatorymodel for many New Source Review (NSR) and other air permitting applications.The ISCST3 model is based on a steady-state Gaussian plume algorithm, andis applicable for estimating ambient impacts from point, area, and volumesources out to a distance of about 50 kilometers. Thus the modeling resultsfor the Class 1 areas examined are essentially meaningless since they aremuch further away than 50 kilometers.
We concur with the National Park Service that a more appropriate modelfor this instance is CALPUFF, which utilizes a Lagrangian puff dispersionalgorithm and is considered valid for long-range transport of pollutantsto receptors 50 km 300 km away.
Concerning fluoride emissions the emissions limit of 0.0053 lbs/MMBTUappear to be essentially a negotiated number and applies to fluoride emittedas HF with a reduction efficiency of 50%. KMP committed to further testingonce the plant is built to determine if additional reduction of HF is possible.Will the Kentucky Division of Air Quality hold KMP to this commitment,and where is an enforceable obligation to conduct such testing and to adoptadditional technology or other measures is further reductions are needed?
11. Opacity is to be measured with a COM (Continuous Opacity Monitor),and if the level of Opacity exceeds 5% above the base level of Opacityestablished during performance testing, an inspection of the COM and controlequipment is required, and possibly a stack test. If the CEM illustratesemissions of SO2 or NOX in excess of those allowedor demonstrated during performance testing, is there a similar requirementfor an inspection of the COM and control equipment and possibly a stacktest? Or is Opacity the only indicator that something is wrong?
12. Concerning upset provisions, under the permit the Particulate andOpacity limitations do not apply during startup, shutdown, or malfunction.Will spans of the CEM equipment be sufficient to measure emissions duringupset conditions?
13. For Emissions Unit 03 Bottom Ash Handling System, Emissions Unit04 Fly Ash Handling System, and Emissions Unit 07 Coal Prep System. Whereis fugitive dust considered? These control efficiencies are mandated at99.9%, but the only method of determining if something is wrong with thecontrol equipment is a visual Opacity (>20%) observation, which would appearinsufficient to demonstrate compliance with the control efficiency.
14. For Emission Unit 05 Limestone Prep System and Emission Unit 06Limestone Mill Dryers, How often will the Division require measurementof PM concentration to ensure the limit of 0.05 gr/dscm is not exceeded?
15. For Emissions Units 08 and 09, Cooling Towers #1 and #2 with controlequipment of 0.01% drift eliminators. Why is fugitive dust considered here?Note that the RBLC limitation is stated to be 0.004% drift on page 46 ofthe Application. Where is the explanation for this difference?
16. For Emissions Unit LS Haul Roads. Since there are not monitoringrequirements, are fugitive emissions estimated for paved roads per AP-42or other methods? It is stated that all roads on site will be paved, butthere are no control methods proposed to keep dust to a minimum on pavedroads. Appropriate controls would include washing down roads after a dryperiod. How many miles of roads are proposed for the site?
17. What triggers the requirement of submitting a Risk Management Plan?Page 38 of draft permit.
18. Concerning the Acid Rain Permit, according to the permit at page5 - units 01 and 02 (CFB's) will be constructed after the SO2allocation date; therefore these units will have no SO2 allowancesallocated by US EPA and must obtain offsets. What exactly does this mean?If all the SO2 allowed for emissions is already allocated, canand will the plant start operating anyway?
19. In the draft permit, page 39, no alternate operating scenarios areconsidered. Yet in the permit application, page 70, two additional alternateoperating scenarios are discussed. These are Fuels and Materials and EquipmentFlexibility. Does this mean that the Division did not accept the applicationalternate scenarios as valid?
These comments and concerns are preliminary, and will be supplementedduring the remainder of the comment period announced in the newspaper.At this point in time, there remain numerous significant unanswered questionsregarding the air quality impacts of the proposed plant, which justifya withdrawal of the draft permit and reassessment of the air quality impactsof sulfur and nitrogen oxides on the Great Smoky Mountains, and of airtoxics, lead, mercury, beryllium, and fine particulates on public healthand the environment.