Prevention of Significant DeteriorationSulfur Dioxide

Baseline Emission Rates

April 2002

ND Department of HealthDivision of Air Quality

1200 Missouri Avenue, Room 304Box 5520

Bismarck, ND 58506-5520

2

Table of Contents

Page

I. Introduction . . . . . . . . . . . . . . . . . . . . . . 5A. Purpose . . . . . . . . . . . . . . . . . . . . . . 5B. Definitions . . . . . . . . . . . . . . . . . . . . 5C. Baseline Areas . . . . . . . . . . . . . . . . . . 11D. Normal Operations . . . . . . . . . . . . . . . . . 12

II. Emission Calculation Methodologies . . . . . . . . . . . 18A. Emission Factors . . . . . . . . . . . . . . . . . 19B. Emissions Testing Data . . . . . . . . . . . . . . 25C. Continuous Emission Monitor (CEM) Data . . . . . . 26

III. Major Source Baseline Emission Rate Calculations . . . . 26A. Beulah Power Plant . . . . . . . . . . . . . . . . 26B. R.M. Heskett Station . . . . . . . . . . . . . . . 32C. Leland Olds Station . . . . . . . . . . . . . . . . 36D. Lignite Gas Processing Plant . . . . . . . . . . . 42E. Mandan Refinery . . . . . . . . . . . . . . . . . . 45F. Wm. J. Neal Station . . . . . . . . . . . . . . . . 51G. Royal Oak Charcoal Briquette Plant . . . . . . . . 55H. Tioga Gas Plant . . . . . . . . . . . . . . . . . . 63I. Williston Refinery . . . . . . . . . . . . . . . . 65J. Stanton Station Unit 1 . . . . . . . . . . . . . . 67K. M.R. Young Station . . . . . . . . . . . . . . . . 71

IV. Oil and Gas Wells Baseline Emission Rate Calculations . 79

Appendix A - Beulah Power Plant DataAppendix B - R.M. Heskett Station DataAppendix C - Leland Olds Station DataAppendix D - Lignite Gas Plant DataAppendix E - Mandan Refinery DataAppendix F - Wm. J. Neal Station DataAppendix G - Royal Oak Briquetting Plant DataAppendix H - Tioga Gas Plant DataAppendix I - Williston Refinery DataAppendix J - Stanton Station Unit 1 DataAppendix K - M.R. Young Station Data

3

List of Figures

Page

1 Heat Input per Operating Hour . . . . . . . . . . . 142 Total Heat Input . . . . . . . . . . . . . . . . . 153 Beulah Station Units 1-2 Sulfur Data . . . . . . . 284 Beulah Station Units 3-5 Sulfur Data . . . . . . . 295 Beulah Station Units 1-5 Heat Input . . . . . . . . 306 R.M. Heskett Station Sulfur Data . . . . . . . . . 337 R.M. Heskett 1 Heat Input . . . . . . . . . . . . . 348 R.M. Heskett 2 Heat Input . . . . . . . . . . . . . 359 Leland Olds 1 Sulfur Data . . . . . . . . . . . . . 3710 Leland Olds 2 Sulfur Data . . . . . . . . . . . . . 3811 Leland Olds 1 Heat Input . . . . . . . . . . . . . 3912 Leland Olds 2 Heat Input . . . . . . . . . . . . . 4013 Wm. J. Neal Station Sulfur Data . . . . . . . . . . 5314 Wm. J. Neal Station Heat Input . . . . . . . . . . 5415 Royal Oak Sulfur Data . . . . . . . . . . . . . . . 5616 Royal Oak Boilers Coal Usage . . . . . . . . . . . 5817 Royal Oak Carbonizers Coal Usage . . . . . . . . . 5918 Royal Oak Briquetting Plant Coal Usage . . . . . . 6019 Stanton 1 Sulfur Data . . . . . . . . . . . . . . . 6820 Stanton 1 Heat Input . . . . . . . . . . . . . . . 6921 M.R. Young 1 Sulfur Data . . . . . . . . . . . . . 7222 M.R. Young 1 Heat Input . . . . . . . . . . . . . . 7323 M.R. Young 2 Sulfur Data . . . . . . . . . . . . . 7424 M.R. Young 2 Heat Input . . . . . . . . . . . . . . 76

4

List of Abbreviations

API American Petroleum InstituteAP-42 Compilation of Air Pollutant Emissions

Factors; Fifth EditionAvg. averagebbl barrelBtu British thermal unitBtu/hr British thermal units per hourBtu/lb British thermal units per poundCEM continuous emission monitorCO carbon monoxideF emission factorFCCU fluidized catalytic cracking unitft3 cubic feetgal gallonH2S hydrogen sulfidelb/day pounds per daylb/hr pounds per hourlb/106 Btu pound per million British thermal unitslb-mole pound molelb/lb-mole pounds per pound molelb/gal pounds per gallonlb/scf pounds per standard cubic footlb/ton pounds per tonMWe megawatts electricityNa2O sodium oxideNDAC North Dakota Administrative CodeNO2 nitrogen dioxideppmv parts per million volumePM10 particulate matter less than 10 micrometersPSD Prevention of Significant DeteriorationS sulfurSO2 sulfur dioxideSO3 sulfur trioxideSOx sulfur oxidesscf standard cubic feetscfd standard cubic feet per dayscf/day standard cubic feet per daySRU sulfur recovery unittons/yr tons per yearWBS Williston Basin Regional Air Quality Studyµg/m3 micrograms per cubic meters

5

Baseline Emission Rates

I. Introduction

A. Purpose:

The purpose of this document is to present the data andthe methodology that were used in establishing theproposed emission rate for sources that contribute to thebaseline concentration. This document presents data foreach source on production rates (heat input, coal usage,processing rates, etc.), fuel and raw materials quality,hours of operation and other pertinent data. Thecalculation of the baseline emission rate is presentedalong with the methodology used in the calculation. Themethodology for the various sources includes the use offactors from AP-42, Compilation of Air Pollutant EmissionFactors, stack test data, or a mass balance approach.Any interpretation of the data or assumptions are alsoexplained.

B. Definitions:

Within the Prevention of Significant (PSD) rules found inNDAC 33-15-15, there are several definitions of terms andthe interpretation of those definitions that are criticalto the establishment of the emission rate that is used toestablish the baseline concentration. These terms, asdefined in NDAC 33-15-15, include:

"Actual emissions" means the actual rate of emissions ofa contaminant from an emissions unit, as determined inaccordance with paragraphs a through d.

a. In general, actual emissions as of a particulardate must equal the average rate, in tons per year,at which the unit actually emitted the contaminantduring a two year period which precedes theparticular date and which is representative of

6

normal source operation. The department may allowthe use of a different time period upon adetermination that it is more representative ofnormal source operation. Actual emissions must becalculated using the unit's actual operating hours,production rates, and types of materials processed,stored, or combusted during the selected timeperiod.

b. The department may presume that source-specificallowable emissions for the unit are equivalent tothe actual emissions of the unit.

c. For any emissions unit (other than an electricutility steam generating unit specified inparagraph 4) which has not begun normal operationson the particular date, actual emissions shallequal the potential to emit of the unit on thatdate.

d. For an electric utility steam generating unit(other than a new unit or the replacement of anexisting unit) actual emissions of the unitfollowing the physical or operational change shallequal the representative actual annual emissions ofthe unit following the physical or operationalchange, provided the source owner or operatormaintains and submits to the reviewing authority,on an annual basis for a period of five years fromthe date the unit resumes regular operation,information demonstrating that the physical oroperational change did not result in an emissionsincrease. A longer period, not to exceed tenyears, may be required by the department if itdetermines such a period to be more representativeof normal source postchange operations.

"Allowable emissions" means the emission rate of astationary source calculated using the maximum ratedcapacity of the source (unless the source is subject to

7

enforceable construction permit conditions which restrictthe operating rate, or hours of operation, or both) andthe most stringent of the following:

a. Applicable standards of performance or emissionlimitations as set forth in this article.

b. The emission rate specified as an enforceablepermit condition.

"Baseline area" means any intrastate area (and every partthereof) designated as attainment or unclassifiable undersection 107 (d)(1)(D) or (E) of the Federal Clean Air Act[Pub. L. 95-95] in which the major source or majormodification establishing the minor source baseline datewould construct or would have an air quality impact equalto or greater than one Fg/m3 (annual average) of thecontaminant for which the minor source baseline date isestablished. Any baseline area established originallyfor the total suspended particulate increments shallremain in effect and shall apply for purposes ofdetermining the amount of available PM10 increments,except that such baseline area shall not remain in effectif the department rescinds the corresponding minor sourcebaseline date. North Dakota is divided into twointrastate areas under section 107 (d)(1)(D) or (E) ofthe Federal Clean Air Act [Pub. L. 95-95]: the CassCounty portion of Region No. 130, the Metropolitan Fargo-Moorhead Interstate Air Quality Control Region; andRegion No. 172, the North Dakota Intrastate Air QualityControl Region (the remaining fifty-two counties).

"Baseline concentration" means that ambient concentrationlevel which exists in the baseline area at the time ofthe applicable minor source baseline date. A baselineconcentration is determined for each contaminant forwhich a minor source baseline date is established andincludes:

8

a. The actual emissions representative of sources inexistence on the applicable minor source baselinedate, except as provided in paragraph b.;

b. The allowable emissions of major stationary sourceswhich commenced construction before the majorsource baseline date but were not in operation bythe applicable minor source baseline date.

The following will not be included in the baselineconcentration and will affect the applicable maximumallowable increases:

a. Actual emissions from any major stationary sourceon which construction commenced after the majorsource baseline date; and

b. Actual emissions increases and decreases at anystationary source occurring after the minor sourcebaseline date.

"Major source baseline date" means:

a. In the case of particulate matter and sulfurdioxide, January 6, 1975; and

b. In the case of nitrogen dioxide, February 8, 1988.

"Minor source baseline date" means the earliest dateafter the trigger date on which a major stationary sourceor a major modification subject to requirements of thischapter submits a complete application under the relevantregulations. The trigger date is:

a. In the case of particulate matter and sulfurdioxide, August 7, 1977; and

b. In the case of nitrogen dioxide, February 8, 1988.

9

The baseline date is established for each contaminant forwhich increments or other equivalent measures have beenestablished if:

a. The area in which the proposed source ormodification would construct is designated asattainment or unclassifiable under section 107(d)(1)(D) or (E) of the Federal Clean Air Act [Pub.L. 95-95] for the contaminant on the date of itscomplete application under this chapter; and

b. In the case of a major stationary source, thecontaminant would be emitted in significant amountsor, in the case of a major modification, therewould be a significant net emissions increase ofthe contaminant.

Any minor source baseline date established originally forthe total suspended particulate increments shall remainin effect and shall apply for purposes of determining theamount of available PM10 increments, except that thedepartment may rescind any such minor source baselinedate where it can be shown by the applicant, to thesatisfaction of the department, that the emissionsincrease from the major stationary source, or the netemissions increase from the major modification,responsible for triggering that date did not result in asignificant amount of PM10 emissions.

Although not defined in the PSD rules, several otherterms are used in this document. For purposes of thisdocument:

“average emission rate” means the average hourly emissionrate within a year based on the annual emission rate andactual hours of operation for the given year.

“baseline emission rate” means the emission rate of asource that contributes to the baseline concentration.

10

“baseline period” means the two year timeframe for whichthe baseline emission rate is calculated.

“baseline source” means a facility of which any portionof its emissions contributes to the baselineconcentration.

“two year average emission rate” means the average hourlyemission rate within two consecutive years based on thetotal annual emissions and total hours of operation forthe two year period. Mathematically, it equals the totalemissions divided by the total hours of operation.

The Department has interpreted the definition of “actualemissions” to mean the two year average emission rate ofthe source which is representative of normal operationsfor a given period (see Summary of Legal Issues Relatingto Administration of the Prevention of SignificantDeterioration Provisions of North Dakota’s StateImplementation Plan - hereafter Legal Summary).

The baseline emission rate is the two year averageemission rate calculated for the baseline period, and isused for determining the baseline concentration for allaveraging periods. The Department has determined that atime period (two or more years) other than the two yearperiod immediately preceding the minor source baselinedate may be used for establishing the baseline emissionrate provided it is more representative of normaloperation (see Legal Summary). This may include a timeperiod after the minor source baseline date. TheDepartment has also determined that any reasonablyanticipated increases or decreases in emissions due toproduction increases, as of the minor source baselinedate, that genuinely reflect normal source operation canbe taken into account when determining the baselineconcentration and baseline emission rate (see LegalSummary).

11

C. Baseline Areas:

North Dakota is divided into two interstate areas underSection 107(d)(1)(D) or (E) of the Federal Clean Air Act(Public Law 9595): The Cass County portion of Region No.130, the Metropolitan Fargo-Moorhead Interstate AirQuality Control Region; and Region No. 172, the NorthDakota Intrastate Air Quality Control Region (theremaining 52 counties). Under the present PSDregulations, North Dakota is divided into two baselineareas (the same two areas).

The minor source baseline dates for the two baselineareas are as follows: 1) Region No. 130: Sulfur Dioxide -November 30, 1979, Particulates - November 30, 1979;2) Region No. 172: Sulfur Dioxide - December 19, 1977,Particulates - January 13, 1978; 3) NO2 - October 31,1989.

The particulate matter and sulfur dioxide minor sourcebaseline dates for Region 130 were established by theapplication of Cargill, Inc. for a sunflower seedprocessing plant.

In Region No. 172, the minor source baseline date forsulfur dioxide was established by the Warren PetroleumCompany application for the Little Knife Natural GasProcessing Plant in Billings County. The minor sourcebaseline date for particulate matter was established byBasin Electric Power Cooperative's application for theAntelope Valley Station Units No. 1 and 2 steam electricgenerating facility in Mercer County. The minor sourcebaseline date for NO2 was established by DakotaGasification Company’s application for an amended Permitto Construct. The minor source baseline dates are thedates of receipt by the Department of the last submittalsof information from the applicants that made theapplications complete.

12

The discussion in this section will be limited to RegionNo. 172 and sulfur dioxide. However, the methodology isapplicable to Region No. 130 and for nitrogen oxides andPM10.

The minor source baseline date for the eastern portion ofMontana that may be affected by sulfur dioxide emissionsfrom North Dakota sources is March 26, 1979. No attempthas been made to determine baseline emission rates forthe Montana minor source baseline date. If no problemswith Class I increment are encountered in North Dakota,none are expected in Montana. The correction of anyincrement problems in North Dakota should resolve anyproblems in Montana.

D. Normal Operations:

The Prevention of Significant Deterioration rules do notcontain a definition of “normal operations”. TheDepartment searched other rules such as the New SourcePerformance Standards (40 CFR 60), National EmissionStandards for Hazardous Air Pollutants (40 CFR 61), theNational Emission Standards for Hazardous Air Pollutantsfor Source Categories (40 CFR 63) and the Acid RainProgram rules (40 CFR 72, 73 and 75) for a definition ofnormal operations. However, a definition and amethodology for determining normal operations were notfound. Therefore, the Department had to establish itsown methodology for determining normal operations for thebaseline sources.

Most of the baseline sources that were evaluated are orwere coal-fired electric utility steam generatingfacilities. Other sources include two natural gasprocessing plants, two oil refineries, and a charcoalbriquette plant. In determining normal operations forthe facilities, the Department decided that actualpollutant emission rates should not be a direct factor inthe decision process. Production rates appear to be thefactor which defines normal operations.

13

The Department evaluated the currently existing coal-fired utility plants based on the annual heat inputs forthe various units. Both total heat input and heat inputper operating hour for a given year were evaluated. Theamount of electricity generated was also evaluated.However, data on electrical generation was only availableto the Department from 1989 to the present. The amountof electricity generated correlated well with the heatinput. Therefore, data on electricity generation is notpresented.

Figures 1 and 2 depict the heat input per hour ofoperation and total heat input for the existing majorbaseline coal-fired utility steam generating facilities.The heat inputs were calculated from information suppliedin the Annual Emission Inventory Reports for the variousfacilities. In reviewing the data, the heat input nearthe minor source baseline date was compared to operationsbefore and after that date. Since emissions for thebaseline period are to be calculated based on the actualhours of operation (lb/hr), the heat input per hour ofoperation was used to define normal operations forexisting coal-fired steam electric utility boilers. Fornearly all sources, there is a maximum two year periodvery near the minor source baseline date that is nearlyequivalent to any other period after the baseline date.The one exception is the M.R. Young Station Unit 2.Beginning in 1990, the heat input to this unit increaseddramatically. The reasons for this increase are unclear.However, based on the twelve years of previous operation,it appears that a two year period near the baseline datecan adequately represent normal operations for this unitas of the minor source baseline date.

Based on the data, it is proposed that the time period1975 through 1980 contains two consecutive years thatwere representative of normal operations for currentlyexisting utility boilers. This time period representsnearly three years before the minor source baseline date(December 19, 1977) and approximately three years after

FIGURE 1HEAT INPUT PER

OPERATING HOUR

1.0E+08

1.1E+09

2.1E+09

3.1E+09

4.1E+09

5.1E+09

1975 1980 1985 1990 1995 2000

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R(B

TU

)

L.O. 1 L.O. 2 M.R.YOUNG 1 M.R. YOUNG 2 HESKETT 1 HESKETT 2 STANTON 1

14

FIGURE 2TOTAL HEAT INPUT

0.0E+00

5.0E+12

1.0E+13

1.5E+13

2.0E+13

2.5E+13

3.0E+13

3.5E+13

4.0E+13

4.5E+13

1975 1980 1985 1990 1995 2000

HE

AT

INP

UT

(BT

U)

L.O. 1 L.O. 2 M.R. YOUNG 1 M.R YOUNG 2 HESKETT 1 HESKETT 2 STANTON 1

15

16

it. This close proximity to the minor source baselinedate maintains some consistency with EPA’s policy ofdetermining actual emissions in the two years precedingthe baseline date but also provides some flexibility fortaking into account production increases that wereanticipated on the baseline date (see F.R. Vol. 45, No.154, p. 52714). The two years immediately prior to thebaseline date (1976-77) were evaluated and eitheraccepted or rejected as being representative of normaloperations. Where a period other than 1976-77 was usedfor the baseline emission rate calculation, the reasonsare given for the choosing a different period.

All other source categories were evaluated independently;however, weight was given to using the same time periodfor consistency. After review of the other sources, itis proposed that two years within this same time period(1975-80) adequately represented normal operations forthese other sources. Again, the 1976-77 time period wasused unless there was sufficient evidence to support useof a different period.

When determining normal operations for a facility, thefuels or raw materials (coal, natural gas, oil, etc.)used in the process must be considered. When changes inthe fuels or raw materials could have been anticipated onthe baseline date, the Department proposes that it isappropriate to take these changes into account whendetermining the baseline emission rate. For the baselinesources, the characteristic that is most important is thesulfur content of the fuel burned or raw materialsprocessed.

For coal-fired steam generating units, the sulfur dioxideemission rate is directly proportional to the sulfurcontent of the coal provided all other conditions remainthe same. Based on discussions with one mine operator,the sulfur content of the coal within mine areas is, toa certain extent, known by the mine operator as well asthe source receiving the coal. Mining plans are preparedat the beginning of the mine operation and amended as

17

conditions warrant. When developing a mine, the qualityof coal (including the sulfur content) determined fromcore sampling is considered. Although the core samplingdata is not comprehensive, it is one of the factorsevaluated. The Department’s experience indicates thatmining plans are developed thirty years or more inadvance and are developed to provide the customer’s (e.g.power plant) needs. Since changes in sulfur contentcould have been anticipated based on data used indeveloping the mining plan, the Department proposes thatit is appropriate to consider the changes whencalculating the baseline emission rate.

Some baseline sources have used coal from the same minesince the beginning of operation. However, other sourceshave changed coal suppliers and mines. A change in thesource of the coal could change the sulfur content andother properties of the coal substantially. The sourceof coal for the Stanton Station switched from theIndianhead Mine in 1992 to the Freedom Mine. In 1993,the Leland Olds Station switched from the Glenharold Mineto the Freedom Mine. Because of the length of time afterthe minor source baseline date, the Department proposesthat it is appropriate only to consider sulfur changes atthe mine that was supplying the coal on the baselinedate.

Figures in Section III, Baseline Emission RateCalculations, present the annual average sulfur contentfor the baseline sources as reported in the AnnualEmission Inventory Reports. Data was generally availablefrom the early 1970's through 2000. The Department hasno information on the sampling and analysis techniquesthat were used to obtain these results. As can be seen,results vary somewhat from year to year. This variationcould be due to actual variations in coal quality, due toa variation in sampling or analysis techniques, or both.Therefore, the Department proposes that the averagesulfur content over the life of the mine which was in useon the minor source baseline date is the appropriatevalue for determining the baseline emission rate. When

18

determining the mine average sulfur content, the averagewas weighted based on the amount of coal burned in aparticular year. The average was also determinedindependent of its effect on the emission rate. In somecases, the mine average sulfur content is greater thanthe 1976-77 average (or other baseline period) and insome cases it is less. There was also no attempt made toaverage the sulfur content over units receiving coal fromthe same mine (e.g. Leland Olds 1 and 2). Other factorsmay have required some coal blending or selective miningfor a particular unit. Leland Olds 1 is a wall-firedunit while Leland Olds 2 is a cyclone unit. M.R. Young1 and 2 are both cyclone units; however, Unit 2 isequipped with a scrubber while Unit 1 is not. Unit 2also has a lower emission limit (1.2 lb/106 Btu vs. 3lb/106 Btu). Any of these factors could have influencedthe quality of coal sent to the particular unit.

II. Emission Calculation Methodologies

There were a number of options available for calculating thebaseline emission rates for the various sources. Theseinclude various emission factors, continuous emission monitordata where available, stack testing data from individualsources and a mass balance approach. Each of these methodshas some advantages and disadvantages over another method.The one common factor that is important to each method is thequality of the fuel combusted or raw material processed. Thequalities of the material in question includes the heatcontent, sulfur content, alkalinity of the coal ash, thesulfur content of the oil that is refined or natural gasprocessed, the density of fuel oil combusted, or the sulfurcontent of the coal carbonized. Where air pollution controlequipment for sulfur dioxide was utilized, the efficiency ofthe unit may be required. In general, not all of this data isavailable for the baseline period. In order to calculate thebaseline emission rates, certain assumptions have to be madeand an appropriate method selected. The following discussesthe merits of each method.

19

A. Emission Factors:

The primary source for emission factors for air pollutionsources is AP-42, Compilation of Air Pollution EmissionFactors. This document is published by the U.S.Environmental Protection Agency’s Office of Air QualityPlanning and Standards. The fifth addition of thisdocument was published in 1995 with updates in 1996through 2000. For lignite combustion sources, AP-42 waslast updated in September of 1998. For other NorthDakota baseline source categories, there have been noupdates since 1995.

AP-42 emission factors represent average emission factorsfor a given source category. The average emission factoris based on the data EPA has evaluated and may notrepresent the actual emission rate for a specific source.However, it does represent EPA’s best estimate of theemissions over the source category.

1. Lignite Combustion

For lignite combustion units, the sulfur oxidesemission factor for all of the baseline sources is30(S). The Heskett Station Unit 2 is currently afluidized bed combustor for which the emissionfactor is 10(S). However, during the baselineperiod, the unit was a spreader stoker combustorwhich has an emission factor of 30(S). Theemission factor indicates that average sulfuroxides emissions in pounds per ton of coalcombusted (lb/ton) will equal 30 times the sulfurcontent. A footnote to this emission factor states“S = Weight % sulfur content of lignite, wet basis.For example, if the sulfur content equals 3.4%,then S = 3.4. For high sodium ash (Na2O > 8%), use22S. For low sodium ash (Na2O < 2%), use 34S. Ifash sodium content is unknown, use 30S”. Anexplanation of the ash sodium content issue isfound previously in AP-42 and states, “The SO2emissions from lignite combustion are a function of

20

the sulfur content of the lignite and the lignitecomposition (i.e., sulfur content, heating value,and alkali concentration). The conversion oflignite sulfur to SOx is generally inverselyproportional to the concentration of alkaliconstituents in the lignite. The alkali content isknown to have a great effect on sulfur conversionand acts as a built-in sorbent for SOx removal.”

An emission factor of 30(S) indicates that 75% ofthe sulfur entering the combustion unit is emittedas sulfur oxides and 25% is captured in the ash(bottom ash or fly ash). An emission factor of22(S) indicates 55% emitted as SOx and 45% iscaptured in the ash. An emission factor of 34(S)indicates 85% emitted and 15% captured. Theemission factor (and emissions) can vary greatlydepending on the sodium oxide (Na2O) content of theash. Little information is available regardinglignite ash sodium content for the baseline unitsduring the baseline period. Given this fact andthat the ash sodium content can vary significantlyfrom one mine to another, and even within a mine,the average emission factor of 30(S) appears to bethe appropriate factor when using emission factorsto calculate emission rates. If additionalinformation on the sodium oxide content of the ashof the coal consumed during the baseline periodwould become available, the emission factor shouldbe reevaluated.

In response to the Department’s letter of July 3,2001, regarding baseline issues, two companies thatoperate coal-fired electric utilities submitted acomparison of 1995-2000 continuous emissionmonitoring data to estimated emissions using theAP-42 emission factor 30(S). The analysis by bothcompanies suggests that the AP-42 emission factor30(S) underestimated the emission rate for theirbaseline source units. One company states theiranalysis showed that a factor of 33.14(S) should be

21

used for annual emissions estimates and 45(S) forshort-term estimates. Any emission factor greaterthan 40(S) is theoretically impossible; however,the analysis may have yielded this result becauseof inaccuracies in stack gas flow measurement. Theother company suggested emission factors of 36.0(S)for one unit and 40.5(S) for another unit.

As discussed earlier, the alkalinity of the ligniteash significantly affects the sulfur dioxideemission rate. No data on the sodium (Na2O) contentof the lignite ash was submitted with the analysesby the two companies. Both companies obtain theircurrent coal supply from the same mine. This mineis different from the mines that supplied coalduring the baseline period. There is insufficientdata to compare the emission factor for thebaseline period to emission factors derived fromrecent continuous emission monitoring data. Unlesscoal ash sodium data is available for comparisonfor both time periods, any analysis is suspect.

AP-42 uses a rating system from A to E for theemission factors listed. The emission factor forsulfur dioxide for lignite combustion units isgiven a rating of C. The ratings are described asfollows:

A - Excellent. Factor is developed from A- and B-rated source test data taken from manyrandomly chosen facilities in the industrypopulation. The source category population issufficiently specific to minimize variability.

B - Above average. Factor is developed from A- orB-rated test data from a “reasonable number”of facilities. Although no specific bias isevident, it is not clear if the facilitiestested represent a random sample of theindustry. As with an A rating, the source

22

category population is sufficiently specificto minimize variability.

C - Average. Factor is developed from A-, B-,and/or C-rated test data from a reasonablenumber of facilities. Although no specificbias is evident, it is not clear if thefacilities tested represent a random sample ofthe industry. As with the A rating, thesource category population is sufficientlyspecific to minimize variability.

D - Below average. Factor is developed from A-,B- and/or C- rated test data from a smallnumber of facilities, and there may be reasonto suspect that these facilities do notrepresent a random sample of the industry.There also may be evidence of variabilitywithin the source population.

E - Poor. Factor is developed from C- and D-ratedtest data, and there may be reason to suspectthat the facilities tested do not represent arandom sample of the industry. There also maybe evidence of variability within the sourcecategory population.

It should also be noted that the AP-42 emissionfactor is based on a number of stack tests at NorthDakota power plants from the 1970's into the1990's. Since only the average emission rate isbeing calculated, any short term variations inemission rates within the two year baseline perioddue to coal ash sodium content may tend to averageout. Unless specific coal ash sodium (Na2O) datacan be obtained for the coal consumed during thebaseline period which would justify a differentemission factor, the Department believes theaverage AP-42 emission factor, 30(S), representsthe best emission factor available. For purposesof the baseline emission rate calculations, all

23

sulfur oxides emissions are considered to be sulfurdioxide.

On March 20, 2002 and March 26, 2002, additionalinformation regarding the sodium content of coalconsumed at two facilities was received. TheDepartment has not had time to evaluate the data.However, an evaluation of the data will be made andany adjustments to the emission factor that arejustified will be made in the final determinationof the baseline emission rates.

2. Fuel Oil Combustion

For fuel oil combustion units, AP-42 lists sulfuroxides emission factors ranging from 157(S) to142(S) (lb/1000 gal) depending on the grade of oilfired. The emission factors indicate that all ofthe sulfur present in the fuel oil is emitted as asulfur compound. AP-42 states “The emission of SOxfrom conventional combustion systems arepredominantly in the form of SO2. Uncontrolled SOx

emissions are almost entirely dependent on thesulfur content of the fuel and are not affected byboiler size, burner design, or grade of fuel beingfired. On average, more than 95 percent of thefuel sulfur is converted to SO2, about 1 to 5percent is further oxidized to sulfur trioxide(SO3), and 1 to 3 percent is emitted as sulfateparticulate. SO3 readily reacts with water vapor(both in the atmosphere and in the flue gases) toform a sulfuric acid mist.”

The sulfur content of the fuel oil is generallyexpressed as a weight percent. The variation inthe emission factors is due to the difference indensity of the various grades of fuel oil.Distillate oil is the least dense fuel and has anemission factor of 142(S). Grade 6 fuel oil is thedensest and has an emission factor of 157(S).

24

Since all of the sulfur in the fuel oil will beemitted, the Department proposes that AP-42emission factors adequately represent emissionswhen firing fuel oil. However, where the actualdensity of any fuel oil combusted is known, a massbalance approach should be used to determine sulfuroxides emissions. For purposes of the baselineemissions calculations, all sulfur oxides emissionswere assumed to be sulfur dioxide.

3. Gaseous Fuels or Waste Gas Combustion

AP-42 states “Emissions of SO2 from natural gas-fired boilers are low because pipeline qualitynatural gas typically has sulfur levels of 2,000grains per million cubic feet. However, sulfur-containing odorants are added to natural gas fordetecting leaks, leading to small amounts of SO2

emissions. Boilers combusting unprocessed naturalgas may have higher SO2 emissions due to higherlevels of sulfur in the natural gas. For theseunits, a sulfur mass balance should be used todetermine SO2 emissions.”

The Department proposes that emissions of sulfurdioxide from baseline sources combusting “pipelinequality natural gas” are insignificant whencompared to emissions due to other fuels and wasteproducts. Therefore, sulfur dioxide emissions fromthe combustion of pipeline quality natural gas arenot calculated and not included in the baselineemission rates. This determination only affectsthe oil refinery and natural gas processing sourcecategories.

For fuel gas combustion and waste gas (tail gas oracid gas) incineration/flaring, the Departmentproposes that a mass balance approach isappropriate and adequately represents the baselineemission rate. The Department assumed that all

25

sulfur in the sour gas was converted to sulfurdioxide and emitted to the atmosphere.

B. Emissions Testing Data:

Emissions testing data represents a good snapshot ofemissions during the testing period provided the testmethods, procedures, and analysis methods are sufficientfor the task. However, emissions testing data may or maynot be representative of emissions over a longer periodof time such as the baseline period. As discussedearlier, emissions can vary significantly depending onthe characteristics of the fuel. The sulfur content oflignite combusted at a power plant in North Dakota canvary by as much as a factor of 4 or more in a given year.Short-term sulfur dioxide emission rates can vary by thesame factor. One or two emissions tests in a year aregenerally not sufficient to establish the emission ratefor the whole year (or baseline period). However, if theemissions testing data can be correlated with sufficientfuel characteristics, it can provide a basis forestablishing emission factors for the source. Also,using emissions testing data may provide a preferableapproach to a mass balance approach especially when coalis burned as fuel or processed.

The Royal Oak charcoal briquetting plant was a uniquefacility because it utilized lignite to make charcoalbriquettes. The Department is not aware of any othercharcoal briquetting facility in the Unites States thatused lignite. Sulfur dioxide emissions from the boilerscan be adequately estimated using AP-42 emission factors.However, there are no AP-42 emissions factors for sulfurdioxide emissions from the carbonizer furnaces(Herreschoff and Lurgi furnaces) that utilized lignite.

As discussed later, stack test data was used to establishan emission factor for the carbonizer furnaces.

26

C. Continuous Emission Monitor (CEM) Data:

The only baseline sources operating CEM systems forsulfur dioxide during the baseline period were MinnkotaPower Cooperative at M.R. Young Unit 2 and Basin ElectricPower Cooperative at Leland Olds Unit 2. Neither ofthese systems provided an emission rate on a mass perunit of time (i.e. lb/hr) basis. There is also no datacurrently available to the Department that would providea comparison of the CEM readings to the sulfur content ofthe coal during the baseline period. Based on thelimited CEM data available from the baseline period, theDepartment proposes that it is not an adequate resourceto determine baseline emission rates.

As discussed earlier, a comparison of current CEM data tocurrent emission rates may be useful provided adequateinformation is available about the current properties ofthe coal and the properties of the coal burned during thebaseline period. Lacking information on the propertiesof the coal burned during the baseline period, anycomparison to current CEM and coal data is of littlevalue.

III. Baseline Emission Rate Calculations

A. Beulah Power Plant:

The Beulah Power Plant, which was located in Beulah,North Dakota, was owned and operated by Montana DakotaUtilities at the time it ceased operation in 1986. Thefacility actually consisted of two plants next to eachother known as the Knife River Station and the DakotaStation. For clarity, the two plants are referred to asthe Beulah Plant. The plant consisted of five differentboilers. Boilers 1 and 2 were installed in 1927, Boiler3 in 1928 and Boilers 4 and 5 in 1948. Boilers 1-3 werechain grate stokers while Boilers 4-5 were spreaderstokers. Boilers 1 and 2 had a nominal rating of 77 x106 Btu/hr each, Boiler 3 was 88 x 106 Btu/hr and Boilers4 and 5 were 91 x 106 Btu/hr each. Boilers 1 and 2

27

exhausted emissions through separate stacks while Boilers3-5 vented through a common stack as of the minor sourcebaseline date. Since Boilers 1 and 2 have been modeledin the past as a single source and boilers 3-5 as anothersingle source, the units are treated the same way in thisanalysis.

The Beulah Plant obtained its coal from the Beulah Mine(North Beulah Mine and later South Beulah Mine). Figures3 and 4 present the coal sulfur content data submitted inthe Annual Emission Inventory Reports for the facility.Based on this data, an average sulfur content, weightedon coal usage, was determined to be 0.70% for Units 1 and2 and 0.71% for Units 3-5.

Figure 5 presents the total heat input for all units.The facility was evaluated to determine normal operationsbased on the total heat input for the facility and notthe heat input per operating hour. Because of theflexibility provided by five boilers at the facility,some boilers were not needed for power generation andmaintained on warm standby. The Annual EmissionInventory Reports for the facility for some years duringthe period evaluated (1972-1986) list total coal usagefor power generation and heating, while only listing thehours of operation for power generation. The total hoursof operation for both heating and power generation isunknown. Therefore, an accurate estimate of the heatinput per operating hour could not be calculated.Emissions were calculated separately for Boilers 1-2 andBoilers 3-5, because they are modeled separately. TheDepartment proposes that the 1976-1977 period isrepresentative of normal operations for the facility onthe minor source baseline date. The Department hascalculated the baseline emission rates based on the AP-42emission factor 30(S) and data in the Annual EmissionInventory Reports.

FIGURE 3BEULAH STATION

UNITS 1 - 2

0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

0.9

1972 1974 1976 1978 1980 1982 1984

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

28

FIGURE 4BEULAH STATION

UNITS 3 - 5

0.4

0.5

0.6

0.7

0.8

0.9

1

1972 1974 1976 1978 1980 1982 1984 1986

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

29

FIGURE 5BEULAH STATION

UNITS 1 - 5

0.00E+00

2.00E+11

4.00E+11

6.00E+11

8.00E+11

1.00E+12

1.20E+12

1.40E+12

1.60E+12

1.80E+12

1972 1974 1976 1978 1980 1982 1984 1986

TO

TA

LH

EA

TIN

PU

T(B

TU

)

30

31

The data and results are as follows:

Units Year

CoalUsage(tons)

MineAvg. SulfurContent (%)

SO2Emissions(tons)

Hours ofOperation(Total forall Boilers)

AverageHours ofOperation

(per boiler)

2-Year Avg.EmissionRate

(lb/hr)

1 & 21 & 2

19761977

2181121322

0.700.70

229.0223.9

7427.65789.1

3713.82894.6

137.1

3 - 53 - 5

19761977

5403368452

0.710.71

575.5729.0

15879.918975.4

5293.36325.1

224.6

Where:

SO2 emissions (tons) = (30)(Mine Avg. Sulfur Content)(CoalUsage) ÷ (2,000 lb/ton)

2-Year Avg. EmissionRate (lb/hr) = [1976 SO2 Emissions (tons) + 1977 SO2

Emission (tons)] (2,000 lb/ton) ÷ [1976Ag. Hours of Operation + 1977 Ag. Hoursof Operation]

32

B. R.M. Heskett Station:

The R.M. Heskett Station is located near Mandan, NorthDakota and is operated by Montana Dakota Utilities. Unit1 began operation in 1954 and Unit 2 in 1963. Unit 1 hasa spreader stoker combustion unit and a nominal heatinput rating of 387 x 106 Btu/hr. Unit 2 was originallyconstructed with a spreader stoker combustion unit butwas converted to a fluidized bed unit in the 1980's.Unit 2 has a nominal heat input rating of 916 x 106

Btu/hr. Each of these units exhaust through a separatestack and are treated as separate units for purposes ofthis analysis.

The Heskett Station has obtained its coal from the BeulahMine (North Beulah Mine and South Beulah Mine) in thepast and continues to receive it from that mine. Figure6 presents the coal sulfur content data submitted in theAnnual Emission Inventory Reports for the facility.Based on the data, a weighted average sulfur content of0.80% was calculated for each unit.

Figures 7 and 8 present the heat input for each unit.Based on this information, the Department proposes thatthe 1976-77 time period adequately represents normaloperations for the facility as of the minor sourcebaseline date. The Department has calculated thebaseline sulfur dioxide emission rate for each sourcebased on the AP-42 emission factor 30(S) and data fromthe Annual Emission Inventory Report. The data andresults are as follows:

Unit Year

CoalUsage(tons)

MineAvg. SulfurContent (%)

SO2Emissions(tons)

Hours ofOperation

2-Year Avg.EmissionRate

(lb/hr)

11

19761977

159196171162

0.800.80

1910.42053.9

74337879

517.8

22

19761977

376017406145

0.800.80

4512.24873.7

76687871

1208.0

Where:

SO2 emissions (tons) = (30)(Mine Avg. Sulfur Content)(CoalUsage) ÷ (2,000 lb/ton)

2-Year Avg. EmissionRate (lb/hr) = [1976 SO2 Emissions (tons) + 1977 SO2

Emission (tons)] (2,000 lb/ton) ÷ [1976Hours of Operation + 1977 Hours ofOperation]

FIGURE 6HESKETT STATION

0

0.2

0.4

0.6

0.8

1

1.2

1974 1979 1984 1989 1994 1999

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

33

FIGURE 7HESKETT 1

1.00E+11

6.00E+11

1.10E+12

1.60E+12

2.10E+12

1975 1980 1985 1990 1995 2000

HE

AT

INP

UT

(BT

U)

1.00E+08

2.00E+08

3.00E+08

4.00E+08

5.00E+08

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R(B

TU

)

TOTAL HEAT INPUT HEAT INPUT PER OPERATING HOUR

34

FIGURE 8HESKETT 2

1.00E+12

1.50E+12

2.00E+12

2.50E+12

3.00E+12

3.50E+12

4.00E+12

4.50E+12

5.00E+12

5.50E+12

6.00E+12

1975 1980 1985 1990 1995 2000

TO

TA

LH

EA

TIN

PU

T(B

TU

)

5.00E+08

5.50E+08

6.00E+08

6.50E+08

7.00E+08

7.50E+08

8.00E+08

8.50E+08

9.00E+08

9.50E+08

1.00E+09

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R

TOTAL HEAT INPUT HEAT INPUT PER OPERATING HOUR

35

36

C. Leland Olds Station

The Leland Olds Station consists of two units and islocated near Stanton in Mercer County. The facility isoperated by Basin Electric Power Cooperative. Unit 1 isa pulverized wall fired unit which began operation in1966. The unit has a nominal heat input rating of 2622x 106 Btu/hr. Unit 2 is a cyclone unit with a nominalrating of 5130 x 106 Btu/hr and a generator nameplaterating of 440 MWe. This unit was put into operation in1975. Each unit exhausts through a separate stack and istreated as separate unit for this analysis.

Coal for the Leland Olds Station was obtained from theGlenharold Mine until the mine closed in 1993. Afterthis date, coal has been obtained from the Freedom Mineand other sources. Figures 9 and 10 shows the variationin the yearly average sulfur content of the coal consumedas reported in the Annual Emission Inventory Reports thatwere submitted for the facility. Based on the data, theDepartment has determined a weighted mine average sulfurcontent for the coal obtained from Glenharold Mine to be0.65% for both units.

Figures 11 and 12 present the heat inputs for Units 1 and2, respectively. The Department proposes that the 1976-77 time period is representative of normal operations forUnit 1. Unit 2 began operation in late 1975 withcommercial operation on December 15, 1975. Commercialoperation was only at 68% of the units rating because ofequipment problems (see 12/2/75 letter in Appendix C).In a May 3, 1976 letter, Basin Electric explained thatUnit 2 had operated at only about 300 megawatts for thefirst couple of months in 1976. In May of 1976 generatorproblems forced the shutdown of unit for an extendedperiod (see 5/26/76 letter in Appendix C). TheDepartment proposes that 1976 is not representative ofnormal operations for this unit. In part of 1976, theunit was still in a startup mode and there were problemsthat forced extended outages. The Department does not

FIGURE 9LELAND OLDS 1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1974 1979 1984 1989 1994 1999

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

GLENHAROLD MINECLOSED (1993)

37

FIGURE 10LELAND OLDS 2

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1974 1979 1984 1989 1994 1999

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

GLENHAROLD MINECLOSED (1993)

38

FIGURE 11LELAND OLDS 1

0.000E+00

5.000E+12

1.000E+13

1.500E+13

2.000E+13

1975 1980 1985 1990 1995 2000

TO

TA

LH

EA

TIN

PU

T(B

TU

)

0.000E+00

5.000E+08

1.000E+09

1.500E+09

2.000E+09

2.500E+09

3.000E+09

3.500E+09

4.000E+09

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R(B

TU

)

TOTAL HEAT INPUT HEAT INPUT PER OPERATING HOUR

39

FIGURE 12LELAND OLDS 2

0.00E+00

5.00E+12

1.00E+13

1.50E+13

2.00E+13

2.50E+13

3.00E+13

3.50E+13

1975 1980 1985 1990 1995 2000

TO

TA

LH

EA

TIN

PU

T(B

TU

)

2.000E+09

2.500E+09

3.000E+09

3.500E+09

4.000E+09

4.500E+09

5.000E+09

5.500E+09

6.000E+09

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R(B

TU

)

TOTAL HEAT INPUT HEAT INPUT PER OPERTING HOUR

40

41

consider initial startup mode to be normal operations fora power plant. Based on the heat input data, theDepartment proposes that the 1977-78 time period isrepresentative of normal operations for the purpose ofcalculating the baseline emission rate.

The baseline emission rate for each unit was calculatedbased on the AP-42 emission factor 30(S) and data fromthe Annual Emission Inventory Reports for the facility.The data and results are as follows:

Unit YearCoal Usage(tons)

MineAvg. SulfurContent (%)

SO2Emissions(tons)

Hours ofOperation

2-Year Avg.EmissionRate

(lb/hr)

11

19761977

12559951306785

0.650.65

12246.012741.2

75537894

3235.2

22

19771978

19646602435160

0.650.65

19155.423742.8

66677445

6079.7

Where:

SO2 emissions (tons) = (30)(Mine Avg. Sulfur Content)(CoalUsage) ÷ (2,000 lb/ton)

2-Year Avg. EmissionRate (lb/hr) = [(2 year total SO2 Emission (tons))

(2,000 lb/ton)] ÷ [2 year total Hours ofOperation]

42

D. Lignite Gas Processing Plant

The Lignite Gas Processing Plant is located near Lignite,North Dakota in Burke County and is currently operated byBear Paw Energy, Inc. The plant was built by Texaco,Inc. in 1962 and was designed to process 21 millionstandard cubic feet of gas per day. The plant wasoriginally built with a 20 long ton per day two bed Claussulfur recovery unit with a design efficiency of 94percent. In 1971, the plant was processing about 8 to 9million cubic feet per day of gas. The processing ratedeclined to 6 million cubic feet per day in 1973, 4million cubic feet per day in 1976 and 2-3 million cubicfeet per day in 1977. The decline in gas volume wasprimarily due to declining production at the wells,deterioration of the gas gathering lines and the removalof several compressors.

In December of 1975, the sulfur recovery unitmechanically failed. Texaco stated that it was noteconomically feasible to repair the sulfur recovery unit.The plant was then operated without the sulfur recoveryunit. The plant was purchased by Darenco, Inc. inDecember of 1976. In September of 1977 the plant wassold to the Energy Operating Company (ENOPCO). The plantwas again sold in January 1980 to Cities Service, Inc.Cities Service replaced the sulfur recovery unit in late1983.

The original Permit to Operate for the facility wasissued in 1980 with an emission limit of 708 lb/hr fromthe acid gas flare. Based on the PSD rules in effect in1975, the shutdown of the sulfur recovery unit was notconsidered a major modification. In June 1983, a Permitto Construct was issued to Cities Service which allowedan increase in SO2 emissions from the flare to 1416 lb/hrprovided a new sulfur recovery unit was installed. OnJuly 11, 1983, the Department issued a Permit toConstruct for the sulfur recovery unit which limitedemissions to 217 lb/hr.

43

In the analysis for the June 1, 1983 Permit to Construct,a determination was made that the baseline emission ratefor the plant was 708 lb/hr. However, this determinationwas based on the potential to emit of the source (beforeair pollution controls) as of the major source baselinedate (January 6, 1975). On August 7, 1980, substantialchanges were made to the PSD rules. These changes renderthe earlier determination of the baseline emission rateinvalid. It is proposed that the baseline emission ratebe based on actual emissions on or near the minor sourcebaseline date.

Up until December 1975, sulfur dioxide emissions from theplant were very small because of the sulfur recoveryunit. An emissions inventory report for 1971 indicatedapproximately 40 tons of sulfur dioxide emissions. Areview of the facility by Pacific Environmental Servicesin early 1975 estimated sulfur dioxide emissions atapproximately 170 tons per year.

In a letter to the Department on February 26, 1976 (seeAppendix D), Texaco estimated that the average emissionsfrom the plant during 1976 would be 35.8 g/sec (284.1lb/hr). This was based on an average processing rate of3.60 x 106 SCFD and an H2S content of the inlet gas of1.12%. In early 1978, the Department conducted aninspection of the facility. The processing rate waslisted as 2-3 million cubic feet per day and the H2Scontent of the inlet gas was 1.64%.

Because of the frequent change in ownership near theminor source baseline, the future of the plant on thatdate was uncertain. If the plant were to continue tooperate, the future production rate was also uncertain.Therefore, the Department proposes that the 1976-77 timeperiod be used to calculate the baseline emission rate.Based on this information, the baseline emission rate wascalculated as follows:

44

SO2 Emissions (1976) = (3.6 x 106 scfd)(0.0112 ft3 H2S/ft3

gas) (1 lb-mole/379.4 ft3)(64 lbSO2/lb-mole)

= 6,801.5 lb/day= 283.4 lb/hr= 1,241.2 tons

SO2 Emissions (1977) = (2.5 x 106 scfd)(0.0164 ft3 H2S/ft3

gas) (1 mole/379.4 ft3)(64 lb SO2/lb-mole)

= 6,916.2 lb/day= 288.2 lb/hr= 1,262.2 tons

Avg. SO2 Emission Rate = (283.4 lb/hr + 288.2 lb/hr) ÷ 2= 285.8 lb/hr

45

E. Mandan Refinery

The Mandan Refinery was built in 1954 and iscurrently operated by Tesoro Refining and MarketingCompany. The plant is located along the MissouriRiver near Mandan in Morton County. The refineryoperates 24 hours per day, seven days a week and 52weeks per year. The refinery has processed up to65,000 barrels per day of crude oil with themaximum capacity unknown. On the minor sourcebaseline date, the refinery consisted of thefollowing source units that emitted significantamounts of sulfur dioxide:

Source Identification

Boilers 1, 2, 3Crude Furnace Crude FurnaceFluid CatalyticCracking Unit FCCU

Ultraformer Furnaces F-100, F1, F2, F3,F4, RegenerationFurnace

Alkylation Unit Furnaces B1, B2

The Department has no information about crude oilprocessing rates for the refinery. Recordsindicates that 1978 was a turnaround year for thefacility and is not considered normal operations.In 1978, the crude furnace was replaced with alarger unit now commonly referred to as the COfurnace. In addition, carbon monoxide from thefluid catalytic cracking unit was now routed tothis furnace. Since carbon monoxide has a heatingvalue when combusted, the amount of fuel oil andfuel gas combusted in this unit droppeddramatically from 1978-1980. With the drop in fuelusage, sulfur dioxide emissions also dropped fromthis furnace. Because of the replacement of thecrude furnace and the change in the method ofoperation of the refinery (i.e. routing off gas

46

from the FCCU to the CO furnace), the Departmentproposes that a time period before 1978 should beused to represent the baseline period. The onlyyear prior to 1978 for which data is available is1976. In a letter to the Department onSeptember 13, 2001 BP (former owner of therefinery) indicated they could not locate emissionsinventories for 1974, 1975 or 1977. They alsoindicated that 1976 data would be fairlyrepresentative of both 1976 and 1977. The oneexception that BP noted was for the Ultraformerfurnaces. BP indicated that the calculations forthe 1976 data, which the Department has on file,did not account for sulfur in the fuel gas from thedesulfizer hydrogen sulfide stripper on theUltraformer. BP estimated the Ultraformer fuel gascontained 500 ppmv H2S and emissions of sulfurdioxide were about 57 tons in 1976 for the heatersand furnaces at the Ultraformer.

The amount of fuel gas combusted in 1976 wasreported in units of standard cubic feet. Tesorohas provided information that indicates thatstandard temperature for the report was 32EF. Thisis the standard temperature used in chemistry butnot the temperature normally used for reporting gasvolumes for air pollution calculations. Based onTesoro’s information, a value of 359 cubic feet/lb-mole was used in the calculations as the molecularvolume of the fuel gas. The sulfur content of thefuel gas was also reported as percent by weight.This is also unusual since it is normally reportedas a volume percentage (or mole percentage). Theinitial data that was supplied for the oil that wascombusted provided the API Gravity (EAPI) from whichthe density of the oil can be calculated. Sincethe density of the oil is known, a mass balanceapproach to calculating emissions from the oil wasused instead of AP-42 emission factors.

47

Based on the data available, the Departmentproposes to use the 1976 data for the determinationof the baseline emission rates. The emissions foreach unit was calculated as follows:

Boilers

C Boiler 1 & 2 (identical fuel usage)

Fuel Gas Combusted - 423.5 x 106 scf (each)Avg. Sulfur Content of Fuel Gas - 1.43% (weight %)Avg. Molecular Wt. of Fuel Gas - 19.30 lb/lb-moleFuel Oil Combusted - 104,715 barrels (each)Sulfur Content of Fuel Oil - 1.66% (weight %)Avg. API Gravity - 11.26 EAPI (8.25 lb/gal)

SO2 (Fuel Gas) = (423.5 x 106 scf)(0.0143 lb S/lbgas)(19.30 lb gas/mole gas)(1 mole/359scf)(2 lb SO2 /1 lb S) ÷ (2,000 lb/ton)

= 325.6 tons (each)

SO2 (Fuel Oil) = (104,715 bbl)(42 gal/bbl)(0.0166 lb S/lboil)(8.25 lb/gal)(2 lb SO2/lb S) ÷(2,000 lb/ton)

= 602.3 tons (each)

Total SO2 = (325.6 + 602.3)(2 boilers)= 1,855.8 tons

C Boiler #3

Fuel Gas Combusted - 132.4 x 106 scfAvg. Sulfur Content of Fuel Gas - 1.43% (weight %)Avg. Molecular Wt. of Fuel Gas - 19.30 lb/lb-moleFuel Oil Combusted - 49,791 barrelsSulfur Content of Fuel Oil - 1.66% (weight %)Avg. API Gravity - 11.26 EAPI (8.25 lb/gal)

48

SO2 (Fuel Gas) = (132.4 x 106 scf)(0.0143 lb S/lbgas)(19.30 lb gas/mole gas)(1 mole/359scf)(2 lb SO2 /1 lb S) ÷ (2,000 lb/ton)

= 101.8 tons

SO2 (Fuel Oil) = ( 4 9 , 7 9 1 b b l ) ( 4 2 g a l / b b l ) ( 8 . 2 5lb/gal)(0.0166 lb S/lb oil)(2 lb SO2/lbS) ÷ (2,000 lb/ton)

= 286.4 tons

SO2 (Total) = 101.8 + 286.4 = 388.2 tons

SO2 (Total for all boilers) = 1,855.8 + 388.2= 2,244.0 tons

Crude Furnace

Fuel Gas Combusted - 837 x 106 scfAvg. Sulfur Content of Fuel Gas - 4.25% (weight %)Avg. Molecular Wt. of Fuel Gas - 20.3 lb/lb-moleFuel Oil Combusted - 69,216 barrelsSulfur Content of Fuel Oil - 1.66% (weight %)Avg. API Gravity - 11.29 EAPI (8.25 lb/gal)

SO2 (Fuel Gas) = (837 x 106 scf)(0.0425 lb S/lb gas)(20.3lb gas/mole gas)(1 lb-mole/359 scf)(2 lbSO2 /1 lb S) ÷ (2,000 lb/ton)

= 2,011.5 tons

SO2 (Fuel Oil) = ( 6 9 , 2 1 6 b b l ) ( 4 2 g a l / b b l ) ( 8 . 2 5lb/gal)(0.0166 lb S/lb oil)(2 lb SO2/lbS) ÷ (2,000 lb/ton)

= 398.1 tons

Total SO2 = 2,011.5 + 398.1= 2,409.6 tons

49

FCCU

AP-42 lists a sulfur dioxide emission factor of 493 lb/1,000barrels of fresh feed. However, this factor does not takeinto account any variation in the sulfur content of the oilfed to the FCCU and there is no explanation of the sulfurcontent on which the factor was derived. Therefore, theDepartment proposes that a mass balance approach is moreappropriate. In the 1976 data, the amount of coke burned offthe FCCU catalyst during regeneration and the sulfur contentof the coke are reported.

Coke Burned during Regeneration - 75379 tonsSulfur Content of Coke - 3.3% (wt. %)

SO2 = (75379 tons)(0.033 lb S/lb coke)(2 lb SO2/1 lb S)

= 4975.0 tons

Alkylation Unit Furnaces

Fuel Gas Combusted - 972 x 106 scfAvg. Sulfur Content of Fuel Gas - 1.95% (weight %)Avg. Molecular Wt. of Fuel Gas - 13.3 lb/lb-moleOil Combusted - 1,490 barrelsAvg. Sulfur Content of Oil - reported as 0%Avg. API Gravity - 22.3 EAPI (7.66 lb/gal)

SO2 (Fuel Gas) = (972 x 106 scf)(0.0195 lb S/lb gas)(13.3lb gas/mole gas)(1 lb-mole/359 scf)(2 lbSO2 /1 lb S) ÷ (2,000 lb/ton)

= 702.2 tons (each)

Ultraformers Furnaces

Total Fuel Gas Combusted - 1345.87 x 106 scfH2S Content - 500 ppmv (per BP 9/13/01)

50

SO2 = (1345.87 x 106 scf gas)(500 scf H2S/106 scf gas)(1

lb-mole/359 scf)(64 lb SO2/lb-mole)(1 mole SO2/moleH2S) ÷ (2,000 lb/ton)

= 60.0 tons

UnitsSO2 Emissions

(tons)

Total Hours ofOperation(all units)

Avg. Hours ofOperation(per unit)

2-year Avg.Emission

Rate(lb/hr)*

Boilers 1-3 2244.0 21624 7208 622.6

Crude Furnace 2409.6 8760 8760 550.1

FCCU 4975.0 8760 8760 1135.8

Alk. UnitFurnaces

702.2 17520 8760 160.3

UltraformerFurnaces

60.0 46986 7831 15.3

*Emission rate is the total for all units.

51

F. Wm. J. Neal Station:

The Wm. J. Neal Station (Neal Station) was located nearVelva in McHenry County. The facility, which wasoperated by Basin Electric Power Cooperative, beganoperation in 1952 and was shutdown in 1985. The facilityconsisted of two pulverized coal-fired boilers with anominal rating of 305.5 x 106 Btu/hr each. Each of theseunits had a separate stack; however, all data todetermine emission rates is reported as a total for thetwo units. For purposes of modeling, the plant has beentreated as one unit. For this analysis, the two unitsare treated as a single entity.

The Neal Station obtained its coal from the Velva Minewhich was located a few miles southwest of the plant.Figure 13 presents the coal sulfur content data submittedin the Annual Emission Inventory Reports for thefacility. Based on the data, a weighted average sulfurcontent was calculated to be 0.32%.

Figure 14 presents the heat input for the facility. TheDepartment proposes that the 1976-77 time periodadequately represents normal operations for the baselineperiod.

The baseline emission rate was calculated using the AP-42emission factor 30(S) and data included in the 1976 and1977 Annual Emission Inventory Reports. The data andresults are as follows:

YearCoal Usage(tons)

Mine Avg.SulfurContent(%)

SO2Emissions(tons)

Hours ofOperation(total)

Avg. Hours ofOperation

(per boiler)

2-Year Avg.EmissionRate

(lb/hr)

1976 249120 0.32 1195.8 14716 7358354.6

1977 327882 0.32 1573.8 16523 8262

52

Where:

SO2 emissions (tons) = (30)(Mine Avg. Sulfur Content)(CoalUsage) ÷ (2,000 lb/ton)

2-Year Avg. EmissionRate (lb/hr) = [1976 SO2 Emissions (tons) + 1977 SO2

Emission (tons)] (2,000 lb/ton) ÷[1976 Ag. Hours of Operation + 1977Ag. Hours of Operation]

FIGURE 13Wm. J. NEAL STATION

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

1975 1980 1985

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

53

FIGURE 14Wm. J. NEAL STATION

1.00E+12

1.50E+12

2.00E+12

2.50E+12

3.00E+12

3.50E+12

4.00E+12

4.50E+12

5.00E+12

1974 1976 1978 1980 1982 1984

TO

AL

HE

AT

INP

UT

(BT

U)

2.50E+08

3.50E+08

4.50E+08

5.50E+08

6.50E+08

7.50E+08

8.50E+08

9.50E+08

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R(B

TU

)

TOTAL HEAT INPUT HEAT INPUT PER OPERATING HOUR

54

55

G. Royal Oak Charcoal Briquette Plant

Royal Oak, Inc. operated a charcoal briquette plantsoutheast of Dickinson in Stark County until 1990. Theplant was built in the 1920's and was operated by severaldifferent companies. The plant initially produced adomestic heating briquette from lignite char. In 1961,Husky Oil Co. bought the plant and converted to theproduction of barbeque briquettes. The main processes atthe facility included lignite crushing, process steam andheat generation, carbonization, pyrite separation,grinding of the char, mixing with a binding agent,briquetting, drying the briquettes and bagging. The nameof the subsidiary that operated the plant was laterchanged to Royal Oak. The facility ceased operation in1990. Of the processes at the facility, only thesteam/heat generation and carbonization generatedsignificant amounts of sulfur dioxide.

Royal Oak obtained its coal from mines in the vicinity ofthe plant. The primary mine was located just south andeast of the plant. The coal that was mined wascharacterized by a high sulfur content, high ash contentand low heating value. Figure 15 presents the coalsulfur content data submitted for the facility in theAnnual Emission Inventory Reports. Based on this data,a weighted average sulfur content of 1.53% was calculatedfor the boilers and 1.45% for the carbonizers.

The steam and heat for the plant was provided by threespreader stoker lignite-fired boilers. Boilers 1 and 2had a nominal rating of 19.65 x 106 Btu/hr each andBoiler 3 had a rating of 57 x 106 Btu/hr. In 1984, theuse of the lignite-fired boilers was significantlyreduced by the installation of a waste heat boiler.

The carbonizer section of the plant consisted of twoLurgi carbonizers until 1976. The Lurgi carbonizers hada nominal rating of 150 tons of char per day. In July of1975, the Department received a Permit to Construct

FIGURE 15ROYAL OAK

0

0.5

1

1.5

2

2.5

3

3.5

1974 1979 1984 1989

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

56

57

application for the addition of a Herreschoff carbonizerfurnace. A Permit to Construct was issued in Septemberof 1975 and the unit started operation in July of 1976.Because of growing demand, Royal Oak applied for a Permitto Construct for a second Herreschoff carbonizer furnacein October of 1976. Due to a court decision, a Permit toConstruct was never issued and the unit began operationin October of 1978. The output of the two Herreschoffcarbonizers was limited to 288 tons per day by the Permitto Operate for the facility.

When determining normal operations for a facility, theDepartment has determined that production increases thatare reasonably anticipated on the baseline date should beincluded when determining the baseline concentration.Because Royal Oak had anticipated production increasesand had initiated or completed construction of two newcarbonizer furnaces to accommodate this productionincrease prior to the minor source baseline date, theDepartment proposes that a two year period after theminor source baseline date is more representative ofnormal operations for the facility. Figures 16 and 17present the coal usage for the boilers and carbonizerfurnaces. Figure 18 presents the total coal usage forthe facility. Because a waste heat boiler was installedin 1984 and heat input is not a value associated withcarbonizer furnaces, the determination of the period thatrepresents normal operations was based on total coalusage for the facility. The Department proposes that the1978-79 period adequately represents normal operation ofthe facility.

When calculating emissions from the source units, thethree boilers were considered one source and the fourcarbonizer furnaces were considered another source.Sulfur dioxide emissions for the boilers were calculatedbased on the AP-42 emission factor 30(S).

No emissions factors were available to determine sulfurdioxide emissions from the carbonizer furnaces. Stacktesting has been done to determine emissions from the

FIGURE 16ROYAL OAK BOILERS

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

1974 1976 1978 1980 1982 1984 1986 1988 1990

CO

AL

US

AG

E(T

ON

S)

58

FIGURE 17ROYAL OAK CARBONIZERS

0

50000

100000

150000

200000

250000

1974 1976 1978 1980 1982 1984 1986 1988 1990

CO

AL

US

ED

(TO

NS

)

59

FIGURE 18ROYAL OAK

BRIQUETTING PLANT

0

50000

100000

150000

200000

250000

300000

1974 1976 1978 1980 1982 1984 1986 1988 1990

TO

TA

LC

OA

LU

SA

GE

(TO

NS

)

60

61

Herreschoff carbonizers. This data was used to developan emission factor for use in the baseline emission ratecalculations.

The Department files contain two emissions tests forsulfur dioxide from the carbonizer furnaces. The firsttest, in 1983, was conducted in the stack of the aftercombustion chamber for the Herreschoff carbonizerfurnaces. The flue gas in the stack was characterized bysevere cyclonic flow and negative velocity pressure atsome sampling points. The testing that was conductedused a “blind man’s” approach. That is, no adjustment ofthe sampling nozzle was made for the angle of the flowwithin the stack. Because the cyclonic flow problems andthe measurement method, the data is considered unreliableby the Department.

In 1984, Royal Oak added a waste heat boiler to theircarbonizer system. The flue gas from the aftercombustion chamber was routed to this boiler to recoverheat and then transmitted to the atmosphere by way of aseparate stack. In late 1984, stack testing wasconducted to determine the sulfur dioxide emissions fromthe system. Based on the Department’s review of thetest, the data is considered valid and appropriate fordetermining emission rates. Based on the test results,an emission factor of 29.6(S) lb/ton was derived for theHerreschoff carbonizer furnaces. This factor is for onlyone test (3 runs) and does not account for the ligniteash sodium content. However, no better data exists fordetermining sulfur dioxide emissions. This emissionfactor was also used for the Lurgi carbonizer furnacessince no data on emissions was available for those units.

The data and calculations are summarized as follows:

62

Unit YearCoal Usage(tons)

Mine Avg.SulfurContent(%)

SO2Emissions(tons)

Hours ofOperation(total forall units)

Avg. Hoursof

Operation(per unit)

2-Year Avg.EmissionRate

(lb/hr)

Boilers 1-3 1978 19300 1.53 442.9 16800 5600172.1

Boilers 1-3 1979 21856 1.53 501.6 16128 5376

Carbonizers1-4

1978 177192 1.45 3802.5 21672 54181562.9

Carbonizers1-4

1979 226637 1.45 4863.6 23289 5822

Where:

SO2 (tons) = F (S)(coal usage) ÷ (2000 lb/ton)F = 30 for the boilersF = 29.6 for the carbonizers

2-Year Avg. Emission Rate = [(1978 SO2 Emissions + 1979 SO2Emissions)(2000 lb/ton)] ÷[1978 Avg. Hours of Operation +1979 Avg. Hours of Operation]

63

H. Tioga Gas Plant:

The Tioga Gas Plant, which is located in Williams Countynear the city of Tioga, is currently operated by AmeradaHess, Inc. The plant was originally constructed in 1954and had a nominal rating of approximately 110x106

scf/day. In 1967, a sulfur recovery unit was added whichhad a nominal rating of 150 long tons per day of sulfur.In 1975, the basic plant operations included compression,dehydration, fractionation, gas sweetening and sulfurrecovery. The only significant source of sulfur dioxidewas from the incineration of the tail gas from the sulfurrecovery unit.

The initial Permit to Operate application was submittedin April of 1978 and the permit was issued in June of1982. This permit limited sulfur dioxide emissions fromthe tail gas incinerator to 1,074 lb/hr.

In 1991, a new sulfur recovery unit was installed. Theunit started up in October, 1991. In September of 1992,the Permit to Operate was modified such that it reducedthe allowable SO2 emission rate from the tail gasincinerator to 575 lb/hr (24- hr avg.) and 671 lb/hr (1-hr avg.). This is the same limit currently in the Title5 Permit to Operate for the facility.

Emission estimates are available for the facility for1971, 1975, 1977, and 1979. All of these estimates list8760 hours of operation per year. The Department is notaware of any changes to the plant that were being plannedon the minor source baseline date that would havesignificantly affected plant operations or sulfur dioxideemissions. The Department proposes that a periodpreceding the minor source baseline date isrepresentative of normal operations.

The annual emission inventory report for 1971 estimatedemissions at 4,560 tons. This estimate is based on an

64

efficiency of the sulfur recovery unit of 95% and wascalculated as follows:

Sulfur in inlet gas - 45,600 tons

SO2 Emission(tons) = (45,600 tons)(1 - .95)(2 tonsSO2/(ton S)

= 4,560 tons SO2

In 1975, Pacific Environmental Services in a surveyreport for the facility listed the emission rate as 4,850tons. In the 1978 Permit to Operate application, sulfurdioxide emissions were estimated at 4849 tons based on1977 data. Both documents did not list sufficientinformation to verify the calculation and referenced amodeling analysis as the source of this estimate. The1979 annual Emission Inventory Report listed thefollowing:

Tail gas Incinerated - 2,528 x 106 scfSulfur content of Tail Gas - 0.00194 lb/scf

SO2 Emissions = (2,528 x 106 scf)(0.00194 lb/scf)(2 lb SO2/lbs) ÷ (2,000 lb/ton)

= 4,904.3 tons

The estimated emissions for the various years appear verysimilar. Since the 1975 and 1977 emissions estimates maybe from the same source, it is proposed that the 1971 and1977 data should be used to calculate the baselineemission rate as follows:

Avg. SO2 Emission Rate (tons) = (4,560 + 4849) ÷ 2

= 4,704.5

SO2 Emission (lb/hr) = [(4,560 + 4,849)(2,000 lb/ton)] ÷(8,760 + 8,760)

SO2 Emission (lb/hr) = 1,074

65

I. Williston Refinery:

The Williston Refinery was located in Williams Countyjust east of the City of Williston. The refinery, whichbegan operation in 1954, was last operated by Flying J,Inc. The facility ceased processing operations in 1984and none of the source units exist anymore. At itsclosing, the facility was capable of processing 5200barrels per day of crude oil.

The plant consisted of several boilers and heaters forprocessing the crude oil. The facility did not have acatalytic cracking unit or other process units that arenormally associated with larger refineries. Prior to1976, the heaters were primarily fired on pipelinequality natural gas. In 1975, Pacific EnvironmentalServices estimated sulfur dioxide emissions from thefacility to be less than 1 ton per year. In 1976, therewas a switch to the use of fuel gas and fuel oil to firethe boilers and heaters.

On April 1, 1977, Thunderbird Resources, Inc. (theoperator at that time) submitted a Permit to Constructapplication to the Department for improvements to therefinery. On December 29, 1977, the Department issued aPermit to Construct for those changes. Because ofeconomic conditions, some of the changes listed in thePermit to Construct were never completed. Therefore, itis proposed that these changes not be considered in thedetermination of the baseline emission rate.

In the 1977 Permit to Construct application, ThunderbirdResources, Inc. provided an estimate of emissions fromthe existing sources based on 1976 data. There is noother data available to estimate emissions from thefacility until 1982. The Department proposes that the1976 data is representative of normal operations. The1976 fuel usage data was presented as a total for allexisting units; however, a breakdown of emissions by unitis made. No sulfur content information is available forthe fuel gas; however, the sulfur content of the oil was

66

listed as 0.81%. The sulfur dioxide emissions estimates,as provided in the application are as follows:

1977 SourceUnits Identification

SO2 Emissions(lb/hr)

SO2 Emissions(tons)

Preflash Heater S-1 7.13 28.7

Crude Heater S-2 7.67 30.8

Thermal CrackingHeater

S-3 0.29 1.2

Charge Heater S-7 0.12 0.5

Reformer Heater S-9 0.46 1.9

Boiler 1 S-11 10.51 42.3

Boiler 2 S-12 10.51 42.3

Boiler 3 S-13 15.02 60.4

These emission rates are proposed to be the baselineemission rates for the Williston Refinery.

67

J. Stanton Station Unit 1:

The Stanton Station power plant is located along theMissouri River in Mercer County near Stanton, NorthDakota. The facility consists of two units referred toas Units 1 and 10. Unit 1 began operation in 1967 and isconsidered a baseline source. Construction of Unit 10began in 1980 with startup in 1982. The emissions fromUnit 10, without consideration of Unit 1, consume sulfurdioxide increment. Therefore, only Unit 1 is addressedin this analysis.

Unit 1 is a pulverized front wall fired unit with anominal rating of 1800 x 106 Btu/hr. The unit has nosulfur dioxide control equipment and vents to a commonstack with Unit 10. Unit 1, along with Unit 10, isoperated by Great River Energy.

Coal for the Stanton Station Unit 1 was obtained from theIndianhead Mine until 1992 when the mine closed. Coal iscurrently obtained from the Freedom Mine. Figure 19presents the coal sulfur content data submitted in theAnnual Emission Inventory Reports for the facility.Based on the data, the Department has calculated aweighted average sulfur content for coal obtained fromthe Indianhead Mine as 0.69%.

Figure 20 presents the heat input for Unit 1 based ondata from the Annual Emission Inventory Reports. Thegraph shows that the heat input to Unit 1 was muchgreater in 1970, 1972 and 1974 than subsequent years. OnFebruary 18, 1977, the Department received a letter (seeAppendix J) from United Power Association (operator atthe time) indicating that over the past few years theyhad experienced difficulty in supplying sufficient steamto operate the turbine at the Stanton Station at thecapacity level for which it was designed. The problemwas directly related to the sodium content of the fuel.The high sodium coal caused fouling of the boiler to thepoint it could not provide enough steam for an extended

FIGURE 19STANTON 1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1974 1979 1984 1989 1994 1999

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

INDIANHEAD MINECLOSED (1992)

68

FIGURE 20STANTON 1

4.00E+12

5.00E+12

6.00E+12

7.00E+12

8.00E+12

9.00E+12

1.00E+13

1.10E+13

1.20E+13

1.30E+13

1970 1975 1980 1985 1990 1995 2000

TO

TA

LH

EA

TIN

PU

T(B

TU

)

8.000E+08

1.000E+09

1.200E+09

1.400E+09

1.600E+09

1.800E+09

2.000E+09

2.200E+09

2.400E+09

2.600E+09

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R(B

TU

)

TOTAL HEAT INPUT HEAT INPUT PER OPERATING HOUR

69

70

period of time. United Power’s solution to the problemwas the construction of Unit 10. Unit 10 provides theadditional steam necessary to run the generator at fullcapacity. Based on United Power’s February 18, 1977letter and the data available to the Department, itappears that the period 1975-1977 is not representativeof normal operations. There are no two consecutive yearsprior to this period for which data is available.Therefore, it is proposed that the baseline period berepresented by the 1978-79 period.

The proposed sulfur dioxide baseline emission rate wascalculated using the AP-42 emission factor 30(s). Thedata and the results of the calculations are as follows:

YearCoal Usage(tons)

Mine Avg.Sulfur

Content(%)

SO2Emissions(tons)

Hours ofOperation

2-year Avg.Emission Rate

(lb/hr)

1978 577004 0.69 5972.0 54662132.1

1979 728136 0.69 7536.2 7205

Where:

SO2 (tons) = (30)(Mine Avg. Sulfur Content)(coalusage) ÷ (2000 lb/ton)

2-Year Avg. EmissionRate (lb/hr) = [(1978 SO2 + 1979 SO2)(2000 lb/ton)] ÷

[1978 Hours of Operation + 1979 Hours ofOperation]

71

K. M.R. Young Station:

The M.R. Young Station is located in Oliver Countyapproximately 5 miles southeast of Center, North Dakota.The facility consists of two units and is operated byMinnkota Power Cooperative. Unit 1 is a cyclone boiler

with a nominal heat input rating of 2500 x 106 Btu/hr.

Unit 2 is also a cyclone boiler with a nominal heat input

rating of 4696 x 106 Btu/hr.

Unit 1 began operation in 1970 and Unit 2 in March of

1977. Therefore, both sources are considered baseline

sources. Each boiler exhausts through a separate stackof different heights and are treated as separate units.Unit 2 is equipped with a wet scrubber for sulfur dioxidecontrol.

Coal for the M.R. Young Station has been obtained fromthe Center Mine since the startup of the facility. Themine is located adjacent to the plant. Figures 21 and 23present the coal sulfur content data submitted in theAnnual Emission Inventory Reports for the facility.Based on this data, a weighted average sulfur content wascalculated to be 0.77% for Unit 1 and 0.80% for Unit 2.

Figure 22 presents the heat input for Unit 1 based ondata in the Annual Emission Inventory Reports. The graphshows higher heat input per operating hour in the early1970's, lower levels in the mid 1970's and higher levelsagain in the late 1970's. The dip in heat input after1978 may be due to Unit 2 coming on line. With Unit 2providing capacity to the load area the demand on Unit 1was lessened. There is no data available to theDepartment for 1973; however, a two year period near thebaseline date could be selected which is indicative ofnormal operations. Because of the variation in heatinput per operating hour during the 1970's, it isproposed that the 1978-1979 period represents normaloperations for Unit 1.

FIGURE 21M.R. YOUNG 1

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1974 1979 1984 1989 1994 1999

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

72

FIGURE 22M.R. YOUNG 1

0.00E+00

5.00E+12

1.00E+13

1.50E+13

2.00E+13

2.50E+13

1972 1977 1982 1987 1992 1997

TO

TA

LH

EA

TIN

PU

T(B

TU

)

1.00E+09

1.50E+09

2.00E+09

2.50E+09

3.00E+09

3.50E+09

4.00E+09

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R(B

TU

)

TOTAL HEAT INPUT HEAT INPUT PER OPERATING HOUR

73

FIGURE 23M.R. YOUNG 2

0

0.2

0.4

0.6

0.8

1

1.2

1978 1983 1988 1993 1998

AV

G.S

UL

FU

RC

ON

TE

NT

(%)

74

75

For Unit 1, the baseline sulfur dioxide emission rate wascalculated based on the AP-42 emission factor 30(S) andthe data from the Annual Emission Inventory Reports. Thedata and the results of the calculations are as follows:

Unit Year

Coal Usage

(tons)

Mine Avg.

Sulfur

Content

(%)

SO2

Emissions

(tons)

Hours of

Operation

2-Year Avg.

Emission

Rate

(lb/hr)

1 1978 1427485 0.77 16587.5 67144649.9

1 1979 1508182 0.77 17419.5 7870

Where:

SO2 (tons) = (30)(Mine Avg. Sulfur Content)(CoalUsage) ÷ (2,000 lb/ton)

2-Year Avg. EmissionRate (lb/hr) = [(1978 SO2 Emissions (tons) + 1979

SO2 Emissions (tons))(2,000 lb/ton)]÷ [1978 hours of operations + 1979hours of operation]

Unit 2 represents a baseline source unlike any othersource in this analysis. Figure 24 presents the heatinput for the unit. The unit is equipped with a wetscrubber for SO2 control and was in operation only ninemonths prior to the minor source baseline date. Inaddition, there were significant problems with the sulfurdioxide removal system. The facility was the subject ofenforcement action by both the U.S. EnvironmentalProtection Agency and the Department due to variousproblems, including availability of the scrubber system.A May 9, 1978 letter to the Department documents theproblems associated with the scrubber to that date (SeeAppendix K). The Department files contain otherdocumentation of scrubber problems after this time.Since the unit was in a startup mode and only operatedfor approximately nine months in 1977, the Departmentdoes not consider 1977 to be indicative of normaloperations. The scrubber problems and other problemswith the boiler and generator limited operation of the

FIGURE 24M.R. YOUNG 2

0

5E+12

1E+13

1.5E+13

2E+13

2.5E+13

3E+13

3.5E+13

4E+13

4.5E+13

1975 1980 1985 1990 1995 2000

HE

AT

INP

UT

(BT

U)

3.00E+09

3.50E+09

4.00E+09

4.50E+09

5.00E+09

5.50E+09

6.00E+09

HE

AT

INP

UT

PE

RO

PE

RA

TIN

GH

OU

R

TOTAL HEAT INPUT HEAT INPUT PER OPERATING HOUR

76

77

unit to only 6,890 hours in 1978 or 79% of the availablehours. The heat input to the unit was only 61% of thenominal rating of the unit. This is contrasted with 1978when the unit operated 8064 hours or 92% of the availablehours. The heat input in 1978 was 82% of the nominalrating of the boiler. It appears that 1978 is notindicative of normal operations because of the somewhatlimited operation. Therefore, the Department proposesthat the 1979-80 time period be used to calculate thebaseline emission rate.

The Annual Emission Inventory Reports for Unit 2 for 1979and 1980 list emission rates that indicate noncompliancewith the allowable emission rate of 1.2 lb/106 Btu ofheat input. Therefore, the Department proposes tocalculate the baseline emission rate for this unit basedon the allowable emission rate (1.2 lb/106 Btu) and theactual heat input for the baseline period selected (1979-80). The data and results are as follows:

YearCoalUsage(tons)

Hours ofOperation

Avg. HeatContent

ofCoal

(Btu/lb)

TotalHeatInput(Btu’s)

AllowableEmission

Rate(lb/106

Btu)

SO2

Emissions(tons)

2-YearAvg.

EmissionRate

(lb/hr)

1979 2508465 8064 6736 3.379 x1013

1.2 20276.44905.6

1980 2410163 7571 6249 3.012 x1013

1.2 18073.3

Where:

SO2 Emissions (tons) = (Total Heat Input)(AllowableEmission Rate) ÷ (2000 lb/ton)

2-Year Avg. EmissionRate = [(1979 SO2 Emissions (tons) +

1980 SO2 Emissions (tons)(2000lb/ton)] ÷ [1979 Hours ofOperation + 1980 Hours ofOperation]

78

Baseline Emission RatesSummary

Source Unit

EmissionRate

(lb/hr)BaselinePeriod

Beulah Power Plant 1 & 23-5

137.1224.6

76-7776-77

R.M. Heskett Station 12

517.81208.0

76-7776-77

Leland Olds Station 12

3235.26079.7

76-7777-78

Lignite Gas Plant SRU Incinerator 285.8 76-77

Mandan Refinery Boilers 1, 2 & 3Crude Furnace

FCCUAlkylation Unit FurnacesUltraformer Furnaces

622.6550.11135.8160.315.3

76-77

Neal Station 1 & 2 354.6 76-77

Royal Oak BriquettingPlant

Boilers 1, 2 & 3Carbonizer Furnaces

172.11562.9

78-7978-79

Tioga Gas Plant SRU Incinerator 1074.0 71 & 77

Williston Refinery Preflash HeaterCrude Heater

Thermal Cracking HeaterCharge HeaterReformer Heater

Boiler 1Boiler 2Boiler 3

7.17.70.30.10.510.510.515.0

76

Stanton Station 1 2132.1 78-79

M.R. Young Station 12

4649.94905.6

78-7979-80

79

IV. Oil and Gas Wells Baseline Emission Rate Calculations

Estimating baseline SO2 emissions from oil and gas wellsinvolves similar difficulties as with estimating baselineemissions from other major SO2 sources. Oil and gas wellemissions must be calculated from other data, and the neededoil and gas data from the period 1976-77 are not very completeor reliable. First, emissions from oil and gas wells must becalculated from monthly values of wellhead gas produced andrelatively sparse data on the H2S content of the wellhead gas.

All oil and gas production data are collected and maintainedby the Oil and Gas Division of the North Dakota IndustrialCommission. The Oil and Gas Division has given the Departmentaccess to the well production data in order to track aircontaminant emissions from oil and gas wells. The Departmentcollects from well operators data on the H2S content of thewellhead gas and adds it to the Oil and Gas Division’sdatabase. In general, a well’s SO2 emission rate iscalculated by multiplying a monthly total of wellhead gasproduced by the percent of H2S in the gas and dividing by thenumber of days of production in the month (as well as someother conversion factors). Data from additional months may beadded together to obtain an average emission rate over severalmonths or a year.

Emissions of SO2 from oil and gas wells typically come fromtwo sources, either treaters or flares. Treaters separate thefluids in the crude oil for later transport or disposal.Flares burn the waste gas, which contains hazardous H2S gas,converting the H2S into less harmful SO2. Since about the mid-1980s, the Oil and Gas Division has been collecting wellproduction data on the amount of wellhead gas flared monthlyand the amount of wellhead gas used in firing the treater orother on-site equipment (lease use). These two monthly gastotals, amount of lease use and the amount flared, are useddirectly by the Department to calculate the SO2 emissions froma well’s treater and flare, respectively.

The Department has been requiring well operators to measureand report the H2S content (percent) of the wellhead gas since

80

the 1980s. However, such H2S measurements generally arerequired only when the well is first completed for productionor when it is recompleted into another geologic formation.Thus, there are generally only one or two values of H2S, atmost, for each well. The H2S data are generally notconcurrent with the gas production data. In addition, manyolder wells have either unreliable H2S data or no H2S data atall. In these cases, it was necessary to substitute for themissing H2S data from a nearby similar well.

The Oil and Gas Division considers its gas production databack to about 1987 to be very reliable, but considers gasproduction data before 1987 to be at least somewhatunreliable. The earlier gas production data are consideredsomewhat unreliable because, until about the early to middle1980s, the main product from the wells was oil, and the gaswas considered a waste product. It was important toaccurately keep track of the valuable oil production for thebenefit of the owners, operators, and the State. However,because the gas was considered a waste product to be disposedof, data on gas production were not consistently and reliablyreported or recorded until the importance of the gas changedin the mid-1980s.

Some data on total wellhead gas production are available backto 1976-77, the two years before the baseline date, but inmost cases the data represent the amount of gas sold to a gasprocessing plant. The amount of wellhead gas used in thetreaters or flared which is necessary for emissionscalculations was not consistently reported before the mid-1980s and was available for only a very few wells in 1976-77.

Because of the unreliability and the lack of gas productiondata before the mid-1980s, the Department did not consider thedata to be good enough for calculating oil and gas wellbaseline emissions. The recent needs for more completely andaccurately determining the state’s compliance status withrespect to Class I PSD increments has necessitated theDepartment reevaluate calculating the baseline emissions ofoil and gas wells. The problem was obtaining reasonablycomplete, reliable gas production data appropriate for 1976-

81

77. The Department’s strategy has been to use the oldestavailable gas production data that is considered complete andreliable and apply it to the wells producing during the period1976-77. Based on the Oil and Gas Division’s judgment thatgas production data before 1987 were not completely reliable,the Department is proposing to use 1987-1988 gas productiondata to calculate the baseline emission rate for wells thatexisted during the 1976-77 time period.

The Department has calculated SO2 emissions for oil and gaswells for 1987-88 that were used in a regional air qualitystudy conducted in 1989-90. The Department conducted a studynamed the “Williston Basin Regional Air Quality Study” (WBS,1990) based on using air quality dispersion models to predictwhat SO2 concentrations were occurring in and around numerousoil and gas well fields in North Dakota.

The Williston Basin Study processed oil and gas H2S andproduction data for all North Dakota wells producing duringthe period November 1987-March 1988. Modeled SO2concentrations were compared to the State and Federal AmbientAir Quality Standards (AAQS) and Class II and Class I PSDincrements. The Department examined impacts in or near 12worst-case oil and gas fields and in four Class I areas inNorth Dakota. Software programs were written and executed byDepartment personnel to calculate SO2 emission rates for alloil and gas wells producing gas during November 1987-March1988 and output the oil and gas well source data in a formatready for input to an air quality dispersion model. Theprograms use a mass balance approach to calculating sulfurdioxide emissions and are based on the assumption that all ofthe H2S is converted to SO2.

With SO2 emissions data calculated for 1987-88, the challengewas in applying this data to wells producing during 1976-77.Increases or decreases in gas production back to 1977 couldnot be reliably and consistently determined, but the data from1987-88 could be applied to the wells producing during 1976-77. Basic identifying information, such as the wells’ names,file numbers, field names, and locations of all wellsproducing during 1976-77 were extracted from the Oil and Gas

82

Division database. Only wells that actually existed and wereproducing during 1976-77 were accepted in a new 1976-77emissions inventory by using this information as a startingpoint. For wells that existed in both time periods, the 1987-88 emissions data were copied over from the WBS inventory tothe new inventory. For wells that produced in 1977 but not in1987, there were no emissions data directly available. Theaverage emission rate over all wells in the same field fromthe WBS was calculated and substituted into the new inventoryfor each well of this type, where data were available. Incases where data were needed for a 1977 well in a field thatdid not produce at all during 1987-88, then a field-averageemission rate from a similar, nearby field was added to thenew inventory for all such wells.

A problem with using the WBS data “as is” is that much of thegas produced during 1987-88 was sold to gas processing plantsand not flared. Many of these gas plants didn’t exist in1976-77, so that all of that sold gas would have been flaredotherwise if not sold. In cases where a 1977 well was notselling gas to a gas plant, all of the gas produced at a wellin 1987, except for lease-use gas, would have been flared in1977. Applying the WBS data to these wells involved addingthe 1987-88 sold gas to the flared gas amount beforecalculating the flare emission rate. These recalculated flareSO2 emission rates for the 1976-77 wells were often muchhigher than the original WBS flare emission rates because ofthe sold gas being included. Any wells that did not sell gasin 1976-77 but did sell gas in 1987-88 would be assigned ahigher flare emission rate including the sold gas. Wells thateither sold gas in both periods or did not sell gas in bothperiods were assigned a flare emission rate unchanged by soldgas amount because the sold gas was already accounted forconsistently.

The remaining task was to determine which wells were sellinggas in 1976-77. Gas processing plants receive their gas fromoil and gas wells through a pipeline, or gas-gathering system,that connects the wells to the gas plant. The gas-gatheringsystem may be relatively small and serve only one oil and gasfield, or it may be large and extensive, connecting to many

83

fields over large distances. In 1977, there were two largegas plants and two smaller gas plants operating in NorthDakota. The Tioga Gas Plant, near Tioga, was connected to thelargest gas-gathering system in the state at the time,reaching from about 20 miles north of Tioga southward tosoutheastern McKenzie County southeast of Watford City. TheTioga pipeline connected to at least 20 separate oil and gasfields. The Lignite Gas Plant, near Lignite, was connected toa somewhat smaller gas-gathering system serving about ninefields in Burke County. Two small gas plants, the Red WingCreek Gas Plant and the Boxcar Butte Gas Plant in westernMcKenzie County, also were operating in 1977 and received gasfrom only two isolated fields.

All other fields in the rest of the state were not served byany gas-gathering systems in 1977 and could not have sold gas.In particular, later development in west-central North Dakotabetween Williston and Belfield triggered the construction ofat least three additional gas plants and an extensive gas-gathering system in this area by 1987. Many wells in thispart of western North Dakota were selling gas by 1987 where nogas plants existed in 1977.

When applying the above procedure for flare emission rates, itwas assumed that any field connected to a gas-gathering systemin 1977 was selling gas in 1977. Therefore, all wells infields connected to the Tioga and Lignite gas-gatheringsystems and the Red Wing Creek and Boxcar Butte gas plantswere assumed to be selling gas both in 1976-77 and 1987-88 andso were assigned flare emission rates from the WBS inventoryunchanged by sold gas. However, many other wells in westernNorth Dakota were not connected to gas-gathering systems in1977 and thus got credit for higher flare emissions in 1977because of sold gas in 1987-88. The result of this procedurewas an SO2 emissions inventory for all oil and gas wellsproducing in 1976-77 that reflected gas production levels backto 1987, using the earliest reliable gas production data, andappropriately accounted for gas sold to gas processing plants.

Appendix A

Beulah Power Plant

Data

Appendix B

R.M. Heskett Station Data

Appendix C

Leland Olds Station Data

Appendix D

Lignite Gas Plant Data

Appendix E

Mandan Refinery Data

Appendix F

Wm. J. Neal Station Data

Appendix G

Royal Oak

Briquetting Plant Data

Appendix H

Tioga Gas Plant Data

Appendix I

Williston Refinery Data

Appendix J

Stanton Station Unit 1 Data

Appendix K

M.R. Young Station Data