Environ. Sci. Technol. 2010, 44, 2709–2714
Carbon Dioxide Emission Factors for U.S. Coal by Origin and Destination JEFFREY C. QUICK* Utah Geological Survey, P.O. Box 146100, Salt Lake City, Utah 84114-6100
Received September 8, 2009. Revised manuscript received February 6, 2010. Accepted March 10, 2010.
This paper describes a method that uses published data to calculate locally robust CO2 emission factors for U.S. coal. The method is demonstrated by calculating CO2 emission factors by coal origin (223 counties, in 1999) and destination (479 power plants, in 2005). Locally robust CO2 emission factors should improve the accuracy and verification of greenhouse gas emission measurements from individual coal-fired power plants. Based largely on the county origin, average emission factors for U.S. lignite, subbituminous, bituminous, and anthracite coal produced during 1999 were 92.97, 91.97, 88.20, and 98.91 kg CO2/ GJgross, respectively. However, greater variation is observed within these rank classes than between them, which limits the reliability of CO2 emission factors specified by coal rank. Emission factors calculated by destination (power plant) showed greater variation than those listed in the Emissions & Generation Resource Integrated Database (eGRID), which exhibit an unlikely uniformity that is inconsistent with the natural variation of CO2 emission factors for U.S. coal.
flue gas method to be more reliable than those calculated using emission factors, whereas the European Union prefers the mass balance method using emission factors or measured carbon contents (8). Ackerman and Sundquist (9) compared results from the flue gas method and emission factor method. They observed that total emissions calculated from flue gas measurements at 828 U.S. electric power plants were 1.1% higher than those calculated using emission factors. Results for individual power plants showed less agreement; the average difference (from the mean of both results) was (17%. The comparatively small (1.1%) difference between total emissions suggests that the methods have a comparable accuracy. The larger difference for individual power plants ((17%) has little significance for national inventories but is important for a market-based emission reduction rule. For example, consider a 1500 MW coal-fired power plant that emits 10 million tons of CO2 per year. At a market CO2 price of $10 per ton, and a 5% emission measurement error, this plant would realize an annual gain, or loss, of five million dollars. Winschel (10) observed that CO2 emission factors for 519 commercial North American coal samples naturally vary by about 10%. Causes of this variation include differences in coal rank, sulfur content, maceral abundance, and mineral content (Supporting Information). Regardless of the cause, the natural variation of CO2 emission factors likely contributes to the large ((17%) difference between CO2 emissions calculated using the flue gas or emission factor methods. In this paper, plant-specific emission factors were determined according to the county origin of the coal. The proposed emission factors are intended to more closely match the natural variation of U.S. coal than emission factors developed for national inventories.
Method Introduction Accurate and verifiable measurements of greenhouse gas emissions are required for an equitable and efficient emission reduction rule. This paper is intended to improve the accuracy and verification of CO2 emission measurements from U.S. coal-fired power plants. Limitations of CO2 emission factors used for these measurements are discussed. A method to calculate locally robust emission factors is demonstrated by calculation of CO2 emission factors for 479 large U.S. power plants. Carbon dioxide emissions from coal combustion are proportional to the mass and carbon content of the coal. Although the tons (Mg) and heating value (GJ/Mg) of coal shipped to U.S. power plants are reported, the carbon content is not (1, 2). Accordingly, CO2 emissions reported in national U.S. inventories (3, 4) are calculated by multiplying the coal energy content (GJ) by a CO2 emission factor (kg CO2/GJ). The resulting national emission estimate is thought to be accurate to within 5% (5). Local, plant-specific CO2 emissions from most large U.S. coal-fired power plants can be calculated from flue gas volume and CO2 concentration measurements, which are reported to the U.S. Environmental Protection Agency (EPA) (6). These plants, which are responsible for about 85% of CO2 emissions from U.S. coal, must use this calculation method to comply with the EPA greenhouse gas reporting rule (7); other coal consumers may report results calculated using emission factors. The EPA considers results from the * Corresponding author e-mail:
[email protected]; phone: 801537-3372; fax: 801-537-3400. 10.1021/es9027259
2010 American Chemical Society
Published on Web 03/16/2010
Data. In-ground coal quality data, including C, S, ash, fixed carbon, and heating values, are from COALQUAL (11), IGS (12), and Keystone (13, 14). Year 1999 county coal production is from EIA (15). Year 1999 coal origin, tonnage, heating value, ash, and S values are from FERC-423 (1), ICR (16), and FERC580 (17). Year 2005 data for delivered coal are from FERC423 (1) and EIA-423 (2). Data used in this study were reported in U.S. customary units as pounds, short tons, and British thermal units. Study results were converted to SI units shown in the manuscript. With noted exceptions, heating values and emission factors are expressed on a gross (higher) energy basis. Calculation of Emission Factors by County-of-Origin. Selected COALQUAL data records were used to calculate county-average heating values and carbon contents of the in-ground coal. Where possible, these in-ground heating values and carbon contents were adjusted to comport with the S content of coal produced from these counties. Carbon dioxide emission factors (kg CO2/GJ) for coal produced from 223 counties were calculated according to kg CO2 /GJ )
1000 C × × 3.6642 MJ/kg 100
(1)
where, MJ/kg is megajoules per kilogram coal, C is wt.% carbon of the coal, and 3.6642 is the gravimetric factor to convert C to CO2. Sulfur Adjustment. The necessity for this adjustment is illustrated in Figure 1, which confirms Winschel’s observation (10) that CO2 emissions decline with increasing coal S. Sulfur values for in-ground U.S. coal (COALQUAL data) average VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2709
for coal from these counties were estimated using empirical relationships between ASTM rank parameters (21), S contents, and CO2 emission factors shown in Figure 2. For high volatile A bituminous and lower rank coal, the CO2 emission factor was estimated as kgCO2 /GJ ) 108.73 - 1.647 × Sef - 1.085 × MJ/kgm,mmf + 2 1.450E-2 × MJ/kgm,mmf (5)
FIGURE 1. Carbon dioxide emission factors for 86 coals from Fayette County, Alabama (b, thick best-fit line) decrease with increasing sulfur. Correction of these emission factors to a sulfur-free basis (O, thin best-fit line) does not fully account for this decrease (eqs 2, 3, and 4 in text). Data from ref 11. 2.0% Sdry, whereas sulfur values for coal shipped to power plants during 1999 averaged 1.4% Sdry (18). Unless this difference is accounted for, emissions factors calculated from unadjusted COALQUAL data will systematically underestimate CO2 emissions from U.S. power plants. To account for the difference between the S content of the in-ground and produced coal, the in-ground C and MJ/ kg values were adjusted to match the S content of the produced coal. First, an adjusted S value (Sadj) was calculated as
Sadj ) Sprod ×
(
) )
Sin-grd 100 × 100 100 - Sin-grd Sprod 100 - 9.42 × × 100 100 - Sprod (2)
MJ/kgin-grd - 9.42 ×
(
MJ/kgprod
where Sprod is the wt.% S of the produced coal, MJ/kgin-grd is the average heating value for the in-ground coal by origin county, MJ/kgprod is the heating value for coal produced from that county, 9.42 is the MJ/kg contribution from sulfur (19, 20), and Sadj is the wt.% S required for the in-ground coal to have the same S emission factor as the produced coal. Changing the wt.% S in coal will change the wt.% C. Accordingly, an adjusted C content (Cadj) was calculated as Cadj ) Cin-grd ×
(100 - Sadj) (100 - Sin-grd)
(3)
where Cin-grd is the county-average wt.% C of in-ground coal, Sin-grd is the corresponding in-ground wt.% S, and Sadj is from eq 2. Because S contributes about 9.42 MJ/kg S, changing the wt.% S also changes the heating value. Accordingly, an adjusted heating value (MJ/kgadj) was calculated as
(
MJ/kgadj ) MJ/kgin-grd - 9.42 ×
)
Sin-grd 100 - Sadj × + 100 100 - Sin-grd Sadj 9.42 × (4) 100
Results from eqs 3 and 4 (Cadj and MJ/kgadj) were used with eq 1 to calculate S-adjusted, CO2 emission factors for coal produced from 182 of the 223 counties considered in this study. Emission factors for 20 counties that lack produced coal quality values were necessarily calculated using unadjusted, in-ground MJ/kg and C values. Estimation Formulas. Twenty-one coal-producing counties lack values for in-ground coal C content. Emission factors 2710
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
where Sef is the S emission factor (kg S/GJ) for the produced coal, and MJ/kgm,mmf is the moist, mineral-matter-free MJ/ kg rank parameter (21) for produced coal. Equation 5 was derived by multivariate regression on 129 county-average values; it has an adjusted R2 of 0.88, a standard error of 0.54 kg CO2/GJ, and the S, MJ/kg, and MJ/kg2 variables are highly significant with t-statistics of -15, -8.2, and 5.8 (respectively). For medium volatile bituminous and higher rank coal the CO2 emission factor was estimated as kgCO2 /GJ ) 144.45 - 0.7647 × Sef - 1.6521 × FCd,mmf + 2 1.228E-2 × FCd,mmf (6)
where Sef is the S emission factor (kg S/GJ) for the produced coal, and FCd,mmf is the dry, mineral-matter-free fixed carbon rank parameter (21). Because fixed carbon is not reported for produced coal, county-average FCd,mmf values for the inground coal were used. Equation 6 was derived by multivariate regression on 26 county-average values where FCd,mmf was greater than 64%. The equation has an adjusted R2 of 0.97, a standard error of 0.48 kg CO2/GJ, and t-statistics for the S, FC, and FC2 variables of -2.0, -8.2, and 9.6 (respectively). Calculation of Emission Factors by Destination. With a few changes, the method used to calculate emission factors for coal produced during 1999 was also used to calculate emission factors for coal delivered during 2005 to 479 power plants (>50MW) in 46 states. The most significant change was the use of both FERC-423 (1) and EIA-423 (2) data, which together provide a better estimate of coal distribution than the FERC-423 data alone. After calculating emission factors for each coal shipment, the resulting data set (23118 records) was sorted by power plant, and an average, weighted heating value, ash yield, S content, and CO2 emission factor was calculated for coal delivered to each power plant.
Results and Discussion Emission factors for coal produced during 1999 by countyof-origin are illustrated in Figure 3, together with the resulting CO2 emissions aggregated by origin state. Emission factors for coal delivered to 479 power plants during 2005 are illustrated in Figure 4, together with the resulting CO2 emissions aggregated by destination state. Sulfur. Figure 2 shows that the effect of S is clearly diminished at high rank (above high volatile bituminous). Less clear is why the effect of S varies. To investigate, I adjusted heating values and C contents for high volatile bituminous coals from Fayette County, Alabama, to a S-free basis using eqs 3 and 4 (Sadj ) 0), and used the results with eq 1 to calculate their S-free, CO2 emission factors. Figure 1 shows that the S-free CO2 emission factors for Fayette County coal still vary with S content. Indeed, only half of the variation of the CO2 emission factor with coal S can be attributed directly to S; the other half is due to something else. Most S in U.S. coal is present in the mineral pyrite (FeS2) or bound to organic matter. Minor amounts of S in sulfate minerals such as CaSO4 · 2H20 are not considered here. Considering heats of formation for FeS2 and combustion products SO2 and Fe2O3 (∆Hf° ) -167.4, -825.5, and -296.8 kJ/mol, respectively), one kg of pyritic S produces 13.08 MJ.
FIGURE 2. Average CO2 emission factors for coal from 147 U.S. counties vary with ASTM rank parameters and coal sulfur content. The effect of sulfur is diminished above high volatile bituminous rank. Lines correspond to eqs 5 and 6 (see text).
FIGURE 3. Left: Carbon dioxide emission factors for coal produced from 223 counties in 26 states during 1999 (b, county average; O, state average). Right: CO2 emissions from produced coal assuming complete combustion. Ignoring the uncertain heats of formation for organic S compounds, similar calculations show 1 kg of organic S produces 9.26 MJ. Consequently, less pyritic S in high rank coal might explain the diminished significance of S at high rank shown in Figure 2. However, the relative abundance of pyritic S does not vary significantly with coal rank; sulfurform assays for coal from 134 counties show pyritic S accounts for 49% of the S in lower rank coal and 47% of S in higher rank coal. Something other than the relative abundance of pyritic S must be responsible for the diminished significance of S at high rank that is shown in Figure 2.
Organic S in coal originates during early diagenesis (22) when, in the absence of ferrous iron, H2S from sulfatereducing bacteria combines with otherwise-labile lipids such as fatty acids or alkenes (23). This natural vulcanization process produces H-rich organic S compounds that resist further microbial degradation and consequently enrich the coal-forming peat in H. Hydrogen in coal reduces the CO2 emission factor (24). Thus, preservation of H associated with the microbial origin of organic S during early diagenesis results in a lower CO2 emission coal. This mechanism is consistent with the results shown in Figure 1, where only VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2711
FIGURE 4. Left: Carbon dioxide emission factors for coal delivered to 472 U.S. power plants in 46 states during 2005 (b, power plant average; O, state average, seven plants not shown). Right: CO2 emissions from delivered coal assuming complete combustion. half of the reduction of the CO2 emission factor can be attributed to the thermal contribution from S. The other half is due to the contribution from H associated with organic S. During subsequent coalification, this associated H is progressively lost as aliphatic S forms decompose to form aromatic thiophenes (25). Consequently, the diminished significance of S at high rank (Figure 2) is explained by the loss of H associated with organic S during coalification. Coal Rank. The EPA greenhouse gas inventory (4) lists CO2 emission factors for coal specified by coal rank class (lignite, subbituminous, bituminous, and anthracite). These rank-specific factors are used for facility-level reporting of CO2 emissions from U.S. power plants (7, 26). Given the local variation of CO2 emission factors shown by previous studies (27, 28), it is worth asking if rank-specific emission factors are suitable for local inventories. For example, examination of 2082 bituminous Kentucky coals led Sakulpitakphon et al. (28) to reject the notion that a single CO2 emission factor can “be used as typical for any given rank of coal.” Results shown in Figure 5 amplify their concern. Figure 5 shows greater variation of CO2 emission factors within coal rank classes than between rank classes for commercially produced U.S. coal. Consequently, accurate estimates of CO2 emissions from individual coal-fired power plants cannot be assured using CO2 emission factors specified by rank. 2712
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
FIGURE 5. Greater variation of CO2 emission factors is observed within rank classes than between rank classes. Circles show year 1999 average values for coal-producing counties. Stars show weighted values for lignite, subbituminous, bituminous, and anthracite rank classes (92.97, 91.97, 88.20, and 98.91 kg CO2/GJ, respectively). Uncertainty and Reliability of Emission Factors. Variation (double standard deviation) within rank classes il-
lustrated in Figure 5 suggest that emission factors for the lignite, subbituminous, bituminous, and anthracite rank classes are reliable to within 2.93, 1.60, 2.49, and 6.42 kg CO2/GJ, respectively, which is better than a single factor for all 212 counties (average 89.58, (5.03 kg CO2/GJ). Emission factors for 18413 coal shipments from 148 counties were calculated with eq 5 to evaluate the reliability of emission factors specified by origin county. The results suggest that emission factors specified by county of origin are typically reliable to within 0.41 kg CO2/GJ; the reliability ranged between (0.08 and (1.80 kg CO2/GJ for 95% of these counties. Other sources of uncertainty are less easily quantified. For example, the S adjustment (described above) increased the county-specific factors by an average +0.09 kg CO2/GJ, with the result ranging from about +1.22 to -0.65 kg CO2/GJ for individual counties. However, as discussed, the effect of H associated with S limits the effectiveness of this theoretical adjustment. Although the empirical eqs 5 and 6 better account for the combined effect of S and H, results from these equations are less reliable than emission factors specified by county origin. Standard errors for eqs 5 and 6 indicate that they are reliable to within 1.09 and 0.96 kg CO2/GJ, respectively. Instances where these equations fail to achieve these reliability thresholds help to explain their limited reliability. For example, the county-specific CO2 emission factor for bituminous coal from Montrose County, Colorado, is 92.56 kg CO2/GJ, whereas eq 5 predicts a significantly lower emission factor (88.31 kg CO2/GJ). Maceral analyses for Montrose County coal (29) report abundant inertinite; these hydrogen-poor macerals are consistent with the higher, county-specific emission factor (27, 28). Conversely, abundant hydrogen-rich liptinite macerals in coal from Webb County, Texas (30), are consistent with the relatively low county-specific emission factor (84.48 kg CO2/GJ), rather than the higher value predicted by eq 5 (89.29 kg CO2/GJ). Although the maceral assemblages in these coals are unusual compared to most other U.S. coal (31), these examples suggest that more subtle and common variation of maceral content limit the reliability of eqs 5 and 6. Unlike emission factors specified by geographic (county) origin, which can be plainly known, emission factors specified by coal rank class require reference to a standard rank classification method, such as ASTM D388 (21). Practical problems, such as seasonal desiccation of produced coal and lack of assay data, add uncertainty when such methods are used to establish coal rank (32, 33). This uncertainty extends to the derivation of rank-specific emission factors. For example, the rank-specific CO2 emission factors listed by the EPA (4) were derived using rank designations listed in COALQUAL, which the U.S. Geological Survey has emphatically stated are not reliable (34). The reliability of rank designations shown in Figure 5 are likewise uncertain and are based on multiple data sets using methods and assumptions described in the Supporting Information. Regardless of whether these methods and assumptions are proper or valid, the inherent uncertainty associated with rankspecific emission factors can be can be avoided by specifying emission factors by county origin. Although CO2 emission factors specified by county origin are more reliable and certain than those specified by rank class, measuring the C content of coal shipments is simpler and likely more reliable. Reporting the C content of coal shipments should not be difficult for most large U.S. power plants. Indeed, considering their existing efforts to measure coal shipment tonnage, collect a representative coal sample, and determine the heating value, ash yield, and S content, the cost of an additional C analysis is minimal. Comparison with eGRID Emission Factors. Figure 6 compares plant-specific CO2 emission factors from this study
FIGURE 6. Carbon dioxide emission factors for coal shipped to 282 large power plants during 2005 (this study) are poorly correlated with year 2005 CO2 emission factors reported with eGRID data (6). Except for a few outliers and plants that burn lignite, emission factors from eGRID are uniformly near 88.22 kg CO2/GJ. Data show plants >50 MW where >99% of the electric generation was attributed to lignite (∆), subbituminous (-), or bituminous ( ×) coal, and where eGRID emission factors are based on measurements made with continuous emissions monitors. One plant is off-scale. with those listed in the Emissions & Generation Resource Integrated Database (eGRID) for 282 power plants with continuous emissions monitors (6). Given the natural variability of U.S. coal, the remarkably uniform CO2 emission factors reported in eGRID are unlikely. Comparison of CO2 emissions from these power plants calculated using emission factors from this study, with those listed in eGRID, is informative. Total emissions calculated using eGRID emission factors were 1.6% less than those calculated using emission factors from this study. Differences for individual plants varied by more than 5%; eGRID results ranged from 4.4% less CO2 to 1.4% more CO2 for 95% of the plants. For these calculations, the same plant-specific energy inputs (GJ) were used so the differences are due entirely to the emission factor. Given the large ((17%) difference between CO2 emissions calculated from mass-balance and flue gas measurements observed by Ackerman and Sundquist (9) other causes, besides the CO2 emission factor, must also contribute to this difference. Notably, methods to measure plant energy consumption (GJ) have not been considered in this study. Plant energy consumption can be calculated from coal tonnage and heating value measurements or from flue gas volume measurements using rank-specific F-factors (35). Comparison of these methods might suggest ways to further improve the accuracy and verification of CO2 emission measurements from U.S. coal-fired power plants.
Acknowledgments Kate Ackerman, Barbara Toole-O’Neil, and Richard (Dick) Winschel provided comment on early drafts. Comment from four anonymous reviewers is also acknowledged and appreciated.
Supporting Information Available Data used in this study are described and discussed. This material is available free of charge via the Internet at http:// pubs.acs.org.
Literature Cited (1) U.S. Energy Information Administration. Form FERC-423 Database, Monthly Cost and Quality of Fuels for Electric Plants Data; EIA: Washington, DC, 2007; www.eia.doe.gov/cneaf/ electricity/page/ferc423.html. VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2713
(2) U.S. Energy Information Administration. Form EIA-423 Database, Monthly Nonutility Fuel Receipts and Fuel Quality Data; EIA: Washington, DC, 2007; www.eia.doe.gov/cneaf/electricity/ page/eia423.html. (3) U.S. Energy Information Administration. Emissions of Greenhouse Gases in the United States, 2005; EIA: Washington, DC, 2006; www.eia.doe.gov/oiaf/1605/ggrpt/index.html. (4) U.S. Environmental Protection Agency. Inventory of U.S. Greenhouse Gases and Sinks 1990-2005, Report 430-R-07002; EPA: Washington, DC, 2007; www.epa.gov/climatechange/ emissions/usinventoryreport.html. (5) U.S. Energy Information Administration. Documentation for Emissions of Greenhouse gases in the United States 2006, Report DOE/EIA-0638; EIA: Washington, DC, 2008, www.eia.doe.gov/ oiaf/1605/ggrpt/documentation/pdf/0638(2006).pdf. (6) U.S. Environmental Protection Agency. eGRID2007, v.1.1, Year 2005 Plant Data; EPA: Washington, DC, 2009; www.epa.gov/ cleanenergy/energy-resources/egrid/index.html. (7) U.S. Environmental Protection Agency. Mandatory Reporting of Greenhouse Gases, Final Rule, Federal Register, October 30, 2009, pp 56260-56519; EPA: Washington, DC, 2009; www. epa.gov/climatechange/emissions/ghgrulemaking.html. (8) Evans, S.; Deery, S.; Bionda, J. How Reliable Are GHG Combustion Calculations and Emission Factors? Proceedings of the Air and Waste Management Association, First International Greenhouse Gas Measurement Symposium, San Francisco, CA, March 2224, 2009, p 12. (9) Ackerman, K. V.; Sundquist, E. T. Comparison of two U.S. powerplant carbon dioxide emissions data sets. Environ. Sci. Technol. 2008, 42, 5688–5693. (10) Winschel, R. A. The relationship of carbon dioxide emissions with coal rank and sulfur content. J. Air Waste Manage. Assoc. 1990, 40, 861–865. (11) Bragg, L. J.; Oman, J. K.; Tewalt, S. J.; Oman, C. L.; Rega, N. H.; Washington, P. M.; Finkelman, R. B. Coal Quality (COALQUAL) Database, version 2.0; Open-File Report 97-134; U.S. Geological Survey, Energy Resources Program: Reston, Virginia, 1997; www.energy.er.usgs.gov/products/databases/CoalQual/ intro.htm. (12) Illinois Geological Survey. Coal Mines, Coal Geology and Resource Data; Illinois Geological Survey: Champaign, IL, 2007; htwww.isgs.uiuc.edu/maps-data-pub/coal-maps/coal-qualitynonconf.xls. (13) Keystone. 1975 Keystone Coal Industry Manual; McGraw-Hill, Inc: New York, 1975. (14) Keystone, 2001 Keystone Coal Industry Manual, ISBN 0-92953151-5; Intertec Publishing Corp.: Chicago, IL, 2001. (15) U.S. Energy Information Administration. Coal Industry Annual EIA: Washington, DC, 1999; www.eia.doe.gov/cneaf/coal/page/ acr/backissues.html. (16) Quick, J. C.; Tabet, D. E.; Wakefield, S.; Bon, R. L. Optimizing Technology to Reduce Mercury and Acid Gas Emissions from Electric Power Plants: A GIS Study of Coal Chemistry, DOE contract DE-FG26-03NT41901; Utah Geological Survey: Salt Lake City, UT, 2005; www.geology.utah.gov/emp/mercury/index.htm. (17) U.S. Energy Information Administration. The Coal Transportation Rate Database; EIA: Washington, DC, 2003; www.eia.doe. gov/cneaf/coal/ctrdb/database.html. (18) Quick, J. C.; Brill, T. C.; Tabet, D. E. Mercury in US coal: Observations using the COALQUAL and ICR data. Environ. Geol. 2003, 43, 247–259.
2714
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
(19) Mott, R. A.; Spooner, C. E. The calorific value of carbon in coal: The Dulong relationship. Fuel 1940, 19, 226-231, 242-251. (20) Given, P. H.; Weldon, D.; Zoeller, J. H. Calculation of calorific values of coals from ultimate analyses: Theoretical basis and geochemical implications. Fuel 1986, 65, 849–854. (21) American Society for Testing and Materials. Method D388-90, Standard Classification of Coals by Rank, Annual Book of ASTM Standards, v.05.05; 193-197, ASTM: West Conshohocken, PA, 1990. (22) Casagrande, D. J. Sulfur in Peat and Coal. In Coal and Coal Bearing Strata: Recent Advances, Geological Society; Scott, A. C., Ed.; Blackwell Scientific: London, 1987; Special Publication 32, pp 87-105. (23) Sinninghe Damste, J. S.; Rijpstra, W. I. C.; Kock-Van Dalen, A. C.; deLeeuw, J. W.; Schenk, P. A. Quenching of labile functionalized lipids by inorganic sulfur species: Evidence for the formation of sedimentary organic sulfur compounds at the early stage of diagenesis. Geochim. Cosmochim. Acta 1989, 53, 143–155. (24) Quick, J. C.; Glick, D. C. Carbon dioxide from coal combustion: Variation with rank of U.S. coal. Fuel 2000, 79, 803–812. (25) George, G. N.; Gorbaty, M. L.; Kelemen, S. R.; Sansome, M. Direct determination and quantification of sulfur forms in coals from the Argonne premium sample program. Energ. Fuel. 1991, 5, 93–97. (26) The Climate Registry. General Reporting Protocol for the Voluntary Reporting Program (draft): TCR: Los Angeles, CA, 2007. (27) Quick, J. C.; Brill, T. C. Provincial variation of carbon emissions from bituminous coal: Influence of inertinite and other factors. Int. J. Coal Geol. 2002, 49, 263–275. (28) Sakulpitakphon, T.; Hower, J. C.; Taulbee, D. N. Predicted CO2 emissions from maceral concentrates of high volatile bituminous Kentucky and Illinois coal. Int. J. Coal Geol. 2003, 54, 185–192. (29) Tremain, C. M.; Wynn Eakins, A.; Schultz, J. E.; Kelso, B. Colorado Coal Quality Data, Information Series 58; Colorado Geological Survey: Denver, CO, 2000; www.geosurvey.state.co.us/pubs/ gis_data/mresources/CGS_IS58.xls. (30) Warwick, P. D.; Hook, R. W. Petrography, geochemistry, and depositional setting of the San Pedro and Santo Tomas coal zones: Anomalous algae-rich coals in the middle part of the Claiborne Group (Eocene) of Webb County, Texas. Int. J. Coal Geol. 1995, 28, 303–342. (31) Waddell, C.; Davis, A.; Spackman, W.; Griffiths, J. C. Study of the Interrelationships among Chemical and Petrographic Variables of United States Coal, Report FE2030-TR9; Pennsylvania State University: University Park PA, 1978; p 240. (32) Mathai, C. V. Comment from Arizona Public Service Co. to EPA proposed rulemaking on mercury emissions from coal-fired power plants; 2005; Regulations.gov: docket OAR-2002-00565494.pdf. (33) Harja, J. Comment from Utah Resource Development Coordinating Committee to EPA on the new source performance standards for mercury emissions from coal-fired power plants; 2005; Regulations.gov: docket OAR-2002-0056-6510.pdf. (34) U.S. Geological Survey. Definitions of Parameters in the U.S. Geological Survey Coal Quality Database; USGS: Reston, VA, 1997; www.energy.er.usgs.gov/products/databases/CoalQual/ Docs/techinfo.pdf. (35) U.S. Environmental Protection Agency. Consolidated Part 72 and 75 Regulations; volumeII;EPACleanAirMarkets:Washington,DC,2008; www.epa.gov/airmarkets/emissions/consolidated.html.
ES9027259
