Prog. Energy Combust. Sci. 1990, Vol. 16, pp. 243-251 Printed in Great Britain. All rights reserved.
0360-1285/90 $0.00 + .50 © 1990 Pergamon Press plc
B E H A V I O R OF BASIC E L E M E N T S D U R I N G C O A L C O M B U S T I O N G. P. HUFFMAN, F. E. HUGGINS,NARESHSHAHand ANUP SHAH CFFLS. University of Kentucky, 233, Mining and Mineral Resource Building, Lexington, KY 40506-0107, U.S.A. Abstract--X-ray absorption fine structure (XAFS) spectroscopy, Mossbauer spectroscopy and computercontrolled scanning electron microscopy (CCSEM) have been used to investigate the reactions of Ca, Fe and alkalies in combustion systems. Ca may either transform to a CaO fume that reacts with SO2 to form CaSO4, or may react with clays, quartz and other minerals to form slag droplets, or flyash. Similarly,pyrite may devolatilize and oxidize exothermically to form molten or partially molten iron sulfide-iron oxide mixtures, or may react with other minerals to become part of the slag. Alkalies in lignites (principally Na) volatize and may react with either SO2 to form sulfates or with clay minerals (principally kaolinite) to form aluminosilicate slag droplets. K in bituminous coal is contained in illite which melts and becomes part of the slag phase. The calcium and alkali sulfates and the iron-rich species are observed to be concentrated in the initial layers of deposits, while the complex aluminosilicate slag droplets collect to form an outer glassy layer. CONTENTS
1. Introduction 2. Experimental Procedures 3. Experimental Results 3.1. Calcium 3.2. Iron 3.3. Alkalies 4. Summary and Discussion Acknowledgements References
243 243 244 244 245 247 248 250 250
briefly below, and with the observations of numerous other investigators.H2
1. I N T R O D U C T I O N
The basic elements in coal ash (alkalies, Ca, Fe) exhibit an interesting dual behavior during combustion. Not surprisingly, this dual behavior is closely related to the way in which these elements are contained in the coal being burned. In this paper, some recent experimental results on the behavior of basic elements in combustion systems are reviewed and qualitative models of their combustion reactions are discussed. It should be emphasized that our intention is not to present a comprehensive review of the rather extensive literature on this topic. For the reader interested in such a review, we recommend the excellent article by Reid, l and the book by Raask, 2 as well as a number of articles dealing with the combustion behavior of iron, 3-5 calcium6-J° and the alkali elements. 1°-12Furthermore, our discussion deals only with high-temperature ( ~ 1600-2000K), pulverized coal combustion and is not relevant to fluidized bed combustion. The goal of the current article, therefore, is to discuss some systematic trends in the behavior of calcium, iron and the alkalies during hightemperature, pulverized coal combustion. These trends are consistent with our own data, summarized
2. E X P E R I M E N T A L P R O C E D U R E S
Much of the data on which the current paper is based is summarized in detail in several recent and forthcoming publications.13-19 Bituminous coals and lignites and a variety of combustion products obtained from both pilot scale combustion rigs and drop-tube furnaces have been investigated. The principal characterization techniques used were XAFS spectroscopy, Mossbauer spectroscopy and computer-controlled scanning electron microscopy (CCSEM). Detailed discussion of these experimental methods are given elsewhere. 19-24 Combustion products from ten different coals have been investigated. Combustion was carried out at several scales in four different systems: two drop-tube furnaces,25'26a small combustor27firing 2.5 kg of coal per hr, and a pilot-scale combustion facility (24mBTU/hr) at the Combustion Engineering Corporation Kreisinger Development Laboratory. 2s Both flyash and deposit samples were investigated. CCSEM was used to determine the mineralogy of all coals investigated. The results for the weight per-
G.P. HUFFMANet aL T A B L E 1. C C S E M
Quartz Kaolinite Illite Chlorite Montmorillonite M i x e d silicates
W. Kentucky # 9*
for ten coals (wt. % of total mineral
W. K e n t u c k y # 11
Illinois # 6*
Eagle Butte 40 13 2
10 7 35 1 1 19
12 6 23 -< I 34
19 8 16 -I 23
20 6 10 < 1 < I 18
12 30 1 . 1 13
Pyrite Jarosite F e 2÷ s u l f a t e s Gypsum
18 < 1 ---
16 < 1 --
24 -< I 1
28 -2 1
Fe-rich minerals Ca-rich minerals Ti-rich minerals
I 2 < 1
3 < 1
< I 4
1 7 --
Misc., mixed 100
Kentucky # 9*
12 2 -.
14 2 3 .
Illinois # 6*
17 8 10
47 3 3
26 3 2 --
4 -< 1 --
29 5 2 8
22 2 I 2
< 1 I --
4 3 I
< 1 8
*Samples of Kentucky # 9 and Illinois # 6 from two different sources were investigated. centages of the total mineral matter observed for the discrete mineral phases in each coal are tabulated in Table 1. Note that CCSEM does not determine the amount of carboxyl-bound calcium and sodium contained in the lignites and subbituminous coals. More detailed information on the inorganic constituents in these coals is given elsewhereJ 5-js'2s'29 In the current paper, our objective is not to present a detailed discussion of the combustion reactions of the inorganics in these coals. That will be done in other papers. Rather, the objective is to present a general description of the behavior of the basic elements in coal: iron, calcium and the alkalies, potassium and sodium. As will be seen, these elements exhibit an interesting dual type of behavior in the combustion process that reflects different forms of occurrence of the element in different coals. 3. EXPERIMENTAL
The principal experimental observations for the dominant basic elements are summarized below.
In lignite, calcium is molecularly dispersed in the coal macerals and is bonded to the oxygen anions in carboxyl groups, while in bituminous coals it is predominantly contained in the discrete mineral, calcite. A systematic change from carboxyl-bound calcium to calcite is observed with increasing rank. 3° In a recent study, 13 the reactions and transformations of carboxyl-bound calcium during pyrolysis in nitrogen and gasification in oxygen and carbon dioxide was investigated. Temperatures ranged from approximately 700 to 1300 K at atmospheric pressure, while reaction times varied from approximately 0.1 sec to I hr. It was
concluded on the basis of XAFS spectroscopy data that the molecularly dispersed calcium in lignite agglomerates and, eventually, forms CaO. '3 A recent XAFS study of the combustion products formed in a pilot-scale combustion rig at the Combustion Engineering Kreisinger Development Laboratory from a Texas lignite and an Ilinois #6 bituminous coal detected two major forms of Ca, Ca incorporated in aluminosilicate glass and CaSO4.17 Solids rapidly extracted from within and above the flame contained all Ca in glass. In both superheater and waterwall deposits, glass was the dominant Ca-bearing phase in the outer (fireside) deposits, while CaSO4 was dominant in the inner or initial deposits. In work currently underway, XAFS and CCSEM are being used to investigate the forms of Ca in combustion products from drop-tube furnance (DTF) tests 26'29 and a small combustion system. 27'29 Most of the coals in Table 1 are being studied, as well as a number of synthetic coals. Most of the Ca in these ash samples is contained in aluminosilicate glass derived from molten slag droplets, but a significant fraction appears to be present as CaO. Typical XAFS data are shown in Fig. 1 for samples of Beulah flyash derived from a small combustor test. The X-ray absorption near edge structure (XANES, top Fig. 1) is similar to Cabearing aluminosilicate glass XANES observed previously but exhibits pre-edge features that may be due to CaO. This is also true for the radial structure function (RSF), derived by Fourier transformation of the extended X-ray absorption fine structure (EXAFS), shown in Fig. 1 (bottom). The major peak is principally the contribution of the nearest-neighbor oxygen shell in the glass phase, while the smaller peak at approximately 3.3 A is believed to arise from the first Ca shell in the CaO component (the peak at 1 A is a mathematical artifact due to truncation of the transform). Note that that the RSF has not been
Basic elements during coal combustion 1.4 1,3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
Energy (eV) 0
seen that there are a substantial number of Ca-rich (principally CaO) particles, but aluminosilicate glass particles with a fairly broad range of Ca contents are the dominant species. These glass particles are derived from molten slag droplets formed primarily by the reaction of carboxyl-bound Ca and Na and pyrite with kaolinite and quartz during combusion of Beulah coal. In Fig. 3, a similar CCSEM ternary diagram is shown for a Beulah ash sample produced in a D T F test. It is seen that more Ca-rich particles (presumably CaO) are present, possibly because of less time in the DTF for reaction of CaO with aluminosilicates to occur. 3.2. Iron
FIG. 1. Typical XANES (top) and RSF (bottom) obtained from a Beulah ash sample.
phase-shift corrected, and the peaks therefore occur at distances slightly less than the true interatomic spacings. CCSEM is being used to examine the compositional variation of the ash particles. Typical results for ash produced during small combustor runs on Beulah lignite are shown in Fig. 2. Here each point represents an ash particle identified by the CCSEM analysis as containing > 80% Ca + Si + AI. The composition of each particle, normalized to three elements, is then plotted on the ternary diagram. It is
Pyrite and its oxidation products are the dominant iron-bearing phases in both low and high-rank coals, although bituminous coals frequently contain substantial amounts of iron in clays and siderite.-'°'2~ If pyrite is contained in a burning coal particle that also contains clay minerals and/or quartz, it is likely to react with these minerals to form an aluminosilicate slag droplet. Mossbauer spectroscopy has demonstrated that much of the iron in flyash samples quenched from within or above the flame of a pilot scale combustor, and much of the iron in the outer layers of deposits, is contained in aluminosilicate glass. ~s~6 Typical spectra are shown in Fig. 4. The labels denote peaks derived from magnetite (M), hematite (H), ferric glass (G3) and ferrous glass (G2). These samples were obtained from a combustion test on a Kentucky No. 9 coal performed at the Combustion Engineering Fireside Test Performance Facility (FTPF). 28 Pyrite particles that are isolated within a coal particle or that are liberated behave much differently.
FIG. 2. Ternary Ca-Si-AI diagram derived from CCSEM data for Beulah ash particles from a small combustor.
FIG. 3. Ca-Si-AI diagram derived from CCSEM data for Beulah ash particles from a drop tube furnace combustion experiment.
101 100 99 98
94 N-FLAME SO
92 100 98
96 94 2: o_ r/i (n Z c¢ I-
98 95 M
VELOCITY,m m / s e c
FIG. 4. Typical spectra 57Fe Mossbauer spectra obtained from ash and deposit samples in the Combustion Engineering FTPF. Recently, Mossbauer spectroscopy and CCSEM measurements were conducted to determine the transformation products formed from pyrite removed from coal in drop-tube furnace tests at gas tempertures of 1311-1727 K and residence times from 0.07 to 1.2 sec. in a 95% N 2: 5% 02 atmosphere. 's Magnetite is the dominant oxide formed, while pyrrhotite
(FeI_~S) is the dominant sulfide. Typical iron phase percentages determined by Mossbauer spectroscopy for samples derived from DTF tests on pure pyrite are shown as a function of residence time in Fig. 5. These tests were conducted at 1560 K on uniformly sized (75-90 #m) pyrite in an atmosphere containing 3% 02 and are reported on in more detailed elsewhere in this
Basic elements during coal combustion
Hydrotherm,,I pyrne 75/80/sm
100 90 80 70 60
50 40 30 20 10 0 0
Ru. time ( ~ )
FIG. 5. Iron phase percentages as a function of residence time for samples from DTF tests on pyrite (1560K, 3%02, 75-90/1m). volume. 3LCCSEM results show evidence of extensive melting, even at the lowest gas temperature. This is consistent with a thermodynamic model 32 which yields particle temperatures significantly in excess of the gas temperature because of the exothermic oxidation of pyrrhotite. These results are also consistent with the observation of iron-rich initial layers observed in deposits 3'4'15''6 that appear to have been formed from the impaction of highly viscous liquid droplets. The inner layer is transformed to hematite over a period of time at the lower temperatures within the deposit near the waterwall or the superheater tubes (Fig. 4, middle spectrum).
3.3 Alkalies The principal alkali species in lignite is Na, believed to be molecularly dispersed through the macerals and bonded to the oxygen anions in carboxyl groups. K is low in abundance but can be inserted into carboxyl bound sites in the coal macerals by cation exchange. In bituminous coals, K is the dominant alkali and it is contained almost exclusively in the clay mineral illite. Illite has a strikingly different XAFS spectrum from that of cation-exchanged K in lignite) 4 Na in bituminous coals is usually contained as NaC1, frequently in solution in moisture adsorbed in coal pores and capillaries.33 The behavior of alkalies during combustion is strongly dependent on their form in the coal. Carboxyl bound alkalies or alkalies in solution volatilize. Several volatile species are possible, but NaOH is the most likely for lignite) z K in illite is likely to remain with the aluminosilicate particle and form a 3PECS 16:4-E
molten stag droplet. Partial melting occurs at 9001000°C because of eutectic regions in the K:O-SiOzAI203 phase diagram.19'34 Melting is readily confirmed by potassium XANES, such as that shown in Fig. 6, which was obtained from a KY ~# 11 ash sample collected on an impactor in the University of Arizona small combustion system. As discussed elsewhere, ~'3s the aluminosilicate glass produced by melting illite yields precisely this type of potassium XANES spectrum. The composition of these molten particles is similar to that of the parent illite, as illustrated by Fig. 7. These are CCSEM ternary diagrams for ash particles from a D T F test on Upper Freeport coal and for clay minerals in the coal; each point represents an ash particle containing > 80% of K + Si + A I on the basis of the EDX spectrum. It is evident that the combustion process has produced little alteration in the composition of the molten slag particles derived from illite. A more detailed discussion of the behavior 1.5 1.4 1,3 1.2 1.1 1 0.9 O.B 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
Energy ( e V )
FIG. 6. XANES of KY # 11 coal ash from University of Arizona combustor.
G.P. HUFFMANet al. SI
Upper Freeport Coal
Upper Freeport Coal DTF Ash
FIG. 7. K-Si-AI diagrams derived from CCSEM data for Upper Freeport coal and ash from a DTF experiment. of iilite in combustion systems is given in another paper in this issue) 6 Volatile alkalies have a strong tendency to be adsorbed by the aluminosilicates derived from clays ~2"~9and significantly lower the melting point of the resulting slag. This is illustrated by Fig. 8 which shows the ternary Na-Si-A1 diagram generated from CCSEM data on ash particles collected during combustion of Beulah lignite in a drop-tube furnace. Each data point indicates the normalized three-element composition of a particle identified in the CCSEM analysis as containing > 80% Na + Si + AI. The Ca-Si-AI diagram of Fig. 3 is from the same sample. It is seen that the NaO2-SiO2-AI203 slag particles fall into a relatively small range of compositions, close to that of nepheline, which is the first phase expected to crystallize out from a NaO-Ai203-SiO2 melt. 81
FIG. 8. Na-Si-A1 diagram derived from CCSEM data for Beulah ash particles from a drop tube furnace combustion experiment.
When volatile alkalies reach the upper regions of the furnace they react with SO2 to form molten alkali sulfates that react strongly with metal superheater tubes. 37 In a limited set of experiments, we have used XAFS and CCSEM to determine that superheater deposits from a boiler firing North Dakota lignite contained a mixture of alkali sulfate and alkalicontaining glass. 38 Similar results have been reported by Tufte and BeckeringJ ° 4. S U M M A R Y AND D I S C U S S I O N
On the basis of the results summarized above, qualitative models of the behavior of the major basic elements during combustion can be deduced. These models are indicated in the schematic diagrams of Fig. 9. The reactions and transformations indicated are appropriate for high-temperature, pulverized coal combustion. Flame temperatures for most of the experiments on which our data is based were typically 1750-2000 K, while gas temperatures in the vicinity of the deposit surfaces was in the range of approximately 1100-1400 K. More detail on combustion conditions is given elsewhere. '8''9 Calcium in lignite is molecularly dispersed and bonded to carboxyl groups in the macerals. It rapidly agglomerates into C a - O clusters and CaO, while the calcite in bituminous coal devolatilizes to form CaO. The CaO may be given off from the burning coal particles as a CaO fume or may remain with the particles and react with the clay, quartz and other minerals to form a molten aluminosilicate slag. The CaO fume apparently reacts with SO2 to form CaSO4 which is part of the initial, presumably sticky, layer in both superheater and waterwall deposits./7 The precise sticking mechanism is not clear at this point. Moreover, it should be noted that it cannot be
Basic elements during coal combustion
SO2 LiEnice, Ca-Carboxyl
Bituminous, > CaO Calcite (Cat03)
> CaSO 4
> Slag (flyash)
< - - E x o V.heraic oxidac J.on...-->
SOz > NaOH (Na20 , ochers) L£guice Na-carboxyl
t clays, quartz, etc.
> Slag (flyash)
FIG. 9. Schematic diagram indicating the behavior of basic elements in combustion systems. concluded from our data exactly where sulfation occurs. Pyrite that is liberated or isolated alone in a coal particle devolatizes to form pyrrhotite which undergoes exothermic oxidation that raises the particle temperature sufficiently far above the gas temperature to exceed the melting point of magnetite.Is'32 The molten iron-rich particles impinging on the walls and tubes may be iron sulfideiron oxide mixtures or magnetite, depending on residence time and temperature. Because of the high
momentum of these particles, they readily impinge on the walls and tubes, forming a major part of the initial sticky layer for pyrite-rich coals. Those pyrite particles that are contained in coal particles containing clay, quartz and other minerals react to form molten aluminosilicate slag or flyash droplets which stick on top of the initial iron-rich layer. There is less good experimental data on alkalies in combustion systems. Nevertheless, some speculations can be made. It appears that sodium is bound to carboxyl groups in lignite and readily volatilizes,
G.P. HUFFMANet al.
presumably to N a O H , although other volatile species are certainly possible. Sodium reacts strongly with clay (mainly kaolinite in lignite) and quartz, both within the burning coal particle and in the vapor phase to form aluminosilicate slag droplets or flyash. The volatile N a O H also reacts with SO2 to form NaSO4 which has a low melting point (884°C) and becomes part of the initial sticky layer. Finally, potassium in bituminous coal is contained in illite which melts and becomes part of the slag or flyash component of the ash stream. While the simple diagrams of Fig. 9 are consistent with the general trend of slagging and fouling behavior observed in high-temperature, pulverized coal combustion systems, it is clear that we have ignored many reactions and made a number of speculations for which there is no firm experimental proof. The current 'model', therefore, should really be considered as a point of departure for future discussion and experimentation.
Acknowledgements--This research was sponsored by the U.S. Department of Energy under DOE Contract No. DEAC22-86PC90751. We would also like to acknowledge a number of colleagues who performed the combustion experiments that produced the samples used for this research: S. Srinivasachar, J. J. Helble and A. A. Boni of Physical Sciences, Inc.: T. W. Peterson and J. O. L. Wendt of the University of Arizona; A. A. Levasseur, J. Durant, and O.K. Chow of Combustion engineering; and A. F. Sarofim of MIT.
I. REID, W. T., Coal ash--Its effect on combustion systems, Chemistry of Coal Utilization, Second Supplementary Volume, M.A. Elliot (Ed.), pp. 1389-1445, John Wiley, New York (1981). 2. RAASK, E., Mineral Impurities in Combustion, Hemisphere (1985). 3. BR~RS, R. W., The physical and chemical characteristics of pyrites and their influence on fireside problems in steam generators, J. Engng Pwr 98, 517 (1976). 4. BORIOR. W. and NARCISO, R. R., The use of gravity separation techniques for assessing slagging and fouling potential of coal ash, ASME Paper 78-WZ/CO-3 (1978). 5. HUWMAN,G. P., HUGGINS,F. E. and DUNMYRE,G. R., Fuel, 60 585 (1981). 6. PROCTER,N. A. A. and TAYLOR,G. H. J., Fuel, 39, 284 (1966). 7. R1NDT, D. K., JONES, M. L. and SCHOBERT, H. H., Investigation of the mechanism of ash fouling in lowrank coal combustion, Fouling and Slagging Resulting from lmpuritites in Combustion Gases, pp. 17-35, Proc. 1981 Engng Found. Conf., R. W. Bryers (Ed.), (1983). 8. KALMANOVICHn. P. and WILLIAMSON,J., Crystallization of coal ash melts, Mineral Matter and Ash in Coal pp. 234-255, K. S. Vorres (Ed.), ACS Symposium Series No. 301 (1986). 9. KALMANOVICH,D. P., ZYGARLICKE,C. J., STEADMAN,E. N. and BENSON,S. A.,ACS Division of Fuel Chemistry Preprints, 34, (2), 318-329 (1989). 10. TUFTE, P. H. and BECKERING W., ASME J. Engng Power, p. 407 (July, 1975). 1!. HALSaXAD,W. D. and RAASK,E., J. Inst. Fuel, 42, 344 (1969).
12. WALLT. F. and WIBBERLEY,L. J., Sticky particles and deposit formation, Fouling and Slagging Resulting from lmpuritites in Combustion Gases, pp. 493-514, Proc. 1981 Engng Found. Conf., R. W. Bryers (Ed.), (1983). 13. HUGGINS,F. E., SHAH,N., HUFFMAN,G. P., JENKINS,R. G., LYTLE,F. W. and GREEGOR,R. B., Fuel, 67, 938-942 (1988). 14. HUFFMAN,G. P., HUt~GINS,F. E. and SHAH,N., EXAFS spectroscopy: A new technique for investigating impurity elements in coal that control slagging and fouling behavior, Engineering Foundation Conference on Mineral Matter and Ash in Coal, Santa Barbara, CA, 1988, K. S. Vorres and R. W. Bryers, (Eds), to be published. 15. HUGGINS,F. E., HUFFMAN,G. P. and LEVASSEUR,A. A., Ash Deposits from Raw and Washed Coals, ibid. 16. HUGGINS,F. E., HUFFMAN,G. P. and LEVASSEUR,A. A., ACS Division of Fuel Chemistry Preprints, 33, (2), 73-80 (1988). 17. HUFFMAN,G. P., HUGGINS, F. E., LEVASSEUR,m. A., LYTLE,F. W., GREEGOR,R. B. and MEHTA,A., Fuel, 68, 238-242 (1989). 18. HUFFMAN,G. P., HUGGINS F. E., LEVASSEUR,A. A., CHOW, O. and SR]NlVASACHAR,S., Fuel, 68, 485-490 (1989). 19. HUFFMAN,G. P., HUOG]NS,F. E., SnOENBERGER,R. W., WALKER,J. S., LYTLE,F. W. and GRF~GOR,R. B., Fuel, 65, 621-632 (1986). 20. HUFFMAN, G. P. and HUGGINS, F. E., Fuel, 57, 592 (1978). 21. HUGGINS, F. E. and HUFFMAN, G. P., M6ssbauer analysis of the iron-bearing phases in coal, coke, and ash, Analytical Methods for Coal and Coal Products, Vol. III, Chapter 50, pp. 371-423, Clarence Karr, Jr. (Ed.), Academic Press (1979). 22. HUGGINS,F. E., KOSMACK,D. A., HUFVMAN,G. P. and LEE, R. J., Coal mineralogies by SEM automatic image analysis, Scanning Electron Microscopy1980/I, pp. 531540. SEM Inc., AMF O'Hare, Chicago. 23. HUGGINS, F. E., HUFFMAN, G. P. and LEE, R. J., Scanning Electron Microscopy-Based Automatic Image Analysis (SEM-AIA) and Mossbauer Spectroscopy: Quantitative Characterization of Coal Minerals, Coal Products: Analytical Characterization Techniques, pp. 239-258, E. L. Fuller, Jr. (Ed.), ACS Symposium Series No. 205 (1982). 25. NSAKALA,N., PATEL, R. and BORIO R., Joint Power Generation Conference, Portland, Oregon, ASME publication 86-JPGC/FACT-L (1968). 26. HELBLE,J. J., SR1NIVASACHAR,S. and BONI,A. A., Prog. Energy Combust. Sci. (submitted, 1989). 27. LINAK, W. P. and PETERSON, T. W., Aerosol Sci. Technol.3, 77 (1984). 28. LEVASSEUR,A. A., CHOW, O. K. and LEXA, G. F., Combustion characterization of Kentucky # 9 coals, Electric Power Research Institute (EPRI) Report, CS4994,(1987); Levasseur, A. A., Goetz, G. J., Lao, T. C. and Patel, R. L. The fundamental role of iron in coal ash deposition, EPRI Report to be issued (1989). 29. BON1,A. A., FLAGAN,R. C., BRYERS,R. W., SAROEIM,A. F., BEER, J. M., PETERSON, T. W., WENDT, J. O. L., HUFFMAN, G. P., HUGGINS, F. E., HELBLE, J. J. and SRINIVASACHAR,S., Transformation of inorganic coal constituents in combustion systems, Proc. U.S. DOE Office of Fossil Energy AR & TD Contractors" Review Meeting, Sept. 6-9, 1988, Pittsburgh, PA, pp. 287-364, NTIS, Springfield, VA (1988). 30. HUFFMAN,G. P. and HUGGINS, F. E., Analysis of the inorganic constituents in low-rank coals, Chemistry of Low-Rank Coals, ACS Symposium Series No. 264, pp. 159-174, Schobert, H. H., (Ed.), (1984). 31. SRINIVASACHAR,S., HELBLE, J. J. and BONI, A. A., Mineral behavior during coal combustion: 1. Pyrite
Basic elements during coal combustion transformations, this volume. 32. SRINIVASACHAR,S. and BONI,A. A., Fuel, 68, 829 (1989). 33. HUGGINS, F. E., HUFFMAN,G. P., LYTLE, F. W. and GREEGOR, R. B., The form of occurrence of chlorine in U.S. coals: an XAFS investigation, ACS Division of Fuel Chemistry Preprints, 34, (2), 551-558 (1989). 34. LEVlN,E. M., MCMURDm, H. F. and HALL, H. P., Phase Diagrams for Ceramists, American Ceramic Society, Columbus, Ohio (1964). 35. SIRO, C., WONG, J., LYTLE, F. W., GREEGOR, R. B., MAYLOTTE,D. H., SAMSON,S. and GLOVER,B., Fuel, 65, 327 (1986).
36. SRINIVASACHARS., HELBLE,J. J., BONLA. A., HUFFMAN, G. P., H'dC,G1NS, F. E. and SHAH,N., Mineral behavior during coal combustion: 2. I11ite behavior, this volume. 37. REID, W. T., Coal ash--Its effect on combustion systems, Chemistry of Coal Utilization, pp. 1389-1446, M. A. Elliott (Ed.), J. Wiley (1981). 38. HUFFMAN,G. P. and HUGGINS, F. E., Reactions and transformations of coal mineral matter at elevated temperatures, American Chemistry Society Symposium on Mineral Matter/Ash in Coal, Philadelphia, PA, 1984, ACS Symposium Series No. 301, pp. 100-113 (1986).