Coal combustion by-product diagenesis ii

1999 International Ash Utilization Symposium, Center for Applied Energy Research, University of Kentucky, Paper #67. Copyright is held by the Authors.
Coal Combustion By-Product Diagenesis II
Gregory J. McCarthy, Dean G. Grier, Marissa A. Wisdom, Renee B.
Peterson, Stephanie L. Lerach, Raquel L. Jarabek, Jeffrey J. Walsh, and
Ryan S. Winburn
Department of Chemistry, North Dakota State University; Fargo, ND 58105 KEYWORDS: cementitious coal combustion byproducts rietveld quantitative X-ray diffraction Class C fly ash, some flue gas desulfurization (FGD) residues, and by-products of many cleancoal technologies are cementitious, which indicates a potential for high-volume utilization incivil engineering. As described in the 1995 and 1997 Symposia,1 the established 18-hour, 7-day,or 28-day regulatory or engineering tests of chemical and physical properties may not berepresentative of the long-term behavior of these materials when exposed to the environment.
Studies of the behavior of disposed CCBs, however, can provide insights into the long-termstability of these materials in natural utilization settings.
The goal of this project is to learn more about the phenomenon of coal conversion by-product(CCB) diagenesis, first described by our group and co-workers at UND EERC seven years ago.
CCB diagenesis is a change over time in the mineralogy that occurs after some CCBs aredisposed in a landfill or utilized for a civil engineering application. This change in mineralogy istypically accompanied by a gain, loss or redistribution of major, minor and trace elements, andalteration of physical properties2-4. To attain this goal, two objectives were defined. The first isto develop improved methodology for quantitating mineralogy of these complex crystalline phaseassemblages, using a modern quantitative X-ray Diffraction (QXRD) technique which hassignification advantages over existing, often inadequate techniques. The Rietveld QXRDmethod has been successfully implemented in quantitative characterization of CCB mineralogy.
The second objective is to investigate the phenomenon of CCB diagenesis further by studyingadditional materials recovered from disposal landfills or civil engineering works. Over 300 feetof core material has been recovered and quantitatively characterized from eight boreholes at foursites representing a range of CCB types.
Core material was recovered from four sites. CCBs derived from a clean coal combustiontechnique (Circulating Fluidized Bed Combustion, CFBC), dry-process flue gas desulfurization(FGD), and Class F and C fly ashes in long-term disposal settings have been studied. Three ofthese sites were disposal landfills adjacent to the power plants that produced the CCBs, and thefourth site (CFBC material) was a road embankment.
Crystalline phase assemblages were identified using X-ray diffraction. The diffractogramspresented here were obtained on instruments utilizing CuKα radiation, and equipped with theta-compensating variable divergence slits and graphite diffracted beam monochromators. Detaileddiscussions of XRD characterization5 and CCB mineralogy6 have been presented previously.
Quantification of crystalline phases identified by XRD was performed by Rietveld analysis, usingGSAS software.
RESULTS AND DISCUSSION Rietveld QXRD analysis. Hydrated CCBs are chemically and mineralogically complex, whichmakes quantitative mineralogy determination by conventional X-ray diffraction unusable orunreliable. The whole-pattern Rietveld quantitative X-ray diffraction (RQXRD) method,however, can overcome many of the problems and seems well suited to improve reliability.
CCB materials typically contain a large number of crystalline phases, with many present in minorquantities. Many of these phases exhibit solid solution behavior and polymorphism, as well asvarying degrees of crystalline order. Additionally, a significant amorphous or glassy content istypically present, further reducing crystalline signals. Many of the phases present contain largenumbers of diffraction peaks which overlap with those of other phases. These factors makeconventional semi-quantitative XRD analysis, which utilizes small numbers (often only one) ofanalyte peaks, and which depends on the availability of representative reference materials,unreliable for precise and accurate quantitative determinations.5 Reference materials used inconventional semi-quantitative XRD are often difficult to obtain and are typically notrepresentative of the actual CCB phases, which, as noted above, are subject to solid solutionbehavior, polymorphism, variable degrees of crystallinity, and severe peak overlap.
Rietveld analysis, originally developed for refinement of known crystal structures, has recentlyprogressed to include fully quantitative multiphase analytical capabilities. This method uses thefull XRD profile to simultaneously refine the crystal structures of all constituent crystallinephases, and can model structural parameters sensitive to chemical variation (e.g., site chemicalcompositions), specimen texture effects (i.e., preferred orientation), and parameters indicative ofcrystalline order (crystallite size and strain, stacking fault defects).7 Successful implementation of Rietveld analysis requires a sophisticated understanding of X-ray
crystallography, qualitative and quantitative XRD analysis, and CCB crystal chemistry and phase
behavior. Based on full characterization of a wide variety of disposed CCBs, including thorough
testing of the applicability of Rietveld analysis to CCBs, this study provides the coal ash
community with not only a generic understanding of long-term behavior of CCBs exposed to the
natural environment, but also a structured set of protocols to follow in use of the Rietveld QXRD
method for many CCBs. General recommendations and procedures have been developed and are
provided for the method overall, along with specific protocols for the freely available public
domain package, GSAS, developed at Los Alamos National Labs.8 These recommendations may
be viewed at [http://qxrd.chem.ndsu.nodak.edu/ccbs].
The first step in utilizing the Rietveld method is selection of crystal structure data for each
crystalline phase in a mixture. Review of the literature and crystallographic databases for the
most reliable structures, and testing with the DOE code GSAS were completed in 1997. The
downloadable GSAS input data sets are available for use by other analysts at our website
[http://qxrd.chem.ndsu.nodak.edu/ccbs/]. Several other Rietveld codes were evaluated for
CCB analysis, but GSAS was found to be the most robust and applicable software for this
application. A GSAS-based RQXRD protocol has been developed, and analytical sensitivity,
precision and accuracy have been determined using standard mixtures of NIST Standard
Reference Materials (SRMs) and other CCB phases. Relative error determined from the standard
mixtures is typically in the ±10-15 wt% range (Table 3). The protocols are in use now for
characterization of CCB samples obtained for this project. The protocol has also been applied to
a group of NIST SRM Fly Ashes as well as the core material in this project.
Recovered core material from the four sites has been successfully modeled using the RietveldQXRD method. Three to eight crystalline components, plus an internal standard added forcrystalline content normalization, were analyzed, as shown in Figure 1. Multiphase peak overlapwas easily accounted for, and peak broadening and preferred orientation were modeled for eachof the phases present, as necessary.
CCB Diagenesis. Detailed discussions of earlier studies of emplaced CCBs have been previouslypresented1-4, 9-13. Briefly, three of the five materials, at three of the four sites studied experiencedsignificant by-product diagenesis, including mineralogical alteration coincident with dramaticchange in physical properties. This includes disposed materials from two advanced combustiontechniques and dry process flue gas desulfurization. The predominant new compound generatedin the three sites during diagenesis, and associated with undesirable engineering properties, wasthe mineral thaumasite (see Table 2 for nominal compositions of cited minerals). Theappearance of thaumasite in two of the test cell core samples was accompanied by loss of muchof the initial strength.2,12 If thaumasite formation is not a major cause of strength loss, it iscertainly associated with it.
Diagenesis was not observed in two of the five CCBs studied. In the case of fluidized bedcombustion disposed under arid conditions, no significant alteration beyond initial hydration wasobserved. This has been attributed to insufficient moisture infiltration. The other case in whichno diagenesis occurred involved a disposed material composed of a blend of AFBC by-productsand fly ash10. The absence of significant by-product diagenesis in this second case may be due tothe additional, less-reactive, C-S-H in the cementitious matrix, or simply to insufficient moistureinfiltration to continue hydration reactions and transport leached constituents within thematerials12.
Of the four disposal sites characterized in the present study, two have shown indications of long-term diagenetic alteration. One of these involves Class F fly ash and other conventional boilerby-products from a disposal site in Kentucky, which contains material emplaced for up to 19years. The material contains predominantly unaltered Class F fly ash components (quartz,mullite, magnetite, hematite, and low calcium aluminosilicate glass) as predicted. However,samples from each of the two boreholes indicate a zone enriched in ettringite, a common component of high calcium CCBs. Occurrence of this phase in this Class F fly ash pond iscurious. The chemical components calcium and sulfur, as well as the extreme (high) pHconditions typically required for formation are not expected in this setting. Representativequantitative results for the materials studied are given in Table 4.
Samples from the Indiana road embankment, which contains a mixture of Stoker ash and CFBCby-products appear to be in the process of alteration seen in previous studies of emplaced FBCmaterials. This is indicated by the presence of thaumasite in a few of the samples in addition tothe typical hydration mineralogy of CFBC materials.
Core samples from the Midwestern FGD site show little change from predicted initial hydrationmineralogy. Assemblages characterized show typical nonreactive high-Ca fly ash phases, andunreacted (portlandite) and reacted (bassanite, gypsum, and hannebachite) scrubber residues.
Phases formed by later hydration or carbonation include ettringite, calcite, and gypsum. The onlylong-term alteration observed in the core samples studied to date involves continued hydration ofbassanite to gypsum in many of the near-surface samples, as well as leaching of gypsum near thesurface of the landfill. No evidence of thaumasite has been found in these samples.
The Class C fly ash materials sampled from the backhaul site in central North Dakota also showlittle change from predicted initial hydration mineralogy. Typical (unreactive) Class C fly ashcomponents (quartz, merwinite, periclase, hematite, high calcium aluminosilicate glass) werefound. Minor amounts of ettringite (<10 wt%) were present in addition to occasional occurrenceof the hydration phases monosulfoaluminate and stråtlingite. XRD results from two of theboreholes from the older site (up to 19 yrs emplacement) sampled at this location also indicatesignificant FGD residue codisposed with the Class C fly ash. The scrubber phase, hannebachite,is present in major quantities (up to 50 wt%). In many cases where hannebachite is present, theettringite phase identified shows significant deviation in diffraction peak position, potentiallyindicating sulfite substitution in the channels typically occupied by sulfate oxyanions.
IMPLICATIONS FOR COAL COMBUSTION BY-PRODUCT UTILIZATION The initial behavior one observes on hydrating the cementitious CCB materials discussed here issimilar to that of a low-strength concrete. However, by-product diagenesis associated with manyof these CCBs has been observed to reduce strength by up to 90%, and increased permeability bytwo orders of magnitude, after just a few years in the natural environment2-4. The characteristicsof the altered byproducts resemble those of soils more than concrete. These results should benoted by those working on utilization of this class of by-products for civil engineering andconstruction applications, manufacturing of aggregates, etc. Initially promising 7-day or 28-daylaboratory tests of strength and permeability may not be characteristic of these materials onexposure to the environment. Blending of CCBs with fly ash to increase the proportion ofcementitious C-S-H, and controlling subsequent moisture additions could minimize deleteriousby-product diagenesis. Alternatively, for some applications, it might be desirable to design asystem where by-product diagenesis is allowed to develop naturally.10 The Rietveld whole-pattern quantitative X-ray diffraction technique has been shown useful indetermining mineralogical abundances of crystalline CCB phases (as well as noncrystallinephases, by difference). While not straightforward in many cases, especially in the presence ofhannebachite, the Rietveld technique has been successfully employed with four distinctlydifferent CCB assemblages. This technique has significant advantages over conventional semi-quantitative methods such as RIR. The applicability of the method to CCBs has been rigorouslytested, allowing for the development of suggested estimated standard deviations for the phasesmost commonly encountered in CCBs.
Research is being supported by the Department of Energy (Federal Energy Technology Center)University Coal Research Program Grant No. DE-FG22-96PC96207, and ND EPSCoR Projectthrough NSF Grant No. OSR-945292.
1. McCarthy, G.J., Grier, D.G., Parks, J.A., Adamek, S.D., Butler, R.D., 1995, Long-Term Stability of Disposed Cementitious Coal Combustion By-Products: 1995 International Ash Utilization Symposium, Internatl. Ash Util.
Symp. Proc. Oct 23-25, 1995, Lexington, KY.
2. Weinberg, A., Petrini, R.H. and Butler, R.D. Advanced Coal Technology Waste Disposal, draft final report
prepared for U.S. Department of Energy, Morgantown Energy Technology Center, 1993; A. Weinberg, Coel,B.J. and Butler, R.D., Field Study for Disposal of Solid Wastes from Advanced Coal Processes: Ohio LIMB SiteAssessment. Final Report, Rept. No. DOE/MC/22118-3969 (DE950000053), Radian Corporation, Austin, TX,October 1994.
3. Butler, R.D., Brekke, D.W., Foster, H.J., Solc, J. and McCarthy, G.J. In: Proc. Seventeenth Biennial Low-rank Fuels Symposium, St. Louis, MO, UND-Energy and Environmental Research Center, Grand Forks, ND, 515,1993; Butler, R.D., 1993, In: Proc. 10th Ann. Intern. Pittsburgh Coal Conf., Pittsburgh, PA, Sept., 1993.
4. Butler, R.D., and Pflughoeft-Hassett, D.F. Diagenesis and Leaching Characteristics of Aged Coal Conversion Solid Residues from Mine Disposal Environments, Rept. ND08, UND Energy and Environmental ResearchCenter, Grand Forks, ND, 1995.
5. Bender, J.A., Solem, J.K., McCarthy, G.J., Oseto, M.C., and Knell, J.E., 1993, Quantitative XRD Analysis of Advanced Coal Combustion Solid Residuals by the RIR Method: Adv. X-Ray Anal., 36 343-353.
6. Solem, J.K. and McCarthy, G.J. In: Advanced Cementitious Systems, Mat. Res. Soc. Symp. Proc. Vol. 245, 71,
Materials Research Society, Pittsburgh, 1992; McCarthy, G.J., Bender, J.A., Solem, J.K. and Eylands, K.E. In:Proc. Tenth Intern. Ash Use Symp., EPRI TR- 101774s, Electric Power Research Institute, Polo Alto, CA, 58/1-14, 1993; McCarthy, G.J. and Solem-Tishmack, J.K. In: Advances in Cement and Concrete, (Eds. M.W.
Grutzeck and S.L. Sarkar), 103, Am. Soc. Civil Engineers, New York, 1994.
7. Hill, R.J. (1991) Expanded Use of the Rietveld Method in Studies of Phase Abundance in Multiphase Mixtures: Powd. Diff., 6(2) 74-77; Hill, R.J., and Howard, C.J. (1987) Quantitative Phase Analysis from Neutron Powder
Diffraction Data Using the Rietveld Method: J. Appl. Cryst., 20 467-474; Bish, D.L., and Howard, S.A. (1988)
Quantitative Phase Analysis Using the Rietveld Method: J. Appl. Cryst., 21 86-91; O'Connor, B.H., and Raven,
M.D. (1988) Application of the Rietveld Refinement Procedure in Assaying Powdered Mixtures: Powd. Diff.,
3(1) 2-6; O'Connor, B.H., Deyu, L., Jordan, B., Raven, M.D., and Fazey, P.G. (1990) X-ray Powder Diffraction
QPA by Rietveld Pattern-Fitting - Scope and Limitations: Adv. X-Ray Anal., 33 269-275, Hill, R.J. (1992)
Applications of Rietveld Analysis to Materials Characterization in Solid-State Chemistry, Physics and
Mineralogy: Adv. X-Ray Anal., 35 25-38; Bish, D.L., and Post, Jeffrey E. (1993) Quantitative Mineralogical
Analysis Using the Rietveld Full-Pattern Fitting Method: Amer. Miner., 78 932-940.
8. Larson, A.C., and Von Dreele R.B., GSAS, General Structure Analysis System, Los Alamos National 9. McCarthy, G.J., Butler, R.D., Grier, D.G., Adamek, S.D., Parks, J.A., and Foster, H.J., 1997, Long-Term Stability of Landfilled Coal Combustion By-Products: Fuel, 76(8)
10. Weinberg, A., and Hemmings, R.T., 1997, Hydration and Weathering Reactions in By-Products from Clean Coal Technologies: Effects on Material Properties: Fuel, 76(8), 705-709.
11. McCarthy, G.J., Adamek, S.D., Butler, R.D., Parks, J.A., Brekke, D.W., Foster, H.J., and Solc, J. In: Microstructure of Cement-Based Systems, Mat. Res. Soc. Symp. Proc. Vol. 370, 179, Materials Research
Society, Pittsburgh, 1995.
12. Weinberg, A., Coel, B.J., Foster, J., and Cornelius, B.J., Field Monitoring of Advanced Coal By-Products: Illinois Site. Final Report, DE-AC21-94MC311932, Radian Corporation, Austin, Texas, February, 1997.
13. Stevenson, R.J., Hassett, D.J., McCarthy, G.J., Manz, O.E., and Groenewold, G.H. Solid Waste Codisposal Study, Final Rept. to the Gas Research Institute, ND Mining and Min. Res. Res. Inst., UND-EERC, 1989.
Table 1. Sites and materials involved in this study.
Coal Type
Landfill Site
Midwest FGD
Selected Mineral Names and Nominal Compositions
_________________________________________________ _________________________________________________ _________________________________________________ Table 3. Suggested estimated standard errors associated with CCB phases analyzed by RQXRD.
anhydrite, calcite, ettringite, gypsum, hannebachite, Brownmillerite, "C3A", melilite, merwinite, periclase, Table 4. Selected results from the sites involved in this study.
Codisposed ND Class C Fly Ash + FGD Residue Hist 1 Lambda 1.5405 A, L-S cycle 439 Obsd. and Diff. Profiles Lambda 1.5405 A, L-S cycle 273 Obsd. and Diff. Profiles Figure 1. Rietveld refinements of two selected CCB materials. "Pluses" are observeddata,continuous line is calculated, hashes indicate possible diffraction peak positions for eachphase modeled, and lower curve shows the residual between observed and calculated.

Source: http://www.flyash.info/1999/durabil/grier2.pdf