Recent advances in the characterization of the glassy aluminosilicate phases in coal fly ashes are reviewed and discussed in terms of the development of new models describing their mechanism of formation, composition, structural relationships and chemical reactivity. Characterization techniques such as electron microscopy, x-ray diffraction, infrared and Raman spectroscopy, nuclear magnetic resonance and Mössbauer spectroscopies, thermal analysis, acid dissolution, silanation, and density fractionation have been used as experimental probes, and reveal a range of information relating to the glass phases from the macro, through the micro, to the nano-structural levels. Complex inter- and intra-particle heterogeneity in ash is observable not only in bulk chemical and mineralogical analyses but also at the molecular level in the constituent glasses. Remarkable relationships between ash particle size, density and glass composition have been observed and lead to a new proposed mechanism for ash morphogenesis based on temperature-viscosity effects in the mineral precursor melts in the boiler flame. A glass-ceramic model for fly ash particle structure is proposed and discussed in terms of phase relations in the CaO–SiO2–Al2O3 and FeO–Fe2O3–A12O3–SiO2 systems, modification of the aluminosilicate glasses by alkali and alkaline earth metal cations, and microheterogeneity involving glass devitrification and phase separation.

The bulk chemistries of five Class F and one Class C pulverized fuel ash (PFA) are compared with the chemical compositions of the glass phase therein, the latter having been determined by analytical electron microscopy. Relative to the bulk, the glasses of Class F ashes concentrate Al2O3 and alkalis but reject Fe2O3. The glass phase is notably inhomogeneous on a micrometer scale; for example, large local variations occur in its Al2O3/SiO2 ratios. The Class C ash contained amorphous material with lower average Al2O3 contents than in Class F, but the extent of compositional homogeneities were greater.

Fly ash consists of mixtures of crystalline substances in a glassy matrix. This matrix is itself inhomogeneous. The combustion process gives rise to compositional fluctuations typically on a micrometer scale; these fluctuations are preserved in the glass and are gradational. However, high-resolution transmission electron microscopy of Class F ash also reveals the existence of interfaces on a nanometer scale. These arise as a consequence of phase separation. Textures and interfaces typical of spinodal decomposition and possibility also suggestive of classical immiscibility have been observed. It is believed that the occurrence of phase separation resulting in nanometer-scale inhomogeneities will be a feature common to most Class F glasses. The consequences of this complex microstructure to reactivity are not as yet known, but some speculations are presented.

Ash density fractions, separated from a high-Ca fly ash, were leached with 6M HCl at 105°C. The density fractions (<0.79 and >2.85 g.cm−3) selected for leaching have been shown in previous work to consist largely of two different glass types, designated Glass I and Glass II. It was shown that both Glass I and Glass II are leachable to yield solutions containing Al and modifier cations. Acid attack on Glass I appears to involve hydrolysis of Si–O–Al (silalane) bridges along with ion exchange at available non-bridging oxygen sites. Attack on Glass II is more extensive, with both silalane and siloxane bridges being hydrolysed, and ion-exchange occurring at non-bridging oxygens. The products of leaching are aluminosilicates. In the case of Glass I, these remain as pseudomorphic, hollow spheres; whereas, for Glass II, substantial reprecipitation of silica (alumina) gel was found.

A low calcium fly ash, rich in iron oxide, was subjected to shaking in 1% HF for up to 20 hours, and measurements were made at intervals of the following parameters: total weight loss, composition of the dissolved material, x-ray peak intensities of crystalline components and residual glass, and internal structures present within the fly ash spheres as revealed by the progressive dissolution treatments. The same fly ash was subjected to similar but longer-term treatment with simulated cement pore solution (potassium and sodium hydroxide). The compositions of the magnetically separated fraction and the non-magnetic residue were also separately determined. As a result of these studies, it was suggested that: (a) the time at which all of the potassium is dissolved might be used as an indicator of the completion of glass dissolution; (b) the position of the x-ray band for residual (more resistant ) glass shifts to higher 2Θ angle before the glass is completely dissolved; (c) in this fly ash, as determined by chemical analysis, all of the iron is in the magnetically-separable fraction, a feature that permitted estimation of the overall glass composition; and (d) this fly ash contains certain poorly understood components including silica-rich spheres, small spheres yielding EDXA signals only for iron, and “patches” of silica and alumina bearing material not of mullite composition, but all of which are highly resistant to HF dissolution.

A brief summary of the use of x-ray powder diffraction for studying the mineralogy of fly ash is presented. Mineralogies of low-, intermediate- and high-calcium fly ashes are discussed and illustrated by results from XRD characterization of U.S. National Bureau of Standards fly ash Standard Reference Materials.

The chemical variation of central North American lignite fly ash can be compared using the major and minor oxide analyses and the distribution pattern of the inter-grain chemistries of the individual fly ash particles. AAS and electron microprobe chemical analysis of nineteen fly ashes have been used to illustrate the variations in bulk sample and individual grain chemistry. This information documents the chemical variability of the lignite fly ashes by geographic region. There is a direct relationship between grains rich in SiO2 and Al2O3 and in the Na2O content of the grains in lignite derived fly ashes. There is also a direct relationship between grains rich in CaP and the MgO and SO3 content of the grains in lignite derived fly ashes.

X-ray powder diffraction has been used to determine the crystalline phase mineralogy in samples of fly ash from each of the lignite mining areas of North America. The characteristic phases of North Dakota lignite fly ashes were periclase, lime, merwinite and the sulfate phases anhydrite, thenardite and a sodalite-structure phase. Mullite was absent in these low-Al2O3 ashes. Montana lignite ash mineralogy had characteristics of ND lignite and MT subbituminous coal fly ashes; mullite and C3A were present and the alkali sulfates were absent. Texas and Louisiana lignite fly ashes had the characteristic mineralogy of bituminous coal fly ash: quartz, mullite, ferrite-spinel (magnetite) and minor hematite. Even though their analytical CaO contents were 7–14%, all but one lacked crystalline CaO-containing phases. Lignite fly ashes from Saskatchewan were generally the least crystalline of those studied and had a mineralogy consisting of quartz, mullite, ferrite spinel and periclase. Quantitative XRD data were obtained. The position of the diffuse scattering maximum in the x-ray diffractograms was indicative of the glass composition of the lignite fly ash.

The physical and chemical properties of fly ash produced at Ottumwa Generating Station have been monitored since April, 1985. The fly ash is produced from burning a low sulfur, sub-bituminous coal obtained from the Powder River Basin near Gillette, Wyoming. One-hundred and sixty samples of fly ash were obtained during the two year period. All of the samples were subjected to physical testing as specified by ASTM C 311. About one-hundred of the samples were also subjected to a series of tests designed to monitor the self-cementing properties of the fly ash. Many of the fly ash samples were subjected to x-ray diffraction and fluorescence analysis to define the mineralogical and chemical composition of the bulk fly ash as a function of sampling date. Hydration products in selected hardened fly ash pastes, were studied by x-ray diffraction and scanning electron microscopy. The studies indicated that power plant operating conditions influenced the compressive strength of the fly ash paste specimens. Mineralogical and morphological studies of the fly ash pastes indicated that stratlingite formation occurred in the highstrength specimens, while ettringite was the major hydration product evident in the low-strength specimens.

Fly ash (FA) is becoming a common addition to ordinary portland cement (OPC) both in Europe and in the USA. Although a great deal of work has been carried out to characterize fly ashes from different sources, less work has been reported in the literature as to the exact nature of the reaction products when fly ash blends are hydrated. Microstructural studies carried out using scanning electron microscopy (SEM) of fracture surfaces have been summarized by Diamond [1]. The results often show a shell of reaction product around the fly ash residues. Some authors have postulated the formation of a duplex film of calcium hydroxide (CH) and calcium silicate hydrate (C-S-H) around the particles. However SEM alone cannot reveal the composition of features in the microstructure and interpretation on the basis of the morphology of the features may be misleading. Some analytical electron microscopy has been carried out using crushed particles of blended pastes [2] but this technique results in a loss of information about the location of analyzed areas.

Particle size distributions of a suite of fly ashes previously analyzed by aclassical sedimentary proced ure have been redetermined using a commercial laser-based particle size measurement system. For most of the fly ashes the concordance between the two sets of data was unexpectedly close, considering the different physical principles involved in the two types of measurement. Laser diffraction instruments (“laser sizers”) appear to have the potential for extremely rapid and accurate size distribution measurement for fly ashes to be used in commercial applications. Such determinations, easily carried out on a routine basis, could serve as important elements in future specifications for acceptance and quality assurance programs.

The reactivity of fly ash is generally difficult to determine because neither a single physical nor chemical property nor any combination of these can be used characterize a particular sample completely. A direct reflection of this is the range of pozzolanicity tests proposed by standards worldwide and the applicability and relevance of such tests is currently being debated.

In many countries, codes of practice and standards relating to fly ashes are currently under review. Some work has been carried out in countries including Australia, United Kingdom, and the United States on commercial fly ashes with a view to possible classification and implementation into standards [1-2]. It is generally recognized that current standards on fly ash do have some inadequacies with regard to performance prediction when included in concrete. Fineness requirements for fly ash for use in structural concrete (BS3892: Part 1) in the United Kingdom are currently being debated [2] with many workers suggesting that it would be possible to produce durable concrete with ashes which would not comply with the requirements of the standard [3].

The development of an improved method of trimethylsilylation [1] made possible the assessment of different polysilicate fractions in hydrated cement pastes, as a means of structural characterization. Several laboratories have presented various gas-liquid chromatographic data on the polysilicate derivatives of C-S-H [2–5]. Identification of the different molecular weight fractions has been made, in most laboratories, by matching the gasliquid chromatographic peaks with the gel permeation chromatographic peaks. The success of this matching process is always dependent on type of instruments, length of columns and temperature range.

In the present investigation, a mass spectrometer, in direct contact with the gas-liquid chromatograph, was used to identify the molecular weight of each of the fractions separated from the chromatograph. The technique was applied to determine the molecular weights and abundances of different polysilicate species in hydrated fly ash pastes.

A major consideration in the development of a coal-fueled turbine engine is to minimize the deleterious effect of products of combustion (POC) on the engine components. Characterization of the POC phases is important for understanding and regulating combustion, emissions, deposition, corrosion and erosion behavior. POC phases deposited from the combustion of beneficiated bituminous coal in turbine simulators were characterized using x-ray techniques and electron microscopy. The phase assemblages include plagioclase feldspars, feldspathoids, spinels, anhydrite, hematite and noncrystalline material. The deposits had the form of a partially sintered powder and a slag. Differences in the phase composition of the powder and slag were determined by xray analysis and observed in the feldspar/spinel lattice parameters. The morphology, phase associations, compositional differences of the powder and slag, and heat treatment experiments indicate that the slag is largely formed by fusion of the powder. The hypothetical fusion process decomposes the CaSO4 in the powder and the released calcia is incorporated into the plagioclase solid solution powder phase. The additional alumina required for charge neutrality in the plagioclase comes from the powder's Al-rich spinel which transforms to an Al-poor spinel. These phase characterization results are useful for assessing and optimizing deposit remediation methods.

The physical and chemical processes underlying the sulfate attack on concrete are briefly reviewed with special reference to blended portland cements containing natural pozzolans, fly ash, or granulated iron blast-furnace slag. Published literature contains empirical guidelines for evaluating the suitability of blending materials for sulfate resistance; however a critical examination shows that some of the guidelines are not satisfactory. Ideas for establishing a scientific basis for determining the sulfate resistance of blended cements are presented in the paper.

Sulphate expansion and strength tests were performed on mortars containing equal proportions of Type 10 (ASTM Type I) Portland cement and fly ash as the cementitious material. Two intermediate-Ca subbituminous ashes, one high-Ca subbituminous ash and one bituminous ash were tested. Control mortars manufactured with Types 10 and Type 50 (Type V ASTM) cements and one mortar containing 50% replacement of cement with limestone powder were also examined. The period and temperature of curing prior to sulphate exposure and the concentration and pH of the attacking solutions were other test parameters. Results indicate that the two intermediate-Ca subbituminous ashes from Alberta Canada, combined with Type 10 cement produce mortars that show excellent long-term strength, acceptable short term strength, and are highly resistant to attack by aggressive sulphate solutions. In some cases durability of these mortars was better than the control mortar made with Type 50 cement. On the other hand, mortar made using the high-Ca ash showed very high susceptibility to sulphate attack; at the same time, mortars made with this ash showed good strength development. Qualitative x-ray diffraction analysis indicates the presence of more lime, magnesium and iron bearing crystalline compounds in the high-Ca ash than in the Alberta ashes. These appear to be responsible for the poor performance in sulphate environments.

A by-product (residue) from the Ames Lime-Soda Sinter Process for recovering alumina from power plant fly ash consists largely of dicalcium silicate and can be used as a raw material for the manufacture of sulfate-resistant (Type V) portland cement. Utilization of the residue will eliminate the need for its disposal and will improve the economic feasibility of the lime-soda sinter process. Laboratory burnability tests were used to identify optimum cement mixes and burning temperatures from both clinker quality and economic perspectives. At a typical kiln temperature of 1450°C, cements that formed concrete that exceeded ASTM specifications for strength were obtained for a limited range of lime-saturation factors and silica ratios. A highly conservative cost estimate for a combined alumina recovery and cement manufacturing facility for a 1000 MWe coal-fired power station gave a 4.7% internal rate of return.

Alternate approaches to conventional control of sulfur dioxide emissions from electric utilities burning coal have resulted in “advanced SO2 control technologies”, among which are furnace sorbent injection, fluid bed combustion, and the use of spray dryers. This summary is a review of laboratory-measured properties of wastes produced from these advanced SO2 control technologies. Chemical composition, leaching properties, physical characteristics, and engineering properties of these wastes were determined using standard methods. These properties were measured to aid in design strategies for subsequent disposal and/or utilization of the wastes under RP 2708, which was funded by the Electric Power Research Institute under the direction of Mr. D.M. Golden. Complete, final reports documenting the sampling and analytical protocol, waste characterization per process, and waste utilization will be published over the next two years by the Electric Power Research Institute.

In primary flue gas desulphurization lime is introduced into the flame, in addition to modifications of the burners and/or equipment. An ash is thus obtained in which the sulphur present in the fuel becomes chemically bound, rather than leaving the heat-generating unit as gaseous SO2, a constituent of the flue gases. In the present paper the composition, particle shape and size distribution of a series of ashes formed in the primary flue gas desulphurization are reported. The options for their utilization need to be evaluated.

Preliminary investigations have shown that the substitutions of up to 10% fly ash (FA), with spray drying absorption products (SDA), in cement mortars (cement: 80% rapid portland cement (RPC), 10–20% FA, 0–10% SDA) results in low early strength of the same magnitude as in mortar with 80% RPC + 20% FA. After storing for 3 months under wet or moist conditions the strengths of some of the SDA-containing mortars had surpassed the strength of the (80% RPC + 20% FA) mortar. Dry stored SDA-containing mortars showed a higher tendency to desiccate. Determination of SO32− showed quick oxidation to SO42− and XRD revealed the formation of ettringite. After a year the strengths of mortars made of 100% RPC, (80% RPC + 20% FA) and (80% RPC + 10% FA + 10% SDA) stored wet or moist were equal. Attempts to use synthetic CaSO3 · ½H2O (CS·½H) as a retarding agent for cement hydration failed, but conductivity measurements of tricalcium aluminate (C3 A)-CS · ½H mixes showed reaction. SEM and XRD investigations identified calcium aluminate sulphite hydrate (C4ASH8). DTA–DTG examinations of synthetic C4ASH8 showed the formula to be C4ASH11. A modification of the instructions for preparation of mortar prisms resulted in the desired retarding reaction between the C3A from portland cement (PC)-clinker and synthetic CS · ½H and SDA-products as well. The resulting C4ASH11 proved to be very unstable and changed to ettringite within a very short time. Use of the modified instructions for preparation of mortar prisms containing SDA resulted in satisfactory early strengths. A series of mortar prisms with increasing content of SDA (x% RPC, (100-x)% SDA in cements; 0<x<80) were prepared and stored under wet, moist and dry conditions. These mortars, now one year old, show a gradual development of properties. Mortars with a SDA content in cement >80% cannot be stored wet. The effects on mortars of the individual constituents of the SDA-products are studied by XRD, development in strength and density. Fragments of 2 year old SDA containing cement mortars and SDA containing concrete from a parking place have been studied.