FINAL TECHNICAL REPORT

September 1, 1995, through August 31, 1996

Project Title: PRODUCTION OF CEMENTS FROM ILLINOIS COAL ASH

DOE Cooperative Agreement Number: DE-FC22-92PC92521 (Year 4)

ICCI Project Number: 95-1/3.1A-13M

Principal Investigator: John C. Wagner, Institute of Gas Technology

Other Investigators: Javed I. Bhatty, Alex Mishulovich, Construction Technology Laboratories, Inc.

Project Manager: Daniel D. Banerjee, ICCI

ABSTRACT

The objective of this program is to convert Illinois coal combustion residues, such as fly ash, bottom ash, and boiler slag, into novel cementitious materials for use in the construction industry. Currently less than 10% of the 3 million tons of these coal combustion residues generated in Illinois each year are utilized in commercial products. These residues are composed largely of SiO₂, Al₂O₃, Fe₂O₃, MgO, and CaO, which are also the major components of cement. The process being developed in this program will use the residues directly in the manufacture of cement products. Since the residues are used as an integral component of the cement and not just as additives to concrete, larger amounts of the residues can be utilized.

The process uses submerged combustion to melt blends of coal combustion residues with lime, clay, and/or sand. The submerged combustion melter utilizes natural gas-oxidant firing directly into a molten bath to provide efficient melting of mineral-like materials. Use of this melter for cement production has many advantages over rotary kilns including very little, if any, grinding of the feed material, very low emissions, and compact size.

During the first year of the program, samples of coal combustion residues were blended and mixed, as needed, with lime, clay, and/or sand to adjust the composition. Six mixtures, three with fly ash and three with bottom ash, were melted in a laboratory-scale furnace. The resultant products were used in mortar cubes and bars which were subjected to ASTM standard tests of cementitious properties. In the hydraulic activity test, mortar cubes were found to have a strength comparable to standard mortar cements. In the compressive strength test, mortar cubes were found to have strengths that exceeded ASTM blended cement performance specifications. In the ASR expansion test, mortar bars were subjected to alkali-silica reaction-induced expansion, which is a problem for siliceous aggregate-based concretes that are exposed to moisture. The mortar bars made with the products inhibited 85 to 97% of this expansion. These results show that residue-based products have an excellent potential as ASR-preventing additions in concretes.

Notice: U.S. DOE Patent Clearance is NOT required prior to the publication of this document.

EXECUTIVE SUMMARY

Background

A significant portion of mined Illinois coal is used in utility plants across the State to generate electricity. There are nearly 30 power utilities in the State of Illinois that consume more than 30 million tons of coal every year. The coal is combusted and the residues generated are primarily fly ash, bottom ash, and boiler slag. Over 3 million tons of these coal combustion residues are generated every year. Generally in the U.S., about 25% of the fly ash, 40% of the bottom ash, and 60% of the boiler slag are consumed, mostly in cement, concrete, and structural fills, leaving about 70% of the residues stockpiled in nearby landfills.

In the State of Illinois, the bulk (nearly 90%) of these residues remains unused. As a result, the utilities are faced with an increasing disposal problem because of both the large volume of these residues and the lack of landfill sites. Furthermore, by virtue of their diverse chemical compositions, these residues are susceptible to ground water leaching and possible contamination. The potential pollution scenarios, pressure for environmentally safe disposal, and the "landfill crisis," combined with political opposition to new landfill sites, have exacerbated the waste problems, posing a technical challenge to develop safe alternative disposal routes.

Approach

The objective of this program is to convert Illinois coal combustion residues, such as fly ash, bottom ash, and boiler slag, into novel cementitious materials for use in the construction industry. These residues are composed largely of SiO₂, Al₂O₃, Fe₂O₃, MgO, and CaO, which are also the major components of cement. Success of this program will lead to significantly higher usage of these residues because they will be used to generate cementitious products instead of being added to concrete mixes as aggregates.

To achieve the above objective, a three-phase program is planned. In the first phase, small samples (100-200 g) of coal combustion residues were blended and mixed, as needed, with limestone, clay, and/or sand to adjust the CaO composition. Six mixtures were melted in a laboratory-scale furnace at CTL. The resulting products, or novel cementitious materials (NCMs), were then tested for their cementitious properties. These tests included x-ray diffraction, x-ray fluorescence, optical microscopy, grindability index, and pozzolanic activity. After testing, the NCMs were blended at various levels with portland cement and subjected to a series of ASTM and CTL-derived tests including hydraulic activity, compressive strength, and ASR expansion.

With optimal mixing and melting conditions established, the second phase will be used to generate larger quantities of NCMs. The three best raw mix formulations as determined in the first phase will be utilized. A pilot-scale submerged combustion melter assembled at IGT will be used for melting the mixtures. The products generated during this phase will be subjected to the same tests as in the first phase. The submerged combustion melter utilizes natural gas/oxidant firing directly into a molten bath to provide efficient melting of mineral-like materials. Use of this melter for cementitious materials production has many advantages over rotary kilns as well as other melters including very little, if any, grinding of the feed material, low emissions, and compact size. Economics are improved because grinding, which is very energy-intensive, is responsible for a substantial portion of the cost associated with production of cement using rotary kilns.

Submerged combustion melters have been in commercial operation in Ukraine and Belarus for several years for the production of mineral wool. IGT has licensed the technology from its developer, the Gas Institute of the Academy of Sciences of Ukraine, for exclusive application outside the former Soviet Union. Other applications for the submerged combustion melter include ash vitrification, glass melting, scrap melting, and direct waste processing.

The third phase will be focused on data analysis and interpretation, process economics, and commercialization considerations. The first phase was completed during this first-year program, while the second and third phases were proposed for the second year. CTL, as a subcontractor, performed the bulk of the effort for the first phase. IGT will perform the substantial part of the effort for the second phase. IGT and CTL will share the effort for the third phase.

Results and Discussion

Initially, two preliminary blends were tested. One blend used fly ash with limestone, while the other used fly ash with cement kiln dust (CKD). Each blend was melted and then quenched, and the resulting product samples were ground to a specific surface area similar to portland cement. Cementitious properties of these product samples were evaluated by accelerated testing of 1-inch cube specimens. The specimens were formed out of cement paste where a certain percentage of the cement paste is displaced by one of the sample products. The specimens were cured for 24 hours at 55C and 100% relative humidity. The specimens made with the product samples obtained 84 and 89% of the compressive strength of a pure portland cement control cube. For comparison, similar (pozzolanic) materials in standard concrete practice are required to have a compressive strength of at least 75% of that of the control.

Based on these initial results, six experimental mix formulations were designed to cover the entire prospective area of compositions. Mixes 1F, 2F, and 3F used fly ash while Mixes 1B, 2B, and 3B used bottom ash. In Mixes 1F, 2F, 1B, and 2B, the proportions of the three major oxides are equal to eutectic points on the SiO₂-Al₂O₃-CaO ternary phase diagram. The composition of Mixes 3F and 3B is approximately half way between those two eutectic points, in an area of higher melting temperature.

Each of the six mixes was melted in an Inconel 601 crucible in an electric muffle furnace and quenched in water. All products were fully vitrified as confirmed by polarized light microscopy. The products were ground in a laboratory mill to the fineness characterized by the Blaine specific surface of about 360 m²/kg, which is typical for portland cements.

To evaluate the hydraulic reactivity of cementitious materials, a number of chemical methods have been used based mainly on certain calculated or measured chemical properties of the materials. However, in this program reactivity was measure more directly by physical testing of paste and mortar cubes prepared from mixtures of the products with cement. Pastes are mixtures of cement and water. Mortars are mixtures of cement and sand.

The products were subjected to a series of standard tests applied usually to blended cements and their ingredients.

Hydraulic Activity By Reaction With Alkali (ASTM C 1073)

For this test, mortar is prepared from a hydraulically active material and sand in a 1:2.75 proportion and mixed with 20% NaOH solution. Hydraulically active materials are those which can react with water to gain strength. Standard 2-in cubes made from the products were tested for compressive strength after 24-hr curing at 55C (131F) and 100% humidity. This test method is normally used for evaluation of the hydraulic activity of slags and similar materials from different sources. As a point of reference, it can be noted that these cementless compositions showed compressive strengths comparable to that of standard (ASTM C 91) mortar cements at a 28-day age (900, 2100, and 2900 psi for different types).

Compressive Strength of Hydraulic Cement Mortars (ASTM C 109)

For this test, blended cements were prepared containing 75% portland cement clinker and 25% products. Characteristics of such cements should conform to the ASTM C 117 specification. Mortar was prepared from cement and sand in a 1:2.75 proportion and mixed with water (0.484 parts per part of cement). Standard 2-in cubes were cast from the mortar and cured under controlled conditions for up to 28 days. Compressive strength of these cubes exceeded ASTM C 117 performance specifications and obtained 76 to 94% of the strength of a pure portland cement control cube. The results of physical testing indicate that products with higher basicity (lower (S+A)/C ((silica+alumina)/lime) ratio) tend to have higher strength. The bottom ash-based products appear to produce better results but, to confirm it, additional testing is required.

ASR Expansion Reduction

In the presence of moisture, concrete can expand due to the alkali-silica reaction (ASR). Low-alkali (<0.6%) cement is specified to be used in concretes for locations with high moisture exposure. In Illinois, ASR is not a serious problem because limestone is the aggregate typically used in concrete. In other areas, particularly western states, siliceous stone is used as the aggregate. There, ASR is a problem.

A further benefit of using coal combustion residue-based cements would be the ability of the cements to inhibit the ASR-induced expansion. The potential of the products to inhibit the ASR in concrete was evaluated using the ASTM C 441 test. In this test, ASR-induced expansion of 10-inch mortar bars containing high-alkali cement and reactive siliceous aggregate is measured. The test results with the coal combustion residue-based products showed ASR expansion reductions of 85% to 97%. These results show these products have an excellent potential as ASR-preventing additions in concretes.

Fiberization of Melts

Usually, slag-like glasses prepared by water-quenching of melts are harder to grind than portland cement clinker. This leads to non-uniformity of particle sizes in blended cements. The prospective commercial melter is likely to utilize a quenching technique based on dispersing the liquid melt by a jet stream of air or rotating metal cylinder. In either case, the final product would form fiberized mass similar to mineral wool. Since the actual pilot-scale melter is not yet available, mineral wool produced from blast furnace slag was used for a preliminary investigation of some aspects of the future technology. Chemical analysis revealed that the composition of the mineral wool was reasonably close to that of coal combustion residue-based products. Microscopic examination confirmed that the material consisted of completely vitrified fibers 7 to 10 m in diameter. Experimental study of mineral fibers proceeded in two directions-intergrinding fibers in blended cements and fiber-reinforced composites.

Grindability of fibers in blended cement mixes was studied by preparing a 75:25 mix of portland cement clinker and fibers and grinding it in a jar mill. Fineness was characterized by Blaine specific surface determined at several points during grinding. Photomicrographs taken at 20-min. increments demonstrated the gradual shortening of fibers, along with comminution of the clinker particles.

Adding fiberized material to mortars and concretes is expected to increase their tensile and flexural strength, as well as control drying shrinkage and cracking. Unlike the polymer fibers currently used, this fiberized mineral material would react with cement paste providing for a stronger chemical bond. Accelerated testing showed that mixtures of fiber and coal combustion residue-based products, when activated by 20% NaOH solution, had compressive strengths equivalent to a pure portland cement control sample.

Conclusions

During this first-year program, it was shown that Illinois coal combustion residues blended with limestone and clay or sand can be successfully used to produce blended cement products. Comprehensive testing of these products using standard cement industry methods has shown strengths meeting or exceeding ASTM performance specifications. The products also have shown an excellent potential to inhibit the expansion induced by the alkali-silica reaction in concretes. Fiberization of the products should lead to reduced grinding cost and increased end-product strength.