Pyrite particles as large as 75 microns should be readily separated from the coal once they are liberated.

In addition to information on the size of the pyrite particles, microscopic examination of polished surface sections of the experimental coals has also revealed much concerning the physical occurrence or arrangement of the mineral matter present. Examination of polished surfaces under low magnification has been reported recently by Deul. The mineral matter, especially pyrite, is largely distributed in planes which are approximately parallel to the bedding plane. These planar distributions of pyrite represent lines of weakness in the bonding between the coal substance and the mineral matter, and as such they offer promise in the development of more effective methods for liberation of the pyrite by mechanical grinding.

Significance of results from further pulverization of experimental coals

In evaluating the effectiveness of different pulverizing procedures, many factors should be considered. Important among these are:

(a) Degree of pyrite liberation achieved.

(b) Concentration of the pyritic sulfur in the various sieve fractions of the product.

(c) Particle size distribution of the product.

(d) Pattern of fracture of coal particles during pulverization.

(e) Degree of liberation of mineral matter achieved.

All of these factors affect the extent to which the pyritic sulfur content of any given coal can be reduced by further processing prior to use. The significance of the data and results obtained to date in this continuing investigation are discussed below.

Liberation of pyrite.-The most important factor in the development of new or improved methods of removing sulfur from coal is the degree of pyrite liberation effected by any given pulverizing procedure.

In many cleaning processes, particles of coal with embedded pyrite behave similar to pyrite-free coal particles that are somewhat larger in size and thus are difficult to remove without excessive loss of the combustible organic material.

At the conclusion of the pulverization tests in the laboratory hammermill and the commercial impact mill, microscopic examination of polished briquets revealed that the pyrite had been effectively liberated in all sizes of the products.

Microscopic examination of the pulverized products also show that the size reduction methods employed are too severe and that less severe pulverizing procedures should yield products with larger pyrite particles. This is particularly advantageous since coal as fired in pulverized form is crushed to about 80 percent minus 200 mesh. Coal could thus be pulverized in stages without an excessive increase in cost of pulverization.

Liberation of nonpyritic mineral matter.-Although the current research is not directly concerned with nonpyritic mineral matter, any process which is effective in liberating pyrite by pulverization will also be effective in the liberation of nonpyritic mineral matter. This is a natural result of the physical similarity of these materials. The nonpyritic mineral matter includes all the clay minerals, carbonate minerals, sulfate minerals, and quartz. Although the amount present of these materials varies from coal to coal, this nonpyritic material constitutes a major source of ash. Elimination of it during cleaning would minimize ash handling problems of powerplant operators.

In the microscopic examination of the pulverized products, it was observed that a significant proportion of the nonpyritic mineral matter was also liberated during the pulverization process.

Particle size distribution.-The size consist or particle size distribution for each hammermill-pulverized product is essentially the same; it does not vary significantly from one experimental coal to the other. (See table II.) The 150 to 250 micron fraction represents about 3 to 10 weight-percent of the product; and, the 74 to 150 micron fraction represents 33 to 38 weight-percent, except for the Ohio Meigs Creek No. 9 seam coal. The 74 to 150 micron fraction of the Ohio Meigs Creek No. 9 seam coal represents about 43 percent of the total. In general, the hammermill-pulverized products are all 50 to 60 weightpercent minus 200 mesh, i.e., less than 74 microns in size. Of this material, the greater than 80 micron fraction represents the largest fraction for three of the five coals, namely, the West Virginia Pittsburgh No. 8, Kentucky No. 11, and Missouri Tebo seam coals.

The size consist of all the one-pass impact mill products is not as fine as that of the hammer mill pulverized samples. Only about 47 to 58 percent of the product is minus 200 mesh or 74 microns in size. Some 8 to 14 percent of the product is plus 60 mesh or 250 microns in size. For three of the coals, the greater than 80 micron fraction is the largest of all the minus 200-mesh fractions; these coals are the West Virginia Pittsburgh No. 8, Ohio Meigs Creek No. 9, and Missouri Tebo seam coals.

The two-pass impact mill products are, as would be expected, considerably finer than either the hammer mill products or the one-pass impact mill products. The range is from 65 to 82 weight-percent minus 200 mesh in size with 1 to 2 weight-percent still greater than 60 mesh in size. For two of the coals; namely, the West Virginia Pittsburgh No. 8 and Kentucky No. 11 seam coals, the 74 to 149 micron fraction is the largest fraction of any, being 27 and 24 percent respectively. For the Ohio Meigs Creek No. 9 seam coal, the greater than 80 micron and 20 to 40 micron fraction are equal in size and together represent 45 percent of the total product; for the Kentucky No. 11 seam coal, the 20 to 40 micron and 40 to 80 micron fraction together represent more than 40 percent of the product. The 20 to 40 micron fraction in every product is a major fraction of the total.

The products from the air jet pulverizer have a considerably coarser size consist than those pulverized in either the hammer mill or the impact mill. This is of especial interest since microscopic examination of the products from these other two mills indicated they had been ground much too fine.

The air jet pulverized products still contain from 16 to 39 weight-percent of plus 16-mesh size material, and contain only 15 to 20 weight-percent of minus 200mesh size material. In general, but depending on the particular coal, the laboratory size one gun jet mill is more effective in pulverizing the experimental coals. In almost every case the product from the one gun unit is much finer in size than that from the two gun unit.

Distribution of pyritic sulfur.-In addition, to the percentage yield by weight of the different particle-size fractions in the pulverized coal products table II also shows the concentration of pyritic sulfur in the individual fractions. The outstanding observation to be made from these histograms is that the greater than 80 micron fraction with only one or two exceptions contains the largest proportion of the total pyritic sulfur of any of the different particle-size fractions. The greater than 80 micron fraction in every case has the highest concentration of pyritic sulfur but it is the largest fraction by weight only in the products from pulverization of the Ohio Meigs Creek No. 9 seam coal in the impact mill. In these two instances, this greater than 80 micron fraction contains about two-fifths of the total pyritic sulfur present.

The procedure for collection of this greater than 80 micron fraction may provide a basis for an effective method for removing a large part of the pyrite from coal. This greater than 80 micron fraction is obtained by air elutriation of the minus 200 mesh sieve fraction in a Roller particle size analyzer; it represents the residue from elutriation of the coal sample under conditions which are calculated to remove all particles of coal (sp. gr. 1.35) less than 80 microns in size. But not all the material in the residue is oversize, i.e., greater than 200 mesh. Actually some of it may be undersize and has reported to the residue during the elutriation because of its higher than normal specific gravity due to the presence of embedded pyrite. In fact, each fraction from the Roller particle size analyzer contains off-size material since the separation process is based on Stokes law which states that the rate of settling of a particle in a fluid medium varies with the square of the radius and with the difference between density of the particle and of the fluid medium.

Microscopic examination of the hammer mill and impact mill products indicates that the pyrite has concentrated in the various particle-size fractions. The minute pyrite particles embedded in the coal substance tend to concentrate in the minus 200 mesh fractions. The larger pyrite particles associated with the mineral matter tend to concentrate in the +60 mesh fractions.

The distribution of pyritic sulfur in the air jet pulverized products has not yet been completed. However, it is apparent that the large particles, i.e., those +16 mesh, contain a higher percentage of pyrite than the coal prior to pulverization. Also, the minus 200 mesh fractions have, in general, an enriched pyritic sulfur content. As these products are evaluated further, conclusions of even greater significance may become evident.

Fracture pattern.-To obtain information on the fracture pattern being produced during pulverization of the coals in the various types of equipment, polished sections of sieve fractions from a single coal were examined under the microscope.

The fracture pattern observed in these products appears quite similar despite the use of different types of pulverizing equipment. The coal substance is readily distinguished from mineral matter with a high content by the relative roundness of the latter material. The sharpness of the breakage is interrupted when a pyrite particle is exposed. In general, the fracture of the coal and mineral matter may be designated as being subangular with medium sphericity.

OPPORTUNITIES FOR FURTHER RESEARCH

The present study pertains to the nature and occurrence of pyritic sulfur in a limited number of high-sulfur bituminous coals and to initial experiments on the liberation of pyrite present in these coals by selected pulverization techniques. The degree of pulverization achieved using a hammer mill and an impact mill, is by design, comparable to that presently required in normal powerplant operation and is indicated to be too fine for optimum pyrite liberation.

The investigation of coarser pulverization, such as that obtained in the initial air jet mill experiments, should be continued and expanded to include stage grinding; i.e., pulverization in two or more steps with intermediate screening to remove the smaller size materials as they are produced.

More intensive studies of the effect of mill design are needed to establish the optimum pulverizing conditions for maximum liberation of pyrite.

Similar research should be conducted on these same coals after they have been subjected to an initial cleaning comparable to that achieved in conventional coal cleaning plants. Results from such studies should indicate whether initial cleaning by presently established procedures is applicable in any new cleaning process for pyrite removal; if so, a reduction in operating costs should be possible.

To date, primary emphasis has been placed on only the liberation of pyrite. The separation of pyrite from coal once it has been liberated constitutes the second half of the overall problem and presents another major area for further research. The cleaning of fine coal, especially steam coal, has not been developed to any degree on a commercial scale.

Air classification, electrostatic separation, and thermomagnetic separation techniques have been suggested as promising possibilities for the dry separation of liberated pyrite. Each of these have been and are being used in other industries on a variety of materials. The effectiveness of each in the removal of pyrite once it is liberated should be evaluated as the second step in the overall removal of pyrite sulfur from coal.

For the successful development of a commercial process for reducing the pyritic sulfur content of steam coals, much additional research and development will be necessary on both pyrite liberation and pyrite separation.

SUMMARY

In summary, research is being conducted to expand our knowledge of the nature and mode of occurrence of pyrite in coals in an effort to gain information which will enable the development of new or improved methods for reducing the sulfur content of coals to levels below that now attainable by conventional coal cleaning techniques.

Application of modern microscopic and chemical techniques in the characterization of a group of selected high sulfur bituminous coals has shown that the major portion by weight of the pyrite present exists as particles large enough to be readily separated once they are liberated.

Results from initial laboratory tests indicate that liberation of the major portion of the pyritic matter and of the nonpyritic mineral matter may be accomplished by controlled pulverization to sizes coarser than that now obtained in modern pulverized coal fired powerplants.

The results also indicate that during pulverization the pyrite tends to concentrate in the various particle-size fractions. The microscopic pyrite particles embedded in the coal substance tend to concentrate in the very fine fractions; that is, fractions below minus 200 mesh. The larger pyrite particles that are associated with the mineral matter tend to concentrate in the greater than 60

mesh fraction. However, for the full benefits of these results to be realized in commercial practice, much additional research will be necessary to develop practical pulverizing machinery and techniques.

Where the pyritic sulfur is mainly associated with mineral matter, there are good prospects of its removal by adapting presently available equipment. Removal of the pyritic grains that are embedded in the organic matter will require the development of improved equipment for processing coal in the minus 200 mesh size.

This paper reports results of research made possible by a cooperative program jointly sponsored by the electric utility industry through the Association of Edison Illuminating Companies and the Edison Electric Institute and the coal industry through Bituminous Coal Research, Inc. Permission of the Joint Research Advisory Committee for this cooperative program to publish the results of this research is hereby acknowledged. The authors also wish to acknowledge the assistance of Maurice Deul and of the other members of the BCR staff in the conduct of this research.

[From Edison Electric Institute Bulletin, October 1963]

PROGRESS REPORT ON EEI-AEIC-BCR COOPERATIVE RESEARCH PROGRAM ON CONTROL OF SULFUR OXIDES IN FLUE GASES

With the growing concern in the adequate control of air pollution, the electric utility and coal industries in July 1958, established a cooperative research program to study the problem. Two research projects have been pursued at the Bituminous Coal Research, Inc., laboratory under joint sponsorship by the two industries-by the electric utilities, through the Edison Electric Institute and the Association of Edison Illuminating Companies, and by the coal industry through BCR. One project consists of engineering evaluation, research, and development on methods for control of sulfur oxide in flue gases; the other consists of research and development for decreating sulfur in steam coals.

Methods for removal of sulfur dioxide from stack gases have been sought by many investigators for a number of years. However, the only approaches which have been attempted on large installations involved the wet scrubbing of the stack gases with water containing alkali, lime, ammonia, or similar materials which reacted chemically with sulfur dioxide. None of these has proved to be feasible and, although scrubbing of stack gases with alkaline estuarine water is being tried at two large powerplants in England, it is not considered to be practical.

In the early stages of the research program, BCR made a complete economic and technical evaluation of the most promising methods for control of sulfur dioxide, selected in cooperation with the Joint Research Advisory Committee which represents the sponsors of the research program. This evaluation led to the conclusion by the committee that the most attractive process for future development was one on which BCR had been conducting some bench-scale exploratory research since 1955. In this method, referred to as the catalytic gasphase oxidation process, the sulfur dioxide in the flue gases is converted catalytically to sulfur trioxide, and subsequently collected as sulfuric acid. The committee felt that if technical problems attendant to the complete development of the process could be overcome, the two primary advantages of the process: namely, the absence of the need for purchase of raw materials and the production of a salable product, would make this process more practical than any of the others evaluated.

DECREASING SULFUR IN STEAM COALS

The second of the two projects being pursued involves research and development on methods for decreasing sulfur in steam coals. The inclusion of research on sulfur reduction in coal prior to combustion with research on sulfur dioxide removal from the flue gases was recommended by the program advisory committee in order to provide the maximum research effort possible. It was also felt that the development of better methods for reducing the sulfur content of coals prior to burning would not only be valuable in control of sulfur dioxide but also would lead to reduction of boiler corrosion and improved boiler efficiency.

At the beginning of this investigation on sulfur reduction methods, it was recognized that current scientific knowledge on coal cleaning did not hold

definite promise for new and economical processes for reduction of sulfur content below that already attainable in conventional coal cleaning practice. Therefore, the general objective of this research investigation has been to expand the knowledge regarding the nature of the occurrence of sulfur in bituminous coals and to utilize this information for development of new cleaning methods in the powerplant as well as at the mine.

CONTROL OF SULFUR DIOXIDE EMISSION

In the initial phases of the experimental work on the catalytic gas-phase oxidation process for recovery of sulfur dioxide from flue gases, bench-scale tests were conducted to determine the optimum operating conditions for the oxidation of sulfur dioxide in simulated flue gases. Also, this work evaluated the efficiency of several commercially available catalysts in promoting this reaction at temperatures in the range 600° to 900° F.

The results of this study indicated that at the test conditions the conversion process was technically feasible. However, many questions remained unanswered including the extent of "catalyst poisoning" by flue gas components, the effect of short contact time, fly ash abrasion, and of significant importance, the pressure drop through the catalyst bed. The pressure drop will control the physical shape of the bed. To minimize pressure drop problems, the most active catalyst was combined with specially designed ceramic shapes to provide an open bed with high catalytic activity and low pressure drop. Currently, these catalyst-carrier shapes are being tested using high gas velocities to determine their usefulness in this process.

ACID RECOVERY METHODS

Concurrent with the catalyst development phase, several approaches to an acid recovery method are being explored. Initial experiments were concerned with removal of the acid produced in the conversion step by scrubbing with hot concentrated sulfuric acid. Current work on this phase emphasizes the recovery of sulfuric acid by condensation in a zoned heat exchanger. The objective of this work is to obtain high efficiency of removal by controlling not only the heat exchanger surface temperature but also by controlling the temperature of the condensed acid. The preliminary experiments on control of acid condensation have been promising.

A laboratory pilot-scale plant combining both the catalytic oxidation step and the acid recovery by condensation step is now being operated at the laboratory. For this stage of the test work, flue gasses from a small pulverized coal burner are being used. It is anticipated that when current test work on this pilot plant is completed, BCR will have developed sufficient design and construction data to enable evaluation of the economics and technical feasibility of a full-scale integrated sulfur dioxide control process for use in powerplants.

REMOVAL OF SULFUR FROM COAL

The sulfur reduction phase of the program was initiated by characterization of samples of six typical high-sulfur coals selected by the Joint Research Advisory Committee. In addition to the usual analytical procedures including float-sink methods, the mode of occurrence, the origin, and the distribution of the pyrite in the coals were studied using modern microscopic techniques. From these analytical studies, especially the microscopic studies, it was concluded that selective pulverization to liberate the pyrite, followed by mechanical and/or electrical removal, merited further investigation.

Following tests of the crushing characteristics of equipment produced by various manufacturers, study of air classification equipment and techniques, and experiments on electrostatic separation of pyrite from coal, a small experimental pilot plant was constructed at the laboratory for further investigation of a process which combines stage crushing and air classification. The pilot plant, with a capacity of 250 pounds per hour, consists of a swing hammer pulverizer, a coal feeder, an air classifier, and the associated electrical drives, air blowers, and controls. Tests currently underway using this pilot plant have the objective of determining the factors affecting the operation of the classifier and its operating efficiency under conditions obtained in a powerplant. Tests to date indicate that if mechanical deficiencies in the system can be overcome, the process will enable substantial reduction of the pyritic sulfur material present in some coals.