£2.622 2 GFERC/IC-75/2 (CON F-750586) TECHNOLOGY AND USE OF LIGNITE PROCEEDINGS OF A SYMPOSIUM SPONSORED BY THE U.S. ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION AND THE UNIVERSITY OF NORTH DAKOTA, GRAND FORKS, NORTH DAKOTA, MAY lL-15, 1975 COMPILED BY WAYNE R. KUBE AND GORDON H, GRONHOVD, CO-CHAIRMEN GRAND FORKS ENERGY RESEARCH CENTER, ERDA GRAND FORKS, NORTH DAKOTA 58202 UNITED STATES ENERGY RESEARCH & DEVELOPMENT ADMINISTRATION OFFICE OF PUBLIC AFFAI RS • TECHNICAL INFORMATION CENTER NOTICE This report was prepared as an account of work sponsored by the United States Government. Neither the United States nor the United States Energy Research and Development Administration, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. This report has been reproduced directly from the best available copy. Available from the National Technical Information Service, U. S. Department of Commerce, Springfield, Virginia 22161 Price: Paper Copy $10.50 (domestic) $13.00 (foreign) Microfiche $2.25 (domestic) $3.75 (foreign) GFERC/IC-75/2 (CONF-750586) Distribution Category 'JC-90 TECHNOLOGY AND USE OF LIGNITE PROCEEDINGS OF A SYMPOSIUM SPONSORED BY THE U.S. ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION AND THE UNIVERSITY OF NORTH DAKOTA, GRAND FORKS, NORTH DAKOTA, MAY lU-15, 1975 COMPILED BY WAYNE R. KUBE AND GORDON H. GRONHOVD, CO-CHAIRMEN GRAND FORKS ENERGY RESEARCH CENTER, ERDA GRAND FORKS, NORTH DAKOTA 58202 UNITED STATES ENERGY RESEARCH & DEVELOPMENT ADMINISTRATION Digitized by the Internet Archive in 2018 with funding from University of Illinois Urbana-Champaign Alternates https://archive.org/details/technologyuseoflOOIign CONTENTS PAGE Abstract. 1 Introduction. 1 Abstracts of papers. 3 SESSION I - COAL COMBUSTION GORDON H. GRONHOVD, PRESIDING An overview of ERDA - The United States Energy Research and Development Administration, by S. William Gouse, Jr. 12 Fluidized-bed combustion of coals, by K.D. Kiang, H. Nack, J.H. Oxley, and W.T. Reid. 36 Scrubber developments in the West, by Everett A. Sondreal and Philip H. Tufte. 65 Status of the Citrate process for S02 emission control, by W.A. McKinney, W.I. Nissen, Laird Crocker, and D.A. Martin. l*+8 Electrostatic collection of fly ash from Western coals: Some special problems and the approach to their solution, by Grady B. Nichols and Roy E. Bickelhaupt. 173 LUNCHEON R.O.M. GRUTLE, PRESIDING Environment vs Western coal, by J. Louis York. 193 SESSION II - COAL CONVERSION ROBERT C. ELLMAN, PRESIDING Overview of coal liquefaction projects, by Sam Friedman, S. Akhtar, and P.M. Yavorsky. 202 Coal gasification now, by Noel Mermer. 222 The Hygas process for converting lignite to SNG, by Bernard S. Lee. 230 C02-Acceptor process pilot plant - 197*+, Rapid City, South Dakota, by C.E. Fink, G.P. Curran, and J.D. Sudbury. 239 The outlook for underground coal gasification, by L.A. Schrider, C.F. Brandenburg, D.D. Fischer, R.M. Boyd, and G.G. Campbell. 25*+ CONTENTS—CONTINUED PAGE BANQUET THOMAS C. OWENS, PRESIDING Coal gasification - when, if ever?, by Martin A. Elliott. 276 SESSION II - GENERAL DONALD E. SEVERSON, PRESIDING Large-scale surface mining on the Northern Great Plains, by Robert E. Murray. 286 Overview of reclamation in the West, by Mohan K. Wali, Philip G. Freeman, Alden L. Kollman, and Wilton Johnson. 29^ Commercial-scale drying of low rank Western coals: Part I - Rail shipment test observations, by Robert C. Ellman, Leland E. Paulson, and S. Alex Cooley. 312 Part II - Utilization feasibility, by Clare Wegert and Harry Jensen. 3Ul Metallurgical applications of lignites and low-rank coals, by Robert S. Kaplan and Ralph C. Kirby. 3^5 TECHNOLOGY AND USE OF LIGNITE Proceedings of a Symposium Sponsored by the U.S. Energy Research and Development Administration and the University of North Dakota, Grand Forks, N. Dak., May 14-15, 1975 Compiled by Wayne R. Kube^ and Gordon H. Gronhovd,^ Co-Chairmen of the Symposium ABSTRACT Sixteen papers concerning the technology and utilization of low-rank coals are presented as the proceedings of the 1975 lignite symposium. The eighth in a biennial series, the symposium was cosponsored by the U.S. Energy Research and Development Administration and the University of North Dakota. INTRODUCTION Since 196l, biennial lignite symposia have been held to disseminate information on recent developments in the technology and utilization of Western low-rank coals, lignite and subbituminous. The U. S. Bureau of Mines and the University of North Dakota cosponsored previous symposia. With the formation of the United States Energy Research and Development Administration (ERDA) and the transfer of energy related research functions from the Bureau to ERDA, cosponsorship of the 1975 lignite symposium was assumed by ERDA. Usually the meetings have been held on the campus of the University of North Dakota, but in 1965 and 1971 they were held in Bismarck, N. Dak. to facilitate field trips to lignite-fired electrical generating stations and operating lignite mines. The present proceedings compile the papers presented at the 1975 lignite symposium which was held at Grand Forks, N. Dak. on May lU-15, 1975* An estimated 500 persons attended the 1975 meeting. Those registered were from many States, the District of Columbia, Canada and other foreign countries and represented middle to upper level management and technical people from energy related organizations. This attendance is indicative of the increased interest in the potential of the low-rank coals from the Northern Great Plains Coal Province as a source of energy for power generation and for conversion processes. Professor of Chemical Engineering, University of North Dakota, Grand Forks, N. Dak.; chemical engineer. Grand Forks Energy Research Center, U.S. ERDA, Grand Forks, N. Dak. ^Director, Grand Forks Energy Research Center, U.S. ERDA, Grand Forks, N. Dak. Presiding at the various functions were: Gordon H. Gronhovd, Director, GFERC, ERDA, Grand Forks, N. Dak.; R. 0. M. Grutle, Vice President - Production, Otter Tail Power Company, Fergus Falls, Minn.; Robert C. Ellman, Research Supervisor, GFERC-ERDA, Grand Forks, N. Dak.; Thomas C. Owens, Chairman, Chemical Engineering Department, University of North Dakota, Grand Forks, N. Dak.; and Donald E. Severson, Professor of Chemical Engineering, University of North Dakota, Grand Forks, N. Dak. The welcome to the symposium was given by Thomas J. Clifford, President, representing the University and S. William Gouse, Jr., Deputy Assistant Administrator for Fossil Energy, representing ERDA. Dr. Gouse also reviewed the objectives, organization, and operation of ERDA; his review is included in the proceedings as a paper. A. M. Souby, Manager of Project Lignite, the University of North Dakota, Grand Forks, N. Dak. gave a short slide presentation at the close of the Wednesday afternoon session illustrating progress in construction and initial operation of a process development unit for continuous solvent liquefaction of lignite with hydrogen and carbon monoxide. Tours of the Grand Forks Energy Research Center and Project Lignite at the University were conducted following the Thursday morning session for interested groups. Proceedings of all previous symposia including the 1958 lignite forum have been published to secure wider dissemination of the information presented at the meetings.8 The co-chairmen of the present symposium acknowledge with appreciation the assistance of Charles C. Boley-, Staff ^Gronhovd, Gordon H., and Wayne R. Kube (comp.). Technology and Use of Lignite. Proceedings: Bureau of Mines-University of North Dakota Symposium, Grand Forks, N. Dak., May 9-10, 1973. BuMines IC 8650, 197*+, 262 pp. Kube, W. R., and J. L. Elder (comp.). Technology and Use of Lignite. Proceedings: Bureau of Mines-University of North Dakota Symposium, Bismarck, N. Dak., May 12-13, 1971- BuMines IC 85*+3, 1972, lU5 pp. Elder, J. L., and W. R. Kube (comp.). Technology and Use of Lignite. Proceedings: Bureau of Mines-University of North Dakota Symposium, Grand Forks, N. Dak., May 1-2, 1969. BuMines IC 8*+71, 1970, 17*+ pp. Kube, W. R., and J. L. Elder (comp.). Technology and Use of Lignite. Proceedings: Bureau of Mines-University of North Dakota Symposium, Grand Forks, N. Dak., April 27-28, 1967* BuMines IC 8376, 1968, 201 pp. Elder, J. L., and W. R. Kube (comp.). Technology and Use of Lignite. Proceedings: Bureau of Mines-University of North Dakota Symposium, Bismarck, N. Dak., April 29-30, 1965- BuMines IC 830*+, 1966, 12*+ pp. Kube, W. R., and J. L. Elder (comp.). Technology and Use of Lignite. Proceedings: Bureau of Mines-University of North Dakota Symposium, Grand Forks, N. Dak., April-May 1963. BuMines IC 823*+, 196*+, 128 pp. Elder, J. L., and W. R. Kube (comp.). Technology and Use of Lignite. Proceedings: Bureau of Mines-University of North Dakota Symposium, Grand Forks, N. Dak., April 1961 . BuMines IC 8 l 6 *+, 1963, 113 pp. North Dakota Economic Development Commission. North Dakota Lignite Forum, Speech Summaries. Bismarck, N. Dak. , 1958, *+9 pp. 2 Engineer, GFERC, for his assistance in compiling this publication. Abstracts of the papers presented follow in order of presentation. 4 ABSTRACTS OF PAPERS An Overview of ERDA - The United States Energy- Research and Development Administration By S. William Gouse, Jr., Deputy Assistant Administrator for Fossil Energy, U. S. ERDA, Washington, DC The origins of the ERDA Fossil Energy Program were reviewed and the current organization presented. The Program strategy seeks to develop a mix of technologies to supply fuel in liquid, gaseous and solid forms so as to provide sufficient energy for heating, power, transportation, and chemical feedstocks. Coal utilization will have to increase significantly by 1985 for replacement of oil and gas and in conversion processes. Laboratory, Process Development Units, Pilot Plants and Demonstration Plants will be used to develop and prove processes. Industrial cooperation will be utilized to share costs and speed commercialization. Requests for funding in FY 1976 are approximately $313 million compared to nearly $195 million in 1975 including MHD projects. Other projects include liquefaction, high- and low-Btu gasification, direct combustion and demonstration plants. Fluidized-bed Combustion of Coals By K. D. Kiang, H. Nack, J. H. Oxley and W. T. Reid, Battelle Columbus Laboratories The development of nonpolluting, higher economy, power-generating systems with coal for fuel has received increasing importance as a national objective during the past few years. A major effort in this area on fluidized-bed combustion will probably be launched in the near future. The concept itself is relatively old. However, it is not sufficiently developed, particularly as applied to minimizing the release of sulfur and nitrogen oxides upon combustion and generation of power in advanced systems, to predict the ultimate impact upon the nation's energy posture. ^Company and trade names are used throughout these proceedings for clarity and identification purposes only, and their use does not imply endorsement or recommendation by either the Energy Research and Development Administration (ERDA) or the University of North Dakota. 3 The current state of the art is reviewed. Potential advantages and problems are discussed. For lignite, advantages for combustion in this system appear to be minimization of sodium fouling, reduction in release of nitrogen oxides, and improved combustion efficiencies and power generation rates. Sulfur removal is also an attractive benefit with lignite, but is probably less important than when burning the high-sulfur content, higher rank coals. Scrubber Developments in the West By Everett A. Sondreal and Philip H. Tufte, U. S. ERDA, Grand Forks Energy Research Center, Grand Forks, N. Dak. Wet scrubbers for Western coals are typified by: (l) the removal of only a few hundred parts per million of SO 2 , (2) the oxidation of sulfite to sulfate in the scrubbing liquor, and (3) the presence of reactive alkaline coal ash. These factors and stringent emission standards at state or local levels will dictate future design requirements for wet scrubbers in the Western U. S. Wet scrubbers for particulate removal are operating at seven power stations in the West. In the future, emphasis will shift to wet scrubbing for SOp removal. Three power stations are operating units based on lime, limestone, and soda ash respectively. No general solution has been found for the problem of calcium sulfate scaling in closed loop operation, but progress has been made by recirculating high levels of suspended solids and by adjusting the pH. The inherent alkalinity in some Western coal ashes has been shown in wet scrubbing operation to remove sufficient SO 2 to meet the Federal emission standard without the use of added reagents. Removal is not sufficient to meet more stringent state and local standards; however, improvement to this level may be possible with some of the fly ashes. h Status of the Citrate Process for SO 2 Emission Control By W.A. McKinney, W.I. Nissen, Laird Crocker, and D.A. Martin Salt Lake City Research Center, U. S. Bureau of Mines, U. S. Department of the Interior, Salt Lake City, Utah Under development by the Bureau of Mines, the citrate process for removing SO 2 from industrial waste gases comprises absorption of SO 2 in a solution of sodium citrate, citric acid, and sodium thiosulfate, followed by reacting the absorbed SO 2 with HpS to precipitate elemental sulfur and regenerate the citrate solution for recycling. Research on the process has progressed to the pilot plant stage. Two operations to assess the feasibility of the process for SO 2 emission control are described. One pilot plant constructed by the Bureau of Mines and operated jointly by the Bureau and the Bunker Hill Co. at a lead smelter in Kellogg, Idaho, treats 1,000 scfm of 0 . 5 -percent SO 2 gas and yields about 1/3 ton sulfur per day. The other pilot plant, designed, assembled, and operated by Arthur G. McGee and Co., Peabody Engineering, and Pfizer, Inc., treated 2,000 scfm of 0 . 1 -percent SO 2 gas from a coal-fired steam generating station at Pfizer's Vigo chemical plant in Terre Haute, Ind. In both plants, exit gases containing 30 - to 50 -ppm SO 2 were obtained, representing SO 2 removal efficiencies of 96 to 99 pet. The status of larger scale citrate demonstration plants to operate on 30- to 60-MW powerplants or steam¬ generating plants of equivalent capacity is summarized. Electrostatic Collection of Fly Ash from Western Coals: Some Special Problems and the Approach to Their Solution By Grady B. Nichols and Roy E. Bickelhaupt, Southern Research Institute, Birmingham, Alabama One of the factors that controls the collectability of fly ash by electrostatic precipitation is the electrical resistivity. In the temperature range where power station precipitators normally operate, conduction is primarily controlled by the interaction between the environment and the ash surface which mobilize alkali metal ions that serve as charge carriers. Many Western coals produce low amounts of sulfur trioxide and ashes having low amounts of iron and alkali metals with moderate to high calcium contents. These ashes have very high resistivity and are not particularly sensitive to the conditioning effect of the low SO 3 environ¬ ment . When the above situation occurs, one of three approaches is usually followed to overcome this difficulty. The precipitator is installed in the conventional location on the cold gas side of the air preheater with sufficient collection electrode area to collect the particulate with the poor electrical conditions; or the precipitator is installed on the hot gas side of the air preheater where the thermal effects reduce the resistivity to an acceptable value; or finally, chemical additives are used to modify the resistivity. 5 The electrostatic process is reviewed, precipitator behavior under normal and adverse conditions is discussed, and factors influencing resistivity examined. Alternative approaches to the problem of high resistivity are discussed and criteria are suggested for the approach to solutions. Environment vs Western Coal \ By J. Louis York, Stearns Roger Incorporated, Denver, Colorado The complex interaction between environmental restrictions and effects on society are stressed. Society allows no one to be self sufficient and changes in one sector often have major influences on many others. No small community can offer a specialty based on other than the two reserves of land or human skills. Most Western communities have only resources of land. The conflict between preservation and utilization requires wise use of the natural resources. Air and water pollution is of primary consideration in utilization of Western coal although ambient conditions are not usually problems. Standards are usually set for emission sources rather than ambient conditions. All possible alternates should be considered to insure that correct solutions to pollution problems are used. Environmental impact studies add two to three years lead time to powerplant construction. Use of scrubbers for emission control presents problems in terms of water usage, waste disposal, operational reliability, and increased non¬ productive costs. Water pollution from power stations consists of chemical and thermal discharges. Effluents must be discharged into air or water or be disposed of on land. Overview of Coal Liquefaction Projects By Sam Friedman, S. Akhtar, and P.M. Yavorsky, U.S. ERDA, Pittsburgh Energy Research Center, Pittsburgh, Pa. The dual objectives of the government’s coal liquefaction development program are, in order of priority, production of low-sulfur, low-ash fuel oil for generating electricity without pollution, and production of upgraded products as gasoline, turbine fuel and petrochemical feed stocks. The synthetic fuel oil is the easiest and cheapest product to make, and would replace petroleum products and natural gas burned for power, eliminating the need for Mideast imports of oil and gas. Current work is reviewed for these four categories of coal liquefaction projects: (l) Direct Hydroliquefaction includes the SYNTHOIL, HCoal, Zinc Chloride, Disposable Catalyst, and CO-Steam processes, differing in reactor style and/or choice of catalyst. (2) Solvent Extraction includes two similar pilot plants; the SRC process by Pittsburgh-Midway Coal Company 6 (OCR) and another by Southern Services, differing somewhat in extraction conditions and size. (3) Carbonization includes COED, Clean Coke, and Hydrocarbonization processes, involving different schemes to upgrade or extend liquid yield by use of byproduct hydrogen from carbonization or char gasification. (U) Indirect Liquefaction involves converting coal to syn-gas (CO + Hp) first, followed by catalytic recombination to methanol fuels, liquid hydrocarbons and possibly ethylene as chemical feed stock. Projects in this category are not yet underway. Higher costs of this route may be justified by higher values of specialized products. Coal Gasification Now By Noel F. Mermer, American Natural Gas Service Co., Detroit, Mich. The specifics of American Natural's coal gasification project are reviewed, including: (l) The project background and choice of lignite, (2) The technical team assembled for purposes of exploring feasibility, (3) The resulting synthetic gas cost and special problems associated with the project's financability, and (4) The conditions under which the synthetic gas can be marketed. Additionally, the project timetable is presented; it indicates that synthetic gas will be available in 1981. The Hygas Process for Converting Lignite to SNG By Bernard S. Lee, Institute of Gas Technology, Chicago, Ill. Operating results and experience in using Montana lignite in the HYGAS pilot plant are discussed. Lignite with its high moisture content and high reactivity represents a unique feed material for SNG production. The HYGAS pilot plant has been operated on a completely integrated basis with lignite including gasification to produce hydrogen and methanization to a SNG of about 1,000 Btu/cu ft. Design and evaluation for a demonstration plant is underway and if conditions are favorable, such a plant could be constructed in this decade. A conceptual commercial plant producing 250 million scf/day of pipeline gas is presented. CO 2 Acceptor Process Pilot Plant - 197^ Rapid City, South Dakota By C.E. Fink, G.P. Curran, and J.D. Sudbury, Conoco, Rapid City, S. Dak. In the past two years 17 runs (Runs 9 through 25) were completed in the pilot plant. During this period, fully integrated plant operation was demonstrated using active acceptor to supply the gasifier heat requirements and lignite coal as feedstock. Balance periods were achieved, which allowed the gathering of process data for which heat and material balances were calculated. A new startup procedure was developed in which dead-burned dolomite rather than active acceptor was used for the system's initial acceptor inventory. The new procedure provided a reliable means of achieving operating conditions. 7 Many process and mechanical problems were also overcome. These include: (l) control of corrosion in recycle gas heater, (2) elimination of regenerator deposits, and (3) control of the accumulation of trash (non-char, non-acceptor) material in the system. Construction of a methanation unit was recently completed. Plans for the coming year include the operation of this unit which features a catalyst-in-tube, liquid-cooled methanation reactor design. The Outlook for Underground Coal Gasification By L. A. Schrider, C. F. Brandenburg, D. D. Fischer, R. M. Boyd and G. G. Campbell, U. S. ERDA, Laramie Energy Research Center, Laramie, Wyo. Past experiments have shown underground coal gasification (UCG) to be technically feasible but not economically competitive. During these tests, stabilization of gas production rates and of gas heating value were not achieved for sustained periods. The Bureau of Mines began UCG experiments at Hanna, Wyoming, in November 1972. On January 19, 1975, this work was transferred to the U. S. Energy Research and Development Administration. The problems of past tests have been avoided and encouraging results have been obtained. No gas leakage from the reaction zone has been observed. Gas production rate and gas heating value were relatively stable for a 5 1/2-month period. During this 5 l/2-month period approximately 20 tons of moisture-free coal were gasified per day, energy balance calculations showed 3-5 times more energy produced than consumed, and comparison with an air-blown surface gasifier showed similar energy recovery efficiencies. A second experiment to further define process feasibility is underway. If results from this experiment are favorable, design and construction of a 15-30 MWe pilot plant will follow. Successful pilot plant operation would lead to design of a commercial demonstration plant by 1980. Coal Gasification - When, If Ever? By Martin A. Elliott, Texas Eastern Transmission Corporation, Houston, Tex. The urgent need for building commercial coal gasification plants to produce substitute natural gas (SNG) from coal is demonstrated in a discussion of some of the basic factors affecting the future producibility of natural gas. In spite of this urgency, there are unconscionable delays at all levels in governmental and other institutional procedures that must be complied with before commercial coal gasification becomes a reality. These delays, coupled with inflation, are resulting in substantial increases in the projected cost of SNG. Hopefully technology under development will help to reduce costs. However, it should be recognized 8 that there are limitations on possible cost reductions and that commer¬ cialization of these development will take time. The road blocks and delays will eventually be overcome, but today they look so formidable that we may understandably raise the question — COAL GASIFICATION - WHEN, IF EVER? Large Scale Surface Mining on the Northern Great Plains By Robert E. Murray, The North American Coal Corporation, Bismarck, N. Dak. The depletion of our oil and gas reserves, coupled with increasing energy needs and the goal to achieve energy independence by 1985, has placed renewed emphasis on coal, our most abundant fossil fuel. It is estimated that the Northern Great Plains is underlain with UU pet of the nation's total recoverable coal deposits. As coal utilization technologies are realized, this area will play an ever-increasing role in meeting future energy requirements. Large-scale surface mining on the Northern Plains encompasses a myriad of factors: Exploration, reserve acquisition, coal sales agreement execution, financing, preparation of specifications, equipment selection, development of plans, obtaining permits, staffing, and the implementation of sound land reclamation programs. Changing social patterns further dictate that developmental objectives include provisions for socio-economic concerns, public input, and interaction with government agencies. An Overview of Reclamation in the West By Mohan K. Wali, The University of North Dakota, Philip G. Freeman, U.S. ERDA, Alden L. Kollman, The University of North Dakota, all of Grand Forks, N. Dak., and Wilton Johnson, USDI, Bureau of Mines, Washington, D.C. This report concerns the current practices and problems of land reclamation in the Western States and is the result of firsthand information obtained from visits to virtually all coal strip mines in the Western United States, including two mines in Arizona, five in Colorado, three in Montana, three in New Mexico, eight in North Dakota, one in Texas, and nine in Wyoming. At each site, the methods currently followed by the operators (under supervision of the State regulating agencies in many cases) were noted. While there appears to be a serious desire on the part of the companies and the State regulatory agencies to revegetate the spoil materials to nearly approximate the original conditions, the efforts to reclaim them seem to be seriously hampered by the lack of comprehensive 9 research data on problems relating to revegetation. Since most of these mining areas lie in arid or semiarid regions, the total precipitation received and its distribution seems to be the key limiting factor. Of equal concern is the problem of erosion by both wind and water. Apprehension of prohibitory regulations on strip mining by regulatory agencies was often voiced by the operators. Commercial-Scale Drying of Low Rank Western Coals Part I. - Rail Shipment Test Observations By Robert C. Ellman, Leland E. Paulson, and S. Alex Cooley, U.S. ERDA, Grand Forks Energy Research Center, Grand Forks, N. Dak. The Grand Forks Energy Research Center and Commonwealth Edison of Chicago jointly conducted tests in which 400 tons each of subbituminous and lignite coals were dried in a commercial scale dryer, oil sprayed and cooled, then shipped from Pekin, Illinois to Grand Forks, N. Dak. and stockpiled. Cars containing raw coal and dried coal which had not been oil sprayed were also transported for comparative purposes. The tests were conducted in August and November of 197^- The subbituminous coal was dried from 26 to l6 pet moisture and upgraded in heating value from 8,^20 to 9,650 Btu/lb. The lignite was dried from 39 to 22 pet moisture and its heating value was increased from 6,U20 to 8,300 Btu/lb. The subbituminous coal before loading was cooled to 115° F and sprayed with oil at a rate of from 2 to 6 gal/ton. Similarly, lignite was cooled to 85° F and oil sprayed at rates of from 1 to 2 gal/ton. The subbituminous coal was subjected to 2 inches of rain but the average moisture content did not change. Dried coal that was oil sprayed had less wind loss than either the raw or unsprayed dried coal. With both dried lignite and subbituminous, a moderate increase in weight was measured during transit. With the subbituminous shipments, ignition occurred at poorly fitted bottom dump doors, but was limited to a very small area near the door. The dried lignite was shipped when ambient temperatures were below freezing, and a 3-inch crust of semi-frozen coal was formed around the edge of the car. Part II. - Utilization Feasibility By Clare Wegert and H.M. Jensen, Commonwealth Edison, Chicago, Ill. The feasibility of drying high-moisture, low-sulfur Western coals is dependent upon the realization of several potential benefits. These benefits include (l) lower transportation costs, both capital and operating expense, (2) delivering a coal with improved burning qualities, (3) reduction of operating and maintenance costs of boilers, and (k) an increase in boiler capability. 10 Because of the newness of the inherent moisture drying process, many of the questions concerned with the development of the process cannot be answered at this time. Metallurgical Applications of Lignites and Low Rank Coals By Robert S. Kaplan and Ralph C. Kirby, USDI, Bureau of Mines, Washington, DC In 1973 more than h2 pet of our iron consumption was derived from approximately 6l million tons of domestically produced taconite pellets, with more than 22.3 x 10^ cubic feet of natural gas for induration being consumed in Minnesota alone. By 1980 pellet production in the United States is expected to reach 90 million tons and require approximately 65 x 109 cubic feet of natural gas. However, the increasing shortage of natural gas projected for the near future would inhibit the ability of our Nation to harden these domestic taconite pellets, thereby forcing us to rely more heavily on foreign sources of iron ore. To counteract this problem, which was brought into sharp focus by the announcement that natural gas supplied to the iron ore companies of Minnesota would be cut off permanently in 1978, the Bureau undertook to demonstrate in a pilot scale kiln that taconite pellets hardened in a solid fuel fired system have properties equivalent to those hardened in a natural gas fired system without contamination from the coal ash. Tests of 120 hours each in duration have been conducted using pulverized lignite, subbituminous coals, and bituminous coals to fire a pilot-scale grate kiln. The coal firing rate was l.h to 1.6 x 10^ Btu per long ton of pellets, or about twice the thermal requirements for commercial requirements for commercial magnetite pellet induration. The balance of heat necessary to sustain the 1,300° C kiln temperature was supplied with a separate natural gas burner. Green pellets of magnetite taconite concentrate were fed to the kiln at an average rate of 825 lbs per hour. The amount of kiln ringing which occurred is thought to depend on both the ash softening temperature and the oxides contained in the ash. Although some ash was picked up by the pellets, this did not significantly affect pellet chemistry. Tests using a cyclone burner for firing to prevent ash from entering the kiln are also discussed. 11 AN OVERVIEW OF ERDA - THE UNITED STATES ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION by S. William Gouse, Jr."^ introduction The Fossil Energy Program in ERDA is a rapidly growing, ongoing program that principally originated in the Department of Interior's Office of Coal Research and the U. S. Bureau of Mines. It involves work on coal, oil shale, petroleum and natural gas. The underlying assumption on which the Fossil Energy research and development program is based is the following: No matter which projection of future energy supply and demand one examines, one concludes we must increase our utilization of coal; we must fully understand the potential of our vast oil shale resources and assure that a technical option to utilize them in an environmentally sound manner is available; and we must develop technology to more efficiently extract oil from known depleted reservoirs and that we develop technologies to extract oil and gas from presently uneconomic resources. The Fossil Energy research and development program moved on to an accelerated planning and implementation basis with a supplemental appropriation in FY 7^« Continued growth was evident in FY 75 as is also evident by our request for FY 76. The basic program structure and strategy is based on the fact that the knowledge base of various fossil fuel resources with respect to extraction and conversion differs; fossil fuel resources in the future must meet the demands of a variety of markets — power, transportation, chemicals, home heating, industrial processing; there is a vast existing fossil energy industry involving electric utilities, pipeline gas industry, chemical industry, transport industry and manufacturing; and that this is currently a private sector operation. Our program strategy is also based on the fact that we have not been rapidly developing techniques for the conversion of coal to other fuel forms, and that we are not moving fast enough to efficiently extract the fossil fuels represented by much of our oil shale resource, coal in the thick, deep seams of the West and natural gas currently in formation which we do not understand how to economically exploit. 1 Deputy Assistant Administrator for Fossil Energy, US ERDA, Washington, D.C. 12 The legislation under which we are now operating clearly instructs us to demonstrate the technical and economic feasibility of new technologies for providing energy from fossil fuels. This means a heavy involvement with various elements of private industry. Our program has been designed and is being implemented with this in mind. We have a balance of architect/engineering firms, designers, chemical companies, power companies, coal and oil companies, etc., involved in our research and development program. The cost sharing for pilot and demonstration plants assures technology transfer to the implementing sector and the involvement in our decisions to scale-up of the people who will be responsible for implementing and operating these technologies. A major initiative in our programs was the recent award of the contract for the phased design, construction, and operation of a clean boiler fuel demonstration plant. This is the first demonstration plant Fossil Energy has undertaken. In FY j6 we will be continuing this effort as well as starting the conceptual design of two more demonstration plants — one in the high-Btu gas area and one in the low-Btu gas area. The remaining appropriations are to continue forward with the activities launched during FY 74 and FY 75 and to expand them where appropriate. Origins of the ERDA Fossil Energy Program Figure 1 indicates the pre-ERDA, Department of Interior Energy organization from which elements were transferred to ERDA. A portion of the Bureau of Mines, including its Energy Research Centers, was transferred. This included 8l2 personnel (ERC, 773 and Washington, 39), and $87-4 million in FY 75 appropriations. The Office of Coal Research was transferred in its entirety. This transfer included a 272 FY 75 year-end employment ceiling and $246.3 million in FY 75 appropriations. The AEC transfer included $9*2 million and 15 personnel. Coal, gas and oil extraction and in situ processing of oil shale programs were transferred from the Office of Coal Research, Bureau of Mines, and the AEC. Current Interim Organization The Fossil Energy Program is currently organized in the manner shown in figure 2. The organization consists of an Assistant Administrator for Fossil Energy, a Deputy Assistant Administrator, and three program divisions Coal Conversion and Utilization, Petroleum, Natural Gas, and In Situ Technology, and Advanced Research and Supporting Technology. Staff offices include Evaluation and Review, Planning, Environment and Safety, Program Support, and an MHD Project. 13 PRE - ERDA DEPARTMENT OF INTERIOR ENERGY ORGANIZATION >- z CJ x g °iu: U- O UJ UJ 5 o => < O h- < z < s H < £ X u. SlU u O CC £ oB o CD UJ < cc _i 3 U GO UJ 0C GO o o < 5 s- C3 uj iu a. S^CE uj O CC Z o z < ID < < 3 CO — < —i GO z LU lU INTERIM ORGANIZATION CHART 1 15 DIVISION OF DIVISION OF DIVISION OF PETROLEUM. NATURAL GAS COAL CONVERSION ADVANCED RESEARCH AND AND IN SITU TECHNOLOGY AND UTILIZATION SUPPORTING TECHNOLOGY Dr. N. Dunning Dr. R. Zahradmk q, ^ Acting Director Acting Director Acting Director Program Strategy The ERDA Fossil Energy Program structure is influenced by many factors, and a program strategy has evolved to account for these factors. The program seeks to develop a mix of technologies because the complex American economy requires enormous amounts of fuel in various forms: natural or synthetic gas for space heating, electric power, and industrial processing; liquid fuels for transportation, electric power, heating and chemical feedstocks, combustible solids for power generation and industrial applications (f'ig. 3). Experience in the petroleum industry with tremendous variations in petroleum feedstocks, production and transportation costs, regional demands for product mix, seasonal cycles, and local prices for competing fuels has resulted in a wide variety of refinery designs. Today large numbers of competing refinery process operations are available to the refinery designer, and even after fifty years of sophisticated technical development, no single process for petroleum refining has emerged. Coal and other fossil fuel feedstocks are even more variable than petroleum feedstocks, and some of their product markets are extremely susceptible to competition by easily transportable liquid and gaseous fuels. Thus, the fossil fuel program strategy is to advance a broad spectrum of processes, specifically suited to wide ranges of feedstocks and markets. The ultimate commercialization of a synthetic fuel process depends upon process efficiency and cost. Small percentage increases in process efficiency result in huge savings because of the large throughput of material in these plants. Consequently, it is necessary to keep the program tracking towards optimum process configurations. The Fossil Energy Program conducts continuing evaluations to provide comparisons of the process alternatives being developed. These studies provide program guidance, as well as information for the development of demonstration plant projects. The strategy is also influenced by the estimate that coal utilization will have to increase significantly by the year 1985, and that on the order of 20 synthetic fuel plants will be on line producing a million barrels per day Btu-equivalent synthetic fuel from coal and shale by 1985 , based on both existing technology and technology in the current research, development and demonstration program. Our program is now accelerated with the overlapping of development steps. The funding levels and activities are different for coal, oil, gas, and oil shale. These differences reflect the state of development of conversion and extraction technologies for the different fossil fuels as well as the differing levels of investment by the private sector in research and development. 16 IMPACT OF ERDA ENERGY PROGRAM ON ENERGY SYSTEM IT There is also recognition of the extent of our resource base for each fossil fuel type. Thus, coal as our most abundant resource, but with the least developed technology base, receives the largest funding share. Program Accomplishments Major accomplishments are the development of new and improved process technology for conversion of coal to synthetic gas and liquids, for combustion of coal in a more efficient and environmentally acceptable manner, for transformation of oil shale to shale oil and refining to clean fuels, and demonstration of improved petroleum and gas recovery techniques. In addition to the six Energy Research Centers, the Fossil Energy Program includes the following major facilities, either underway or in the design phase: Laboratory/Process Development Units Pilot Plants Demonstration Plant 11 9 1 These major projects include pilot plants in coal gasification and liquefaction, large process development units in fluidized bed combustion, comprehensive component test facilities and a liquefaction demonstration plant. Geographically, these energy activities are located throughout the country, as shown in figure U. The start of major operations is shown in figure 5- To date these fossil energy conversion facilities have processed several tens of thousands of tons of feedstock and produced a spectrum of clean gaseous, liquid and solid products. These products have been utilized in large scale tests in utility boilers. Navy destroyers, gas turbine generators and chart gasifiers. Comprehensive process designs have been made in the development of an information base to support a coal conversion industry. Coal Conversion and Utilization Coal is our most abundant domestic fossil fuel resource and reserve. Because of numerous delays in bringing nuclear electric plants on line, limited domestic supplies of oil and gas, insecurity of supplies of imported oil and gas, and the size of projected import requirements to close the various projected demand-supply deficits, the country has no alternative but to accelerate its investment in making coal available as a major energy source. 18 ENERGY ACTIVITY LOCATIONS 19 • UNIVERSITIES + RESEARCH CENTERS ■ CONTRACTORS ★ MAJOR PROJECT SITES ERDA FOSSIL ENERGY MAJOR PROJECTS 00 >- < o < X o (- O « z I >- z X 00 N CM LU cr < 5 LU O -t o ° UJ m O Q a x S < CO < cd 00 o cc o > < 5 55 u -i ^ -o a: uj i a cj go a o o o cc a > CO cZ LU x £ o ! 3 > o' -J X C_> 00 GO LU h- X -J Q- S a CD Z CD <■ < X »— X Z 2 > < 00 Z CG < x a a “J > ^ X LU GO i" < a. ** CD Q Q LU LU O co co 2 a a o UJ LU q- 2 2 u < < X X X ^ h- h- a o o — Q O £ a < z o cc o u. X oo uj lli X X < (~ X X F— CJ < X < > c\» ro ^intflp^coa)0*-c\in^ tr> ld r-. m ai o f- c\j n r- i- (\J (NJ (N ('J (VJ OO (NJ CO cm n n CD ►— > X z X X < UJ F— —J > CJ x Q_ O < »— CJ X GO UJ LU LU X X CD LU UJ Z Q —j > GO o X LU < X LU GO o _J O. F— O X X < GO ►— LU r^» lD cn rsj ro ro ro CD Z h- X X LU < < -J Q F— CJ X _J Q_ < O O < GO X X X )— LU LU — LU CO o £ X X CJ o G LU J a_ z h- X o GO < < X O GO GO LU £ X LU Q_ a GO 1— DC > < < a_ X CD LD o CO CNJ ro ro ro h- Z GO GO LU CJ < —i Q- Z UJ < o cc o X X LU 1 — GO GO a. £ CJ a UJ o LU z o h- < X GO “O Z < GO X LU X h- GO LU X CSI O cc < _J CJ cu CJ h- o LD 03 ro r-* r _ cm rsi ro ro cino 20 Increasing quantities of coal or coal-derived products will be needed for electric utility and industrial boilers, refinery feedstock, pipeline gas distribution systems, and chemical feedstock. Accelerated use of coal requires resolution of many potential problems. The knowledge base for most aspects of the process of coal conversion and utilization is not adequate for the fossil fuel industry in all its parts to meet the challenge before it, even though some technologies are ready for immediate implementation. The overall program goal is to provide the knowledge for the economic and environmentally sound utilization of our vast coal resources to help meet national energy needs. This goal includes developing and demonstrating technical methods and processes for: The production of clean gaseous and liquid fuels from coal, suitable for power generation, transportation, residential, industrial and chemical uses. Direct combustion of coal in an environmentally acceptable manner. To achieve this goal we estimate for 1976 operating cost of $279*5 million and plant and capital equipment obligations for clean boiler fuel demon¬ stration plant of $20.0 million. The operating costs and obligations as described in the budget submission break down in the following way: COAL PROGRAM PROGRAM ESTIMATE (In thousands) Sub-program 1974 1975 1976 Operating Costs: Liquefaction $ 19,764 $ 54,632 $ 96,897 High-Btu Gasification 29,415 57,841 42,838 Low-Btu Gasification 8,442 22,308 51,671 Direct Combustion 3,509 20,681 32,645 Advanced Research and Supporting Technology 1,463 14,780 32,061 Demonstration Plants — — 18,100 Advanced Power Systems 20 3,957 5,26l (MHD) (2,815)* (7,584)* (13,773)' Subtotal, Operating Costs 62,613 174,199 279,473 $(65,428)* $(181,783)* $(293,246)' Plant and Capital Equipment Obligations — 13,150 20,000 TOTALS 62,613 187,349 299,473 (65,428)* (194,933)* (313,246)- *includes MHD projects presented in the budget under the Solar, Geothermal and Advanced Energy Systems activity. 21 I •would like to describe the objectives of each of these program elements, to list the major ongoing projects and to describe how the requested appropriation will be used. Liquefaction The prime objective of coal liquefaction research is to provide technology which is economically competitive and environmentally satis¬ factory to convert coal to a clean liquid fuel for electric power generation, transportation and heating for industry and homes. Major ongoing projects in this area are given in table 1. TABLE 1. - Coal Liquefaction Projects Major projects Contract value $M (cost share)— Contractor Location Key events Coal Oil Energy Development (COED) 21.0 FMC Princeton, NJ Pilot operations complete FY 75 Solvent Refined Coal (SRC) U2.0 PAMCO Tacoma, Wash. Pilot operations started mid FY 75 H-Coal 8.1 HRI (2.7) Trenton, NJ Cattletsburg, KY PDU runs FY 75; pilot plant decision mid FY 76 Clean Coke 6.5 U.S. Steel (1.9) Monroeville, PA PDU complete FY 75 pilot plant decision FY 76 Synthoil 6.9 Foster (l.l) Wheeler PERC Bruceton, PA RFP for construction June 75 1 _/ Contract values as of 6/30/75. The solvent refined coal process can produce such a fuel in the near future. If pilot plant work presently underway continues to be successful, it should be possible to move rapidly with an early commercial¬ ization of this process by private industry. At the same time, technology will be developed for converting this heavy semi-solid product into a crude oil from which transportation fuels, chemicals, and home heating oils may be obtained. The goal is to have synthetic liquid fuels from coal produced at the rate of at least one-half million barrels per day by 1985 in commercial plants. 22 High-Btu Gasification The major objective of the high-Btu gas program is to provide improved technology for the manufacture of pipeline quality gas from coal; improve¬ ments which, firstly conserve coal as they permit greater efficiency, and secondly decrease the cost of manufacture of gas by about 20 pet compared to present technology. The major ongoing projects in this area are given in table 2. TABLE 2. - High Btu Gasification Program Major projects Contract value $M (cost-^/ share) - Contractor Location Key events CO 2 Acceptor Process 2.0 ( 1 . 0 ) Conoco Coal Dev. Co. Rapid City, SD Methanation plant construction complete FY 75 Hygas Process 39-3 ( 10 . 1 ) Institute of Gas Technology Chicago, IL Steam oxygen system construction complete FY 75 Liquid Methanation 1.9 (.7) Chemical Systems Inc. TBD Complete pilot plant construction FY 75 Ash-Agglomer¬ ating Process 8.8 ( 1 . 6 ) Battelle Columbus West Jefferson, OH Complete pilot plant construction FY 76 Steam-Iron- Process 18.1 (7.9) Institute of Gas Technology Chicago, IL Complete pilot plant construction FY 76 Bi Gas 69.6 (11.5) Bituminous Coal Research/ Conoco Coal Dev. Co. Homer City, PA Complete pilot plant construction FY 76 Synt'nane 9.6 Rust Engrg. Lumus Corp. PERC Bruceton, PA Complete construction FY 75 1/ Contract values as of 6/30/75. 23 This program, the furthest along technically of all our programs, will he pursued with vigor, and four pilot plants will be operating in the 1975-76 period. The best process or combination thereof will be used to design a demonstration plant; construction of such a demonstration plant is targeted to begin about 1977* The goal is to be able to begin construction of full scale commercial plants with new technology in the early 1980*s. Low-Btu Gasification The objective of the low-Btu gas program is to develop at the earliest possible date one or more gasifier systems which are economically applicable for the use of coal as a substitute for oil or natural gas for power generation, industrial heating, and chemical feedstock. Hie major ongoing projects in low-Btu gasification are given in table 3. TABLE 3- - Low-Btu Gasification Projects Maj or projects Contract value $M (cost share)— Contractor Location Key events Advanced 13.9 Westinghouse Waltz Mill, PA PDU Gasification Sys. for Electric Power Generation (k.2) Electric Operational FY 75 Entrained Bed 9.0 Foster Wheeler Sioux Falls, Pilot plant (Pressurized) (3.0) Northern States Power SD design FY 75 Fluidized Bed 2.5 Bituminous Coal Research Monroeville, PA PDU Operational FY 76 Molten Salt 6.9 Atomic Norwalk Harbor, PDU complete (2.3) International/ Rockwell CT FY 76 Entrained-Bed 21.9 Combustion Windsor, CT Pilot plant design (Atmospheric) (6.9) Engineering & systems fabrica¬ tion FY 77 1/ Contract values as of 6/30/75 - 2k Improvements to low-Btu gasification processes are needed. New fluidized-bed technology and improvements in older fixed-bed technology offer economic opportunities. Hot-gas clean up offers improved efficiency. The next level of improvement of low-Btu gasification would be to pressurize the gasifier so that the fuel can be burned in a gas turbine coupled with a conventional steam boiler. Such a combined cycle power plant has the potential of increasing overall thermal efficiency of conversion to electricity by 25 pet. Much of the gasifier development work in our high-Btu gas program will be useful in moving the low-Btu gasifier program forward. The FY 1976 request for the low-Btu gasification sub-program, on the basis of budget authority, represents a decrease over FY 1975- This decrease is primarily attributable to a lessening requirement for selected support studies and engineering activities which were conducted in 1975 - Direct Combustion The prime purposes of the direct coal combustion program are: (l) to develop first atmospheric and then pressurized systems capable of burning high sulfur coals of all degree of rank and quality in fluidized bed combustors in an environmentally acceptable manner; (2) and to improve the reliability and efficiency of present boilers. The most efficient way to utilize coal is by its direct combustion as any process to upgrade coal requires expenditure of energy. Thus, the highest overall efficiency of coal utilization is direct burning. Processes to permit combustion in fluidized bed boilers have a high potential payoff and will be pursued. Major direct combustion projects are listed in table k. TABLE U. - Direct Combustion Projects Major proj ects Contract value $M (cost share)— Contractor Location Key events Fluidized Bed (Atmospheric) Ik.3 Pope, Evans, Robbins Rivesville, WV Startup FY 76 Fluidized Bed (Atmospheric) 3.5 Combustion Power Co. Menlo Park, CA FY 76 — Dry hot gas cleanup development Fluidized Bed (Pressurized) 6.7 (2.0) RFP Out TBD Select contractor; Response due 31 Jan., 1976 1/ Contract values as of 6/30/75. 25 Advanced Research and Supporting Technology Advanced Research and Supporting Technology serves as the central research arm for all program areas of fossil energy and has three main objectives: (a) to provide support technology to assure reliable and efficient process operations, (b) to initiate development of new and improved extraction, conversion and utilization process, and (c) to assure an adequate supply of trained technical personnel. The successful conversion of coal and oil shale to clean fuels, and thermal and electrical energy, depends upon efficient and reliable plant operations. Equipment must operate under hostile erosive and corrosive conditions at high temperature. A strong material and components research program has been initiated to insure reliable operation of the plants. An active university research program is underway and expanding rapidly. About Uo universities are now involved. An essential element of this program is the training of technical personnel, furthermore, the universities have a unique potential for innovative and fundamental research which can be important to achieve major improvements. Advanced research, especially for synthetic fuels from coal, is underway in government, institutional, and industrial laboratories. Examples of projects include use of novel catalysts and refining of synthetic fuels from coal and oil shale to clean products, as well as process research to provide for a supply of chemical feedstocks. An important component of Advanced Research and Supporting Technology is science and analysis. Intensive studies on coal and oil shale structure and reaction mechanism and how sulfur and nitrogen are bound into the structure can lead to new, more efficient conversion processes. Insight into properties of coal and oil shale, combustion chemistry and thermo¬ dynamics of hydrocarbons are likewise the subject of research. The refining of synthetic fuels and testing of fuels, alternative to conventional gasoline is underway and is expected to lead to optimization of processes and products or, indeed, new fuels for transportation. 26 Major ongoing projects are shown in table 5- TABLE 5* - Supporting Science and Technology Projects Maj or projects Contract value $M (cost . share)— Contractor Location Key events Fireside corrosion 7.0 TBD TBD Response to RFP by 31 Jan. Gasification materials 3.3 NBS Argonne MPC/IITRI Washington Chicago, IL Chicago, IL — University contracts 9- 0/yr (1.0) 40 Various — Industry contracts 9.0 15 Various — 1/ Contract values as of 6/30/75. Demonstration Plants A major program in the Division of Coal Conversion and Utilization is the Demonstration Plant program. Rapid commercialization of processes for removing sulfur from coal can be expected to have early impact on the national energy problem by providing acceptable alternatives to oil and gas. In addition, these processes can be used as a first step toward the production of a synthetic oil. The hydrogenation of coal produces byproduct gas as well, which can be reformed to yield a substitute natural gas (SNG). Future commercial plants which are anticipated from planned demonstration facilities can be expected to produce clean liquids and gases as co-products with further potential relief for the national energy situation. In July 197^, an RFP was issued for clean boiler fuel demonstration plant. On January 17, 1975, the Office of Coal Research awarded a contract to Coalcon for the phased design, construction and operation of a 2,600 ton/day demonstration plant using a hydrocarbonization process for producing 3,900 barrels/day of liquid product, and 22 million cubic feet daily of pipeline quality gas. The first two phases of the project, conceptual design and detailed design, will provide a continuing opportunity to further evaluate the optimum plant configuration to be constructed. This plant will fulfill a near term objective of the coal program to provide a fuel substitute from coal for a portion of higher grade oil and gas used in the generation of electricity and in industrial heating processes. 27 This plant is the first demonstration plant Fossil Energy has undertaken. It is the last step before commercialization and is the result of work through the pilot plant stage. While the design and engineering is being funded by the government, the industrial partner will share the construction and operating costs on a 50-50 basis. The plant schedule is being accelerated, and we hope to have the plant begin operations in FY 80. In order to meet national needs and continue acceleration of coal conversion technology, we have requested funding to initiate conceptual design activities during FY 76 for a low-Btu gasification demonstration plant, and for a high-Btu gasification demonstration plant. These plants will build on experience of a coal conversion technology. Products produced in demonstration plants will be tested by potential users to insure compatability with all elements of the product supply system. Advanced Power Systems The objective of the advanced power program is to increase the efficiency of coal burning electrical power generators. The program focuses largely on developing advanced gas turbines capable of operating on coal derived fuels, and on developing magnetohydrodynamic systems which utilize coal directly. Major advanced power projects are listed in table 6. TABLE 6. - Advanced Power Systems Projects Maj or projects Contract value $M (cost-jy share) — Contractor Location Key events MHD 3.6 (2.U) AVCO Everett, MA FY 77 Component test MHD 8.3 Univ. Term. Space Inst. Tullahoma, TN FY 78 Direct coal fired operation ECAS 3.0 (2.0) GE Westinghouse NASA/Lewis Various Complete Dec. 75 Open Cycle Gas Turbine h.O TBD TBD RFP Issued Feb. 75 1/ Contract values as of 6/30/75- 28 Technology for turbines, incorporating materials and cooling design to avoid corrosion, integrated with heat recovery from gasifier, and operating on coal derived fuels is anticipated by 1980. Petroleum and Natural Gas Petroleum and natural gas will continue to be the nation’s main fossil energy resources for many years. This program is directed toward increasing the production of oil and gas from both on-shore and off-shore areas by advanced production and recovery techniques and in improving the efficiency of petroleum use and re-use. To these ends we estimate in FY 76 $23.6 million in operating costs and $100,000 for plant and capital equipment obligations as follows: PETROLEUM AND NATURAL PROGRAM ESTIMATE GAS PROGRAM (in thousands) Sub-program 197^ 1975 1976 Operating Costs: Oil and Gas Extraction $ 6,695 $ 16 , 21+2 $ 22,065 Oil and Gas Utilization 1,182 1,025 1,582 Subtotal, Operating Costs $ 7,877 $ 17,267 $ 23,6U7 Plant and Capital Equipment Obligations 13U 35 100 TOTALS $ 8,011 $ 17,302 $ 23,7U7 The budget request for this sub-program, measured on a budget authority basis, is less than FY 1975. This change is mainly attributable to a decreasing level of follow-up analysis and interpretation activities associated with previously conducted tests of nuclear explosives to stimulate natural gas. Gas and Oil Extraction The Oil Extraction Program emphasizes demonstrations of existing and improved secondary and tertiary recovery techniques rather than new refinery technology. Because industry has a broad technological base in refinery technology, it would be difficult for the Government to make substantial contributions in this area. 29 Budgets for research on oil production are much smaller than those for refining. Also, small independent operators traditionally have contributed much to oil a.nd gas production. Government participation with industry will ensure rapid development of enhanced oil recovery technology and rapid transfer of this technology to those who can use it. Presently, the economics associated with advanced recovery projects are uncertain. It is probable that the oil industry would eventually implement competitive advanced recovery techniques. However, time is the critical element, and is the reason that the Government must take the development lead. The natural gas stimulation program is designed to stimulate the commercial production of natural gas from formations containing vast quantities of natural gas but having natural permeabilities (rate of flow of fluids through porous rock formations) so low that commercial production to date has not been feasible. Experimental methods include massive hydraulic fracturing, combinations of hydraulic and chemical-explosive fracturing, and fracturing wells deviated from normal to intersect natural fractures. The major ongoing projects in this area are given in table J. 30 TABLE 7- - Petroleum and Natural Gas Major Projects Contract value $M Major projects (costly share)— Contractor Location Key events Micellar- Polymer Flood Field Test 7*0 (4.0) Cities Service, Inc. Eldorado, KS Injection tests - Oct. 74. Injection- Salinity Adj . - Apr. 75. CC >2 Injection 6.7 (5*9) Williams Bros. Eng. Co. Lincoln City, MS Begin injection - Nov. 75. COg Injection 3.2 (2.0) Guyan Oil Co. Lincoln City, WV Redrill wells - July 75. Injection - Jan. 76. Thermal Recovery 6.8 (4.8) Husky Oil Co. Paris Valley, CA Ignition - Apr. 76 . Econ./Tech. Eval. - Dec. 76 . Thermal Recovery 3.1 (2.1) Hanover Petrol. Co. Zavala City, TX Ignition - Apr. 75* Chem. Explos. Fract. 0.4 (Wells, etc. ) Petroleum Tech.. Inc. Romney, WV Inj. Equip. - Feb. 75. 1st Explos. test - Feb. 75. Chem. Explos. Fract. 0.2 (Wells, etc. ) Talley Frac., Inc. Mineral Wells, TX 1st Explos. test - Nov. 74. 2nd Explos. test - Jan. 75. Hydraulic Fract. 2.3 (1.0) CER Geonuclear, Inc. Rio Blanco, CO 1st Drilling - Nov. 74. Fracturing - Nov. 74. 1/ Contract values as of 6/30/75* 31 Fluid injection and fracturing methods for gas and oil extraction will he used in field demonstration projects. Solvent recovery methods for heavy oil as well as in situ combustion methods for tar sand will be developed and demonstrated. Gas and Oil Utilization Increased efficiency in using our petroleum and natural gas is an obvious method of extending available energy supplies. This program element contributes to increasing end-use efficiency by characterization and process improvement of petroleum and residuum oil or tar from naturally occurring heavy-oil reservoirs and crude liquids from oil shale or coal. Major ongoing projects are also shown in table 7. We would seek, wherever possible. Government-industry cooperation. Oil Shale The Government has prime responsibility to ensure efficient and environmentally sound utilization of the nation's enormous oil shale resource. The focus of the ERDA Fossil Energy portion of the oil shale program is threefold: reducing the water requirements of the oil shale industry through in situ processing; increasing the recoverable reserve base through improved production technology; and ensuring that environmental safeguards are built into the in situ oil shale process as an integral part of the process development. The Oil Shale Program focuses on in situ retorting rather than surface retorting, which is considered a known technology. In addition, laboratory and bench scale studies on composition and conversion for clean fuels from oil shale have been initiated to provide a technology base for improvements and new process development. Advancing in situ production of shale oil to commercial feasibility is targeted for the early 1980*s. This will require expansion of ongoing in situ retorting activities. These larger tests will be performed on a contract basis, starting with design and preparatory work in FY j6. In order to achieve this goal we request FY 76 operating costs of $8.1^7 million and $325,000 for plant and capital equipment, broken down in the following way: 32 OIL SHALE PROGRAM PROGRAM ESTIMATE (in thousands) Sub-program 197^ 1975 1976 Operating Costs: In Situ Processing $ 2,026 $ 2,903 $ 7,03^ Composition and Characterization 75^ 551 1,113 Subtotal, Operating Costs $ 2,780 $ 3,W $ 8,lU7 Plant and Capital Equipment Obligation 25 75 325 TOTALS $ 2,809 $ 3,529 $ 8,^72 In Situ Processing The requested appropriation includes investigations leading to the production of liquid fuels with field tests of one to ten acres each planned to serve as an equivalent process development unit effort. Both combustion and circulating hot gas would be used as heating methods. Expanded gasification work would also be performed with design of an aboveground facility to begin FY 76, and actual construction to begin in FY 77. Operation of this facility should provide data to permit an in situ field demonstration to begin in 1980. Major projects are shown in table 8. TABLE 8. - Oil Shale Major Projects Major projects Contract value $M ( COSt-^/ share) — Contractor Location Key events Small area test of in-situ extraction In-house LERC Rock Springs, WY Test eval. thru 75. Pilot field tests Est. 10- 20 M. w/50% sharing Unknown Rock Springs, WY Design and Prep FY 76-78 "Best" tech, test - FY • 80 - 8 ; 1/ Contract values as of 6/30/75. 33 Special Foreign Currency Program Under this program the Administration will provide support to selected coal energy research projects now underway in foreign nations that will complement current domestic research efforts. Payments for these projects will be made in currencies of those nations in which the research takes place, and which the U.S. Treasury determines to be in excess of our nation’s normal requirements. The 1976 budget request provides $ 6.65 million to initiate projects in Poland dealing with the hydrogenation of coal ($5-35 million) and the production of clean gas from coal for use in MHD power systems ($1.30 million). The specific goal of this program will be to utilize, through international cooperation, the technological expertise developed by Polish specialists to corroborate or complement research activities being conducted by the Energy Research and Development Administration. Poland is an obvious choice to provide coal research and development information. Poland has a large and growing coal industry with extensive reserves but only small amounts of oil and gas. This has led to a strong research program to develop the needed technology for the conversion of coal to liquids and gases. Conclusion I would like to note in conclusion that the Fossil Energy Program currently has over 125 contracts outstanding with total value of nearly $600 million. Industry is contributing over 25 pet of these funds: We hope to maintain the present standard for cooperative joint funding which anticipates one-third private, two-thirds Federal, for pilot plants and 50-50 for demonstration plants. Our participants include many aspects of American industry: oil and gas companies, chemical companies, coal companies, architect engineer and manufacturing companies, utilities, universities, research centers, and trade associations. In addition to the contractors performing the major project work I have described, we have contractors performing engineering evaluations, systems studies, planning studies, and environmental analyses. We have several interagency agreements including agreements with the Department of Commerce, Corps of Engineers, and the Department of the Interior, as well as strong intra-agency support. We have international agreements with the United Kingdom, Russia, Poland and Germany. We are constantly advised of work in the private sector and adjust our planning accordingly. We cooperate with both American industry and American utilities (through A.G.A., EPRl) to develop national plans which integrate privately sponsored and Federally sponsored R&D. Table 9 is a summary displaying by program and subprogram the appropriations required for operating expenses. 3b TABLE 9 . - U.S. Energy Research and Development Administration Fossil Energy Development co ■d ft d co 2 O -ft &H G •H CO ft CCj r—I i—I o FQ CO p CO o o d G d i>s P •H ft O -ft p ft < P cd bD d ft pq bD G •H P d ft CD ft O VO D— o\ i—I >H ft CD P d e •H P co ft LTV c— ov >H ft (L) d 6 •H p CO ft p- b— On i—I >h ft r—I ccj ft P o < ft- CO H i —1 LTN i —1 o OO LTV CVJ ft- p CO IH- ft- co OV OO ft VO P vo o ft- vo CO P OO rH -d" vo P OO CO VO CVJ vo o 1—1 p o LfN vo o i —1 i —1 CVJ CO r r r n n n n n n n n n n n n o VO Cvi H lf\ OJ CVJ CO Ov CVJ i —1 co ft- i —1 co i —1 o ONP UA 00 OO i—i ft- CVJ CVJ l —1 CVJ co -ea -ea -ea -ca -ca -ca -ea >5 OJ p Ov H VO OO o UA ov ft- vo o LfN UA vo p VO VO co o Ov Ov o o CVJ OV CVJ CVJ VO co i—i p •rH ltv co co vo o OO o p ov b- ft- ft- OJ Ov i—i CD ft n n n n n n r\ n n n n n n n bD o b- 00 UA VO OO LTV ft- co i —1 vo ft- 1—1 OO OV d -ft OvVO P OO OO CO CVJ CVJ CVJ UA d P CO OO FQ d •€A- -ea -ea -ca -ca ca < CM H OO ft- 1—1 o o Ov CVJ Lf\ ft- 00 1—1 o co 1 CO p o UA OO OO ov p CVJ vo o LTV LTN CVJ P vo co co CAVO ft- i—i CVJ O CVI ov LTV Ov co nr'#' n n n n n n n n n n o P t— OJ co o vo l—1 r- CVJ co o LTV UA CM CVI l—1 c— i— i i— i Ov i—i <—1 -ea -ea -ta¬ -ca -ca ca IA IA O ft- ft- Lf\ o OV Ov Q\ co CVI OJ p i—i p P O OO Ov CO OJ CO vo CO LTN vo vo CVJ CVJ P •H ft- 00 P O CO OO OO OO D— •— i ft- ON ft- CVJ CD ft n n n n n n n n n n n n n bD O p o\ vo -H’ UA 00 vo i —1 OO co ft- d PI OV UA UA OO OJ ft- CVJ CVJ o d P CVJ OO PQ d -£a -ea -ea -ca -ca -ca -ca < P UA CVJ o OV OO o 00 UA CVJ ft- vo o o co VO i — 1 -P - CVI o vo i —i Ov co t— cvj ir\ OO ft- p ft- P P UA vo vo 1 — 1 OO o t— ft- CVJ CO r< r* r* n n n r n n n n n O OV Ov OO OO 1 — 1 CVI vo 1 — 1 ft- CVJ CVJ 00 O H CVJ vo ft- -ta -ea •ea -ca -ea •ca ca i>2 i—1 CO i—1 o ft- VO o co oo LTN co co C\J o vo P OV OO CVJ P o p co VO P o o OJ OO ft- P •H CVJ OO H Lf\ UA Ov t— CVJ ft- 00 ON CVJ vo CD n n n n n n n n n n n n n bD o vo OO CVJ i—1 Lf\ ov co ft- l—1 CO CVJ co o d PI P OO CVJ i—1 CVJ d p I—1 r~1 PQ d -ea •ea -ca- -ca -ca -ca -ea- < • • • • • 1—1 • • • • • • • • cd • • • • • • • • • • • • • • • • • • • • • • • r H • • • • • • • • • • • p • • £ • • • • • d • • cd • • • • • G S • ft • • co • bD • G O • bD • • G • S • d O • CD O •H Td • O • • O G 0) • G 1-1 CO cd •H p G • ft • • •H O p • d O p CD O p d d • CQ • • p •H co • G G ft o tsi • bD G • • d p >> G rG ft d bD 1—1 d •H • G O CD • O d CQ O o CJ 1—1 O d ft i—1 P • •H •H 1-1 •H o •H ft CD CQ ft ft P •H CD CO Ti p d •

d CD bD O 1—1 cd ft bD o •H O cd CO O ft 03 r-. 60 X OO r-* OO x X l i a . <■ CD u 3 U5 E cn x QJ cn N z CN a> •H CN > E CO rH X •H G /T-s O' X X X CJ ^ CJ ^ E 03 Q X cq CQ G cn pu CQ CJ - X ~ 3 nj ^ tH W CNI 3 m u rH v —' X o z z z r «■ 00 X XJ 5 X G U PC < W 3 X 0) PC i X C 0Q cd /■“N X X c u cn o C 3 * oc o O c 2 CD 3 •H •G C CJ (D W • cn PC o C X X X C X CQ a. a- CO c X 00 X cn E G G X w o ^ • o CO X X o O 2 cu 3 CJ CJ < w f*4 CJ 'V 'O oo I 1850 F) and low oxygen level (<0.5 percent). On the other hand, the carbon burnup cell (^9) needed to be operated at high temperature (1950 to 2050 F) and high oxygen level (>3 percent) to achieve over 99 percent overall combustion efficiency. Addition of salt into the fluidized boiler increased slightly the SO 2 capture effectiveness of the system. However, corrosion would certainly be expected to be more of a problem with salt additions. The experimental efforts at ANL in a 6-inch diameter atmospheric combustor and 6-inch and 3-inch pressurized combustors were aimed at providing fundamental information on combustion efficiency, S0 9 and NO^ emission, particulate emission, and sorbent regeneration and attrition in both atmospheric and pressurized fluidized-bed combustion. The atmos¬ pheric data (50) suggested an optimum temperature (1400 to 1600 F) for sulfur retention depending on coal and sorbent types. For 90 percent sulfur retention, the Ca/S ratio must be maintained over 3. In contrast, the pressurized data (10 atm)(51) indicated little temperature dependence for sulfur capture (1450 to 1650 F) and a Ca/S ratio of only two for over 90 percent sulfur removal. NO emission in pressurized combustion (120- 270 ppm) was also much lower than in the atmospheric system (215-350 ppm). Argonne is also one of the few organizations that have conducted tests with lignite. (52 53) Research activities at Esso R&E ’ included bench-scale inves¬ tigation using 3-inch diameter atmospheric and pressurized combustors and regenerators, and a 0.65-MW pressurized continuous miniplant. The miniplant consists of a 12.5-inch diameter combustor and a 5-inch diameter regenerator capable of operating at temperatures up to 2100 F (combustor) and 2000 F (regenerator) and velocities up to 10 fps. Both ER&E and ANL data indicated NO emission in fluidized-bed combustion is derived from nitrogen in coal rather than from nitrogen fixation. Regeneration studies at both ANL and ER&E revealed technical difficulties with the pressurized system. The Esso miniplant has been in operation recently without the regeneration unit. The present goal of Esso's investigation is to improve fluidization quality with deep-bed operation and demonstrate the feasi¬ bility of deep-bed combustion with immersed cooling coils. WRL has been designing a turbine blade cascade to be appended to the miniplant. 154) The EPA-NCB joint study , which began June 1970 and ended July 1971, was very comprehensive and included pilot-plant tests of American coals in five British test rigs at BCURA and CRE for atmospheric and pressurized fluidized-bed combustion. For the atmospheric system, the gas velocity was varied from 2 to 11 fps, temperature from 1420 to 1680 F, and bed depth from 2 to 7 feet. For pressurized combustion at BCURA (5 atm), operating conditions were 2 fps, 1470 F, and bed depth of 4 feet. In general, these data using much larger combustors (48-inch x 24-inch pressurized combustor at BCURA, 27-inch diameter at BCURA, and 36-inch, 12-inch, and 6-inch at CRE for atmospheric combustors) verify the findings at ANL and ESSO R&E. The BCURA data at low temperature (1470 F) and low gas velocity revealed little sintered deposits or erosion on the turbine blade cascade at the outlet of the pressure combustor. BCURA(55) ]_ ater (August 1972 to September 1973) extended investigation into the high temperature ranges (1650 to 1750 F) under a joint contract between National Research Development Corporation (NRDC) and the Office of Coal Research of the U.S. Department of Interior. These later studies (55) revealed that deposition on turbine blades cascade was not significant at bed temperatures less than 1600 F but was sufficiently extensive at a bed temperature of 1750 F that blade cleaning by injection of a proprietary fruit stone material was required. Based on data accumulated from experimental work at BCURA, NCB- CRE, PER, ANL, and ER&E, Westinghouse Research Laboratories (WRL) under an EPA contract evaluated the technical and economic feasibility of such systems(56). With subcontracts to United Engineers and Constructors, and Foster-Wheeler, WRL prepared conceptual designs of a 250,000 lb/hr industrial boiler plant, and 635-MW atmospheric and pressurized fluidized- bed boiler utility plants in 1971. The industrial boiler design was based on a gas velocity of 12.5 fps and bed depth of 2.5 feet; the atmospheric utility boiler on 10-15 fps and bed depth of 2.5 feet; and pressurized utility boiler on 6 to 9 ft/second and bed depth of 10 to 15 feet. The capital cost for the fluidized-bed industrial boiler was $7.40 lb/hr steam as compared to $7.60 lb/hr for a conventional coal-fired plant with scrubber and $2.44 lb/hr for a conventional gas/oi 1-fired plant. The marginal economic advantage of the fluidized-bed industrial boiler and the availability of clean fuel at the time of the Westinghouse study prompted WRL to recommend that development of the industrial fluidized- bed boiler be suspended until future availability of clean fuel can be assessed. For utility application, the pressurized fluidized-bed combustion boiler plant (with sorbent regeneration) represented a capital cost saving of 18 percent and power cost saving of 7 percent over the conventional plant with a stack gas clean-up system. The potential advantages of the pressurized system as revealed by WRL's evaluation thus shifted the EPA's contract activity at ANL and ER&E to emphasize the pressurized systems. Continued effort of WRL for EPA culminated in a three-volume report on the evaluation of the fluidized-bed combustion process in 1973.(52) In this later report, WRL advocated the once-through pressurized fluidized boiler plant for the first generation utility application and presented a preliminary design of a 30-MW pressurized fluidized- bed boiler development plant. Alternate pressurized fluidized-bed combustion systems such as the adiabatic combustion concept and recir¬ culating fluidized-bed combustion concept were also analyzed and recommended for further development. 50 The Combustion Power Company, Menlo Park, California, has been involved in R&D on fluidized-bed combustion of low-grade fuels (such as municipal waste) and coal. They have a 7 ft-l-inch inside diameter pilot plant reactor operated at 5 atmospheres, and a superficial velocity of 6 ft per sec, A 22-inch I.D. atmospheric test unit is used for support studies. The pressurized combustor delivers hot gas (about 1500 F) to a gas turbine (after gas cleaning). The turbine drives an electric gener¬ ator and the fluidizing air compressor. In addition, CPC Ms developed a high performance fluidized-bed incinerator designed to burn a wide variety of liquid and solid wastes separately or together. The pilot scale unit is 3 ft I„D. and is operated at a high superficial velocity (up to 15 ft per second) at atmospheric pressure. The process has been used for disposal of municipal wastes, sewage sludge, and industrial wastes. In some cases a chemically active bed material was employed to react with undesirable products of combustion. The Bergbau-Forschung (a research organization for German coal industry, Steinkohlenbergbauverein) in Essen, Germany, has been working on fluidized-bed combustion of coal in their 16-inch x 3.1-inch atmospheric combustor for two years. The coal feed rate is about 100 lb/hr. The combustion conditions investigated include temperature from 1470 to 1560 F and superficial gas velocity from 4 to 10 ft/sec. For a 2 percent sulfur coal a Ca/S ratio of four is normally required for over 90 percent sulfur removal, although ratios sometimes as low as 1.3 to 2.0 have been achieved. A special feature of this test rig is the unique grid design which utilizes metal bars for directing combustion air downward first into inverted cone-shaped plenums to avoid shifting of coal ash. Bergbau- Forschung is planning an 8-atm pressure fluidized-bed combustion unit for a capacity of 10 to 15 MW with an experimental gas turbine. The turbulent layer furnace (fluidized-bed furnace) developed by Lurgi Gesellschaft fur Chemie und Huttenwesen mbH has been used for combustion of low-grade fuels such as carboniferous tailings resulting from coal preparation, coal-containing clays, low-grade lignites and lignite-containing strippings, oil sands, oil shales, and bituminous marls which have heating values as low as 1,800 Btu/lb. For fuels with higher heating value, the extraction of heat from the fluidized combustor is necessary. The power output for a 1,000 ton/day plant in the order of 10 MW is in commercial practice. The combustion residues could be used for the production of building aggregates. Lurgi has also been developing the fast-bed combustion technique of Dr. Reh for the chemical processing of fine materials. For possible application to fluidized-bed combustion of coal, Lurgi has indicated that considerable research needs to be done in regard to particle agglomeration, elutriation, particle strength, and erosion problems. We have also tried to trace back Battelle's efforts in this area and believe the first work was for the Delaware Clay Company of Delaware, Ohio, in 1949. This was a program led by John Foster, who many of you may remember as the Chairman of the ACS Fuels Division back in 1952. ILLINOIS STATE GEOLOGICAL SURVEY LIBRAS* 51 The bench-scale reactor used is shown schematically in Figure 9. The pur¬ pose was to attempt to bloat shale to produce an expanded aggregate. Later work involved roasting of ores^°\ oxidation of waste pulping liquors (59-60) ^ and more recently the combustion of coal. Figure 10 shows a 6-inch diameter unit built under support of the Battelle Energy Program (kO. two pilot-plant reactors of 10 and 24 inches in diameter for operation at atmospheric pressure are shown in Figure 11; and Figure 12 shows the agglomerating ash fluidized-bed combustor which will be used to generate heat for the Battelle-Union Carbide gasifica¬ tion process^^). Finally, Figure 13 shows a photograph of Battelle's fast bed model which is being used to explore the ramifications of Dr. Reh's^3) and the City College of New York's (^4) wor j c this area. Application to Lignite It would not be appropriate, at a symposium on lignite, not to review the special characteristics of these coals which might make them attractive for use in fluidized-bed combustors. First, of course, is the tremendous reserves of this low-rank coal, justifying the development of combustion systems uniquely suited to the characteristics of lignites. Second, some lignites, although having a high grindability index, prove difficult to pulverize in conventional equipment as used for bituminous coal. The Hardgrove grindability can vary substantially with moisture content of some lignites, so that grindability data must be used with caution. Fluidized-bed combustors handle crushed coal satisfactorily without the need for pulverizing, hence a wide variety of lignites can be burned in fluidized beds without major concern for mill capacity, moisture content, or size consist. Third, the fouling characteristics of lignite, because of its high sodium content, may be minimized in fluidized-bed combustors. Experience has demonstrated in pulverized-coal-fired boiler furnaces that the fouling of heat-receiving surfaces becomes much more troublesome as the alkali in the fuel is greater. Although the exact mechanism is not well defined, it is believed generally that sodium and potassium in coal, often as salts of carboxylic acids but also present in mineral forms such as halite or feldspars, are volatilized in the pulverized-coa1 flame where temperatures can exceed 3000 F momentarily. These volatilized metals then convert to Na 20 and K 2 O, and eventually to Na 2 S 04 and I^SO^ which condense on the surface of fly-ash particles suspended in the gas stream. This alkali-rich surface thus provides a sticky layer on the ash particle to enhance the build-up of deposits on heat-receiving surfaces such as wall tubes and on superheater and reheater tube banks. Ash deposition, then, becomes a function of alkali content, and it has been shown conclusively that such fouling when burning lignite depends strongly on the sodium content of the lignite when burned in pulverized form. At the relatively low temperature 52 Raw shale feed Fume hood Sillimanite rammed ning Steel pipe Fluidized shale Perforated bed plate Downflow tube y —Natural y Fuel-air mixing chamber Product collection pan FIGURE 9. - FIuidized-bed furnace for bloating shale. 53 FIGURE 10. Battelle's 6-inch combustor fluidized bed. 5U FIGURE 11. - Pilot-scale fluidized-bed experimental facility. 55 FIGURE 12. - Process development unit for fluidized-bed coal gasification program at Battelle. Burner vessel is being hoisted into place. 56 FIGURE 13. Battelle's 4-inch fast bed apparatus. 57 of fluidized-bed combustors, the amount of sodium vaporized might be only one ten-thousandth that in a pulverized-coa1 flame. Thus fluidized-bed combustors may be more tolerant of high-alkali fuels than are the higher temperature combustion systems. This is particularly important for the lignites containing more than 0.5 percent Na 20 in the coal and 10 percent in the ash. Fourth, most lignites contain appreciable quantities of calcium and magnesium, probably again as salts of carboxylic acids but also present in mineral forms such as calcite and dolomite. These are potent fluxes to reduce the fusion temperature of the ash so that most lignites produce ash with a strong tendency to form a highly fluid molten slag, or at least to have a high fouling potential because of low ash-fusion temperature.(65) Again, at the low temperatures in fluidized-bed combustors, ranging say from 1500 F to 1800 F, the inorganic matter in lignite will not fuse and so fouling and slagging problems are not likely to occur. Fifth, this high calcium and magnesium content of lignites coupled with relatively low-sulfur content provides a built-in sulfur-capture system if the combustion temperature is kept low enough. This was first pointed out by Mr. Gronhovd at the previous Lignite Symposium(66) . On the assumption that a lignite contains 1 percent sulfur, 10 percent ash, and with 25 percent CaO and MgO in the ash taken as CaO, and assuming further that the combustion temperature remains below 1600 F so that any CaSO^ formed is not thermally dissociated, there is about 1.4 times more CaO available than is necessary to react with all the sulfur to form CaSO^. On this basis, lignite ash rather than limestone is suitable as the inert material in a fluidized bed with capture of essentially all the sulfur that otherwise would be emitted as SO 2 . Such capture will not occur at high furnace temperatures since CaSO^ is unstable above about 1800 F. This feature of fluidized beds has not been exploited because most bitum¬ inous coals contain less than 5 percent CaO and MgO; it is a promising advantage for high-CaO lignites. Sixth, since fluidized-bed combustors operate normally with low combustible content in the bed, they are particularly suited to burning high-ash fuels. Most lignites from North Dakota contain less than 12 percent ash, but lignites elsewhere in fields not being exploited as yet contain as much as 30 to 40 percent ash. Although these lignites can be burned in pulverized form, they will be penalized by a high-ash burden in the flue gas, requiring exceptional efforts to keep heat-receiving surfaces clean and to handle ash settling out of the flue gas. These problems would not occur in fluidized-bed combustors operating at low gas velocity. Whether or not fluidized-bed combustion will be preferable with such high-ash lignites remains to be demonstrated, but the principles appear sound. 58 Conelusions In contrast to the current arguments over scrubber technology, the feasibility of generating power by fluidized-bed combustion seems almost uncontested. Fluidized-bed combustion technology is operable and appears capable of reducing both sulfur and nitrogen oxides emissions in large-scale applications. The economic picture still remains unsettled, but preliminary cost estimates of the technology as compared to those for conventional boilers plus scrubbers tend to favor fluidized-bed combustion. The relative costs of fluidized-bed technology as compared to conventional firing without scrubbers remains in doubt, but it is possible that fluidized systems would be favored in the long run in advanced power systems even without environmental regulations. Remaining problems involve control of particulate emissions, reduction of limestone requirements, and erosion and corrosion of materials. In addition, further attention is needed on the difficulties of pressurized operation, on the development of process controls, on further reduction in gaseous emissions, and on the regeneration and dis¬ posal of solids. Finally, support work on the development and use of fundamental information to assist in commercialization is generally lacking. In addition, a word of caution on possible new environmental prob¬ lems caused by the process itself should be made. We are not yet sure about possible emissions of increased amounts of organic material due to low temperature operation, the generation of ultrafine particles by the bed itself, the release of trace but potentially harmful materials by heating large volumes of various grades of limestones or dolomites, nor the disposal of solids which may be in a more leachable condition after conditioning in a chemically active bed. The widespread ramifications of fluidized-bed combustion, not only for direct power generation and pollution control, but also for energy- related applications in gasification, carbonization, synthesis, and possibly liquefaction, as well as for waste disposal, appears to justify a major research and development investment in this technology. Acknowledgments We wish to express our appreciation to the many organizations that have supported our work in this area, but especially to the EPA and OCR (now ERDA) during the past two years. 59 References (1) Stratton, J. F., Power, 68., 486 (September 1928). (2) Godel, A., and Cosar, P., AIChE Symposium Series 116, Volume 67, 210 (1971). (3) Novotny, P., Sb. Prednasek 50 (Padesatemu) Vyroci Ustavu Vysk. Vyuziti, Paliv, 104-11 (1972). (4) Frank-Kamenetskii, D. A., ,f Diffusion and Heat Transfer in Chemical Reaction", Academic Press, USSR, Moscow, 1947. (5) Ghosh, B., Sc. D. Thesis, Carnegie Institute of Technology, 1951. (6) Spaulding, D. B., in "Fourth Symposium on Combustion", pp. 847-64, Waverly Press, Baltimore, 1953; J. Inst. Fuel, 26., 289 (1953). (7) Putnam, A. A., "Combustion-Driven Oscillations in Industry", Fuel and Energy Science Series, American Elsevier Publishing Company, Inc., New York (1971). (8) Thring, M. W., and Essenhigh, R. H., in "Chemistry of Coal Utiliza¬ tion, Supplementary Volume", H. H. Lowery, Editor, pp 754-766. (9) Essenhigh, R. H., Ph.D. Thesis, University of Sheffield, 1959. (10) Powell, A. R. , Ind. Engng. Chem. , _L2, 1969 (1920). (11) Khundkhar, M. H., J. Indian Chem. Soc., 24, 407 (1947). (12) Nishihara, K., and Kondo, Y. , Ryusan, 11., 43, 89 (1958). (13) Shmuk, E. I., Izv. Akad. Nauk SSSR, otd. Tekh. Nauk, Met. i Toplivo p. 177 (1959) (Chem. Abstr. 54: 1654). (14) Schwab, G. M., and Philinis, J. , J. Am. Chem. Soc., 69_, 2588 (1947) (15) Ogale, B. S., and Krishnaswami, K. R., Curr. Sci. , _14_, 21 (1945). (16) Malin, K. M., Zh. khim. Prom., _16, 4 (1939). (17) Mendelsohn, N., Pincovschi, E., and Pintilie, S., Revue Chim. Buc., 10, 199 (1959). (18) Charrier, J., Bull. Soc. Hist. Nat. Toulhouse, 85, 317 (1950) (19) Flint, D., and Copson, A. J., CRURA Inf. Circ., 1955. (20) Kopp, 0. C., and Kerr, P. F., Am. Miner, 43, 1079 (1958). 60 (21) Karavaev, N. M., and Amagaeva, V. N., Khim. Kl. Iskop Uglei (Chemistry and Classification of Coals), p. 164 (1966). ( 22 ) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) Malet, A. M., Khim. Prom., 5, 329 (1964). Given, P. H. , and Wyss, W. F., BCURA Monthly Bull., 2_5, 165 (1961). Given, P. H. , and Jones, J. R. , Fuel, Lond. , 45^, 151 (1966). Oxley, J. H„, Ph.D. Thesis, Carnegie Institute of Technology, 1956. Blum, I., and Cinda, V. , Pop. Romine Inst. Energ. Studii, 11 , 325 (1961). Muntean, V. C., Akad, Rep. Populare Romine, Studii Cercetari Met., 8, 331 (1963). (Chem. Abstr. 60: 2689). Sinha, R. B., and Walker, P. L., Jr., "Removal of Sulfur from Coal by Air Oxidation at 350-450 C", Fuel, 51 , 125 (1972). Harrington, R. E., Borgwardt, R. H., and Potter, A. E., Amer. Ind. Hyg. Ass. J., 2_9, 52 (1968). Falkenberry, H. L. , and Slack. A. V. , Chem. Eng. Progr., 6_5 (12), 61 (1969). Coutant, R. W., Barrett, R. E., and Lougher, E. H., "SO 2 Pickup by Limestone and Dolomite Particles in Flue Gas", American Society of Mechanical Engineers Preprint No. 69-WA/APC-l (1969). Potter, A. E., Amer. Ceram. Soc. Bull., 48 (9), 855 (1969). Attig, R. C., and Seder, Paul, "Additive Injection for Sulfur Dioxide Control", Babcock and Wilcox Co. Research Center Report 5460, Research Center, Alliance, Ohio, 1970. Davidson, D. C., Small, A. W., 2nd International Conference on Fluidized Bed Combustion, Hueston Woods, Ohio, October 4-7, 1970. Harvey, R. D., "Petrographic and Mineralogical Characteristics of Carbonate Rocks Related to Sorption of Sulfur Oxides in Flue Gases", Interim Report to the National Air Pollution Control Administration, Contract Number CPA 22-69-65, June 22, 1970. Borgwardt, R. H., and Harvey, R. D., Environ. Sci. Technol., 6_ (4), 350 (1972). O'Neill, E. P., Keairns, D. L., and Kittle, W. F., "Kinetic Studies Related to the Use of Limestone and Dolomite as Sulfur Removal Agents in Fuel Processing", 3rd International Conference on Fluidized Bed Combustion, Heuston Woods, Ohio, November 1972. 6l (38) Wen, C. Y., and Ishida, M. , "Reaction Rate of Sulfur Dioxide with Particles Containing Calcium Oxide", Environ. Sci. Tech., 1_ (8), 703 (1973). (39) Hsieh, B. C., Ashworth, R. A., and Switzer, G. W., Jr., "An Analysis of Chemistry and Mechanisms for High Temperature Desulfurization of Low Btu Gas When Using Lime or Limestone", Prepared for the Office of Coal Research, Washington, D.C., Contract No. 14-32-001-1236, by Gilbert Associates, Inc., May 31, 1974. (40) Skinner, D. G., The Fluidized Combustion of Coal , Mills & Boon Limited, London (1971). (41) Robison, E. B., et al., "Development of Coal Fired Fluidized-Bed Boilers", OCR R&D Report No. 36, Vol. 1, II (1972). (42) Coates, N. H., and Rice, R. L., "Proceedings of the Second Inter¬ national Conference on Fluidized-Bed Combustion", sponsored by NAPCA, Houston Woods, Ohio (1970). (43) Zielke, C. W., et al., "Sulfur Removal During Combustion of Solid Fuels in a Fluidized-Bed of Dolomite", Journal of Air Pollution Control Association, 2_0, 3 (1970). (44) CSIRO, Journal of Fuel and Heat Technology, 15_ (5), 11-13 (1968). (45) Henschel, D. B., "Status of the Development of Fluidized-Bed Boilers", 1971 Industrial Coal Conference (October 1971). (46) Carls, E. L. , "Review of British Program on Fluidized-Bed Combus¬ tion", Report of the U.S. Team Visit to England, ANL/ES-CEN 1000 (1969). (47) Robison, E. B., et al., "Characterization and Control of Gaseous Emissions from Coal-Fired Fluidized-Bed Boilers", PER Report to EPA (1970). (48) Gordon, J. S., et al., "Study of the Characterization and Control of Air Pollutants from a Fluidized-Bed Boiler--The SO.-, Acceptor Process", PER Report to EPA (1972). (49) Robison, E. B., et al., "Study of Characterization and Control of Air Pollutants from a Fluidized-Bed Combustion Unit—The Carbon- Burnup Cell", PER Report to EPA (1972). (50) Jonke, A. A., et al., "Reduction of Atmospheric Pollution by the Application of Fluidized-Bed Combustion", Annual Report ANL/ES- CEN 1004 (1971). 62 (51) Jonke, A. A., et al., "Reduction of Atmospheric Pollution by the Application of Fluidized-Bed Combustion", Annual Report (1974). (52) Skopp, A„, et al., "A Regenerative Limestone Process for Fluidized- Bed Coal Combustion and Desulfurization", Final Report to EPA by Esso R&E (1971). (53) Hoke, R. C., et al., "A Regenerative Limestone Process for Fluidized- Bed Coal Combustion and Desulfurization", EPA-650/2-74-001, Esso R&E Report to EPA (1974). (54) Archer, D. H., et al., "Evaluation of Fluidized-Bed Combustion Process, Vol. II", Westinghouse Report to EPA, PB212, 960, November 1971. (55) Gordon, J. S., et al., "Study of the Characterization and Control of Air Pollutants from a Fluidized-Bed Boiler--The S0 ? Acceptor Process", PER Report to EPA, EPA-R2-72-021, July 1972. (56) Archer, D. H., et al., "Evaluation of the Fluidized-Bed Combustion Process", Volumes I, II, and III. A Report to EPA by Westinghouse Research Laboratories (1971). (57) Keairns, D. L., et al., "Evaluation of the Fluidized-Bed Combustion Process", Volumes I, II and III, Westinghouse Report to EPA (1973). (58) Stephens, F. M., "The Fluidized Bed Sulfate Roasting of Nonferrous Metals", Chem. Engr. Prog., 4_9 (9), 455 (1953). (59) Copeland, G. G., and Hanway, J. E OJ Jr., "Fluidized Bed Oxidation of Waste Liquors Resulting from the Digestion of Cellulosic Materials for Paper Making", U.S. Patent No. 3,309,262, March 14, 1967. (60) Smithson, G. R., Jr., and Hanway, J. E., Jr., "Process of Converting Sodium Sulfate to Sodium Sulfite, Particularly for Pulping Processes", U.S. Patent No. 3,397,957, August 20, 1968. (61) Locklin, D. W., Hazard, H. R., Bloom, S. G., and Nack, H., "Power Plant Utilization of Coal", A Battelle Energy Program Report, Columbus, Ohio (1974). 96 p. (62) "Status of the Battelle/Union Carbide Coal Gasification Process Development Unit Installation", Carder, W. C., and Goldberger, W. M. , Sixth Synthetic Pipeline Gas Symposium Proceedings , Chicago, Illinois, October 28-30, 1974, 21 pp, American Gas Association. (63) Reh, L., "Fluidized Bed Processing", Chem. Eng. Progr., 6_7 (2), 58-63 (1971). 63 ( 64 ) Yerashalmi, J., Mclver, A. E., and Squires, A. M., "The Fast Fluidized Bed", GVC/AIChE Joint Meeting, Munich, Germany, September 17-20, 1974. (65) Reid, W. T., "External Corrosion and Deposits - Boilers and Gas Turbines", Fuel and Energy Science Series, J. M. Beer, Editor, Elsevier, New York, 1971. (66) Gronhovd, G. H., Tufte, P. H., and Selle, S. J., "Some Studies on Stack Emissions from Lignite Fired Power Plants", Grand Forks Energy Research Laboratory. Presented at 1973 Lignite Symposium May 9-10, 1973. 6h SCRUBBER DEVELOPMENTS IN THE WEST by Everett A. Sondreal—^ and Philip H. Tufte—^ INTRODUCTION Most of the 30 commercial size scrubber modules operating in the Western United States have been installed for particulate removal, where their selection over electrostatic precipitators was, in many cases, motivated by the poor performance of the first precipitators operated on low-sulfur Western coals. Only six commercial-scale scrubber modules have been installed in the West specifically to remove SOp from powerplant stack gas, and these must be considered to be largely experimental units. These installations represent three different reagents—1-ime, limestone, and soda ash—and they include a diversity in design. The particulate scrubbers, operated without an added reagent, also remove an appreciable fraction of the flue gas SO 2 , due to the inherent alkalinity of the fly ashes produced from Western coals. The deliberate use of ash alkalinity for flue gas desulfurization has been investigated in three separate pilot-scale studies, and commercial installations are being built. It is the purpose of this paper to describe the special requirements that need to be considered in the design of wet scrubbers for Western coals, and to present supporting information on the scrubbers that are being operated. Hopefully, the judgments that led to the selection of the current designs and the experiences arising from their operation can be used to guide future developments and to avoid costly mistakes. There is a clear consensus among utilities endeavoring to operate scrubbers that substantial improvements are essential. 1 J Research Supervisor, Grand Forks Energy Research Center, U.S. Energy Research and Development Administration, Grand Forks, N. Dak. 2/ Chemical Engineer, Grand Forks Energy Research Center, U.S. Energy Research and Development Administration, Grand Forks, N. Dak. 65 The text of this paper discusses the properties of Western coals, their emissions, emission standards, and the design and performance of operating scrubbers. Any value judgment that is stated or implied in the text is the opinion of the authors and not that of the parties who provided the information for this paper, unless otherwise stated. An appendix is provided listing data on operating scrubbers so that interested persons will be assisted in reaching their own conclusions. ACKNOWLEDGMENTS Information contained in this paper was obtained from published sources and, in part, from inquiries directed to utilities operating wet scrubbers on boilers burning Western coals and to vendors. Persons who have contributed information include Messrs. Lyman Mundt, Gilbert Gutierrez, Samuel Bayless, and Aubrey Parsons of Arizona Public Service Company; Messrs. Peter Smith and'Herbert Braden of Research Cottrell; Messrs. James Zornes and David Barneby of Nevada Power Company; Mr. Thomas Ashton of Pacific Power and Light; Mr. George Green of Public Service of Colorado; Messrs. Conrad Aas and John Noer of Northern States Power Company; Mr. Eldon Kilpatrick of Minnesota Power and Light; and Mr. Carlton Grimm of Montana Power Company. The contributions of these and others who have freely exchanged information in numerous past contacts concerning scrubbing are gratefully acknowledged. WESTERN COALS Western coal production has increased rapidly over the past five years as shown in table 1. The Western reserve base for measured and indicated coal in place, as defined by the U.S. Bureau of Mines, totals 2l6 billion tons (iO— and is distributed by state as shown in table 2. Montana, Wyoming, North Dakota, and Colorado top this list in reserves. Production in 197^ was at a rate greater than 2 million tons/yr in each of 11 Western states, with Wyoming producing the largest tonnage. Wyoming and Montana are expected to experience very large increases in mine capacity. Firm plans for mine expansion will raise capacity in the West above 200 million tons per annum by 1983 (3_), not including tentative plans for a number of coal gasification plants. Considering these projections, stack gas cleaning technology for burning Western coals will assume much greater importance in the future than at present. 3/ Underlined numbers in parentheses refer to items in the list of- references at the end of this report. 66 TABLE 1. - Total Western coal production— Millions of Percent of Year short tons U.S. total 1970 46 7.6 1971 52 9-4 1972 62 10.5 1973 76 12.8 1974 85 l4.4 1/ For the : states that are included, see table 2. TABLE 2. - Coal in the Western U.S., reserves and production by state Millions of short tons New mine—^ State In-place coal Estimated 1974 capacity reserve (l) production (2) by 1983 (3) Arizona 350 3.2 8 o / Colorado 14,870 6.9 NA- Kansas 1,388 .8 .5 Missouri 9,488 4.3 NA Montana 107,727 13.6 21 New Mexico 4,394 9.5 5 North Dakota 16,003 7.2 NA Oklahoma 1,294 2.4 NA South Dakota 428 .0 NA Texas 3,272 6. o .1 Utah 4,042 6.5 1.3 Washington 1,954 3.9 1 Wyoming 51,228 20.5 82.5 216,439 84.8 119.4 1/ New capacities reported by 1983 are those given by a recent Keystone survey (3_) and are representative of firm plans announced by major producing firm. The values given do not reflect plans for coal gasification. 2/ NA indicates that no value was given in reference 3. Some of the states affected are known to be experiencing major expansions in coal production. 67 Properties of Western Coals The problems of stack gas cleaning start with the characteristics of the coal. Western reserves include lignite, subbituminous, and bituminous coal, with the lower rank coals predominating. An important property of almost all Western coals is that they contain far less sulfur than the 2 and 3 pet typical of Eastern and Central coals. Unfortunately, sulfur content in Western coals averaging .7 pet is not generally low enough to meet new source emission standards. On average, 30 to hO pet removal of SOp is required to meet the Federal standard of 1.2 lb SO 2 /MM Btu, and higher removals are required to meet some more stringent state and local standards. The lower sulfur content does, however, make stack gas cleaning potentially easier to achieve and less costly, provided that design innovations capitalize on the advantages of Western coals and emission standards are not raised to cancel out the advantage. Since sulfur oxide emission standards are based on heat released, variations in heating value according to rank have a pronounced effect on the coal sulfur content that is equivalent to the emission standard, as shown in table 3. The nominal TABLE 3. Coal sulfur content equal to emission standards Coal Higher heating value, Btu/lb Coal sulfur equal to the Federal standard of 1.2 lb S0 2 /MM Btu, pet Coal sulfur equal to the Clark Co., Nevada standard of 0.15 lb S0 2 /MM Btu, pet North Dakota lignite 6,800 0 .U 1 0.05 Montana subbituminous 8,600 .52 .06 Arizona ^, (Black Mesa)— bituminous 11,000 . 66 0 Co 1/ Coal from the Black Mesa Mine is burned at the Mohave Station of the California Edison Company in Clark County, Nevada. average of 0.7 pet sulfur in Western coals does not satisfy the Federal standard, and it is an order of magnitude higher than that allowed by the Clark County, Nevada standard. 68 Ash content in Western coals varies greatly between mines, and between locations within mines, with the U to 20 pet shown in table 4 quite representative of the overall range. Dust loading in stack gas depends on boiler design as well as on coal ash content. For a pulverized coal-fired boiler, a rough estimate of dust loading leaving the boiler, in gr/sefd, can be obtained by halving the numerical percentage of coal ash. Thus, dust loadings are typically 2 to 10 gr/sefd. An important characteristic of many Western coal ashes is their high content Of the alkali oxides NapO, MgO, and particularly CaO (table h). Alkali content tends to be highest in the lowest rank coal, lignite, and progressively less prevalent in the subbituminous and bituminous coals. This trend is related to the greater ion exchange capacity of low-rank coals, compared with the higher rank. Variations in alkali content are also influenced by the minerology of the overburden and the course of ground water movement. Alkali content in Western coal ash varies from under 10 to over 50 pet, with important variations occurring within individual mines. A guideline for assessing the importance of the amount of alkali in Western coal is the ratio of the alkali to coal sulfur. For a coal containing 7*5 pet ash and 20 pet alkali in the ash, the total alkali is chemically equivalent to slightly more than 120 pet of a 0.7 pet sulfur content. For some lignites, the alkali/sulfur ratio can be several hundred percent. Thus there is ample alkali to interact importantly with sulfur oxides in a wet scrubber in burning many Western coals. Emission Standards Seven Western states have adopted state-wide S0p emission standards that are more restrictive than the Federal standards (table 5)* No statewide restrictions on particulate is more severe than the Federal. In many states, the severity of the standard depends on the size of the generating unit. Table 5 applies for a size of 500 MW. County restrictions are, in some cases, far more stringent than those applying statewide. The most important example is Clark County, Nevada, where the standards require emission levels below .15 lb SOp/MM Btu and a Ringleman opacity of 1, which is estimated to be equivalent to a dust loading of about 0.02 gr/sef for conditions applying at the Reid Gardner Station of Nevada Power Company. In addition, some states have passed or are considering stricter regulations which become effective 69 -P O co tJ g G p G on * CO o co US G H M O O > G o G ‘H m -P o •H o >> pq •H G X rH O CL) G H • S -H CO P > CJ G S H1 s CD o i—1 o G •H o O X • r "3 OJ CJ G G • S > OJ o G G CVJ G £ S G G -P S CD p G co & G G co co G G G P •H O G S o G O LfN rG •H tSI p H i — 1 • -p W P ■H CJ G b- G G G G G O G -P *G H o CJ O P w G G • • • G . . (L> • rH i—1 -P P g G G 6 o -P -H G o CO S CO CO CO LTN CO b-VO ON vo LTV o G oo 1-1 LTV rH G CVJ 1-1 -p G G O P o G cd b- vo o o O UA CVJ VO CO CO G in 1 —1 C— I-H VO i—1 1—1 Lf\ G LT\ CVJ Ph bD G G G W vo OJ H VO UO ON oo I/O VO CVI in x LTV VO VO CO 1-1 CO G Lf\ CVJ O Ph Ti ^-7 G P G V—X G O G O G •rH o 1—1 b— CO VO CO -4" CO cvi ,G -p ■P G CVJ CO Lf\ b— CVJ 1-1 CO G 1—1 1-1 •P -P cd co •H G G 1-1 •H •H P Tl b— Ov vo L/NVO LTN cvj LTN o G < b- CVJ oo VO L/N OJ b- G -p OJ 1—1 1—1 1-1 1—1 P G O G s G p O O 1-1 cd G LT\ b- oo b- CO VO P b— P P -p > G G UV CO b- U~\ p 1-1 CO T* tt CO 1—1 1-1 1—1 6 o G G cd P ,g tJ o G G -P G b— 1—1 1-1 CO vo On LTV P ITS O G G in ON 1—1 Ov P VO VO ON rH G 1-1 1—1 CVJ 1—1 G K in >3 G bl) G G G G > G W co G G co • >3 -P 1—1 G P G O • on co . < CVJO O OJ Lf\ • O • • o o. CVJ o O O O CVJO CO •rH H G -H OJ G bD cd CVJO CO P Eh A O SI s CO H | 70 TABLE 5- - Emission standards for new coal-fired powerplant in the Western U.S.l/ SO2 Particulate N0 X State lb/MM Btu lb/MM Btu lb/MM Btu Federal Standards 1.2 0.10 0.7(£/ State Standards Wyoming 0.20^/ o.iojj/ .15V 0.70 New Mexico .3b .1+5 Nevada .k 0 .13 5/ Arizona .80 .13 • 70 Montana 1.00 .lU 5/ Missouri 1-17 1 —1 5/ Arkansas 1.20 .10 .70 Oklahoma 1.20 .12 .70 Oregon 1.20 .20 5/ South Dakota 1.20 1.282/ .10 .70 Colorado .10 .70 Minnesota 1-75 .Uo 5/ Washington 2.33 .2° .15^ 5/ Nebraska 2.50 .67 Idaho 2.78 5/ Kansas 3.00 . 27 P • 90 North Dakota 3.00 5/ Texas Iowa 3.00-, 6 . 00 + .30 . 60 5/ 5/ California No statewide standards; local standards vary. Utah 7/ 1/ Standards given are those in effect in May 1975 for a 500 MW plant. All standards have been converted to units of lbs/MM Btu. Conversions when performed were based on combustion of coal having a moisture-free analysis of 65 pet carbon, 1 pet sulfur, and 10,800 Btu/lb, burned with 30 pet excess air. 2/ The Federal Standard for N0 X does not apply to lignite. 3/ This standard is being contested in court. bj The units presented in the regulation are "lb/hr/MM Btu". 5/ Indicates there is no statewide standards. 6/ Effective January 1, 1980, SO 2 is restricted to 0.35 lb/MM Btu. 7/ Powerplants are required to use the best available technology for removal of sulfur dioxide and particulates. Guidelines are for removal of 85 pet of particulate and sufficient sulfur dioxide to attain an equivalent emission to combustion of 1 pet sulfur coal. 71 at a later date, as in Colorado where the standard for SO 2 is .35 lb/MM Btu after January 1, 1980 . Thus, utilities installing stack gas cleaning equipment are sometimes selecting designs which exceed present standards in anticipation of more stringent requirements in the future. Required Removal Efficiencies As already stated, the average sulfur content of Western coals, 0.7 pet, does not permit such coal to be burned without flue gas desulfurization. Some coals contain less than the average sulfur content, and retention of sulfur oxides on ash during combustion may lower the SO 2 emission, by 10 to Uo pet for lignites (5.). Thus there are some Western coals which will meet the Federal standard. There are none that will meet the standards of Clark County, Nevada, of New Mexico or Nevada, or of Colorado after 1980; it is doubtful that any would reliably meet the Arizona statewide standards. The removal efficiencies required to meet the more stringent standards are illustrated in figure 1 for a Western subbituminous coal. At the 0.7 pet coal sulfur level, the required SO 2 removal is increased from about 30 pet to meet the Federal standard to 90 pet to meet the Clark County, Nevada standard. At an inlet dust loading of 5 gr/sefd, corresponding to the approximately 80 pet of the ash from a 10 pet ash coal that would be released as fly ash using pc firing, the required particulate removal would be increased from 99.0 pet to meet the Federal standard to 99-8 to meet the Clark County, Nevada standard. The capital and operating costs for stack gas cleaning must be expected to rise steeply as higher percentage removals are required for substances that are initially present at low concentrations. For an approximate estimate based on principles of engineering design, it can be assumed that equipment size and power requirements for wet scrubbing will increase in proportion to the logarithm of one over the required exit concentration (size and power oC log — exit), and capital cost will increase as the 0.6 power of size. Under these assumptions, equipment size would triple between 50 pet and 90 pet removal, and would double again at 99 pet removal. Capital cost would first double and then rise by a further 50 pet for the same increases in removal. These figures are hypothetical and they make no allowance for improved design. However, they are close enough to reality to validly demonstrate that high costs will have to be paid to achieve improved control of stack gas emissions. 72 REQUIRED REMOVAL, percent REQUIRED REMOVAL, percent Figure I. - Removal efficiencies required for stack gas cleaning of western coal. 73 FULL-SCALE PARTICULATE SCRUBBERS There are 2h commercial size scrubber modules operating in the Western U.S. for the purpose of removing particulate (table 6 ). Sulfur dioxide removal in these units is incidental, not intentional. All units except one are retrofitted installations; and in some cases, the retrofitted scrubbers are in series with a previously installed mechanical ash collector or electrostatic precipitator. Dust loadings entering the scrubbers vary widely, from 0.3 to 12 gr/scfd, because of the series cleaning equipment and large differences in coal ash content. Entering SO 2 levels fall in the range of 500 to 800 ppm. Particulate scrubbers are operated at seven different power stations and by four different utility companies. Three designs are represented; venturi scrubbers (Chemico), spray towers with mobile bed packing (UOP), and high-pressure spray impingement scrubbers (Krebs). The combined capacity of all units is 2,2h0 MW, which is 76 pet of total utility scrubber capacity in the West, including the units operating to remove SO 2 . All units were installed within the last five years, and have startup dates between June 1971 and July 197*+• Arizona Public Service Company, Four Corners Plant (j, 8) The Four Corners Plant at Farmington, NM has three pc- fired boilers (2-175 MW, and 225 MW) equipped with Chemico venturi scrubbers for particulate removal, with two scrubber modules on each boiler. Two additional boilers of 755 MW each are equipped with electrostatic precipitators (ESP's). The first scrubbers began operation in December 1971. Total cost of the scrubber system was $30 million, or $52 per kw. The subbituminous coal burned at the Four Corners Plant, supplied by the Navajo mine, has as its outstanding characteristic an unusually high ash content of nominally 22 pet. This results in a high dust loading in the flue gas leaving the boiler, 12 gr/scfd, and initial operation of the plant using mechanical collectors for fly ash removal resulted in dense plumes from the stacks, which dispersed to restrict visibility over a large surrounding region. Since scrubbers and ESP's have been operating, the plumes are greatly improved, although still visible. The Navajo coal averages 0.68 pet sulfur, which produces an SO 2 content of approximately 650 ppm. The alkalinity of the fly ash is low for a Western coal, with U pet CaO in the ash. TABLE 6. - Condensed summary of operating vet scrubbers in the Western U.S. •h .o .. I- 3 U < ft to O ft i CO aJ ° s •h a) oj w i 3 a> : & co o P aJ TJ P a) P fe § S § ■§■2 a. p . i ft 4J O TJ P c o c >» ^ a) £ o w ® S H CO -H CO U 3 aJ p < -H p O CO u CM-h aJ 3 • 3 P S p t) ^ a p to u O S -h c aJ cd co o 3 ft) > • • cr > w s ■a xj w. p t, 0) -H P i co a) (h a» oj p 3 t? a co p p a* I -H ft) O C I ft ffi O 4) O 1, >0) O CO 104 CM CM U*n 4j 4> 4T CM >» >. IA (0 (O r—I ® D LTV ft >> >> CO CO < (I) d Z >» >» CM CO CO » >* CO CO CO O j h ® ti m ft >» >> LTV CO O CM CO VO t— OJ c lA UA >» O O t- o a 4) o H l. ^ ft) ft) 3 ft P V, ft P o •H 3 co 3 CO P ft ft C d >i a) ft CC CQ U VO CM O • • CM O LTV t— • • l o o VO OJ O . . CM o crv ov co C O OJ i/v • 4J ir\ cm • O CO CO CM VO O CM <=9 ° -=* ' CO • 3 P cr o » u u. ft> ftj o o p u ft) a; o 5 < w w O Ov o o c— i P4 04 (MOW • U-V On P ft -X Os P O V* OJ CO CO o On CM O PO CO On O On OJ • O ft O O O O CAN CO CM CO o o o o ^ CM lA CM ft VO • O Ov CA C 4) • O U J • 4) OOP - p P ft o 75 However, the amount of calcium entering the scrubber system is still relatively large owing to the high percentage of ash. The calcium is chemically equivalent to approximately 75 pet of the coal sulfur content; and total alkalinity including small amounts of Na o 0 and MgO exceeds 100 pet of sulfur equivalence. A detailed description of the Four Corners scrubber installation is given in Appendix I and in figure A-l. Flue gas from the air heaters enters the venturi and then passes consecutively through a mist eliminator, a wet ID fan, another mist eliminator, and a steam reheater. There is no bypass. Turndown is 50 pet. Reheaters have been removed and the units operated with wet stacks during the past year. Scrubber liquor is continuously recycled from the cyclone separator back to the venturi, and blowdown from the cyclone is sent to a thickener. Lime is added in the thickener. The ash settles well. Thickener under flow is diluted and pumped to ash ponds. Sludge presently is being allowed to accumulate in the ponds, but eventually it may be dredged and returned to the mine. Key operating variables are a liquid to gas ratio, L/G, of 9 gal/1,000 acf and a total pressure drop of 28 inches of water. The pH of the venturi recycle liquor is 3.2 to 3.5- In the thickener, pH is U to 5 without lime, but it is presently being maintained at 7.5 with addition of lime. Operating costs are not available. The major operating requirements are electrical power equal to 3 to U pet of generating capacity, an estimated water usage of 5-9 acre ft/MW/yr, manpower includin 8 operators plus maintenance and supervision, and 10 tons of lime per day for control of pH. Operation of the venturi scrubbers has been satisfactory from the standpoint of meeting the particulate removal goal of 99-2 pet. Sulfur dioxide removal is 30 to 35 pet without addition of lime. The present lime addition rate is equivalent to 7 pet of the SOp in the flue gas being treated, and would be expected to improve SO 2 removal by an estimated 5 pet; however, no measured values are available with addition of lime. Availability for the scrubber system is currently estimated by plant personnel at about 80 pet. This is the fraction of the time that a boiler is operating or could be operating that the scrubber modules are also operative. Since there is no bypass, this level of availability involves appreciable loss in power generation. With two scrubbers per boiler, lost generation can involve either reduction in load when one scrubber is inoperative or complete boiler shutdown when both scrubbers are inoperative. 76 Principal operating problems have involved solids buildup in blowdown lines, corrosion and leakage in lines and vessels where coatings on carbon steel have failed, and most importantly- scaling. Scaling has occurred on most wetted surfaces, and it is not yet under control. Control measures include the use of an appreciable amount of blowdown (open-loop operation), a recent increase in the percentage of fly ash solids in the recirculating scrubbing liquor (from 2 to 6 pet), and addition of a new lime add system to maintain the pH at 7.5 in the thickener. It is not clear why this pH adjustment insures improved control of scaling, since in a system where the state of oxidation is high, with most dissolved sulfur present as sulfate, calcium sulfate would not be expected to be precipitated in the thickener by the rise in pH (6^). The measures being investigated for scale control at Four Corners, if successful, should find wide application in scrubbing in applications involving low-sulfur Western coals. Pacific Power and Light, Dave Johnston Plant (8, 9) The Dave Johnston Plant at Glenrock, Wyoming has one 330 MW pc-fired boiler equipped with three parallel Chemico venturi scrubbers. Initial cost was $8 million, or $2U/kw. A reported $5 million has been spent on improvements ( 10 ), bringing total cost to $39/kw. Startup was in April 1972. Coal burned at the Dave Johnston Plant is Wyoming subbituminous coal from a captive mine. Sulfur content is 0.5 pet, resulting in an SO 2 content of 500 ppm in the untreated flue gas. Coal ash content is 12 pet, and the CaO content of the ash is approximately 20 pet. The calcium is chemically equivalent to 275 pet of the sulfur content. Inlet dust loading is k gr/sefd. A detailed description of the Dave Johnston scrubber installation is given in Appendix II and figure A2. Flue gas from the air heaters enters the venturi and then passes consecutively through mist eliminators, to a wet ID fan, and on to a wet stack. No reheat is used; there is no bypass. Turndown is to approximately 30 pet of scrubber design capacity. Scrubbing liquor is continuously recycled from the bottom of the venturi scrubber back to the plumb bob and to the deflector surrounding the bob that was installed to prevent solids buildup. Blowdown from this loop is pumped directly to two fly ash 77 settling ponds; no thickener is used. Overflow from the settling pond is sent to a clear pond. Some lime is added to the scrubber for pH control. Clear liquor from this pond is pumped back to the recycle loop. Ash is dredged from the settling ponds once each year and is hauled away for land fill. Key operating variables are an L/G of 13 gal/1,000 acf and a total pressure drop of 15 inches of water. The pH leaving the scrubber is 5, without lime addition. Operating costs are not available. Major operating requirements are electrical power equal to 2.3 pet of generating capacity, an estimated water requirement of 3.6 acre ft/MW/yr, lime for control of pH, and manpower. Particulate removal efficiency is over 99 pet, meeting the design goal of .Ob gr/sef. Preliminary values for SOp removal are 40 pet without lime and somewhat higher with lime addition. Availability is characterized by PP&L as being less than adequate for utility use, but no company sanctioned percentages are available. Availability depends on the amount of blowdown and fresh water irrigation that are employed. Operation is characterized as "intermittent open loop," meaning that operation with a minimum of blowdown is attempted as the normal mode of operation, with much larger amounts of blowdown and makeup used periodically to irrigate the system. Operating problems are detailed in Appendix II. The major problem of scaling has been improved but not eliminated by use of lime for pH control and of ligno sulfonate to alter the hardness of deposits. Public Service of Colorado, Valmont, Cherokee, and Arapahoe Stations (ll) Public Service of Colorado has 12 similar TCA scrubber modules installed on pc-fired boilers for particulate control. P-signed by Uni err 1 °il Products, these units consist of tiir.-^ stages of iuob.i.1'' racking, or "ping pong balls," with .,pray it ^et^G uownwe.rd through the balls and gas passing T>-r>T.Tpy(J t 78 Installation of scrubbers by Public Service of Colorado represents a last stage of improvement in particulate control, after previous installations of both mechanical collectors and electrostatic precipitators (ESP's). All of the scrubber- equipped boilers are still serviced by the previously installed equipment. At the Valmont Station, flue gas from the mechanical collector on a 196 MW boiler is split into two parallel stream, with 60 pet sent to the scrubbers and 40 pet to the ESP. On three boilers at the Cherokee Station (115 MW, 170 MW, and 375 MW) and on one boiler at the Arapahoe Station (112 MW), scrubbers are installed in series after a mechanical collector and an ESP. Coal burned at the Valmont and Arapahoe Stations is Wyoming subbituminous, having 0.6 pet sulfur, 5-2 pet ash, and 20 pet CaO in the ash. Calcium content is chemically equivalent to 99 pet of sulfur content. At Cherokee, Colorado bituminous coal is burned, having 0.7 pet sulfur, 9*4 pet ash, and 5 pet CaO in the ash. Calcium content in this coal is chemically equivalent to 38 pet of sulfur content. Dust loadings, given in Appendix III, range from 0.4 to 0.8 gr/sefd. All units have inlet SO 2 levels of nominally 500 ppm. Detailed descriptions of the Public Service Company of Colorado installations are given in Appendix III and figure A- 3. Flue gas from a booster fan is directed through the scrubber, to the chevron mist eliminators, and on to a reheater. Reheat on most units is accomplished by heating the flue gas directly with steam coils; the Cherokee No. 4 unit uses externally heated air. All units have bypasses. Typical turndown capability is from 47 to 105 pet of rated scrubber capacity. Scrubbing liquor is recycled from the bottom of the scrubber back to the spray header above the mobile bed. Blowdown is mixed with bottom ash, and with lime as needed to bring pH into range of 6.5 to 8.5, and is then sent to settling ponds. Ash sludge is dredged periodically for landfill. Clear effluent from the ponds, or from cleanup clarifies at the Cherokee Station, is discharged under permit from the state of Colorado. Operation is "open loop". Key operating variables are a high L/G of 50, a total pressure drop of 10 to 15 inches of HpO, and pH of 7 to 9 entering the scrubbers and 2.8 to 3 leaving. 79 Operating costs are not available. Operating requirements are electrical demand equal to 4 pet of the power generated (very high) and steam for reheat (amounts in Appendix III). Water requirements are approximately 2.8 acre ft/MW/yr. No reagents or addit-ives are normally used. Manpower for scrubber operation is not identified by the company separate from other plant operations, but their estimate is 1-1/2 to 2 man per scrubber per shift for operation and 4 man per scrubber per day for maintenance. These manpower requirements are high. Particulate removals meet the design goal of 0.02 gr/sefd, requiring 95 to 98 pet removal. Sulfur dioxide removal is 40 to 45 pet at the Valmont and Arapahoe Stations, burning Wyoming coal, and 20 pet at the Cherokee Station, burning Colorado coal. The difference in SOo removals is caused by difference in ash alkalinity. Availabilities for individual units are listed in Appendix III. Performance by this measure has ranged from poor (20 to 40 pet availability on Arapahoe No. 4) to moderately good (85 pet on Cherokee No. 4). Operating problem include wear and periodic replacement of the mobile balls, erosion of linings, corrosive failure of reheaters, and, of course, scaling. Scaling and plugging has occurred at the wet/dry zone, on the first stage grids, and in the reheaters. Various additives have been tried for control of scaling, including phosphated esters, but without significant success. Blowdown is maintained at an adequate level, but otherwise no chemical control measures are currently practiced. With 870 MW of installed scrubber capacity. Public Service Company of Colorado is the largest user of scrubbers in the West. Capital cost of the TCA scrubbers, averaging $33/kw, has been moderate. Operating cost, including an electrical requirement of 4 pet of generating capacity, is judged to be high. Availability is not adequate by the standards of electrical utilities. The company is engaged in research to convert present TCA scrubber to lime or limestone scrubbing, but no results are yet available. 80 Minnesota Power and Light Company, Clay Boswell and Aurora Stations (12) Minnesota Power and Light has Krebs-Elbair spray impingment scrubbers installed on two retrofitted 58 MW boilers at Aurora (startup June 1971) and on one new 350 MW unit at Cohasset (startup May 1973). Two additional retrofit installations of 70 MW each are under construction at Cohasset. The Elbair scrubber is a stainless steel box containing nozzles which direct high pressure spray against baffles consisting of either vertical rods or a punch plate (figure A-4). The impingement of the spray against the baffles causes it to be finely atomized, and the induced turbulence promotes effective scrubbing of particulate. It is the theory of this device to substitute power input in the high pressure spray in place of pressure drop in the gas stream, presumably affecting a net saving in power and cost. Another important feature is a nozzle tree design which permits sections of nozzles to be removed for maintenance without shutdown. A design limitation is the apparent inability of this scrubber to tolerate more than very low levels of recirculated ash solids because of erosion and plugging of the high pressure nozzles. Capital costs for the MP&L installations are not available, but in 1975 dollars total cost for this type of installation is estimated to be between $40 and $50 per kw. Coal burned at both the Aurora and Cohasset Stations is Montana subbituminous from the Big Sky mine, containing 0.8 pet sulfur, 9 pet ash, and 9 to 13 pet CaO in the ash. These levels of calcium are chemically equivalent to 58 to 84 pet of the coal sulfur. Inlet dust loadings are 2 gr/sefd at Aurora and 3 gr/sefd at Cohasset. A typical inlet SO 2 level is 800 ppm. Data on the MP&L installations are given in Appendix IV-A and IV-B and in figure A-4. Flue gas from the air heaters passes through three concurrent sprays: a quench spray, the main high-pressure spray, then through the mist eliminator, and finally through a post humidification spray. This last spray washes the wet ID fan, which discharges flue gas to a wet stack. 81 The flow circuit for scrubbing liquor for the Cohasset scrubber is shown in figure A-4. Liquid is pumped from a seal tank at the bottom of the spray chamber to two clarifiers. Overflow from the clarifiers is combined with makeup water and pumped back as spray. The spray washing the ID fan is makeup water only. Blowdown from the clarifiers is sent to an 80 acre ash pond. Operation is not closed loop. At Aurora, there are no clarifiers, and clear scrubbing liquor is instead returned from the ash pond. In other respects the circuit is similar to that at Cohasset, shown in figure A- 4. Key operating variables at both stations are an L/G of 8.3 gal/1,000 acf and a total gas stream pressure drop of 4 inches of water. High pressure spray enters at 200 psi. The pH leaving the scrubber is typically 4.4. Operating costs are not available. Operating requirements are electrical power equal to 0.86 pet of generating capacity (which is low), water requirements of 4.3 acre ft/MW/yr at Cohasset and 30 acre ft/MW/yr at Aurora, and a high labor requirement for operation and maintenance. Particulate removal efficiency is about 98 pet at Aurora and 99 pet at Cohasset. Sulfur dioxide removal is typically 20 pet at both stations. Availability was not obtainable in terms of effective scrubber operation during times that boilers were operational. Little down time on boilers would be experienced using this type of scrubber, since many problems can be repaired without shutdown. Massive plugging or problems involving the wet ID fan would, however, necessitate shutdown. The major problems that have occurred are stack gas mist carryover and scaling in the scrubber and liquid circuit. Both these problems are more severe at Cohasset than at Aurora, owing to operation closer to rated load and restriction in amount of blowdown. A long term solution to the problem of disposing of sulfate-laden blowdown water is particularly difficult to find in this region if such water cannot be discharged to aquifiers, since there is no net evaporation from ponds in this climate. The option of operating strictly closed loop except for water evaporated in the scrubber and that leaving with ash sludge is probably not tenable unless a breakthrough occurs in methods for control of scaling. The remedy of circulating ash solids is probably not applicable because of erosion and plugging of high pressure nozzles. 82 FULL-SCALE S0 2 SCRUBBERS Six scrubber modules have been operated in the Western U.S. for the purpose of removing S0 2 from powerplant stack gases, with particulate removal as a secondary goal. Reagents used include lime, limestone, and soda ash. Designs represented include venturi scrubbers with either a wash tray or a packed tower in series, a horizontal cross flow-spray scrubber, and a TCA (mobile bed) scrubber. All units are in some sense experimental. Southern California Edison, Mohave Generating Station (13, lM Southern California Edison operates two 790 MW pulverized coal fired boilers at the Mohave Generating Station burning Arizona bituminous coal transported 275 miles from the Black Mesa Mine by pipeline. The coal is unusually low in sulfur content, averaging 0.38 pet, which results in an average SO^ content of 200 ppm in the stack gas. This uncontrolled average represents less than Uo pet of the allowable emission under the Federal standard. Ash content is 9 pet, with 15 pet CaO in the ash. The CaO content is chemically equivalent to 200 pet of the coal sulfur content. However, the ash alkali is not utilized in the scrubber because particulate emissions are currently controlled by electrostatic precipitators to an exit dust loading of nominally 0.07 gr/sefd. It is ironical that a station which is discharging stack gas that satisfies the Federal emission standards by a wide margin has become the focal point of the most extensive flue gas desulfurization program in the West. Spurred by the Clark County, Nevada standard, a program has been implemented in cooperation with other private and governmental agencies to develop an optimum scrubber for the Mohave Station, and presumably for conditions pertaining to Western coals generally. Work has been reported on operation of eight small pilot scrubbers (l MW) operating on four different reagents —lime, limestone, soda ash, and ammonia (13_). From the results of this work, two designs were selected for scaleup to full commercial size of nominally l60 or 170 MW, which would permit each of the Mohave boilers to be serviced by h or 5 modules. One module of each design has been built and tested to determine the final design for a plant-wide system. One unit, a four stage TCA system designed to operate on limestone, was badly damaged by fire shortly after startup in January 197^+. The system has been repaired, and operation recommenced in October 197^. No results have been reported. This paper does not discuss the design of this unit. 83 A second unit, called the Horizontal Cross Flow Scrubber and designed to operate on lime, was also commissioned in January 197*+» and operated at Mohave until recently when work was started to dismantle the unit and ship it to the Four Corners Station of Arizona Public Service for further testing. The Horizontal Module design is proprietary to Southern California Edison, and is based on a pilot design executed by Stearns-Rogers Inc. The 170 MW unit (figure A-5 in Appendix V) consists of approximately 50 feet of unobstructed horizontal duct work (cross sectioned 28 ft wide by 15 ft high) separated into four sections, each section or stage having its own spray header and drain. Scrubbing liquid is pumped consecutively from one stage to the next in the direction countercurrent to gas flow, and then falls into a reaction or recycle tank to begin its journey again. Gas flow passes consecutively through the scrubber and the mist eliminators, and is then reheated by mixing with heated air. The somewhat involved flow network is described in simplified form in Appendix V, and in greater detail in reference lb. Key operating variables are an L/G of nominally 20 gal/l,000 acf per stage and a total pressure drop of 6 inch of HgO. The transfer of scrubbing liquor from one stage to another results in pumping requirements equal to an L/G of 80; but in terms of solution chemistry, it is more useful to think in terms of an L/G of 20 since there is no appreciable hold time between stages. The pH of the slurry in the scrubber is not available, but it can be assumed to be high. A pH between 6 and 7 would be expected operating on lime. Operating and capital costs are not available. Operating requirements are electrical power equal to 1.6 pet of generating capacity, reheat steam equivalent to 1.2 pet of generating capacity, a water requirement of 1.3 acre ft/MW/yr, and two operators and one foreman per shift. The calculated lime requirement for reducing SO 2 from 200 ppm entering to Uo ppm is 8 tons/day. Removal efficiencies for SO 2 are reported between 70 and 97 pet, depending on the level of SO 2 entering, gas flowrate, L/G, and number of stages operating. A 90 pet removal of 200 ppm entering is reported at an L/G of 17*5 using all k stages. Removal drops to 70 pet using 2 stages. Particulate removal is reported to be 98 pet at 1.0 gr/sef entering and JO pet at 0.01 gr/sef entering. 8U The Mohave operation is characterized by Southern California Edison as "closed loop," meaning that all possible water is returned to the scrubbing circuit. This is the only operation that is so characterized by an operating utility among all scrubbers operating in the West. The makeup water requirement is imposed 93 pet by water evaporated in the scrubber and 7 pet by sludge loss and pond evaporation. The water requirement of 1.29 acre ft/MW year is the lowest of all Western scrubbers. Evaporation per MW in the scrubber is comparable to the amount evaporated in the Krebs scrubber at the Cohasset Station of Minnesota Power and Light. The amount lost in sludge and pond evaporation is less than the sludge loss calculated for a typical Western coal containing 8 pet ash, the reason being the very low inlet dust loading at Mohave. Thus the total water loss from the system is conservatively low for this size unit operating on typical Western coals in any climatic region, and the "closed loop" status should not be challenged on the basis of Mohave's arid climate. What can be challenged, or questioned rather, is whether the closed loop operation could have been achieved without severe scaling at a higher coal sulfur level, with its attending increase in ppm S02 and sulfate loading. The normal 200 ppm SO 2 at Mohave is very low even by the standards of Western coals; and higher levels such as 800 ppm obtained burning some Montana coals might make this system far more prone to scaling at the makeup rates used in this test. This question may already have been answered by the Mohave tests, but has not been published. If it is unanswered, testing at Four Corners should seek an answer. Scaling was not a serious problem in the Mohave tests, judging by the record of the problems that were encountered, which included "removal of two hard hats from the thickener (lU)." Sulfate scaling in the lime slacker was eliminated by switching to station service water for slacking. Various mechanical problems did have to be solved. The Mohave Horizontal scrubber appears overall to have high marks. Availability was reported to be 85 pet of the time the boiler was operating. Performance in terms of removal was up to the required specifications for the Mohave application. Perhaps most importantly, the concept of a long empty box with multiple sprays is an uncomplicated approach to obtaining freedom from internal plugging together with operating flexibility, redundancy, and an extended gas-liquid contact and residence time (a large number of transfer units). It is nox reported whether spray headers can be removed without shutdown but this should be possible since it is done on the Krebs-Elbair scrubber. 85 Arizona Public Service Company, Cholla Station ( 15 ) The Cholla Station at Joseph City, Arizona has one 115 MW wet-bottom boiler retrofitted with a limestone scrubber system designed by Research Cottrell. Two parallel scrubbers are installed in series with a mechanical dust collector. One scrubber train consists of a venturi followed by a packed tower and is designed for removing both particulate and SO 2 . The parallel train is a venturi only, with an empty non-functioning tower, and is designed for particulate removal only. Capital cost was $57/kw. Startup was October 1973. Coal burned is New Mexico bituminous from the McKinley mine. Sulfur content is O.H to 0.5 pet; average ash content is 9.6 pet. Inlet SO 2 level is U00 to 500 ppm, and dust loading is 1.2 gr/sefd. Data on the Cholla scrubber design are given in Appendix VI and figure A-6. Flue gas from the mechanical dust collector passes through a booster fan and then either to the scrubber or via a bypass to the stack. In the SO 2 scrubber train, the flue gas passes consecutively through the venturi, a cyclone separator, a "conical slurry separator" for keeping the venturi and tower slurry streams separated, through the tower packing, the mist eliminators, a steam reheater, and on to the stack. Separate liquid recycle circuits are maintained for the tower and the venturi. Fresh limestone slurry is admitted into the tower circuit, where most of the SO 2 absorption occurs. Spent slurry containing a high percentage of calcium sulfate solids is bled to the venturi recycle tank where it is diluted with some makeup water. System blowdown in pumped from the venturi recycle tank to a hold tank, and the sludge is periodically pumped out to the ash pond. No liquid is returned from the ash pond but evaporation prevents accumulation. In response to inquiries, both Arizona Public Service and Research Cottrell describe operations at Cholla as "open loop." The amount of makeup water used per MW is less than at any other installation except Mohave (table 6). If water usage were ratioed against the level of SO 2 entering or that absorbed, the ratio for Cholla would be lower than that for Mohave. Liquid to gas ratio, L/G, is 15 gal/1,000 acf to the venturi and U 5 to the tower. Total pressure drop is 20 inches of water. The pH level is not controlled; pH is about 6.5 into the tower and 5.2 into the venturi. 86 Operating cost is estimated to be 0.6 mills/kw hr. When fixed costs are added, total cost is about 3 mills/kw hr. Operating requirements include electric power equal to 2.h pet of generating capacity, steam for reheat equivalent to approximately 1.6 pet of generating capacity, makeup water of 1.8 acre ft/MW/yr, one operator per shift plus 30 hours of direct maintenance per day, and 15 tons of limestone per day at $20 per ton. Sulfur dioxide removal efficiency is 90 pet in the scrubber train with the packed tower; 20 pet in the unit with the venturi only. Overall removal is about 60 pet. Particulate removal in the scrubber is 99 pet, after 80 pet prior removal in the mechanical ash collector. Exit dust loading is 0.026 gr/sefd. Availability for both scrubber trains considered together is 91-5 pet, as a.fraction of boiler operating time. Arizona Public Service indicated that this was achieved only by a "very high level of effort." Availability of the scrubber train having the packed tower was higher (95 pet) that for the unit with the venturi only (86 pet) because of corrosion and fouling in the stainless steel reheater under the more acid conditions that prevailed when more SO 2 remained in the flue gas. Erosion has occurred at the throat of the stainless steel venturi. No serious scaling or plugging has occurred. The liquid in the recycle to the packed tower is maintained below saturation with respect to calcium sulfate by transfer of slurry to the venturi circuit and addition of makeup water at the mist eliminators. The solution in the venturi recycle is saturated with calcium sulfate, but scaling is controlled by recirculation of 10 pet ash solids and by use of an adequate level of blowdown to avoid a critical level of supersaturation. Degree of oxidation of sulfite to sulfate is low in the tower and high in the venturi. Calcium sulfite is precipitated in the tower circuit, but the soft deposits formed do not plug the system. The Cholla operation is quite successful from the standpoint of removal efficiencies and availability for the conditions applying at this location. The process has not, however, been validated as "closed loop" in the general sense in which the term is applied. 87 Nevada Power Company, Reid Gardner Station The Reid Gardner Station has two 125 MW pc-fired boilers that are retrofitted with scrubbers operating on soda ash (Na 2 CC> 3 ). The scrubbers were designed by Combustion Equipment Associates and consist of a venturi followed by a tower with a flooded tray. Scrubber installation was in series with mechanical ash collectors, one scrubber module on each of the two retrofitted boilers. Capital cost including ponds was $11 million, or $44/kw. Startup was in March and April 1974. The Utah bituminous coal burned contains 0.6 pet sulfur, 9 pet ash, and 8 to 18 pet CaO in the ash chemically equivalent to 69 to 154 pet of the 0.6 pet sulfur content. Inlet SO 2 level is 400 ppm and dust loading is 0.3 to 0.6 gr/sefd. Data on the Reid Gardner scrubbers are given in Appendix VII and figure A-7- Flue gas from the mechanical collector can be either bypassed to the stack or sent to a booster fan for forced draft entry into the scrubber. The gas passes consecutively through the venturi, a cyclone separator, a concurrent upward spray, a flooded tray, a mist eliminator, and after mixing with heated air for reheat, to the stack. Here, as at Cholla, separate liquid recycle circuits are maintained for the tower and the venturi. The tower circuit contains no reagent, only recycled makeup water. Spray directed at the bottom of the tower tray drains as bleed into the venturi recycle tank. Soda ash reagent is added in the return flow to the venturi at a point just beyond the blowdown from this circuit. The blowdown is neutralized to pH 7 and pumped to settling ponds, from which ash is dredged periodically for landfill. Clear liquor is sent to a 47 acre evaporation pond, which is monitored to detect loss of salt into the surrounding ground water. No leakage has been detected in six months of operation. As is evident, the system is open loop. The sodium carbonate reagent is converted in the overall system to a 6 pet solution of sodium sulfite and sodium sulfate, which is disposed of in the evaporation pond. The scrubbing solution is far below saturation with respect to calcium sulfate, although saturation could occur because of calcium derived from ash if blowdown were restricted. Reported water requirements are not greatly different than for the Cholla station on the basis of acre ft/MW/yr, which is somewhat surprising. The reason why more difference is not observed is that evaporation in the scrubbers tends to dominate water usage in both cases. 88 Liquid to gas ratio in the venturi is 9-5 gal/1,000 acf, and approximately 1 gal/l,000 acf in the tray tower. Total pressure drop is l8 inches of H 2 O. The pH is 6.8 entering the venturi, 5-8 to 6 .b leaving. pH in the tray recycle tank is 3 to 5 • Operating costs are not available. However, total cost including fixed charges is h to 6 mills, judging by an environmental surcharge that is being added to utility bills. Operating requirements include electric power equal to 2 pet of generating capacity, reheat steam equivalent to 2 .b pet of power generation (calculated), 2.2 acre feet of water per MW per year, 1 operator per shift and U maintenance and instrumentation personnel per day. Reagent required is 10,000 tons of soda ash per year at $75/ton, or 15,000 tons of Trona per year at $U0/ton. Trona is 66 pet Na 2 C 03 mixed with smaller percentages of NaCl, Na2S0^, and sand. The design goal of 8U pet SO 2 removal is easily met, and removals of 95 pet or higher are attainable by adding more soda ash. Particulate removal is 97 pet in the scrubber and 99-^ pet overall, meeting the 0.02 gr/sef standard. Typical availability as a percentage of boiler operating time is 90 pet. During March 1975, one unit achieved 99*^ pet availability. Problems include plugging of liquid lines, corrosion of piping where coatings have failed (largely "infant mortality"), difficulty in feeding Trona, and plugging due to sand found in Trona. No chemical scaling has occurred. Prevention of chloride stress corrosion requires expensive alloy construction. In summary, the Reid Gardner installation is an effective and relatively trouble free operation. Unfortunately, the success of the process depends strongly on the special circumstances existing in the locality of this plant, which are an arid climate, a ground hydrology which minimizes seepage, and a supply of sodium carbonate. The technology therefore is probably not widely applicable to general scrubbing problems in the West. PILOT PLANT STUDIES ON FGD BY SCRUBBING WITH ALKALINE ASH The deliberate utilization of alkali in Western coal fly ash in wet scrubbing for flue gas desulfurization has been studied independently by three groups since the early 1970's. This is in addition to work incidental to particulate scrubbing. 89 already reported. Work has been done at the Grand Forks Energy Research Center of the Energy Research and Development Administration; at Montana Power in Billings, with Bechtel Corp., Combustion Equipment Associates, A.D. Little, and York Engineering also involved; and at Minneapolis by Northern States Power Co. in cooperation with Combustion Engineering, and Black and Veach Consultants. Results from all three studies indicate that alkali in the fly ashes studied could provide 70 to 100 pet of the alkali required to meet Federal emission standard of 1.2 lbs SO 2 /MM Btu, without significant modifications of lime/limestone technology. Typical conditions of testing are given in table 7* The major advantage is a significant saving in reagent costs. In addition, it has been found that for this method ash recirculation controls calcium sulfate scaling very effectively. The throwaway product is highly oxidized from sulfite to sulfate and settles readily. A complication, as opposed to reagent based scrubbing, is the variability of coal ash quantity and quality available for the process. Grand Forks Energy Research Center The Grand Forks Energy Research Center of the Energy Research and Development Administration undertook research on ash alkali scrubbing in 1971- Since that time, two papers have been published on the subject (l6, 17_). A summary of the work presented in these is given here. The pilot plant facility (figure 2) was a 120 scfm flooded disk venturi scrubber, with cyclone mist eliminator, reaction mix tanks and settling tanks. The flue gas was produced by burning natural gas to which SOp was added. The fly ash to be evaluated and supplementary lime when used were added directly to mix-reaction tanks. The principal objectives of the program were to determine SO 2 removal and gypsum scaling rate as a function of SO 2 level, ash amount, alkali in fly ash, supplementary lime, suspended solids level, L/G, amount of makeup water and total dissolved solids. Dust removal efficiency was not investigated. Fly ashes showing high alkali (24 pet CaO) and low alkali (9*3 pet CaO) were investigated at addition rates equivalent to 1.3 to 4.0 gr/sef. The SO 2 level was varied from 335 to 1,150 ppm in several combinations with fly ash type and dust loading. As a general indication of scrubber performance, SO 2 removal efficiency for a high-alkali fly ash, 24 pet CaO, at a high dust loading, 4 gr/sef, is typically 62 pet of 840 ppm SO 2 . Removal for a low-alkali ash, 9-3 pet CaO, at the same dust loading is typically 30 pet of 840 ppm SO 2 . 90 TABLE 7- - Typical conditions in pilot plant tests on ash alkali scrubbing Grand Forks Energy Research Center Montana Power Company Northern States Power Company Coal MT Subbit MT Subbit MT Subbit S, pet ND Lignite .8 ash, pet - 8 8i/ CaO in ash, pet 9-3 - 2k 18 11 - 23 Flue gas flow 120 scfm 2,700 scfm 12,000 acfm. S0 2 , ppm 335-1150 800-1000 130° F 480-920 dust loading, gr/sefd 1.3-4 2-b 1.6-2.8 Removal goal S0 2 to 425 ppm 50 pet removal Particulate - .03 gr/sefd .04 gr/sefd Auxiliary reagent, pet limestone lime limestone stoichiometric 0-150 0-25 0-100 Venturi L/G, gal/1,000 scf 22-83 15 17 AP, inches H 2 0 10-20 17 7 Spray tower or bed L/G, gal/1,000 scf - 20 15 AP, inches H 2 0 - - 7 1/ Estimated from mine sampling data. 91 SO? sampling ports 92 Figure 2 . - Pilot plant scrubber, Grand Forks Energy Research Center. Increasing L/G increased SO 2 removal marginally. Increasing blowdown and makeup tended to decrease scrubbing efficiency. The level of suspended solids did not greatly affect SO^ removal. A large increase in dissolved solids during the approach to steady state operation increased scrubbing efficiency significantly. Supplementary limestone markedly affected SO 2 removal for a low alkali ash, going from 30 pet without limestone to 71 pet with 100 pet stoichiometric added. For a high alkali ash the effect was less marked, with an increase from 62 pet without limestone to 88 pet with 100 pet stoichiometric limestone, showing little increase after reaching 25 pet stoichiometric. The relationship between inlet SO 2 , total alkali added, and pet SO 2 removal is shown in figure 3. Rate of gypsum scale formation was measured by periodically weighing the accumulation on the inner surface of a short section of pipe located in the liquid drain from the demister to reaction mix tank. The principle variable affecting scale formation was the amount of suspended solids in the scrubbing solution. As suspended solids increased from 0.02 to 1.5 pet, the scaling rate decreased nearly tenfold, and thereupon remained very low up to the highest level tested, 7 pet. Increasing L/G from 22 gal/1,000 scf to 82 gal/1,000 scf caused a threefold decrease in scaling rate. Increased blowdown reduced scaling by diluting the scrubbing liquor below saturation. Scaling was not affected by the level of total dissolved solids. The system was entirely scalefree only during transcient periods when the scrubbing solution was below saturation with respect to calcium sulfate. Fly ashes derived from some Western coals may contain significant amounts of soluble magnesium and sodium in addition to CaO. As the solubilities of these species are high, high levels are obtained in solution. To investigate system behavior under these conditions, epsom salt and NagSO^ were added to the solution to give the following levels, which were believed representative of steady-state closed-loop operation on the North Dakota lignite fly ash tested: PPM Magnesium 30,000 ill, 000 Sodium Calcium Sulfate 900 192,000 93 00 1 o o o CD ^ CVJ O o ro O CVJ o in 6 cr w CD CD Z) DC O CO o I- < * CM O CD O I- < or o » c o a. CO o > o E a> a> •o K O ■o Z3 CD i ro cu k_ =j cn |U0OJ0d * “IVA0W3d 2 0S 9 U O GO As mentioned previously, SO2 removal was favorably affected byincreased dissolved solids. Scaling was reduced due to desupersaturation caused by loss of sulfate at this high concentration level through inefficient mist elimination. The general conclusion to be drawn from the Grand Forks tests is that alkali from fly ashes tested was sufficient under favorable conditions to meet the Federal emission standard. Under circumstances of low dust loading, supplementary limestone was shown to raise removals to acceptable levels without greatly aggravating scale. Effective control of scaling was achieved by circulation of ash solids. High dissolved solids in solution could lead to problems in waste disposal, or to sulfate loss in mist. Research to improve the utilization of ash alkali is continuing. In addition, the Grand Forks Energy Research Center is cooperating with the Square Butte Electric Cooperative in larger tests for scaleup for conditions of cyclone firing. Montana Power Company (l8) A pilot plant program investigating ash alkali scrubbing was undertaken in 1973 for Montana Power by Bechtel Corporation in cooperation with Combustion Equipment Associates who designed the scrubber. Arthur D. Little, Inc. also participated in the test program, and SOp and particulate testing services were provided by York Research Corp. Based on generally favorable results in the pilot plant program, Montana Power is planning to use ash alkali scrubbing in two new pc-fired generating units at Colstrip rated at 362 MW each, which are scheduled for startup in 1976 and 1977- The information presented on these tests was obtained primarily from Dr. Carlton Grimm of Montana Power Company, and from a report published by Combustion Equipment Associates (l8). The present description of the tests as pieced together from the various sources is solely the responsibility of the writers of this paper. The 3,000 acfm pilot plant (figure U) consisted of a venturi section for particulate removal, a spray tower for SO2 removal, followed by mist eliminators and a reheat section. Provisions were made for alkali makeup as Na2C03 or lime and the use of cooling tower blowdown as makeup water. 95 0 ) jQ k_ O -O < 96 Figure 4 . - Pilot plant scrubber, Montana Power Company. The principal objectives of the study were to: 1. Demonstrate CEA's guarantees on SOp and particulate removal. 2. Determine level of supplemental alkali required, if any, for SOp removal. 3. Optimize variables influencing system performance. 4. Investigate use of cooling tower blowdown as a demister wash spray. The goal for particulate removal was an outlet dust loading of 0.03 gr/scf. Results at an inlet dust loading of 2 gr/scf indicated that the 0.03 gr/scf exit loading could be met at a venturi AP of approximately 12 inches of H 2 O, and 0.02 gr/scf at about IT inches of H 2 O. A AP of IT inch H 2 O was selected for the 4 to 6 gr/scf anticipated in the full scale units. The SOp removal requirement was to comply with a standard of 1 lb SO 2 /MM Btu, or a level of 425 ppm. Compliance required removals in the range of 50 to 60 pet. This was accomplished with ash alone (without supplemental alkali) supplied at a dust loading of 2 gr/scf. At this dust loading, however, a pH of 4 in the scrubbing liquor entering the venturi was judged lower than desirable for scale control. Using a simulated grain loading of 4.0 gr/scf, a higher pH of 5 to 6 was achieved and the removal criterion was also met. An approximate plot of SO 2 removal percentage as a function of stoichiometric ratio alkali/SO^ (inlet) is shown in figure 3, for ash alone and with lime added. A suspended solids level of 12 pet was thought to be adequate for SO 2 removal and scale control, but higher levels would furnish more residence time for the ash to react and provide additional nuclei for precipitation, to control scale. Suspended solids level was found to have an important positive effect on SO 2 removal in the range of 3 to 12 pet. Increasing L/G in the spray tower also increased SO 2 removal. L/G and A.P in the venturi had little effect on SOg removal. Tests were run in which ^200^ was added for pH control. Sulfur dioxide removal was improved by 60 pet or more of the stoichiometric equivalent of the Na 2 C 03 added. There are no plans to use soda ash in the full scale units at Colstrip. 91 In other tests, lime was added to supplement the ash alkali. Sulfur dioxide removal was improved by 60 pet or more of the stoichiometric equivalent of lime added. (See figure 3 for approximate results.) Use of lime is planned in full scale units to control pH during periods when fly ash loadings are less than h gr/sef or when fly ash reactivity decreases. Scaling in the scrubber and liquid circuit was controlled adequately, mainly by the recirculation of ash. Some fouling did occur in the wet-dry zone, and loose scale deposit accumulated in the reheater. Major emphasis was given to testing the chevron mist eliminators to establish their effectiveness and the wash conditions required to prevent plugging. Three approaches indicated that there was little measurable liquid entrainment through the demister. First, direct measurement of the degree of saturation of the flue gas indicated no carryover. Second, outlet particulate loadings with variable suspended and dissolved solids in the liquid showed no increase in outlet loading. Third, outlet particulate loadings did not change upon increasing or reducing demister wash sprays. However, deposits did form on the reheater, indicating that some carryover was occurring. Simulated cooling tower blowdown was found to be unacceptable for washing the mist eliminators, because of scaling that occurred. To combat this scaling, a Koch bubbler tray was added ahead of the mist eliminator and reheater. Fresh makeup water used to wash the mist eliminator dropped into this tray, and reentrainment of this relatively clean water from the Koch tray lowered suspended and dissolved solids in the mist by dilution and "exchange," aiding significantly in keeping the mist eliminator clean. Plans call for using this feature in the full scale units. The level of dissolved solids in recycled scrub liquor was influenced by the amount and solubility of cations derived from the ash and by the amount and quality of makeup water added. Analysis at the end of a prolonged period of closed loop operation indicated that dissolved solids reached 26,000 ppm. Dissolved ion concentrations were as follows: Magnesium U,000 Chlorine 400 Sulfate 18,000 Sulfur trioxide 3,000 Calcium Uoo 98 The principal findings of the Montana tests were that acceptable SOp removal could be achieved using fly ash alone as the source of alkali, and acceptable particulate removal could be obtained. Supplementary alkali could be used to control pH if fly ash quantity or quality was reduced. The optimum L/G for spray tower operation was determined to be in the range of 15 to 20 gal/1,000 scf. Scale control was achieved principally through the recirculation of ash solids at 12 pet and above. Control of pH also aided in scale control. Other measures found helpful were fresh water washing of the mist eliminator and the use of the Koch bubbler tray. Cooling tower blowdown was found unacceptable as a source of makeup water for washing this mist eliminator. Northern States Power Company (19) A study on ash' alkali scrubbing was undertaken by Northern States Power at their Black Dog Plant near Minneapolis to establish the design of scrubbers for the new Sherburne County Generating Plant where two 680 MW units are scheduled for completion in 1976 and 1977. Other parties involved in the test program included Combustion Engineering, the scrubber vendor, and Black and Veatch, consulting engineers. The test facility was a marble bed scrubber with demister and reheat capability, having a capacity of 12,000 cfm. Attendant equipment included a reaction tank, thickener and ash pond, makeup and limestone tanks. Arrangement of equipment is shown in figure 5. The principal objectives of the test program were to: 1. Demonstrate capability for an SO 2 removal of 50 pet or greater operating on alkaline ash only or on ash and limestone. 2. Demonstrate the guaranteed particulate removal capability. 3. Demonstrate reliable operation with acceptable maintenance. k. Determine optimum operating conditions, including blowdown requirements, and instrumentation and control functions. 99 100 To obtain the guaranteed dust removal of 0.04 gr/scfd outlet dust loading or 99 pet removal, it was necessary to modify the marble bed scrubber by installing a venturi rod section at the inlet. By thereby increasing the total pressure drop from 6 inches H 2 O to a AP exceeding 13 inches, the particulate criterion was met. Optimum system L/G was about 32 gal/1,000 acf. A series of tests was made wherein inlet SOg and supplementary limestone were varied while the input fly ash remained relatively constant. Inlet SO 2 was varied from 482 ppm to 919 ppm and the limestone from 9 pet to 100 pet of stoichiometric. The SO 2 removal percentage varied from 48 to 88 pet under these conditions. The relationship between the stoichiometric ratio of input alkali to SO 2 and the SO 2 removal percentage is shown in figure 3. The SO 2 removal criterion was met with as little as 5 pet stoichiometric limestone added. The calcium in the fly ash played a major part in the removal of SO 2 , representing TO to 80 pet of the alkali reagent. Some scaling and plugging which occurred in the pilot program was remedied in the course of testing. Heavy mud deposits up to two inches thick appeared on the mist eliminator and scrubber wall surfaces, requiring shutdown every two days for washing. Deposits in both the mist eliminator and reheater were similar in chemical composition to the spray water solids, consisting primarily of fly ash with CaSO^, CaSO^, and CaCO^. Deposits on the ID fan were chemically similar but physically more amorphous. Hard calcium sulfate scaling occurred on overflow pots. A redesign of the mist eliminator washing equipment was made to prevent plugging at that point. In addition, three modes of sulfate scale control were utilized. Ash solids were recirculated to provide ash and gypsum seed crystals, with a 10 pet level of ash found to be optimum. Supplementary limestone was reduced to a level giving adequate SO 2 removal while minimizing scale, with a level of approximately 15 pet of stoichiometric or less found to be optimum. Oxidation of sulfite to sulfate was increased by bubbling air into the reaction tank to help prevent supersaturation with calcium sulfite in the scrubber proper. Oxidation level was increased from 36 pet to 98 pet and above with four times stoichiometric air. 101 After 40 days of operation, the following concentrations of dissolved species in the spray water were measured: Sulfite Sulfate Chloride Nitrate Calcium Magnesium Silica Sodium A demonstration test of with 99 pet availability of ' PPM 0 30,000 550 bOO U00-500 7,000 200 200 continuous days was completed two stage scrubber system. THE FUTURE OF SCRUBBING IN THE WEST Table 8 lists plans for U5 FGD scrubber installations to be built in the West between now and 1983. Many are in the early stages of planning. Among those listing the intended process, the largest number, 19, are committed to limestone scrubbing, followed by alkaline ash with lime backup, 8, straight lime, 5, and soda ash, 1. Most of the units will probably be designed as venturi-tower combinations judging by the vendors that are listed. With few exceptions, future plans for installing scrubbers are not for the primary purpose of particulate removal, as was true in the majority of past installations. A major reason is the necessity for having flue gas desulfurization for most low sulfur Western coals under the developing regulatory pattern in the West. Besides this, recent ESP installations on some Western coals have performed adequately and particulate control can be expected to swing in this direction. A series combination of an ESP and a scrubber, although costly, has the appeal of eliminating the visible fly ash plume with a relatively reliable device, the ESP. If the scrubber can then be bypassed for reasonable periods for maintenance, a high plant availability can be maintained. It is significant that all Western FGD installations, existing or planned, use throwaway type processes. This is consistent with the West's remoteness from most large markets for sulfur or gypsum. Sulfuric acid could possibly find sizable markets in the West as an ingredient in fertilizer manufacture. 102 P p VO vo p C\ d\ t— CO CP t— t— C— P t— t— t— t— c— "— C— G- CD ■—^ —^ o CM -P MO VO VO VO VO 1—1 P oo -3 CO 1/ • P p p p CO Cm c p G G G 1 • O E o G O O O i—1 D p -p O P P P P P -p cd P CD CD CD CD r a c P< cd p cd P P P p p p P P >> E E E E E E P p E E p EH CO •H •H •H •rH •H •H o CJ •H •H p rH rH rH i—1 rH P C p rH i—I CD P p G T3 x : P P p P P p P CJ c o ,g i—1 X-. i—1 G 1—1 ,c rH P p •H G o i—1 o i—1 CJ r—1 o rH *—1 o p o o c p P p P p p p P P CJ p CJ CJ c p P P p P p p p P CD •H CD •H •rH o > P P p P p p p P E E E •H CD -P CD P CD p CD P P P P P P p CO P o P o P o P O O 6 o JS x: o s CP o PS o PS cj PS CJ Gs C o o G w Ei E- 1 -p CO cd >H G CO p o G i—t OO OO o El Pi p • • < iH o O o -3 -3 P > G o 1 1 -3 ~3 CO OO CO oo P o CO tfV ir\ • • • • • • "G s c • • o o o o o o c w M o o 3 PS p CM o o CO p p p p •a •rH •pH •rH •rH p P p, Pi Cm Cm g p o o o O M rH p p Cl p P § •J •H CO o > ir\ > o > o £ CM g LTV p LTV p CP p e P •H M o •H o p o p ir\ p CO p t— p t- P t— p CM p o P P g FQ CO CM G CM c CM c CM c OO G 1 — 1 p p p CM p oo P cd C o •H 4-> P i—i CM oo CM OO p- CD ID CD CD r—l P P P p p c • . • . • P P P p -p o o o O o o G c c c cd •H GS s GS GS GS P p p p c p O o o o •H p P p P p p o o CJ CJ -p •G x : i—1 1—I Ip 1—1 CM OO -3 P CO O CJ i — 1 1—1 1—1 p p p p p P P o o o G • G • G • G . p P, P x : 8 x : o o O o o o O O •2 < < o o CP GD E. s (p GS E. G’C p o CO i P p CJ CJ o CJ O CJ O P p •H •H •rH •H •H •H •rH • > > i — 1 i — 1 i — 1 P p P i — 1 oo O o x> X> P P x> P x> CP CP G G G G G G G w p CP CP CP CP CP CP CP IG •H CJ CJ PQ rH P *H p •H P p P p P P P p P p P P p P < •H G p c p G V G CJ G CJ C CJ c CJ c O c CJ H P o -p o p O •H O •H O •H O *H o •H o P o •H to CJ CO CJ CO > CO > to e to > CO > to > co > •rH p •H p •H p •H p •H p •H p •H P •rH P •H P P iH P 1 — 1 P p P p p p P P P P p P p P < w < w < CO < CO < CO < CO 0) a; QJ QJ QJ QJ QJ CP Cm p P p p P P P P P • o E o o o o O O o G O p -p P 4~> p P P P P P P P to to to to to to to to to p P- to p P p P p P p p p O CJ t) o CJ o CJ o •H o •rH p o P rO •rH rO •H •H -P P P P P P P E bi E bL p> C Q o o o o o O O o p O P o p p p p p P P o w o W p pi p t4 4-5 00 to c xn P c P rH o pi Cm cd < rH O L/V LTV LO LTN p > P O -M 0) c 00 • • • • X 2 C o o o O c w i —\ p (X p C\) o o 00 P P X) g •rH •rH p p Cm Cm p Eh p #■ O O g PH 1 — 1 p P P 3 PJ •rH to o > O U e > o £ c > Ovl P C P o > o > rH PH c •rH urv p to p trv P LTV P LTV QJ o QJ LTN QJ o QJ o 0.' p g CP CO LEV p y\ c LTN t— -3 P P CNJ P CM P -H P E~ p to G C •rH -p P P c P •H •rH •rH 1 — 1 to to to r-H CM 1 — 1 P P P i —t Cvj p c CP PP CP • • P o . • o o to •rH •r4 •H •rH o o s 53 c -p P P P s s h >. •rH P P P P p P P P P o i —t O CV! o m bi bL QJ QJ QJ r — 1 QJ CM p CO to to to •H •H x X V. Cm P to • to to . cd cd J>. Cm • s ST K so o o cp K ►T) s PC P P O to 1 o cj o •rH •H •rH • p p p W K K W P P oo P ■p 4-5 fc) E: Eh QJ P >. O o O b b > > w p> P 0) P O p o p PI •H i —i rH s — 1 o CJ o o o O o o (X o: (X x: CO rH W w w X •H X •H X •H X »H b: bL < •rH cd p a! P P P P P to •rH to •H Eh -p P p p P p P P p P P P cd P PI •rH -H •r4 O CJ O CJ o CJ o CJ to to (A tn to i—I a) rH 0) rH QJ rH QJ P p X cd cd cd o rH o rH o rH O rH P g p g PD PQ o w C_) W O CO o w tP cd 3 10U Montana Dakota Lewis and 50 0.50 Research ash alkali Utilities Co. Clark retrofit Cottrell and lime- 8/75 stone p p t ir\ VO CO o\ o 1—1 X CO uo c l— E- E- oo OO OO oo X o P -V •— Ov '■'V. o P l/\ LTV e~ i—1 vo vo vo vo VO CO CO • •H •H rH i—t OJ cd OJ cd 01 1 1 i 1 1 1 p R O E E X E 1—1 i—1 i—i 1—1 r—1 1—1 p P 0) rH •rH X •H oj 0) 0) a; 0J 0J E p (D oj p cd r—1 cd rH CO X CO X CO X CO X CO X CO X 3 o -p X co 0) oj 01 0) OJ 0) •H rp co 2 x : X X r O p p p p p p p p p p p p X p (D Eh CO (0 p co P o o o o o V o o o o o o o p S cd cd p 3 p a; p 0) p 0) p oj p 0) P OJ CO o OJ x: p X X X X X X p 0) 0) 0) OJ OJ 0J •H C P p p p p p p p p o P o p o o o CJ o o o p c o •H P •H p oj OJ X •H X •H X •H P E p E p p p p p p p E p P CO O cr O cr o o o o o o o cr p 2 o w o w p p p p p p o w co w p E-< o CO o 2 CO P p P rH CO CO CO CO o 0) X P • • • • . X < 1—1 o rH r—1 rH rH X c > P o OO oo C— t— | i i 1 1 3 o CO • • • • e- f— t— C— I/O 2 c o o o o . • • • • p w M o o o o o o X x OJ oj o p CO p p •H 3 2< P (—i Eh OJ o p X rH oj p X •H N o > o > o > o > o > o > o > o > uo p co X o •rH VD 0J VO oj o oj o 0) CO oj uo 0J C/0 0J UO OJ X OJ c o g PQ CO CO p oo p E- p E- p uo p uo p uo p UO c *— 1 p •H p cd r — 1 rH X C\J oo X P P • • • • P P p p p CO o O o o O O o o 0J p c S3 S3 S3 S3 2. P 2 2 P 2 p •H o P p X •H P< P x X 3 3 3 3 p P p •H •H •H •H o o CJ o p oj cd p p p p o X p p p p p > 1 — 1 > X > CO > CO X CO co to CO CO o O o o X p 1 — 1 1 — 1 1 — 1 1 — 1 p • p • p • p • •H • p o O o o p o P O p o P o 0J o o CO 1 o o O o < S3 < S3 < s < a X S3 CO p p p p w a) oj 0) 0) p p p p p X > £ > > 0) OJ 0J OJ 0) CQ o o o o > > > > > < p (X (X X X o o o o o EH •H X X X X X rH cd p cd cd •H c p c p cd p p p p p cd 3 3 3 X X X X X X p p p p cd cd p p p C p p p > > > > > o o o o 0) 0) OJ OJ 0J 2 2 2 2 s S3 s S3 S3 105 Northern States Sherburne 680 1.0 Combustion ash alkali 5/76 Power No. 1 new Engineering and lime¬ stone -p p o o CO ZD £ d -P d C! d ,P -P P •H p o •H -p o e p> d p o o p (D d P 3 P o d (U P § rH p •H i—! P -P d P •H p a; 43 rQ 2 P o co i ao W PI § Eh P< P b— VO ■P t— t— b- OO o MO b- b- P b— t— t— CD b- \ b- cri -v o "V •v o ■P on rH 1—1 id ID, CO 1—i oo CO •H (—i 1 d d Q) d V-, rc d P l 1 p P p o E X E O i—i r-H o a o ID <—I •rH P d d -p -p -p dl -P P rH d V d d d d d d d d P d P d d d d d d d d d d h 43 T2 o E E -p -p +-> p> E E E E E E E- CO CO p p •H •H o d O d •H •H •H •H •H •H p s d i—! rH p d P d i—1 rH (—1 r-C i—1 rH d d d d d d & d 1 1 d d d d d p p P P P -p -p -p •p -p p o •H O d d o d d d d o •r*f p d £ > d d d d d d -P d i —1 o o i—i rH r-C i —i rH P d d d O- P- d d d d d P •H >. E bf p p d > w •p +3 -P P> ■p CO o P o p P p p O o o o o d O W P o bC o bi P p p p p w Eh CO CO P P i—1 CD OO OC CO P] Ph P . • . . ■3 f—1 o O o o o > p o o 1 CD CD CD CD 1 1 1 o CO . LTV • • • . LD ID LD p 1—1 • o o o o • • • w PH 1—1 o o o o K CM o CO g p -P -p •H •H •H >H P p-. o p irv £ o -P o > o -P o > o > o > 1—1 o •H OO d o d b- d d o a> o d LD d ID d LD d S CQ CO VO P 1—1 p oo P CO P LTV p LD P t~- p t— p C- P CM rH CM OO hL 6 • • • • rH CM oo PC o o o o p P3 Jp fp • • • o O o o o •H P p p p p s JP s -p P p p p p p p P p p p p p o o o -p rO CM o Hh *"5 d * r T •d •d CO P £ a3 p p O • rH c P p p > > > 43 O P 3 P p p p p p CO P3 > CO CO CO CO s s s d o d d d d d d d p o P o o o O d o d o p •H P •H o •H o •H d •h d p > o > •rH > •rH > •H r> *H i>; CO p rH p X p X p X P X p p p -P d o d d d d d d d d d d d •H p CO o CO CO s CO s co s: > > > i — 1 p •H -p •H -p •rH -p •H d o o > d ? d > PC d K d cc d -p 43 P •H o •H d •rH d •H d •h d d d d n> p d rH rH S •H rH J3 rH S -p •f-H -P •r~2 ■p •d p > O • fp 43 43 rH o rH o rH o o o P o p p

1-1 P • •H G cd • rH CD CD ID CD CD W G P CO

5 E 43 P E E E E E P CD G P c CD CO E P CD P CD P CD P CD CD •H i—i •rH g M rH o > P •H CD P 3 o 1 CO -d -d -d _d i t3 CD Q CO UA • • • • • CO CD P S c • o O O O O • 1 — 1 O p c w P5. M o o •rH Pi c « (D< CD o E P s p CM o O O o 03 p P CO MW P •H CD § § cd >H P CO c EH cd * O O r. CD c M 1—I CO P •H N o P o > on > on > on > on > o > P E- •rH i—i M o •H -d CD o CD On CD On CD On CD ON CD o ID ON P P Eh pq CO VO P -d G c G t— G t— G r— G co G CD r—t cd b P P CO cd P CD G CD p. O CD P o •H be p O O P c H CD CD CD ID XI •H cd p G O • o • ■s i IS ■3 o I—1 ID t O CM •H i — 1 •H CM •H on •H -d •H on CD W CO P CO cd P p p p p P P •rH 33 CD P-, G < •H CD CD P 43 E-c P 43 •H P P CO CO CO w CO Eh i ZD P c cd O cd cd cd cd cd G p G CD \ 1 CO Cm CO W Eh Eh Eh Eh Eh rH 107 Future trends in design of scrubbers cannot be predicted with accuracy, but some of the possibilities can be reviewed. The three most significant factors to be considered in designing for Western coals are: l) the low concentration of SOp to be removed, 2) the alkalinity of Western coal fly ashes, and 3) the tendency to operate at a high state of oxidation, producing sulfate and not sulfite. A direct means of achieving savings in Western installations is to treat only a portion of the flue gas. Optimization will require balancing the savings of treating a smaller volume of gas against the cost of removing a higher percentage of the SO2 from the fraction treated. If standards are made too stringent, the option of treating a partial flow would, of course, cease to exist. At low concentration of SO2, it can be argued that gas film diffusion should be the rate controlling step, since the equilibrium partial pressure of SO2 for alkaline slurry is low and the capacity of the slurry to absorb SO2 during passage through the scrubber is not taxed if the amount absorbed is small. A sufficient L/G of course plays a part in validating this argument. If gas diffusion does control, design should maximize gas-liquid contact and residence time. Long residence time necessitates a large volume; and good contact requires either multiple sprays or tower packing. If scaling can be resolved chemically, packing is probably the economical choice. If scale has a tendency to form, the large empty volume with multiple sprays will be the better option. It is the opinion of the authors that if blowdown is sufficiently restricted, scale will indeed tend to form. The problem of scaling is inexorably tied in with the question of what constitutes "closed loop" operation. The practical answer to the latter is that a system is "closed" if no liquid blowdown is deliberately removed and disposed of. Inadvertent loss in sludge cannot be eliminated. Beyond this, pond evaporation of a saturated scrubbing liquor does remove sulfate from the system, even though liquor may be returned from the pond. Since the sludge and pond losses will vary with design and climate, every system will be "closed" to a different extent, and arguments based on absolutes lose their meaning. What then is the "closed loop" question? It is simply the question of whether or not one has truly solved the disposal problem for the locality involved. 108 If we assume that discharge of sulfate laden waters to aquifers will be prohibited in the long run and that seepage into ground water will become an increased concern, then we can also assume that lime/limestone/alkaline ash scrubbers for Western coals will be operated saturated with calcium sulfate. This can be debated from more viewpoints than can possibly be discussed here. However, a high state of oxidation and the slight control that can be afforded by manipulating pH in a sulfate system offer little hope that unsaturated operation is possible for Western coals during closed loop operation. The oddity that this can be accomplished for high sulfur Eastern coals depends on a low state of oxidation, with consequent precipitation of calcium sulfite and coprecipitation of calcium sulfate from a solution that is not saturated with CaSO^ { 6 ). These conditions do not appear to exist for Western coal operations. Control of scale formation, in the opinion of the writers, will depend most directly on ability to circulate a sufficiently high level of suspended solids and to operate at a constant pH, whether high or low. Depending on the cost of reagent and the properties of the waste products produced, there may be more or less motivation to improve the utilization of alkalinity in Western fly ashes in scrubbing systems. Laboratory tests at the Grand Forks Energy Research Center have shown that utilization should improve very significantly as the pH is dropped from 5 to 3. Other operating variables would have to be changed along with the pH, including the flow circuit, gas-liquid contact, L/G, and slurry reaction times. Another form of optimization is believed necessary for fly ashes having relatively high concentrations of sodium and magnesium along with calcium. In conclusion, significant developmental work remains ahead on scrubbers for the West. Involvement of alkaline ash in a scrubbing system implies a new variable which must be controlled, and the analyses of Western ashes vary sufficiently to rule out one tailor-made solution. As esh varies, the characteristics of sludge will vary, and therefore multiple studies on properties of sludge including leaching of major and trace elements will need to be performed. As emission regulations are tightened, a closer look should be taken at the fine mist that is produced in a scrubber, to see if and how it survives passage through a mist eliminator and a reheater. This last topic is related to the question of submicron particulate, where the first priority is to find a satisfactory way to 109 measure it with sufficient agreement to proceed to the consideration of control. Experiences on operating scrubbers indicate that work is needed on materials of construction, and on component design to improve reliability. This list could, of course, be continued. The solution of stack cleaning problems will require continued development effort so long as there is a conscious desire for improvement; and the present state of scrubbing art contrasted with the desire for non-degrading power generation assures a long uphill course. REFERENCES 1. U.S. Bureau of Mines, Division of Fossil Fuels. Coal— Bituminous and Lignite in 1973. Mineral Industry Surveys, January 4, 1975, p 5* 2. Coal Age. Estimated 1974 Production by Region, v 80, February 1975, P 113. 3. Nielsen, G. F. Coal Mine Development Survey. Coal Age. v. 80, February 1975, PP 130-139* 4. Energy Research and Development Administration. Open file report. Survey of Coal and Ash Composition and Characteristics of Western Coals and Lignite. Grand Forks, ND, 1975* 5. Gronhovd, G.H., P.H. Tufte, and S.J. Selle. Some Studies on Stack Emissions from Lignite Fired Power Plants. BuMines IC 8650, 1974, pp 103, 133. 6. Borgwardt, R.H. EPA/RTP Pilot Studies Related to Unsaturated Operation of Lime and Limestone Scrubbers. 7. Facts Sheet. Four Corners Powerplant, Farmington, NM, April 1973. 8. Quig, R. H. Chemico Experience for SO2 Emission Control on Coal-Fired Boilers, presented at the Coal and the Environment Technical Conference, Louisville, KY. Oct. 23, 1974. 9. Ashton, T.M. Operating Experience Report, Flue Gas Scrubbing System, Dave Johnston Steam-Electric Plant Unit 4, Pacific Power and Light Company, presented at the American Society of Mechanical Engineers National Symposium, Philadelphia, PA. April 1973. 10. The Mcllvaine Company. Mcllvaine Scrubber Manual. Northbrook, IL, 1974, Chapter IX, Section 4911-900, p 175*0. 110 11. Green, G.P. Operating experience with Particulate Control Devices. Presented at the American Society of Mechanical Engineers National Symposium, Philadelphia, PA, April 1973. 12. Kilpatrick, E.R., and H.E. Bacon. Experience with a Flue Gas Scrubber on Boilers Burning Subbituminous Coal. American Society of Mechanical Engineers Winter Annual Meeting, New York, NY, November 197*+. Paper No. 7*+-WA/APC-3. 13- Weir, A., and L.T. Papay. Scrubbing Experiments at the Mohave Generating Station. Proceedings: Flue Gas Desulfurization Symposium - 1973, May 1973, New Orleans, LA, pp 357-392. 1*+. Weir, A., J.M. Johnson, D.G. Jones, and S.T. Carlisle. The Horizontal Cross Flow Scrubber. Presented at the Flue Gas Desulfurization Symposium - 197*+, November 197*+, Atlanta, GA. 15. C&E News. Stack Gas Scrubber Makes the Grade. January 27, 1975, PP 22-2*+. 16. Tufte, P.H., E.A. Sondreal, K.W. Korpi, and G.H. Gronhovd. Pilot Plant Scrubber Tests to Remove SO 2 Using Soluble Alkali in Western Coal Ash. BuMines IC 8650, 197*+, pp 103-133. 17. Sondreal, E.A., P.H. Tufte, and S.J. Selle. Wet Scrubbing of SO 2 with Alkali in Western Coal Ash. Paper No. 7*+-272, 67 th Annual Meeting of the Air Pollution Control Association, June 9-13, 197*+, 31 pp. 18. LaMantia, C.R., and I.A. Raben. Some Alternatives for SOp Control. Presented at Coal and the Environment, Technical Conference sponsored by the National Coal Association, October 22-2*+, 197*+. 19. Noer, J.A., D.O. Swenson, and K.W. Malki. Results of a Prototype Scrubber Program for the Sherburne County Generating Plant. Presented at the IEEE-ASME Joint Power Generation Conference, Miami Beach, Florida, September 15-19, 197*+. 20. The Mcllvaine Company. Mcllvaine Scrubber Manual. Northbrook, IL, 197*+, Chapter IX, Section *+911-1000, pp 176.1-176.91- Ill Appendix I Scrubber Design and Operation (7_, 8_) Four Corners Plant Arizona Public Service Company LOCATION 1. Farmington, New Mexico. 2. Elevation is 5300 feet. 3. Atmospheric pressure is 12.1 psi. U. Annual precipitation is 8 inches. 5. Water supply for the plant comes from Morgan Lake, a man-made reservoir filled from the San Juan River. SCRUBBER APPLICATION 1. Particulate removal, retrofit. 2. Boilers equipned with scrubbers. - Two 1T5 MW Riley pc-fired boilers (Units 1 and 2). - One 225 MW Foster Wheeler pc-fired boiler (Unit 3). 3. Service date: Units 1 and 2, December 1971; Unit 3, January 1972. 4. Fuel is New Mexico subbituminous coal from the Navajo mine. - 8900 Btu/lb. - 12 pqt moisture. - 0.68 pet sulfur. - 22 pet ash. - 4 pet CaO in ash. 5. Flue gas entering the scrubbers. - Boilers 1 and 2. - 8l4,000 acfm. - 3U0° F. - 650 pnm SO 2 . - 12 gr/sef particulate. - Boiler 3. - 1 , 030,000 acfm. - Conditions are the same as on units 1 and 2. 7. The particulate removal goal was set by the project at 99.2 pet. SCRUBBER DESCRIPTION 1. Two venturi scrubbers on each of boilers 1,2, and 3. 2. Vendor, the Chemico Air Pollution Control Company. 3. Capital cost is $30 million, or $52/kw. 4. Operating costs are not available. 5. Materials of construction. - Scrubbers are carbon steel with stainless steel or plastic lining. - Outlet ducts were stainless steel lined, later lined with plastic over the stainless steel. - Liquid lines and pumps are rubber lined. - Process vessels are plastic lined. - Reheaters were 3l6 L stainless steel. - Wet fans are inconel. 112 To stack Figure A-l - Simplified flow diagram for the Four Corners fly ash scrubbers . 113 6 . No bypass. 7. Turndown is to approximately 50 pet of rated scrubber capacity. 8 . Chevron mist eliminators have 6 stages. 9. Wet fan. 10. Reheaters that heated flue gas directly with steam coils failed because of corrosion. The reheat units were removed about one year ago, and no reheat has been used since. Indirect reheat by mixing with heated air is being considered. SCRUBBER OPERATING DATA 1. L/G is 8.5 gal/1000 acf, or l8 gal/1000 scf. 2 . AP is 20 to 22 inches P^O across the venturi, 28 inches overall. 3. "Open loop." Total makeup water for the system is 1700 to 2000 gpm. U. Gas residence time in the scrubber is not available. 5. Liouid delay time in the venturi recycle loop is about 2 minutes. 6 . Liquid temperature leaving the scrubber is 120° F. 7. Solids recirculated has recently been increased from 2 pet to 6 net. 8 . pH of the recycle loon on the scrubber is 3.2 to 3.5. pH at the thickener is U to 5 without lime added. A level of pH 7.5 at the thickener is to be maintained with lime addition. 9. Scrubbing liquor analysis is not available. 10. State of oxidation is not available. 11. Degree of supersaturation is not available. CDPERATING REQUIREMENTS 1. Lime is added at the rate of 6 tons/day for pH control. 2. No dispersing agent is being used. 3. System makeup water requirements are about 3^00 acre ft/vr. U. Power requirements. - Electrical requirements are 3 to h pet of generating capacity. 5• Manpower. - 8 operators. - Maintenance personnel not available. OPERATING RESULTS 1. Particulate removal meets the goal of 99.2 pet. 2. SO,-, removal is 30 to 35 pet without lime. - No typical SO 2 removal has been determined with lime. 3. Availability overall is estimated at 80 pet. b. Scaling has occurred on most wetted surfaces. 5. Methods used for scale control. - The level of recirculated solids was recently increased from 2 to 6 pet. - A lime system has recently been installed to maintain pH at 7.5 in the thickener. - The amount of blowdown used helps to control scaling. 6 . Problems. - The principal problem is that scaling is not under control. The effects of high pH in the thickener and 6 pet solids in the recirculation scrubber liquor have not yet been assessed because of their recent implementation, and future experience may be improved. llU - Corrosion with resulting leakage has occurred where coatings on mild steel have failed. - Solids buildup has occurred in blowdown lines. - Deterioration of stack linings. 7. Disposal of sludge. - Sludge settles well without a flocculating agent. The sludge is concentrated to 30 or 35 pet solids in the thickener underflow and is pumped after some dilution with blowdown to decanting ponds. Ash at present is left to accumulate in the ponds, but it may be dredged and returned to the mine. TRANSFERABLE TECHNOLOGY With 575 MW of installed scrubber capacity, this is the largest scrubber installation at any single location in the Western U.S. A con¬ tinuing program of large-scale innovation has been carried out in an effort to prevent scaling at reduced levels of system blowdown. Scaling is not yet under control despite very strenuous efforts. The effects of the last two innovations, which are recirculation of 6 pet solids and increased use of lime in the thickener, have not yet been evaluated. If successful, methods developed here should be widely applicable. The apparent rationale for adding lime in the thickener circuit would appear to be to precipitate calcium at a high pH in the thickener, and to use this liquor to dilute the more acid venturi recycle liquor and thereby remain below saturation in the scrubber. Somewhat similar methods reported for high-sulfur coals (6) depend on maintenance of a low state of oxidation to sulfate, with resulting precipitation of calcium sulfite and coprecipi¬ tation of sulfate. State of oxidation at Four Comers is not known. It should be noted that the 6 tons per day of lime used in this system represent only an estimated 7 pet stoichiometric equivalence to the SOg in the flue gases being treated. Thus, even with lime addition, this system is nowhere near representative of lime or limestone scrubbing for SOg. The lime added would, however, cause an appreciable increase in SO,-, removal, which feeds back through the system as an increased load of dissolved sulfur in sulfite or sulfate forms. The success of the lime-add scale control method depends on more than compensating for increased sulfate loading through an appropriate balancing of pH levels, dilutions, and hold times throughout the system. 115 Appendix II Scrubber Design and Operation (8_, 9) Dave Johnston Plant— Pacific Power and Light LOCATION 1. Glenrock, Wyoming 2. Elevation approximately 5000 feet. 3. Atmospheric pressure 12.3 psi. h. Annual precipitation is lU inches. 5. Plant water supply is the North Platte River. SCRUBBER APPLICATION 1. Particulate removal, retrofit. 2. One 330 MW Combustion Engineering pc-fired boiler (Unit no. 4). 3. Three parallel scrubbers. 4. Scrubber startup was April 1972. 5. Fuel is Wyoming subbituminous coal from a captive mine. Typical analysis is: - 7430 Btu/lb. - 26 pet moisture. - 0.5 pet sulfur. - 12 pet ash. - 20 pet CaO in ash. 6. Flue gas entering scrubber. - 1,500,000 acfm. - 270° F. - 500 ppm S0p. - 12 gr/sef (design). - 4 gr/sef (actual). 7. Removal goal. - 99.7 pet removal, or 0.04 gr/sef exit dust loading. SCRUBBER DESCRIPTION 1. Three identical scrubbers in parallel. 2. Venturi scrubbers. 3. Vendor, the Chemical Construction Company (Chemico). 4. Initial capital investment was $8 million, $24/kw. Costs incurred since startup have increased this amount significantly. 5. Operating cost is not available. 6. Materials of construction. - Scrubbers, vessels, outlet duct, and stack are polyester lined steel. - Piping and fan housings are rubber lined. - Fan wheels are Inconel. 7. No bypass. 8. Turndown is to approximately 30 pet of rated scrubber capacity. 9. Chevron mist eliminator. 10. Wet fans, no reheat. 116 Flue gas from air heaters 117 Figure A-2 - Simplified flow diagram for the Dave Johnston fly ash scrubbers. SCRUBBER OPERATING DATA 1. L/G is 13.3 gal/1000 acf, or 22 gal/1000 scf. 2. /ip is 10 inches of H O across the venturi, 15 inches total. 3. Intermittently "open loop." Normal operation is attempted at a makeup rate of 500 gpm, which compensates for evaporation and loss in sludge. The unit has teen operated at times with 3000 gpm fresh water makeup to flush out scale. *4. Gas residence time in the venturi section of the scrubber is estimated at about 1 second. 5. Liquid exit temperature is 126° F. 6. Liquid delay time in the venturi recycle loop is 2 to 3 minutes. 7. Solids recirculated are 2 pet of scrubbing liquor. 8. pH leaving the scrubber is about 5 without lime. Tests have been run at various pH ranging from 5 to 7 with lime addition. 9. Scrubbing liquor analysis is not available. 10. Degree of supersaturation is from 1.0 to 1.3. OPERATING REQUIREMENTS 1. Lime is added for pH control. 2. Additives tested include ligno sulfonate and hexameta phosphate. 3. Water requirements are approximately 800 acre ft/yr in "closed loop" mode. Actual requirements are greater because of occasional flushing. h. Power requirements. - Electrical power requirement is 7 to 8 MW, or 2.3 pet of generating capacity. - No steam is used for reheat. 5. Manpower requirements are not available. OPERATING RESULTS 1. Particulate removal meets the outlet grain loading goal of 0.0L gr/sef. 2. SO 2 removal (preliminary values). - 35 to L0 pet without lime. - Lime addition results in modest increase in SO 2 removal. Exact value has not been determined. 3. Availability not available, but it is not considered adequate for a utility power source. U. Scaling and plugging. - Solids buildup has occurred at the wet-dry interface. - Hard gypsum scale has formed in the scrubber vessels and piping. - Solids plug bleed and recycle lines. 5. Methods for controlling scaling and plugging. - Lime for pH control has resulted in reduction but not elimination of scaling. - Ligno sulfonate addition has resulted in a less adherent or friable wet-dry buildup. - Effects of hexamata phosphate on scaling and wet-dry buildup have not been determined. 118 - Continuous fan wash has essentially eliminated buildup on fans. - Fresh water washing has been required to flush ash and scale deposits from the scrubber vessel. 6. Additional problems. - Recycle pump errosion. - "Silting" during shutdown. 7. Disposal of sludge. - Bleed from the scrubber circuit is sent directly to two ash -ponds, from which overflow flows to a clear pond for recycle to the scrubber circuit. Each ash pond is dredged once per year, and the solids hauled out for land fill. Excess water resulting from periods of flushing is discharged to the North Platte River under a variance from the State of Wyoming. TRANSFERABLE TECHNOLOGY Severe operating problems at the Dave Johnston installation have been partially solved by strenuous development efforts, but the pivotal problem of controlling scale without flushing and producing blowdown has not been satisfactorily solved. As in the case of the Four Comers installation, solutions developed here should find wide application. 119 Appendix III Scrubber Design and Operation (ll) Valmont, Cherokee, and Arapahoe Stations Public Service Company of Colorado Note: Public Service Company of Colorado has a total of 12 similar scrubber modules of TCA design installed on five boilers at three stations. Because of the close similarity between these install¬ ations, they will be discussed collectively rather than individually. LOCATIONS Valmont _ Cherokee _ Arapahoe _ Southwest Denver, Elevation, ft. Atm. P, psi Annual rainfall, in. Plant water supply SCRUBBER APPLICATION 1. All units are retrofits for particulate removal. 2. Boilers equipped with scrubbers (all PC-fired). 3. Valmont #5, 196 MW, 2 scrubber modules, Nov. 71. Cherokee #1, 115 MW, 1 scrubber vessel, 2 modules, June 73. Cherokee #3, 170 MW, 1 scrubber vessel, 3 modules, Nov. 72. Cherokee 375 MW, 1 scrubber vessel, U modules, July 7^. Arapahoe #U, 112 MW, 1 scrubber module, Sept. 73. U. Arrangements of particulate cleaning equipment. - At Valmont, flue gas from a mechanical collector is split into two parallel streams, with 60 pet sent to the scrubbers and kO pet to an electrostatic precipitator (ESP). - All other units have a mechanical collector, an ESP, and scrubber(s) in series, with all flue gas entering the scrubber(s). 5. Coal burned at Valmont and Arapahoe is Wyoming subbituminous. - 8300 Btu/lb. - 29 net moisture. - 0.6 pet sulfur. - 5.2 pet ash. - 20 pet CaO in ash. 6. Coal burned at Cherokee is Colorado bituminous coal. - 11,000 Btu/lb. - 9.8 pet moisture. - 0.7 pet sulfur. - 9•^ pet ash. - 5 pet CaO in ash. Boulder, Colo. 5300 12.1 lU Hillcrest Lake North Denver, Colo. Est. 5200 12.1 Ik South Platte River Colo. 5600 Est. 12.1 Ik South Platte River 120 Flue gas to reheater it Flue gas from electrostatic precipitator Clear effluent discharge Figure A-3. - Typical scrubber installation at Valmont, Cherokee, and Arapahoe Stations, Public Service Company of Colorado. 121 7. Flue pas entering the stack pas cleaninp train. Valmont ff 5 Cherokee ff 1 Cherokee ff 3 Cherokee ffb Arapahoe ffb Flow, acfm U 63,000 520,000 610,000 1 , 520,000 520,000 T, ° F 271 285 272 267 305 SO 2 , ppm (est) 500 500 500 500 500 Dust, gr/sefd .8 .8 .b .7 .8 8. Removal poals for particulate. - The applicable Colorado State Standard is 0.1 lb/MM Btu, or about .05 pr/scf. - The coirroany's desire for clean stacks requires a poal of .02 pr/scf. SCRUBBER DESCRIPTION 1. All units are Turbulent Contact Absorbers, consisting of three stapes of mobile packing, or "ping pong balls," with spray directed downward through the balls and gas passinp countercurrent upward. 2. Vendor: Air Correction Division, Universal Oil Products Company. 3. Capital cost: (Average for system is $33/kw): Valmont #5 Cherokee ffl Cherokee #3 Cherokee ffb Arapahoe ffb $3,600,000 $3,810,000 ,237,000 $12,200,000 $U,560,000 $32/kw $33/kw $25/kw $33/kw $Ul/kw b. Operating costs not available. 5. Materials of construction (typical). - Scrubbers are rubber-lined steel with stainless steel grids. - Exit ducts are mild steel. - Slurry piping is rubber lined. - Pumps are rubber lined. 6. Bypasses on all units. T. Typical turndown is bj to 105 pet. 8. Mist eliminators have 2 stages, 7 passes. 9. Demisters for Valmont if 5 and Cherokee #1 and ff2 are flake reinforced polyester. Demisters for Cherokee fib and Arapahoe ffb are stainless steel. 10. All units except Cherokee ffb reheat the flue pas directly with steam coils; Cherokee ffb uses extemallv-heated air. 11. All units have dry fans that are forced-draft with respect to the scrubber. SCRUBBER OPERATING DATA 1 . 2 . L/G is typically 50. A P is approximately 10 inches to 15 inches of HgO depending on design and operating conditions. - 3 stages of mobile packing - 9 inches . - Mist eliminators. - Flake reinforced polyester, 1 1/2 to 3 inches. - Stainless steel - 1/2 inch. 122 - Reheat. - Direct steam coil reheater, 1 1/2 to 4 inches. - Hot air. - negligible. - Transition ductwork - 1/2 inch __ - Total 10 inches minimum to l 6 1/2 inches maximum 3. "Open loop." Amounts of makeup water from cooling tower blowdown are as follows: Valmont #5 230 gpm Cherokee ft 1 Cherokee #3 Cherokee #4 203 gpm 380 gpm 7 ^ gpm Arapahoe tib 203 gnm 4. Gas residence time in the scrubber is 3.8 to 5 seconds. 5. Liquid temperature leaving the scrubber is 125° F. 6 . Liquid holdup time in the recycle circuit is very short, estimated at ten seconds. 7. pH is 7 to 9 entering; 2.8 to 3 leaving the scrubber. 8 . A scrubber liquor analysis for Valmont (not known to be representative) - Ca - 590 ppm - Mg - 350 - Na - 1 - SO^ - 10,000 9. State of oxidation of dissolved sulfur is not available. 10. Degree of supersaturation is not available. OPERATING REQUIREMENTS 1. No lime or other reagent or additive is normally used. 2. Water requirements in acre feet/year (approximate). 3. Valmont #5 Cherokee til Cherokee ti3 Cherokee #4 340 300 Power requirements. Electric Power; 550 1100 o p . nsi : Arapahoe #4 300 Valmont ti 5 Cherokee til Cherokee #3 Cherokee #4 Arapahoe 6 MW 4 .6 MW 6.8 MW 15 MW 4.5 MW 5.1 % 4.0 % 4.0 % 4.0 % 4.0 1 Steam for •: 60,000 reheat: 46 ,680 60,000 110,000 60,000 700 420 715 634 360 490 300 300 1975 150 4. Manpower for scrubber operation is not identified by the company as a separate category from plant operators. Their estimate of scrubber manpower is 1 1/2 to 2 men per scrubber per shift for operation and 4 men per scrubber per day for maintenance. OPERATING RESULTS 1. Particulate removal achieves an outlet grain loading of .02 gr/scf. 2 . SO 2 removal. - 40 to 45 pet at Valmont and Arapahoe, burning Wyoming coal. - 20 pet at Cherokee, burning Colorado coal. 123 3 . Availability. Valmont #5 - 80 net for 12 months operation. - Cherokee #1 - 55 pet for 12 months operation. - Cherokee #3 - 80 pet for 12 months operation. - Cherokee - 85 pet for 3 months oneration. - Arapahoe - 20 pet until recently; about U0 pet currently. U. Problems. - Scaling and plugging has occurred at: - the wet/dry zone - the first stage grid - the reheater. - Corrosion has caused major failures of reheaters at the Cherokee Station. - Erosion and failure of linings have not been a serious problem, except in recirculating rumps. - Wear on the balls used for racking requires replacement after 6000 hours or less, at a cost of about $70,000 per unit for about 1 million balls per scrubber on all scrubbers except Cherokee , where cost is $ 220,000 for about 3 million balls. 5. Measures for control of scaling. - Additives tried have not worked, including phosphated esters. - Blowdown must be maintained at an adequately high level, but otherwise no specific methods are being used. 6 . Sludge disposal. Fly ash slurry is mixed with bottom ash to achieve some neutralization and lime added if required to brine- the pH into the range of 6.5 to 8 .5- The ash sludge is sent to settling ponds, from which it is periodically dredged for use as landfill. Clear effluent from the ponds is discharged under permit from the state of Colorado, to the South Platte River at the Cherokee and Arapahoe Stations, and to the cooling pond at the Valmont Station. At the Cherokee Station only, additional clarifiers are used to further clear up effluent before its discharge to the river, owing to the heavy loading on the ponds caused by discharge from the three scrubber-equipped boilers at this station. TRANSFERABLE TECHNOLOGY Public Service Company of Colorado has a total of 870 MW of installed scrubber capacity, all of the mobile packing type Turbulent Contact Absorber design. Particulate removal meets desired specifications, but as elsewhere scaling is an unsolved problem. Capital costs averaging $ 33 /kw are moderate. Electric power requirements at h pet of generating capacity are relatively high. Availability is not adequate by the standards of an electrical utility. The company is engaging in research to convert units of TCA type to lime or limestone scrubbing, but no results are available. Results on increased SO 2 removal and improved scale control on this type of scrubber should be widely applicable if they demonstrate reliability. 12h Appendix IV-A Scrubber Design and Operation (12) Clay Boswell Plant Minnesota Power and Light LOCATION 1. Cohasset, Minnesota. 2. Elevation approximately 700 feet. 3. Atmospheric pressure lU.3 psi. U. Annual rainfall approximately 25 inches SCRUBBEE APPLICATION 1. Particulate removal for a new plant. 2. 350 MW Combustion Engineering pc-fired boiler. 3. Scrubber startup May 1973. U. Fuel is Montana subbituminous coal from the Big Sky Mine. - 8800 Btu/lb. - 2k pet moisture. - 0.8 pet sulfur. - 9 pet ash. - 9 to 13 pet CaO in the ash. 5. Flue gas entering the scrubber. - 1,300,000 acfm. - 25k° F. - 800 ppm SOg. - 3 gr/sef. 6. Removal goals. - Minnesota particulate standard is 0.6 lb/MM Btu. - Scrubber guarantee is 0.03 gr/sef or 0.078 lb/MM Btu. SCRUBBER DESCRIPTION 1. A single Elbair spray-impingement scrubber. 2. One stage of high pressure spray is directed concurrent with gas flow against punch plate baffles to be atomized into fine droplets. 3. Vendor, Krebs Engineers. U. Size of scrubber is nominally 350 MW or 1.3 x 10 b acfm. 5. Capital cost is not available. 6. Operating cost is not available. 7. Materials of construction. - Scrubber is 3l6 LC stainless steel. - Outlet ducts are flake-polyester coated carbon steel. - Piping is fiberglass or rubber lined. - Pumps are rubber lined. 8. No bypass. 9. Turndown is 0 to 110 pet. 10. Demistor consists of one bank of vertical chevrons. 11. No reheat, wet fan. 125 SCRUBBER OPERATING DATA 1. L/G — 8.3 gal/1000 acf, or 13.3 gal/1000 scf. 2 . AP is 2.L inches HoO across scrubber, U inches total. 3. Not "closed loop" although scrubbing liquor is recycled from the clarifier back to the scrubber. Makeup water is approximately 1000 gpm. Of this, 280 gpm is used to compensate for evaporation in the scrubber and 720 gpm is blowdown to the ash pond. h. Gas residence time in the scrubber is 3 seconds. 5. Liquid delay time in one clarifier is 2 hours, or 1* hours if two clarifiers are on line. 6 . Solids circulated. - .02 pet entering scrubber. - .75 pet leaving the scrubber. 7. pH L. 5 in , U . U out. 8 . Scrubbing liquor analysis. - Ca -- 600 ppm. - Mg — 200 ppm. - Na — 15 ppm. - S0^ — 2300 ppm. 9. State Of oxidation of dissolved sulfur is high, estimated over 90 pet sulfate. 10. Degree of supersaturation is not measured, but it is evident that it is variable depending on the amount of soluble alkali in the ash, the amount of SOp absorbed, and the amount of blowdown removed from the system. OPERATING REQUIREMENTS 1. No reagent is regularly used. 2. Hydrochloric acid has been used to control pH as a means of controlling heavy scaling enisodes. 3. Water requirements are about 1500 acre ft/vr. *J. Power requirement. - Electrical power is about 3 MW, or .86 pet of net generating capacity. - No steam is used for reheat. 5. Mannower requirements are not available, but maintenance is known to be very high because of a continuous schedule of cleaning on nozzles. As of July 197*4, 85 to 100 manhours were spent each week removing fly ash and calcium sulfate scale deposits. OPERATING RESULTS 1. Particulate removal is close to 99 pet; exit dust loading is 0.03 gr/sef. 2 . SO 2 removal, occurring incidental to particulate removal by reaction with the alkaline fly ash and by sulfate removal in blowdown, is typically 15 to 20 pet. 3. Availability. - The Elbair scrubber, consisting of a stainless steel box containing high pressure spray nozzles that are removable for maintenance section by section, can remain on line without a bypass, even though the spray system is not operating. However, effective operation requires a very high level of maintenance effort. A percentage availability in terms of effective operation was not obtainable. 126 Figure A-4 - Simplified flow diagram for the particulate scrubber at the Clay Boswell station. 127 U. Scaling and plugging. - Scaling and plugging occurs in nozzles, nozzle trees, strainers, on punch plate baffles, in the wet-dry zone, and on the fan and mist eliminator; and deposits fall into the drains at the bottom of the scrubber. - Scaling is aggravated by any increase in CaO content in the ash of the coal beiner burned, which causes increased SCU removal and an increase in the level of Ca ++ and SO^ ions in solution. Under this condition, saturation is exceeded for the previously adequate level of blowdown, the solution pH rises and it becomes milky with a fine suspended precipitate of calcium sulfate, and scaling is greatly aggravated. In such crisis episodes, hydrochloric acid has been added to reverse these effects, apparently successfully. 5. Measures used for scale control. - Substantial amounts of blowdown are removed from the system to remain below saturation with CaSO^ - A very high level of cleaning and maintenance is carried out continuously. Nozzles and nozzle trees are removed and cleaned on a rotating schedule, with all nozzles being cleaned once per week. 6. Disposal of ash and spent scrubbing liquor. - Blowdown from two clarifiers, containing typically 5 or 6 pet fly ash, is sent to an 80-acre ash pond. In this region, there is no net evaporation but rather an accumulation of about 10 inches of water per annum. T. Problems. - No adequate solution to the problem of disposing of sulfate-laden blowdown water has been found. The two possible solutions would involve either discharge of diluted blowdown to streams or closing the loop in the scrubber circuit to eliminate blowdown. A break¬ through in methods of scale control would be required to eliminate blowdown. - A stack mist problem results primarily from washing of the wet fan to remove scale buildup. Attempts to use steam soot blowers in place of washing and to apply non-stick coating have been unsuccess¬ ful. Some improvement has been achieved by reducing the amount of fan washing. More extensive changes that have been considered but not adopted are installation of reheat or construction of a new low-velocity stack. TRANSFERABLE TECHNOLOGY The Elbair scrubber, it is generally agreed, must be operated on scrubbing liquor that is relatively clear and free of solids to avoid erosion of high pressure (200 psi) nozzles and physical plugging. Its application for particulate removal burning Western coal having a calcium-rich ash leads inevitably to a buildup of dissolved calcium sulfate in recirculated scrubbing liquor, which in the absence of blowdown reaches and exceeds the saturation level and results in severe chemical scaling. Inability to recirculate solids through the system eliminates one effective means of 128 reducing scaling. If large amounts of sulfate-laden blowdown cannot be safely discharged to aquifers, the classical dilemma between scaling and water pollution applies in its full force. The choice of the Elbair scrubber was motivated chiefly by its ability to remove particulate without a large pressure drop in the flue gas stream, and therefore without a large power requirement. With an electrical power usage of 0.86 pet of generating capacity, this advantage has been borne out in practice. The mist carryover problem results from one or more of the following factors: a one stage mist eliminator, a wet fan requiring washing, a high velocity stack, and lack of reheat. None of these relate specifically to the main scrubber design, and the oroblem should be solvable. If this installation were representative, design with a wet fan and without reheat would not be recommended. 129 Appendix IV-B Scrubber Design and Operation (12) Aurora Station Minnesota Power and Light LOCATION 1. Aurora, Minnesota. 2. Elevation 1500 feet. 3. Atmospheric pressure lU.l psi. U. Annual rainfall approximately 25 inches. SCRUBBER APPLICATION 1. Particulate removal, retrofit. 2. Two 58-MW pc-fired boilers. 3. Scrubber startup June 1971. U. Fuel is Montana subbituminous coal from the Big Sky Mine. - 8800 Btu/.lb. - 2h pet moisture. - 0.8 pet sulfur. - 9 pet ash. - 9 tol3 pet CaO in ash. 5. Flue gas entering the scrubber on each boiler. - 291,160 acfm. - 3^0° F. - 800 ppm SO 2 . - 2.06 gr/sef 6. Removal goals. - Minnesota particulate standard is 0.6 Ib/MM Btu. - Scrubber guarantee is 0.03 gr/sef or 0.078 Ib/MM Btu. SCRUBBER DESCRIPTION 1. One Elbair spray-impingement scrubber for each of the two boilers. 2. One-stage of high pressure spray is directed concurrent with gas flow against vertical rods to be atomized into fine droplets. 3. Vendor, Krebs Engineers. U. Size of each scrubber is nominally 60 MW or 300,000 acfm. 5. Capital cost is not available. 6. Operating cost is not available. 7. . Materials of construction. - Scrubber is 3l6 ELC stainless steel. - Outlet ducts are flake-polyester coated carbon steel. - Piping is fiberglass, or rubber lined. - Pumps are rubber lined. 8. No bypass. 9. Turndown is 0 to 110 pet. 10. Demistor consists of one bank of vertical chevrons. 11. No reheat, wet fan. 130 SCRUBBER OPERATING DATA 1. L/G is 8.3 gal/1000 acf, or 13.3 gal/1000 scf. 2 . AP is 2.5 inches of HgO across scrubber, U inches total. 3. Not closed loop, although scrubbing liquor is recylced from the ash pond back to the scrubber. No clarifier is used. Makeup water is approximately 1200 gpm for each unit. L. Gas residence time in the scrubber is approximately 3 seconds. 5. Liquid delay time in the recycle circuit is not available. 6 . Solids circulated. - .02 pet entering scrubber. - .75 pet leaving scrubber. 7. pH estimated at U.5 in, U.L out. 8 . Scrubbing liquor analysis is not available. 9. State of oxidation of dissolved sulfur is high, estimated over 90 pet sulfate. 10. Saturation of scrubbing liquor does not occur at the level of blowdown used. OPERATING REQUIREMENTS t 1. No reagent is used. 2. Scrubber water requirements are about 3500 acre ft/yr. 3. Power requirement. - Electrical power is about .5 MW per unit, or .8 pet. - No steam is used for reheat. U. Manpower requirements are not available. OPERATING RESULTS 1. Particulate removal is about 98 pet; exit dust loading is .0U to ,0L6 gr/sef. 2 . SO 2 removal, occurring incidental to particulate removal by reaction with the alkaline fly ash and by sulfate removal in blowdown, is typically 20 pet. 3. Availability. - The Elbair scrubber, consisting of a stainless steel box containing high pressure snray nozzles that are removeable section by section for maintenance, can remain on line without a bypass, even though the spray system is not operating. However, effective operation requires a high level of maintenance effort. A percentage avail¬ ability in terms of effective operation was not obtainable. U. Scaling and plugging. - Scaling and plugging has not been too severe, owing to the amount of blowdown used. 5. Disposal of ash and spent scrubbing liquor. - Blowdown, estimated to contain 1 pet solid fly ash, is sent an ash pond. Overflow from the pond is neutralized with lime before disposal. 131 6. Problems. - This unit would suffer from the same scaling and plugging problems as the Clay Boswell scrubber of similar design, described in the preceding section, if the recycle loon were closed to a similar extent. As operated, it is less troublesome. - The mist carryover problem is less severe than at Clay Boswell owing to operation at partial load with a resultant lower stack velocity. TRANSFERABLE TECHNOLOGY The application of the Elbair scrubber to particulate removal burning Western coals is discussed in the preceding section. The same assessment applies for this unit as well. FLOW DIAGRAM The flow diagram for the Aurora scrubber is similar to that for the Cohasset unit drawn in figure A-U, except that: 1. No clarifiers are used, and recycle liquor is brought back directly from the ash pond. 2. Rods are substituted for the punch plate. 132 Appendix V Scrubber Design and Operation (13, ll+ ) Mohave Generating Station Southern California Edison Company LOCATION 1. Clark County, Nevada. 2. Elevation TOO feet. 3. Atmospheric pressure lU.3 psi. 1+. Annual rainfall approximately 7 inches. 5. Water supply for the plant is the Colorado River. SCRUBBER APPLICATION (Horizontal Scrubber Unit) 1 . SO 2 removal, retrofit. 2. One nominally 170 MW scrubber is installed on a slip stream from a 790 MW Combustion Engineering pc-fired boiler. 3. The boiler is equipped with 98 pet efficient, cold-side, electrostatic precipitators. 1+. Scrubber startup was November 1973. 5. Fuel is Arizona bituminous coal from the Black Mesa Mine. - 11,000 Btu/lb. - 10 pet moisture. - 0.38 pet sulfur average. - 9 pet ash. - 15 pet CaO in ash. 6 . Flue gas entering the scrubber. - 1 + 75,000 scftn nominal. - 280° F. - 200 ppm SO 2 average. - 0.07 gr/sef particulate nominal. 7. Removal goals were set by Clark County, Nevada standards. - SO 2 - 0.15 lb/MM Btu, or 50 ppm - Particulate - Ringleman 1, or about 0.02 gr/sef. - Project goals were to exceed these standards by a safe margin. SCRUBBER DESCRIPTION (Horizontal Scrubber) 1. Note: A second scrubber, a Turbulent Contact Absorber desiemed by Universal Oil Products, is not described in this paper. This unit was started up in January 1971+, but was damaged by fire January 2l+, 1971+. Operation was recommenced in October 1971+, but no operating results are available. 2. Horizontal Cross Flow Scrubber. - This unit consists of approximately 50 feet of horizontal duct¬ work (cross section is 28 ft wide by 15 ft high) separated into four sections, each stage having its own spray header and drain. Scrubbing liquor from the Mix-Reaction tank is pumped to the 133 fourth staae spray header, and drainage from each stage is punned as snrav to the preceeding stage until the first stage drains back to the Reaction tank. The unit uses low pressure soray nozzles that permit recirculation of solids. 3. Reagent is lime. b. Vendor: The 170 MW scrubber was designed, constructed, and tested as oart of the Nava.1 o/Mohave Test Module Program, representing six utilities and the Bureau of Reclamation. Smaller 1 MW and 10 MW horizontal scrubbers of similar design were sponsored by Southern California Edison and designed and built by Stearns-Roger, Inc. 5. Size is nominally 170 MW or 450,000 scftn. 6 . Capital cost is not available, but it is estimated to fall close to the $ 50 /kw level that is presently characteristic of lime scrubbers. 7. Operating costs are not available. 8 . Materials of construction are not available. 9. Bypass capability is inherent in this experimental installation, which treats a 20 pet slip stream from the boiler. A different kind of "bypass" is the capability to continue operation on three, two, one, or presumably no operating stages. This redundancy permits repairs to be postponed until a scheduled shutdown. The amount of on-line maintenance that is possible on headers and spray nozzles is not known, but with proper design, it should be possible to perform most scrubber maintenance without shutdown. 10. Turndown has been demonstrated from 30 pet to 120 pet of the nominal rating of ^50,000 scfm, with SOj removals over 90 pet. 11. Two demistors of undesignated design are shown on the publish flow diagram. Pressure drop for demisters at design operating conditions is listed for only one demisting stage. 12. Dry booster fan, forced draft with respect to the scrubber. 13. The reheater heats outside air for mixing with exiting flue gas. SCRUBBER OPERATING DATA 1. L/G is nominally 20 gal/1000 scf for each stage, but operating data are reported for values from 10 to 25. An L/G of 12.5 meets a criteria of Uo ppm SOg exiting burning 0.38 pet sulfur coal (200 ppm SO 2 ). An L/G of 17.5 meets this criteria for 0.83 pet sulfur coal, which is the highest sulfur level ever excepted for the Black Mesa coal. 2 . AP totals 6 inches of water under design operating conditions. - Inlet ductwork - 1 inch of water. - Scrubbing chamber - 1 inch of water. - One demisting stage - .5 inches of water. - Reheater mixing chamber - 2.5 inches of water. - Outlet ductwork - 1 inch of water. TOTAL - 6 inches of water. 3. "Closed loop" operation. Makeup water rate is 152 gpm, which is made up 82 pet of cooling tower blowdown having 15 to 20 pet dissolved solids and l8 pet service water. Losses that balance makeup are 93 pet evaporation in the scrubber and 7 pet sludge entrainment and pond evaporation. U. Gas residence time in the scrubber is 2 to 3 seconds. 13U o o e q- O O - £ W CL ° o o> •— ^1 ^ / \ — d 3 \ -a c o '■4— o o c CD k- >< 5 o 135 Figure A-5 - Simplified flow diagram for the Mohave scrubber . 5. Liquid delay time in the recycle loop through the Mix-Reaction tank is not known. 6. Liquid and gas temperature leaving the scrubber is approximately 130° F. 7. The concentration of solids in the recirculated scrubbing slurry is not available, but the system design is believed to permit substantial levels of solids. 8. pH in the scrubbing circuit is not available. 9. Scrubbing liquor analysis is not available. 10. State of oxidation is not available. 11. Degree of supersaturation is not available. OPERATING REQUIREMENTS 1. Lime required for the 170 MW scrubber is calculated to be 8 tons per day (about 3000 tons/yr) based on 99.5 pet lime utilization, as published, and removal of SO 2 from 200 ppm entering to U0 ppm at the outlet. 2. Information on additives used, if any, is not available. 3. Water requirement is approximately 220 acre ft/yr. U. Power requirement. - Electrical power averaged 2.66 MW, or about 1.5 pet generating capacity. - Reheat steam is ^0,000 lb/hr at U00° F and 350 psig, which is calculated to be equivalent to 2 MW, or 1.2 pet of generating capacity. - Total power requirements are 2.5 to 3 pet of generating capacity. 5. Manpower. - 2 operators and 1 foreman per shift. - Maintenance - 136 man-hours per week. - Multiple commercial units would be expected to require less manpower. OPERATING RESULTS 1. S0 2 removals of 70 to 97 pet are reported depending on the SO 2 level entering, flue gas flowrate, number of stages operating, and L/G. A typical value is 90 pet removal of 200 ppm SO 2 entering at an L/G of 17.5 using U stages of scrubbing. With only two stages of scrubbing, a typical removal is 70 pet. 2. Particulate removal. - 98 pet at 1.0 gr/sef entering dust loading. - 70 pet at 0.01 gr/sef entering dust loading. 3. Availability was 85 pet relative to the time that the boiler was operating. U. No serious scaling occurred in the scrubber. 5. Methods that were used to control scale are not available. 6. Sludge disposal. - Thickener underflow is pumped to a sludge disposal pond and clear liquor is pumped back to the scrubber circuit. The system is designed for maximum utilization of waste water. 7. Problems. - Scrubber debris solids plugged 1/U inch inlet screens to the mix tank; replacement with screen having 3A inch openings solved this problem. 136 - Hard calcium sulfate scale formation in the lime slaker was eliminated hy using station service water for slaking. - Various mechanical problems have been solved. TRANSFERABLE TECHNOLOGY The Horizontal Cross Flow Scrubber is the culmination of an extensive developmental program involving eight smaller pilot plant scrubbers and four different reagents—lime, limestone, soda ash, and ammonia. This particular design was tested at 1 MW and 10 MW. The technical information generated in the program should find wide application to scrubbing in the Western U.S. The design philosophy applied in constructing a long empty box with a series of spray stages was to obtain maximum freedom from internal plugging with provision for an extended and variable gas-liquid contact and residence time. The design was motivated in part by a desire to maintain a low pressure drop, to reduce both capital and operating costs. The extended gas-liquid contact time was incorporated to compensate for the low SOp partial pressure that is available to act as a driving force for absorption where high percentage removal is required starting with an inlet level of a few hundred pnm SOj- This program must be judged highly successful overall, compared with many other scrubber operations. The relative freedom from serious scaling under conditions that are the closest approach to closed loop operation that is likely attainable in an important accomplishment. Unfortunately, the exact conditions of liquid residence times, state of oxidation, pH, and stream mixing that may have contributed to this success are not available. The unobstructed design of the scrubber interior is a contributing factor, but it does not affect scaling elsewhere in the system. The availability of 85 pet that has been published is not adequate by utility standards. However, if the problems that have been published are representative, improvement seems possible. With a modular desiCT, it should be possible to perform many maintenance operations relating to the liauid spray system without scrubber shutdown. 137 Appendix VI Scrubber Design and Operation (l5_) Cholla Station Arizona Public Service Company LOCATION 1. Joseph City, Arizona. 2. Elevation approximately 5000 ft. 3. Atmospheric pressure approximately 12.3 psi. U. Annual rainfall 7 inches. 5. Plant water supply from deep wells. SCRUBBER APPLICATION 1. Particulate and SOp removal, retrofit. 2. Startup October 19(3, commercial operation December 1973. 3. 115 MW Combustion Engineering wet bottom (slapping) boiler. U. In series with a mechanical ash collector. 5. Fuel is New Mexico bituminous coal from the McKinnley Mine. - 10,U00 Btu/lb - 0.U to 0.5 pet sulfur - 9.6 pet ash 6 . Flue pas entering scrubber. - U80,000 acfm - 260 to 270° F - U00 to 500 ppm SO 2 - 1.2 gr/sef particulate 7. Removal goals set by project. - 0.2 lb particulate/MM Btu - 1.0 lb S0 2 /MM Btu SCRUBBER DESCRIPTION 1. Two scrubbers in parallel. - Both have a flooded disk venturi for particulate removal. - One has a packed tower for SOp removal. - Second scrubber has an empty non-functioning tower. 2. Reagent is limestone. 3. Vendor, Research Cottrell. U. Capital cost is $57/kw. 5. Operating cost not including capital cost is 0.6 mills/kw hr. - Total cost estimated to be 3 mills/kw hr. 6 . Materials of construction. - Flooded disk venturi and absorption tower are 316 L stainless steel. - Process vessels and outlet ducts are flake lined steel. - Some liquid lines are plastic (fiberglass). - Valves and pumps are rubber lined. - Reheater is 316 stainless steel. - Fan is carbon steel. 138 To stack Figure A-6 - Simplified flow diagram for the Cholla scrubber having a functional packed tower. 139 T. Bypass — Either of the two scrubbers can be bypassed independently. 8. Turndown to UO MW on one unit or to 80 MW on both, to about 60 pet of rated scrubber capacity. 9. Demistors. - Cyclone on venturi. - 2 stage chevrons plus slat-type on towers. 10. Reheat (U0° F rise) by direct heating of flue gas with steam coils. 11. Dry fan is forced draft with respect to scrubber. SCRUBBER OPERATING DATA 1. L/G 15 gal/1000 acf to flood disk venturi. U 5 gal/1000 acf to packed tower. 2. AP Design was 15 inches H 2 O across venturi and 20 inches total. Actual is 6 to 10 inches across venturi and 20 inches total. The high total is due to a high AP across the retrofit ductwork and the reheater. 3. "Open loop." U. Makeup water is 60 gpm to each flooded disk and 20 gpm to the functioning tower. No recycle is returned from the disposal pond. 5. Gas residence time in the tower is approximately .25 sec. 6 . Liquid delay time in the tower recycle tank is 10 minutes. Delay time in the venturi tank is 5 minutes. 7. Liquid temperature leaving scrubber is 125 to 130 F. 8 . Liquid holdup time in tower recycle tank is 8 minutes. 9. Scrubbing liquor analysis. - Venturi recycle. - 1.5 % CaS0 3 - 0.3 1 CaSOjJ - 0.5 1 CaC0 3 - 10 l fly ash solids - Solution is saturated with CaSO^ - Tower recycle. - 5 % CaS0 3 - 0.1 % CaSOjj - 3 ! CaC0 3 - 0.5 % fly ash solids - Solution is not saturated with CaSO^ 10. pH not controlled. - 6.5 into tower. - 5.2 into venturi. 11. State of oxidation is high in the venturi recycle; low in the tower recycle. 12. Degree of supersaturation is not available. OPERATING REQUIREMENTS 1. Limestone used is 15 tons/day at full load,at $20/ton. 2. Additives — none. 3. Water requirement with both venturi scrubbers and the one tower operating continuously is 205 acre feet/year. 1 U 0 4. Power requirements. - 2.8 MW electric power or 2.4 pet. - 1.8 MW equivalent for 18,000 lb steam/hr for reheat. - Total power requirement is 4 pet of net generation capacity. 5. Manpower. - 4 operators (l per shift). - 30 hours direct maintenance per day. - Supervision not estimated. - Scrubber operation has no full time engineer assigned at the plant OPERATING RESULTS 1. SO 2 removal. - 90 pet in packed tower scrubber unit. - 20 pet in the unit with a non-functioning tower. - About 60 pet overall. 2. Particulate removal. - 80 pet in mechanical ash collector. - 99 pet additional removal in scrubber. - 0.026 gr/sef exit dust loading. 3. Availability. - Overall average of 91.5 pet for both units together, achieved by a "very high level of effort." - Availability is lower (86 pet compared with 95 pet) on the unit with the non-functioning empty tower because of corrosion and fouling of the reheater and outlet duct work under more acid conditions. 4. Scaling and plugging. - No chemical scale because of "open loop" operation. - No plugging on polypronylene tower packing. - Soot blowers are used to control fouling of reheaters. 5. Disposal of sludge. - The blowdown stream from the venturi recycle loop, containing 15 pet solids, is sent to a sludge storage tank, which is periodically pumped out to the ash pond. No liquid is returned from the ash pond, where accumulation is controlled by evaporation 6. Problems. - Corrosion and fouling of reheaters is a continuing problem at the relatively more acid conditions in the unit with the non-functioning tower. Consideration is being given to converting this unit to have a functioning packed tower. - Erosion has occurred on the throat of the stainless steel venturi. Silicon carbide brick or rubber lining is proposed as a solution. - Sulfite solids form in the tower circuit, but these do not adhere to surfaces to cause severe scaling or plugging. TRANSFERABLE TECHNOLOGY The Cholla Station scrubber has operated quite successfully, and most results should be widely applicable. A possible crucial exception is the relative freedom from chemical scaling in the scrubber, including the packed tower, which at Cholla is favorably influenced by a degree of "open loop" operation. Open loop in this case means a significant amount of fresh water makeup used to offset blowdown that can be evaporated in this arid climate. The extent to which a similar operation would remain scale free if the scrubber loop were more stringent, as would be dictated in areas of plentiful rainfall, can be questioned. Scaling is controlled in the venturi recycle circuit, which operates at a high state of oxidation and at saturation with respect to calcium sulfate, by recirculation of ash solids and a liquid delay time sufficient to provide a degree of desupersaturation by precipitation. The tower circuit operates at a low state of oxidation and is maintained below calcium sulfate saturation with makeup water. The finding that fouling and corrosion of a stainless steel reheater is reduced by removal of larger amount of SO 2 (in the packed tower unit) should be generally transferable result for any similar reheater design. lU2 Appendix VII Scrubber Design and Operation Reid Gardner Station Nevada Power Company LOCATION 1. ^5 miles northeast of Las Vegas, Nevada. 2. Elevation approximately l600 feet. 3. Atmospheric pressure approximately lU psi or more. h. Annual rainfall (Las Vegas) 5 inches. 5. Plant water supply from deep wells and the Muddy River. SCRUBBER APPLICATION 1. Particulate and SO 2 removal, retrofit. 2. Startup March 197^ on unit 1, April 197^ on unit 2. 3. Two 125 MW Foster Wheeler pc-fired boilers. k. In series with mechanical ash collectors of 80 pet efficiency. 5. Fuel is Utah bituminous coal. - 12,300 Btu/lb. - 5 pet moisture. - 0.6 pet sulfur. - 9 net ash. - 8 to l8 pet CaO in ash. 6 . Flue gas entering the scrubber on each boiler. - U73,000 acfm. - 350° F. - hOO ppm SOj. - 0.3 to 0.6 gr/sef particulate. 7. Removal goals are determined by Clark County, Nevada Emission Regulations. - Ringleman 1 or about 0.02 gr/sef. - 0.15 lb S0 o /MM Btu or 50 ppm SO 2 . - Mass emission restriction for particulate based on plant heat input (Logarithmic scale). SCRUBBER DESCRIPTION l. Two similar scrubbers, one for each 125 MW boiler. 2. Reagent is soda ash (Na2C02) or Trona (66 pet Na2C0^ plus NaCl and Na 2 S 0 ^) and insolubles (sand). 3. Venturi plus a "separator" with one flooded tray (sieve tray). H. Vendor, Combustion Equipment Associates. 5. Size for each scrubber, 1.25 MW or ^73,000 acffri, including disposal/ evaporation ponds. 6 . Capital cost is $11 million total, $UL/kw. 7. Total cost per kw hr is approximately U to 6 mills based on the average environmental surcharge on utility bills. 8. Materials of construction. - Venturi — Incoloy 825. - Sieve plate, valves, and mist eliminator — 3l6 LC stainless steel. - Outlet ducts — original Ceilcote failed due to excessive temp¬ erature (U00° F). Plasite U00U-5 epoxy substituted successfully. - Vessels and piping — rubber lined steel. 9. Bypass 100 pet (fully automated). 10. Turndown to 30 pet of full load (with seive tray out of service). 11. Demistors are of radial vane design (airfoil shapes). 12. Reheat to 169° F by mixing flue gas with air heated by steam coils. 13. Dry fan is forced draft with respect to scrubber. 3000 HP/scrubber fan. SCRUBBER OPERATING DATA 1. L/G is 9.5 gal/1000 acf, or 12.5 gal/1000 scf (venturi). Tray is approximately 1 gal/1000 acf. 2. AP is 15 inches H 2 O for venturi; 18 inches total. 3. "Open loop" by virtue of 280 gpm evaporation in the two scrubbers plus 100 gpm from ponds. No liquid is recycled back from the ponds. U. Gas residence time in scrubber not available, but Judged unimportant in a scrubber operated on soda ash. 5. Liquid temperature leaving scrubber is 135° F. 6. Liquid holdup time in recycle loop is 6 minutes. 7. Suspended solids recirculated 5 pet. 8. pH is 6.8 entering the venturi; 5.8 to 6.U leaving. pH in the flooded tray recycle tank ds 3 to 5. 9. Scrubbing liquor to venturi. - 5.3 pet fly ash solids. - 8.U pet dissolved solids. - 5.6 pet Na2S03. - 0.7 pet NagSOli. - l.U pet NaCl. - 0.7 pet other. 10. State of oxidation — 11 pet sulfate, 89 pet sulfite. 11. Scrubbing liquor is well below saturation with calcium salts. 12. Scrubber operation is fully integrated with boiler automated purge sequence and digital interlock system. OPERATING REQUIREMENTS 1. Reagent for both units — 15,000 tons Trona/yr @ $U0/ton, or 10,000 tons soda ash @ $75/ton, 0.3 to 0.U mills/kw hr. 2. Additives — makeup water treated with Nalco 32A09 (scale inhibitor). 3. Water use is 550 acre ft/yr based on evaporation. To stack Figure A-7 - Simplified flow diagram for Reid Gardner station. one scrubber at the 4. Power requirements. - Electric power is 2.4 MW ner unit or 2 pet; reheat steam equivalent to 3 MW per unit or 2.4 pet. 5. Manpower. - 4 operators (l shift). - 3 maintenance mechanics. - 1 instrument man. OPERATING RESULTS 1. SO 2 removal. - 84 pet design removal level is met. - 95 pet removal or higher can be attained by adding more soda ash. 2. Particulate removal. - 80 pet in mechanical collector. - 97 pet of remaining in scrubber. - 99.4 pet overall, meeting the 0.02 gr/sef standard. 3. Availability. - Until recently, availability was limited by lack of soda ash. In last three months, monthly availabilities on the individual units has ranged from 70 to 94 net as a percentage of boiler operating time. A tyrical value is 90 pet. During March, one unit achieved 99.4 pet availability. 4. Scaling and plugging. - No chemical scale. - Fly ash plugging in liquid lines and valves has been corrected by increasing the size of control valves to permit full flow through valves without bypass, to achieve higher velocities. 5. Disposal of ash and spent scrubbing liquors. - Blowdown from the venturi recycle tank is neutralized to pH 7 and pumped to two 4-acre lined settling ponds, from which ash is dredged periodically for land fill. - Clear liquor from the settling ponds is sent to a 47-acre, unlined evaporation pond, which is isolated laterally from the surrounding area by a trenched ring of clay-sand slurry. Downward mobility of salt is limited by several clay horizons. A series of 24 perimeter wells are monitored weekly to detect any lateral movement of salt. No leakage has been detected in six months of operation. 6. Problems. - Ash plugging in liquid lines; solved by valve and piping modifications. - Corrosion of rubber lined pining. The failed areas of duct lining have not experienced significant corrosion because of adequate reheat of stack gas. Ceilcote has been replaced with an epoxy coating. Piping failures have been largely "infant mortality" and are not recurring failures. - Plugging in piping due to flaked-off rubber lining. - Plugging of the clarifier underflow system by sand found in trona. - Difficulty in feeding trona into mix tank due to inadequate drive torcue. 146 TRANSFERABLE TECHNOLOGY The Reid Gardner installation as operated depends stroneTy on special circumstances existing in the locality of this plant. The non repenerat inp sodium scrubbing method employed would not be widely applicable. Use of the method depends on an economic source of soda ash, an arid climate for evaporation of scent sodium sulfate solution, terrain and pround hydrolopy that remits lonp-tem containment of lame quantitities of soluble sodium sulfate, and a low sulfur content in the fuel. Where applicable, the process has a potential for 95 pct+ SO 2 removal and can be expected to be relatively trouble free. In particular, it would not suffer hard chemical scale unless bleed were reduced to a point where ash-derived calcium caused saturation with respect to calcium sulfate or sulfite. Corrosion has been a problem; however, this difficulty is judped to be a startup problem and it is solveable by pood maintenance and substi¬ tution of improved materials at critical points. Prevention of chloride stress corrosion crackinp requires expensive alloy construction. This process, as well as all others which produce sodium sulfate waste, would find extended application if an economic means were found to fix the waste in an insoluble form. An economic method is judped to be unavail¬ able at the present time. Brine concentrators, evaporator crystallizers, and flash evaporators are available and were considered for use on the Nevada Power project but were evaluated out due to increased cost over evaporation ponds. They could be used and the product mipht be marketable. Beyond this, some research is beinp done on other projects to trap sodium sulfate waste in an insoluble "concrete-like" matrix, but no commercial process is available. lUT STATUS OF THE CITRATE PROCESS FOR S0 2 EMISSION CONTROL^/ by W. A. McKinney ,—I W. I. Nissen,^/,Laird Crocker, and D. A. Martin?/ INTRODUCTION Research has been conducted in this country and abroad for nearly 50 years to develop technology for control of sulfur dioxide emissions from coal-burning powerplants. Of many processes investigated, five wet-scrubbing systems have received the most attention for commercial application. These processes are (l) wet limestone or lime scrubbing, (2) double alkali scrubbing, ( 5 ) magnesium oxide scrubbing, (4) cata¬ lytic oxidation, and (5) sodium-base scrubbing. The first two processes are nonregenerable and produce a throwaway sludge. In the remaining three processes, the sulfur dioxide absorbent is regenerated, and marketable sulfuric acid or elemental sulfur is produced. Reacting sulfur dioxide with limestone or lime slurries in scrubbing towers to form a calcium sulfate-sulfite sludge is the most thoroughly studied of all S0 2 control systems. Lime or limestone scrubbing has the advantage of removing particulate matter as well as sulfur dioxide, and capital costs are low compared with those of other processes. Principal disadvantages are (l) the system is nonregenerative requiring continued replacement of absorbent; (2) considerable scaling, plugging, erosion, and corrosion can occur resulting in poor reliability and high operating costs; (3) a throwaway sludge is produced requiring large areas for impoundment, settling, and stabilization; and (4) water pollution problems can result from the leaching of soluble salts from the sludge. X7Presented at the 1975 Lignite Symposium, Grand Forks, N. Dak., May 14-15, 1975. 2/ Research director. 3/ Metallurgist. hj Chemical engineer. All authors are with Salt Lake City Metallurgy Research Center, Bureau of Mines, U.S. Department of the Interior, Salt Lake City, Utah. Within the scrubber itself, the plugging and scaling problems associated with limestone and lime scrubbing are almost completely eliminated in the double alkali process. This system uses a clear sodium or ammonium alkaline liquor to absorb the sulfur dioxide, and the resulting solution is treated outside the scrubber with either limestone or lime to regenerate the absorbent solution and produce a throwaway sludge of calcium sulfate and sulfite. This system still has the disadvantage of producing a throwaway sludge with its attendant disposal problems. It also generates a purge stream of sodium or ammonium sulfate that must be disposed. The cost of sludge disposal may be a deterrent to consideration of either limestone-lime scrubbing or the double alkali systems. If waste disposal costs are above $3 to $4 per ton of wet sludge, otl covery processes may be less costly than throwaway processes Magnesium oxide scrubbing comprises reaction of sulfur dioxide with a magnesia slurry followed by thickening, centrifuging, and calcining the resulting magnesium sulfite and sulfate with added carbon to regenerate the magnesium oxide and evolve dilute sulfur dioxide for conversion to sulfuric acid. The process has the advantage of regen¬ erating the absorbent at a separate site away from the scrubbing units, possibly at a centrally located regeneration facility servicing several plants. Major disadvantages are (l) high heat requirement for drying and regenerating the magnesia; (2) potentially high costs because of many stages involved such as thickening, fly ash separation, centrifuging, magnesium salt crystallization, and thermal decomposition; (3) questionable marketability of sulfuric acid; and (4) costly alternative reduction of dilute S0 2 gas produced to elemental sulfur. In the catalytic oxidation process, S0 2 -bearing flue gas is thoroughly cleaned of dust in hot precipitators and then is passed over a catalyst to convert the sulfur dioxide to sulfur trioxide, which combines with moisture to form a 80-percent sulfuric acid as the end product. This process has the advantage of simple operation, and no cooling of the gas is required. Drawbacks are the thorough cleaning of the flue gas required to keep particulates out of the catalyst bed, the low economic value of the dilute sulfuric acid produced, and potential problems with disposal of large quantities of this acid. The best known sodium-base scrubbing regenerable systems are the Wellman-Power Gas sodium sulfite-bisulfite process and the Bureau of Mines sodium citrate process. In the Wellman-Power Gas process, sodium sulfite-bisulfite solution absorbs sulfur dioxide from stack 57 Underlined numbers in parentheses refer to items in the list of references on page 172. gas and converts sulfite to bisulfite. The loaded liquor is processed in a steam-heated evaporator-crystallizer to recover strong sulfur dioxide and regenerate sodium sulfite for recycling. The sulfur dioxide can be used to make sulfuric acid or can be reduced to elemental sulfur using natural gas and a modified Claus process. Advantages of this sodium-based system are as follows: (l) No scaling or plugging problems are present because of the use of clear absorbent liquor, (2) absorbent is regenerated, and ( 3 ) marketable sulfuric acid or elemental sulfur is produced. Drawbacks are (l) appreciable oxidation of sulfite to sulfate with possible problems in disposal of bleed stream, (2) questionable availability of natural gas for initial reduction of sulfur dioxide to sulfur, and ( 3 ) costly intermediate Claus plant required for production of elemental sulfur. The Bureau of Mines citrate process for removal of sulfur dioxide and recovery of elemental sulfur from waste gases comprises absorbing the S0 2 in a sodium citrate solution, reaction of the absorbed S0 2 with gaseous H 2 S in the aqueous medium to precipitate sulfur and regenerage the absorbent liquor, separating and melting the sulfur, and recycling the regenerated citrate solution to the S0 2 absorption step. The H 2 S for absorbent regeneration and sulfur precipitation can be obtained by reacting a bleed stream of product sulfur with natural gas and steam over an alumina catalyst. If byproduct H 2 S is available from nearby petroleum or sour gas refineries, the purchase of H 2 S may be more economical than onsite generation. Research on techniques for removing S0 2 from waste gases was initiated at the Bureau of Mines Salt Lake City Metallurgy Research Center in 1968 . After screening many possible reagent combinations of inorganic and organic solutions, researchers established that a solution of citric acid and sodium citrate was a very effective absorbent for S0 2 and had most of the desirable characteristics that had been sought. Among the factors affecting the choice of citrate were chemical stability, low vapor pressure, adequate pH buffering capacity, and the purity and physical character of the precipitated sulfur. Because the metallurgy research program of the Bureau of Mines is oriented toward improving metals and minerals processing technology, including the reduction or elimination of air, water, and land pollution from mineral processing operations, this work was directed toward abatement of S0 2 emissions from nonferrous smelters. Preliminary Bureau of Mines research on the citrate process attracted considerable industrial interest. As a result, a small pilot unit to process up to 300 cfm of reverberatory furnace gas was placed in operation in November 1970, jointly by the Bureau of Mines and Magma Copper Co. at the San Manuel smelter in Arizona. Purchased H 2 S was 150 used to precipitate sulfur. Owing in part to hasty procurement and assembly, intermittent operation of the pilot plant over a 6-month period was troubled by failures of the gas cleaning system, pump breakdowns, and plugging of flow lines by precipitated and melted sulfur. Useful data on consumption of citric acid and other reagents were not obtained, but the S0 2 absorption and regeneration system proved readily manageable for removal of 93 to 99 percent of the S0 2 from the smelter gas. Findings of the initial laboratory and pilot unit research were reported in 1970 and 1971 (3> 8). The preliminary research demonstrated that the citrate process is capable of substantially complete removal of S0 2 from industrial waste gases. Use of the clear citrate absorbent liquor prevents scaling or plugging problems. Because of the formation of an oxidation inhibitor in the citrate solution, most of the S0 2 is converted to sulfur with only about 1 percent converted to sulfate regardless of the S0 2 and oxygen content of the feed gas, thus considerably reducing a sodium sulfate purge stream. The circulating citrate solution has a high capacity for short-term overloads of either S0 2 or H 2 S because the citrate acts as a buffer. Simple direct regeneration of the absorbent liquor and precipitation of sulfur is obtained in a single step with¬ out intermediate stripping of S0 2 and separate reduction to sulfur. The process produces an end product of elemental sulfur that can be marketed or readily stored with a minimum of environmental distrubance. The principal questions concerning the process relate to the generation of H 2 S used in the absorbent regeneration-sulfur precipitation step. The cost of H 2 S generation by the sulfur-natural gas-steam reaction is not yet known; the availability of natural gas for this reaction is uncertain, and technology for H 2 S generation from sulfur using reductant other than natural gas has not yet been demonstrated. However, a preliminary cost estimate made by the Environmental Protection Agency indicated that the citrate process might be up to 20 percent less costly than currently available processes such as Wellman-Power Gas and limestone scrubbing (7)• Two pilot plant investigations were undertaken recently to further test the process and obtain useful data for engineering evaluation and cost estimates. This report describes the Bureau of Mines pilot plant operation at The Bunker Hill Co. lead smelter near Kellogg, Idaho, and briefly summarizes the Pfizer-McKee-Peabody pilot plant operation treating stack gas from a coal-fired, steam-generating station in Terre Haute, Ind. Information on H 2 S generation from the Bunker Hill pilot plant is being gathered. This report also describes some laboratory work on the generation of H 2 S by reacting high-sulfur petroleum coke with steam and citrate process sulfur. Other process development research briefly described relates to steam-stripping of the S0 2 -loaded citrate solution as an alternative to H 2 S stripping and 151 elemental sulfur production. Such an alternative would be applicable in areas where a market for sulfuric acid exists because steam stripping produces strong S0 2 that can be converted directly to sulfuric acid. Research with glycolic acid in place of citric acid to produce the buffered S0 2 absorption system also is described. PROCESS DESCRIPTION The citrate process as shown in figure 1 comprises the following steps: 1. The S0 2 -bearing gas is cooled to between 45° and 65° C (113° and 149° F) and cleaned of H 2 S0 4 mist and solid particles. 2. The S0 2 is absorbed from the cooled and cleaned gas by a solution of sodium citrate, citric acid, and sodium thiosulfate. 3. Absorbed S0 2 is reacted with H 2 S at about 65 ° C (l49° F) and atmospheric pressure, thus precipitating elemental sulfur and regenerating the solution for recycle. 4. Sulfur is separated from the solution by oil flotation and melting. 5. The H 2 S for step 3> if not otherwise available, is made by reacting two-thirds of the recovered sulfur with natural gas and steam. CHEMISTRY OF THE PROCESS Absorption of S0 2 in aqueous solution is pH-dependent, increasing at higher pH. Because dissolution of S0 2 forms bisulfite ion with resultant decrease in pH by the following reaction, S0 2 + H 2 0 - HSO 5 + H*, (1) the absorption of S0 2 in aqueous solution is self-limiting. However, by incorporating a buffering agent in the solution to inhibit pH drop, high-S0 2 loadings and substantially complete S0 2 removal from waste gases can be attained. The principal function of the citrate or other carboxylates that have been tested is to serve as a buffering agent during S0 2 absorption. 152 153 FIGURE I-Generalized citrate process flowsheet. The chemistry for the production of sulfur and regeneration of absorbent by reacting H 2 S with the S0 2 in the aqueous solution is complex, but the overall reaction is as follows: S0 2 + 2H 2 S _ 3S° +• 2H 2 0. (2) Actually, thiosulfate and polythionates are found in solution at equilibrium concentrations after several S0 2 -absorption and H 2 S- regeneration cycles. Oxidation of S0 2 in the aqueous solution is sharply depressed by complexing of HSOg and H 4 " from reaction 1 by the thiosulfate ion, according to the following reaction: + HSO 3 + S 2 0| ^ (S0 2 *S 2 0 3 ) = -t- H 2 0. ( 3 ) Reaction 4 shows how this complex might react with H 2 S to form elemental sulfur and thiosulfate ion: (S0 2 -S 2 0 3 ) = +2H 2 S - 3S° + 2H 2 0 + S 2 0 3 . (4) To insure satisfactory operation of the system on startup, sodium thiosulfate is added to the initial absorbing solution. Hydrogen sulfide for regenerating the absorbent and precipitating elemental sulfur can be produced by reacting sulfur with methane and steam as shown in reaction 5 • CH 4 + 4S + 2H 2 0 _ C0 2 +■ 4h 2 s. ( 5 ) Other reducing gases such as hydrogen and carbon monoxide can be used in place of methane. More detailed information on the chemistry of the citrate process is provided in a Bureau of Mines publication ( 9 ) and a paper presented at the American Chemical Society National Meeting in April 1974 (4 ). BUNKER HILL CITRATE PILOT PLANT Nominal capacity of the Bunker Hill pilot plant is 1,000 scfm of 0.5 percent S0 2 gas yielding about 1/3 ton of sulfur per day. Operation of the pilot plant was planned in three phases. Because persistent mechanical failures of the gas cleaning system at the San Manuel copper smelter pilot plant were a principal cause of intermittent operation, Phase I of the Bunker Hill plant was designed to treat cleaned, 4 to 5 percent S0 2 gas diverted from the Lurgi sintering furnace feed to the lead smelter acid plant and diluted with air to 0.5 percent S0 2 . Commercially produced hydrogen sulfide from a tank trailer was used for the sulfur precipitation reaction. In the Phase II operation, an H 2 S generation plant producing 76 to 78 percent H 2 S gas by reacting product sulfur with natural gas and steam is the source of H 2 S. For Phase III, the lead smelter sinter plant tail gas, which contains dust, acid mist, and from 0.3 to about 1 percent S0 2 , is used as pilot plant feed. This sinter tail gas presently passes through a baghouse and then is discharged to the atmosphere through Bunker Hill's main stack. To simulate conventional lead smelter practice, the Phase III operation recovers most of the valuable dust from the tail gas in a baghouse, cools the gas in a packed scrubber, and removes H 2 S0 4 mist and traces of particulate matter with a wet electrostatic precipitator. One of the goals in the pilot plant operation is to determine the minimum gas cleaning requirement compatible with the citrate process. A block diagram of the complete Bunker Hill pilot plant is shown in figure 2. Phase I Operation Figure 3 shows the flowsheet for the Phase I operation. Strong gas containing about 4.5 percent S0 2 from the Bunker Hill Lurgi updraft lead sintering furnace passes through a baghouse, scrubber, and wet electrostatic mist precipitator for removal of particulate matter and H 2 S0 4 mist before entering the lead smelter acid plant. Clean acid plant feed gas for the citrate plant was drawn from either of two connections, one between the mist precipitator and the acid plant drying tower and the other downstream from the drying tower. This gas was diluted tenfold with air to provide 1,000 scfm of gas containing approximately 0.5 percent S0 2 . Humidification of dilution air and the dry gas from the acid plant was required to prevent excessive evaporation of citrate solution in the absorption tower. During humidification in the packed tower, the cool air-gas mixture was heated to between 45° and 65° C (113° and 149° F) to correspond with the temperature range to which the lead sintering furnace tail gas would have to be cooled before treatment in the citrate plant. The gas stream then passes upward through a 2.5-foot-diameter by 30-foot-high Fiberglass/ reinforced polyester (FRP) packed absorption tower countercurrent to the citrate solution, in which over 95 percent of the S0 2 is absorbed. The absorption tower contains three 6-foot sections packed with 1-inch polypropylene Intalox saddles and a stainless steel mist eliminator. Other gas-liquid contacting techniques may be applicable, but nearly all of the authors' experience in the laboratory has been with packed towers. From the absorption tower, the citrate solution flows by gravity to closed, stirred vessels for reaction with 2/3 ton of H 2 S per day to form 1 ton per day of elemental sulfur. Three 100-gallon stainless 17 Reference to specific trade names is made for identification only and does not imply endorsement by the Bureau of Mines. 155 2.5’ft-diam by 18-ft-high packed scrubber tower --► Particulate sludge to £ > o o S S! c a> o o> CO OJ I c 0 ) o E ro N OJ O' jl) c O m >* ° X3 O a> 0 0 c a> 0 w 0 0 O' 0 1 cr O m tO 32 W o ^ E cO a> O 1 | ro . \ CO O o 156 FIGURE 2-Bunker Hill pilot plant. c 3 CD i ro L±J cr Z> o Li- 157 steel reactor vessels arranged for countercurrent flow of citrate solution and H 2 S gas are available for the sulfur precipitation step. During Phase I operation when nearly pure commercial-grade H 2 S was used, the 10-minute retention time available in one reactor usually was sufficient for the sulfur precipitation step. When the H 2 S-C0 2 mixed gas prepared from recycled sulfur is used, as in the Phase II operation, three reactors probably will be necessary to insure adequate retention time for contact of the gas and loaded citrate solution. A slurry containing 1- to 3-percent solids of elemental sulfur in the regenerated citrate solution overflows the reactors and passes through a common header to a stainless steel reactor effluent tank. From this tank, the dilute slurry is pimped to a 100-gallon FRP conditioner tank where kerosine or other hydrocarbon oil is added for the sulfur flotation-separation step. The oil-conditioned slurry flows to a specially-designed sulfur flotation skimming device that resembles an Esperanza drag classifier. The sulfur separates from the citrate solution by floating to the surface as a 35- to 45-percent-solids product, leaving clear regenerated citrate for recycle to the absorption tower. The sulfur is skimmed off the surface of the citrate solution, pulled up an inclined chute, and discharged to a conical storage bin. Regenerated citrate solution from the feed well of the sulfur skimmer passes to a 300-gallon FRP absorber feed tank. Citrate solution from the absorber feed tank is pumped through parallel backwash clarification filters and a water-cooled heat exchanger to the absorption tower. On day shift only, the sulfur float product is withdrawn from the storage bin and pumped by a Moyno positive displacement auger-type pump through a single-tube, steam-jacketed heat exchanger where the sulfur is melted at about 135° C (275° F). Molten sulfur and citrate solution pass into a closed settler tank at 135° C under a pressure of 35 psig. Part of the molten sulfur is tapped from the bottom of the autoclave settler and cast in 100-pound blocks. The Phase II operation requires a bleed stream of molten sulfur to flow to the H 2 S generating plant. Citrate solution and the oil used for flotation are withdrawn from the top of the settler and then pass through a sulfur knockout pot and a water-cooled heat exchanger into a decanting vessel for separation and reuse. Citrate solution from this tank drains to the absorber feed tank. An incinerator is provided outside the main building to burn H 2 S vented from the system or released under emergency or upset conditions. In a commercial plant, the offgas from the incinerator would be returned to the gas stream entering the absorber. The pilot plant is completely instrumented and controlled from a panel mounted in a 40-foot-long instrument trailer that is connected to a pilot plant building and also serves as an onsite laboratory. 158 Operation of the Bunker Hill citrate pilot plant was started, on February 15, 1974. The plant operated for a total of 1,900 hours through December and produced about 66 tons of bright yellow, high- quality sulfur. Because of interruptions resulting from mechanical failures, unavailability of feed gas, and changing work crews, the longest continuous operation was about 160 hours. Citrate loss over this period was 7*5 pounds per net long ton of sulfur recovered from feed gas. Sulfur dioxide removal from feed gas containing 0.3 to 0.5 percent S0 2 ranged from 96 to 99 percent when operating at the design gas flow rate with varying gas temperature and citrate concentration. The precipitated sulfur was successfully recovered as a high-purity product by oil flotation and melting. Table 1 summarizes results obtained under reasonably steady-state continuous operation at a gas flow rate of 1,000 scfm and S0 2 concentrations of 0.3 to 0.5 percent. Gas temperatures ranged from 35° to 65° C (95° to 149° F). Citrate solution flow rate in the absorber was 10 to 11 gallons per minute, citrate concentration was 0.5 M for most of the tests, the sodium-to-citric acid molar ratio in the citrate solution was 2:1, and the pH of the citrate feed solution to the absorption tower was about 4.5. TABLE 1. - Results of Bunker Hill citrate pilot plant operation, February-December 1974 Feed gas concentration, pet S0 2 Gas temperature, °F Feed Exhaust Citrate solution loading, fi/1 so 2 Offgas concen¬ tration, ppm S0 P so 2 removal, pet 0.32 85 95 6 -5l/ 52 99.0 .41 100 100 8 . 81 / 57 98.6 .47 114 123 9.8 28 99-4 .49 131 155 10.1 51 98.9 .47 148 149 7.5 162 96.5 T/ 0.25 M citrate solution. The test results show that under the design conditions (1,000 scfm of 0.5 percent gas at 114° F and 10 gallons per minute of 0.5 M citrate solution), an S0 2 removal efficiency of 99-4 percent and an offgas containing less than 30 ppm S0 2 can be obtained. The solution loading of 9*8 grams S0 2 per liter represents 55 percent of the maximum equilibrium loading of 0.5 M citrate solution at the feed gas temperature of 114° F. The test results also show that the S0 2 removal efficiency exceeded 96 percent and the offgas contained less than 200 ppm S0 2 even when the feed gas temperature was increased to 65 ° C (149° F). Excellent S0 2 absorption was obtained with the more dilute citrate solution at the lower temperatures, even though the solution loading represents 65 percent of the maximum loading of the 159 0.25 M citrate solution. At the design capacity of the plant, a gas flow rate of 1,000 scfm and a solution flow rate of 10 gallons per minute, the total pressure drop through the absorption tower was 6 inches of water. During continuous campaigns in late summer 1974, the pH of the recycled citrate liquor started dropping from the desired 4.5, and S0 2 absorption efficiency subsequently decreased. As the pH dropped from 4.5 to about 4.0, the S0 2 removal efficiency at the design gas flow of 1,000 scfm of 0.5 percent S0 2 feed gas at 45° C(ll3° F) dropped from 99 to 85 percent. During this time, offgas from the pilot plant increased from about 30 ppm S0 2 to 750 ppm. Laboratory tests on samples of the pilot plant citrate solution showed that a combination of low pH, high polythionate content, and low thiosulfate content resulted from incomplete regeneration of the absorbent liquor. The principal cause of incomplete regeneration was found to be insufficient retention time of H 2 S gas in the single sulfur precipitation reactor. Apparently, slightly more contact time than that provided in one reactor was necessary for effective utilization of the H 2 S. If the H 2 S-S0 2 reaction for producing elemental sulfur is not allowed suffi¬ cient time, more polythionates, principally in the form of polythionic acids, are produced permitting the pH of the solution to drop. Low solution pH causes two problems, the more critical of which is a drop in S0 2 absorption efficiency. In addition, thermal decomposition of thiosulfate in the sulfur melting system accelerates at pH 4 or below, thus increasing the sulfate concentration of the solution. In practice, this would require additional sodium carbonate for neutrali¬ zation and purging of the additionally formed sodium sulfate from the system. The problem of decreasing S0 2 absorption in the pilot plant was solved by (l) increasing the contact time of H 2 S with the loaded liquor by using two reactors with countercurrent flow of H 2 S and citrate solution, (2) adding sodium thiosulfate to bring the concentration back up to the desired level, and ( 3 ) adding sodium carbonate to neutralize the sulfate in the plant liquor. These measures were successful in regenerating the pH at 4.5 and restoring high S0 2 absorption capability to the solution. In current campaigns, the indicators of incomplete regeneration (low pH, high polythionates, and low thiosulfate) were closely monitored. During precipitation of sulfur with the commercially produced, nearly pure H 2 S, sulfur buildup along the walls of the stainless steel reactors or on the impellors was minimal when the tip speed of the impellors was at least 900 feet per minute. In the early stages of plant operation, excess H 2 S absorbed in the citrate solution resulted in cloudy recycle 160 solution recovered from the kerosine flotation step, apparently due to delayed precipitation of colloidal sulfur. Some of the absorbed H 2 S escaped at times from the sulfur skimmer into the plant building; this was corrected by using a second stirred reactor as a delay tank to allow more contact time with the H 2 S and bypassing about 5 volume - percent of the S0 2 -loaded liquor for the absorption tower to the reactor effluent tank, which reacted with the excess absorbed H 2 S. These measures have resulted in consistently clear citrate solution from sulfur flotation for recycling to the absorption tower and have stopped the escape of H 2 S from the sulfur flotation unit. Few problems have been encountered in plugging of lines between reactors and the reactor effluent tank. A problem existed with sulfur buildup in the automatic level-controlled reactor effluent tank until a small agitator was installed to keep the sulfur in suspension. The kerosine conditioner tank has operated well with no sulfur buildup at the design liquid flow rate when the impellor tip speed is at least 700 feet per minute. Initially, some trouble was experienced with holdup of floated sulfur in the freeboard of the tank. This required an occasional cleanout of the 3-inch-diameter overflow line to the skimmer. However, the addition of a second impellor operating just beneath the liquid surface prevented buildup of the large lumps of powdery floated sulfur that were blocking the overflow line. A powder-like sulfur product of about 50 percent solids has been obtained by adding between 35 to 40 pounds of kerosine per ton of sulfur floated. About 20 percent of the kerosine added for flotation has been recovered from the melting operation for reuse. Most of the kerosine loss can be attributed to volatilization from the hot sulfur slurry in the kerosine conditioner and sulfur skimmer. Because of this high volatilization loss, kerosine consumption during the pilot plant operation to date has averaged 90 pounds per net ton of sulfur recovered from the feed gas. Laboratory tests have indicated that this hydro¬ carbon consumption can be reduced considerably by using motor oil in place of kerosine. In a 20-scfm continuous test plant at the Salt Lake City Metallurgy Research Center, the use of SAE 10 motor oil resulted in a sulfur float product equivalent to that produced with kerosine, and the oil consumption was one-fourth that of kerosine. Various oils will be investigated in future campaigns at the Bunker Hill pilot plant. The sulfur melting step has functioned satisfactorily at the design capacity. Some plugging problems have been encountered in the citrate liquid lines from the autoclave settler, apparently because of sulfur being dissolved in the kerosine flotation reagent and then crystal¬ lizing out upon cooling. Possibly a dual filter downstream of the liquid cooler or substitution of motor oil for kerosine will solve this problem. l6l The sulfur produced by the Bunker Hill citrate pilot plant has been bright yellow and of better than 99*5-percent purity. Carbon content has ranged from 0.2 to 0.3 percent. In the laboratory, continuous test plant sulfur recovered by motor oil flotation contained 0.01 percent carbon. During operation of the Bunker Hill citrate pilot plant, the rate of oxidation of S0 2 to S0| was determined to be about 1.3 percent. This is quite low considering that the feed gas, which is predominately air, contained about 20 percent 0 2 . This amount of oxidation requires a sodium carbonate addition of about 90 pounds per ton of sulfur recovered from the gas. To date, the highest sulfate concentration in the plant has been about 50 grams per liter. However, this figure is not representative of the greater sulfate buildup expected because of a' few plant upsets resulting in large solution losses, which required a makeup of fresh citrate solution. These losses occurred when the reactor effluent tank or kerosine conditioner tank plugged causing citrate solution to back up and flow through vent lines to the H 2 S incinerator. The modifications to the reactor effluent tank and kerosine conditioner have considerably reduced solution backup through the vent lines. In addition, collection tanks have been installed in the vent lines to the incinerator whenever plugging problems do occur. In future tests with solution losses minimized, it will be possible to determine the maximum concentration of sulfate that can be maintained in the recirculating citrate solution before sodium sulfate removal is necessary. Papers providing more detail on the design and preliminary operation of Phase I section of the Bunker Hill pilot plant were presented at the 1974 AIME Annual Meeting in Dallas, Tex., and the 1974 EPA- sponsored Flue Gas Desulfurization Symposium in Atlanta, Ga. (_5, 6). Phase II Operation In Phase II of the Bunker Hill pilot plant operation, the H 2 S generation plant will be operated utilizing both a one-stage and two-stage procedure for production of H 2 S by the sulfur-methane-steam reaction to provide data for engineering evaluation of this important step in the citrate process. The gas produced should contain 76 to 78 percent H 2 S on a dry basis with most of the remainder being C0 2 . The H 2 S generation plant also will be operated in conjunction with the sulfur dioxide absorption and sulfur recovery section to determine the influence of the impure H 2 S gas on sulfur precipitation and S0 2 removal efficiency. The rated capacity of the generation plant is 1.25 short tons of H 2 S per day. A flow diagram for the H 2 S generation section is shown in figure 4. Molten sulfur from the absorption and sulfur recovery section of the 162 Molten sulfur from Natural gas Water -Flow during single-stage reaction ^2^ ^2 Product gas to citrate plant sulfur precipitation reactors FIGURE 4.-Bunker Hill citrate pilot plant-H 2 S generation section. 163 pilot plant is transferred to a sulfur feed tank from which it is pumped through a filter to remove impurities and then to a gas-fired superheater. The sulfur is vaporized and superheated to about 730° C (1,350° F). Natural gas containing about 92 percent methane and 6 percent heavier hydrocarbons heated to 650° C (1,200° F) in another gas-fired superheater joins with the hot sulfur vapor ahead of reactor 1. The first stage of the sulfur-methane-steam reaction to produce H 2 S takes place in the presence of a catalyst in this reactor according to the following equation: CH 4 + kS - CS 2 + 2H 2 S. ( 6 ) The H 2 S-CS 2 reactor product gas is first air cooled to about 315° C (600° F) and then further cooled in a steam heat exchanger to about 150° C (300° F). Excess sulfur is condensed in the reactor product cooler and is removed from the gas mixture in the first sulfur knockout vessel. Steam is superheated to k2^° C (800° F) in a gas-fired steam superheater and then blended with the cooled H 2 S-CS 2 gas prior to entering reactor 2. In this reactor, all of the CS 2 is hydrolyzed to H 2 S and C0 2 in a catalyst bed at 370° C (700° F) according to the following reaction: (7) CS 2 + 2H 2 0 - C0 2 + 2H 2 S. A steam-cooled heat exchanger cools the reactor product gas to 150° C (300° F) to condense any free sulfur. The condensed sulfur is removed in two additional knockout tanks. Sulfur collected in all three knockout vessels is periodically drained to the sulfur feed tank. Product gas from the second reactor containing H 2 S, C0 2 , and some water vapor is cooled to about 65° C (150° F) in a water heat exchanger and then flows to the S0 2 absorption and sulfur recovery section of the citrate pilot plant. Provision has been made in the H 2 S generation section to feed super¬ heated steam along with sulfur vapor and high-temperature natural gas to the first reactor, thus bypassing the second reactor and carrying out the entire reaction in one stage according to equation 5* This has been demonstrated successfully in a continuous single-stage H 2 S generator at the Salt Lake City Metallurgy Research Center. This unit has an H 2 S production capacity of 0.7 scfm and is used in conjunction with the 20-scfm continuous citrate test plant. The H 2 S generation section of the Bunker Hill citrate pilot plant was started up in late September 197^-. Initially, some problems were encountered in maintaining temperatures high enough for the reaction 16U of sulfur, natural gas, and steam to produce H 2 S. In addition, much of the first stage of the reaction to produce CS 2 and H 2 S took place prematurely upstream of the reaction vessel. In spite of these problems, adequate temperatures were attained over a 24-hour period and 78 percent H 2 S gas was produced while the plant was operated at 1.5 tons H 2 S per day or 125 percent of the design capacity. The H 2 S generator was .shut down in early October for modification to reduce heat losses. Additional insulation was added to piping and vessels, and reactor 1 was moved closer to the sulfur and natural gas superheaters to utilize the heat of the exothermic reaction. The plant was restarted, and successful turndown to about 0.6 ton H 2 S per day was achieved. The product gas from the first reaction between sulfur and natural gas to form H 2 S and CS 2 contained, in volume- percent on a dry basis, 66 H 2 S, 26 CS 2 , 4.7 C0 2 with the remainder consisting of CE 4 , COS, and CO. After hydrolysis of the CS 2 to H 2 S in the second reactor, the final product gas contained, in percent 78.3 H 2 S, 19*5 C0 2 , 0.8 CS 2 , 0.6 COS and no CH 4 . Although the combined CS 2 and COS was slightly higher than the 1.2 percent hoped for, the product gas should be suitable for sulfur precipitation. When the citrate pilot plant is treating 1,000 scfm of 0.5 percent S0 2 feed gas, 1/3 ton sulfur per day is removed and this requires 2/3 ton H 2 S per day for the sulfur precipitation step. Although the H 2 S generator operated successfully at this production rate, further turndown is necessary to accommodate lower strength S0 2 feed gas. In November 1974, when continued turndown of the H 2 S production rate to a specified minimum 0.4 ton per day was attempted, the sulfur vaporizer plugged and the plant was shut down. Preliminary sampling of the plug material indicated a corrosion product consisting of iron, nickel, and chromium sulfides caused by sulfidization of the stainless steel. This corrosion was believed to be caused by excessively high tempera¬ tures in the sulfur vaporizer during the initial startup--not by operation at the low production rate. The high temperatures were attained because of incorrectly adjusted temperature transmitters. In late December, the plug in the sulfur vaporizer coil was finally removed after several weeks of mechanical reaming with a drill and flexible coupling. During initial operation of the H 2 S generator when the sulfur vaporizer operated at excessively high temperatures, some sagging of the coil occurred. Refractory bricks were placed inside the superheater at several points to support the coil. However, these bricks deflected the burner flame to several points along the surface of the coil. When the H 2 S generator was restarted in January of this year, hot spots developed at these points already weakened by the mechanical reaming, and the coil burned out. A new vaporizer coil nas been purchased, and H 2 S generator operation is expected to resume in May. 165 Phase III Operation As previously mentioned and shown in the block diagram of the complete Bunker Hill citrate pilot plant (figure 2), the gas cooling and cleaning section used for the Phase III operation treats lead smelter sinter plant tail gas containing 0.3 to 1 percent S0 2 . Flue dust is removed in a baghouse, the gas is cooled in a packed tower where dust not collected in the baghouse is removed as a sludge, and S0 3 is removed in a wet electrostatic mist precipitator as sulfuric acid mist. The gas cooling and cleaning section was started up in late February 1975* Shakedown runs made over a period of 5 days demonstrated that the various units of the system functioned satisfactorily. The dust loading of the tail gas from the Lurgi sintering furnace averaged 3 grains per cubic foot over one 60 -hour run and most of this was removed in the baghouse. In these preliminary runs, the Lurgi tail gas feed to the citrate plant analyzed only 0.4 milligram S0 3 per standard cubic foot instead of the nearly 10 milligrams it was purported to contain. Although this low S0 3 content is equal to the guaranteed S0 3 outlet of the electrostatic mist precipitator, the precipitator removed about two-thirds of the S0 3 from the Lurgi tail gas leaving only 0.15 milligram S0 3 per standard cubic foot of feed gas going to the S0 2 absorption section of the plant. After the shakedown runs, the gas cooling and cleaning section was successfully integrated with the S0 2 absorption and sulfur recovery section. The plant was operated for about 200 hours at the design capacity of 1,000 scfm of Lurgi tail gas averaging 0.5 percent S0 2 . This operation substantiated prior results obtained when treating diluted acid plant feed gas from the Lurgi sintering furnace. With an absorption tower temperature of 35° C (95° F), 99 percent of the S0 2 was absorbed in 0.5 M citrate solution at liquid loadings of 10 to 11 grams S0 2 per liter. Operation of the gas cooling and cleaning section to determine gas cleaning requirements will resume after the H 2 S generation section has been brought on stream. THE TERRE HAUTE CITRATE PILOT PLANT The Ffizer-McKee-Peabody citrate pilot plant at Terre Haute, Ind., is described in three publications (l, 4, 10). Briefly, the skid- mounted unit treated 2,000 scfm of gas from a coal-fired spreader stoker-type boiler with a 25 , 000 -pound-per-hour stream-rated capacity. The flue gas,' containing 0.1 to 0.2 percent S0 2 , was adiabatically cooled with quench water and passed into a Venturi-type water scrubber to remove fly ash. No other gas cleaning was done. Absorption of S0 2 took place in an impingement plate scrubbing tower using an aqueous solution of sodium citrate and citric acid. The S0 2 -rich 166 citrate solution flowed from the absorber through a steam-heated heat exchanger to a three-stage continuous stirred tank reactor system countercurrent to a flow of pure H 2 S from a tank trailer. The S0 2 was reduced to sulfur, and the citrate solution was regenerated. Sulfur slurry was pumped from the reactors to a surge tank and then to a sulfur flotation separation system. Citrate liquor was recycled from the flotation unit to the absorption system. The sulfur flotation product was pumped as a slurry through a heater at a temperature above 125° C (257° F) and a pressure of 35 psig to melt the sulfur. Liquid phases were separated in a pressure decanter where the bottom layer was drawn off as high-quality molten yellow sulfur and the citrate solution top layer was discharged to a flash drum at reduced pressure. Although generally similar to the Bureau of Mines' Kellogg, Idaho, pilot plant, there were two major design differences in the Terre Haute plant. An impingement plate tower was used rather than a packed tower. The sulfur separation was based on a flotation principle, but no hydrocarbon addition was made. In a paper presented in December 197*+ at the American Institute of Chemical Engineers Annual Meeting in Washington, D.C., Srini Vasan of Peabody Engineered Systems reported on the Terre Haute pilot plant operation (10). Under the final equipment configuration, the pilot plant operated from March 15 to September 1, 197 *+, logging 2,300 total operating hours. The longest sustained run was 180 hours. Results of the 5-l/2-month operation are summarized in table 2. TABLE 2. - Summarized results of Terre Haute citrate pilot ~plant operation, March - September 197*+ Gas flow Feed gas Gas Offgas S0 2 rate, c one ent ration, temperature, concentration, removal, scfm pet S0 2 0 F ppm S0 2 pet 2,000 0.10-0.20 120-140 25-50 95-97 The test results show that an S0 2 removal efficiency of at least 95 percent and offgas containing 50 ppm S0 2 or less were obtained during long-term operation of the pilot plant. The tray scrubber operated during the entire run without any malfunction or lowered efficiencies. Pressure drop through the 5-tray, 2.5-foot-diameter by 10-foot-high absorber was 8 inches of water. The trouble-free operation of the absorber was highlighted during one phase of the program. Although clear liquor normally returned to the absorber from the sulfur- separation step, for a period of weeks a 1-percent sulfur slurry was deliberately returned. No problems were encountered. Tests also showed, that the circulating liquor has a high capacity for short-term overloads of S0 2 or H 2 S, thus eliminating the need for precise instantaneous control of H 2 S feed rate to match variations of S0 2 content of the flue gas. As a result, operational control was simplified., and. the plant operator performed routine checks for pH and. thiosulfate levels only for monitoring purposes. Sulfur produced at the Terre Haute pilot plant analyzed. 99*96 percent sulfur content with 0.03 percent carbon. The low carbon content is attributed, to the lack of hydrocarbon addition for the flotation- separation step of the process. Vasan's paper also provides a comparison of the economic and process features of the citrate process with a wet limestone scrubbing system for treatment of flue gas from a 200-MW powerplant burning 3-5 percent sulfur coal. This comparison is summarized in table 3* TABLE 3• - Comparison of citrate process and limestone scrubbing for S0 2 removal from 200-MW coal-fired boiler offgas Citrate process Limestone scrubbing SOo removal. 95 90 Offgas concentration. Investment costs (battery 25-50 200-300 limits--November 1974).. 40-50 60-70 Operating costs. 2 - 2.5 2.5-3.0 Operating costs. .dollars/ton coal.. 7 8 This economic evaluation indicates that the annual operating cost of a citrate process plant could be about 15 percent less than those of a wet limestone scrubbing plant. PROCESS DEVELOPMENT Research continues at the Salt Lake City Metallurgy Research Center to develop and improve the buffered S0 2 -H 2 S (citrate) process for removing S0 2 from industrial stack gases. Three areas of this process development research that look encouraging are (l) H 2 S generation using high-sulfur petroleum coke in place of natural gas, (2) steam stripping, as an alternative to H 2 S stripping, to produce strong S0 2 gas for feed to a sulfuric acid plant, and ( 3 ) substitution of glycolic acid for citric acid as the buffering agent in the buffered S0 2 -H 2 S process. 168 H 2 S Generation Using Petroleum Coke Research has been initiated to investigate generation of H 2 S for the buffered S0 2 -H 2 S process by reacting a 5-percent sulfur petroleum coke with steam and citrate-produced sulfur. The manufacture of H 2 S from this raw material, typical of high-sulfur coke produced at oil refineries, would eliminate the dependence of the citrate process on natural gas. Utilization of this high-sulfur petroleum coke, which cannot be burned in many areas because of air pollution standards, would help in solving some of the stockpiling problems presently developing at many oil refineries. Initial research has been directed toward a two-stage procedure, whereby the coke is first reacted with steam in a vertical tube furnace at 800° to 900° C to produce a water gas containing predom¬ inantly H 2 and CO according to the following reaction: C + H 2 0 _ H 2 + CO. (8) The water gas is then reacted with vaporized sulfur and more steam in the horizontal tube reactor of the single-stage laboratory H 2 S generator where H 2 S and C0 2 is produced by the following reaction: H 2 + CO + 2S + H 2 0 _ C0 2 * 2H 2 S. (9) In the best test to date, reaction of the sulfur-bearing coke and steam at 860° C (1,580° F) resulted in a gas containing, in percent on a dry basis, 52 H 2 , 42 CO, 4 C0 2 , 2 H 2 S, and a trace of CH4. Reaction of this gas with sulfur vapor and more steam over an alumina catalyst at 550° C (1,020° F) in the laboratory H 2 S generator produced a gas containing, in percent, 68 H 2 S, 31 C0 2 , and the remainder CO, COS, and CH 4 . Such a gas would be suitable for precipitation of elemental sulfur and regeneration of absorbent liquor in the citrate process. Steam Stripping Investigations have been initiated to determine the feasibility of utilizing steam stripping in the buffered S0 2 -H 2 S process as an alternative to H 2 S stripping for removing S0 2 and regenerating the absorbent liquor. Strong S0 2 gas would be recovered as an end product in place of elemental sulfur. Such a procedure might be applicable where markets existed for S0 2 or H 2 S0 4 , or where acid plant tail gas could be reduced to acceptable levels by the citrate process and the strong S0 2 gas produced could be recycled to the acid plant feed gas stream. 169 Continuous tests have been initiated in a bench-scale, integrated absorption-stripping apparatus to treat gases containing 0.25 and 0.5 percent S0 2 . This covers the S0 2 concentration range normally found in acid plant tail gas. Absorption of S0 2 is carried out at 45° to 50° C using a 0.5 M citrate solution at pH 4. The steam stripping column is operated at a temperature of 9^° C. Preliminary tests indicated high steam consumptions, but with heat losses in the system reduced by added insulation, near isothermal conditions maintained in absorption and stripping columns, and improved utilization of the steam, better results are being obtained. In the best test to date, 95 percent of the S0 2 was removed from feed gas using about 12 pounds of steam per pound of S0 2 recovered. Sodium thiosulfate is added to the citrate solution to inhibit oxidation of S0 2 to SO 4 . To date, S0 2 oxidation in the system has been about 1.3 percent, which is comparable with the oxidation rate obtained when H 2 S is used for regeneration of citrate solutions. The results of Bureau investigations indicate that steam stripping might be an alternative to H 2 S regeneration in the citrate process when extremely high absorption is not needed and inexpensive steam is available. Glycolate System for S0 2 Absorption In the early stages of Bureau reasearch to develop the buffered S0 2 -H 2 S process for removal of S0 2 from stack gases, laboratory tests indicated that other carboxylate solutions could be substituted for citrate solution as the buffering agent. Citric acid was initially chosen for development of the process because of its chemical stability, low vapor pressure, good pH buffering capacity, and the purity and physical character of the precipitated sulfur. Other carboxylates tested appeared to have about the same properties as citric acid and costless in some cases, but in the closed system projected, with slight solution or decomposition loss, the carboxylate price seemed relatively unimportant. However, if reagent losses were to become excessive and the price of citric acid remains high, the use of less costly carboxylates might become worthwhile. One of the more promising of the other carboxylate systems investi¬ gated was a solution of sodium glycolate, glycolic acid, and sodium thiosulfate. Glycolic acid is one-third the molecular weight and less than half the cost of the citric acid. However, citric acid is nontoxic and is, in fact, used in many food products, but the biological effects of glycolic acid are not well understood. 170 The 20-scfm continuous test plant at the Salt Lake City Metallurgy Research Center has been operated with glycolate and citrate solutions to compare absorption efficiencies and to test various H 2 S precipita¬ tion reactor configurations. Regardless of the method used for H 2 S precipitation and solution regeneration, no significant difference was found in absorption efficiency, reaction rate, or quality of sulfur between citrate and the lower molecular weight glycolate.when compared on the basis of weight of carboxylate in the absorbent liquor, i.e., i«5 M glycolate solution compared with 0.5 M citrate solution. DEMONSTRATION PLANT Plans are underway for a large-scale plant to demonstrate the citrate process for S0 2 emission control at a 75 to 100-MW electric generating facility burning high sulfur coal. The demonstration plant will be provided and operated under a cooperative arrangement and cost-sharing basis between the Bureau of Mines and the Environmental Protection Agency and interested industrial firms. Proposals for the demonstration plant, including preliminary engineering estimates, are to be submitted by late October 1975- A contract will be awarded, after negotiations, probably in March 1976. Work on the citrate demonstration plant contract will be divided into four phases. Phase I, consisting of process design and a definitive cost estimate, should be completed by the fall of 197o. Phase II, which includes detailed engineering design, construction, and mechanical acceptance of the plant, should be complete! by the end of 1978. Phase III, consisting of startup and performance acceptance testing, would take place at the conclusion of ihase II. This would be followed by Phase IV, which consists of a comprehensive emission testing program to be conducted at the demonstration plant by an independent contractor for 1 year. 171 REFERENCES 1. Chalmers, Frank S., Louis Korosy, and A. Saleem. The Citrate Process to Convert S0 2 to Elemental Sulfur. Pres, at Industrial Fuel Conf., Purdue University, West Lafayette, Ind., Oct. 3? 1973? 6 pp. (Available upon request from Arthur G. McKee & Co., Cleveland, Ohio.) 2. Elliot, T. C. S0 2 Removal From Stack Gases, A Special Report. Power, v. 118, No. 9? September 1974, PP- 5-24. 3. George, D. R., Laird Crocker, and J. B. Rosenbaum. The Recovery of Elemental Sulfur From Base Metal Smelters. Min. Eng., v. 22, No. 1, January 1970, pp. 75-77. 4. Korosy, L., H. L. Gewanter, F. S. Chalmers, and Srini Vasan. Chemistry of S0 2 Absorption and Conversion to Sulfur by the Citrate Process - . Pres, at 167th ACS Meeting, Los Angeles, Calif., Apr. 5? 1974, 32 pp. (Available upon request from Pfizer, Inc., New York.) 5. McKinney, W. A., W. I. Nissen, and J. B. Rosenbaum. Design and Testing of a Pilot Plant for S0 2 Removal From Smelter Gas. Pres, at AIME Ann. Meeting, Dallas, Tex., Feb. 23-28, 1974, AIME Preprint A-74-85, 12 pp. 6. McKinney, W. A., W. I. Nissen, D. A. Elkins, and J. B. Rosenbaum. Pilot Plant Testing of the Citrate Process for S0 2 Emission Control. Pres, at Flue Gas Desulfurization Symp., Environmental Protection Agency, Atlanta, Ga., Nov. 4-7, 1974, 19 PP- (Available upon request from the Salt Lake City Metallurgy Research Center, Salt Lake City, Utah.) 7- Rochell, G. T. Economics of Flue Gas Desulfurization. Proceedings: Flue Gas Desulfurization Symposium--1973• Office of Research and Development, National Environmental Research Center, U.S. Environmental Protection Agency, Research Triangle Park, N.C., December 1973? PP- 103-132. 8. Rosenbaum, J. B., D. R. George, and Laird Crocker. The Citrate Process for Removing S0 2 and Recovering Sulfur From Waste Gases. Pres, at AIME Environmental Quality Conf., Washington, D.C., June 7-9? 1971? 26 pp. (Available upon request from the Salt Lake City Metallurgy Research Center, Salt Lake City, Utah.) 9- Rosenbaum, J. B., W. A. McKinney, H. R. Beard, Laird Crocker, and W. I. Nissen. Sulfur Dioxide Emission Control by Hydrogen Sulfide Reaction in Aqueous Solution - The Citrate System. BuMines RI 7774, 1973? 31 PP- 10. Vasan, Srini. The Citrex Process for S0 2 Removal. Chemical Engineering Progress, v. 71, No. 5, May 1975, PP- 61-65. 172 ELECTROSTATIC COLLECTION OF FLY ASH FROM WESTERN COALS: SOME SPECIAL PROBLEMS AND THE APPROACH TO THEIR SOLUTION by Grady B. Nichols and Roy E. Bickelhaupt* I. INTRODUCTION The electrostatic precipitator is the primary air pollution control device for removing particulate material from effluent gas streams from coal-fired power boilers. The operation of this device is dependent upon three steps in the process of collection: particle charging, particle collection, and the removal and disposal of the collected material. These three steps must be performed at near optimum conditions for the effi¬ cient operation of the device. The theory of electrostatic precipitation is covered in detail in a number of publications, and will not be developed in this presentation. However, this theory does directly apply to the development and will be freely called on. Particle Charging The electrical charging function is dependent upon two variables directly associated with the precipitation - the electri field at the point of charging and the free ion density in the neighborhood of the particle. The charging function is also related to the physical characteristics of the dust particle. These items combine to yield an expression for the charge as a function of time for field dependent charging as shown in equation 1 q = 127T£ 0 a 2 —E 0 _L_ (1) 0 e+2 t+x and the charging time constant is related to the free ion density by 4 e T ~ n 0 ey *Grady B. Nichols Head, Environmental Engineering Div. Dr. Roy E. Bickelhaupt Head, Ceramics Section Southern Research Institute 2000 Ninth Avenue South Birmingham, Alabama 173 The primary point is that the charge is proportional to the electric field and the charging rate is proportional to the free ion density. Particle Collection The particle collection function is directly related to the electrical force that results from the action of a charged particle in the presence of an electric field. This force is opposed by the viscous drag force of the gas stream to yield a term referred to as the electrical migration velocity as given by W = 67rari which if we use the charge shown in equation 1 we have Ta7 _ 2aE n E e 0 e t W “ 0 P 0 _ _ (2) n e+2 t+x The fundamental point here is that the electrical migration velocity is proportional to the electric field where the particle was charged (E 0 ) and the electric field at the plate where it is actually collected (Ep). The collection function of the precipitator is related to the electrical migration velocity for each particle size as shown above and to the remixing caused by the turbulent gas flow conditions in the precipitator. These factors combine to yield the familiar Deutsch-Andersen efficiency equation given in almost all precipitation texts. n = 1 - exp - (A w) (3) V This equation specifically applies to the collection efficiency for a group of particles with a known and constant electrical migration velocity. Equation 2 above clearly shows that from theoretical considerations this migration velocity is directly proportional to particle size. Particle Removal The particle removal is a two step process. First the particles must be dislodged from the plate where they will hopefully fall into the collection hopper and at some time later they will be removed from the hopper for disposal. The particles are retained on the collection electrode by a number of forces: Van der Waals, mechanical and electrical. The rapping function must overcome these forces at some point in the dust layer to provide adequate removal. The rapping function in a precipitator is at this point in time, handled as an empirical factor rather than one with sound theoretical support. In general, the rapping requirement is such that the acceleration imparted to the dust-collection plate combination be sufficient to just dislodge the dust layer without actually breaking it into a powder, thus inhibiting excessive reentrainment. These items form the basis for electrostatic precipitator behavior, and so long as the particulate matter exhibits characteristics that are "normal", efficient collection will result. The desired precipitator behavior is typically associated with applied voltages on the order of 40 to 50 kilovolts and current densities of about 40 microamperes per square foot, resulting in a power density of about two watts per square foot. These are typical upper limit values for a well designed and installed precipitator operating with intermediate resistivity (10 10 ohm-cm) fly ash. II. NORMAL PRECIPITATOR BEHAVIOR If we select the above operating conditions and apply them to a precipitator collecting a representative fly ash, utilize a computer systems analysis approach to apply the electrical conditions to each particle, and then integrate over the particle size range to predict the collection area, we predict the performance shown in Figure 1, Curve a. This curve represents the type of behavior expected with no problems such as high resistivity, misalignment, poor gas flow, excessive reentrainment, or others. III. REDUCED COLLECTION EFFICIENCY WITH LOW CURRENT - LOW VOLTAGE OPERATION If we repeat the computer analysis for the same dust conditions used above but for a current density of three and one half microamps per square foot and an applied voltage of 30 kilovolts, we find that the predicted behavior drops from the previous value to that shown by Curve d, Figure 1. The point of reference here is that if the normal precipitator achieved an efficiency of 99.5%, the one operating with reduced voltage and the same collection electrode area would only attain an efficiency of 94%, with an accompanying error in emission of a factor of twelve. (Intermediate values of current density are also shown in Curves b and c). The change in resistivity that results from the combustion of low sulfur Western coals can cause this order of magnitude change in the behavior of a precipitator. 175 Collection Efficiency Specific Collection Electrode Area/ Volume Flow Ratio, ft 2 /kcfm. Figure 1. Projected Relationship Between Collection Efficiency and Specific Collection Electrode Area for Various Densities in the Absence of Reentrainment. 176 IV. HOW HIGH RESISTIVITY DUSTS LEAD TO POOR PERFORMANCE The high resistivity dust particles place a limit on the useful operating current density (and applied voltage) by either causing heavy sparking at low current density or back corona at less than normal current density and low voltage. Electrical current conduction in flue gases is primarily associated with the flow of ions. These ions are produced in a normally operating precipitator near the corona electrode. The corona electron avalanche occurs where the electric field exceeds the electrical breakdown strength of the gas. This process is shown schematically in Figure 2. A similar physical process can occur in the dust layer if the electric field in the deposit exceeds the breakdown strength. The electric field in the deposit is proportional to the current density and resistivity as shown in equation 4. E = jp (4) If the electric field in the dust layer approaches the electrical breakdown strength of the gas contained in the voids of the dust layer, a similar electrical breakdown will result. This breakdown will cause either electrical sparking at reduced current density or back corona depending upon . several factors. These disruptive effects in the voltage vs current curves are illustrated in Figure 3. The back corona situation is typically associated with the collection of fly ash from the combustion of the low sulfur western coals - at temperatures around 300-320°F where the maximum point in the resistivity curve typically occurs. The collection efficiency of a precipitator operating in a back corona mode is limited by two factors; first, as indicated by curve d in Figure 3, the applied voltage is limited by the back corona, forcing the power supply to become current limited before a sufficiently high voltage (and electric field) is applied. Second, the back corona at the surface of the dust layer gives rise to the formation of positive ions that are projected back into the inter¬ electrode region. Consider how this disrupts the precipitator performance. The free electrons emerging from the corona region together with the negative ions formed by the inter¬ action of the free electrons and the electronegative gases in the effluent stream cause a negative charge to be imparted to the dust particles. The electric field interacts with the charged particles to drive them toward the collection electrode. ITT Electric Field FIGURE 2. SCHEMATIC OF THE ELECTRON AVALANCHE PHEONOMENON. 178 Current Density Nanoamperes per Cm Applied Voltage Kilovolts Figure 3. Volt-Ampere Characteristic for an Electrostatic Precipitator Plate Spacing 10 inches. Wire Diameter 0.109" for Various Dust Resistivities. 179 These negatively charged particles will move into a region of space adjacent to the back corona region where they will be subjected to large quantities of positively charged ions. The positive ions will interact with the negatively charged particles and tend to neutralize their charges. Since the collection force is proportional to both the value of the electric field and to the charge on the particles, both of which are reduced by the back corona situation, the operation of the electrostatic precipitator is reduced somewhat as a squared function of the limiting high resistivity characteristics. V. FACTORS INFLUENCING RESISTIVITY The fly ash resulting from the combustion of coal in a conventional steam electric generating plant consists of incompletely burned coal, certain crystalline compounds, and spherically shaped, glassy particles. The latter usually represents from 80 to 95% of the total fly ash and forms a continuous matrix in the ash layer through which conduction occurs. Resistivity is influenced by the chemical and physical characteristics of this major constituent. It is also affected by temperature, voltage gradient through the ash layer, and the type and concentration of certain species present in the flue gas. The degree of influence of the individual factors and the number of factors that need to be simultaneously con¬ sidered varies with the temperature of interest. To review this subject, it is helpful to use a resistivity- reciprocal temperature plot for hypothetical ashes of two resistivity levels. Figure 4, and separately consider volume or bulk resistivity and surface resistivity. The right or high temperature leg of the inverted V and the dashed extension represents volume resistivity. The left leg and its extension represent surface resistivity. Between points B and C of curve 1, resistivity is determined by the combined effects of volume and surface conduction. Volume Resistivity Volume resistivity is principally affected by the tempera¬ ture and the chemical composition of the ash. Conduction occurs by an ionic mechanism in which the alkali metal ions, principally sodium, are the charge carriers. The relationship among resistivity, chemical composition, and temperature can be expressed with an Arrhenius equation, in logarithmic form: l8o RESISTIVITY , OHM-CM 140 233 360 540 827 °F TEMPERATURE Figure 4. Resistivity vs Reciprocal Absolute Temperature for Two Hypothetical Ashes. l8l log p = log p 0 + [(0/k) log e] (1/T) (5) where 0 = experimental activiation energy p = resistivity p Q = a complex material parameter T = absolute temperature From equation 5 the inverse linear relationship between log p and absolute temperature shown in Figure 4 is anticipated. As the temperature is increased, the carrier ions become more mobile due to the increased thermal energy and resistivity decreases. It can be seen that the temperature encountered in a conventional hot side precipitator is sufficient to pro¬ duce two orders of magnitude decrease in resistivity. The number of mobile carrier ions is part of the complex material parameter, p 0 . Therefore, a direct relationship between log p and log (number of mobile carrier ions) is expected from equation 5. This has been demonstrated experi¬ mentally for a large number of ashes. The volume conduction- ash chemistry relationship is somewhat complex. The lithium and sodium ions are quite mobile, while the potassium ions are relatively immobile. Furthermore, the iron concentration affects conduction probably by its effect on the structure of the glassy phase. At a given temperature for ashes of otherwise uniform composition, one can show approximately a two order of magnitude change in resistivity for a one order of magnitude change in the combined atomic percentage of lithium and sodium. This effect is illustrated by curves 1 and 2 in Figure 4. This suggests two practical points of interest: 1) a rather subtle change in ash composition can produce a high resistivity problem and 2) additions can be made to the coal to increase the con¬ centration of lithium and/or sodium and thereby lower resistivity. Volume resistivity is also influenced by the porosity of the ash layer and the voltage gradient through the ash layer. As the voltage gradient is increased and/or porosity is decreased, resistivity decreases. Since the precipitator is operated at the sparking level and since a given ash will have an inherent precipitated porosity, these factors are mainly of interest with respect to the technique of measuring resistivity and to the analysis of laboratory data. 182 Surface Resistivity Surface resistivity is influenced by several factors including some that must be considered collectively and some that have an indirect rather than a direct effect. The factors of primary importance are: temperature, flue gas chemistry, ash composition, field strength, surface area of the ash, and chemical durability of the ash. There are at least two viewpoints regarding the mechanism of surface conduction. It has been generally accepted that conduction takes place by an electrolytic or ionic mechanism in which water molecules, sulfuric acid, etc. are adsorbed on the surface of the ash particles forming a conductive film. Without additional definition, this implies that the charge carriers result from the chemical electrolysis of the species forming the adsorbed film. An alternative ionic mechanism has been proposed recently that suggests conduction results from the reaction between the environmental species (water, sulfuric acid, etc.) and the ash surface thereby mobilizing the alkali metal ions to serve as charge carriers. Experimental evidence has been presented to demonstrate the substantial migration of the alkali metals under the influence of an electric field using conditions suitable for surface conduction only. This mechanism will be used to discuss the factors influencing surface resistivity. As the effluent is cooled from the temperature range in which volume conduction is predominant, the relative concen¬ trations of water vapor and sulfuric acid vapor formed from the available SO 3 are increasing. The water and acid react with the ash surface promoting surface conduction, and the resistivity-temperature relationship deviates from a straight line at point c, curve 1, Figure 4. Additional lowering of the temperature should decrease the rate of reaction between ash and environment; however, this lowering of temperature also increases the relative concentrations of the attacking species causing an increase in surface conduction. At some temperature depending on the concentrations of SO3 and water present, the sulfuric acid condenses, and the interaction between this species and the ash surface is increased. The reaction between the ash surface and the attacking medium is believed to be one of ion exchange and dissolution of the ash surface. By ion exchange and/or dissolution, the alkali metal ions in the ash are mobilized at the surface to serve as charge carriers for the surface conduction mechanism. The effect of the factors influencing surface conduction can be considered relative to this reaction. An increase in absolute concentrations of water and S0 3 will increase the reaction between the ash and the environment at a given temperature and increase the temperature at which sulfuric 183 acid will become available for reaction in the condensed form. Without the acid and water the reaction will not take place and surface conduction will be negligible. Little quantitative information is available; however, for some ashes an attenuation of resistivity of one order of magnitude has been observed by increasing the water concentration from 5 to 15 volume percent. The ash itself affects the reaction in several ways, chemically and physically. If all other factors are constant, an ash of finer particle size will present more surface area for reaction with the environment. This promotes the reaction, and consequently more carrier ions are mobilized causing a reduction m resistivity. As would be intuitively suggested, a linear relationship was found between resistivity and a surface area parameter. Ash composition plays an important role regarding the reaction between the ash and the environment. The alkali metals offer reaction sites for the ion exchange process and promote the dissolution of the ash. Therefore, the greater amount of these elements, the less the resistance to chemical attack and the lower the resistivity is because of the presence of mobilized carrier ions. The overall resistance of the ash to attack by the environment might be related to several facets of the ash composition. It is believed that an elevated iron concentration reduces the chemical durability of the ash, promot¬ ing dissolution and thereby decreasing resistivity. Other ash constituents may increase the resistance to chemical attack; for example, Si0 2 and CaO. In addition to increasing the resistance t ^ e r , ash to chemical attack, another effect has been suggested ror CaO. The CaO that is not part of the glassy ash could serve as a getter for S0 3 , sulfuric acid or water preventing these species from reacting with the glassy ash. Limiting the reaction m this fashion would produce high resistivity. It has been observed that surface resistivity decreases with increasing field strength. It is conceivable that the voltage gradient in the ash layer could affect the reaction; however, this point warrants additional thought and experi¬ mentation. ^ Based on the above discussion, the general approach to attenuating high surface resistivity would include: the reduc¬ tion of the chemical durability of the ash, the introduction of additional alkali metal ions to serve as charge carriers, the selection of operational conditions promoting a maximum amount of attacking medium, and the selection of conditioning agents that most readily attack a given ash. VI. ALTERNATIVE TECHNIQUES FOR COLLECTING FLY ASH FROM LOW-SULFUR COALS There are five basic techniques available for overcoming the problems presented by the high resistivity fly ash. These are: Brute Force Flue Gas Conditioning Source Conditioning Operating at Elevated Temperatures Operating at Depressed Temperatures The last four techniques are methods utilized to modify the resistivity while the first merely attempts to accept the existing conditions and work within that framework. Brute Force Technique The brute force approach is essentially the technique utilized by the Australians and by some of the current western utilities. The high resistivity fly ash reduces the effective¬ ness of an electrostatic precipitator by limiting and upsetting the electrical conditions within the unit. This limiting action does not negate the collection function, it merely reduces it. Therefore, if a sufficiently large precipitator is utilized, effective precipitation will result. / High resistivity acts to limit the voltage and current as suggested previously in Figure 3. This limitation will result in depressing an allowable current density of about forty micro¬ amperes per square foot that is associated with low and inter¬ mediate resistivity fly ashes to perhaps three or four micro¬ amperes per square foot in the extreme cases of high resistivity. i The effect of this current and voltage depression was shown in Figure 1 which was generated with the aid of a computer systems model of an electrostatic precipitator. This systems model closely approximates the behavior of full-scale electro¬ static precipitators. The curves shown in Figure 1 were generated by supplying as an input to the system, a representative particle size distribution. The overall collection efficiency was pre¬ dicted as a function of specific collection area for the four values of current density that represent varying degrees of difficulty in the dusts. The curves in Figure 1 are useful in two ways: first, if an existing precipitator is operating at one current density with a given efficiency, the expected behavio] for a different current density can be reasonably estimated by projecting vertically from one current density to the other. As an example utilized before, consider a precipitator operating at an efficiency of ninety-nine and five-tenths percent. If this unit is switched to a high resistivity ash that limits the useful current density to three and one-half microamperes per square foot, the efficiency would be expected to drop to about ninety-five percent, resulting in an increase in the emissions by a factor of ten. 185 A second use is to predict the specific collection electrode area required to attain a given efficiency with a given current density when the efficiency is known for a different current density; consider a ninety-nine and one-half percent collector with a current density of forty microamperes per square foot with an SCA of three hundred square feet per thousand cfm. The SCA required to attain this efficiency at three and one-half microamperes per square foot is projected to be about eight-hundred-fifty square feet per thousand cfm. The above discussion describes the range of variation in collection efficiency expected for the range of current densities associated with the intermediate and high values of resistivity. If the brute force technique is employed, a sufficiently large precipitator is installed and the adverse electrical conditions are allowed to limit the behavior. The advantage of this approach is that no particular or unusual operating procedures are required for the precipitator. Flue Gas Conditioning Flue gas conditioning is a technique used to increase the reaction between the ash and the environment by introduc¬ ing, for example, water vapor or S0 3 to the effluent stream. The surface resistivity of the fly ash can be decreased by introducing sulfur trioxide into the flue gas. This addi¬ tional sulfur trioxide together with the inherent sulfur trioxide and water will react with the surface of the fly ash to provide an increased carrier mobilization. The chemistry of the fly ash as well as the temperature of the flue gas influ¬ ences the amount of sulfur trioxide required for effective conditioning. Typically, the operating temperature is above the sulfuric acid dew point such that a concentration limited adsorption rate prevails. This operation above the acid dew point is important for other reasons, with corrosion one of the primary concerns. The chemical composition of the fly ash and its resistance to attack by the environment also influence the amount of conditioning agent required. For example, fly ash with high free lime contents require greater concentrations of flue gas conditioning than those with low lime. It is thought that the lime reacts with the sulfuric acid to form calcium sulfate. This chemical reaction must continue until the free lime is neutralized, before the S03 can accomplish its intended purpose. The change in resistivity as a function of sulfur trioxide injection rate for a high and low lime content fly ash is given in Figure 5. It is hopeful that additional research will yield other flue gas conditioning agents that are more efficient for certain types of ashes. 186 Resistivity ohm-cm SO 3 INJECTION RATE - PPM FIGURE 5 - RESISTIVITY VS S0 3 INJECTION RATE FOR HIGH AND LOW LIME CONTENT FLY ASH 187 Source Conditioning Source conditioning is the technique utilized to modify the chemical composition of the fly ash. Certain additives blended with the coal feed will combine with the glassy ash to alter the chemical composition of the major constituent. The objective of this approach is to reduce the chemical durability of the ash and/or increase the number of available charge carrying ions. Since sodium serves as a charge carrier for both volume and surface conduction, compounds containing this element can be considered as source conditioning agents. Furthermore, the chemical durability of ash should be decreased by additions of this element. The two curves shown in Figure 4 can be used to illustrate the effect of sodium concentration. All other factors equivalent, the ashes represented by curve one and curve two contain respectively a few tenths of a percent and 2 to 2.5% Na 2 0. The effect of sodium additions to a particular coal for a given set of precipitator operating conditions is shown in Figure 6. From the above, it would also seem possible to increase the precipitator inlet grain loading with a particulate high in carrier ions that would be susceptible to attack by the environment. Properly dispersed in the ash layer and subject to attack by the environment, the additive would release- charge carrier ions producing a reduction in resistivity. Flue Gas Temperature Modification The remaining technique for modifying the fly ash resis¬ tivity is associated with the selection of the operating temperature of the precipitator. Figure 4 shows the resistivity versus temperature dependence for fly ash for a specific set of conditions. At temperatures above six hundred degrees fahrenheit, the volume resistivity is sufficiently low for effective collection of most ashes. In a power station, the flue gas temperature between the economizer and the air preheater falls within this range, thus hot-side precipitators will generally provide effective collection. However, there are two negative factors to consider in hot-side precipitators. First, the volume of gas handled is about fifty percent greater than for cold side, and second, thermal expansion problems are increased. The resistivity can also be reduced by operating at reduced temperatures. The low temperature behavior is some¬ what more variable than the high temperature. In addition to the variation in carrier ion concentration, one must also be concerned with changes in S0 3 and water vapor concentrations. 188 Resistivity ohm-cm Sodium Oxide Concentration by Weight Percent FIGURE 6 - RESISTIVITY VERSUS SODIUM OXIDE CONCENTRATION FOR A FLY ASH FROM A LOW SULFUR COAL T%320°F. 189 The variability of coal within the seam may require modifica¬ tions to the temperature on a day-to-day basis. In one case, it was necessary to drop the temperature to about two-hundred-seventeen degrees for effective collection. In other instances, temperatures in the range of two-hundred-fifty degrees will suffice. Summary When inadequate precipitator performance is caused by high resistivity, each operator is confronted with problems related to the ash produced from a particular coal and the general economics of the given situation. Therefore, one cannot suggest a single approach that will solve every problem. The foregoing approaches may be used alone or in combina¬ tion and must be considered based initially at least on rather broad guidelines: for example, a) existing versus new precipitator b) chemical and physical uniformity of ash produced c) ability to alter existing operating conditions d) the degree of the resistivity problem e) field and laboratory data available Design Parameters and Test Results from Electrostatic Precipitators Collecting Fly Ash from the Combustion of Low Sulfur Western Coals The performance of the precipitators collecting the fly ash from the combustion of the lignites indicates that the ash formed does not have particularly high resistivity. The chemical analysis indicates a sodium oxide concentration typically ranging from 2.5 to 8% by weight. Therefore, the good performance is to be expected. The design parameters and test results for several precipitators are given in Table 1. The lignite installations indicate that resistivity is not severely limiting the performance with the possible exception of the Lelands Olds 1 installation. The Jim Bridger unit is collecting high resistivity ash and the San Juan Station is a hot side unit. However, there is the potential for high resistivity for the lignites when the sodium contents are low. 190 c o c 4 J G »d CP id P G • p • ■H X »p CJ o e • • o * O CP £ P z G a* VO o o CP id «d in ip P ° £ -p o in o 0) Eh a) (d e o G P CP m id Eh JZ in < in P CP p X S a* e U £ e •rt O •H P H -H i p •H •H P 0) > > • 0) M G c c d) T3 O O O £ 4H id -P -P -P •H d) in P P G p O O 04 CO ». in ro rH H CP d) o p cq o 0) 33 (X4 in IN rH p rH d) d) id > Cm •H • «d o3 CJ in • • P G d> 04 -p CP £ mm 0 p •H o Q) G re Cm rH G CP • id G a G P • O d) o z O G >H z O «h 1 rH G p rH o O o o >1 ■P -p o rH G p •H a) id £ cj OQ CN . Q in • TJ TJ Z r—1 rH o d) o «. p G G id 0 TJ 0 53 <—l C -P u 1 rH id -P co o VO r- 00 o IN co CP CP CP o m H O m H O in CN e 00 H • H in Q) in P a + CP CP CP CP in CP CP G Cm G Uh o o 0 O G •H •H \ >n •rH TJ mh -p •P 0~ TJ O TJ id d) P MM d) \ rH \ in rd o id P P d) p rd mm rH P HJ P id rH 0 o •H rH a id p d) G CP P CP

p CO a p CO ai P •P i—i X 42 i—! •p •p 1 •H u •P O o •rl CO d bC CO o p CO o a> CO S > IP <2 CO d <2 S 0 ) <2 g p o •p p CU p o CO 6 d 0) <2 o 0 ) 42 u 22 H I - 1 351 TABLE 5 Ash Analyses and Fusion Temperature 352 Coal fed to the pulverizer was a nominal minus 3/4-inch size. The coal pulverizer was an oversized unit which could not be throttled down to operate continuously at the lower coal rates employed in these investigations. Hence, after intermittent pulverization to about 60 percent minus 200-mesh and 90 percent minus 100-mesh, the coal was collected in a dry cyclone and a baghouse, and stored in a bin made inert with CO2 gas. Indirectly heated air from the kiln hood was used to reduce the moisture content of the coal during the grinding operation. This increased heats of combustion about 1,500 to 2,000 Btu/lb. The pulverized coal was then fed from the bin by a constant weight feeder to a venturi, where the coal was picked up and transported to the burner pipe by an air stream equivalent to 15 to 18 percent of stoichio¬ metric requirements. The coal so transported was injected into the kiln through a simple 1-inch-diameter stainless burner pipe. Tables 8 and 9 give operating data for the burners and for the rotary kiln. TABLE 6. Chemical -and Size Analyses of Magnetite Taconite Concentrate Chemical analysis, dry basis, vt-pct Fe.... Fe 2+ .. Si02- . ai 2 o 3 . CaO... MgO. NapO k 2 o. S. .. 66.3 21.6 5.7 0.1 0.5 0.3 0.012 0.01U 0.015 Sieve size mesh, vt-pct Plus 100 mesh. 0.3 Minus 325 mesh. 83.7 Density, g per cc. 4.88 Blaine No., sq. cm per g. 1,849 TABLE 7. Properties of Green and Grate Discharge Pellets 1 Green Pellets (Range) Moisture, pet. 6.9 - 7.2 Compressive strength, (wet) lb. 6.8 - 7.8 l8-inch drops. 5-6 Grate Pellets (Range) Degree of oxidation, pet. 66 - 69 Compressive strength, lb. 88 - 103 Tumbling index, pet minus 28-mesh. 36-38 Bulk density, lb/cu ft. 125 - 128 ^"Represents average of tests 1-4 353 TABLE 8* Natural Gas-Coal Burner Operating Data 1 Test Oil 1 2 3 4 5 6 Fuel being tested. Natural Gas Kentucky Bituminous Colorado Bituminous Montana Sub B Montana Sub B Lignite Heat value.. 995 13,750 13,420 10,280 10,280 8,930 Natural gas-air premix burner Gas rate, scfm. 55 23 18 32 23 25 Air rate, scfm. 477 219 169 325 224 248 Heat input, MM Btu/hr. 3.28 1.37 1.07 1.94 1.35 1.52 2 Coal burner Coal rate, lb/hr. - 92 92 111 50 70 Coal rate, MM Btu/hr. - 1.26 1.23 1.14 0.51 0.63 Coal rate, MM Btu/l.t. pellets. “ 3.7 3.6 3.4 1.4 1.7 Primary air to burner pipe Volume flowrate, scfm. 20 33.3 33.3 33.3 41.7 20.4 Mass flowrate, lb/hr. - 150 150 150 188 92 Temp. 8 C. - 41 38 42 38 49 Ratio, lb air/lb air. - 1.6 1.6 1.4 3.8 1.3 Pet combustion air. - 15.7 15.8 17.9 48.1 19.7 Velocity at burner tip, ft/min. 5,925 5,868 5,934 7,354 3,725 Secondary air from pellet cooler Volume flowrate, scfm. 311 431 467 476 261 270 Mass flowrate, lb/hr. 1,403 1,942 2,104 2,148 1,178 1,219 Cooling rate, lb air/lb pellet. 1.8 2.6 2.7 2.8 1.4 1.4 2 Heat value units: natural gas - Btu/cu ft, coal - Btu/lb. Refractory screen burner: (Port area + 3.73 sq in). Burner pipe diameter: 1.05 in (Area = 0.006 sq ft). TABLE 9. Operating Data - Rotary Kiln * Test Oil i 2 3 4 5 6 Fuel being tested. Natural Gas Kentucky Bituminous Colorado Bituminous Montana Sub B Montana Sub B Lignite Grate pellet feed to kiln, lb/hr 765 758 770 762 815 835 Total heat input to kiln MM Btu/hr. 3.28 2.63 2.30 3.09 1.86 2.14 Heat distribution to kiln, pet.. Natural gas burner. 100 52 47 63 73 71 Coal burner 48 53 37 27 29 Kiln temperature, 0 C Distance from kiln discharge, ft 1.4. 1,240 1,194 1,189 1,188 NA NA 4.5. 1,300 1,300 1,302 1,301 NA NA 11.4. 1,278 1,248 1,289 1,296 NA NA 17.0. 1,167 1,124 1,048 1,134 NA NA 24.9. 1,079 1,040 1,018 1,054 NA NA 30.3. 971 917 904 957 NA NA 34.1. 918 808 790 880 NA NA Kiln off-gases Volume flowrate, scfm. 862 707 688 868 NA NA Mass flowrate, lb/hr. 3,795 3,148 3,070 3,856 2,485 2,572 Temp. ° C. 918 808 790 880 674 723 Velocity, ft/min. 554 414 394 540 280 303 Oxygen, vol-pct. 6.2 6.7 6.4 7.0 6.1 6.1 Carbon dioxide, vol-pct.... - 10.1 11.1 9.9 10.7 11.0 Excess air, pet. " 44 42 47 38 32 Kiln pressure (discharge end)... +0.05 -0- +0.02 + .06 -0.01 +0.02 Rotary kiln: 34-inch ID, length - 35 ft,volume = 220 ft? All cooler air diverted to kiln. Rotary kiln: 34-in. ID, 35 ft long, 220 cu. ft. volume 35^ RESULTS AND DISCUSSION Problems resulting from the presence of the coal ash (that is kiln ringing, pellet quality, and fouling of the grate to kiln transfer point) did occur as expected. Other characteristics determined were kiln temper¬ ature gradients, pellet quality, and sulfur distributions. Kiln Ringing In test 1 in the pilot plant kiln, with 100 percent natural gas firing, a uniform coating about ^-inch thick was formed on the kiln lining. No evidence of a ring buildup was observed during this 120-hour test period. However, during pellet induration tests with solid-fuel firing, signs of a ring formation began to show within 24 to 36 hours of steady state operations. The rings continued to grow at a comparatively slow rate, but, except for Colorado coal, did not affect pellet flow during these relatively short runs. This condition of ring growth was most severe with Colorado coal, less severe with lignite, still less with subbituminous coal, and least with Kentucky coal. The Colorado coal produced a pronounced buildup, averaging about 9 inches in height, and with irregular accretions almost blocking the kiln's interior in 80 hours. The kiln ringing had so effectively blocked passage of pellets that the test had to be terminated at that time. Lignite firing caused a buildup about 2.5 feet in length and 5 inches deep. Photographs of the kiln interior taken during the last three consecutive days of operation with lignite firing (figure 2 - A, B, C) show the progressive ring growth over this period. A view of the kiln interior after the test (figure 2-D) indicates the extent of the buildup by the end of the 120-hour test. A profile analysis of the kiln lining depicted in figure 3 (the temperature gradients are discussed below) shows that the maximum ring occurred at a distance of about 10 to 11 feet from the discharge end of the kiln, at a point a few feet beyond the coal flame. The temperature at this point (1,160° C) was approximately 140° C lower than the maximum kiln temperature. From the composition of the ring materials given in table 10, it is apparent that much of the ash generated from the lignite combustion was deposited at this point and contributed to the ring formation. The ring was composed mainly of pellet chips and fines and contained approximately 5 percent lignite ash. A piece of the ring structure removed from near the thermocouple well is shown in figure 4. The photomicrograph of this material indicates that the microstructure is similar in appearance to a self-fluxed pellet. The hematite particles (white) appear semirounded to rounded and bonded together by a slag matrix (light grey), as well as by ore bridges. The less severe ringing problems that occurred with subbituminous-coal firing (average thickness 3 inches, with irregular accretions reaching 6 inches) could be attributed to the lower calcia and soda contents of that coal's ash. The fact that the ring accretion contained mostly iron oxide and only a small percentage of ash is an indication that the fluxing action of these particular ash elements may be more important to ring formation than the ash-fusion temperature. 355 356 DISTANCE FROM WALL, in FIGURE 3. - Profile analysis of kiln lining showing location of maximum ring formation (test 6). 0 I 23456789 10 I_I_ i i l _I_l_I_I_I_I Scale, cm FIGURE b. - Photomicrograph showing micro¬ structure of ring buildup near thermocouple (test 6). 357 TABLE 10. Analyses of Material Deposited on Kiln Lining Q-l 3 -a •H d I B •pH x c0 B B o 3 e •H (U l-l 4J o o r—4 4-1 CU M CO a> 3 i—i tO > 3 O l-i •H CO 3 O 3 V4 CU 4-4 4-4 CU B CO 40 cfl > pq cO •rH cu 4-1 co B X t-4 oo OO co c •i—l O -l-l WO -pi CM CO 358 Table 10 gives partial chemical analysis of deposited materials taken from interior kiln walls at both the hot zone (1,300° C) and the point of maximum buildup after tests 1 through 5. In both instances, the major and unreported constituent is iron oxide or pellet fines imbedded in molten or semimolten ash. However, at the point of maximum buildup the ash represents a much smaller proportion of the total mass. Figure 5, in addition to showing the extent of kiln ringing (for tests 1 through 4) also shows that the maximum buildup in each instance occurred at the point where the temperature was between 1,150° to 1,200° C—closely corresponding to the initial deformation temperature of the ash of both Colorado and Montana coals. Also of some interest, figure 5 attempts to show that the length of flame obtained in burning these different coals varies somewhat. An additional factor that may also have contributed to the less severe ringing was the higher burner tip velocity used in the test with subbituminous coal. Typical characteristic flames from burning of both lignite and sub- bituminous coal are shown in / the photographs of figure 6. The higher velocity subbituminous flame was more compact and intense than the lignite flame, and the ash particles may have been propelled further from the kiln hot zone where they would be less likely to contribute to ring buildup. Whether or not the burner tip velocity is an important factor in ring formation will be determined in future tests. With regard to buildup, one might expect that some indicator could be found, such as sodium content, ferrous iron content, or basicity of ash, that might help in screening out coals that would be poor performers. However, at this point there is no single indicator that can be pinpointed—except that the highest ash fusion temperature obtainable appears to be preferable. Even in this instance however, present evidence is not conclusive because of differences in behavior noted in the Montana and Colorado coals, which are closely matched with respect to ash fusion temperatures. Still more confusing is the lesser amount of kiln ringing caused by Montana coal as compared to lignite in spite of the lignite's higher ash fusion temperature (see table 5). Pellet Quality Pellets discharging from the kiln were of moderate strength and abrasion resistance as shown in table 11. Crushing strengths ranged from 550 pounds for pellets fired with natural gas down to 360 pounds for those fired with the Kentucky coal. This variation seems to reflect the degree of pellet oxidation, as measured by ferrous iron content, and the degree of pellet oxidation, in turn, is indicative of different stoichiometric requirements of oxygen for burning each of these fuels. Chemical analyses of both green and fired pellets are given in table 12. Since one of the concerns over coal firing is that the pellets will suffer chemical degradation, it is gratifying to note that, with the exception of about 0.20 to 0.25 percent increases in silica content, coal firing had an insignificant effect in this 359 V .c u c < $ O tr <_> z < (- to Q 15 10 5 0 15 AVERAGE KILN TEMPERATURE, °C 1,200 l.300 (|320) 1,280 1,180 1,050 940 850 t -r Coal flame 3 Coal flame i-1-1 10 - 5 0 15 10 5 0 ■ Average bulk □ Irregular accretions Coal flame I-1—H J_L J- Mi Gas flow Pellet flow Colorado bituminous Montana subbituminous - Jr Kentucky bituminous _L _L _L _L _L _L 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 DISTANCE FROM DISCHARGE END, feet FIGURE 5. - Flame characteristics and kiln ringing when firing with bituminous and sub- bituminous coals. Subbituminous B coal flame Lignite flame (V = 7,354 ft/m in) (V = 3,725 ft/m in) FIGURE 6. - Typical burner flames produced from combustion of subbituminous coal and lignite. 360 TABLE 11. - Physical Properties of Pellet's n ft J>S Cm Si ft 1-1 •H ft 0 CO 0 0 LTN ft CO 0 -=t - 3 - m G 1—1 1—1 1—1 1—1 1 —1 CD ft ft 1-1 G ft O CO •H X co a> ft a On -4- CO 0 00 ctj O ft ft ft cO ft co _=r CO ft •H cm < 1 ft to ON -=f LT\ On ft G ft t"— VO Lf\ CO 0 4 ) , possibly through the reaction of sulfur dioxide and sodium oxide in the gas phase. This is evidenced by the fact that much of the sodium found in electrostatic precipitator ash is in the form of sodium sulfate (3). The remaining extraneous or inherent ash constituents, (SiC^, A^O^, CaO, MgO, and Fe 203 ) become free to coalesce and deposit onto the pellets or kiln lining. Deposition of fly ash in the grate-to-kiln transfer chute was observed when firing with Montana subbituminous coal. Partial fusion of fly ash particles from this fuel was noted. Firing Characteristics The effect of coal firing on flame length has already been discussed. However, flame lengths also affect temperature profiles. In tests 2 and 3, using the Kentucky and Colorado coals, substantially less total heat input (less natural gas required) gave a temperature profile at the hot end similar to that of baseline test 1, but the kiln exit gas temperatures were more than 100° C lower than for all gas firing (see figure 10). This resulted because the control thermocouple in the hot zone sensed the localized high temperature coal flame and automatically reduced the heat input from the gas burner. In test 4 the localized coal-flame hot zone was forward of the control thermocouple so that more natural gas was automatically introduced, while maintaining coal input at about the same Btu/hr level. This thereby raised the total thermal input to approximate that of the baseline test. The net effect was a flattening of the temperature profile and of raising the temperature of kiln exist gases. Kiln temperature profiles for tests 5 and 6 are shown in figure 11. Although both tests were run under similar conditions, somewhat higher temperatures were sustained in test 6. 365 TABLE 13. Ash Balance for Coal-Firing Tests cO g 3 G U •H g 00 •H o cO 4-1 CO Q cn 4-1 0) •H G O' O) o 4-1 G 00 G •r^ •iH G O CJ CM 366 Other--includes ring formation, adhesion to kiln lining, and unaccounted losses. KILN TEMPERATURE, FIGURE 10. - Temperature profiles when firing with "bituminous and suhbituminous coals. 367 KILN TEMPERATURE, 1,400 2,552 o O 2,372 2,192 2,01 2 1,832 1,652 1,472 1,292 I, I 12 0 3 6 9 12 15 18 21 24 27 30 33 36 DISTANCE FROM KILN DISCHARGE,ft FIGURE 11 - Temperature profiles when firing with subbituminous coal and lignite. 368 KILN TEMPERATURE, OTHER FIRING TECHNIQUES Cyclone Burner The cyclone burner is a slagging type reactor. The furnace is a compact, horizontal cylinder that is water cooled to freeze a thin slag layer on its interior surfaces. Minus 1/4-inch coal and primary air at about 94° C are introduced tangentially at the front of the burner and pass into the cyclone in a whirling motion. High velocity secondary air at 400° C, also introduced tangentially, burns the coal particles rapidly and completely to reach temperatures in excess of 1,650° C. Molten slag is drained through a tap hole; approximately 75 percent of the total ash is removed in this manner. Contamination of pellets by molten and fly ash should be minimized, but may not be eliminated entirely. Heat is lost in the molten slag and to the cyclone-cooling water. In comparison to gun type burners, the cyclone burner does not require such fine pulverization of the coal, but energy savings accrued in this manner are probably expended by the additional capacity fans required to deliver the secondary combustion air to the cyclone. In a sense the cyclone burner is an external combustion chamber, but its compact configuration may offer advantages, especially in existing installations where space for modifications is limited. Barring other con¬ siderations, the cyclone burner might be adaptable to shaft, grate, or grate-kiln systems and should be the most satisfactory alternative for coal firing for existing shaft furnaces. However, the shaft furnace would appear to be less adaptable to coal firing than grates or grate-kilns because of the possibility that solid particulates might choke the entrance ports and deny access of the hot gases to the pellet bed. There have been no tests using the cyclone burner on any pelletizing system yet. However, the Twin Cities Metallurgy Research Center constructed a small cyclone burner for installation on its grate-kiln system, and evaluation of this unit is underway. External Combustion Chamber Research in progress during the last half of 1974 by the Dravo Corp. (4) suggests at least one other mode of an external combustion chamber, with the specific intention of providing the thermal requirements of the straight grate system. In place of multiple burners, this concept envisions two combustion chambers approximately 22 to 24 feet high on either side of the grate strand. A gun-type burner would be mounted at the top of the chamber, with the flame directed downward. Combustion temperatures are maintained sufficiently high to slag a substantial portion of the ash constituents; the molten slag is discharged through the bottom of the chamber, water quenched, and discarded. The hot gas, still containing some suspended ash and volatile consitutents, is vented to the grate hood, where streams of gas are tempered and routed to various points along the pellet bed. 369 Although Dravo's effort has been directed to applications to the straight grate system, some modification of the external combustion chamber concept should be equally applicable to the grate-kiln system. As yet, there has been no known coupling of the combustion chamber to an actual pelletizing machine, so there is no available information on how much dust, or volatile constituents are carried over in the gas streams to the pellet bed, and what effect these constituents might have on the operation, or on pellet quality. However, it is logical to assume that the problems are less acute than when no ash is removed. With an external combustion chamber a certain fraction of the heat will be lost in discharge of the molten slag. Further, some care must be taken to select refractories able to withstand the corrosive contact with molten slag. Coal pulverization to the same degree of fineness as required for direct-type gun firing would be necessary. Coal Gasification Coal gasification also could be coupled with pellet induration. In this concept hot, raw gas and desulfurized gas will be tested as potential fuels for direct firing of the grate-kiln system. Expected heating values of the fuel gas will be between 150 and 400 Btu/scf. CONCLUSIONS Tests conducted at the Bureau of Mines Twin Cities Metallurgy Research Center to evaluate the use of pulverized coal as a substitute for natural gas in the induration of iron oxide pellets show: 1. When coal is fired directly into the rotary kiln of a grate-kiln unit, the major problem is the formation of coalescent masses or rings in the kiln interior. The severity of ring formation has not been traced to any one source, but the best performance can be expected from coals having an ash-fusion temperature above the range of temperatures at which induration occurs. Dust deposition in constricted passages, as in grate-to-kiln transfer chutes, may also be expected using Montana subbituminous coal. 2. Pellet contamination, including sulfur pickup, from coal combustion products and residues is not a serious problem. 3. Pellets made in the pilot plant from commercial mangetite con¬ centrates and indurated in a grate-kiln system with bituminous coal, sub- bituminous coal, or lignite had mechanical properties equal to or better than their commercial counterparts. 4. Pellet contamination from the alkalies, Na20 and K^O, in the fuel is less than that contributed to the green pellet by bentonite additions. 370 5. Temperature profiles obtained with coal firing show steeper descent from burner to feed end of the kiln than do those with natural-gas firing. Because of the localized luminous flame obtained with coal, temper ature control based on kiln exit gases, as is the case with natural-gas firing, may result in excessively high temperatures at the hot end. 6. The fluxing interaction of coal or lignite ash with iron ore bentonite mixtures plays an important role in promoting ringing at, or near the induration temperature. Fuel selection will have to include considera¬ tion of ash fluxing behavior as well as fusion temperatures. 7. If fuel selection cannot control ringing, cyclone burners, other external coal combustion chambers, or coal gasification may be required, but costs would be expected to be higher than those for direct firing. 8. There are strong indications that magnetite pellets undergo some reversion from their partially oxidized state as they pass through the coal flame. This condition may require some adjustments in standard operating procedures in order to insure that the required degree of pellet oxidation is reached and maintained. REFERENCES 1. Frommer, D. W., and J. C. Nigro. The Bureau of Mines Looks at Coal Firing for Induration of Iron Ore Pellets. Paper presented at the 48th Annual Meeting, Minnesota Section, AIME, Duluth, Minnesota, January 15, 1975. 2. Frommer, D. W., and J. C. Nigro. Practical Aspects of Coal Firing in the Induration of Iron Ore Pellets. Paper presented at and published in the Proceedings of the AIME Ironmaking Conference, Toronto, Ontario, Canada, April 1975. 3. Gronhovd, G. H., W. Berkering, and P. H. Tufte. Study of Factors Affecting Ash Deposition From Lignite and Other Coals. Preprint of presentation at ASME Winter Annual Meeting, November 16-20, 1969, Los Angeles, California. 4. DeKlaver, M. A., and Traveling Grates. Minnesota Section, G. P. Leighton. Coal Firing of Dravo-Lurgi Paper presented at the 40th Annual Meeting, AIME, Duluth, Minnesota, January 15, 1975. 371