EP1.23/8: 600/7-77- 054 U.S. Environmental Protection Agency Office of Research and Development Industrial Environmental Research Laboratory Research Triangle Park, North Carolina 27711 EPA-600/7-77-054 May 1977 PRELIMINARY ENVIRONMENTAL ASSESSMENT OF COAL-FIRED FLUIDIZED-BED COMBUSTION SYSTEMS Interagency Energy-Environment Research and Development Program Report W C GALEGAR RSKERL US EPA AOA * OK 7482 Research reports Environmental Pr These seven broa development and of traditional g transfer and a m are: 1. Environ 2. Environ 3. Ecologi 4. Environ 5. Socioec 6. Scienti 7. Interag This report has pment, U.S. to seven series, litate further gy. Elimination ster technology The seven series (STAR) Development GY-ENVIRONMENT RESEARCH AND DEVELOPMENT series. Reports in this series result from the effort funded under the 17-agency Federal Energy/Environment Research and Development Program. These studies relate to EPA's mission to protect the public health and welfare from adverse effects of pollutants associated with energy systems. The goal of the Program is to assure the rapid development of domestic energy supplies in an environmentally—compatible manner by providing the necessary environmental data and control technology. Investigations include analyses of the transport of energy-related pollutants and their health and ecological effects; assessments of, and development of, control technologies for energy systems; and integrated assessments of a wide range of energy-related environmental issues. REVIEW NOTICE This report has been reviewed by the participating Federal Agencies , and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Government, nor does mention of trade names or commercial products constitute endorsement or recommen¬ dation for use. This document is available to the public through the National Technical Information Service, Springfield, Virginia 22161. EPA-600/7-77-054 May 1977 PRELIMINARY ENVIRONMENTAL ASSESSMENT OF COAL-FIRED FLUIDIZED-BED COMBUSTION SYSTEMS by Paul F. Fennelly, Donald F. Durocher, Hans Klemm, and Robert R. Hall GCA Corporation GCA/Technology Division Bedford, Massachusetts 01730 geolog c URVEY LI DR * 24 ' :r*v. • Contract No. 68-02-1316, Task No. 15 Program Element No. EHE623A EPA Task Officer: D. Bruce Henschel Industrial Environmental Research Laboratory Office of Energy, Minerals, and Industry Research Triangle Park, N.C. 27711 Prepared for U.S. ENVIRONMENTAL PROTECTION AGENCY Office of Research and Development Washington, D.C. 20460 Digitized by the Internet Archive in 2018 with funding from University of Illinois Urbana-Champaign Alternates https://archive.org/details/preliminaryenvirOOfenn M&/ 7 -,7-osy ABSTRACT This report provides a preliminary evaluation of the potential pollutants which could be generated in coal-fired fluidized bed combustion processes. Because S0„ and NO formation have already received considerable attention Z. X from a number of investigators, the primary emphasis here is on the so- called "other” pollutants — namely, trace elements, organic compounds, inorganic compounds (other than S0„ and NO ) and particulates. z x Using available bench scale or pilot plant data and/or simple thermodynamics and empirical correlations with data from other combustion systems, order of magnitude estimates have been made of the concentrations of various elements and compounds in either the flue gas, the solid waste, or the water discharge. The results suggest that, in general, no special environmental problems should result from coal-fired fluidized bed combustion. But the results also indicate that better data are required in several areas, particularly with regard to particle size distributions, possible organic compounds, and the fate of elements such as Be, As, U, Pb, Cd, Ni, Cl, Se, F and their compounds. iii CONTENTS Page Abstract iii List of Figures viii List of Tables x Acknowledgments xii Sections I Summary 1 General Comments 1 Estimated Concentration Ranges of Potential Pollutants From Coal-Fired Fluidized Bed Combustion 5 Flue Gas 5 Solid Waste 5 II Introduction 7 References 8 III Overview of Fluidized Bed Combustion Systems 9 Types of Fluidized Bed Combustion Systems 11 Potential Effluent Streams From Fluidized Bed Combustion 15 Principal Differences Between Fluidized Bed Combustion and Conventional Combustion 24 References 25 v CONTENTS (continued) Sections Page IV Potential Pollutants From a Coal-Fired Fluidized Bed Combustor 28 Introduction 28 Fluidized Bed Combustion 29 Trace Elements 29 Gaseous Organic and Inorganic Compounds 45 Particulate Emissions 56 References 62 V Potential Pollutants From Auxiliary Processes Associated With Fluidized Bed Boilers 67 Limestone Regeneration 67 One-Step Regeneration 67 Two-Step Regeneration 68 Solid Waste Disposal 74 Experimental Studies 74 Trace Metal Leaching 75 Fuel Storage and Handling 80 Coal Storage Requirements 80 Coal Pile Drainage and Leachates 80 Air Emissions From Coal Storage and Handling 82 Coal Drying 82 Cooling Systems 83 References 87 vi CONTENTS (continued) Sections Page VI Suggested Control Technology For FBC Systems 91 Introduction 91 Flue Gas Treatment: And-On Control Technology 91 Particulate Control Equipment for Atmospheric Pressure Combustion 92 Particulate Control for Pressurized Combustion Systems 105 Control of Gaseous Emissions 109 Pollution Control Via Process Modifications: Some Consideration Based Upon Fluidization Fundamentals 110 Bed Depth 111 Bed and Boiler Tube Geometry 112 Fluidized Grid Design 113 Particle Size 114 Fluidization Velocity 116 Excess Air 117 Mechanism of Coal Injection 119 Pressure 119 Control of Pollutants From Spent Stone Disposal 120 Solid Waste Control Methods 122 Commercial Uses for Solid Waste By-Products 124 References 127 Appendix Preliminary List of Conceivable Pollutants 132 vii FIGURES No. Page 1 Schematic of Fluidized Bed Boiler 10 2 Selected Fluidized Bed Combustion System Options 12 3 Schematic Diagram of an Atmospheric Pressure Fluidized Bed Combustion System 16 4 Schematic Representation of Coal Combustion 46 5 Schematic Representation of Coal Pyrolysis 48 6 Pyrolytic Synthesis of B(a)P 49 7 Variation in Hydrocarbon Concentration with Flue Gas Oxygen Content in the Fluidized Bed Module (FBM) 51 8 Typical Particle Size Distribution of Elutriated Material Collected in Primary Cyclone, Secondary Cyclone, and Fil¬ ter Bag During Period of Additive Injection 58 9 Fly Ash Size Distribution from Pope, Evans and Robbins, Inc., Atmospheric Pressure Fluidized Bed Combustion (AFBC) 59 10 Solids Loading of Flue Gas Leaving the Combustor in Argonne National Laboratories Pressurized Fluidized Bed Combustion (PFBC) 61 11 M. W. Kellog One-Step Regeneration Scheme 69 12 M. W. Kellog Two-Stage Regeneration Scheme 71 13 Solubilities of Trace Metals - (Free Aqueous and Mono- Hydroxo Complexes Only Considered) 77 14 Control of Flue Gas Emissions From an Atmospheric Pressure FBC Boiler 93 viii FIGURES (continued) No. Page 15 Control of Flue Gas Emissions From a Pressurized FBC Boiler 94 16 Particle Size Distribution Before Final Control Device 97 17 Dew Point Elevation as a Function of SO^ Concentration 99 18 Measured Fractional Efficiencies for a Hot Side Electro¬ static Precipitator, with the Operating Parameters as Indicated, Installed on a Pulverized Coal Boiler 101 19 Penetration Calculated From a Venturi Scrubber Model as a Function of Pressure Drop and Particle Aerodynamic Diameter 103 20 Median Fractional Collection Efficiency for 22 Tests 104 21 Schematic Representation of Aerodyne Particulate Separator 108 22 Quality of Fluidization as Influenced by Type of Gas Distributor 114 23 Comparison Between Calculated and Experimental Entrainment at Various Pressures 121 ix TABLES No. Page 1 Atmospheric Pressure Fluidized Bed Combustion System Material Flows and Characteristics 17 2 Estimated Trace Element "Worst Case" Emission Factors for Bituminous Coal 32 3 Estimated Trace Element "Worst Case" Emission Factors for Lignite 33 4 Typical Values of Trace Elements in Limestone and Coal (ppm) 35 5 The Separation of Elements in the Geochemical Classification Scheme 36 6 Comparison of Exxon and Argonne Data on Trace Element Recoveries 37 7 Projected Atmospheric Emissions of Trace Elements From Conventional and Fluidized Bed Combustors Expressed as a Percentage of the Element Entering the System 38 8 Comparison of Estimated Trace Element Concentrations (Class I Elements) With an Environmental Index Based on Threshold Limiting Values 40 9 Boiling Points of Compounds Often Found in Coal 41 10 Comparison of Estimated Trace Element Concentrations (Class II Elements) With an Environmental Index Based on Threshold Limiting Values 42 11 Calculated Equilibrium Concentration for Selected Species Produced by Incomplete Combustion of Coal 55 12 Probable Chemical Form of Trace Elements in the Regenerator - Extrapolated From Fuel Gasification, Free Energy Minimiza¬ tion Calculations 72 x TABLES (continued) No. Page 13 Volume of Spent Bed Plus Ash Produced Per Year By a 635-MW FBC Plant 74 14 Coal Ash Contamination of Beneficated Lime/Anhydrite 76 15 Relative Solubilities in Weakly Alkaline Solutions 79 16 Composition of Drainage From Coal Piles 81 17 Chemicals Used in Recirculative Cooling Water Systems 85 18 Cooling Tower Corrosion and Scale Inhibitor Systems 86 19 Distribution by Particle Size of Average Collection Effi¬ ciencies for Various Particulate Control Equipment 95 20 Summary of Potential Particulate Removal Systems 107 21 Spent Bed Plus Ash Produced by a 635-MW Once-Through Sorbent FBC Plant 120 22 Ash Collection and Utilization, 1971 125 23 Preliminary List of Possible Pollutants From Fluidized Bed Combustion of Coal 133 xi ACKNOWLEDGMENTS The authors would like to acknowledge helpful discussions during the preparation of this report with Mr. D. Bruce Henschel of the U.S. En¬ vironmental Protection Agency. The cooperation of the technical staff of the fluidized bed combustion projects at Argonne National Labora¬ tories, Argonne, Illinois; Exxon Research and Engineering Company, Linden, New Jersey; and Westinghouse Research Laboratories, Pittsburgh, Pennsylvania, is gratefully acknowledged. Other members of the GCA/ Technology Division who provided assistance were Mr. Mark Bornstein, Mr. Richard Wang, Ms. Dorothy Sheahan, Ms. Sandra Sandberg, and Ms. Josephine Silvestro. xii SECTION I SUMMARY GENERAL COMMENTS This report provides a preliminary evaluation of the potential pollutants which could be generated in coal-fired fluidized bed combustion processes. Because S0„ and NO formation in fluidized bed combustion have received z x considerable attention from a number of investigators, the primary empha¬ sis here is on the so-called "other" pollutants - namely, organic com¬ pounds, inorganic compounds (other than SO 2 and N0 X ), trace elements and particulates. The major purpose in a sense has been to serve as a "devil's advocate" with respect to pollutant generation. The aim is to focus atten¬ tion on potential environmental problems as early in the development cycle as possible. Accordingly, conclusions are often based on limited data; in some cases, no data are available and one must rely on scientific and engineering estimates. The results are intended primarily to stimulate interest in potential problem areas and to assist in the design and plan¬ ning of future experimental programs. Based on data from bench scale reactors, fluidized bed combustion offers significant potential for reducing atmospheric trace element emissions in comparison with conventional coal combustion systems, but data on trace element emissions from larger fluidized bed pilot plants are still lacking. Also lacking are data with respect to trace element composition as a function of particle size. Based on "worst case" analyses with bituminous and lignite coal as feed materials, elements whose emission pathways warrant further attention are: Be, As, U, Pb, V, Cl, and F. 1 Limited experimental data on particulate emissions from atmospheric pres¬ sure bench scale and pilot plant operations indicate that control devices such as cyclones, fabric filters or electrostatic precipitators should be able to control emissions to a level similar to that attained in conven¬ tional combustion systems. However, the compatibility of particulate control devices with full-scale fluidized bed boilers has not yet been tested; the need for experiments in this area is very important. Control¬ ling particulates in pressurized fluidized bed is a difficult matter. Pilot plant experiments indicate the use of two stage cyclones will not meet current New Source Performance Standards. As yet, no satisfactory third stage device which can operate at high temperature and pressure has been demonstrated. There is virtually no experimental information available concerning poten¬ tial organic pollutants which could form in coal-fired fluidized bed com¬ bustion. An evaluation based on chemical engineering experience and simple thermodynamic considerations indicates no special problems should occur from gaseous organic pollutants; however, experimental verification of estimated emission rates should receive high priority. The following generic classes of "potential" pollutants have been consid¬ ered in this report: • Acids and Acid Anhydrides — Or ganic Acids - Compounds such as carboxylic acids, dicarboxylic acids, and sulfonic acids could conceiv¬ ably form from incomplete combustion of hydrocarbons; however, at the temperatures involved (~850°C), com¬ pounds of this type should quickly decompose to form small hydrocarbons and CC^. — Inorganic Acids - The predominant inorganic acids should be HC1 and HF. Sulfuric, sulfurous, nitric and nitrous acid should not form until the flue gas has cooled to temperatures less than 240°C. 2 • Carbon Compounds The major carbon compounds, as expected, will be CO 2 and CO. Soot (solid carbon) could be a problem since it would be emitted as small particles (< 0.1 pm) which could pass through most particulate control devices. Calcium and mag¬ nesium carbides could also form in small quantities in the ash formed in the combustor, the carbon burn-up cell or the regenerator. After ash disposal, these compounds could re¬ lease acetylene upon contact with water. The quantities generated, however, should pose no special problems. • Halogen Compounds Halogens should be emitted primarily as HX (where X = F, Cl, Br). Experience in coal combustion chemistry indicates that the presence of chlorine in coal en¬ hances condensation reactions via elimination of HC1; hence, species such as chlorinated hydrocarbons are not favored. Furthermore most chlorinated aliphatics are unstable at the temperatures prevailing in the combustor. • Hydrocarbons Long chain aliphatics and cyclic hydrocarbons should decompose within the bed to form H 2 and smaller hydro¬ carbons. Some of these smaller hydrocarbons could condense to form polycyclic species such as pyrene, anthracene, etc. Hydrocarbon concentrations are strongly dependent on the amount of excess air. With excess air levels of about 20 percent, total hydrocarbon levels (measured as methane) of less than 100 ppm are attainable. Based on a comparison with data from conven¬ tional combustion systems, 1 ppb of benzo(a)pyrene (or similar compounds) could be present in fluidized bed combustion. With excess air levels of 10 percent, benzo(a)pyrene could reach 10 ppb. • Nitrogen Compounds The predominant nitrogen compound should be NO. Trace amounts (< 1 ppm) of HCN, (CN )2 and azoarenes may be present. Species such as amines, pyridine, pyrroles, and nitrosamines, which could form as combustion inter¬ mediates, should quickly decompose within the bed to form hydrocarbons, NO and HCN. 3 Oxygen Compounds Oxygenated hydrocarbons such as furans, ethers, esters, aldehydes, etc. could form as combustion intermediates. These species, however, are unstable at temperatures on the order of 850°C, and they should decompose before escaping from the bed. Ozone, if formed, would also decompose at these temperatures. Particulates As mentioned earlier, although there are only very limited data available on particulate concentrations and particle size distributions, it seems that no special problems should result from particulate loadings, provided the conventional process cyclones and a control device such as an electro¬ static precipitator or a fabric filter are used. The actual performance of these devices on full-scale fluidized bed boilers, however, has not yet been tested. Radioactive Isotopes Based on a "worst case" analysis with lignite coal, uranium could be emitted at significant levels, but the radioactive isotopes of uranium as well as those of other radioactive elements should not be present in high enough quantities to cause concern. Sulfur Compounds SO 2 and SO 3 will be the major sulfur compounds formed. The presence of limestone in the bed, however, should keep these emissions below present emission standards. COS could form in trace quantities (-1 ppm). Compounds such as thiophenes or mercaptans, if formed as combustion inter¬ mediates, will decompose to hydrocarbons and H 2 S; the H 2 S will then be oxidized to form SO 2 . Trace Elements Data should be acquired with respect to the fate of trace elements such as Be, As, Pb, V, Ni and Cl. Also needed are data with respect to chemical composition as a function of particle size. For the most part, the other trace elements commonly encountered in coal com¬ bustion are captured in the coarse solids and remain in the bed. The high pH of the bed material is advantageous in that it tends to retard the leaching of the heavier metals ESTIMATED CONCENTRATION RANGES OF POTENTIAL POLLUTANTS FROM COAL-FIRED FLUIDIZED BED COMBUSTION Using available data and/or simple thermodynamics and empirical correla¬ tions, estimates have been made of the concentrations of various elements and compounds in either the flue gas or the solid waste. In some cases, these estimates are based on simple and tenuous assumptions; but, in gen¬ eral, they should be good to within an order of magnitude. The main pur¬ pose of estimates such as these is to help define experiments which can be used to test the environmental acceptability of fluidized bed combustion. Flue Gas One hundred parts per million: (100 ppm) CH^, CO, S0 2 , NO Ten parts per million: (10 ppm) S0 3 , C 2 H a , C 2 H 6 , HC1 One part per million: (1 ppm) HF, HCN, NH 3 , (CN) 2 , COS, H 2 S, H 2 SO 4 , HNO 3 , Elemental Vapors, As, Pb, Hg, Br, Cr, Ni, Se, Cd, U, Be, Na One part per billion: (1 ppb) Diolefins, Aromatic Hydrocarbons, Phenols, Azoarenes One-tenth part per billion: ( 0.1 ppb) Carboxylic Acids, Sulfonic Acids, Alkynes, Cyclic Hydrocarbons, Amines, Pyridines, Pyrroles, Furans, Ethers, Esters, Epoxides, Alcohols, Ozone, Aldehydes, Ketones, Thiophenes, Mercaptans, Chlorinated Hydrocarbons Solid Waste 0.1 to 1 percent Al, Ca, Fe, K, Mg, Si, Ti, Cu, Na 5 Cu, Ni, Co, Pb, As, U Ten part per million (10 ppm) One part per million (1 ppm) Zn, V, F, Br, Cl One part per billion Ba, Co, Mn, Rb, Sc, Sr, (1 ppb) Se, Be One-tenth part per billion: (0.1 ppb) Eu, Hf, La, Sn, Ta, Th. Sb, 6 SECTION II INTRODUCTION The overall technical objective of this project is to provide a prelimi¬ nary evaluation of the potential pollutants which could be generated in all variations of coal-fired fluidized bed combustion processes. In per¬ forming a preliminary environmental assessment, one's major role in a sense is to serve as a "devil's advocate" with respect to pollutant genera¬ tion. The aim is to focus attention on potential environmental problems as early in the development cycle as possible. This provides maximum lead time to gather the technical data on which decisions regarding con¬ trol technology or process modifications (should they be needed) can be based. This particular project was divided into three phases. The first phase was to provide a review of fluidized bed combustion tech¬ nology and to identify conceivable pollutants. An overview of fluidized bed combustion systems appears in Section III. The identification of conceivable pollutants was based on the materials involved (coal, bed material, etc.) and pertinent process parameters (temperature, pressure, etc.); a list was made of possible pollutants which could be emitted from coal-fired fluidized bed combustion processes. This list, which is provided in the Appendix, served as the focal point of the second phase, which consisted of an engineering and scientific evaluation to determine at what concentration levels the pollutants identified in phase 1 could exist in the various effluent streams (air, water, solid waste) of a fluidized bed combustion system. The methods used to estimate the various pollutant levels for the major unit operations of a fluidized bed 7 combustion system are described in Section IV which forms the main body of the report. The primary emphasis was on trace elements, inorganic compounds, organic compounds and particulates; SO 2 and N0 X have already received considerable attention. This evaluation included calculations or extrapolations based on available laboratory or pilot plant data. In cases where data were lacking, simple kinetic or thermodynamic estimates were used. No emission measurements were made as part of this program. The third phase, which is described in Section VI, involved the providing of suggestions for appropriate control measures to reduce any emissions which may exist at undesirably high levels. This project was part of EPA's Program for Environmental Characterization of Fluidized Bed Combustion Systems. It is the first step in a long-range environmental assessment program being carried out by a number of con¬ tractors which will include: more detailed process stream information, a comprehensive analysis of emissions, control technology assessment and environmental impact analysis. The overall program strategy has been recently described.^ REFERENCES 1. Henschel, D. B. The U.S. Environmental Protection Agency Program For Environmental Characterization of Fluidized Bed Combustion Systems. (Presented at the Fourth International Conference on Fluidized Bed Combustion. Sponsored by U.S. Energy Research and Development Administration. McLean, Virginia. December 1975.) 8 SECTION III OVERVIEW OF FLUIDIZED BED COMBUSTION SYSTEMS A fluidized bed boiler can be simplistically depicted as an enclosed cavity containing boiler tubes and a bed of granular solids, to which fuel is added. As shown in Figure 1, the solids are supported on a grid at the bottom of the boiler through which combustion air is passed at high veloci¬ ties, typically 2 to 5 feet per second. The solids are held in suspension by the upward flow of the air and a quasi-fluid is created which contains many properties of a liquid. The most important liquidlike property to the boiler designer is the fact that the bed material is exceptionally well mixed and flows throughout the system without mechanical agitation. Fluidized bed coal combustion systems for the production of steam and/or electricity have several advantages over conventional combustion sys¬ tems. 1 ’^ Capital and operating costs should be lower for the following reasons: • High heat transfer coefficients and volumetric heat release rates will reduce the boiler size by 1/2 to 2/3 or more compared to a conventional unit.-*- • Capital costs will be reduced due to the size reduction and the potential for shop fabrication instead of field 1 ? 1 construction • First generation plants^*- > 2,3,4,5 should achieve fuel-to- electricity efficiencies comparable to the best con¬ ventional systems (36 to 38 percent) 1-5 while second gen¬ eration plants may achieve higher efficiencies of 40 to 47 percent. 2 The higher efficiencies will be achieved by operating at higher steam temperature and pressure than is possible with a conventional system. 9 flue gas to Figure 1. Schematic of fluidized bed boiler 10 Flue gas emissions of S0 o and NO will be reduced without the use of aux- Z x iliary control equipment. A limestone bed will collect the SO^ while the NO emissions will be low due to the inherent ability of fluidized bed x y combustors to achieve low NO emissions. The mechanism for reducing NO x x emissions is not completely understood. It involves low thermal fixation of atmospheric nitrogen as well as low conversion of fuel nitrogen to NO . X Cooling requirements will be reduced for those fluidized bed systems that use gas turbines to extract part of the energy. Potential problem areas include environmental impacts of limestone regenera¬ tion and/or solid reuse disposal and unknown flue gas emissions. Potential pollutants which could form in FBC flue gases but have not yet been measured are discussed in depth in later sections of this report. TYPES OF FLUIDIZED BED COMBUSTION SYSTEMS A fluidized bed combustion system is defined by a combustor, an energy use or conversion cycle, and a spent stone regeneration or disposal method, as well as numerous auxiliary systems. Some major options regarding combustor design conditions include operation of the bed at atmospheric versus elevated pressure and the presence or abscence of heat transfer surfaces in the bed. Energy may be generated as process steam or converted to electric energy through steam turbines, gas turbines, and combined sys¬ tems. Stone processing options include: (1) no processing (once-through operation) and (2) one-step or two-step regeneration followed by sulfur recovery, sulfuric acid recovery, or SO^ scrubbing. Figure 2 presents some of the most probable options for selected variables in a complete fluidized bed combustion system. Three basic combustion types can be defined by the operating pressure and excess air. These are atmospheric combustion, pressurized combustion (15 to 25 percent excess air) and adiabatic combustion (pressurized and 300 percent excess air). The energy output from the atmospheric combustor will be steam from tubes in the bed for process use or conversion to 11 COMBUSTION SYSTEMS OL > o UJ co > CM O X <-> UJ a: - a: UJ > CO o o i a; 3 z UJ -J CO o < _i CO a U UJ UJ u a 3 a: x >- U a m h X UJ UJ u. .c o z < _i N CO CD 3 H CO o < Ul o in 3 CE a X 3 a 0. O o UJ — CM ro ♦ ir> co CO >- UJ CJ o p o X X _l < > u o UJ z z 2 UJ 0. y -J UJ >• o t- o J- UJ X 1 o < < o z X X CO H UJ UJ UJ UJ -J jjg z UJ z UJ _l a UJ o u> >- 3 U. UJ UJ A. u U. UJ X X iZ u> CO CO o u x M- O o UJ O CD if) > ^ < ' *2 H ® z oo UJ — H z a: o o o t- co X CD 28 -* o 3 co ? _i < < 2 O O UJ X — H CM U. CO O z a: O 3> f= $ «q cr II 85 g j Q < UJ (r UJ u U- Q. o ° m — CD £ < UJ “ _l a o . - c n o >- o in uj -* _j < 5t CD Z o ~ a ? cr 3 o •H 4-) a. o S a) ■u CD >s CO c o CO 3 XI E o a x) 00 00 00 00 0O 00 o X CL g 01 H o o o O CD X t—( rH o o o o o o X X 00 00 00 CO i—i i —i r-. o o o o x Liquid 00 r-~ o 1 1 I 1 >1 1 III II II 1 II o m m 3 3 o o X in cd 1 i 1 i 1 1 C\1 1—1 l CN 1 1 o o 1 IN O' CJ • • • • • • • rH rH rH o o rH co rH 1 CO 3 o CO CO 3 o X o C\| O'. o o o i—1 o i O' o O' O CN o CN CN 1 o • • r-H • • • • • • • • H rH o o o r-H rH o o o o o o o O O CN X 3 3 3 3 X 3 X O 3 X 3 i 3 •H G CJ 1 — 1 rH to 4J 3 3 4-> co cd 3 w 3 3 G X 3 X) o 3 3 O rO 3 3 G •H 3 CJ 0) a 3 3 3 3 3 3 X x o r-H 3 3 X X 43 G Cn 3 3 3 3 3 3 00 3 G XXl 3 43 G 01 3 G •H 3 X G G O 3 X 4-4 G cd G o 3 s 3 O o 00 0) o 3 X CO o •H 4-1 3 i — 1 o 3 3 •H o G •H o X 4-) 3 cd kO 3 CL 3 3 3 3 3 rH o 3 O O 3 O 3 1—1 01 4-1 3 3 4-4 T—1 3 4H CL 40 3 3 rH 4-4 44 4-J Cl X 3 4-1 43 X 0) •H X 0) cd 3 G O ±J CL 3 3 3 4-4 3 3X CO 00 cd co 3 o CO 3 4-J O 3 3 CL 3 3 G X O i—1 O 3 0) 3 3 3 O 3 O 3 O 3 3 3 3 O 3 3 3 O o CJ rH •H cd 4-J 4-4 OO O 3 X 04 O 4h 4-J O 4-J X 4-4 G 3 3 O CL 4-J 3 X X 3 4—J •H 3 •H o •H 3 O •H 4-4 o •H 3 4-4 3 X •H 3 3 X) 3 3 CO 4-J 01 X CO 43 1 - 1 31 X co 3 1 - 1 X 3 x 3 3 3 3 N 'w' 3 3 3 3 3 o CO co i — 1 3 co o 3 co o 3 4-4 3 3 3 3 3 •H 3 X 4-> 0) o 4-J •H •H 0) •H 3 3 3 3 •H 3 3 3 3 O •H 3 O 3 X 3 4-J G cd 4-4 4-) cn G rH cl 4-1 G 3 4-J •H CM G 3 •u 4-4 3 G 04 G 3 X 00 •H rH 3 o 3 3 cd a) 3 3 3 kO 3 X •H 3 3 O 3 3 i — 1 G a r-H 3 i — 1 o 1 — 1 i — 1 3 O rH 3 O rH •H 3 3 4-J 3 3 rH 3 cd G 3 3 a cd cd 3 X 3 3 3 3 X 3 3 rH CL S 3 3 G 3 4-1 3 X o o •rH O •H o o •H O O •H O o •rH •H •H 3 o o o o o o 3 o o o o 1 o o o o o • • • • • • • o o o o o o o rH X m c^ 00 rH i—1 CM t—1 X CO 00 O' CO O' 3 o o 1—1 CM 00 rH X o CO 4-J o CM o CM o o 1 o rH o rH o o o • • • • • • • • • • • H o O o O o o o o o o o O OJ 3 1 <1- 3 1-1 OJ O 3 rH H O •H 3 rH 3 4J ca O CJ X + iH • CJ CJ rH >x rH 3 ca 0 OO X CJ OJ X 0 a 3 00 0 0 •H c ca CJ pu X rH ca >x •H 3 •H rH 3 E 3 /-~N X CJ 1 —1 ca a E 3 3 4-> 01 01 3 0 0 O CJ 0 0 0 0 3 O 3 4-J ca 4-1 c OJ 4-1 X m 4-» E 0 E CM E 3 H 4—1 4-J 4-J 0 5 ca £ 0 c^ •u 4-J 3 u 0 0 0 0 4-J •H 3 ca ca /T-V o E 0) rH <—1 4-1 0 ca 0 X rH 3 >x CM 3 • 3 ca E 0 E rH 1—1 ■U 0 ca iH Xl u X ca 4-J 3 4-1 0 X rH 03 3 3 CO 4-J T) ca Oh + iH 0 3 •H ca • 3 X ca X X 3 4-J ca 3 PO CJ ca • iH 0 rH •H 3 ca E O ca ca X •H ca 3 3 0 s—' + rH JO ca CD 3 0 4-> 3 0 ca 0) 3 •H 3 3 0J 3 w 3 ca X 3 ca O rH O 3 • a 4-J u 4-1 O 3 O 3 O 0) 3 rH 3 O' •H E 00 4-J •H 0 X 3 ca •H O. 0J •H O •H 0 4-J 0 E O 3 0 CM i-i 0 c U-J ca 0) C ca x 3 CJ 0 E ca 3 ca rH ca 4-J ca 4-J 0 4-J 4-J 0) u •H rH 0 ca c 33 3 3 •H ca 3 O ca 0 ca E 3 3 X 0 3 + 4-1 4-J rH ca •H 0 3 •H 1 O 0 4-J X •H 3 rH •H >x •H CM CM 4-1 3 X 4-J 4-J ca "O •H e e rH 3 rH 3 4J 4-J ca 0) g 0 O E 0 E E 3 w •H rH 2: ca 3 3 3 3 0 O 0 3 ca 3 X <3 4-J >. 0) OJ • O • ca X 4-J X CM 4-J ca 0) ca >X rH ca 3 0) rH X OJ CJ 0 0 4-J O 4J 3 0 3 H c JO 4-1 'w' 3 O O v-/ X E ca E CJ 3 X N-/ 3 4-J 3 3 4-J 3 3 ca 4-J rH CJ + 01 ca •H >x •H 0 O X •H •H •H 0 3 3 > X < u X 0 CJ CJ < < < X > M # # • # # # O' 0 rH CM CO i o co 3 e oo •H to G 03 CD J3 H 03 O O 4-1 O 03 3 3 O CX P G • X p 3 rH O 03 -C •H P P G G P X 0) £ fH 03 44 O O O 03 4-1 3 O 3 O CO (X 03 3 G 3 P O 3 X CO O 5 O O 00 f-l *• 4-1 00 CM i—I i—I CO 3 G P •• *ft cnj 3 cc G G P P 3 M P 3 a p 3 O X 3 O 03 rH G 3 3 G 3 -H CQ P 3 P 3 P CQ O o o CO > x x • r* o z 6 3 G CO X 00 p 3 -O 3 •H IX 3 O O 21 X w H co >- co Z o H CO da pa X o o CO Q CO w c_> PP M H O CD pd M CO pd M pd Q H M U da < Pd CJ Od da Q CO z CO < w Pd CO o O h4 M pH Pd Pd ,-j 3d C P^ M CO Od O Pd X H H < < X t3 QJ 3 C •H 4-1 c o u qj ■—' -O cO H O Pd .X pd 3 •u C 3 O I c o co 3 pa E O u CNI II o co 3 O 0) pd cO ■u id o o. co e o d cw CO co u T3 0J e 3 CO CO CO ON UP cn pd O pa QJ c o 4-) co cu E •H ►d CJ CO •H d OJ 4J co E cO O C o cO D- CO O OJ PO cO d O 4-1 CO CO pd o pa d CO O d Pd Pp Pd -o pd cO CU •O ■H X o id o E c o pa d cO c_> QJ d •H CO CO CO 01 u X 01 c 0) o d QJ a. 00 c 0) QJ 4-1 Pd toO •d Pd toC 22 'Twenty-five (25) percent excess air. • Coal Storage and Handling • Coal Drying • Cooling Operations The latter operations, coal storage, coal drying and cooling will be similar for all coal combustion systems. The total cooling required, how¬ ever, will vary with the energy feed to the steam turbine and the effi¬ ciency of the turbine. In most cases, cooling equivalent to about 60 per¬ cent of the energy input to the steam turbine will be required. Effluent Guidelines and Standards for Steam Electric Power Generation 16 will require recirculative cooling for almost all new plants. A utility will maintain a coal supply of 90 days while industry commonly maintains a supply of 30 days. 15 Wind erosion and handling cause a minor amount of particulate emissions. Leachates from coal storage can be a more serious problem and, as such, are regulated, 16 and control measures are applied. Particulates are emitted from many coal drying operations, but they can be controlled by conventional equipment. Water treatment is another conventional unit operation. Make-up cooling water is commonly treated by coagulation and clarification with aluminum and iron salts or lime. Polyelectrolytes are used to increase coagulation and sedimentation rates. In some cases, more sophisticated water treatment methods may be needed, depending on the water supply. Waste streams generated by cooling tower make-up water treatment are generally just water streams containing higher concentrations of impurities in the feed water. Even with recirculative cooling, large amounts of make-up water are needed and supply can be a problem. There are many solids transfer operations involved in operating a fluid¬ ized bed combustion system. Conventional practice is to use cyclones for initial solids separation in pneumatic transfer systems with fabric filters as secondary collectors. In the design of the PER system, ma¬ terial regenerators and transfer systems will use cyclones vented to the main high efficiency electrostatic percipitator as shown in Figure 3. 23 The PER system is slightly unusual in that all solid waste products are transferred dry. Wet sluicing is the most common practice at conven¬ tional plants, and ash slurring water discharge can be a major water pollution problem. No significant emissions will occur from coal crushing. In all fluidized bed systems coal will be crushed to about one-quarter inch and less. Coal crushing is a common and conventional unit operation, and the equipment as used by electric utilities is entirely self-contained. The primary focus of this study is the primary fluidized bed combustor, the carbon burn-up cell (if used), and the regeneration system (if used). The fluidized bed combustion process and factors affecting pollutant forma¬ tion will be discussed in depth in Sections IV and V of this report. These systems will be the most important potential emission sources. Potentially very large amounts of solid waste may be produced if regen¬ eration is not used. At a Ca:S ratio of 2:1, the PER system generates about 0.35 to 0.4 pound of waste per pound of coal burned. This is more than a conventional system and would amount to about 700,000 tons per year at a typical 500 MW plant. PRINCIPAL DIFFERENCES BETWEEN FLUIDIZED BED COMBUSTION AND CONVENTIONAL COMBUSTION Many of the differences and similarities between fluidized bed combustion systems and conventional combustion systems have been described in the previous paragraphs. The discussion here focuses mainly on the primary fluidized bed combustor and the carbon burn-up cell. Perhaps the most important difference from the standpoint of pollutant formation is the low combustion temperature in the primary combustor, most typically 1550 to 1650°F (850 to 900°C). Depending on furnace design, conventional systems may operate at temperatures of over 3000°F (1650°C). The lower tempera¬ ture of fluidized bed combustion contributes to reduced formation of NO x and improves the limestone bed reaction with the S0 9 , but could cause less 24 efficient combustion and possibly larger emissions of organics and CO. In most fluidized bed systems, material containing unburned carbon is recycled either to a carbon burn-up cell or back to the main cell. Con¬ ventional utility coal combustors operate with less than 1 percent unburned carbon losses at excess air in the range of 15 to 22 percent. 18 The low operating temperature in fluidized bed combustion also may reduce volatili¬ zation of trace metals as discussed in later sections. Excess air in most designs is very similar to conventional systems — 15 to 25 percent. Higher amounts of excess air in conventional systems lead to greater heat losses in the stack gas and lower efficiencies. The 300 percent excess air used in the adiabatic system greatly exceeds the amount used in any conventional system. That high level of excess air is more typical of refuse incinerators. Other differences in the design of fluidized bed systems are mentioned at appropriate points throughout this report. REFERENCES 1. Henschel, D. B. The Environmental Control Potential of Fluidized-Bed Coal Combustion Systems. (Presented at the Second Seminar on Desul¬ furization of Fuels and Combustion Gases. Washington, D.C. Novem¬ ber 11-20, 1975.) 2. Archer, D. H., D. L. Keairns, J. R. Hamm, et al. Evaluation of the Fluidized Bed Combustion Process. Volume I, Summary Report. Westing- house Research Laboratories. Prepared for U.S. Environmental Pro¬ tection Agency, Research Triangle Park, North Carolina. Publication Number APTD 1165, PB 211 494. November 1971. 3. Keairns, D. L., D. H. Archer, E. J. Vidt, and E. F. Sverdrup. Evalua¬ tion of the Fluidized Bed Combustion Process. Volume III. Westinghouse Research Laboratories. Prepared for U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publica¬ tion Number EPA-650/2-73-048c. December 1973. 4. Environmental Impacts, Efficiency, and Cost of Energy Supply and End Use, Volume II. Hittman Associates. National Technical Information Service, Washington, D.C. Publication Number PB 239 159. January 1975. 25 5. Steam Electric Plant Construction Cost and Annual Production Expenses. Federal Power Commission, Washington, D.C. Publication Number FPC-S- 237. April 1974. 6. Energy Conversion From Coal Utilizing CPU-400 Technology. Combustion Power Company Contract No. 14-32-001-1536. Office of Coal Research, Washington, D.C. R&D Report No. 94 - Interim Report No. 1. January 1975. 7. Keairns, D. L., et al. Fluidized-Bed Combustion Utility Power Plants - Effect of Operating and Design Parameters on Performance and Economics. Proceedings of Third International Conference on Fluidized-Bed Com¬ bustion. U.S. Environmental Protection Agency, Washington, D.C. Publication Number EPA-650/2-73-053. December 1973. 8. Multicell Fluidized-Bed Boiler Design, Construction and Test Program. Pope, Evans and Robbins, Inc. Contract No. 14-32-0001-1237. Publica¬ tion Number PB 236 254/AS, Office of Coal Research, Washington, D.C. R&D Report No. 90 - Interim Report No. 1. August 1974. 9. Robison, E. B., et al. Study of Characterization and Control of Air Pollutants From a Fluidized-Bed Combustion Unit - The Carbon Burnup Cell. Pope, Evans and Robbins, Inc. Prepared for U.S. Environmental Protection Agency. Publication Number APTD 1170. February 1972. 10. Energy Users Report. August 7, 1975. 11. Hoke, R. C., et al. Exxon Research and Engineering Company, Linden, N. J., Proceedings of EPA Symposium on Particulate Control In Energy Processes, U-S- Environmental Protection Agency. Report No. EPA-600/ 7-76-010. 1976. 12. Reduction of Atmospheric Pollution, Volume I-III. National Coal Board, London, England. Fluidized Combustion Control Group. U.S. Environmental Protection Agency. Publication Numbers APTD 1082, APTD 1083, APTD 1084. September 1971. 13. Hoke, R. C., et al. Exxon Research and Engineering Company, Linden, New Jersey. Private Communication. 14. Smith, J. R., J. R. Hamm, D. L. Keairns. Pressurized Fluid Bed Boiler Power Plant Operation and Control. Combustion. January 1975. 15. Surprenant, N., R. Hall, S. Slater, T. Susa, M. Sussman, and C. Young. Preliminary Emissions Assessment of Conventional Stationary Combustion Systems. Volume II - Final Report. GCA/Technology Division, Bedford, Massachusetts. Prepared for U.S. Environmental Protection Agency. Publication No. EPA-600/2-76-046b. March 1976. 26 16. Effluent Guidelines and Standards for Steam Electric Power Generating Point Source Category. Fed Regist. October 8, 1974. 17. Heist, J. A., R. H. VanNote, J. W. Kluesener. A Demonstration of the Use of Properly Treated Sewage for Cooling Water at Fifteen Cycles of Concentration. (Presented at the Second National Conference on Water Reuse. Chicago, Illinois. May 1975.) 18. Steam. New York, Babcock and Wilcox. 1972. 27 SECTION IV POTENTIAL POLLUTANTS FROM A COAL-FIRED FLUIDIZED BED COMBUSTOR INTRODUCTION As mentioned earlier in Section III, the fluidized bed combustion system has been categorized into the major unit operations or activities listed below: • Fluidized bed combustion • Limestone regeneration • Solid waste disposal • Fuel storage and handling • Coal drying • Cooling operations. Potential pollution problems from each of these operations are discussed here as separate subsections. The major emphasis has been on analyzing the fluidized bed combustion unit — and to a lesser extent, the limestone regenerator. These are the operations which are unique to a fluidized bed combustion system. In a sense, the solid waste disposal associated with fluidized bed combustion may also be a unique operation in that the leach¬ ing properties of the ash may be different than that from conventional sys¬ tems, but to date, there is relatively little data available in this regard The environmental hazards which may result from fuel storage, coal drying, and cooling should differ very little from those also encountered in con¬ ventional combustion systems. 28 FLUIDIZED BED COMBUSTION Since S0 9 and NO formation in coal-fired fluidized beds has been exten- Z X sively investigated by the U.S. Environmental Protection Agency; Exxon; Argonne National Laboratory; Pope, Evans and Robbins, Inc.; Westinghouse; Battelle; and BCURA; among others, the primary interest here was the so- called "other" pollutants — namely, organic compounds, trace elements, inorganic compounds (other than S0 9 and NO ), and particulates. Based on available data and/or simple thermodynamics and chemical experience, estimates have been made of the concentrations of various elements and compounds in either the flue gas or the solid waste. In general, the estimates for the trace elements and particulates are probably good to within an order of magnitude, while those for the organic compounds are probably good only to within two orders of magnitude. Trace Elements Trace elements and their compounds are of concern because some of these materials can vaporize and exit with the flue gas. Because vaporized trace elements would be in the gas phase, they would not be captured by particle collection devices. There is also concern about the enrichment of trace elements on fine particles in combustion processes. Studies have indicated that certain elements can concentrate in selected size ranges of particulates. 1-3 For some elements, such as lead and cadmium, these sizes tend to be less than a few microns in diameter. Such small particles are of special en¬ vironmental concern because they are difficult to remove from the flue gas and, once emitted, they can be readily embedded in the lung. k Numbers refer to references at the end of Section IV, page 62. 29 In addition, trace elements in the solid residue from the process could present a leaching problem. This possibility is discussed in a later sub¬ section on solid waste disposal. To assess the importance of trace element emissions in coal-fired fluid¬ ized bed combustion, a "worst case" analysis approach was used. For both bituminous and lignite coals, ranges for the heat content and the concen¬ tration of trace elements were obtained. Assuming that all of the elemental material would exit with the flue gas as either a vapor or a particulate, a "worst case" emission factor was calculated (lbs/10 6 Btu/hr) using the lowest heating value and highest trace element content of the coals. (Calculations to be described later in this section indicated that the trace element contribution from the limestone sorbent is insignificant.) Based on this emission factor, stack gas concentrations were calculated and diluted by a factor of 10 3 to account roughly for dispersion in the atmosphere. These "ambient" concentrations were then compared to indus¬ trial hygiene threshold limiting values (TLVS). Although industrial hygiene threshold limiting values cannot be used to assess the absolute environmental impact of pollutants, they do provide a useful framework in which pollutants can be rank ordered according to their toxicity. Any element whose predicted "worst case" ambient concentration was within a factor of 100 of the industrial threshold limiting value was considered to be potentially harmful. This safety factor of 100 was arbitrarily chosen to account conservatively for the effects of long-term exposure. (Industrial hygiene limits are usually based on exposures to healthy adults over an 8-hour period.) The results of this investigation suggested that the following trace elements could pose potential environmental problems: Be, As, Pb, Cr, V, Cl, U. But again, the results are based on worst case estimates. The evaluations from this study also suggested that in comparison with a conventional coal combustion unit, fluidized bed combustion (FBC) might reduce some trace element emissions. Therefore, the application of FBC technology could present an attractive alternative for the control of some 30 trace element emissions. The major parameters in FBC which will tend to mitigate trace element emissions are expected to be: • Temperature - The range of combustion temperatures in FBC is 700-1000°C, with most units operating between 800-900°C. In conventional combustion, temperatures are of the order of 1650°C. Therefore, many of the trace elements that oxidize, enrich, or vaporize in a conventional system should be less active in a FBC unit. • Pressure - FBC units can operate at pressures as high as 20 atm. High pressure can raise the melting and boiling points of the trace element compounds and therefore may potentially decrease air emissions. However, no measurements have been undertaken to con¬ firm this particular aspect of FBC technology. • Coal size - The coal size used in FBC is larger than that used in conventional combustion. Therefore, fine particulate emissions may be reduced. • Limestone sorbent - Limestone can modify trace element emissions by acting as a sink for certain trace ele¬ ments (e.g. , Fe) . Trace Elements in Coal and Limestone and "Worse Case" Emission Rates - Coal contains, at trace concentrations (-100 ppm or less), virtually all elements below atomic number 92. For this discussion, it has been as¬ sumed that the coals to be used predominantly in FBC are lignite and bituminous coal. The ranges of concentrations of trace elements in these coals in terms of lb/10 6 Btu are given in Tables 2 and 3. Included in Tables 2 and 3 are the maximum emission factors calculated from the lowest heating value of the fuel and the highest trace element content. Because it is impossible to deal with all variations in trace element concentration, this "worst case" emission factor will be used to determine which trace elements may be of environmental concern. The concentration range quoted spans 90 percent of the variations found in different seams. Not all trace elements are included in Tables 2 and 3, because composition data are not available for some elements. 31 Table 2. ESTIMATED TRACE ELEMENT "WORST CASE" EMISSION FACTORS FOR BITUMINOUS COAL a Average Minimum Maximum "Worst case" emission factor (no emission control) Heat content, Btu/lb 12,000 10,500 14,500 Sulfur, lb/10 6 Btu 1.67 0.58 3.75 4.29 Moisture, lb/10 6 Btu 0.33 2.50 15.0 17.14 Nitrogen, lb/10 6 Btu 1.08 0.83 1.33 1.52 Ash, lb/10 6 Btu 11.67 4.17 20.83 23.81 Minor elements, lb/10 6 Btu A1 1.42 0.44 4.31 4.92 Ca 0.25 0.10 5.33 6.09 Fe 1.55 0.29 6.42 7.33 Mg 0.06 0.01 0.50 0.57 K 0.16 0.04 0.69 0.79 Si 2.62 0.69 6.66 7.6 Na 0.04 0.03 0.46 0.53 Ti 0.04 0.01 0.62 0.70 Trace elements. Amount * 10“ 4 lb/10 6 Btu Sb 0.42 As 25 2.50 50.00 57.14 Ba 83.33 < 33.33 > 150 171.43 Be 2.08 0.50 6.67 7.62 Bi 0.08 Bo 41.67 3.33 166.67 190.48 Br 12.5 Cd 0.33 Cl 1250.0 Cr 11.67 3.33 41.67 47.62 Co 3.33 0.42 8.33 9.52 Cu 10.83 2.50 33.33 38.09 F 66.67 8.33 158.33 180.95 Pb 7.50 3.33 11.67 13.33 Mn 41.67 3.33 75.00 85.71 Hg 0.17 0.06 0.42 0.48 Mo 3.33 0.33 7.50 8.57 Ni 11.67 1.67 33.33 38.09 Se 2.50 Te 0.25 Th 0.08 Sn 0.83 u 12.50 8.33 66.67 76.19 V 25.00 1.67 38.33 66.67 Zn 6.67 < 0.83 41.67 47.62 /.r 41.67 a Values taken from references 4 through 11. 32 Table 3. ESTIMATED TRACE ELEMENT "WORST CASE" EMISSION FACTORS FOR LIGNITE 3 Average Minimum Maximum "Worst case" emission factor (no emission control) Heat content, Rtu/Lb 6,900 6,300 7,500 Sti 1 f nr , Lh/ L0 r ’ Bt u 1.01 0. 29 4.35 4.76 Moisture, lb/10 f ’ Btu 50. 72 28.99 57.97 63.49 Nitrogen, lb/10 f ' Btu 1.45 0.72 2.17 2.38 Ash, lb/10 1, Btu 14.49 7.25 21.74 23.81 Minor elements, lb/10 6 Btu AL 1.15 0.46 3.0 3.28 Ca 3.31 1.85 8.0 8.79 Fe 0.92 0.15 5.17 5.67 Mg 0.78 0. 39 1.83 2.0 K 0.08 0.017 0.23 0.26 SI 1.55 0.61 4.09 4.48 Na 1.61 0.03 4.83 5.21 Ti 0.04 Trace elements. Amount x 10 _ ' lb/10 6 Btu Sb 0.58 As 11.59 Ba 405.79 376.81 434.78 476.19 Be 2.17 0.14 5.80 6.35 Bi 0.14 Bo 173.91 115.94 289.86 317.46 Br C<1 0.29 CL 1449 28 72.46 2898.55 3174.60 Cr 10.14 4.35 28.99 31.75 Co 4.35 1.01 10.1 11.1 Cu 21.74 4.35 23.19 25.40 F 86.96 Pb 10.14 Mn 55.07 4 3.98 66.67 73.02 Hg 0.16 0.10 0.13 0.14 Mo 2.46 0.14 4.93 5.40 N1 10.74 2.17 21.74 23.81 Se 1.88 To 0.16 Th < 0.14 Sn 1.3 0.14 8.12 8.89 u 217.39 72.46 347.83 380.95 V 23.19 7.25 43.48 47.62 7.ii 17.39 0 Zr 14.49 'vnlucs taken from references 4 through 11. 33 The use of limestone or dolomite also adds to the trace element loadings in FBC. However, there are very few analyses available of trace elements contained in the sorbents used in FBC. Table 4 contains data for repre¬ sentative types of dolomite and limestone. Also given in Table 4 is the average trace element content of the fuels. The data show that the trace element concentration of limestone is generally equal to or less than that of the coal feed. The mole ratio Ca:S for most FBC processes will be about 2. Because the coal sulfur content will be approximately 3 percent by weight, one is dealing with weight ratios of coal to limestone on the order of 13 to 1. Therefore, trace element loadings from the limestone sorbent should be small compared to the fuel. Furthermore, because trace elements in the sorbent are contained in a limestone matrix as the fairly unreactive oxide or carbonate, they will probably have much lower emis¬ sion factors than the more volatile forms of trace elements (such as sul¬ fides) encountered in coal. Geochemical Classification of Elements in Coal - There have been very few studies performed of trace element emissions from FBC of coal. Therefore, to estimate these emissions, a comparison with trace element behavior in conventional combustion is useful. The primary concern is to identify which trace elements will be emitted as vapors and which will be enriched on small particulates. One can use as a basis for these predictions a geo¬ chemical classification of elements. 12 This classification scheme has been used successfully in predicting the emissions from the Allen Steam Plant. 2 Element volatilities and enrichment behavior have also been determined on the basis of elemental or oxide boiling points. 3 In the geochemical scheme, trace elements in coal are separated into four classes: I. lithophile, II. chalcophile. III. volatile elements, and IV. unclassed elements exhibiting the properties of either Class I or II. Trace elements in each class are listed in Table 5. 34 Table 4. TYPICAL VALUES OF TRACE ELEMENTS IN LIMESTONE AND COAL (ppm) Klement Argonne dolomite 3 Tymochtee dolomite* 3 Limestone 0 Lignite'* Average or typical bituminous'* Afl 1.9 b 0.566 ± 0.17 < 6 8 30 Bn 5 b 30-300 280 100 Be 2 C < 2 1.5 2.5 Br 2 C 6.75 ± 1.A < 0.3 15 Cd lA b < 0.3 0.2 0. A Ce 0.9 C < 3 Cu 1.03 ± 0.21 < 2 3 A Cr A.23 ± 0.85 < 20 7 1A Cs 0.A39 ± 0.091 < 0.06 Dy Eu 0.0598 ± 0.013 < 1 Fe 5.6 x 10 3 3 32A0 ± 650 200-2000 63AA 1.86 x 10 4 Hf Hg 0.2 K A.6 x 10 3 ^ 2180 ± AA0 100-1000 0.1 La 3. A a 0.3-3 551 1927 Mn 55 3 A2 ± 8.A 6-60 38 50 Na 368 3 303 + 61 10-100 4 1 x 10 A81 Ni < 6 7 1A Rb 12.2 + 2.5 < 2 Pb < 3 7 9 Sb 0.0527 ± 0.015 < 0.3 0. A 0.5 Sc 1 . 5 3 0.952 ± 0.19 < 0.3 Se < 3 1.3 3 Sm 0.658 t 0.13 < 1 Sr 130 ± 29 100-1000 Ta Te < 0.3 0.11 0.3 Tb 2.81 + 0.63 < 0.2 Th 0.58 ± 0.12 < 0.1 0.1 Yb Zn < 30 12 8 U 2.23 ± 0.A5 < 0.6 150 15 V 0.06-0.6 16 30 Reference 13. Reference 1A. Reference 15. References A through 11. 35 Table 5. THE SEPARATION OF ELEMENTS IN THE GEOCHEMICAL CLASSIFICATION SCHEME 12 Class I Class II Class III Class IV Al Mn As Hg Cr Ba Rb Cd Cl Cs Ce Sc Cu Br Na Co Si Ga F Ni Eu Sm Pb U Fe Sr Sb V Hf Tu Se K Th Zn La Ti Mg Trace elements listed in Class I are lithophiles and are associated with aluminosilicate minerals in coal. As such, they are high boiling com¬ pounds and do not decompose on combustion. They usually melt and coalesce to form the fly ash and slag. Elements in this class are not enriched during combustion. Class II elements are generally present in coal as sulfides. These sul¬ fides themselves may be fairly volatile or, upon combustion, the sulfides decompose and the elements themselves are produced in the vapor phase. These volatile sulfides or elements can then condense on the extensive surface area presented by particulates thus leading to a surface enrich¬ ment. This enrichment is usually most prevalent in the fine particle fraction (i <_ 3 ym) of the total particulate loading. Generally, elements could be placed in Class II if: Enrichment factor = (x) fly ash/(x) fuel > 3 (EF) where (x) is the concentration in weight percent. Class III elements boil below the furnace and flue gas temperatures and can exit from the stack as vapors. 36 Of the Class IV elements, only Cr and Ni tend to show chalcophile (or volatile) characteristics. Fate of Trace Elements in FBC - There have been several studies of the behavior of trace elements in FBC. The most complete study has been performed on Argonne National Laboratory's bench scale pressurized com¬ bustor. ’ ’ Trace element studies have also been initiated at Exxon's pressurized bench scale combustor. 14 Table 6 is a comparison of Exxon's and Argonne's results, and the agreement is encouraging. Table 6. COMPARISON OF EXXON AND ARGONNE DATA ON TRACE ELEMENT RECOVERIES 3 Element Recovery, % Exxon Argonne As 86 85 Br no data 18 Fe 80 100 K 75 90 Mn 96 130 Na 88 96 Sc 85 97 a Recovery = percentage of element entering combustor that can be accounted for in solids leaving combustor. Based on their bench scale tests, workers at Argonne as shown in Table 7, have indicated that FBC, compared to conventional combustion, has a pro¬ pensity for reducing trace element emissions. 16 Argonne has also performed several experiments to determine the vola¬ tility of trace elements during coal ashing. These experiments involve subjecting a previously formed low temperature ash to elemental analysis after exposure to successively higher temperatures."*^ Their results 37 indicate that Fe, Al, Na, K, Mg, Ca, Ti, Zn, Mn, Ni, Co, Cu, Cr, Li, and V all remain in the ash at FBC temperatures which suggests that, if enrichment or volatilization of these elements does occur, it must result from reactions of compounds in the coal and not the ash. Caution is required, however, in extrapolating these results to larger systems because the heating rates may not be typical of those encountered in commercial combustion systems. Table 7. PROJECTED ATMOSPHERIC EMISSIONS OF TRACE ELEMENTS FROM CONVENTIONAL AND FLUIDIZED BED COMBUSTORS EXPRESSED AS A PERCENTAGE OF THE ELEMENT ENTERING THE SYSTEM 16 Element Conventional combustion 3 Fluidized bed combustion Hg 90 80 F 90-100 (estimated) 40 Br 100 (estimated) 65 As 50-60 15 Pb 0-60 0-20 Be Not available 20-40 Sc 10 0 Cr 0 25 Co 10-20 0-20 Na 20 5 K 30 10 Fe, La, Mn 0 0 Projected from data in the literature on trace element emis¬ sions from conventional power plants - see reference 16 for further references. "Worst Case" Emission Factors - The preceding discussions indicate that elements in Class I should not be enriched or volatized during FBC or conventional combustion. Therefore, using the "worst case" emission factors (i.e., all of the trace element in the feed is emitted to the 38 stack as particulate) from Tables 2 and 3 and an assumed collection effi¬ ciency of 99 percent for a particulate control device (e.g., an electro¬ static precipitator or fabric filter), pollutant loadings at the top of a stack can be calculated for the trace elements in Class I. The results of this calculation are presented in Table 8. The emission rates in Table 8 have then been divided by 1000 to account for atmospheric dilution. To assess the environmental impact of these estimated ambient concentrations, occupational hygiene threshold limiting * values (TLV) are used. As shown in Table 8, these are arbitrarily divided by 100 to provide an environmental index which accounts for long-term exposure. While these cannot be used to assess the absolute environmental impact of pollutants, they do provide a useful framework in which pollu¬ tants can be rank ordered according to their degree of toxicity. The minor elements present in coal, with the exception of Na, are all in Class I and therefore are not enriched or vaporized in FBC; hence, it can be concluded that FBC of coal should present no problems for atmospheric trace element emissions of Class I elements. Trace element emissions for elements in Classes II, III, and IV are more difficult to predict because of their possible volatility which could lead to vapor phase emission which escape through a particulate control device or to enrichment on fine particulates which are less efficiently collected. Table 9, for example, which provides cut-off data on the boiling points of a number of compounds which could either be contained in coal, or formed as combustion intermediates, can be used to provide some insight on the par¬ titioning of elements between the solid (ash) and vapor phase. Table 10 presents "worst case" emission estimates for Class II, III, and IV elements calculated in the same manner as the Class I elements shown earlier in Table 8. Because of the potential volatility of these Class II, III, and * American Conference of Governmental Industrial Hygienists — Threshold Limiting Values. 18 39 Table 8. COMPARISON OF ESTIMATED TRACE ELEMENT CONCENTRATIONS (CLASS I ELEMENTS) WITH AN ENVIRONMENTAL INDEX BASED ON THRESHOLD LIMITING VALUES A B C D Emission factor after control device Estimated ambient Environmental Ambient concentration/ (assuming 992 control concentration: index (D > 1.0 indicates a potential of particulates) (A/1000) (TLV/1000) environmental problem) Element Bituminous 3 . . . a Lignite Bituminous Lignite Bituminous Lignite A1 72.4 48.31 0.075 0.048 _ _ Ba 0.25 0.70 2.5 x 10" 4 7.0 x 10~ 4 5 x 10 3 0.05 0.14 Ca 89 129 0.089 0.129 0.05 1.8 2.6 Ce 0.04 0.02 4 x 10 -5 2 x 10 -5 — _ _ Co 0.01 0.01 1 x 10' 5 1 x 10 -5 1 x 10 -3 0.01 0.01 Eu 1.4 x 10 _3 2.8 x 10 -4 1.4 x 10' 5 2.8 x 10 -6 - - - Fe 108 84 0.108 0.084 0.15 0.72 0.56 Hf 5.6 x 1C' 3 -4 5 x 10 5.6 x 10 -5 5 x 10“ 6 0.005 0.01 0.001 K 11.6 3.8 11.6 x 10~ 3 3.8 x 10' 3 — — — La 0.052 0.004 5.2 x 10' 5 4 x 10 -6 - - - Mg 8.4 29.5 0.008 0.03 0.10 0.08 0.3 Mn 0.1 0.1 0.001 0.001 0.05 0.02 0.02 Rb 0.21 0.002 2.1 x 10" 4 2 x 10” 6 _ _ — Sc 0.03 0.01 3 x 10~ 5 1 x 10" 5 - - - SI 112 66 0.11 0.06 0.1 1.0 1.6 Sn 0.003 0.01 3 x 10" 6 1 x 10 -5 0.02 0.0002 0.0005 Sr 0.03 0.66 3 x 10' 5 6.6 x 10 -4 — — — Ta 0.0003 — 3 x 10" 7 — 0.05 6 x 10 -6 — Th 0.0001 0.0002 1 x 10 -7 2 x 10 -7 1 x 10' 3 -4 1 x 10 -4 2 x 10 Ti 10.3 1.6 0.01 0.001 0.1 0.1 0.01 a Based on References 4 through 11. 40 Table 9. BOILING POINTS OF COMPOUNDS OFTEN FOUND IN COAL Boiling or sublimation points < 1000°C Boiling or sublimation points > 1000°C A1C1 3 A1 4 C 3 Sb 2 0 5 Sb 2°3’ Sb 2 S 3 As 2 S 3 , As 2 S 2> AsH 3 BeCl 2 BeO Cr(CO) 6 CrCl 3 Co(CO) , 6 CoCl 2 CuS Cu 2 0, CuCl FeCl 3 , Fe(CO)^ FeCl 2 , FeO PbCl 2 Pb 3 0 4 , PbS Hg, HgCl 3 MoS 2 Mo 2 0 3 , Mo0 3 Ni(CO) 4 , NiCl 2 NiO Se0 2> SeCl^, Se, SeS 2 TeCl 2 BaO, BaCl 2 CdS, CdO CaO, CaC 2 MgO, MgCl 2 MnCl 2 KC1, K o S0, 2 4 NaCl Ti0 2 V 2°5 41 Table 10. COMPARISON OF ESTIMATED TRACE ELEMENT CONCENTRATIONS (CLASS II ELEMENTS) WITH AN ENVIRONMENTAL INDEX BASED ON THRESHOLD LIMITING VALUES v | W| 00 Q| o o 1^ vO o o <~>l com w E ^ c *•>. o 1 MO E E m C ^ ^ o > U X »-4 0J H > -o ^ C C UJ -fH o *-H X o o c o «l o c rH O I rH O H X X O X o rH X o fH X o d o i—4 X I I o o rH X X o m <\ O <0 o «o o. x: ~ un td •HEP O O 04 (T3 O d vO o o SO CH oo oc e D. <0 a s r j a. co to m co 00 r-4 X c_> 42 IV elements, no degree of control can be suggested a priori based on total particulate. Accordingly, we have indicated the degree of control that would be necessary to reach an acceptable environmental index as defined in the table. Based on the "worst case" analysis, the elements listed below could be of concern. The key question is in what chemical form will these elements appear in either the combustion bed or in the flue gas. Where information of this type is available, it has been indicated. • Arsenic (As) The volatility of As depends on the Ca content of the coal. With a large Ca content. As may be bound as the arsenate/arsenite and not volatized as AS 2 O 3. 1 This is the reason for the high retention of As when using a limestone/dolomite sorbent. On the basis of the emission factor for As in Table 10, however, As could still be emitted in harmful quantities. For example, in the case of bituminous coal, the uncon¬ trolled emission factor is equivalent to approximately 10 ppm As (by weight). • Beryllium (Be) Be is highly toxic (TLV = 0.002 mg/m 3 ). Its behavior in FBC is not yet understood. Argonne data suggest that beryllium is not enriched ; 16 however, as shown in Table 10, there is some evidence that some Be O 1ft could exist in the gas phase. BeC^ boils at 519 C, hence, if Be is present as the chloride, it could vaporize. Therefore, mass balances of Be should be of high priority in FBC systems. Uncontrolled emission factors could be as high as 1 ppm Be by weight as indicated in Table 10. • Lead (Pb) In conventional combustion lead is definitely enriched in fine particulates . 23 Pb 02 is not stable at combus¬ tion temperatures (decomposes at 288°C). PbO is stable to 882°C, but its vapor pressure is low (< 10 mm Hg ). 20 PbCO^ decomposes at 316°C. 213,22 Pb may exit as elemental lead, PbS, PbC^, or PbSO^. 43 Nickel (Ni) Ashing experiments suggest that nickel is not volatized at FBC temperatures. 17 The presence of highly toxic nickel carbonyl is possible because of the high CO con¬ tent in FBC flue gas. Its presence has been postulated in conventional combustion. 24 Because the suggested atmospheric limit for the control of exposure to nickel carbonyl is quite low, 0.3 ppb (2.1 yg/m 3 ), 25 nickel carbonyl should be studied experimentally. In fluidized bed gasification experiments, it has been found that 75 percent of the Ni in the fuel is tied up in the stone. 2 ® Similar data are not available for coal combustion. • Uranium (U) Thermodynamic data show the oxide to be the stable form. 22 Uranium is not volatized in conventional combustion but there are no data on enrichment factors. Because of the concentrations of U in lignite coal, its pathways in FBC should be checked experimentally. As indicated in Table 10, uncontrolled emission rates could be as high as 11 and 56 ppm in bituminous and lignite coals respectively. • Vanadium (V) The enrichment of V is small for conventional combustion, and hence should not be of concern in FBC. Thermodynamic data show that the oxide form is favored over the sulfate. In gasification experiments, Westinghouse has found 100 per¬ cent retention of V on the spent limestone. This V concen¬ tration in the stone can range up to 1 weight percent. 26 This phenomenon should be studied for coal combustion. • Halogens: Fluorine, Chlorine, Bromine Argonne has shown that F is captured by the limestone/ dolomite bed by demonstrating that F retention was 5 to 23 percent with sorbent present. It is probably captured as CaF 2 (m.p. 1360°C). 20 Cl may be trapped as CaCl 2 (bp 1593°C). Chlorine will be emitted in the exhaust gas as HC1 43 and possibly in the form of NaCl. However, HC1 emissions have not been noted at high concentrations. British studies show that Cl content of the exhaust gas is 20 ppm (W/W). 29 44 Br may be captured in the bed as CaBr2 (bp 806°C). Its lower boiling point could explain the higher emission factor noted by Argonne for Br than F. Influence of Selected Process Options on the Fate of Trace Elements - This section discusses process variations and their possible effect on trace element emissions. Because this is not a well documented area, most of the discussion comprises recommendations for further research. • Pressure and Temperature Pressurized FBC could modify trace element behavior. The vaporization of various elements or compounds could change significantly as pressure is increased. These changes could result from phenomena such as boiling point suppression, or shifts in chemical equilibrium concentrations. The difference in potassium loadings from the Argonne and British experiments may be a manifestation of the effect of pressure. • Carbon Burn-Up Cell (CBC) Fly ash from the fluidized bed can contain as much as 20 percent unburned carbon. In some FBC designs, fly ash from the cyclones in the flue gas system will be returned to a carbon burn-up cell (CBC). The CBC can affect trace metal emissions because, it operates at a higher temperature than the fluidized bed (1093°C ver¬ sus 816°C). At the higher temperature it is expected that some of the more active trace elements (Hg, Cl, Se, etc.) will be revolatilized or those in the unburned fly ash volatilized. Gaseous Organic and Inorganic Compounds Data on the specific organic or inorganic compounds formed during coal com¬ bustion are scarce. Analysis of combustion gases is usually limited to species such as CO, CO 2 > SO 2 , NO^ and, in some cases, total hydrocarbons. In addition to these combustion "end products," however, an extremely wide variety of other organic compounds could also form - especially during 45 transient operating conditions which often foster incomplete combustion. Predicting these products in the case of coal combustion is a difficult task. Detailed thermodynamic or kinetic calculations are of limited value because, for the most part, the actual reacting species are a matter of speculation and the extent to which true equilibrium is attained is often questionable. Some insight into potential organic pollutants, however, can be gained on the basis of a simple coal combustion model, the present understanding of chemical reactions in fluidized beds, and some simple thermodynamic calculations. Simple Combustion Model - Conceptually, the combustion of a coal particle can be viewed in two steps as shown in Figure 4. Figure 4. Schematic representation of coal combustion 46 In step I, volatile hydrocarbons are ejected from the coal particle; they mix with oxygen and burn in a cloud surrounding the particle. Both 31 32 theoretical and experimental evidence indicate that this first step is completed in several tenths of a second or less. After devolatiliza¬ tion is complete, oxygen molecules attack the remaining char in step II, with burning here usually completed in times on the order of several seconds. The solid char in step II is predominantly carbon, although 33 Sternling and Wendt indicate that much of the chemically bound nitrogen in coal also winds up in the char. This char nitrogen probably burns to form NO, although the combustion mechanisms for heterogeneous combustion of nitrogen are not nearly as well studied as those for carbon. The main reaction product from heterogeneous carbon combustion is CO. The CO subsequently burns within the bed or in the freeboard to form CO^. Some of these char particles are ejected from the bed but they are captured by the flue gas cyclones and returned for combustion in the carbon burn-up cell (CBC). Consequently, the crucial step for forming the more complicated organic species would be step I. Volatile Products From Coal Decomposition - Potential organic pollutants can form because some of the products from step I could pass through the bed either completely or partially unburned. To estimate the extent to which this can occur, one must (1) determine the chemicals released from coal devolatilization, and (2) estimate the extent to which they will survive in a hot fluidized bed. Figure 5 provides a convenient summary of the types of reactions that occur during coal decomposition. Coal has no unique structure; generally, it is viewed as a network of aromatic carbon compounds interspersed with various heterocyclic compounds containing oxygen, nitrogen, or sulfur. These heterocyclic compounds are less stable than aromatics and during pyrolysis these bonds tend to break first, as shown in Figure 5. 47 48 Figure 5. Schematic representation of coal pyrolysi The product distribution in coal pyrolysis is temperature dependent. At temperatures on the order of 900°C (similar to those proposed for fluid¬ ized bed combustion) the predominant reactions are ring closures, con¬ densation, and aromatization reactions. The main products tend to be polynuclear ring compounds with occasional nitrogen, oxygen, or sulfur 34 substitution and simple compounds such as , t^S, NH^, CH^, etc. The overall synthesis of polycyclic organic compounds is shown schemat¬ ically in Figure 6. 35 Coal combustion can be a source of these com¬ pounds 36 ’ 37 since, as seen previously in Figure 5, many of the products of coal decomposition are equivalent to the advanced stages of pyrene synthesis, shown below. (The extent to which these compounds might escape unaltered from the hot reaction zone of a fluidized bed will be discussed in the next section.) naphthalene 6ENZ0 (0) PYRENE Figure 6. Pyrolytic synthesis of B(a)P 35 49 During periods of start-up or shut-down, products from low temperature pyrolysis might also be encountered in a fluidized bed. These compounds tend to be single aromatic rings or heterocyclic compounds with alkyl 34 side chains. Examples shown below are substituted benzenes, phenols, pyridines, thiophenes, quinolines, where R = CH^, etc - OH Using the above chemicals as starting materials, estimates can be made of which classes of compounds might survive in the reactive environment of a fluidized bed. Combustion of Hydrocarbons - The predominant reaction of the volatile hydrocarbons, of course, will be combustion; the main question is to what extent combustion will be complete. As mentioned previously, both theoretical and experimental evidence indicate that combustion of vola¬ tiles will occur on the order of 100 ms. This is more than an order of magnitude less than the residence time in the bed; hence, there is cer¬ tainly ample time for complete combustion. However, phenomena such as bubbling, slugging, uneven gas distribution, or localized reducing areas near the points of fuel injection could produce oxygen deficient zones from which unburned or partially burned hydrocarbons could escape. The extent to which this will occur will depend on the boiler design and could vary from reactor to reactor. Experimental tests indicate that the amount of unburned hydrocarbon in the flue gas has a strong depen- 38 dence on the amount of present. Some of the experiments are sum¬ marized in Figure 7. Note that at 1 percent 0^ in the flue gas (5 percent excess air), hydrocarbon concentrations are greater than 2000 ppm. With 50 HYDROCARBON(METHANE) CONCENTRATION IN FLUE GAS,ppm 3000 2500 2 000 1500 1000 500 0 Figure 7. Variation in hydrocarbon concentration with flue gas oxygen content in the fluidized bed module (FBM)-^ ILLINOIS CEO' 0 gyRVEY L ; -'' £4137 51 EXCESS AIR, percent 0 9 flue gas concentrations on the order of 3 percent (17 percent excess air), hydrocarbon concentrations are reduced to 50 ppm. It is important to note, however, that these experiments were performed on a fluidized bed module (FBM) designed primarily for investigating heat transfer phenomena. The unit had a limited freeboard and was not necessarily designed for optimum combustion conditions. Hence, the results may be upper limits, but they do provide some insight into the generation of un¬ burned hydrocarbons in FBC. Products of incomplete combustion - Some insight as to the chemical com¬ position of the products of incomplete combustion is provided below; the next section indicates methods which can be used to estimate concentra¬ tions of potential compounds. • Hydrocarbons Fluidized bed reactors have long been used in the petroleum industry to "crack" or thermally decompose high molecular weight hydrocarbons. The process is shown schematically below: HOT SURFACE (700 - 850°C) h 2 ,ch 4 ,c 2 h 4 , c 2 h 6 etc. In these reactors, sand is often used as the bed material, sometimes with added catalysts such as oxides of V, Ni, and Co. In some respects, a coal- fired fluidized bed combustor may be similar to a com¬ mercial cracker (including trace metals present as catalysts) ; hence one might expect that most unburned hydrocarbons would be extensively "cracked" by the time they leave the bed. Theoretical estimates of the first order rate constant for the cracking of a hydrocarbon of molecular weight 225 indicate that half the material will be cracked in 30 ms at 727°C.^^ In some respects, this could be a significant advantage of fluidized bed coal combustion versus conventional coal combustion. Coal-fired units (particularly small industrial or residential units) are significant sources of polycyclic organic compounds. In a fluidized bed, the increased gas-solids contact may enhance the tendency for these species "crack" to form compounds such as CH^, C 2 H 4 , C2H6, etc. If, in fact, this is so, the probability of finding or ganic sulfur and nitrogen compounds such as thiophenes, mercaptans 52 carbazoles, etc. should also be very small. These species will most likely decompose to form small hydrocarbons and species such as H 2 S, HCN, and COS. Pyridine, for example, decomposes readily at tempera¬ tures on the order of 900°C to form HCN and hydrocarbons • Oxygenated Hydrocarbons Organic chemicals, such as ethylene oxide, phthalic an¬ hydride, naphthaquinones and aromatic carboxylic acids, can also form via partial oxidation of hydrocarbons. An example is shown below: 0 II 0 naphthalene phthalic anhydride These reactions, which are often used in organic syn¬ theses, usually proceed only under very controlled process conditions and at temperatures on the order of 200 to 400°C; hence, their occurrence in a combustor operating around 900°C seems very unlikely. There could be some possibility of their occurrence during start-up and shut-down when temperatures are lower, but it seems unlikely. A tentative identification of diphenylene oxide in particulates from acetylene-oxygen and ethylene oxygen flames, however, has been recently reported.^0 • Carbon Monoxide (CO) Carbon monoxide forms in the bed, but it usually burns there or in the freeboard to form CO 2 . At 10 percent ex cess air, CO usually drops below 1000 ppm in atmospheric pressure fluidized bed units and below 200 ppm in pres¬ surized units. High CO levels usually indicate signi¬ ficant gas-bypassing within the bed. • Soot Fuel-rich conditions lasting for only a short period of time can lead to the formation of a fairly signifi¬ cant quantity of soot which can take a long time to burn away again. 41 Soot particles tend to be exceed¬ ingly small and would not be collected by the cyclones or the electrostatic precipitator. 53 Carbides It is also conceivable that metal carbides could form in a fluidized bed combustor. Calcium carbide (CaC 2 ), in particular, is formed by heating lime and carbon. Most carbides would not be volatile at the prevailing temperatures; hence, they should remain in the bed. If formed, they could present problems upon disposal of the stone since carbides release C 2 H 2 upon hydrolysis. • Halogenated Hydrocarbons There is considerable chlorine content in coal itself plus there are some indications that NaCl may be added to fluidized beds to enhance SO 2 scrubbing;^ hence, the possible halogenation of hydrocarbons must be con¬ sidered. At temperatures encountered in a fluidized bed combustor, however, chlorinated hydrocarbons most likely will not be stable. The effect of chlorination on the pyrolysis products of British coals has been investigated^ and it has been found that at high tem¬ peratures, chlorinated tars are not produced. Practi¬ cally all of the chlorine appears as HC1. Concentration Estimates of Organic and Inorganic Compounds - Equilibrium calculations can be used to predict the composition of the combustion gases The most commonly used method is the minimization of the chemical system's 44 free energy. This method can provide useful concentration estimates but is probably most useful for predicting trends or concentration ratios. In the case of fossil fuel, one usually uses the elemental composition as the starting material and the combustion products are usually limited to simple molecules (3 atoms or less). Calculations of this type have not been extensively applied to coal combustion, al- 45 though they have been used in coal gasification analyses. Because gasification is essentially incomplete combustion of coal, these analy¬ ses can be used to provide very conservative upper limits for the con¬ centrations of "reduced" species such as COS, NH^, and H^S which may be present in combustion gases. Estimates based on gasification equilib¬ rium calculations are shown in Table 11. 54 4 Table 11. CALCULATED EQUILIBRIUM CONCENTRATION FOR SELECTED SPECIES PRODUCED BY INCOMPLETE COMBUSTION OF COAL 45 Coal analysis: Oxygen present: Temperature: C - 68.5%, H - 5.3%, 0 - 8.5%, N - 1.4%, S - 4.1% 59% of stoichiometric requirements 7 60°C Species N 2 CO C0 2 CH. 4 h 2 s COS NH 3 C(g), HCN, CS 2 c 2 h 2 , c 2 h 4 , (cn) 2 s 2 , so 2 , so 3 NO, N0 2 Mole fraction 0.36 0.28 0.15 0.06 0.10 0.06 C.008 0.0005 0.0007 As the amount of oxygen in the combustor is increased, the concentration of "reduced" species decreases. The equilibrium calculations include a mass balance for each of the elements; hence, an extrapolation of the calculations to excess air levels on the order of 20 percent, where S 0 2 is approximately 500 ppm, indicates that compounds such as H 2 S, COS, and CS 2 should decrease by about two orders of magnitude. Similar considera¬ tion sould apply to "reduced" nitrogen compounds; hence, a conservative upper limit for the concentration of compounds such as H^S, COS, CS 2> S0 3 , HCN, (CN) 2 , and NH 3 » under typical combustion conditions, is 1 ppm, which does not pose significant environmental problems. 55 Free energy minimization calculations for the more complicated hydrocarbons (e.g., polynuclear aromatics) are impractical because of the complexity of the chemicals involved. To estimate the concentrations at which these types of compounds might exist in the FBC flue gas, one can use empirical correlations between benzo(a)pyrene and CH^ concentrations from measurements in conventional coal-fired combustion systems. Fig¬ ure 7 indicates the variation of total hydrocarbons (as CH^) as a func¬ tion of 0 concentration from one set of fluidized bed combustion ex- ^38 periments. Under normal operating conditions, about 3 percent 0^ in the flue gas (20 percent excess air), the concentration of hydrocarbons (as CH^) is about 100 ppm (volume/volume - V/V). Although emissions can often vary between different fluidized bed systems, 100 ppm provides a convenient reference value. Previous measurements with conventional coal-fired systems indicate that compounds such as benzo(a)pyrene are typically 10 times less than the concentration of total hydrocarbons 46 47 as CH^, ’ Using our reference value of 100 ppm (CH^), this implies that in a fluidized bed system polynuclear aromatic hydrocarbons (PAH) could exist in the flue gas at concentrations (V/V) on the order of 1 part per billion (ppb). Considering that flue gases are typically diluted by about a factor of a thousand when they are emitted from the stack, this implies that ambient concentrations of PAH near FBC facili¬ ties would be on the order of 1 part per trillion. This corresponds to 3 about 0.6 ng/m which is roughly comparable to the natural background 46 47 concentration ranges found in rural areas. ’ Accordingly, it seems that polynuclear aromatic hydrocarbon concentrations should not be high enough to cause problems. Particulate Emissions Only limited data on particulate emissions from fluidized bed combustion are currently available. 4 ^“ 50 Preliminary data, using more sensitive particle sizing techniques than used previously, indicate that the mass median diameter of the flue gas particles (50 percent of the mass of the 56 particles are above that size) is about 7 pm in a pressurized system . 14 This means that significant concentrations of fine particles could exist in the flue gas. Further experiments on particle size distribution and chemical composition as a function of particle size should receive high priority. The following discussion summarizes currently available data on particulate loadings in FBC. Atmospheric Pressure Fluidized Bed Combustion (AFBC) - For the process parameters shown below, Argonne Laboratories measured an average dust loading leaving the secondary cyclone of 0.06 grains/scf and a maximum loading of 0.22 grains/scf. Flue gas flow rate: Coal feed rate: Additive feed rate: Primary cyclone: Secondary cyclone: Dust loading (combustor exit): Combined cyclone efficiency: 8-14 cfm 4-7.8 lb/hr 1 . 1 - 2 .3 lb/hr 6-5/8 in. diameter 4-1/2 in. diameter 0.16-1.78 grains/scf Approx. 90%. Figure 8 shows the particle size distribution obtained from these experiments. Experiments at the National Coal Board in England indicated dust loadings of 0.1 to 0.6 grains/scf when using primary and secondary cyclones having collection efficiencies of 90 percent at 10 pm . 49 These experiments also indicated that the particulate contained 5 to 15 percent carbon and 85 to 95 percent ash and additive. Pope, Evans and Robbins made initial investigations of particle size dis¬ tributions in FBC, as shown in Figure 9. In some of their experiments, NaCl (salt) was used as an additive to enhance SC^ removal; the addition of NaCl also affected the particle size distribution as shown in Figure 9. No mechanism explaining the influence of NaCl on particle formation was postulated. 57 001 o o o o o o oooooooooo (DONiOiO^flON — UJ M CO UJ _J o cc 2 3ZIS 310liavd a31V9ICINI NVH1 d31V3U9 !N30d3d 1H9I3M 3AllV3nWflO S° Figure 8. Typical particle size distribution of elutriated material collected in primary cyclone, secondary cyclone, and fil¬ ter bag during period of additive injection 48 Figure 9. Fly ash size distribution for Pope, Evans and Robbins, Inc., atmospheric pressure fluidized bed combustion (AFBC) 14 59 At the FBC demonstration plant currently under construction in Rivesville, West Virginia, Pope, Evans and Robbins will use a hot electrostatic pre¬ cipitator to reduce particulate loading below the EPA limit of 0.04 grains per scf (0.1 lb/10 Btu). A hot (approximately 316°C) electrostatic pre¬ cipitator is used because the high carbon content (as high as 20 percent) of the fly ash causes a high resistivity which makes operation of a cold precipitator inefficient. Pressurized Fluidized Bed Combustion (PFBC) - A comprehensive study of the influence of selected process parameters on grain loadings in pres¬ surized fluidized bed combustion was performed by workers at Argonne National Laboratories. 51 Their results for particulate loading as a function of fluidizing gas velocity and as a function of Ca/S mole ratio are shown in Figure 10. The Argonne experiments indicated that, after passage through two cyclones, the solids loading in the flue gas ranged from 0.3 to 2.1 grains/scf. By adding a final filter, flue gas loadings were brought below 0.04 grains/ scf — the EPA emission limit. The above results suggest that particulate removal devices, in addition to the process cyclones, normally used, will be required in order for fluidized bed combustion to meet EPA particulate emissions standards. As noted earlier, Pope, Evans and Robbins will use a hot electrostatic precipitator at their demonstration plant in Rivesville, West Virginia. Exxon is incorporating a granular bed filter in the pressurized mini¬ plant which shows promise of reducing fine particulate loadings substantially. 60 LOADING OF FLUE GAS LEAVING COMBUSTOR , grains/ft 3 Ca/S MOLE RATIO Figure 10. Solids loading of flue gas leaving the combustor in Argonne National Laboratories pressurized fluidized bed combustion (PFBC)-^ 61 REFERENCES 1. Kaakinen, J. W., R. M. Jorden, M. H. Lawasani, and R. E. West. Trace Element Behavior in Coal Fired Power Plant. Environ Sci Technol. 9:862, September 1975. 2. Klein, D. H. et al. Pathways of Thirty-Seven Trace Elements Through Coal-Fired Power Plant. Environ Sci Technol. 9:973, October 1975. 3. Natusch, D. F. S., J. R. Wallace, and C. A. Evans, Jr. Science. 183, 202. 1974. 4. Abennethy, R. F., M. J. Peterson, and H. Gibson. Major Ash Constitu¬ ents in the U.S. Coals. U.S. Bureau of Mines. R.I. 7240. 1969. 5. Magee, E. M., H. J. Hall, and G. M. Varga, Jr. Potential Pollutants in Fossil Fuels. U.S. Environmental Protection Agency, Research Tri¬ angle Park, North Carolina. Publication Number EPA-R2-73-249. June 1973. 6. Surprenant, N., R. Hall, S. Slater, T. Susa, M. Sussman, and C. Young. Preliminary Emissions Assessment of Conventional Stationary Combustion Systems. Volume II - Final Report. GCA/Technology Division, Bedford, Massachusetts. Prepared for U.S. Environmental Protection Agency. Publication No. EPA-600/2-76-046b. March 1976. 7. Steam. New York, Babcock and Wilcox. 1972. 8. Devitt, T. W., R. W. Gerstle, N. J. Kulujian. Field Surveillance and Enforcement Guide: Combustion and Incineration Sources. U.S. En¬ vironmental Protection Agency, Research Triangle Park, North Carolina. EPA Publication Number APTD-1449. June 1973. 9. Kessler, T., A. G. Sharkey, and R. A. Friedel. Analysis of Trace Elements in Coal by Spark-Source Mass Spectrometry. U.S. Bureau of Mines, Pittsburgh, Pennsylvania. R.I. 7714. 1969. 10. Coal-Fired Power Plant Trace Element Study, Volume IV, Station III. Radian Corporation. Prepared for U.S. Environmental Protection Agency, Region VIII, Denver, Colorado. September 1975. 11. Steam. New York, Babcock and Wilcox. 1972. 12. Mason, B. Principles of Geochemistry, 3rd Edition. New York, John Wiley and Sons. 1966. p. 57. 13. Vogel, G. J. et al. A Developmental Program on Pressurized Fluidized Bed Combustion. Quarterly Report. Argonne National Laboratories. Publication Number ANL/ES-CEN-1009. October 1, 1974 - January 1, 1975. 62 14. Hoke, R. C., R. R. Bertrand, M.S. Nutkis, D. D. Kinzler, L. A. Ruth and M. W. Gregory. Studies of the Pressurized Fluidized Bed Coal Combustion Process. Exxon Research and Engineering Co., Linden, New Jersey. Prepared for U.S. Environmental Protection Agency under Contract Nos. 68-02-1451, and 68-02-1312. U.S. EPA Report, EPA-600/ 7-76-011. September 1976. 15. Jennings, K. Esso Research Ltd. Esso House, Steventon, Abingdon, Oxford, England. Private Communication. 16. Swift, W. M., G. J. Vogel, A. F. Panek. Potential of Fluidized Bed Combustion for Reducing Trace-Element Emissions. Argonne National Laboratories. (Presented at 68th Annual Meeting of the Air Pollution Control Association. Boston. June 15-20, 1975.) 17. Jonke, A. A. A Development Program on Pressurized Fluidized Bed Combustion. Monthly Progress Report. Argonne National Laboratories. Prepared for ERDA under Agreement No. 14-32-000101780 and for U.S. Environmental Protection Agency under Agreement No. IAG-D5-E681. September 1975. 18. Threshold Limit Values for Chemical Substances and Physical Agents in the Workroom Environment with Intended Changes for 1974. American Conference of Governmental Industrial Hygienists. Cincinnati, Ohio. Copyright 1974. 19. Cowherd, C., M. Marcus, C. M. Guenther, J. L. Spigarelli. Hazardous Emission Characterizations of Utility Boilers. U.S. Environmental Protection Agency. Publication Number EPA-650/2-75-066. July 1975. 20. Handbook of Chemistry and Physics, 5th Edition. Weast, R. E. (ed.). The Chemical Rubber Company, Cleveland, Ohio. 1975. 21. Vogel, G. J. et al. A Developmental Program on Pressurized Fluidized Bed Combustion. Quarterly Report. Argonne National Laboratories. Publication Number ANL/ES-CEW-1010. January 1, 1975 - April 1, 1975. 22. Lowell, P. S., T. B. Parsons. Identification of Regenerable Metal Oxide SO 2 Solvents for Fluidized-Bed Coal Combustion. Radian Cor¬ poration. Prepared for U.S. Environmental Protection Agency. Pub¬ lication Number EPA-650/2-75-065. PB 244 402/4G1. July 1975. 23. Trace Elements in a Combustion System. Final Report. Battelle- Columbus Laboratories. Prepared for Electric Power Research In¬ stitute, Palo Alto, California. EPRI Publication Nos. 122-1, 614-229, 3151. April 1975. 24. Davison, R. L., D. F. S. Natusch, J. R. Wallace, C. A. Evans, Jr. Trace Elements in Fly Ash - Dependence of Concentration on Particle Size. Environ Sci Technol. 8(13):1107, December 1974. 63 25. Medical and Biologic Effects of Environmental Pollutants - Nickel. Committee on Medical and Biologic Effects of Environmental Pollutants National Academy of Sciences, Printing and Publishing Office, Wash¬ ington, D.C. 1975. 26. Newby, R. A., D. L. Keairns, and E. Y. Vidt. Residual Oil Gasifica¬ tion/Desulfurization at Atmospheric Pressure - Clean Power From Existing Boilers. Westinghouse Research Laboratories. (Presented at 67th Annual AIChE Meeting. Washington, D.C. December 1974.) 27. Andren, A. W., D. H. Klein, and Y. Talrai. Selenium in Coal-Fired Steam Plant Emissions. Environ Sci Technol. 9(9):856, September 1975 28. Vogel, G. J. et al. A Developmental Program on Pressurized Fluidized Bed Combustion. Quarterly Report. Argonne National Laboratories. Publication Number ANL/ES-CEN-1008. July 1, 1974 - October 1, 1974. 29. Murthy, R. S. et al. Engineering Analysis of the Fluidized Bed Com¬ bustion of Coal. Final Report. Battelle-Columbus Laboratories. Prepared for U.S. Environmental Protection Agency, Control Systems Laboratory under Contract No. 68-02-1323, Task Order No. 6. May 1, 1974. 30. Robison, E. B. et al. Study of Characterization and Control of Air Pollutants from a Fluidized Bed Combustion Unit. Pope, Evans and Robbins, Inc. Prepared for U.S. Environmental Protection Agency, Division of Control Systems, Office of Air Programs under Contract No. CPA 70-10. February 1972. 31. Field, M. A., D. W. Gill, B. B. Morgan, and P. G. W. Hawksley. Combustion of Pulverized Coal. British Coal Utilization Research Association, Leatherhead, 1967. p. 178. 32. Beeston, G. and R. H. Essenhigh. The Kinetics of Coal Combustion: The Influence of Oxygen Concentration on the Burning-Out Times of Single Particles. J Phys Chem. 67:1349, 1963. 33. Sternling, C. V. and J. 0. L. Wendt. Kinetic Mechanisms Governing the Fate of Chemically Bound Sulfur and Nitrogen in Combustion. U.S. Environmental Protection Agency. Publication Number EPA-650/ 2-74-017. August 1972. 34. The Chemistry of Coal Utilization, Lowry, H. H. (ed.). New York, John Wiley and Sons, 1963. p. 340. 35. Edwards, J. B. Combustion: The Formation and Emission of Trace Species. Ann Arbor, Michigan, Ann Arbor Science Publishing, 1974. p. 64. 64 36. Smith, W. S., C. W. Gruber. Atmospheric Emissions From Coal Combus¬ tion - An Inventory Guide. Environmental Health Series. U.S. De¬ partment of Health, Education and Welfare. Public Health Service Publication No. 999-AP-24. 1966. 37. Cuffe, S. T., R. W. Gerstle. Emissions from Coal-Fired Power Plants: A Comprehensive Summary. Environmental Health Series. U.S. Depart¬ ment of Health, Education and Welfare. Public Health Service Publi¬ cation No. 999-AP-35. 1967. 38. Robison, E. B., A. H. Bagnulo, J. W. Bishop, and S. Ehrlich. Char¬ acterization and Control of Gaseous Emissions from Coal-Fired Fluidi¬ zed Bed Boilers. Pope, Evans and Robbins, Inc., Alexandria, Virginia. Prepared for Division of Process Control Engineering, National Air Pollution Control Administration (now U.S. Environmental Protection Agency). Publication Number APTD-0655, PB 198 413. 1970. p. 103. 39. Gibbs, B. M. and J. M. Bder. Concentration and Temperature Distri¬ butions in a Fluidized Bed Coal Combustor. In: Combustion Institute- European Symposium 1973, Weinberg, F. J. (ed.). London, Academic Press, 1973. p. 627. 40. Crittendon, B. D., R. Long. Diphenylene Oxide and Cyclopentacenapthy- lene(s) in Flame Soots. Environ Sci Technol. 7:743, 1973. 41. Field, M. A. et al. Op cit. p. 182. 42. Gordon, J. S., R. D. Glenn, S. Ehrlich, R. Ederer, J. W. Bishop, and A. K. Scott. Study of the Characterization and Control of Air Pollu¬ tants from a Fluidized Boiler - The SO 2 Acceptor Process. Pope, Evans and Robbins, Inc., Alexandria, Virginia. Prepared for U.S. Environ¬ mental Protection Agency. Publication Number EPA-R2-72-021. 1972. 43. Lowry, H. H. Op cit. p. 376. 44. Zaggeren, F. Van. and S. H. Storey. The Computation of Chemical Equilibria. London, Cambridge Press, 1970. 45. Stinnett, S. J., D. P. Harrison, and R. W. Pike. Fuel Gasification: The Prediction of Sulfur Species Distribution by Free Energy Minimi¬ zation. Environ Sci Technol. 8:441, 1974. 46. Hangebrauck, R. P., D. J. von Lehmden, J. E. Meeker. Emissions of Polynuclear Aromatic Hydrocarbons and Other Pollutants from Heat Generation and Incineration Processes. J Air Pollut Control Assoc. 14:267, 1964. 47. Hangebrauck, R. P., D. J. Von Lehmden, J. E. Meeker. Sources of Polynuclear Hydrocarbons in the Atmosphere. U.S. Department of Health, Education and Welfare. Publication Number PHS 999-AP-33. 1967. 65 48. Vogel, G. J. et al. Fluidized Bed Combustion. Annual Report. Argonne National Laboratories. Publication Number ANL/ES-CEW-1001. July 1968 - July 1969. 49. Reduction of Atmospheric Pollution (Via Fluidized Bed Combustion). Volumes I-III. Final Report. National Coal Board, London, England. Submitted to U.S. Environmental Protection Agency, Office of Air Programs. Publication Numbers APTD-1082-1084. September 1971. 50. Multicell Fluidized Bed Boiler Design, Construction and Test Program. Pope, Evans and Robbins, Inc. Publication Number PB 236 254/AS, Office of Coal Research, Washington, D.C. R&D Report No. 90 - In¬ terim Report No. 1. August 1974. 51. Vogel, G. J. et al. Reduction of Atmospheric Pollution by the Appli¬ cation of Fluidized Bed Combustion and Regeneration of Sulfur-Containinj Additives. Argonne National Laboratories (ANL/ES-1007). Prepared for U.S. Environmental Protection Agency. Publication Number EPA-650/2-74- 104. September 1974. 66 SECTION V POTENTIAL POLLUTANTS FROM AUXILIARY PROCESSES ASSOCIATED WITH FLUIDIZED BED BOILERS LIMESTONE REGENERATION The purpose of this discussion is: (1) to describe the methods of re¬ generating sulfated limestone produced in coal-fired fluidized bed com¬ bustion; and (2) to discuss the effects of regenerator operating vari¬ ables on possible pollutant formation. At present, both one- and two- step processes are being considered; the one-step process can operate at either atmospheric or higher pressure, while the two-step system is associated with high pressure operations. Effluents from the regeneration process include flue gas, particulate matter in the flue gas, spent stone, and any emissions from the associated sulfur recovery plant. Leachates resulting from disposal of the spent stone could also be important. One-Step Regeneration Calcium sulfate is reduced by CO and via the following reaction: ["H 1 CaSO. + 4 2 CO -*■ CaO 4- S0 2 + -h 2 o " C0 2 ( 1 ) The reaction proceeds rapidly at temperatures of about 1100°C and atmo¬ spheric pressure. (Regeneration is generally carried out at temperatures above 1100°C if operated under pressure (10 atm).) 67 The off gases from the regenerator can be sent to a Claus Plant to pro¬ duce elemental sulfur, a sulfuric plant or a scrubber. The following reaction occurs in the Claus Plant: 2 + S0 2 —*• 2 H 2 0 + 3S (2) A portion of the sulfur can then be reacted with methane to produce which is reused in reaction (2). Figure 11 provides an example of the overall process flow for a one-step regeneration scheme based on a design by M. W. Kellog. 1 Westinghouse Research Laboratories have also designed a one-step regeneration scheme which is similar to that of Kellog, except coal is used as the source of reducing gas for the regenerator, instead of natural gas. 2 (Natural gas is a cleaner fuel than coal and would be expected to reduce the impact of any possible pollutants from regenerator; however, in the future it is expected to become increasingly scarce and possibly be unavailable for such applications.) Two-Step Regeneration The first step of this scheme involves the reduction of CaSO^ to CaS with CO and H 2 : CaSO. 4 + 4 CO L H 2 *>• CaS + 4 (3) The reaction is generally carried out in the temperature range of 870 to 930°C and under high pressure. Calcium sulfide produced via reaction (3) is reacted with C0 2 and steam to produce CaCO^: CaS + C0 2 + H 2 0 ■+ CaCO^ + H 2 S . (A) 68 RECYCLE FROM SULFUR PLANT (/) < o o £ CO o I—< at a OJ X o CO c o •H ■u d 5-i a) d oj 50 OJ a OJ ■U cn i 2H 2 0 + 3S (Claus reaction) . (6) Figure 12 shows a process flowsheet for a two-stage regenerator designed by M. W. Kellogg. 1 It illustrates the material flows involved; a similar design has also been proposed by Westinghouse Research Laboratories. 2 Potential Pollutants From Limestone Regeneration Because the regenerator is also a fluidized bed, it will behave in the same manner as the combustor with respect to operating variables. The main differences between the fluidized bed boiler and the regenerator is both the chemical nature of the reactants and the reaction conditions; i.e., chemical reduction as opposed to combustion. The feed to the re¬ generator will consist of CaO, CaSO^, coal ash, and a reducing gas. Pro¬ posed sources of reducing gas have been natural gas or a mixture of CO, H 2 and CH^ from gasified coal (i.e., incomplete combustion of coal). The use of natural gas should pose no significant environmental problems, but the future availability of natural gas for this type of operation is questionable, because of projected fuel shortages. The use of a "coal gas" for the regenerator could pose environmental problems. In using coal, especially under conditions of incomplete combustion, one has to be con¬ cerned about potential pollutants such as: trace elements, polycyclic aromatic hydrocarbons, fine particulates and inorganic compounds such as nh 3 , h 2 s, COS, HCN, cn, cs 2 . Some insight to fate of trace elements can be gained from equilibrium free energy calculations. Reactions of trace elements will be considerably 70 oc o 44 Simple cyclone 7.5 22 43 80 90 Multiple cyclone (12-in. diameter) 25 54 74 95 98 Multiple cyclone (6-in. diameter) 63 95 98 99.5 100 For electric utility coal-fired boilers, electrostatic precipitators 3 (ESP) are the most common final control device. ESP's have proved to be reliable, have a low operating cost, and perform at reasonably high efficiencies. Venturi scrubbers are used much less frequently because the efficiency tends to be low for fine particles, unless high pressure drops and thus large amounts of energy are used. Fabric filters have 4 5 been used in only two utility coal boilers over the past 2 to 3 years, ’ with a third larger unit scheduled to be installed on a 350-MW boiler. Fabric filters are attractive because of very high efficiencies (99.8 45 7 to 99.9 percent) ’ at competitive costs. In the past, however, fabric filters were not used on coal-fired boilers because of questionable re¬ liability as a result of fabric deterioration. In addition, particulate control devices with maximum efficiencies of 99.9 percent were not needed and are still not required in many cases. New power plants are now being designed to achieve particulate control efficiencies in the 99 to 99.5 percent range; hence, a reasonable goal for FBC systems would be control efficiencies of at least 99 percent and probably 99.5 percent. In FBC, the most important coal ash properties affecting fine particulate control are the particle size distribution, the mass concentration, the SC >2 concentration, and the particle resistivity (for ESP). Generally 95 the smaller the particle, the more difficult it is to collect. The depen¬ dence of collection efficiency on particle size is most pronounced for scrubbers; it tends to be less restrictive for fabric filters and elec¬ trostatic precipitators. Figure 16 presents the predicted particle size 8 9 distribution at the inlet to the final control device ’ of an overall fluidized bed combustion system (shown previously in Figure 14). Large variations in size distribution may result from coal ash properties and the design of the cyclone precollectors. A composite of the particle size distribution from many conventional pulverized coal systems having only cyclone or similar mechanical control systems is also presented in Figure 16. Because the particle size data are extremely limited for fluidized bed systems, one can conclude only tentatively from Figure 16 that the particle size distribution in FBC is not radically different from a conventional system, although there could be more fine particles in atmos¬ pheric FBC emissions. The FBC data should be considered tentative because it is taken from relatively small test systems compared with the size of conventional systems. Particulate mass emissions reaching the final control device of a FBC system will probably be lower than in a conventional system. However, this concentration will depend on both the amount and size distribution of the particulates reaching the cyclone precollectors as well as the design of those precollectors. The amount of fine particulates reaching the cyclones can only be determined experimentally, and care should be exercised to distinguish between particle mass and particle number. The design plan for the Pope, Evans and Robbins - FBC system indicates an 3 expected mass concentration of 0.7 grains/ft reaching the final control device.^ Particulate emissions reaching the final control device in a 3 conventional system are higher - typically 2 to 3 grains/ft . 96 PARTICLE SIZE, MICROMETERS 9 10 20 30 40 50 60 70 80 90 96 WEIGHT PERCENT SMALLER THAN STATED SIZE Figure 16. Particle size distribution before final control device 97 The concentration of SC^, SO^ and water vapor in the flue gases can also affect the selection of particulate control equipment. Water in flue gases from combustion is primarily a result of the fuel hydrogen content and it produces a dew point of 50 to 65°C at normal excess air. However, as illustrated in Figure 17, the SO^ concentration (usually 1 to 2 per¬ cent of the S0 9 concentration) in a conventional coal-fired boiler raises the flue gas dew point.''"'*' Equipment designed to collect dry particulate (fabric filters and dry electrostatic precipitators) must operate above the acid dew point. Most conventional coal-fired plants maintain flue gas temperatures between 150 and 180°C to avoid corrosion problems. 12 Robinson et al. found that the Pope, Evans and Robbins - FBC system produced an SO^ concentration of 39 ppm when sorbent was not used, and no measurable SO^ when sorbent was used. (Note: These early SO^ results represent limited data and must be confirmed by further SO^ analyses on other fluidized bed combustors.) Such a very low SO^ concentration in the presence of sorbent, if confirmed, means that flue gases might be cooled to 95°C or below for dry particulate collection and increased heat recovery. The major problem in using fabric filters on coal-fired boilers has been SO^ and H2S0^ induced deterioration of the fabric. Therefore, fabric filter technology may also be readily applicable to fluidized bed combustion systems if the low SO^ concentrations are confirmed. The concentration of SO^ and the acid dew point will also affect the performance of electrostatic precipitators. Burning low sulfur western coal has resulted in decreased electrostatic precipitator performance, 13 and SO^ has been added to the flue gas as a conditioning agent. Electrostatic Precipitators - An electrostatic precipitator can be applied before or after the final flue gas heat recovery unit as previously illus¬ trated in Figure 14. In the design of the Pope, Evans and Robbins 30-MW demonstration plant, an electrostatic precipitator was selected to oper¬ ate at 315 to 370°C before the final heat recovery (a hot side precipita¬ tor) because of the high resistivity of the particulates at low tempera¬ tures."*^ Therefore, the unit must handle a larger gas volume than if low 98 DEWPOINT 330 310 — 290 — 270 — 250 230 210 90 0.0 0.1 1.0 10 S0 3 (H 2 S0 4 ) IN FLUE GAS, ppm 100 Figure 17. Dew point elevation as a function of S0 3 concentration^ 99 temperature operation were possible. With good design and operation, hot side electrostatic precipitation can collect fine particulates at high efficiencies as illustrated in Figure 18. 14 Similar performance can be attained with cold side precipitation when the aerosol properties are suitable. However, there have been some problems in designing precipi¬ tators to operate at high efficiencies. Data published by Benson and Corn^ show that among 14 large electrostatic precipitators, having an average design efficiency of 98.2 percent, the actual average operating efficiency was only 89.0 percent. Accurate data on particle size distri¬ bution, mass concentration, and in situ resistivity are needed for proper electrostatic precipitator design for FBC systems. Designing precipita¬ tors for FBC systems based on analogies with previous operating experi¬ ence in conventional combustion systems can be very risky. There have been many examples of precipitators designed to operated at 95 to 99 per- 16 cent efficiency which actually operated at 50 to 90 percent efficiency. Electrostatic precipitators have been very widely accepted by the utility industry because they have been applied over many years with minimal main¬ tenance problems. However, the size and capital cost rise sharply as design efficiencies are increased to 99 percent and above. Fly ash re¬ sistivity is a very important factor in ESP performance and may need more study with respect to fluidized bed combustion. However, methods to com¬ bat resistivity problems are being studied, including the use of hot side precipitators, and fly ash conditioning with SO^, ammonia, and sulfamic acid. If resistivity problems require operating temperatures of 370°C instead of 95 to 150°C, then many normally condensable organics and con¬ densable trace metal compounds will not be collected. Nevertheless, one significant advantage of electrostatic precipitators is that the power requirements are lower than other control devices. Wet Scrubbers - Most wet scrubbers that are commonly used collect par¬ ticulates primarily through inertial mechanisms and thus require large amounts of power to collect fine particulates at high efficiencies. Venturi scrubbers are used with very high pressure drops to collect fine 100 o luoojad ‘ADN3I0IJJ3 N0I103TI00 j-> >—i O 3 4-1 O 3 3 4- 1 •H XI CL 0) •H N i—I 3 3 •H CL 4-1 3 3 4- 1 c n C o o 5- > 4-1 X3 u a) 3 r*H l“H 1—I <3 3 4-1 3 3 -a c •H -H 3 4J X3 O 3 J3 4-1 3 3 3 ■H >-4 X) o c >4-1 -H 3 3 3 3 •H 3 3 c 5-> 3 3 •H 4J 3 3 •H E >4—1 3 >4-1 (-1 3 3 CL t—I 3 00 C C O -H •H 4J 4- 1 3 3 3 3 3 5- > CL >4-1 o XI 3 o k. a> CL d oc LU z UJ CL 99.0 99.1 99.2 99.3 9 9.4 99.5 99.6 99.7 99.8 09.9 c a V. D CL o UJ o UJ Ll li_ UJ PARTICLE SIZE , /x m Figure 20. Median fractional collection efficiency for 22 tests 4 104 Fabric filters have several advantages over other control systems. Col¬ lection efficiency does not depend on particle resistivity. Very high efficiencies of 99.4 to 99.9 percent for fine particulates can be routinely achieved by fabric filters. These efficiencies are out of the range of venturi scrubbers and can only be achieved by ESP through enlarged precipitators and sharply increased costs. One disadvantage of suggesting fabric filters for particulate control in FBC systems is that fabric filters have only recently been used on utility boilers and some operators may doubt their reliability and may be unwilling to use them. Particulate Control for Pressurized Combustion Systems There are two pressurized FBC systems under consideration. One operates at 15 to 25 percent excess air, while the other, the adiabatic system, operates at 300 percent excess air. Because the cost of large scale particulate control equipment is proportional to the actual gas volume handled, the capital and operating cost of particulate control equipment for an adiabatic system would be approximately triple the cost for a normal excess air system to achieve the same number of lb/10 6 Btu of emissions. Control devices for pressurized fluidized bed systems should operate at high temperature as well as high pressure if placed upstream of the gas turbine, in order to maximize energy recovery from the turbine. A par¬ ticulate control system operating at 800°C would fail to collect condensable materials in the flue gas that could be collected at temperatures of 375°C by hot electrostatic precipitators or at 95 to 150°C by other con¬ trol devices. Thus, it may be necessary to place a control device for particulates or other pollutants in the low temperature ducting downstream of the gas turbine in order to meet environmental requirements. 105 Particulate control equipment for high temperature and pressure operation is not yet commercially available and considerable development is required before any such equipment will be available. 10 The problem of controlling particulates at high temperature had led to a preliminary review of con¬ cepts for cooling, cleaning, and reheating the flue gases before applica¬ tion of the gas turbine. 20 Table 20 is a summary of potential high tem¬ perature particulate removal systems. Molten salt scrubbers cannot be applied to pressurized FBC systems because the alkali metal vapors would cause gas turbine corrosion. The Aerodyne Tornado cyclone has been seriously considered for applica¬ tion of FBC systems. Figure 21 depicts the operation of the Aerodyne Tornado cyclone which is claimed to have unusually good collection effi- 19 ciency for fine particulates compared to conventional cyclones. It appears unlikely, however, that cyclones — even of advanced design — will be able to achieve the high efficiencies for fine particulates that may be needed for operation in the mid 1980's era. Some type of device in addition to cyclones will be needed even to meet the current New Source Performance Standards for particulates from large coal boilers. Electrostatic precipitators could encounter severe operating problems at high temperatures and pressures, particularly in maintaining stable coronas. Fabric filters are also an option for high pressure operation, but there is still a great deal of development work needed. A basic problem is the need for a fabric capable of maintaining its physical integrity under fabric filter cleaning procedures, hence there is now considerable research underway to develop "advanced" fibers. Some of the most efficient fabrics are composed of very fine fibers similar to Brunswick Metal fibers (~ 8 to 12 microns in diameter). However, corrosion and oxidation of these metals is greatly enhanced at high temperatures. 106 Table 20. SUMMARY OF POTENTIAL PARTICULATE REMOVAL SYSTEMS 107 Exhaust (Clean Gas) Secondary Gas Inlet Inlet IDirty Gas) Falling Dust is Deposited in Hopper Secondary Air Pressure Maintains High Centrifugal Action Secondary Airflow Creates Downward Spiral of Dust and Protects Outer Walls From Abrasion Dust is Separated From Gas By Centrifugal Force, is Thrown Toward Outer Wall and into Downward Spiral , 19 Figure 21. Schematic representation of Aerodyne particulate separator # 108 Gravel bed filter technology is promising, but further development is also needed here. Exxon is in the process of acquiring a Ducon granular 21 bed filter for their FBC miniplant. The filter should be placed in operation during 1976. Data on gravel filters are very limited, and the results of the Exxon studies will be needed for a reasonably accurate assessment of particulate control efficiency. Control of Gaseous Emissions Some of the gases (other than SO and NO ) which could be present in quan- tities possibly high enough (e.g. > 10 ppm) to require some sort of control measures include: HCl, CO, unburned hydrocarbons (HC), and possibly COS. Each is discussed briefly below. HCl - HCl could be troublesome, especially if NaCl is used to enhance SO^ sorbent performance as proposed by some workers. If could also be a problem with high chlorine content coals. Aside from restricting the use of the above materials, the best control option for HCl seems to be scrub¬ bing with an alkaline spray. Such sprays are sometimes used in solid waste incinerators; candidate scrubbing solutions are carbonates and bi¬ carbonates. For maximum economy, the spray device would also have to be useful for particulate removal (as discussed more fully in the previous section). CO - High CO levels should not be routinely encountered based on available data from existing FBC units. High CO levels, if they do occur, should be able to be brought under control by increasing the extent of combustion within the burner. This can be accomplished in several ways: increasing excess air, increasing bed residence time, or by addition of secondary air. The choice between these options is not necessarily straightforward and will depend on the economics and design features of specific units. 109 Hydrocarbons - Hydrocarbons (HC) are also products of incomplete combus¬ tion and their emissions should be able to be reduced by the same techniques listed above for CO. Studies on HC levels as a function of excess air indicate very low emissions (< 50 ppm) provided excess air levels are kept above 15 to 20 percent. Should special circumstances warrant air levels below 20 percent, one also has the option of collecting some of these compounds as particulates after cooling the flue gas. Most of the high molecular weight compounds (especially polycyclic compounds) condense below 315°C. COS - COS emissions could conceivably be a problem; if present. COS can be removed via conventional H S scrubbing techniques such as the Rectisol process (a methanol scrubbing solution at -55°C), but these are not prac¬ tical for combustion gases. Alternate control methods, such as catalytic destruction, should be investigated. Zinc oxide is very effective in removing trace quantities of sulfur compounds from hydrocarbon gases at high temperature. Treatment with zinc oxide in the temperature range 215 to 425°C can produce a product gas having less than 0.1 ppm sulfur, as COS, H 2 S, or CS^ 22 POLLUTION CONTROL VIA PROCESS MODIFICATIONS: SOME CONSIDERATION BASED UPON FLUIDIZATION FUNDAMENTALS The following design and operating parameters can influence pollutant formation in fluidized bed combustion: • Bed depth • Bed and boiler tube geometry • Fluidizing grid design • Particle size • Fluidization velocity • Excess air • Mechanism of coal injection • Pressure. 110 Most of the above factors are interrelated; e.g., fluidization velocity and bed geometry affect the quality of fluidization, thus affecting gas- solid contact, heat transfer, and temperature distribution. The purpose here is to discuss the manner in which these parameters may affect pollu¬ tant formation. This information can then possibly serve as a starting point for pollution control via modification of appropriate process parameters. It is important to note that much of the discussion here is based on exist¬ ing models for low temperature fluidized bed systems. There may be diffi¬ culties in extrapolating some of the conclusions to the higher temperature regime of combustion; nevertheless, the discussion provides a useful starting point for considering some basic problems. Bed Depth Bed depth will influence gas residence time, bed pressure drop, and the quality of fluidization. Its most important influence on pollutant formation, if any, will probably be in terms of its effect on the residence 2 3 time of gas within the bed. Zenz has indicated the fraction of gas by¬ passing the bed decreases as bed depth is increased and attributed this to increases in the gas residence time. Because a bubble remains in the bed for a longer time as the bed depth is increased, it has more time in which to interchange gas with the emulsion phase. As gas exchange increases, bypassing is reduced. Increased gas exchange, should increase combustion; hence, it should tend to reduce the formation of compounds such as CO and unburned hydrocarbons. Increased gas exchange will enhance in situ cap¬ ture of SO^. As mentioned above, bed depth will also affect the quality of fluidiza¬ tion. Poor fluidization, such as slugging or bubbling, can increase particle elutriation and emissions of unburned hydrocarbons. According 9 / to the relationship of Broadhurst and Becker the minimum slugging 111 velocity varies inversely as the bed height to 0.9 power for settled bed- height to diameter ratios less than 3. For height to diameter ratios greater than 3, Steward predicts the minimum slugging velocity will be 25 independent of bed height. Bed and Boiler Tube Geometry Bed diameter and boiler tube configuration can influence gas mixing, slugging characteristics, and bubble flow. Zenz and Othmer report a twentyfold increase in mixing length (or intensity of mixing) due to 2 6 an increase in diameter from 1 to 6 inches. An approximately equivalent effect on eddy diffusivity was noted when diameter increased from 6 to 18 inches. Studies by the National Coal Board (London), however, have shown that bed cross-sectional geometry had little effect on their re- 27 suits. The effect of geometry most likely will vary from unit to unit; but, in general, the quality of fluidization increases with increasing bed diameter, which is encouraging since it implies that full-scale units may be better behaved than their bench-scale predecessors as far as potential pollutant generation via phenomena such as bypassing or elu- triation are concerned. The boiler tubes can serve to break up bubbles, providing smoother fluid¬ ization. However, if they are too densely packed, the tubes can prevent good mixing in the bed. The influence of boiler tube configuration on solids-gas mixing (or quality of fluidization) was observed by Exxon Research and Engineering. 28 They noticed flatter temperature profiles, an indication of good mixing, when the tubes were oriented vertically rather than horizontally. The configuration of the boiler tubes could present a problem in scale-up. Tube spacing, pitch, and size affect the mixing characteristics in the bed and thus affect the temperature profile. Packing tubes too closely will cause large temperature gradients. Exxon noted channeling after installing vertical tube bundles. 29 Upon examination of their tubes, they noticed distinct areas of corrosion. 29 112 This was attributed to impingement of high velocity gas (channels). They indicated, however, that channeling could be eliminated by proper design and operation. Fluidizing Grid Design Grid design is an extremely important factor in providing good fluidiza¬ tion. Unevenly distributed gas can cause channeling and subsequent de¬ activation of portions of the bed, causing potential release of substances such as S0_, NO and hydrocarbons. It is generally recommended that the z x grid pressure drop be 40 percent of the bed pressure drop to ensure uniform distribution. A better quality fluidization can also be obtained by 30 decreasing hole size and increasing their number. The effect of grid design is illustrated in Figure 22. 31 Sintered plate distributors, al¬ though providing the best fluidization, are generally not used in large commercial operations because of their fragility. However, the combina¬ tion of a sintered disk and a sturdy support presents an attractive option. Bottom has found that a large portion of the chemical reactions occurring 32 in a fluidized bed takes place in the grid region. Between 30 and 50 percent of the fast first order reaction which he studied took place in that zone. Cooke et al. noticed a rapid reaction between oxygen and coal in the first 9 inches of a fluidized bed carbonizer and very little above 33 this height. Behie and Kehoe also found the grid zone to be of major 34 importance for fast reactions. Several studies have shown that NO or o £ formation in fluidized beds occurs mainly near the distributor. * Very little additional chemical change occurs in the rest of the bed. Accordingly, changes in operating variables such as fuel/air ratio, which affect the chemistry in the distributor zone, strongly influence the com¬ bustion chemistry; hence, this magnifies the importance of grid design. 113 Poor quality; much fluctuation in density with channelling ~ and slugging f 1 Better quality; less Jluctuation in density, less channelling and slugging Single orifice plate Multiorifice plate Sintered plate Figure 22. Quality of fluidization as influenced by type of gas distributor 31 Particle Size The particle size distribution within the bed can influence not only the physical characteristics but also the chemical nature of the bed. It will influence the velocity needed for fluidization, the height of the dense and dilute phases, elutriation, the quality of mixing, com¬ bustion, and channeling and slugging phenomena. The size distribution may thus affect emissions of SO , NO and hydrocarbons. Z X In general, the quality of fluidization is better for smaller average particle sizes and larger particle size distributions. For high per¬ formance chemical reactors (reactors not necessarily designed for com¬ bustion reactions), the optimum diameter ratio of the largest to smallest 17 particle should be between 12 and 21. Elutriation is a function of both velocity and particle size. If the fluidizing gas travels above the terminal velocity of a given particle, any decrease in size of the particle will cause it to be carried out of the bed. This is assuming that there is no particle/particle inter¬ action which would cause the minimum fluidization velocity to differ from the particle's terminal velocity. Due to the complex nature of a fluidized system, however, particles whose terminal velocities are greater than the fluidizing velocity can be carried out of the bed. (A further description of elutriation is presented later in this section.) Vogel et al. studied the effect of coal feed particle size on S0„ reten- 38 tion. They observed similar retentions (about 78 percent) for coal feeds ground to -12, +50, -50, and -14 mesh. Retention was found to be better (81 percent) for the coarsest feed. They also found that addi¬ tive particle size had only a moderate effect on sulfur retention. 27 Wright has found that decreasing limestone size increases SO 2 reduction. A similar decrease in dolomite size had no effect on SO 2 reduction, which suggested that access to internal surface was not a limiting factor for dolomite. An additive too finely ground may have too short a residence time to achieve a high degree of SO^ removal. Thus, at high velocity, the effect of increased surface area may be counteracted by a decrease in residence time. Data obtained from British Studies support this claim. As mentioned earlier, particle size will influence phenomena such as bubbling, slugging, and channeling. The minimum slugging height is pre- 39 dieted to change with the -0.3 power of particle diameter. The velocity at incipient slugging increases as the square root of particle diameter for height to diameter ratios less than 3. For height to diameter ratios greater than 3, the minimum slugging velocity increases with an increase in minimum fluidizing velocity (and thus particle size). 27 115 Fluidization Velocity The fluidization velocity will influence gas and particle residence time, quality of fluidization, elutriation, temperature distribution, and chemistry. The quantity, size, and velocity of bubbles, as well as the amount of gas bypassing as bubbles, is influenced by the superficial ve¬ locity. (The fluidization velocity is often expressed in terms of the superficial velocity, which is the velocity the fluid would have if the reaction chamber were empty.) The fluid residence time can be easily determined if plug flow conditions are assumed and the volumetric flow rate and free volume (total volume less particle and tube volume) are known. However, due to the turbulent nature of fluidized bed, a great deal of backmixing occurs. Also, bubbles pass through the bed at velocities greater than those of the fluid in the emulsion phase. The fluid, therefore, cannot be characterized by a single residence time. In the absence of residence time distribution data, a residence time could be calculated for each phase within the bed; i.e., emulsion, bubble, and dilute. In the case of the emulsion and dilute phases, the residence time can be found by dividing the height of the phase by actual velocity (superficial velocity/void fraction). Bubble residence time can be calculated in the same manner; however, the bubble velocity must be known. Davidson and Harrison have presented a relation to determine bubble velocity as a function of bubble diameter and super- 40 ficial and minimum fluidizing velocities. In the absence of any other information, the maximum stable bubble diameter can be used to determine the bubble velocity. The maximum stable bubble diameter can be found 40 from knowledge of particle size, particle density, and fluid density. In general, an increase in velocity causes a corresponding decrease in residence time. In some pilot plant experiments, it has been noted that by increasing the superficial velocity, the primary combustion zone moved from an 116 area near the coal injection point to a zone near the top of the bed. 28 Increasing the velocity also had the effect of slightly increasing temperatures within the bed (presumably because increased velocity also meant an increase in excess air; hence, better combustion). Estimating the fraction of gas bypassing the bed is a relatively simple procedure if the bubble velocity and size are known. Relationships have been developed for calculating both bubble velocity and bubble size. 40 ’ 41 However, experimental verifications of these models were performed under conditions very dissimilar to those in a fluidized bed coal combustion unit. Hence, although methods are available for estimating the maximum stable bubble size, which in turn could be used to determine the fraction of gas bypassing, the results at best should be viewed as upper limits. Studies of heat transfer coefficients in fluidized beds indicate that the fraction of gas bypassing can sometimes range between 40 and 70 percent. The extent of gas and solid mixing is also affected by superficial veloc¬ ity. Internal solids circulation is extremely rapid and increases with velocity. During experiments in a 5-foot diameter, 32-foot deep bed of catalysts fluidized at 0.8 fps to determine mixing rates, it was found that 50 grams of powder was essentially completely mixed into 15 tons of 42 powder in less than 1 minute. Excess Air The amount of excess air will have a direct effect on the chemistry of the bed; a change in excess air will also affect the bed's physical properties indirectly at a given coal feed rate by changing the superficial velocity, the effects of which have been discussed previously. The effect of excess air on SC^ reduction with limestone, at constant superficial velocity, has been studied at Argonne National Laboratories. Sulfur reduction 117 increased from 67 to 75 percent when the oxygen level in the flue gas in¬ creased from 0.7 to 5.6 percent, which corresponds approximately to an increase in excess air from 5 to 35 percent. (As discussed previously in Section III, total hydrocarbon emissions can be reduced to less than 50 ppm provided excess air levels of 15 to 20 percent are maintained.) An increase in excess air increases the amount of energy leaving the com¬ bustor as the sensible heat of the off gases. At low levels, this in¬ crease is compensated by the increase in combustion efficiency. At higher levels, however, the heat removed by the flue gases exceeds that produced by higher combustion efficiencies. To maintain constant bed temperature, some heat transfer surface must be removed. If this is not possible, the bed temperature will drop. Increasing excess air levels even further (to about 300 percent) will result in a situation where all the heat transfer surface must be removed to maintain a constant bed temperature. At this point, the bed becomes an adiabatic combustor. Any increase beyond this point will lower the bed temperature. The elutriation rate from a fluidized bed will determine the load on any particle collection equipment (cyclones) and also the amount of particu¬ late emissions. It is affected by such factors as particle size distribu¬ tion, particle density, bed cross section, and terminal and superficial velocity. The entrainment rate decreases as freeboard height increases. If the height is increased sufficiently, a height will be reached at which the entrainment rate is constant. This height is termed the transport dis¬ engaging height (TDH) and is defined as the point at which steady flow conditions are established. Merrick and Highley have developed both an analytical and numerical model J kU for particle size reduction and elutriation for a fluidized bed process. 118 Data to obtain empirical constants for their relation were taken from the results of the British Coal Utilization Research Authority's 3- by 3-foot combustor. Mechanism of Coal Injection The manner in which coal is injected into the bed has an effect on combus¬ tion efficiency, temperature distribution, and particle residence time. In some experiments with their batch combustor, Exxon Research and Engi- 28 neering noted burning in the freeboard zone. Although velocity also had an effect on the degree of combustion in the freeboard, it was determined that the mechanism of coal feeding was the major source of the problem. % Combustion above the bed was eliminated by changing the position and ori¬ entation of the coal feed probe from a position 25 cm above the fluidizing grid and a 45° angle upward to a position just above the grid and a 45° angle downward. The observed combustion in the freeboard was attributed to a reduced particle residence time, resulting from the original coal feed orientation. Also of interest was the fact that no hot spots were observed with the new orientation. Pressure Pressure will influence gaseous equilibrium concentrations and will also influence the potential vaporization of various mineral compounds. Pres¬ sure will also influence the degree of slugging via fluid density changes. For height to diameter ratios greater than 3, the minimum slugging veloc¬ ity decreases with decreasing minimum fluidization velocity. Because the velocity at incipient fluidization decreases with increasing fluid den¬ sity and thus pressure, the minimum slugging velocity should decrease with increasing pressure. Viscosity also changes with pressure; for a change in pressure from 1 to 10 atm, the velocity at incipient fluidization de¬ creases by about 0.3 percent. Thus, the change in minimum slugging 119 velocity due to the contribution from the viscosity effects should be negligible for height to diameter ratios greater than 3. The minimum slugging velocity increases with pressure at values of height to diameter less than 3. A tenfold increase in pressure produces an increase in minimum slugging velocity of about 10 percent. Pressure also has an effect on entrainment via fluid density changes. A tenfold increase in pressure causes approximately a tenfold increase in density (for gases). The terminal velocity of a particle decreases with increasing fluid density. Therefore, at constant superficial veloc¬ ity an increase in pressure will cause larger particles to be elutriated. This effect is shown in Figure 23. 45 CONTROL OF POLLUTANTS FROM SPENT STONE DISPOSAL Solid waste could prove to be one of the most significant pollutants associated with fluidized bed combustion simply because of the poten¬ tially large quantities of waste produced by once-through sorbent oper¬ ation. Table 21 presents estimates of the solid waste disposal require¬ ments for a once-through fluidized bed combustion system as a function 6 of the Ca/S mole ratio and the coal sulfur content. Table 21. SPENT BED PLUS ASH PRODUCED BY A 635-MW ONCE-THROUGH SORBENT FBC PLANT 3 ’ 46 (10^ tons/year) 1% sulfur in coal — 2% sulfur in coal 3% sulfur in coal 4% sulfur in coal Ca/S = 2 300 433 569 704 Ca/S = 1.2 251 337 424 510 Coal feed rate = 430,000 pounds per hour; Ash content = 12 percent, 165 x 10^ tons/year; 90 percent SO 2 removal; 73 percent load factor. Assumptions 120 SOLIDS TRANSPORTED , F s /A t u 0 gm/cm 3 GAS 3 X I O' 3 O 2XIO -3 I X IO'° - Figure 23. Comparison between calculated and experimental entrainment at various pressures^ (solid lines are calculated) 121 The amount of solids produced at low Ca/S ratios would be similar to that from limestone scrubbing systems but the total mass of waste would be 50 percent less because the scrubber sludge contains 50 percent water while the FBC spent stone is dry. Potential pollution from solid waste disposal includes leachates (trace elements and organic compounds) and possibly unsightly and large-scale land fills. As indicated earlier in Table 14 based on an assumed density of 100 lb/ft 3 (65 percent of the theoretical spent stone density), a 635-MW plant burning coal with 3 percent sulfur and 12 percent ash, and using a Ca/S ratio of 2, would require 260 acre-feet per year for spent stone disposal. Over a 30-year plant life, a landfill area of 260 acres would have to be 30 feet deep to accommodate this one plant. Solid Waste Control Methods The most effective method to reduce the volume of solid waste and the overall solid waste problem is to regenerate and recycle the sorbent. As previously discussed, several regeneration options are being actively investigated. Keairns et al. have found from laboratory thermogravimetric testing that the chemistry of regeneration appears to be favorable but cautioned that other factors such as attrition, coal-ash sorbent agglomera tion, tar deposition on the sorbent or eutectic formation could limit the number of regeneration cycles in a commercial system. 47 Based on regenera tion kinetics experiments, they suggest that 0.18 moles of Ca may effect¬ ively remove one mole of sulfur by being recycled 20 times. Therefore, compared to a once-through system, the solid waste burden attributable to sulfur removal would be reduced by 85 percent at a 1.2/1 ratio and by 91 percent if a 2/1 ratio were required for a once-through system. Very little experimental information specific to FBC systems is available with respect to stone disposal or its utilization. However, some insight to the problem can be gained from a discussion of current methods for coal ash or scrubber sludge disposal. Due to its high water content 122 (over 50 percent), sludge from limestone scrubbing will almost always be ponded, either temporarily or permanently. Ponding presents the maximum potential for ground and surface water contamination, as large quantities of saturated water are present. Materials including PVC, Hypalon, con¬ crete, clay, and asphalt are available for lining ponds to prevent ground- 48 water pollution. Most electric power plants transfer coal ash as a wet slurry for convenience and for the same reason wet transfer of spent FBC stone may be considered. If disposal will occur via landfilling techniques, a landfill site should be selected and managed to minimize water pollution through leachates or surface runoff. Geological and hydrological evaluations are particularly important in selecting a site to minimize water pollution. Typical leach¬ ate migration rates range from 0.5 to 30 meters/yr and ground water pollu- 49 tion problems may take several years to develop. Therefore, initial site selection and immediate pollutant monitoring are very important in developing new landfill sites and preventing difficult to correct future pollution problems. The landfill should be located above the natural water table and groundwater flows including springs should be rerouted around the landfill areas. 50 A hydrogeological investigation is an essential part of selecting a land¬ fill site in all areas. Boring is required to obtain soil samples 52 and determine water table levels. The purpose of the survey is to de¬ termine the potential for leachates to reach the groundwater. A general recommendation that landfills be located at least 3 to 10 feet above the water table is based on the natural containment concept. Natu¬ ral containment depends on the ability of the soil to attenuate pollutants contained in the leachate. The best soils for pollutant attenuation, those which are fine grained; these are usually clays, silts or granular soils with high clay content.Physical properties of soils meeting the requirements for natural attenuation are typically: permeability _ & 10 cm/sec or less, at least 30 percent passing a 200 mesh sieve, a 123 liquid limit greater than 30 percent and a plasticity index greater than 15."^ Mechanisms causing pollutant attenuation include dispersion, dilu¬ tion, filtration, retention, ion exchange and biological breakdown. In a study of four landfills in Illinois it was observed that chlorides mi¬ grated farthest from the landfill site; similar results have been obtained 53 54 by other investigators. ’ Soil attenuation of selenium and boron was reported to be poor.'*'’ Semi-artificial containment can be accomplished by selecting a disposal site where extensive clay deposits are present and the soil is relatively impervious. Pond liners can be used to artificially contain leachate. In either case the leachate should be collected and treated with at least 56 conventional water treatment techniques. Commercial Uses for Solid Waste By-Products Commercial uses for the spent stone material are also being investi¬ gated. The major obstacles to using the spent stone will be the very large volumes generated (market saturation), transport costs and availability of competitive raw materials. As illustrated in Table 22, fly ash and bottom ash from conventional boilers are used in a number of products but only about 20 percent of the total is utilized. 57 This indicates that the market in this area for spent stone may be limited. New uses that have been considered for spent stone include autoclaved products — bricks, hot press sintering-pipes and metal coatings; gypsum products — wallboard and plaster; and mineral recovery. Gypsum from SO^ scrubbing is used extensively in Japan, j8 but no large markets exist in the United States because of an abundant supply of natural deposits. Treatment of acid mine drainage is also a potential use for spent stone. Spent stone is also being considered as an agricultural supplement; pilot studies have shown successful application in peanut farming. 124 Table 22. ASH COLLECTION AND UTILIZATION, 1971 57 Fly ash, tons Bottom ash, tons Boiler slag (if separated from bottom ash), tons Ash utilized: Mixed with raw material before forming cement clinker 104,222 NA 91,975 Mixed with cement clinker or mixed with cement (pozzolan cement) 16,536 NA NA Partial replacement of cement in: Concrete products Structural concrete Dams and other massive concrete 177,166 185,467 71,411 35,377 NA NA 76,563 NA NA Lightweight aggregate 178,895 13,942 NA Fill material for roads, construction sites, etc. 363,385 533,682 2,628,885 Stabilizer for road bases, parking areas, etc. 36,939 7,880 49,564 Filler in asphalt mix 147,655 2,833 81,700 Miscellaneous 98,802 475,417 428,026 Total ash utilized 1,380,478 1,069,131 3,356,713 Ash removed from plant site at no cost to utility but not covered in categories listed under Ash utilized 1,872,728 542,895 381,775 Ash removed to disposal areas at company expense 24,497,848 8,446,941 1,232,298 Total ash collected 27,751,054 10,058,967 4,970,786 NOTE: NA - Not applicable. 125 If stone regeneration is used, the final amount of spent stone might be reduced by 85 to 90 percent. In addition, a low sulfur (0.5 percent) lime product could be produced if the waste stone is bled out of the circulation loop after regeneration. This low sulfur lime product could sell for $20/ton, and at that price would be an economically attractive by-product. Several projects funded by ERDA and by EPA are underway to investigate means for utilization of FBC solid residue. Among the larger utilization projects aimed specifically at FBC residues are two being funded by ERDA, one with the U.S. Department of Agriculture (agricultural uses) and one with IU Conversion Systems, Inc. (nonagricultural uses). 126 REFERENCES 1. Danielson, J. A. (ed.). Air Pollution Engineering Manual. 2nd Edition. U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publication Number AP-40. May 1973. p. 95-98. 2. Compilation of Air Pollutant Emission Factors. U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publi¬ cation Number AP-42. April 1973. p. A-3. 3. Cowherd, C. et al. Hazardous Emission Characterization of Utility Boilers. Midwest Research Institute. Prepared for U.S. Environ¬ mental Protection Agency, Research Triangle Park, North Carolina. Publication Number EPA-650/2-75-066. July 1975. 4. Bradway, R. M. and R. W. Cass. Fractional Efficiency of a Utility Boiler Baghouse - Nucla Generating Plant. GCA/Technology Division, Bedford, Massachusetts. Prepared for U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publication Number EPA-600/2-75-013a. August 1975. 5. Cass, R. W., R. M. Bradway, and N. F. Surprenant. Fractional Effi¬ ciency of a Utility Boiler Baghouse - Sunbury Steam Electric Station. GCA/Technology Division, Bedford, Massachusetts. Prepared for U.S. Environmental Protection Agency. Publication Number EPA-600/2-76- 077a. March 1976. 6. Business Briefs. J Air Pollut Control Assoc. 25(11):1163, November 1975. 7. McKenna, J. D. et al. Performance and Cost Comparisons Between Fabric Filters and Alternate Particulate Control Techniques. Pro¬ ceedings: Symposium on the Use of Fabric Filters for the Control of Submicron Particulates. U.S. Environmental Protection Agency. Publication Number EPA-650/2-74-043. May 1974. p. 15. 8. Hoke, R. C. et al. Studies of the Pressurized Fluidized-Bed Coal Combustion Process. Exxon Research and Engineering Co., Linden, New Jersey. Prepared for U.S. Environmental Protection Agency. EPA Report No. EPA-600/7-76-011. September 1976. 9. Shannon, L. J. and P. G. Gorman. Particulate Pollutant System Study: Volume II - Fine Particle Emissions. Midwest Research Institute. Prepared for U.S. Environmental Protection Agency, Durham, North Carolina. Publication Number APTD-0744. August 1971. 127 V 10. Multicell Fluidized Bed Boiler Design, Construction and Test Program. Pope, Evans and Robbins, Inc. Publication Number PB 236 254/AS, Office of Coal Research, Washington, D.C. R&D Report No. 90 - In¬ terim Report No. 1. August 1974. 11. Lisle, E. S. and J. D. Sensenbaugh. The Determination of SO^ and Acid Dew Point in Flue Gas. Combustion. January 1965. 12. Robison, E. B., A. H. Bagnulo, J. W. Bishop, and S. Ehrlich. Char¬ acterization and Control of Gaseous Emissions from Coal-Fired Fluidi¬ zed Bed Boilers. Pope, Evans and Robbins, Inc., Alexandria, Virginia. Prepared for Division of Process Control Engineering, National Air Pollution Control Administration (now U.S. Environmental Protection Agency). Publication Number APTD-0655, PB 198 413. 1970. p. 103. 13. Cook, R. E. Sulfur Trioxide Conditioning. Symposium on Electro¬ static Precipitators for the Control of Fine Particulates. U.S. Environmental Protection Agency, Research Triangle Park, North Car¬ olina. Publication Number EPA-650/2-75-016. January 1975. p. 315. 14. McCain, J. P., J. P. Gooch, and Wallace B. Smith. Results of Field Measurements of Industrial Particulate Sources and Electrostatic Precipitator Performance. Symposium on Electrostatic Precipitators for the Control of Fine Particulates. U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publication Number EPA-650/2-75-016. January 1975. p. 79. 15. Benson, J. R. and M. Corn. Costs of Air Cleaning With Electrostatic Precipitators at TVA Steam Plants. J Air Pollut Control Assoc. 24:339, 1974. 16. White, H. J. Role of Electrostatic Precipitators in Particulate Con¬ trol - A Retrospective and Prospective View. Symposium on Electro¬ static Precipitators for the Control of Fine Particulates. U.S. En¬ vironmental Protection Agency, Research Triangle Park, North Carolina. Publication Number EPA-650/2-75-016. January 1975. p. 17. 17. Cooper, D. W., L. W. Parker, and E. Mallove. Overview of EPA/IERL-RTP Scrubber Programs. GCA/Technology Division, Bedford, Massachusetts. Prepared for U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publication Number EPA-600/2-75-054. Septem¬ ber 1975. 18. Billings, C. E. and J. Wilder. Handbook of Fabric Filter Technology. GCA/Technology Division, Bedford, Massachusetts. National Air Pollu¬ tion Control Administration. Report Number APTD 0690. December 1970. 128 19. Rao, A. K., M. P. Schrag, and L. J. Shannon. Particulate Removal From Gas Streams at High Temperature/High Pressure. Midwest Re¬ search Institute. Prepared for U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publication Number EPA-600/ 2-75-020. August 1975. 20. Keairns, D. L. et al. Fluidized Bed Combustion Process Evaluation: Phase II - Pressurized Fluidized Bed Coal Combustion Development. Westinghouse Research Laboratories. Prepared for U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publica¬ tion Number EPA-650/2-75-027c. September 1975. 21. Hoke, R. C. Exxon Research and Engineering Co., Linden, New Jersey. Private Communication. 22. Purification of Hot Fuel Gases From Coal or Heavy Oil. Stone and Webster Engineering Corp. Prepared for Electric Power Research Institute, Palo Alto, California. Interim Report No. EPRI 243-1. (Available from NTIS.) November 1974. 23. Zenz, F. A. and D. F. Othmer. Fluidization and Fluid Particle Systems. New York, Reinhold, 1960. 24. Broadhurst, T. E. and H. A. Becker. Onset of Fluidization and Slug¬ ging in Beds of Uniform Particles. AIChE J. 21(2):238-247, 1975. 25. Stewart, P. S. B. and J. F. Davidson. Slug Flow in Fluidized Beds. Powder Tech. 1:61, 1967. 26. Zenz and Othmer. Op cit. p. 301. 27. Wright, S. J. The Reduction of Emissions of Sulfur Oxides and Nitrogen Oxides by Additions of Limestone or Dolomite During the Combustion of Coal in Fluidized Beds. Proceedings of 3rd Inter¬ national Conference on Fluidized Bed Combustion. Publication Number EPA-650/2-73-053. December 1973. p. 1-4-1. 28. Hoke, R. C. Exxon Research and Engineering Co., Linden, New Jersey. Private Communication. 1975. 29. Hoke, R. C. Exxon Research and Engineering Co., Linden, New Jersey. Private Communication. 30. Zenz and Othmer. Op cit. p. 260. 31. Kunii, D. and 0. Levenspiel. Fluidization Engineering. New York, John Wiley and Sons, 1969. 32. Bottom, R. Chem Eng Prog. Symposium Series No. 101. 66:8, 1970. 129 33. Cooke, M. et al. (Paper presented at Tripartite Engineering Con¬ ference. Montreal, Canada. 1968.) 34. Behie, L. A. and P. Kehoe. The Grid Region in a Fluidized Reactor. AIChE J. 19(5):1070-1072, September 1973. 35. Parks, D. J. and E. A. Fletcher. Formation and Emission of NO in Fluidized Bed Combustion. Environ Sci and Technol. 9:749, 1975. 36. Pereira, F. J., J. M. Beer, B. Gibbs, and A. B. Hedley. N0 X Emissions From Fluidized Bed Combustion. (Presented at 15th International Sym¬ posium on Combustion. The Combustion Institute, Pittsburgh. 1974.) 37. Zenz and Othmer. Op cit. 38. Vogel, G. J. et al. Bench Scale Development of Combustion and Addi¬ tive Regeneration in Fluidized Beds. Proceedings of 3rd International Conference on Fluidized Bed Combustion. U.S. Environmental Protection Agency. Publication Number EPA-650/2-73-053. December 1973. 39. Zenz and Othmer. Op cit. p. 82. 40. Davidson, J. F. and D. Harrison. Fluidized Particles. Cambridge, University Press, 1963. 41. Zenz and Othmer. Op cit. p. 278. 42. Orr, C., Jr. Particulate Technology. New York, The MacMillan Com¬ pany, 1966. 43. Robison, E. B. et al. Study of Characterization and Control of Air Pollutants From a Fluidized-Bed Combustion Unit - The Carbon Burnup Cell. Pope, Evans and Robbins, Inc. Prepared for U.S. Environmental Protection Agency. Publication Number APTD 1170. February 1972. 44. Merrick, D. and J. Highley. Particle Size Reduction and Elutriation in a Fluidized Bed Process. AIChE Symposium Series. 137(70):366-378, 1974. 45 . Zenz, F. A. and N. A. Weil. AIChE J. 4:472, 1958. 46. Murthy, K. S. et al. Battelle Memorial Research Institute, Columbus, Ohio. Private Communication. 47. Keairns, D. L., E. P. O'Neill, and D. H. Archer. Sulfur Emission Control With Limestone/Dolomite in Advanced Fossil Fuel Processing Systems. Symposium Proceedings: Environmental Aspects of Fuel Con¬ version Technology. U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Publication Number EPA-650/2-74-118. October 1974. 130 48. Rossoff, J. and R. C. Rossi. Disposal of By-Products From Non- Regenerable Flue Gas Desulfurization Systems: Initial Report. Prepared for U.S. Environmental Protection Agency. Publication Number EPA-650/2-74-037a. May 1974. 49. Chian, Edward S. K. and Foppe B. DeWolle. Compilation of Methodology Used for Measuring Pollution Parameters of Sanitary Landfill Leachate (Preliminary Report). Environmental Engineering Section, Department of Civil Engineering, University of Illinois, Urbana, Illinois. Pre¬ pared for U.S. Environmental Protection Agency Office of Research and Development under Contract No. 68-03-2052. 1974. 50. Vandy, P. Design, Environmental Management and Economic Considerations for Sanitary Landfills. Waste Age. 5:21, January/February 1974. 51. Garland, G. A. and D. C. Mosher. Leachate Effects of Improper Land Disposal. Waste Age. 6:40, March 1975. 52. Leonard, J. F. and K. Acimovic. Landfills - Design Based on Hydro- geological Investigation. Waste Age, 1975. p. 16. 53. Hughes, G. M., R. A. Landon, and R. N. Farvolden. Hydrogeology of Solid Waste Disposal Sites in Northeastern Illinois. Prepared for U.S. Environmental Protection Agency, Washington, D.C. Report Num¬ ber SW-12d. 1971 54. Ham, R. K. The Generation, Movement and Attenuation of Leachates From Solid Waste Land Disposal Sites. Waste Age. 6:50, 1975. 55. Rossoff, J., R. C. Rossi, L. J. Bornstein, and J. W. Jones. Disposal of By-Products From Nonregenerable Flue Gas Desulfurization Systems. Proceedings: Symposium on Flue Gas Desulfurization. U.S. Environ¬ mental Protection Agency. Publication Number EPA-650/2-74-126a. December 1974. p. 399. 56. Nordstedt, R. H., L. B. Baldwin, and L. M. Rhodes. Land Disposal of Effluent From a Sanitary Landfill. J Water Pollut Control Fed. 47:1961, 1975. 57. Brackett, C. E. Production and Utilization of Ash in the United States. In: Proceedings: Third International Ash Utilization Sym¬ posium. Bureau of Mines Information Circular 8640. 1973. 58. Ando, J. Utilizing and Disposing of Sulfur Products From Flue Gas Desulfurization Processes in Japan. Proceedings: Symposium on Flue Gas Desulfurization. U.S. Environmental Protection Agency. Publi¬ cation Number EPA-650/2-74-126b. December 1974. p. 955. 131 APPENDIX PRELIMINARY LIST OF CONCEIVABLE POLLUTANTS Based on preliminary considerations of the prevailing temperatures and the chemical make-up of a coal-fired fluidized bed combustion system (coal, limestone, combustion or scrubbing additives, etc.), a list of potential pollutants has been prepared and is shown in Table 23. This list is based on the major elements, C, H, 0, S, N and the principal trace elements commonly found in American coals. The organic compounds (including sulfur and nitrogen compounds) are those which could result from incomplete combustion of coal. Only classes of organic compounds are listed since further detail would make the list too unwieldy. The trace element compounds are representative of species expected in the coal or limestone as well as in the combustion products. Chlorine compounds of these elements are included since, in some cases, NaCl is used to increase the efficiency of the sorbent bed. 132 Table 23. PRELIMINARY LIST OF POSSIBLE POLLUTANTS FROM FLUIDIZED BED COMBUSTION OF COAL ACIDS AND ACID ANHYDRIDES 1. Organic Acids a. Carboxylic acids: b. Dicarboxylic acids: c. Sulfonic acids: 2. Inorganic Acids a. Sulfuric: h 2 s ° b. Sulfurous: h 2 so c. Nitric: HN0 3 d. Nitrous: hno 2 e. Phosphoric: HP0 3 f. Hydrofluoric: HF g- Hydrochloric: HC1 HALOCARBONS 1. Chlorinated Aliphatics: R - X e.g., chloromethane, chlorobenzene 2. Chlorinated Biphenyls: foToJ Cl 0 II R-C-OH e.g., formic acid, benzoic acid, etc. 0 0 If II HO-C-R-C-OH e.g., phthalic acid, succinic acid 0 I R-S-OH e.g., benzenesulfonic acid t 0 133 Table 23 (continued). PRELIMINARY LIST OF POSSIBLE POLLUTANTS FROM FLUIDIZED BED COMBUSTION OF COAL HYDROCARBONS 1 . Alkynes: -C-C- e.g., acetylene, butyne, propyne, etc. 2 . Diolefins: -C-C-C-C- e.g., butadiene, pentadiene, octadiene, etc. 3. Olefins: -C-C- e.g., ethylene, propylene, butene etc. 4. Aromatics: 0 e.g., benzene, toluene, etc. 5. Polynuclear Aromatics: J0L e.g., anthracene, pyrene, 000 / phenanthrene, etc. 6 . Cyclic Hydrocarbons: 1 j I—] e.g., cyclopentane, cyclo- 00 00 pentadiene, etc. 7. Aliphatic Hydrocarbons I 1 : -C-C- e.g., methane, ethane, propane, ' ' etc. NITROGEN COMPOUNDS 1 . NO 2 . no 2 3. HCN 4. (cn ) 2 5. nh 3 6 . Amines: * R-N e.g., methylamine, ethylamine, aniline, etc 7. Pyridines: 0 N e.g., pyridine, quinoline 134 Table 23 (continued). PRELIMINARY LIST OF POSSIBLE POLLUTANTS FROM FLUIDIZED BED COMBUSTION OF COAL 8 . Pyrroles: u N 9. Nitrate Salts: MNO^ (where M is any cation) 10. Nitrite Salts: MN0 9 (where M is any cation) 11. Nitrosaraines: R - N - N = 0 12. Azoarenes: - N = N OXYGEN COMPOUNDS 1. Furan: 2. Ethers: R-O-R e.g., phenyl ether, anisole, etc. 0 II 3. Esters: R-C-OR e.g., phenylacetate, benzylacetate, etc. 4. Epoxides: -C-C- e.g., ethylene oxide, propylene oxide, etc. V 5. Alcohols: R-OH e.g., methanol, phenol, etc. 0 II 6 . Aldehydes: R-C- e.g., formaldehyde, benzaldehyde, etc. 0 II 7. Ketones: R-C-R e.g., acetone, benzophenone, acetophenone PARTICULATES 1. Size Distribution a. < 2 pm b. 2 pm < > 20 pm c. > 20 ym 135 Table 23 (continued). PRELIMINARY LIST OF POSSIBLE POLLUTANTS FROM FLUIDIZED BED COMBUSTION OF COAL RADIOACTIVE ISOTOPES 1. Uranium-235, Polonium-210, Lead-210, Radium-226, Bismuth-210, Thorium-227, Radon-222. SULFUR COMPOUNDS i. so 2 2. S0 3 • 3. H 2 S 4. cos 5. CS 2 6. S X 7. Thiophenes: u s 8. Mercaptans: R-SH e.g., methyl mercaptan, phenyl mercaptan 9. Sulfates: MSO. 4 (where M is a metal ion) e.g., FeSO^, PbSO^, etc 10. Sulfites: mso 3 (where M is a metal ion) e.g., CaSO^ TRACE ELEMENTS AND THEIR COMPOUNDS 1. Nickel: NiS, NiO, Ni 2 0 3 , Ni(C0) 4 , NiS0 4 , NiC0 3 , NiCl 2 , Ni(CN) 2 , Ni(OH) £ 2. Cadmium: CdS, CdO, CdS0 4 , CdC0 3 , Cd(CN) 2 , CdCl 2 , Cd(OH) 2 3. Mercury: HgS, HgO, Hg 2 0, Hg(CN) 2 , HgS0 4 , Hg 2 S0 4 , HgC0 3 , HgCl, HgCl 2 136 Table 23 (continued). PRELIMINARY LIST OF POSSIBLE POLLUTANTS FROM FLUIDIZED BED COMBUSTION OF COAL 4. Zinc: ZnS, ZnO, Zn0 2 , Zn(SO)^, ZnCO^, ZnCl 2 , Zn(CN) 2 , Zn^, Zn(0H) 2 5. Lead: PbS, PbO, Pb 2 0 3 , Pb 3 0 4 , Pb0 2> Pb(CN) 2 , PbS0 4 , PbC0 3> PbCl 2 , PbN 6 , Pb(OH) 2 , PbOH 6. Sodium: NaCl, Na 2 0, Na 2 S0 4 , Na 2 C0 3> NaNH 2 , NaN 3 , NaCN, NaH, Na 2 0 2 , Na 2 S 2 0y, Na 2 S 4 0^, Na 2 S 2 0^, Na 2 CS 2 , Na 2 Si0 2 , NaHS0 3 , NaHS0 4 , NaHCO , NaOH 7. Potassium: KC1, K_0, K o S0., K o C0-, KCNO, KCN, KHS, K o S o 0 o , 2 24 23 228 K~CS„, K S o 0„, KHC0„, KHSO., K.Si.O,, K_Si_0^, KOH 2 3 2 2 3 3 4 2 2 5 2 3 7 8. Vanadium: V 2°5’ V 2°3’ V 2°4’ V 2 S 3’ V0S °4’ Na 2 ° * V 2°5 9. Cobalt: CoS, Co 3 0 4 , CoO, Co 2 0 3 , CoS0 4 , CoC0 3 , CoCl 2 , Co(CO) 4 , Co(CN) 2 , Co(OH) 2 10. Molybdenum: MoS 2 , Mo 2 0 3 , Mo 0 3> Mo(OH )2 11. Copper: CuS, Cu 2 S, Cu 2 0, CuO, CuC0 3> CuS0 4 , CuCl, CuC^, CuCN, Cu N, CuOH, Cu(OH )^ 12. Beryllium: BeO, Be 0 C, Be(S0.) o , BeC0 o , BeCl OJ BeH , Be„N„ l 42 3 2 2 32 13. Selenium: SeS 2 , SeS, Se0 2 , H 2 Se0 3> H^eO^ Se 2 Cl 2 , SeCl 4 , SeCl 14. Arsenic: As 2 S^, As 2 S 2 , As 2 S 2 , As 2 0 3 , As 4 0^, H 3 As 0 4 , AsH ? , AsH 3> AsC1 3 15. Antimony: Sb 2 S 3 , Sb 2 S 3 , Sb 2 0 3 , Sb 2 0^, Sb 2 (S0 4 ) 3 , SbCl 3 , SbCl 5 , Sb(OH) 3 16. Fluorine: F 2 0 2 , F 2 0, F 2 , N0 3 F, HS0 3 F, fluorinated hydrocarbons 137 Table 23 (continued). PRELIMINARY LIST OF POSSIBLE POLLUTANTS FROM FLUIDIZED BED COMBUSTION OF COAL 17. Chlorine: Cl o 0, C10 o , HCl0 o , HC1, Cl„, Cl o 0^, MC10. (where M 2 2 3 2 27 4 is a metal), chlorinated hydrocarbons 18. Thallium: T1 2 S, T1 2 0, T1 2 0 3 , T1 2 SO a , T1 2 C0 3 , T1C1, T10H 19. Manganese: MnS, MnS 9 , Mn 3 0 3 , Mn 3 0 4 , Mn0 3 , MnSO^, MnC0 3 , MnC^, Mn(CO) 3 , MnSi0 3 , Mn(0H) 2 20. Iron: FeS, FeO, Fe20 3 , Fe 3 0 4 , FeSO^, FeC0 3 , FeC^, FeCl 3 , Fe(CO)^, H 2 Fe(C0) A , Fe(0H) 2 21. Barium: BaS, BaO, Ba02» BaC0 3 , BaSO^, BaS0 3 , Ba(N0 3 )2> BaCl 2 , Ba(CN) 2 , BaS 2 0 6 22. Tellurium: TeS 2 , Te0 2 , H 6 Te0 6 , TeCl^, TeCl 2 , H 2 Te °3 23. Titanium: Ti0 2 , TiH 2 , TiCl 2 , Ti0S0 4 , Ti(S0 4 ) 3 , TiCl 3 , TiCl 4 24. Silicon: Si0 2 , H 2 Si0 3 , SiC, SiH 4 , SiO, SiS 2> SiCl 4 25. Aluminum: A1 2 S 3 , A1 2 (S0 4 ) 3 , A1 2 0 3 • 3Si0 2 , Al^, A1C1 3 , A1H 3> AIN, A1 2 0 3 , Al(OH) 3 26. Magnesium: MgO, Mg0 2 , Mg 2 Si 3 0 8 , Mg 2 Si, MgS0 4 , MgS0 3 , Mg(OH) 2> MgS 2 0 3 , MgCl 2 , Mg(NH 2 ) 2 , MgH 2 27. Calcium: CaS, CaC 9 , CaC0 3 , CaC^, CaNCN, Ca(CN) 2> CaH 2 , CaO, Ca0 9 , Ca 2 SiO^, Ca^i^Oy, Ca^Si^Og, Ca 4 (l^S^O^) CaSO^, CaS 2 0 3 , Ca(OH) 2 MISCELLANEOUS POLLUTANTS 1 . Biological Oxygen Demand (BOD) 4. Noise 2. Chemical Oxygen Demand (COD) 5. Odor 3. Thermal Discharge 6. Biological Agents - bacteria virus, spores, etc. 138 TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) 1. REPORT NO. 2. EPA-600/7-77-054 3. RECIPIENT'S ACCESSION' NO. 4 title and subtitle PRELIMINARY ENVIRONMENTAL ASSESSMENT OF COAL-FIRED FLUIDIZED-BED COMBUSTION SYSTEMS 5. REPORT DATE Mav 1977 6. PERFORMING ORGANIZATION CODE 7 . authors Paul F. Fennelly, Donald F. Durocher, Hans Klemm, and Robert R. Hall 8. PERFORMING ORGANIZATION REPORT NO. GCA-TR-75-37-G 9. PERFORMING ORGANIZATION NAME AND ADDRESS GCA Corporation GCA/Technology Division Bedford, Massachusetts 01730 10. PROGRAM ELEMENT NO. E HE 62 3A 11. CONTRACT/GRANT NO. 68-02-1316, Task 15 12. SPONSORING AGENCY NAME AND ADDRESS EPA, Office of Research and Development Industrial Environmental Research Laboratory Research Triangle Park, NC 27711 13. TYPE OF REPORT AND PERIOD COVERED Preliminary; 9/75-6/76 14. SPONSORING AGENCY CODE E PA/600/13 15. supplementary notes IERL-RTP task officer for this report is D. Bruce Henschel, Mail Drop 61, 919/549-8411 Ext 2825. is. abstract -p he re p 0r j- gi ve s results of a preliminary evaluation of potential pollutants which could be generated in coal-fired fluidized-bed combustion (FBC) processes. Because 802 and NOx formation already have received considerable attention from many investigators, the primary emphasis here is on the ’other’ pollutants: organic compounds, trace elements, inorganic compounds (other than S02 and NOx), and particulates. Using available bench scale or pilot plant data and/or simple thermo¬ dynamics and empirical correlations with data from other combustion systems, order of magnitude estimates were made of the concentrations of various elements and compounds in the flue gas, in the solid waste, and in the water discharge. Results suggest that, in general, no special environmental problems should result from coal- fired FBC, but that better data are required in several areas, particularly with regard to particle size distributions, possible organic compounds, and the fate of such elements as Be, As, U, Pb, Cd, Ni, Cl, Se, F, and their compounds. 17. KEY WORDS AND DOCUMENT ANALYSIS a. DESCRIPTORS b.IDENTIFIERS/OPEN ENDED TERMS c. COSAT 1 Field/Group Air Pollution Flue Gases Coal Wastes Combustion Waste Water Fluidized Bed Organic Compounds Processing Inorganic Corn- Size Screening pounds Air Pollution Control Stationary Sources Environmental Assess¬ ment Trace Elements Particulate 13B 2 ID 2 IB 07C 13H ,07A 07B 13. DISTRIBUTION STATEMENT Unlimited 19. SECURITY CLASS (This Report) Unclassified 21. NO. OF PAGES 151 20. SECURITY CLASS (This page) Unclassified 22. PRICE EPA Form 2220-1 (9-73) 139 PUBLICATION NO. EPA-600/7-77-054 \ a 2 h r 1 h-( tu M H- cr rt H o' 05 3 H- 3 3 3 3 05 0 ) 01 o H - l-i H- ^ W M I— 1 (0 C/5 i- 1 cn rt H- O 05 3 3 rr 0 3 3 H- O 05 3 O 3 3 O ff>W K I—‘CO 00 H- 0Q O I— 1 H- M CL O H- 3 3 I— 1 00 /; > z t. c CO m O C > u m Z > r H < o D T) O jo H r Z o ■n ■n < m £ "0 r m C V) m r OD C (/> z m U) 0) o S CO O o ir 3 XI O O 3 O 3 3 3 O a> - ZV JO £ <3 m C/5 - 7 . 3 05 O 3" o 3 o' o zr 3 3 „ T> JO 3 Q. O o 3 Ol oo 05 CO 00 si O m a> o i ± ° 9 T3 Z 3 3 3 C it) 3 z (/) < T3 7) 3 3 O 0 n Z U) 5 £ m H > Z 0 ■n m -1 3 a) 0 3 t> > 0 c > r z 0 CJ 3 0 X y 1 O u tn X 0 -t ■n 3 m 0) m it) 1/1 n -t 3 > JO 01 6 z 0 it > 0 3 z o < > O m z o < < ■ 1 111 *