United States Environmental Protection Agency Air and Energy Engineering Research Laboratory Research Triangle Park NC 27711 Research and Development EPA/600/S7-85/031a Sept. 1985 Project Summary An Evaluation of Full-Scale Fabric Filters on Utility Boilers: SPS Harrington Station Unit 3 John W. Richardson, John D. McKenna, and John C. Mycock The objective of this program was to evaluate and characterize the perfor- mance of full-scale fabric filter units in- stalled on 100 MW or larger coal-fired power plants. This document reports results of total mass and fractional size particulate emission tests at South- western Public Service's Harrington Station Unit 3 from July 8 to July 11, 1981. Three outlet and one inlet mass and fractional size emission tests were performed. Due to the absence of inlet ports, inlet testing was done by bypass- ing the baghouse and testing at the outlet ports of the fabric filter. The fab- ric filter is a shake/deflate unit with 32 compartments. Each compartment has 204 bags, 30 ft 6 in.* long and 11.5 in. in diameter. Design air/cloth ratio is 2.8. Average outlet concentration was 0.007 lb/106 Btu. Inlet loading was 2.0 lb/106 Btu, giving a 99.65% collection effi- ciency. Particle sizing tests indicated that the mass geometric mean diame- ter for the three outlet tests ranged from 7.2 to 13 /jm with an inlet mass diameter of 60 /urn. This Project Summary was devel- oped by EPA's Air and Energy Engi- neering Research Laboratory, Research Triangle Park, NC, to announce key findings of the research project that is fully documented in a separate report of the same title (see Project Report ordering information at back). *Readers more familiar with the metric system may use the conversion factors at the end of this Summary, Introduction EPA funded a program in 1980 to evaluate and characterize the perfor- mance of full-scale fabric filter units in- stalled on 100 MW or larger coal-fired power plants. The program required particulate total mass and fractional size efficiency testing and collection of per- formance, operation, maintenance, and problem information. Southwestern Public Service's Harrington Station Unit 3 was selected as one of the sites to be tested. Four outlet particulate total mass tests and one inlet particulate total mass test were performed. Run 1 was voided because an incorrect stack moisture content was assumed which resulted in the wrong size sampling nozzle being used. Runs 2, 3, and 5 (outlet tests) are considered valid. Run 4 was the inlet particulate test which was conducted by bypassing the baghouse and testing at the stack outlet ports. These data are summarized in Table 1. Sampling was conducted at four ports spaced equidistant around a 20.8ft diameter stack. Three traverse points were assigned to each port, re- sulting in a 12-point traverse particulate test. Process and Control Device Description SPS's Harrington Station Unit 3 con- sists of a tangentially fired, Combustion Engineering steam generator, capable of producing 2,682,393 Ib steam/hr at ------- Table 1. Data Summary (Southwestern Public Service, Harrington, Unit 3) OUTLET EMISSION DATA July 8-11, 1981 Orsat Run & Date Paniculate Emissions Ib/W6 Btu Ib/hr gr/acf MW Production % Opacity Flow acfm dscfm Temp. °F CO2 % 02 % CO Baghouse Differential Pressure in. H2O Stack Gas Moisture, % 1 7/8/81 2 7/9/81 3 7/9/81 5 7/10/81 •Void- 0.009 35.91 0.0024 0.009 0.003 34.55 0.0022 11.7- 0.0008 340.5 5.5 349.7 5.5 349.4 5.6 1712460 944245 1737950 911922 1707690 938412 323.2 323.2 322.6 11.6 11.6 11.6 6.5 0.0 6.5 fl.t?1 • 6.5 0.0 East BH: 6.98 West BH: 6.26 East BH: 8.1 West BH: 6.7 East BH: 7.1 West BH: 6.6 10.6 10.2 aINLET EMISSION DATA Run & Date 4 7/9/81 Paniculate Emissions Ib/W6 Btu Ib/hr gr/acf 2.0 7958.5 0.5167 MW Production % Opacity 349.3 98.1 Orsat Flow acfm Temp. CO2 O2 CO dscfm °F % % % 1777280 -..„ 11.6 6.5 0.0 947257 34t° Baghouse Differential Pressure in. H2O East BH: 5.4 West BH: 4.9 Stack Gas Moisture, % 10.7 "Tests conducted by bypassing the baghouse. 2500 psig, 1005°F superheat, and 1005°F reheat. Pulverized Western coal with average parameters of 8475 Btu/lb, 0.3% sulfur, and 5.5% ash is burned. Paniculate emissions are controlled by a Wheelabrator-Frye, Inc., baghouse designed to operate at a flue gas flow of 1,650,000 acfm at 313°F, with a mini- mum design efficiency of 98.6%. Design air/cloth ratio is 2.81 gross, 2.90 with one compartment down, and 3.0 with two compartments down. There are two baghouse systems at Harrington Unit 3, east and west, each with its own operating control system and bypass dampers for start-up, emer- gency operation, and shutdown. This shake/deflate cleaning system consists of 6528 bags: 32 compartments with 204 bags per compartment. Testing Methodology, Sampling Equipment, and Procedures Particulate emission tests were con- ducted according to U.S. EPA Reference Method 5 procedures in conjunction with Methods 1, 2, 3, and 4. Each test included a 12-point traverse with a 10- minute sampling duration for each point. Assembly and use of the impactor train followed state-of-the-art protocol and general Method 5 sampling train procedures. Special precaution was taken to avoid rough handling of loaded impactors, overloading of the impactor, and in the performance of hot leak tests. The particulate sampling equipment used is referred to as the "EPA Method 5 Particulate Sampling Train," designed and developed by EPA. The apparatus consisted of a stainless steel sampling nozzle, a Method 5 filter holder containing an 87 mm Schleicher and Scherell #1-HV high-purity glass fil- ter, a series of four Greenburg-Smith impingers, a check valve, a leakless vac- uum pump, a dry gas meter, and a cali- brated orifice. The impingers and con- necting tubes were made of Pyrex glass and were connected with glass ball- and-socket joints. The probe was Type 316 stainless steel. Using the type "S" pitot tube, a veloc- ity traverse was performed along each traverse axis during each particulate run. The velocity pressure at each sam- pling point was measured using an in- clined manometer. Prior to, and at the conclusion of, each run, the complete sampling train, in- cluding probe and nozzle, was leak- tested by plugging the nozzle with a rubber stopper and applying a vacuum of 15 in. Hg to the system. At the completion of each test the sampling nozzle, the inside of the probe, the inside of the thimble holder, and the front-half of the glass fiber filter holder were washed with acetone. The wash- ings were collected and stored. Tests to determine carbon dioxide, oxygen, and carbon monoxide were conducted using an Orsat analysis ac- cording to EPA Method 3. Discussion of Results During this test series, there were no ------- deviations from normal operating con- ditions for Unit 3 that could be deter- mined from the control room, baghouse control room data, or conferences with the boiler operators. Control room data monitored during the test period com- pared closely with previous data. Emis- sion rates in pounds per million Btu were calculated using average F factors derived from coal analysis performed on SPS coal. Outlet paniculate emis- sions at Harrington Unit 3 averaged 0.007 lb/106 Btu (27.39 Ib/hr) for the three outlet tests. The single inlet emis- sion test, performed by bypassing the baghouse, resulted in an emission rate of 2.0 lb/106 Btu (7958.5 Ib/hr). Emission rates for outlet Runs 2 and 3 were very close: 0;009 lb/106 Btu (35.9 Ib/hr), and 0.009 lb/106 Btu (34.6 Ib/hr), respec- tively. Outlet Run 5 resulted in an unex- pected low emission rate of 0.003 lb/106 Btu (11.7 Ib/hr). The inlet run performed at the outlet stack ports also resulted in a lower emission rate than expected when compared to previous test results conducted at Unit 3. Cascade impactor sampling was per- formed under the supervision of Re- search Triangle Institute, July 8 and 9, 1981. Three outlet and one inlet im- pactor tests were performed. Emission rates for the impactor and Method 5 tests are compared in Table 2. After comparing previous ash analy- ses with the analysis made on the ash generated during this test series, it was concluded that little difference existed between the typical coal burned and the coal burned during this testing project. The coal burned at Southwestern Public Service Co. is a western coal, high in calcium and silica, mined from the Pow- der River Basin near Gillette, Wyoming. Bag analyses were performed on four bags (two used and two new). The tests performed included permeability, ten- sile strength, MIT flex, and Mullen burst. Overall, the bag analyses indicated a greater percentage loss of strength in terms of MIT flex, Mullen burst, and ten- sile strength than previous tests. The cleaning procedures differed between the two testing laboratories but, in gen- eral, indicated approximately the same permeability improvement after clean- ing. Table 2. Gram Loading from Method 5 and Impactor Tests (SPS-Harrington Unit 3) July 8-11, 1981 Run & Date Impactor Loading Rates gr/acf Method 5 Paniculate Loading gr/acf Outlet 7/8/81 Outlet 7/8/81 Outlet 7/9/81 Outlet 7/10/81 alnlet 7/9/81 1 0.0004 2 — 3 0.0007 5 0.0015 4 1.06 VOID 0.0024 0.0022 0.0008 0.5167 "Test performed by bypassing the baghouse. Metric Conversions This Summary includes certain non- metric units for the reader's conve- nience. Those more familiar with the metric system may use the following conversion factors. Nonmetric Times Yields Metric Btu °F ft ft3 gr in. in.2 Ib 1.055 5/9 (°F-32) 30.48 28.32 0.065 2.54 6.45 0.454 kJ °C cm 1 9 cm cm2 kg ------- J. Richardson, J. McKenna, andJ. Mycock are with ETS, Inc.. Roanoke. VA 24018. Dale L. Harmon is the EPA Project Officer fsee below). The complete report, entitled "An Evaluation of Full-scale Fabric Filters on Utility Boilers: SPS Harrington Station Unit 3," (Order No. PB 85-235 513/AS; Cost: $16.00, subject to change! will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Air and Energy Engineering Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Official Business Penalty for Private Use $300 EPA/600/S7-85/031a 0CQ0329 PS U S ENVIR PROTECTION AGENCY REGION 5 LIBRARY 230 S DEARBORN STREET CHICAGO IL «Q*G4 ------- United States Environmental Protection Agency Air and Energy Engineering Research Laboratory Research Triangle Park NC 27711 Research and Development EPA/600/S7-85/030 Nov. 1985 SEPA Project Summary Steam Stripping of Fixed-Bed Gasification Wastewaters F. D. Skinner and B. J. Hayes Laboratory- and bench-scale steam stripping tests were conducted using gas liquor from a fixed-bed coal gasifier at the Department of Energy's Morgan- town Energy Technology Center. The gas liquor was pretreated by solvent extraction (for phenol removal) and fil- tered prior to stripping. This report pre- sents the results of the wastewater stripping tests and provides engineer- ing and environmental data for the de- sign of steam strippers for fixed-bed gasification wastewaters. The labora- tory tests were performed primarily to determine the effect of pH on contami- nant removals. During the bench-scale tests, samples of influent, effluent, and overhead vapor and condensate were analyzed for a number of species of po- tential environmental concern (dis- solved gases, sulfur and nitrogen spe- cies, trace metals, organics, and other water quality parameters). Mass trans- fer coefficients for ammonia, carbon dioxide, and hydrogen suffide stripping were calculated. This Project Summary was devel- oped by EPA's Air and Energy Engineer- ing Research Laboratory, Research Tri- angle Park, NC, to announce key findings of the research project that is fully documented in a separate report of the same title (see Project Report or- dering information at back). Introduction The raw gas liquor resulting from fixed-bed coal gasification processes contains a number of contaminants, some of which are present in relatively high concentrations. These include tars and oils, dissolved organic (especially phenols), dissolved gases (e.g., NH3, HCN, H2S, C02), and both suspended and dissolved inorganics. Removal of these contaminants in a multiple step wastewater treatment system has been included in many proposed commercial coal gasification plant designs. Steam stripping for the removal (and, in some cases, recovery) of dissolved ammonia and acid gases is a treatment process which is common to most of these designs. Capital and operating costs for steam stripping systems can account for a significant portion of the overall cost of wastewater treatment in coal gasification plants. There is a need, therefore, to develop design data that can be used to maximize the cost effec- tiveness of this process. In addition, the outlet streams need to be characterized to evaluate the effects of contaminants on downstream process performance. This report presents the results of laboratory- and bench-scale stripping work, using wastewater obtained from a fixed-bed gasifier at the Department of Energy's Morgantown Energy Technol- ogy Center (DOE-METC). Objectives and Approach The principal objectives of the wastewater stripping study were: • To provide data characterizing the various stripper outlet streams for species of environmental interest with respect to potential impacts on downstream process performance and environmental effects. • To develop mass transfer data that could be used in the design of a steam stripper for wastewater from fixed-bed gasifiers following pre- treatment by solvent extraction, fil- tration, and pH adjustment (if needed). ------- To meet these objectives, two series of tests were performed. First laboratory- scale screening tests were conducted to determine the effect of wastewater pH on the removal of dissolved ammonia, H2S, C02, HCN, and the residual phenol remaining after pretreatment by solvent extraction using methyl isobutyl ketone (MIBK). These tests were conducted by heating a measured quantity of pH- adjusted wastewater to 95-100°C in a distillation flask. Gas (air or nitrogen) was sparged through the contents of the flask, and the ammonia concentra- tion and pH of the wastewater were measured over a 270-minute period. After each run the contents of the flask were analyzed for pH, conductivity, total alkalinity, ammonia, cyanide (free and total), phenol, sulfide, sulfite, sulfate, and thiocyanate. The results of the laboratory tests were used to determine the influent pH for the second series of tests, performed using a bench-scale steam stripping ap- paratus. The stripper was a 0.1 m (4 in.) diameter stainless steel column packed with 1.2 m (4 ft) of 0.6 mm (1/4-in.) ce- ramic Intalox saddles. Solvent- extracted wastewater was preheated to 90°C in an electric heater and pumped into the top of the column. Stripping steam entered below the packing and flowed countercurrent to the waste- water. Concentric tube heat exchangers were used to condense and cool the overhead vapor and to cool the stripped effluent. Samples were obtained peri- odically during each run of the various inlet and outlet streams for analysis of the environmental and performance parameters of interest. The principal in- dependent variable during these tests was the steam to wastewater influent flow ratio. Results and Conclusions Laboratory-Scale Results • Greater than 90% removals of dis- solved ammonia and alkalinity (due to dissolved C02) were obtained by stripping the extracted METC wastewater at a pH of 8.6 (existing after solvent extraction) and higher. Ammonia removal increased from about 92% to over 99.9% as the ini- tial wastewater pH was raised from 8.6 to 11.0. A decrease in C02 re- moval efficiency from over 99 to 96% was observed when increasing pH from 8.6 to 11.0. • Dissolved H20 removals decreased from 80 to 50% as the initial pH was increased from 8.6 to 11.0. Cyanide removals were between 10 and 20%; most of the cyanide content of this wastewater was present as fixed (and therefore non-strippable) cyanide at the pHs evaluated. It is likely that some of the free cyanide initially in the wastewater had been converted to fixed cyanide and/or thiocyanate, and the removal may be higher for "fresh" wastewater. • Removal of the small amount of phenol (total) remaining in the ex- tracted wastewater was found to be less than 20%. No clear trends were observed as the pH was increased. Technical questions remain regard- ing phenol stripping for unextracted wastewater or wastewater having phenol levels closer to those ex- pected from commercial gasifier systems. • Because of its buffering capacity (due to HC03/C03 alkalinity), the wastewater required a significant quantity of lime to raise its pH from 8.6 to 11.0. In order to go from pH 8.6 to 9.5, 470 milliequivalents (meq)of lime per liter was required: to go from pH 8.5 to 11, nearly 1200 meq of lime per liter was needed. The buffering capacity of the waste- water is readily reduced by steam stripping of the dissolved C02. • Two-stage stripping would likely be required to remove all (or nearly all) of the dissolved ammonia and acid gas species (expecially H2S and HCN) from this wastewater. • Stripping the wastewater at a pH of 8.6 produced significant quantities of solids that collected on the sur- faces of the equipment and led to plugging problems. Increasing the pH to 9.5 or higher by lime addition significantly reduced the plugging. The solids are likely ammonium salts, possibly ammonium carbon- ate or ammonium carbamate; how- ever, the solids were not analyzed. Bench-Scale Tests: Environmental • Thiocyanate, sulfate, fluoride, and chloride are not removed by steam stripping. These contaminants will be found in the stripper effluent stream. • Trace elements detected in stripper outlet streams appear to be largely system contaminants, possibly from the column, ceramic packing, and the lime added for pH adjust- ment. It appears that some of the volatile trace elements (e.g., ar- senic, selenium, and antimony) are stripped to some extent. This has implications for the potential envi- ronmental impacts of the stripper overheads and effluent streams. For example, it may be possible to re- duce the amounts of some toxic trace elements thafmight otherwise concentrate in brines produced by downstream evaporators; however, this potential was not investigated. • Phenols were the major organic species found in the wastewater. 2, 4-dimethyl phenol was largely stripped and was found principally in the overhead condensate. Other phenols (e.g., phenol, cresol, and other xylenols) were only partially stripped and are found in both the effluent and overhead condensate. • Hydrocarbon analyses of the over- head vapor were hampered by the relatively high concentration of residual methyl isobutyl ketone (MIBK) from the solvent extraction process. Toluene and xylene were not detected in any of the samples, and benzene was detected (at 1.1 ppmv) in only one set of the samples collected on charcoal. The residual solubility of MIBK in water is significant (reportedly about 2% by weight). Some other organics may be present in the MIBK layer produced as a result of condensing the stripper overhead stream. The solvent layer was not analyzed in this work. • The presence of significant quanti- ties of solvent vapor in the stripper overhead vapor stream has poten- tial impacts on the downstream pro- cesses that may be used to remove H2S and other acid gas species from this stream.The residual solvent concentration after extraction/inert gas stripping seen in this study is probably not representative of com- mercial operations. In a commercial extraction system, solvent recovery would be more efficient, not only to reduce the possibility of problems with downstream processes, but also to reduce solvent makeup re- quirements. However, more effi- cient solvent stripping would likely produce additional streams contain- ing species stripped from the raffi- nate (including ammonia and hy- drogen sulfide). • Carbonyl sulfide was detected in all ------- Table 1. Component Removal Summary for Bench-Scale Stripping Tests3 % Removal Run Date 9/25 9/27 11/1 11/2 n/5" Steam/Influent kg/m3 133 ± 7 298 ± 31 282 ±44 459 ± 49 297 ± 34 Influent PH 9.03 ± 0.05 9.02 ±0.15 9.14 ± 0.03 9.17 + 0.06 8.45 ± 0.04 NH3 83.6 ± 3.8 94.3 ± 1.5 91.6 ± 4.5 95.0 ± 1.7 31.0 ± 25.6 CO2 93.3 ± 0.5 98.3 ± 0.6 98.6 ± 0.2 99.0 ± 0. 1 8T.6 + 5.1 Sulfide 30.9 ± 17.9 17.7 ±38.7 65. 1 ±9.1 69.4 ± 16.6 -7.6±S1.8 Total Cyanide 18.9 ± 4.2 23.8 ± 13.6 59.7 ± 3.5 16.6 ±41.1 23.7 ± 21.0 Total Phenols -5.2 ± 5.4 44.9 ±9.7 -72.8 ±28.6 4.2 ± 37.8 46.7 ± 12.4 aValues shown are mean ± sample standard deviation. All runs performed using 1.2 m (4 ft) of packing. bn/5 run performed using effluent collected from previous stripping runs at similar steam/influent ratios. overhead vapor samples at concen- trations of about 0.04 ppmv in the two-pass stripper run and from 1 to 5 ppmv in the single-pass runs. Car- bon disulfide was the only other sul- fur species detected (1 to 32 ppmv). Bench-Scale Tests: Performance • Contaminant removals consistent with the results of the laboratory- scale tests were found for ammo- nia, C02, and H2S. Data scatter pre- cluded the development of meaningful correlations for HCN and phenol (total) removal as a function of the steam to influent ratio. The component removals are summarized in Table 1. • Contaminant removals were found to increase with increasing steam to wastewater ratio up to 250-300 kg steam/m3 wastewater. Higher ratios produced no statistically significant improvement in contaminant re- movals. There would appear to be little incentive to operate at a steam to wastewater ratio higher than about 250 kg/m3. • Overall volumetric mass transfer coefficients (KLa) were calculated for steam stripping of NHj, CO2, and H2S for the wastewater. For ammo- nia, KLa increased from 1.8 to 6.6 hr~1 as the liquid mass velocity in- creased from about 550 to 2000 kg/ m2hr. Over this same range KLa for CO2 increased from 2.2 to 4.5 hr1. KLa for H2S was found to be approx- imately constant at 0.6 hr~1 over this range of liquid flow rates. ------- F. D. Skinner andB. J. Hayes are with Radian Corporation, Austin, TX 78766. William J. Rhodes is the EPA Project Officer (see below). The complete report, entitled "Steam Stripping of Fixed-Bed Gasification Wastewaters,"(OrderNo. PB85-247450; Cost: $16.95, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Air and Energy Engineering Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 United States Environmental Protection Agency" Center for Environmental Research Information Cincinnati OH 45268 Official Business Penalty for Private Use $300 EPA/600/S7-85/030 0030329 PS U S ENVIR PROTECTION AGENCY REGION 5 LIBRARY 230 $ DEARBORN STREET CHICAGO li. 60604 ------- |