EMBReport88CEPI3 Neshap Screening Method Chromium Emission Test Report Roll Technology Corporation Greenville, South Carolina U. S. ENVIRONMENTAL PROTECTION AGENCY Office of Air and Radiation Office of Air Quality Planning and Standards Research Triangle Park, North Carolina 27711 AUGUST 1983 ------- EPA Contract No. 68-02-4346 Work Assignment 4 December 8, 1988 Final Report Determination of the Efficiency of a Four-Stage Mist Eliminator Candidate Plant ROLL Technology Corporation Greenville, South Carolina Prepared for U.S. Environmental Protection Agency Emissions Measurement Branch Research Triangle Park, North Carolina 27711 Prepared by PEER Consultants, P.C. 4134 Linden Ave., Suite 202 Dayton, Ohio 45432 ------- CONTENTS SECTION PAGE 1.0 INTRODUCTION 1-1 2.0 PROCESS OPERATION ..... 2-1 PROCESS DESCRIPTION 2-1 AIR POLLUTION CONTROL 2-2 PROCESS CONDITIONS DURING TESTING 2-5 3.0 SUMMARY OF RESULTS 3-1 INTRODUCTION 3-1 HEXAVALENT CHROMIUM EMISSION RESULTS 3-1 Inlet to the Moisture Extractor 3-3 Inlet to the M1st Eliminator 3-7 Outlet from the Mist Eliminator 3-8 Plating Tank Solutions 3-11 MOISTURE EXTRACTOR RINSE 3-11 MIST ELIMINATOR RINSE 3-11 4.0 SAMPLING LOCATIONS AND TEST METHODS 4-1 PLATING TANK EMISSIONS 4-1 Location of Measurement Sites 4-1 Test Methods 4-5 EMISSION SAMPLE ANALYSIS 4-7 5.0 QUALITY ASSURANCE 5-1 INTRODUCTION 5-1 FIELD QUALITY ASSURANCE PROCEDURES 5-1 Sample Blanks 5-1 Duplicate Samples 5-2 Standards 5-2 Chain of Custody 5-2 Sample Transfer 5-4 SAMPLING TRAIN COMPONENTS 5-4 VERIFICATION OF CALCULATIONS 5-4 Emission Calculations 5-4 Chromium Concentration Calculations 5-4 ------- FIGURES Number Page 1-1 Location of sample sites 1-2 2-1 Schematic of control device system 2-3 2-2 Cross section of mist eliminator 2-4 2-3 Overlapping-type blade design for chevron-blade mist eliminators 2-6 TABLES Page 2-1 Average Operating Parameters During Each Mass Emission Test Run 2-8 2-2 Total Current Supplied to Tank No. 6 During Each Mass Emission Test Run 2-8 3-1 Schedule of Activities 3-2 3-2 Summary of Flue Gas Conditions 3-4 3-3A Summary of Sample Volumes, Analytical Results and Emission Rates for the Moisture Extractor Inlet . . . . 3-5 3-3B Summary of Screening Method Results at Moisture Extractor Inlet 3-5 3-4 Summary of Sample Volumes, Analytical Results and Emission Rates for the Mist Eliminator Inlet 3-9 3-5A Summary of Sample Volumes, Analytical Results and Emission Rates for the Mist Eliminator Outlet 3-10 3-5B Summary of Screening Method Results at Mist Eliminator Outlet 3-10 3-6 Summary of Cr+6 Removal Efficiencies 3-12 3-7 Summary of Plating Solution and Rinseate Analytical Results 3-12 4-1 Sample Traverse Point Locations for Moisture Extractor Inlet and Mist Eliminator Outlet . . 4-3 4-2 Sample Traverse Point Locations Mist Eliminator Inlet . . . 4-4 5-1 Summary of Analytical Results for QA/QC Samples and Blanks 5-3 IV ------- SECTION 1.0 INTRODUCTION During the week of August 8, 1988, an emission measurement program was conducted at ROLL Technology Corporation in Greenville, South Carolina. Primary purpose of this program was to collect data to determine the efficiency of a four-stage horizontal-flow mist eliminator. Based on this determination, it may be necessary to develop a regulatory alternative based on this type of system. The principal reason for selecting ROLL Technology was that the plant had recently installed a new fume control system. This new mist eliminator system combines both single chevron blade and the mesh pad configuration into one unit. (ROLL had participated in this EPA test program in the past when the EPA tested the packed-bed scrubber/mist eliminator on another plating tank.) The capture and control system servicing Tank No. 6 consists of a double-sided down-draft hood ducted to a moisture extractor followed by a mist eliminator unit containing two sets of overlapping-type blades followed by two mesh pads. Hexavalent chromium emissions were measured at three locations along the duct. These locations are identified in Figure 1-1 as: (1) IA - inlet to the moisture extractor, (2) IB - inlet to the mist eliminator, and (3) 0-outlet from the mist eliminator. Figure 1-1 is a schematic showing the sample locations. The emission samples were collected using the Modified Method 13B (MM13B) sample train and the EPA Screening Method. These methods will be discussed in Section 4.0. The samples were analyzed for Cr+6 concentration using the diphenylcarbazide colorimetric method. This method will also be discussed later in Section 4.0. 1-1 ------- Outlet (O) 96" \ Mist Eliminator 75- Flow En let (IB) _J \\\\\\\\\\\\\\\\\\ \ Mist Elimlnator Rinse Return Line Moisture Extractor Drian Line Roof —7 rt- Moisture Extractor 108" Inlet (IA) Tank No. 6 C " D Figure 1-1. Location of sample sites. ------- PEER Consultants, P.C., located In Dayton, Ohio was responsible for developing the test protocol, conducting the field test, on-site analysis of sample.s and the preparation of draft and final reports. PEER was supported by its subcontractor, Pacific Environmental Services, Inc. located in Cincinnati, Ohio. Midwest Research Institute, located in Raleigh, North Carolina, was responsible for monitoring the process operation, and the EPA conducted Screening Method testing and monitored the implementation of the test protocol. 1-3 ------- SECTION 2.0 PROCESS OPERATION 2.1 PROCESS DESCRIPTION Roll Technology, Inc., is a job specializing in precision finishing and refinishing of industrial rolls. Operations performed at this facility include hard chromium plating, sulfamate nickel plating, machining, grinding, and mirror finishing. The plant plates rolls that are used primarily in the paper manufacturing, roofing, laminating, and coating industries. There are seven hard chromium plating tanks at this facility. On the average, the tanks are charged for a total of 20 hours per day. Approximately 4 hours per day are required for the change-over of rolls. During a change-over, the roll that has been plated is raised out of the plating tank, rinsed with water from a hose, and transferred to the grinding area. Then, the roll to be plated is cleaned with an abrasive cleaner and lowered into the plating solution. Typical plating times range from 12 to 36 hours per roll. Rolls that require longer plating times typically are plated overnight, and rolls that require shorter plating times are plated during the day when personnel are available to perform the change-over. Tank No. 6 was tested during this source test program. Tank No. 6 is 3.65 m (12.0 ft) long, 0.91 m (3.0 ft) wide, and 2.9 m (9.6 ft) deep and holds approximately 9,270 liters (2,450 gallons) of plating solution. The plating solution contains chromic acid in a concentration of 250 grams per liter (g/2.) (33 ounces per gallon) of 2-1 ------- water. Sulfuric acid is used as a catalyst at a bath concentration of 2.5 g/fi. (0.033 oz/gal). The temperature of the plating solution is maintained between 57° and 60eC (135° and 140°F). Tank No. 6 is used to plate small industrial rolls, aircraft engine pistons, and rotary pumps. A manual hoist transfers parts in and out of the plating tank. In addition, the plating tank is equipped with a timer which allows plating operations to be completed during evenings and weekends when no personnel are available to turn off the rectifier. The typical current and voltage applied to Tank No. 6 is 8,000 amperes and 12 volts. Tank No. 6 is typical of other hard chromium plating tanks used in the electroplating industry, based on operating parameters such as current, voltage, plating time, temperature, and chromic acid concentration. Although the composition of the plating solutions remains constant, the operating voltage and current vary with each roll that is plated. 2.2 AIR POLLUTION CONTROL The capture and control system on Tank No. 6 consists of a double-sided lateral hood ducted to a moisture extractor followed by a mist eliminator unit containing two sets of overlapping-type blades and two mesh pads. Figure 2-1 presents a schematic of the capture and control system on Tank No. 6. The fan used in the ventilation system is rated at 255 cubic meters per minute (9,000 cubic feet per minute). The four stage mist eliminator unit was fabricated and installed by KCH Services, Inc., in June 1988. This unit replaced the scrubber that was formerly used to control chromic acid mist from the plating tank. Figure 2-2 presents a cross-sectional view of the mist eliminator unit. This unit has a design airflow rate of 280 m3/min (10,000 ft3/min) and a design pressure drop of 0.62 kilopascal (2.5 inches of water column) at a velocity of 140 meters per minute (450 feet per minute). The blade section consists of two sets of overlapping-type blades. 2-2 ------- fmr/M 22-1 SUCK FtH M7/F. •/roc mzj tLm* lurto *t son a* *i 4.y sunc am , IW7. nm * 10 n> ?JO*6D44D TOOK miat 300 t. HOT CLOUIUm MO F4H MUStW TO If OF . j. omf--nrf i nc »t*f *m ICM f . a uc sr*i mints 4. IMSTAllBS: Of X OfOff ROOT 9M(l flF OOItt(V TO 41LO* IffCfO SP*C{ BfnffH HUIPX*r AfO P4MPII Hit a- CM. OH. OISTM OOF opeiite ou. msnif armcm — mSHfe AFKNOT [tHHJSl HOOD LIB VAT I ON VIEV Figure 2-1. Schematic of control device system. 2-3 ------- MESH-PAD HIST ELIMINATION SECTION 54*0 DOUBLE BAFRE SECTION Figure 2-2. Cross section of mist eliminator. 2-4 ------- Catchments are located along the overlapping edges of the blades and act as collection troughs that provide a central location for droplet collection and facilitate gravitational drainage of the droplets Into a collection sump. Figure 2-3 presents a schematic of this type blade design. Two sets of spray nozzles (three nozzles per set) are located prior to each set of blades and are activated periodically to wash down the blades. The wash down water is drained to a holding tank and recirculated to the plating tank to replace plating solution evaporation losses. The mesh pad section consists of two mesh pads in series. The mesh pads are manufactured by Kimre, Incorporated. Each mesh pad is about 1.4 m (4.5 ft) high, 1.45 m (4.75 ft) wide, and 0.15 m (0.5 ft) deep. Each pad consists of eight layers of mesh. Each layer consists of interlocked polypropylene filaments. Each filament is 0.094 cm (0.037 m) in diameter. The first two layers of each pad have a void space of 97 percent, and the remaining six layers have a void space of 94 percent. The 22-inch diameter moisture extractor is located in the ductwork near the ceiling of the plating shop. Because moisture extractors are designed for the removal of large droplets that also would be collected in the first stage of the mist eliminator unit, the overall performance measured during testing would be equal to the average performance of the mist eliminator unit used alone. During testing, the airflow rate at the outlet of the mist eliminator averaged 195 m3/min (6,880 ft3/min) and the pressure drop was measured at 0.84 kPa (3.4 in. of water column). 2.3 PROCESS CONDITIONS DURING TESTING Mass emission tests were conducted at the following locations: (1) the inlet of the moisture extractor, (2) between the moisture extractor and mist eliminator unit, and (3) the outlet of the mist eliminator unit to characterize the performance of the control devices independently and in series. 2-5 ------- OVERLAPPING BLADES SERVE AS CATCHMENTS MIST-LADEN GAS STREAM CONTROLLED GAS STREAM DROPLETS TO COLLECTION SUMP Figure 2-3. Overlapping-type blade design for chevron-blade mist eliminators. 2-6 ------- Process parameters recorded during each test run were the plating solution temperature, operating voltage, and operating current. Process data sheets documenting the process paramenters monitored during testing are presented in Appendix A. Data on the average operating parameters recorded for each test run are presented in Table 2-1. The process was operating normally during emission testing. The plating tank was plating two industrial rolls during each source test. The two rolls were identical in size. Each roll measured 69 cm (27 in.) long with a diameter of 41 cm (16 in.). The total current supplied to the tanks during each test run was calculated in terms of ampere-hours and is reported in Appendix A. A summary of the total current values is presented in Table 2-2. Grab samples from the plating tank were taken during each test run to determine the chromic acid concentration of the plating solution during emission testing. The mist eliminator was washed down with clean water at the beginning of each day, and grab samples of the mist eliminator washdown water were collected. The chromic acid concentrations of the grab samples are reported in Section 3 of this report. Test run No. 1 was 3 hours in duration and two subsequent runs were 2 hours in duration. Each test run was interrupted 10 to 15 minutes to change test ports. Test run No. 1 was interrupted for 14 minutes due to a power loss to the meter boxes. However, no other process interruptions occurred during the test runs. 2-7 ------- TABLE 2-1. AVERAGE OPERATING PARAMETERS DURING EACH MASS EMISSION TEST RUN Run No. 1 2 3 Operating current, amperes 4,800 5,200 5,200 Operating voltage, volts 6.8 7.0 7.3 Temperature of plating solution, °C (°F) 54 (130) 54 (130) 54 (130) TABLE 2-2. TOTAL CURRENT SUPPLIED TO TANK NO. 6 DURING EACH MASS EMISSION TEST RUN Run No. Test time, h Total current, ampere-hours 1 2 3 3.0 2.0 2.0 14,400 10,400 10,400 2-8 ------- SECTION 3.0 SUMMARY OF RESULTS INTRODUCTION Modified Method 13B (MM13B) samples and the Screening Methods (SM) Samples were collected In triplicate at each sample location. All of the emission samples were analyzed on site for Cr+6 concentrations using the procedures outlined in "Draft Method-Determination of Hexavalent Chromium in Dry Particulate Emissions from Stationary Sources". This analytical method is presented in Appendix D. In addition to the emission samples, grab samples of the plating bath were composited during each MM13B run and analyzed using the same procedures as for the emission samples. Prior to beginning the test program, the moisture extractor and the mist eliminator units were rinsed with freshwater. Following the daily test program, these two units were rinsed again and samples of the rinseate were collected in a polyethylene sample bottle. These samples were also analyzed for Cr+6 concentrations. Table 3-1 presents a schedule of the activities during the test program, the results from the aforementioned analysis are presented in the remainder of this section. HEXAVALENT CHROMIUM EMISSION RESULTS Emission samples were collected isokinetically using a method 13B sample train that had been modified by removing the glass fiber filter and placing 100 ma of 0.1N NaOH in each of the first two impingers. The impinger solutions were recovered into tared polyethylene sample bottles and the total volume of the recovered samples was determined 3-1 ------- TABLE 3-1. SCHEDULE OF ACTIVITIES Date (1988) 8/9 8/9 8/9 8/9 8/9 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/11 8/11 Sample Type MM13B SM SM SM SM Plating Sol. Moist. Ext. Rinse Mist Elim. Rinse MM13B SM SM SM SM Plating Sol. MM13B SM SM SM SM Plating Sol. Moist. Ext. Rinse Mist Elim. Rinse Run No. Test Time IA-1, IB-1, 0-1 192 min. 1A series IB series 11A thru 14H 15E thru 18H 1 1 1 IA-2, IB-2, 0-2 120 min. 2A series 2B series 21A thru 24D 25E thru 28H 2 IA-3, IB-3, 0-3 120 min. 3A series 3B series 31A thru 34D 35E thru 38H 3 2 and 3 2 and 3 Parameter Measured Cr+6 Cr+6 Cr*6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr*6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 Cr+6 3-2 ------- gravimetrlcally. Following recovery of the samples, an alipuot-of the solution was analyzed for Cr+6. The following subsections present the flue gas data and analytical results for each sample location. Inlet to the Moisture Extractor Modified Method 13B— A summary of the flue gas conditions at this location are presented in Table 3-2. The volumetric flowrates were consistent and averaged 171 dry standard cubic meters per minute (dscmm), (6041 dry standard cubic feet per minute, (dscfm). The flue gas temperature averaged 34°C (93°F) and the moisture content averaged 2.52 percent. The flue gas was essentially ambient air and was assigned a dry molecular weight of 28.95 Ib/lb mole. The isokinetic sampling rates were within the allowable limitations for these sample runs. Prior to sampling, it was decided that the first MM13B should be run at 8 minutes per point for a total sample time of 192 minutes. This sample time ensured the collection of a detectable concentration of Cr"1"6. Following the analysis of the sample, it was determined that the sample time per point could be reduced to 5 minutes for a total sample time of 120 minutes. The uncontrolled emissions for each MM13B were consistent and averaged 3.073 milligrams per dry standard cubic meter (mg/dscm). A summary of the MM13B sample volumes, analytical results and emission rates for this location is presented in Table 3-3A. Screening Method— A summary of the sample volumes, analytical results and emission rates for the Screening Methods conducted at this location is presented in Table 3-3B. 3-3 ------- TABLE 3-2. SUMMARY OF FLUE GAS CONDITIONS Run No. IA-1 IB-1 0-1 IA-2 IB-2 0-2 IA-3 IB-3 0-3 Date 8/9/88 8/9/88 8/9/88 8/10/88 8/10/88 8/10/88 8/10/88 8/10/88 8/10/88 Volumetric Flowrate ds cm/mi n dscf/min 177 6.248 167 178 166 176 171 170 168 176 5,906 6,279 5,868 6.207 6,053 6.006 5.918 6.229 Temperature 8C °F 34 36 37 32 34 36 35 37 37 93 97 98 90 93 96 95 99 99 % Moisture 2.32 2.50 2.43 2.62 2.80 3.04 2.62 2.59 2.50 % Isokinetic 96.1 . 101.9 96.7 98.3 98.6 97.1 96.2 99.2 94.4 3-4 ------- TABLE 3-3A. SUMMARY OF SAMPLE VOLUMES, ANALYTICAL RESULTS AND EMISSION RATES FOR THE MOISTURE EXTRACTOR INLET Run No. IA-1 IA-2 IA-3 Vol ume dscm 6.482 3.891 3.897 Metered dscf 228.891 137.395 137.602 Total Mass Cr+6. ma 18.3771 12.9125 11.9374 Concentration *ma/dscm 2.835 3.319 3.064 ar/dscf 0.0012 0.0015 0.0013 Emission Rates kq/hr 0.030 0.033 -0.031 Ib/hr 0.066 0.073 0.069 TABLE 3-3B. SUMMARY OF SCREENING METHOD RESULTS AT MOISTURE EXTRACTOR INLET Screening Methods 1-A-l l-A-2 1-A-3 1-A-4 l-B-1 l-B-2 l-B-3 1-B-4 2-A-l 2-A-2 2-A-3 2-A-4 2-B-l 2-B-2 2-B-3 2-B-4 3-A-l 3-A-2 3-A-3 3-A-4 3-B-l 3-B-2 3-B-3 3-B-4 Vol ume Samol ed 8.856 8.390 8.754 7.509 8.856 8.390 8.754 7.509 8.698 8.241 8.599 7.375 8.698 8.241 8.599 7.375 8.922 8.453 8.819 7.564 8.922 8.453 8.819 7.564 Stack DSCFM 6245 6245 6245 6245 6245 6245 6245 6245 5873 5873 5873 5873 5873 5873 5873 5873 6006 6006 6006 6006 6006 6006 6006 6006 Screen Ma/m3 0.3935 1.331 0.5337 2.114 1.745 1.870 2.127 1.438 2.520 3.144 3.213 4.937 2.063 2.404 2.377 4.293 1.381 1.377 1.880 2.001 1.864 1.815 5.999 5.831 "MM13B Ma/m3 2.84 2.84 2.84 2.84 2.84 2.84 2.84 2.84 3.32 3.32 3.32 3.32 3.32 3.32 3.32 3.32 3.07 3.07 3.07 3.07 3.07 3.07 3.07 3.07 * of MM13B 13.86 46.87 18.79 74.44 61.44 65.85 74.89 50.63 75.90 94.67 96.77 148.70 62.14 72.41 71.60 129.31 44.98 44.85 61.24 65.18 60.72 59.10 195.41 189.93 * Slight discrepancy due to difference in ejaculation procedure 3-5 ------- Sample volumes for the, screening method were controlled by critical orifices made from hypodermic needles. The needles were checked for flow rates in the lab by using a wet test meter in series with each needle. Numbers obtained in the lab were used to determine sample volumes in the field by applying the ideal gas law. Unfortunately, the ideal gas law is not the correct way to determine sample volumes when using hypodermic needles as critical orfices. When components are assembled as they are in the screening train, the ideal gas law will not give satisfactory volumes. As air goes through the system and reaches the hypodermic needle, it is cooled. This changes the diameter of the orfice (hypodermic needle) and the cooler temperature may also change the speed of sound. The upstream and downstream pressures also influence the volume, and the simple calculations of the ideal gas law do not give precise volumes. In order to keep the screening method simple, samples volumes should be determined empirically at the site. The ideal gas law was used to determine sample volumes at the Roll Technology sites, and was adjusted for temperature differences for the speed of sound. While this seemed like a logical approach at the time, the volumes are not correct and the true volumes will probably never be known. The data, however, are still useful. The same error was applied to all samples, so at least the samples give some indication as to the precision of the technique. The reproducibility of the data, especially at the outlet is encouraging. The screening method tests run at Saint Clair Shores gave values equivalent to about 70% of the values measured by the MM13B tests, and this was without the use of a filter in the screening trains. The addition of a filter was expected to increase the yield at the Roll Technology test, however, such was not the case. There are some possible explanations for this result. 3-6 ------- First, sampling was done at a single location during both tests. Although the flow through the ducts was fairly uniform, this does not necessarily mean that the distribution of chromic acid 1s uniform throughout the stack gas. Thus the concentration sampled could be high, low, or average, but still 1s probably biased in one (unknown) direction. Another biasing characteristic could be the sample volumes used 1n the calculations. Calculations performed on a smaller than actual volume will bias the results high; calculations performed on a larger than actual volume will bias the results lower. While the results from the screening method were somewhat disappointing, the next test will incorporate stack traverses into the SM sampling. It is hoped that the next phase of the test will provide results that more closely agree with the MM 13B trains. Inlet to the Mist Eliminator After completion of the cyclonic flow check at this location it was determined that special sampling procedures would need to be implemented. The discussion of those sampling procedures will be presented in Section 4.0 of this report. Modified Method 13B— A summary of the flue gas conditions at this location is presented in Table 3-2. The volumetric flowrates were consistent and averaged 170 dry standard cubic meters per minute (dscmm), (6010 dry standard cubic feet per minute, (dscfm). The flue gas temperature averaged 36°C (96°F) and the moisture content averaged 2.63 percent. The flue gas was essentially ambient air and was assigned a dry molecular weight of 28.95 Ib/lb mole. The isokinetic sampling rates were within the allowable limitations for these sample runs. In order to meet the requirements of the cyclonic flow sampling procedure employed, the sampling times at each traverse point were adjusted. Sampling times ranged from 5 minutes, 28 seconds to 11 3-7 ------- minutes, 55 seconds, based on a nominal sampling time of 8 minutes per point, a total sampling time of 187 minutes, 40 seconds was attained. This sample time ensured the collection of a detectable concentration of Cr+6. Following analysis of the sample it was determined that the nominal sampling time could be reduced to 5 minutes per point which corresponded to actual sampling times ranging from 3 minutes, 31 seconds to 5 minutes, 27 seconds, for a total sampling time of 117 minutes, 11 seconds. The emissions for each MM13B run were consistent and averaged 0.436 milligrams per dry standard cubic meter (mg/dscm). A summary of the MM13B sample volumes, analytical results and emission rates for this location is presented in Table 3-4. Outlet from the Mist Eliminator Modified Method 138— A summary of the flue gas conditions at this location is presented in Table 3-2. The volumetric flowrates were consistent and averaged 175 dry standard cubic meters per minute (dscmm), (6187 dry standard cubic feet per minute, (dscfm). The flue gas temperature averaged 37°C (98°F) and the moisture content averaged 2.66 percent. The flue gas was essentially ambient air and was assigned a dry molecular weight of 28.95 Ib/lb mole. The isokinetic sampling rates were within the allowable limitations for these sample runs. Prior to sampling it was decided that the first MM13B should be run at 8 minutes per point for a total sample time of 192 minutes. This increased sample time ensured the collection of a detectable concentration of Cr+6. Following the analysis of the sample, it was determined that the sample time per point could be reduced to 5 minutes for a total sample time of 120 minutes. The uncontrolled emissions for each MM13B were consistent and averaged 0.0409 milligrams per dry standard cubic meter (mg/dscm). A summary of the MM13B sample volumes, analytical results and emission rates for this location is presented in Table 3-5A and screening method results are presented in Table 3-5B. 3-8 ------- TABLE 3-4. SUMMARY OF SAMPLE VOLUMES, ANALYTICAL RESULTS AND EMISSION RATES FOR THE MIST ELIMINATOR INLET Run No. IB-1 IB-2 IB-3 Vol ume dscm 4 3 2 .740 .041 .919 Metered dscf 167 107 103 .357 .397 .069 Total Mass Cr+6. ma 1 1 1 .644 .610 .257 Concentration mq/dscm qr/dscf 0.347 0.00015 0.529 0.431 0 0 .00023 .00019 Emission ka/hr 0.0035 0.0056 0.0043 Rates Ib/hr 0.0076 0.0123 0.0095 3-9 ------- TABLE 3-5A. SUMMARY OF SAMPLE VOLUMES, ANALYTICAL RESULTS AND EMISSION RATES FOR THE MIST ELIMINATOR OUTLET Volume Metered Total Mass Concentration Run No. 0-1 0-2 0-3. Screenina Methods 1-1 -A 1-2-B 1-3-C 1-4-D 1-5-E 1-6-F 1-7-G 1-8-H 2-1 -A 2-2-8 2-3-C 2-4-0 2-5-E 2-6-F 2-7-G 2-8-H 3-1 -A 3-2-B 3-3-C 3-4-D 3-5-E 3-6-F 3-7-G 3-8-H dscm 6.298 3.810 3.811 TABLE Vol ume Samol ed 10.306 14.751 14.665 15.181 10.306 14.751 14.665 15.181 10.449 14.954 14.867 15.391 10.449 14.954 14.867 15.391 10.697 15.310 15.220 15.756 10.695 15.310 15.220 15.756 dscf 222.371 134.539 134.555 Cr+6. ma *ma/dscm 0.189 0.030 0.165 0.043 0.180 0.047 3-5B. SUMMARY OF SCREENING METHOD MIST Stack DSCFM 6279 6279 6279 6279 6279 6279 6279 6279 6051 6051 6051 6051 6051 6051 6051 6051 6218 6218 6218 6218 6218 6218 6218 6218 ELIMINATOR OUTLET Screen ar/dscf 0.00001 0.00002 0.00002 RESULTS AT mq/m3 *MM13B mq/m3 0.01984 0.01740 0.01979 0.01866 0.01096 0.01815 0.01780 0.01584 0.02193 0.01925 0.02126 0.02203 0.01910 0.02000 0.02252 0.02317 0.01898 0.01845 0.01947 0.02017 0.01700 0.01785 0.01680 0.01952 0.0305 0.0305 0.0305 0.0305 0.0305 0.0305 0.0305 0.0305 0.0442 0.0442 0.0442 0.0442 0.0442 0.0442 0.0442 0.0442 0.0481 0.0481 0.0481 0.0481 0.0481 0.0481 0.0481 0.0481 Emission ka/hr 0.0003 0.0004 0.0005 THE Rates Ib/hr 0.0007 0.0010 0.0011 * of MM13B 65.05 57.05 64.89 61.18 35.93 59.51 58.36 51.93 49.62 43.55 48.10 49.84 43.21 45.25 50.95 52.42 39.46 38.36 40.48 41.93 35.34 37.11 34.93 40.58 * Slight discrapency due to difference in calculation procedure. 3-10 ------- The Cr+s removal efficiencies for the system were consistent and averaged 98.6X. A summary of removal efficiencies for the system is presented in Table 3-6. PLATING TANK SOLUTIONS During each MM13B run, grab samples of the plating bath solution were collected and composited. The samples were analyzed for Cr"1"6 concentration. The results from these analyses are presented in Table 3-7. MOISTURE EXTRACTOR RINSE Prior to the start of sampling program the moisture extractor was rinsed with fresh water. The moisture extractor was rinsed daily and a sample of the rinseate was collected and analyzed for Cr"1"6 concentrations. The result of these analyses were presented in Table 3-7. MIST ELIMINATOR RINSE Prior to the start of the sampling program the mist eliminator was rinsed with fresh water. The mist eliminator was rinsed daily and a sample of the rinseate was collected and analyzed for Cr+6 concentrations. The result of these analyses were presented in Table 3-7, 3-11 ------- TABLE 3-6. SUMMARY OF Cr+6 REMOVAL EFFICIENCIES Run No. IA-1 0-1 IA-2 0-2 IA-3 0-3 Average Cr+« Emission Rate Ib/hr 0.0664 0.0007 0.0729 0.0010 0.0689 0.0011 Or*6 Removal Efficiency 98.9% 98.6% 98.4% 98.6% TABLE 3-7. SUMMARY OF PLATING SOLUTION AND RINSEATE ANALYTICAL RESULTS Run No. Cr+6 Concentration, Plating Solution IA-1 145,872 IA-2 115,385 IA-3 119,308 Moisture Extractor Rinseate 8/10/88 3,046 8/11/88 6,551 Mist Eliminator Rinseate 8/10/88 461 8/11/88 620 3-12 ------- SECTION 4.0 SAMPLING LOCATIONS AND TEST METHODS PLATING TANK EMISSIONS Location of Measurement Sites EPA Reference Method 1 "Sample and Velocity Traverse for Stationary Sources" was used to select representative measurement sites. The measurement site at the moisture extractor was located in a 21.5 inch ID circular vertical duct. Two 3-inch ID holes were cut in the duct at 90 degrees from each other. The measurement site was 53 inches (2.5 stack diameters) downstream from the nearest flow disturbance (bath) and 55 inches (2.6 stack diameters) upstream from any flow disturbance (90° bend). The measurement site for the mist eliminator inlet was located in a 21.5 inch ID circular horizontal duct. The measurement site was located 50 inches (2.3 stack diameters) downstream from any flow disturbance (90° bend) and 20 inches (1.2 stack diameters) upstream from the mist eliminator. The measurement site for the outlet of the mist eliminator was located in a 21.5 inch ID circular vertical duct. The measurement site was located 75 inches (3.5 stack diameters) downstream from the mist eliminator outlet and 20 inches (0.9 stack diameters) upstream from the atmosphere. According to EPA Method 1 criteria, each site required 24 sample traverse points, 12 along each diameter. With the EPA Task Manager's 4-1 ------- approval, the point closest to the port was not sampled because the high static pressure at that point might result in erroneous measurements. Thus it was decided instead to sample point #2 twice. The point closest to the bottom of the duct at the inlet to the mist eliminator was not sampled in order to prevent contamination of the sample due to the presence of liquid accumulated on the duct wall. The EPA Task Manager was consulted and approved this change. Table 4-1 and 4-2 show the traverse points used. Prior to sampling, verification of the absence of cyclonic flow at each sample traverse point was assessed based on procedures described in EPA Reference Method 1. In this method the face openings of the Type-S pitot tube are aligned perpendicular to the duct cross-sectional plane, designated "0-degree reference." Null (zero) pitot readings obtained at 0-degree reference indicate an acceptable flow condition at a given point. If the point reading were not zero at 0-degree reference, the pitot was rotated until a null reading was obtained. The value of the rotation angle (yaw) was recorded for each point and averaged across the duct. Method 1 criteria stipulate that average angular rotations greater than 20 degrees indicate cyclonic (nonaxial) flow conditions in the duct. The average of the angular rotation for the moisture extractor inlet was 3 degrees, and the average angular rotation at the mist eliminator outlet was 8 degrees. Both of these sites indicated acceptable flow patterns so that extraction of representative samples from these sites was performed using normal sampling procedures. The average of the angular rotation at the mist eliminator inlet was 24 degrees indicating the presence of cyclonic flow, by exceeding the 20 degree limit allowed by Method 1. With the approval of the EPA Task Manager it was decided that the alignment approach for sampling in cyclonic flow be used. In the alignment method, the sampling rate must be based on the total velocity at each sampling point in order to maintain isokinetic 4-2 ------- TABLE 4-1. SAMPLE TRAVERSE POINT LOCATIONS FOR MOISTURE EXTRACTOR INLET AND MIST ELIMINATOR OUTLET Traverse Point No. 1 2 3 4 5 6 7 8 9 10 11 12 Location Moisture Extractor Mist Inlet 0.5a 1.44 2.54 3.81 5.38 7.65 13.85 16.13 17.69 18.96 20.06 21.00a Eliminator Outlet 0.5a 1.44 2.54 3.81 5.38 7.65 13.85 16.13 17.69 18.96 20.06 21.00a a Relocated to the minimum distance from the inside stack wall. 4-3 ------- TABLE 4-2. SAMPLE TRAVERSE POINT LOCATIONS-MIST ELIMINATOR INLET Traverse Point No. 1 2 3 4 5 6 7 8 9 10 11 12 Traverse 0.5a 1.44 2.54 3.81 5.38 7.65 13.85 16.13 17.69 18.96 20.06 21.00a Anale of Vertical (degrees) C40 C45 C40 C30 C20 C5 CIS CC25 CC20 CC15 CC15 CC20 Flow Deviation Horizontal (degrees) C20 C20 C15 CIS CIS C21 C18 CC 3 CC18 CC44 CC50 CC40 a Relocated to the minimum distance from the inside stack wall. C - Clockwise CC - Counter Clockwise 4-4 ------- sampling conditions. Since the angle between the flow direction and stack axis varies across the stack, the sampling velocity is not weighted proportionally to the axial, velocity component. Proportional sampling requirements can be satisfied by adjusting the sampling time for each sampling point such that the volume of sample collected at each point is related by a constant to the axial velocity component at each point. Thus, *2 = *1 C°S * where t - nominal sampling time per point t o actual sampling time per point * = angle between flow direction and stack axis The nominal sampling time per point should be of sufficient duration to ensure collection of sufficient sample volume to provide accurate mass concentration measurements since the application of the above weighting procedure will reduce the actual sampling time. When sampling to determine a mass emission rate, the volumetric flow rate is determined as: TS PeMe N I ( i &P\ COS $i> N If negative velocities are encountered at particular sampling points, then no sampling should be conducted at those points and the measured negative axial volume flow rate should be subtracted from the measured positive axial volume flow rate. Test Methods Velocity and static pressures, moisture content, and temperature were measured prior to sampling, in order to define sampling rates and 4-5 ------- nozzle sizes as described 1n the EPA Reference Methods 1, 2 and 4. The stack gas molecular weight was not determined by procedures outlined In EPA Method 3. Alternatively, the molecular weight was assigned the value of 29.0 Ib/lb mole, as stated 1n the EPA Method 2, paragraph 3.6. An EPA MM13B sample train was used to collect the Cr*6 samples. The sample train consisted of a 316 stainless steel button-hook design nozzle, an unheated Pyrex glass-lined probe, and a series of four impingers. The first, third and fourth impingers were Greenburg-Smith design, modified by replacing the tip with a 1/2-in. inside diameter glass tube extending to 1/2-in. from the bottom of the flask. The second impinger was a Greenburg-Smith Impinger with the standard tip. The first and second impingers contained lOOmfi, of 0.1N NaOH. The third impinger was empty and the fourth impinger contained approximately 200 grams of silica gel. On Run IA2, 182 and 0-2, an additional silica-gel filled impinger was added to the sampling train. The balance of the sampling system consisted of a vacuum pump, dry gas meter, calibrated orifice and related temperature and pressure indicating apparatus to determine dry gas sample volume, stack gas temperature, volumetric flow rate and isokinetic sampling rates. During sampling, stack gas temperature and the gas temperature exiting the last Impinger were monitored with calibrated thermocouples. The sampling time was varied from 8 minutes per point (192 minute total sample time) to 5 minutes per point (120 minute total sample time) since the concentration of Cr+6 was such that good analytical results could be obtained using the shorter time. The sampling times calculated for the alignment approach varied from 5.63 to 8.75 minutes per point, for an eight minute nominal sampling time and 3.52 to 5.46 minutes for a five minute nominal sampling time. The impingers were weighed before and after each test to determine the moisture content of the flue gas stream. The contents of the 4-6 ------- 1mp1ngers were placed 1n a tared polyethylene container. All connecting glassware, the nozzle and probe were rinsed with 0.1 N NaOH and combined with the impinger solution in the polyethylene sample bottle. In recovering, the mist eliminator outlet samples the probe and nozzle rinse were separated from the impinger catch in an effort to concentrate the Cr*6 in a smaller sample volume. The liquid level was marked on each sample bottle and each bottle was marked indicating the run number and bottle contents. Appropriate blank solutions were collected. The polyethylene containers were all tared before their use and weighed after the collection of the sample. The volume of each solution was determined by multiplying the specific gravity of the solution times the net weight of the solution. Each sample, including blanks, was analyzed for Cr+6 concentration using analytical methodology recently developed by the EPA. EMISSION SAMPLE ANALYSIS The MM13B samples, the Screening Methods samples, the plating solution and the emission control unit rinses were analyzed for Cr+6 concentration. The analyses were conducted on site in the plant's analytical laboratory. Immediately following the sample recovery, the samples were submitted to the analyst and the analyses and the calculations were performed the same day. The analytical results were performed on the Hewlett Packard 41CV computer that was set up in the on-site computer center. The calculations were also performed by the EPA Task Manager. The analytical method entitled "Draft Method - Determination of Hexavalent Chromium in Dry Particulate Emissions from Stationary Sources" was used as a "guideline" in conducting the analyses. This method is currently under development by the EPA and is presented in Appendix D. 4-7 ------- There were several variations between the Draft Method and the Analytical Method that was performed In the field. They are described as follows: 1. The collected samples were not digested in an alkaline solution. Aliquots of the recovered samples were pipeted directly from the sample bottle and prepared as 1n paragraph 5.7.1 of the Draft Method. 2. The pH of the sample aliquot was monitored with a pH meter while adjusting the pH of the aliquot to 2 ± .5. 3. The spectrophotometer was calibrated with standards containing 2 mi, 5 ml, 7 mft, 10 ml, 15 mfi, and 20 mfi, of the 5 jig/mft, working standard. The spectrophotometer calibration factor, K , was calculated as follows: Aj. + 2.5A2 + 3.5A3 + 5A4 + 7.5A5 + 10A6 Kc - 10 222222 AI + A2 + A3 + A4 + A5 + A6 4. The value of this calibration factor was calculated using a computer program that was developed by the EPA Task Manager for the HP41 calculator. 4-8 ------- SECTION 5.0 QUALITY ASSURANCE INTRODUCTION The goal of the quality assurance activities for this project is to ensure, to the highest degree possible, the accuracy of data collected. The procedures contained in the "Quality Assurance Handbook for Air Pollution Measurement Systems," Volume III, "Stationary Source Specific Methods," EPA-600/4-77-027B served as the basis for performance of all testing and related work activities that were undertaken in this testing program. In addition to the quality assurance measure guidelines presented above, specific quality assurance activities were conducted for several of the individual testing activities, as performed; these are presented in the paragraphs that follow. FIELD QUALITY ASSURANCE PROCEDURES In order to assure a high level of quality control while sampling to allow the comparison of data from these two methods, a field quality assurance program was followed during the test program. Methods used to obtain the required level of quality assurance are itemized below. Sample Blanks Reagent Blanks— The 0.1N NaOH absorbing solution was transported to the field in its "as-purchased" container. When in the field, the 0.1N NaOH was transferred to a polyethylene wash bottle. From the wash bottle, the NaOH solution was used for sample train preparation and recovery. A 5-1 ------- blank sample was collected from the solution In the wash bottle. This sample was given to the on-site laboratory personnel with the emission samples, and analyzed in the same manner. Results of the blank analyses are presented in Table 5-1. H 0 Blanks— A distilled water blank was obtained from the wash bottles and analyzed in the same manner as the emission samples. Duplicate Samples One sample for every 10 samples analyzed was a duplicate, e.g., if 24 samples were analyzed, 3 duplicate samples would be analyzed. The analytical results for the duplicated samples are presented in Table 5-1. Standards Daily, throughout the analysis of the samples, standards were set up as a spot check of the spectrophotometer calibration. The results of these checks are presented in Table 5-1. Chain of Custody In an effort to maintain the integrity of all samples taken at the test facility, a chain of custody procedure was followed. A copy of the "Chain-of-Custody" data sheets is included in Appendix E. These sheets include the sample identification, date of sample recovery, name of person who performs the recovery, place of recovery as well as the name of the responsible person from the analytical group who is taking custody of the samples. Once the samples were placed in custody of the analytical group, that group provided for safe storage and maintenance of records sufficient to maintain sample integrity. 5-2 ------- TABLE 5-1. SUMMARY OF ANALYTICAL RESULTS FOR QA/QC SAMPLES AND BLANKS Sample No. 10 pg/mQ. 10 jig/mi 75 yig/mfi, l-A-2 50 yg/mfi. 10 jig/mi 25 pg/mfi, I-A-2 75 pg/mfi. 3-B-4 3-B-3 50 jig/ml Plating Sol. Run 2 Blanks H2° 0.1N NaOH Date (1988) 8/9 8/9 8/9 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/10 8/29 8/29 8/9 8/10 Tvoe of Sample Duplicate Standard Total pa Cr*6 X 10.04 X 10.14 X 74.73 X 344.04 X 53.19 X 10.70 X 26.57 X 12915.53 X 77.28 X 1295.45 X 1565.31 X 47.10 X 114,430 Jig/ml 0.00 0.392 "(316.21) "(12,001.58) "(1248.93) "(1498.24) «(n5,385ng/mfi,) * Original value aginst which duplicate is to be compared 5-3 ------- Sample Transfer All MM13B samples collected during testing remained in the custody of PEER Consultants, P.C. All Screening Methods samples were returned to the EPA after analysis. All samples were secured in the mobile laboratory while in the field. SAMPLING TRAIN COMPONENTS The equipment used in this test program, including nozzles, pi tot tubes, dry gas meters, orifices, and thermocouples were uniquely identified and were calibrated in accordance with calibration procedures specified in the applicable EPA Reference Method prior to, and at the completion of the testing program. The calibration sheets are presented in Appendix F. VERIFICATION OF CALCULATIONS Emission Calculations Dry gas volumes, percent moisture of the stack gas, gas flow rates, and Cr+6 emission rates were calculated using a Hewlett Packard 41CV programmable calculator. The programs used can be found in the document: "Source Test Calculation and Check Programs for Hewlett Packard 41 Calculators" (EPA-340/1-85-018). The results were checked and verified by the contractor task manager. Chromium Concentration Calculations All absorbance data for blanks, standards, samples and QA/QC samples were documented in a notebook. The percent absorbance was calculated from the percent transmittance and subsequent calculations were carried out as described in the draft method for hexavalent chromium analysis. 5-4 ------- |