United States Environmental Protection Agency Office of Air Quality Planning and Standards Research Triangle Park NC 27711 EMB Report 81-PLY-4 January 1982 Air Plywood/Veneer Emission Test Report Georgia-Pacific Springfield Plant Springfield, Oregon ------- PLYWOOD/VENEER EMISSION TEST REPORT GEORGIA-PACIFIC PLYWOOD PLANT SPRINGFIELD, OREGON JUNE 1981 Environmental Consultants, Inc. EMB Report 81-PLY-4 ESED Project 80/02 EPA Contract No. 68-02-3543 Work Assignment No. 1 TRC Project No. 1460-E80-51 Prepared for: C.E. Riley Task Manager Prepared by: Peter W. Kalika, P.E. Program Manager Eugene A. Brackbill, P.E. Work Assignment Manager John H. Powell Project Scientist Eric A. Pearson Project Scientist January 1982 800 Connecticut Blvd. East Hartford, CT 06108 (203) 289-8631 ------- This report has been reviewed by the Emission Standards and Engineering Division, Office of Air Quality Planning and Standards, Office of Air, Noise and Radiation, Environmental Protection Agency, and approved for publication. Mention of company or product names does not constitute endorsement by EPA. Copies are available free of charge to Federal employees, current contractors and grantees, and nonprofit organizations - as supplies permit - from the Library Services Office, MD-35, Environ- mental Protection Agency, Research Triangle Park, NC 27711. Order: EMB Report 81-PLY-4 ------- PREFACE The work described herein was conducted by personnel from TRC - Environ- mental Consultants, Inc., Research Triangle Institute (RTI), Del Green Asso- ciates (DGA); CH~MHill, Engineers, Planners, Economists and Scientists; Pollution Control Science, Inc., (PCS); Georgia-Pacific (G-P) in Springfield, Oregon; the National Council of the Paper Industry for Air and Stream Improve- ment, Inc. (NCASI); and the United States Environmental Protection Agency (EPA) Emission Measurement Branch (EMB). The scope of work was issued under EPA Contract 68-02-3543, Work Assign- ment 1. The work was performed under the supervision of Eugene A. Brackbill, P.E., TRC work assignment manager, and John H. Powell, TRC field crew chief. Robert L. Chessin of RTI monitored process operations and was assisted by Paul Willhite of DGA. RTI was responsible for preparing Section 3 and Appen- dix I of this report, both of which deal with process descriptions and opera- tions. Mark S. Boedigheimer supervised Method 5X analyses performed by CHJIHill. David Robinson supervised Method 25 analysis performed by PCS. Victor Dallons supervised NCASI sampling and analysis activities as well as providing helpful suggestions and comments in support of the test program. Mitch Steffensen and Pete Fetter of Georgia-Pacific, provided invaluable assistance and guidance to TRC, EPA and RTI in the performance of the test program. Clyde E. Riley, Office of Air Quality Planning and Standards (OAQPS), Emission Measurement Branch, EPA, served as task manager and was responsible for coordinating the test program. Edwin J. Vincent, OAQPS, Chemical and Petroleum Branch, EPA, served as project lead engineer. He was also responsible for coordinating and directing the process operations monitoring. -11- ------- TABLE OF CONTENTS SECTION PAGE PREFACE ii 1.0 INTRODUCTION 1-1 1.1 Background 1-1 1.2 Summary of Process and Emissions 1-2 1.3 Applicability of EPA Reference Test Methods 1-3 1.3.1 EPA Method 5X (Provisional) 1-3 1.3.2 EPA Method 25 1-5 1.3.3 Comparability of Test Methods 1-6 1.4 Measurement Program Summary 1-7 1.4.1 Scrubber Inlet 1-7 1.4.2 Scrubber Outlet 1-7 1.4.3 Georgia-Pacific Scrubber System 1-8 1.4.4 Fugitive Emissions 1-8 1.4.5 Ambient Air Measurements 1-8 1.4.6 Clean-Up Evaluations 1-9 1.5 Report Sections 1-9 2.0 SUMMARY AND DISCUSSION OF RESULTS 2-1 2.1 Background and Definitions 2-1 2.1.1 Particulate Emissions 2-1 2.1.2 Condensible Emissions 2-2 2.1.3 Noncondensible Emissions 2-2 2.1.4 Total Organic Emissions 2-2 2.2 Method 5X - Particulate/Condensible Organics Emission Tests 2-3 2.2.1 Scrubber Inlet 2-6 2.2.2 Scrubber Outlet 2-8 2.3 Method 25 - Total Organic Tests 2-13 2.3.1 Scrubber Inlet 2-16 2.3.2 Scrubber Outlet 2-16 2.4 Visible Emissions 2-21 2.5 Scrubber Operational Summary 2-21 2.6 Summary of Fugitive Emissions 2-23 2.7 Ambient Air Measurements 2-23 2.8 Clean-Up Evaluation 2-27 2.9 Possible Test Interferences 2-27 3.0 PROCESS DESCRIPTION AND OPERATIONS 3-1 3.1 Process Equipment 3-1 3.2 Emission Control Equipment 3-1 3.3 Production and Control Equipment Monitoring 3-2 3.4 Process Operating Conditions During Test Program. . . 3-2 4.0 DESCRIPTION OF SAMPLING LOCATIONS 4-1 4.1 Scrubber Inlet 4-1 4.2 Scrubber Outlet 4-1 4.3 Scrubber Operational Measurement Locations 4-5 4.4 Fugitive Emissions 4-5 -111- ------- TABLE OF CONTENTS (Continued) SECTION PAGE 5.0 SAMPLING AND ANALYTICAL METHODS 5-1 5.1 EPA Reference Methods 5-1 5.2 Preliminary Measurements 5-2 5.3 Measurements for Particulate, Condensible and Noncondensible Emissions 5-2 5.3.1 EPA Method 5X (Provisional) - Particulate and Condensible Organic Compounds 5-2 5.3.2 EPA Reference Method 25 - Condensible and Noncondensible Organic Compounds 5-10 5.4 Preliminary Moisture Determination 5-22 5.5 Preliminary Velocity Determination 5-22 5.6 Visible Emissions 5-23 5.7 Pressure Drop Measurements 5-23 5.8 Scrubber Solution Samples 5-23 5.9 Fugitive Emissions 5-24 5.10 Ambient Temperature and Relative Humidity 5-25 6.0 QUALITY ASSURANCE 6-1 6.1 Method 5X 6-1 6.2 Method 25 6-3 6.3 Method 9 6-3 -IV- ------- LIST OF FIGURES SECTION PAGE 1-1 Veneer Dryer Exhaust and Scrubber System 1-4 4-1 Veneer Dryer Exhaust and Scrubber System Sampling Locations 4-2 4-2 Scrubber Inlet Sampling Port Configuration and Traverse Point Locations 4-3 4-3 Scrubber Outlet Sampling Port Configuration and Traverse Point Locations 4-4 5-1 Modified EPA Particulate and Condensible Organics Sampling Train 5-4 5-2 Method 25 Sampling Train 5-11 5-5 Method 25 Flow Control Assembly Adjustment 5-13 5-6 TRC Nonmethane Organic Analyzer 5-18 5-7 TRC Condensate Recovery and Conditioning Apparatus . . 5-21 LIST OF TABLES 2-la Summary of Method 5X Particulate and Condensible Organic Collection Efficiency for Georgia-Pacific Scrubber System (English Units) 2-4 2-lb Summary of Method 5X Particulate and Condensible Organic Collection Efficiency for Georgia-Pacific Scrubber System (Metric Units) 2-5 2-2 Summary of Method 5X Particulate and Condensible Organic Measurements for Gases Entering the Georgia-Pacific Scrubber System 2-7 2-3 Summary of Method 5X Particulate and Condensible Organic Measurements for Gases Exiting the Georgia-Pacific Scrubber System 2-9 2-4 Comparison of Particulate and Condensible Measured Emissions with Calculated Emissions 2-10 2-5a Summary of Method 25 Total Organic Collection Efficiency for Georgia-Pacific Scrubber System (English Units) 2-14 2-5b Summary of Method 25 Total Organic Collection Efficiency for Georgia-Pacific Scrubber System (Metric Units) 2-15 -v- ------- LIST OF TABLES (Continued) SECTION PAGE 2-6 Summary Method 25 Individual Total Organic Measurements for Gases Entering the Georgia-Pacific Scrubber System 2-17 2-7 Summary Method 25 Individual Total Organic Measurements for Gases Exiting the Georgia-Pacific Scrubber System 2-18 2-8 Summary of Method 25 Individual Total Organic Trap, Tank Measurements for Gases Entering the Georgia-Pacific Scrubber System 2-19 2-9 Summary of Method 25 Individual Total Organic Trap, Tank Measurements for Gases Exiting the Georgia-Pacific Scrubber System 2-20 2-10 Georgia-Pacific Scrubber System Operational Data Summary 2-22 2-lla Fugitive Emission Summary 2-24 2-llb Fugitive Emission Summary 2-25 2-llc Fugitive Emission Summary 2-26 2-12 Method 5X Clean-up Evaluation Results 2-28 -vi- ------- 1.0 INTRODUCTION 1.1 Background Section 111 of the Clean Air Act of 1970 charges the administrator of the United States Environmental Protection Agency with the responsibility of establishing Federal Standards of Performance for New Stationary Sources (SPNSS) that may significantly contribute to air pollution. When promulgated, these standards of performance for new stationary sources are to reflect the degree of emission limitation achievable through application of the best dem- onstrated emission control technology. Emission data collected from con- trolled sources in the plywood industry will provide a portion of the database used by EPA to develop SPNSS. EPA's Office of Air Quality Planning and Standards selected the Georgia- Pacific (G-P) plywood plant in Springfield, Oregon, as a site for an emission test program because it is considered to employ process and emission techno- logy representative of modern plywood manufacturing plants. The test program was designed to determine the emission rate of particu- late matter and condensible and noncondensible organic material emitted from the veneer drying operation. A second objective was to measure the collection efficiency of the Georgia-Pacific scrubber system for condensible and non- condensible organic emissions. TRC - Environmental Consultants, Inc. was retained by the EPA Emissions Measurement Branch (EMB) to perform emission measurements at the G-P plywood plant in Springfield, Oregon. Testing was performed during the week of June 8, 1981 on the veneer dryer emissions and their pollution control, a G-P scrubber system. This report has been prepared in accordance with EPA Con- tract No. 68-02-3543 under the provisions of Work Assignment No. 1. 1-1 ------- The Research Triangle Institute (RTI), the New Source Standard (NSS) con- tractor/ was responsible for coordinating the overall test program with G-p personnel and for assuring that process and control equipment operating con- ditions were suitable for testing. All process data were monitored and recorded by RTI. Fugitive emissions from the veneer dryers/ ambient air tem- perature and relative humidity were monitored and recorded by RTI. Additional testing for total organic compounds was performed by the National Council of the Paper Industry for Air and Stream Improvement, Inc, (NCASI) simultaneously with the TRC test program. This testing was performed at the request of the American Plywood Association (APA) for research purposes and to provide an additional measure of quality assurance. 1.2 Summary of Process and Emissions The G-P Springfield plant is a combination veneer and lay-up facility, and is considered to employ process and emission control technology representative of modern plywood manufacturing plants. The plywood production rate for the drying operation is approximately 800,000 square feet (3/8-inch basis) per 24-hour day. The veneer drying operation begins after the veneer has been peeled from the log at the lathe operation and is transferred to the drying operation. Here, the veneer is continuously hand-fed onto the dryer feed conveyor and into the dryer. The purpose of the operation is to thermally drive the mois- ture out of the veneer in preparation for the lay-up and laminating operations which follow. During the drying operaion, organic compounds are driven out of the veneer. These organic compounds are the emissions of interest. 1-2 ------- The G-P Springfield plant has four veneer dryers. Each is a steam heated, multideck unit, with the number of drying zones varying. Each dryer has two exhausts from the heated zones, except dryer 4, which has three exhausts. The nine exhausts are ducted to a common manifold which carries the dryer emis- sions to the G-P scrubber system. A schematic drawing of the veneer dryer exhaust system is presented in Figure 1-1. 1.3 Applicability of EPA Reference Test Methods EPA is required to publish a national reference test method for each reg- ulated source category and pollutant for which a New Source Performance Standard (NSPS) is established. Reference test methods are usually specified by a State regulatory agency during the State Implementation Planning process and may be different from national reference test methods. The purpose of establishing a national reference test method is to ensure that emission data collected from a specific source is representative of that source and comparable to data collected at other designated sources. The pri- mary purpose of this test program was to collect emission data using standard- ized test methods which allow the data to be evaluated to develop a national SPNSS. Two different test methods were selected by EPA to measure emissions from plywood veneer drying operations. These methods are briefly described in the following subsections and are described in detail in Section 5. 1.3.1 EPA Method 5X (Provisional) Provisional Method 5X is similar to the Oregon Department of Environmental Quality (ODEQ) Method 7 used to measure condensible organic emissions. EPA Method 5X measures particulate matter and condensible organic matter. "Par- ticulate matter" is defined as any finely divided solid or liquid material, 1-3 ------- Spray-JJ Nozzles Dryer 4 Dryer 2 Dryer 1 Figure 1-1. Veneer Dryer Exhaust and Scrubber System Georgia-Pacific Plywood Plant Springfield, Oregon 1-4 ------- other than uncombined water, that condenses at or above the filtration temper- ature range of 350 +25 F (177 +_14 C) , and is collected by the probe and filter (front half of the sampling train). "Condensible organic matter" is defined as any material remaining after extraction, filtration and ambient evaporation of the ether-chloroform extract of the impinger portion of the sampling train. Particulate matter and condensible organic matter are quanti- fied gravimetrically and results are expressed as the mass of collected mater- ial. The purpose of the 350 F filtration temperature is to precondition the Method 25 slipstream sample being withdrawn from the Method 5X sample stream. This temperature was selected on the basis of average veneer dryer operating temperatures throughout the industry. This temperature condition excludes from the Method 25 samples only matter than can condense at or above 350 F. It does not affect Method 5X results because the remaining sample is caught in the condenser portion of the train. 1.3.2 EPA Method 25 EPA Reference Method 25, as promulgated in the October 3, 1980 Federal Register (volume 145, no. 194, 65959 ff.), applies to the measurement of organic compounds as total gaseous nonmethane organics (TGNMO). Emissions are expressed as equivalent carbon (C,) mass. Method 25 sample fractions are separated by a gas chromatographic column, oxidized to carbon dioxide (C0_), and reduced to methane (CH.) prior to analysis by flame ionization detector (FID). Since all the sample organic compounds are reduced to CH , the pro- blems associated with the variable FID response characteristic for different organic compound structures is eliminated. This allows comparison of emission data on a uniform C basis. Method 25 is discussed in greater detail in Section 5 of this report. 1-5 ------- Major procedural modifications made to Method 25 were required to measure accurately emissions from plywood veneer drying facilities. These modifica- tions are discussed in Section 5. An additional condensate trap immersed in water ice was placed in the sampling train ahead of the standard dry ice im- mersed condensate trap. The purpose of the additional trap is to condense moisture that would freeze in the dry ice immersed trap and cause a premature sample flow stoppage. In this manner gas stream moisture content, which may range from 30 to 60 percent by volume, may effectively be reduced to 3 percent or less before entering the dry ice immersed trap. The use of the Method 5X sampling train as a sample preconditioner also represents a major modification. In addition to the 350 F sample stream temperature, isokinetic sample extraction from the source using Method 5X was also deemed necessary to obtain a representative Method 25 sample. This is particularly the case when moisture-saturated gas streams, such as those following wet scrubbing devices, are being sampled. Entrained water droplets may contain organic materials that would not be collected using the normal Method 25 constant sampling rate procedure. 1.3.3 Comparability of Test Methods Methods 5X and 25 are not related and measured results can not be compared under any circumstances. Condensation temperatures differ by more than 100 F between the two methods, and consequently different condensible com- pounds are collected by each method. In addition, it has been demonstrated that Method 5X has limited collection capabilities for organic compounds with high-vapor pressures. A loss of organic material is experienced even during normal Method 5X sample recovery and analysis operations. 1-6 ------- 1.4 Measurement Program Summary The measurement program was conducted at the G-P Springfield facility dur- ing the week of June 8, 1981. The emission tests were designed to measure the veneer dryer organic emissions and to determine the collection efficiency of the G-P scrubber system for those emissions. Tests were performed at the veneer dryer exhaust duct (scrubber inlet) and at the outlet of the scrubber system. All emission testing was performed by TRC and NCASI personnel. RTI per- sonnel monitored process operating conditions, while DGA personnel monitored fugitive emissions/ ambient temperature and relative humidity. Scrubber oper- ational data and solution samples were collected by TRC personnel. 1.4.1 Scrubber Inlet Preliminary Measurements Preliminary testing was performed on June 8 to determine volumetric flow rate and stack gas moisture content. Method 5X - Particulate and Condensible Organics Tests Three Method 5X tests were performed, one each on June 9, 10, and 11, con- currently with tests performed at the scrubber outlet. Method 25 - Total Organic Tests Eighteen Method 25 samples were taken at this location concurrently with the Method 5X tests performed. Six Method 25 samples were taken concur- rently with each Method 5X test. 1.4.2 Scrubber Outlet Preliminary Measurements Preliminary tests were performed on June 8 to determine volumetric flow rate and stack gas moisture content. Method 5X - Particulate and Condensible Organics Tests Three Method 5X tests were performed at this location, one each on June 9, 10, and 11 concurrently with tests performed at the scrubber inlet. 1-7 ------- Method 25 - Total Organic Tests Eighteen Method 25 samples were taken at this location concurrently with the Method 5X samples (six per test run), and simultaneously with Method 25 samples taken at the scrubber inlet. Method 9 - Visible Emissions Scrubber outlet visible emissions were not monitored as planned because of overcast sky background conditions. The scrubber outlet plume was bluish- white and was therefore indistinguishable from the overcast sky. Overcast skies were present on June 8, 9 and 10. During the last test day, June 11 / the sky began to clear and only scattered clouds were present in the afternoon. However, the final test sequence was nearly completed by this time. Consequently, no visible emission observations were recorded. Although the scrubber outlet stack had an attached steam plume, the Method 9 observations were not cancelled because of this condition. The method provides for attached steam plumes by requiring that observations be made at the point where the condensed water vapor is no longer visible. 1.4.3 Georgia-Pacific Scrubber System Static pressure upstream and downstream of the system induced draft fan was measured with U-tube water manometers and recorded at 30-minute inter- vals. These measurements were used to calculate pressure drop (AP) across the scrubber system. Scrubber solution samples were taken every 30 minutes during the scrubber outlet Method 5X test period. One-hundred-mi samples were collected from the scrubber recirculation tank every 30 minutes during each test. The individual samples collected during each test were composited for analysis. 1.4.4 Fugitive Emissions Fugitive emissions from the veneer dryers were monitored by DGA during each Method 5X test. 1.4.5 Ambient Air Measurements Ambient air temperature and relative humidity were monitored and recorded by DGA at the beginning and end of each Method 5X test. 1-8 ------- 1.4.6 Clean-Up Evaluations Prior to any emission testing, two Method 5X sampling trains were prepared and charged, ready to perform a test. The unexposed trains were then cleaned according to the method and samples recovered. The samples were analyzed to establish background and/or contamination levels from the .sample collection equipment. 1.5 Report Sections The remaining sections of this report present the Summary and Discussion of Results (Section 2), Process Description and Operations (Section 3), Description of Sampling Locations (Section 4), Sampling and Analytical Pro- cedures (Section 5), and Quality Assurance (Section 6). Descriptions of methods and procedures, field and laboratory data, and calculations are pre- sented in various appendices as noted in the Table of Contents. 1-9 ------- 2.0 SUMMARY AND DISCUSSION OF RESULTS A summary of all emission measurements and collected data is presented in this section. Section 2.1 provides a brief background discussion and defini- tions of measured parameters. Section 2.2 presents Method 5X particulate/ condensible organics results with a complete breakdown and discussion of par- ameters at both sampling sites. Method 25 total organic emission results are described in detail in Section 2.3, which includes a discussion of emissions at both sampling sites as well as a breakdown of major analytical data. Section 2.4 discusses visible emissions observations. A summary of scrubber operational data is presented in Section 2.5. Fugitive emissions are dis- cussed in Section 2.6. A summary of ambient air measurements is presented in Section 2.7. A full discussion of the Method 5X clean-up evaluation and results may be found in Section 2.8. 2.1 Background and Definitions The test program was designed to measure particulate matter/ condensible and noncondensible organic material emitted from veneer dryers, and to deter- mine the collection efficiency of the G-P scrubber system as a control for those emissions. 2.1.1 Particulate Emissions Particulate emissions are defined as any finely divided solid or liquid o matter, other than uncombined water, that condenses at or above 350 +25 F (177 _+14°C) and is collected in the probe and filter (front half) of the Method 5X sampling train. 2-1 ------- 2.1.2 Condensible Emissions Condensible emissions are defined differently in Methods 5X and 25. Although called by the same name, these two sample fractions differ signifi- cantly in content and composition and may not under any circumstances be com- pared. Method 5X condensibles are collected in glass impingers containing deion- ized distilled water and immersed in a water ice bath, and on a back-up filter following those impingers. Any material remaining after extraction, filtra- tion and ambient evaporation of the impinger solution, plus any material collected on the desiccated back-up filter, is defined as Condensible organic matter. Quantification of this matter is done gravimetrically. Method 25 condensibles are collected in two stainless-steel traps, one immersed in water ice followed by another packed in dry ice. Material collected in the traps is oxidized, reduced and analyzed by flame ionization. Results are expressed as a concentration of carbon (C,). 2.1.3 Noncondensible Emissions Noncondensible emissions are measured by Method 25 only and are those that pass through both ice traps to the evacuated sample tank at the end of the Method 25 train. These samples are oxidized, reduced and analyzed by FID. Results are expressed as concentrations of carbon (C,). 2.1.4 Total Organic Emissions Total organic emissions are those collected by the complete Method 25 sam- pling train drawing a preconditioned sample slipstream from a Method 5X train. These emissions include Condensible and noncondensible emissions as defined above. 2-2 ------- 2.2 Method 5X - Particulate/Condensible Organics Emission Tests A summary of Method 5X data collected at the scrubber inlet and outlet is presented in Tables 2-la (English units) and 2-lb (metric units) . These tables include relevant emission data: stack gas temperature, moisture con- tent and volumetric flow rate; veneer dryer production rate; and a summary of the total measured particulate/condensible emissions by concentration, mass emission rate, and emission rate per unit production. Emission data are presented for the three test series. Testing was per- formed concurrently at the scrubber inlet and outlet. Emissions at the scrub- ber inlet averaged 18.3 Ibs/hr (8.29 kg/hr) or 0.53 lbs/1000 ft2 veneer on a 3/8-inch basis (2.56 kg/1000 m2 on 9.5 mm basis) for the three tests. Emissions from the scrubber outlet averaged 14.9 Ibs/hr (6.77 kg/hr) or 0.43 lbs/1000 ft2 veneer (2.10 kg/1000 m2) for the three tests. The concentra- tions of the emissions from the two sources, however, differed markedly. The average scrubber inlet concentration was 0.164 gr/DSCF (0.376 g/NM3), while the scrubber outlet averaged only 0.103 gr/DSCF (0.236 g/NM3) for the three tests. The removal efficiency of the G-P scrubber system for particulate/ condensible organics averaged 16.4 percent for the three tests. Efficiencies ranged from 29 percent during test 1 to 6 percent during test 3. Detailed summaries of this test data are presented in Sections 2.2.1 and 2.2.2 and in Appendix A. Sample equations and calculations are presented in Appendix B. Field data sheets appear in Appendix C. Sampling logs and sum- maries are shown in Appendix D. Calibration data for the Method 5X sampling train are found in Appendix F. Laboratory analysis data are presented in Appendix G. 2-3 ------- TABLE 2-la (English Units) GUMMARY OF METHOD SX PARTICULATE AND CONUENS1BLE ORGANIC COLLECTION EFFICIENCY FOR UEORGIA-PACIFIC SCRUBBER SYSTEM Georgia-Pacific Plywood Plant, Springfield, Oregon Run Number Date Emission Description! Volume Gas Sampled (DSCF)a Stack Gas Flow Rate (DSCFM)b Stack Temperature (°F) Percent Moisture of Volume0 Percent isoklnesis M Scrubber Pressure Drop (inches lljO) "•*• Production Rate (1000 ft«/hr)d Particulate-Condensible Results Total Catch gr/DSCF Ibs/hour lbs/1000 ft1 Scrubber System Collection Efficiency Run 1 6-9-81 Uncontrolled Controlled 45.3 38.3 12400 15200 303 166 31.9 38.3 07.4)e 116.5 120.4 14.2 35.7 0.198 0.115 23. Of 16. 3£ 0.644 0.456 (») 29.1 Run 2 6-10-B1 Uncontrolled Controlled 42.1 37.3 12000 17500 314 166 34.0 37.5 (37.4)e 111.4 102.1 15.0 34.4 0.169 0.107 18. 5£ 15.9 0.538 0.462 14.1 Run 3 6-11-81 Uncontrolled Controlled 42.4 37.4 12600 16100 309 168 31.8 39.6 (39.3)e 99.2 111.3 13.9 34.1 0.124 O.OB6 13.4 12. 6£ 0.393 0.370 5.97 Average Uncontrolled Controlled 43.3 37.7 12300 16300 309 167 32.6 3U.5 (38.0)a 109.0 111.3 14.4 J4.7 U.164 0.10J 18.3 14.9 0.525 0.429 16.4 a Standard conditions! 29.92 inches Hg at 6B°F. b Outlet volumetric flows (DSCFM) suspected to be slightly biased high due to possible wind Interference. 0 Outlet moisture results calculated with psychometric equation. (See Section 5.) d 1000 square feet veneer per hour on 3/8 inch basis) includes trim factor; does not account for redry material. e Theoretical moisture content at saturation ( )) all results based on this. f Results are average of concentration and area ratio methods due to high isokinetic values. (Bee Table 2-4.) ------- TABLE 2-lb (Metric Units) SUMMARY OP METHOD 5X PARTICULATE AND CONDENSIBLE ORGANIC COLLECTION EFFICIENCY FOR GEORGIA-PACIFIC SCRUBBER SYSTEM Georgia-Pacific Plywood Plant, Springfield, Oregon ro I ui Run Number Run 1 Date 6-9-81 Emission Description! Uncontrolled Controlled Volume Gas Sampled (NM')a 1.28 1.08 Stack Gas Flow Rate (NM»/Min)b 351 430 Stack Temperature (°C) 151 74.4 Percent Moisture of Volume0 31.9 38.3 (37.4)e Percent Isokinesis 116.5 120.4 Scrubber Pressure Drop (mm HjO) 55.9 Production Rate (1000 m'/hr)d 3.32 Particulate-Condenslble Results Total Catch g/NM» 0.454 0.263 kg/hour 10. 4£ 7.39£ kg/1000 ft» 3.14 2.23 Scrubber System Collection Efficiency % 28.9 Run 2 Run 3 6-10-81 6-11-81 Average Uncontrolled Controlled Uncontrolled Controlled Uncontrolled controlled 1.19 1.06 1.20 1.06 1.23 1.07 340 495 357 456 348 461 157 74.4 154 75.6 154 75.0 34.0 37.5 31.8 39.6 32.6 38. 5 (37.4)e (39.3)e (3B.O)8 111.4 102.1 99.2 111.3 109.0 111.3 59.1 54.7 56.7 3.20 3.17 3.22 0.387 0.245 0.284 0.197 0.376 0.236 8.39£ 7.21 6.08 5.72£ 8.29 6.77 2.62 2.26 1.92 1.80 2.56 2.10 14.1 5.92 16.3 a Standard conditional 760 nun llg at 20°C. b Outlet volumetric flows (NM'/min) suspected to be slightly biased high due to possible wind interference. c Outlet moisture calculated with psychometric equation. (See Section 5.) d 1000 square meters veneer per hour on 9.5 millimeter basisi included trim factor) does not account for redry material. e Theoretical moisture content at saturation ( )j all results based on this. £ Results are average of concentration and area ratio methods due to high isoklnetic values. (See Table 2-4.) ------- 2.2.1 Scrubber Inlet A summary of Method 5X data collected at the scrubber inlet is presented in Table 2-2. Data presented include sample volume; stack gas flow rate, tem- perature, and moisture content; isokinesis for each test; veneer production rate; front half (particulate) and total (particulate/condensible) emissions. Tests 1, 2 and 3 were performed at the scrubber inlet on June 9, 10, and 11, respectively. Measured particulate emissions ranged from 0.77 to 2.93 Ibs/hr (0.02 to 0.08 lbs/1000 ft2 veneer), averaging 1.83 Ibs/hr (0.05 lbs/1000 ft2 veneer). Total particulate/condensible emissions ranged from 13.4 to 23.0 Ibs/hr (0.39 to 0.64 lbs/1000 ft2 veneer) for an average of 18.3 Ibs/hr (0.53 lbs/1000 ft* veneer). Particulate matter accounted for approximately 10 percent of the total sample weight while the remaining 90 percent of the catch was condensible organics. Measured particulate grain loadings averaged 0.016 gr/DSCF for tests 1, 2 and 3; ranging from 0.007 to 0.025 gr/DSCF. Total particulate/condensible grain loadings ranged from 0.124 to 0.198 gr/DSCF, for a three-test average of 0.164 gr/DSCF. The bulk of the total emission concentration was accounted for by condensible organics (90 percent). The average stack gas temperature was 309 F with an average moisture content of 32.6 percent. Moisture content varied from 31.8 percent to 34.0 percent over the three tests. The average stack gas flow rate was 12,300 DSCFM and did not vary significantly among the three tests. Isokinesis averaged 109 percent for the three tests performed. Isokinesis for test 1 was 116.5 percent due to a nomograph calculation error, while test 2 was high at 111.4 percent due to a higher than expected gas stream moisture content. Isokinesis was acceptable for test 3 at 99.2 percent. Leak checks were performed at the conclusion of each test and leak rates were acceptable at less than 0.02 cfm. 2-6 ------- TABLE 2-2 SUMMARY OF METHOD SX PARTICIPATE AND CONDENSIBLE ORGANIC MEASUREMENTS FOR GASES ENTERING THE GEORGIA-PACIFIC SCRUBBER SYSTEM Georgia-Pacific Plywood Plant, Springfield, Oregon to -4 Run Number Date Volume Gas Sampled (DSCF)a Stack Gas Flow Rate (DSCFM) Stack Temperature (°F) Percent Moisture by Volume Percent Isokinesis Production Rate (1000 ft'/hr)b Particulate-Condensible Results Front Half Catch (Probe and Filter) milligrams gr/DCSF 1 be/hour 1000 ft* Total Catch milligrams gr/DSCF Iba/hour lbs/1000 ft1 Percent Condensible Emissions Run 1 6-9-B1 45.3 12400 303 31.9 116.5 35.7 74.2 0.0253 2.93C 0.082 582. 3 0.198 23.0° 0.644 87.3 Run 2 6-10-81 42.1 12000 314 34.0 111.4 34.4 44.2 0.0162 1.67° 0.049 459.6 0.169 18.5° 0.538 90.4 Run 3 6-11-81 42.4 12600 309 31.8 99.2 34.1 19.7 0.0072 0.77 0.023 340.6 0.124 13.4 0.393 94.2 Averaqe 43.3 12300 309 32.6 109.0 34.7 46.0 0.0162 1.71 0.049 460.8 0.164 18.3 0.525 90.0 a Standard Conditionsi 29.92 inches llg at 68°F. b 1000 square feet veneer per hour on 3/8 inch baslsi Includes trim factor) does not account for redry material. c Results are average of concentration and area ratio methods due to high isokinetic values. (See Table 2-4.) ------- The mass emission rates for tests 1 and 2 were recalculated using the area ratio method because of the unacceptable anisokinetic conditions. The results are presented in Table 2-4 and are only slightly higher than those obtained from the concentration method, which is the normal approach. This result is probably due to the small percentage of particulate matter in the gas stream which would escape collection by the sampling nozzle under superisokinetic sampling conditions. An explanation of the area ratio method for calculating mass emission rates is presented in Section 5.3.1.4 of this report. Mass emission rates presented in Tables 2-1 and 2-2 represent the average of the two calculation methods for tests 1 and 2. 2.2.2 Scrubber Outlet A summary of Method 5X data collected at the scrubber outlet is presented in Table 2-3. Data presented include sample volume; stack gas flow rate, tem- perature, and moisture content; isokinesis for each test; veneer production rate; front half (particulate) and total (particulate/condensible) emissions. Three emission tests were performed at the scrubber outlet. Testing was performed concurrently with tests at the scrubber inlet on June 9, 10 and 11. Measured particulate emissions for tests 1, 2 and 3 ranged from 2.59 (0.08 lbs/100 ft2) to 3.70 Ibs/hr (0.11 lbs/1000 ftM, averaging 3.20 Ibs/hr (0.09 lbs/1000 ft2 veneer). Total measured particulate/condensible emissions ranged from 12.6 Ibs/hr (0.37 lbs/1000 ft2) for test 3 to 16.3 Ibs/hr (0.46 lbs/1000 ft2) for test 1. The average total emission rate was 14.9 Ibs/hr (0.43 lbs/1000 ft2 veneer). Particulate material collected during these three tests accounted for approximately 21 percent of the total emissions on the average, while the remaining 79 percent was condensible organics. 2-8 ------- to I TABLE 2-3 SUMMARY OF METHOD 5X PARTICULATE AND CONDENSIBLE ORGANIC MEASUREMENTS FOR GASES EXITING THE GEORGIA-PACIFIC SCRUBBER SYSTEM Georgia-Pacific Plywood Run Number Date Volume Gas Sampled (DSCP)a Stack Gas Flow Rate (DSCFH)b Stack Temperature (°F) Percent Moisture by Volume0 Percent Isoklnesia Production Rate (1000 ft»/ht)d Particulate-Condensible Results Front Half Catch (Probe and Filter) milligrams gr/DCSF Ibs/hour lbs/1000 ft1 Total Catch milligrams gr/DSCF Ibs/hour lbs/1000 ft1 Percent Condenslble Emissions Plant, Springfield, Oregon Run 1 Run 2 Run 3 6-9-81 6-10-81 6-11-81 38.3 15200 166 38.3 (37.4)e 120.4 35.7 56.9 0.023 3.30f 0.092 286.9 0.115 16.3* 0.457 80.2 37.3 17500 166 37.5 (37.4)a 102.1 34.4 41.9 0.017 2.59 0.075 257.6 0.107 15.9 0.462 83.7 37.4 16100 168 39.6 (39.3)e 111.3 34.1 61.6 0.025 3.70f 0.109 208.8 O.OB6 12. 6f 0.370 70.5 Average 37.7 16300 167 38.5 (38.0)e 111.3 34.7 53.5 0.022 3.20 .092 251.1 0.103 14.9 0.430 78.7 a Standard Conditionsi 29.92 inches Hg at 68°F. b Outlet volumetric flows (DSCFM) suspected to be slightly biased high due to possible wind interference. c Actual measured moisture adjusted to saturated conditions using psychometric equation. (See Section 5.) d 1000 ft' per veneer hour on 3/8 in. basis; includes trim factor; does not account tor redry material. 0 Theoretical moisture content at saturation ( ), all results based on this. f Results are average of concentration and area ratio methods due to high isokinetlc values. (See Table 2-4.) ------- TABLE 2-4 COMPARISON OF PARTICULATE AND CONDENSIBLE MEASURED EMISSIONS WITH CALCULATED EMISSIONS (Concentration Method vs. Area Ratio Method) Georgia-Pacific Plywood Plant, Springfield, Oregon Emission Rate (Ibs/hr) Sample No. Sample Fraction Concentration Method Area-Ratio Method Average 5X-l-Inlet (%I* = 116.5) 5X-2-Inlet (%I* = 111.4) 5X-l-Outlet (%I* = 120.4) 5X-3-Outlet (%I* = 111.3) Front half Back half Total Front half Back half Total Front half Back half Total Front half Back half Total 2.69 18.4 21.1 1.67 15.7 17.4 2.98 12.0 15.0 3.51 8.39 11.9 3.16 21.6 24.8 1.88 17.7 19.6 3.59 14.5 18.1 3.89 9.30 13.2 2.93 20.0 23.0 1.78 16.7 18.5 3.30 13.3 16.3 3.70 8.85 12.6 * isokinesis 2-10 ------- Particulate grain loadings measured at the scrubber outlet averaged 0.022 gr/DSCF for these tests, ranging from 0.017 gr/DSCP to 0.025 gr/DSCF. Total grain loadings (particulate/condensible) ranged from 0.086 to 0.115 gr/DSCF, averaging 0.103 gr/DSCF for the three tests. The average stack gas temperature measured during the three tests was 167 F. The measured moisture content of the gas stream averaged 38.5 per- cent for the three tests with slight variation. The moisture content of the gas stream as measured during each test exceeded saturation at the measured temperature. This phenomenon was not surprising since entrained water droplets were observed in the gas stream by TKC and EPA personnel. In accordance with EPA Method 4, the gas stream moisture content was recalculated assuming saturation of the gas stream at the average stack gas temperature for each test. An explanation of this procedure is presented in Section 5.3.1.4. Recalculated gas stream moisture contents ranged from 37.4 percent for tests 1 and 2 to 39.3 percent for test 3, averag- ing 38.0 percent. These saturation moisture values were carried through the remainder of the emission calculations. Measured stack gas flow rates ranged from 15,200 DSCFM to 17,500 DSCFM, averaging 16,300 DSCFM. The average outlet stack gas flow rate was measured to be approximately 25 percent greater than that measured entering the scrubber system. One reason for the difference in the flow rate measurements may be leaks within the system (around spray nozzles, in the cyclone ductwork, and before the fan) which would draw ambient air into the system. Another reason may be the inherent inaccuracy of EPA Method 2 at stack gas velocities of approximately 500 feet per minute (fpm) . A third possible reason is wind interference. Since sampling was performed only 9 feet from the top of a 9 foot i.d. stack, even a small gust of wind created turbulence within the 2-11 ------- duct. Momentary negative flow rates were indeed observed during the test pro- gram. These reasons may account for the differences between the measured inlet and outlet flow rates as well as the wide variation in measured flow rates from test to test. Isokinesis was acceptable only for test 2 at 102 percent. Isokinesis was unacceptable for tests 1 and 3 at 120.4 percent and 111.3 percent, respective- ly. The average isokinesis for the three tests performed was 111.3 percent. Varying stack gas moisture content was a major factor affecting the unaccept- able isokinetic conditions. The preliminary determination indicated 15.8 per- cent moisture/ but measured moistures were more than twice this during the subsequent tests. It was later discovered that different grades of veneer with varying moisture contents were dried throughout the test program, as shown in Table 3-1. The mass emission rates for tests 1 and 3 were recalculated using the area ratio method because of unacceptable superisokinetic conditions. The results are presented in Table 2-4 and are only slightly higher than those obtained from the concentration method, which is the normal approach. This fact is probably due to the small percentage of particulate matter in the gas stream which would escape collection by the sampling nozzle under anisokinetic sampl- ing conditions. An explanation of the area ratio method for calculating mass emission rates is presented in Section 5.3.1.4 of this report. Mass emission rates presented in Tables 2-1 and 2-3 are the average of the two calculation methods for tests 1 and 3. Leak checks were performed at the conclusion of each test and leak rates were acceptable at less than 0.02 cfm. Some diffi- culty was encountered maintaining probe and filter outlet temperatures at 350 o +25 F during these tests. Further discussion is presented in Section 5.3.1.1. 2-12 ------- 2.3 Method 25 - Total Organic Tests A summary of the Method 25 total organic data (condensible and nonconden- sible) collected at the scrubber inlet and outlet is presented in Tables 2-5a (English units), 2-5b (metric units). These tables include TRC, PCS, and NCASI average emission data: stack gas flow rate, moisture content and tem- perature; veneer drying production rate, and a summary of the total organic emissions by concentration, mass emission rate, and emission rate per unit production. All emissions are expressed as carbon (C,). NCASI calculates the emission rate as Ibs/hr equivalent methane (CHJ instead of carbon (C ). Their data in the tables have been converted to Ibs/hr C to present the data on a consistent basis, conforming with Method 25. Emission data are presented for the three test series. Testing was per- formed simultaneously at the scrubber inlet and outlet on June 9, 10 and 11. Total organic emissions entering the scrubber system ranged from 23.8 Ibs/hr (10.6 kg/hr) or 0.67 lbs/1000 ft2 veneer (3.20 kg/1000 m2) to 35.8 Ibs/hr (16.2 kg/hr) or 1.05 lbs/1000 ft2 (5.10 kg/1000 ra2), averaging 30.2 Ibs/hr (13.6 kg/hr) or 0.87 lbs/1000 ft2 veneer (4.22 kg/1000 m2). Emissions exiting the scrubber system ranged from 30.9 Ibs/hr (13.8 kg/hr) to 43.6 Ibs/hr (19.6 kg/hr) or 1.22 lbs/1000 ftz (5.87 kg/1000 m2) for an average emission rate of 38.8 Ibs/hr (17.5 kg/hr) or 1.12 lbs/1000 ft2 veneer (5.39 kg/1000 m2). The collection efficiency of the system was measured to be less than zero for tests 1 and 2 and 13.7 percent for test 3. The average collection efficiency of the scrubber system was less than zero. Detailed summaries of these test data are presented in Sections 2.3.1 and 2.3.2, and in Appendix A. Sample equations and calculations are presented in Appendix B. Field data sheets appear in Appendix C. Sampling logs and sum- maries are shown in Appendix D. Laboratory analysis data are presented in Appendix G. 2-13 ------- TABLE 2-5a (English Units) SUMMARY OF METHOD 25 TOTAL ORGANIC COLLECTION EFFICIENCY FOR GEORGIA-PACIFIC SCRUBBER SYSTEM Geocgia-Pacific Plywood Plant, Springfield, Oregon Run Number Date Emission Description! Stack Gas Flow Rate (DSCFM)3' b Stack Temperature (°F) Percent Moisture by Volume0 Scrubber Pressure Drop (inches H2O) Production Rate (1000 ft'/hr)a to 1 £ Total Organic Results8 parts/million, Cj gr/DSCF, Cj Ibs/hour, Cj lbs/1000 ft1 , Cj System Collection Efficiency (%) Run 1 Run 2 Run 3 6-9-81 6-10-81 6-11-81 Uncontrolled Controlled Uncontrolled Controlled Uncontrolled Controlled 12,400 15,200 12,000 17,500 12,600 16,100 303 166 314 166 309 168 31.9 37.4 34.0 37.4 31.8 39.3 14.2 15.0 13.9 35.7 34.4 34.1 1,027 1,536 1,389 1,286 1,524 1,030 0.22 0.33 0.30 0.28 0.33 0.22 23.6 43.6 31.1 42.0 35.8 30.9 0.67 1.22 0.90 1.22 1.05 0.91 <0 <0 13.7 Average Uncontrolled 12,300 304 32.6 14. 34. 1,313 0.21) 30.2 0.87 <0 Controlled 16,300 167 38.0 4 7 1,284 0.28 38.8 1.12 a Standard conditions! 29.92 inches Hg at 68°F. b Outlet volumetric flows (DSCFM) suspected to be slightly biased high due to possible wind interference. c Actual measured moisture adjusted to saturated conditions using psychometric equation (controlled emissions only). d 1000 square feet veneer per hour on 3/8 inch basis) includes trim factor) does not account for redry material. e Emission results calculated and reported as Cj. Does not include front half results from Method 5X collector, and cannot be compared with Method 5X mass determinations. ------- TABLE 2-5b (Metric Units) SUMMARY OF METHOD 25 TOTAL ORGANIC COLLECTION EFFICIENCY FOR GEORGIA-PACIFIC SCRUBBER SYSTEM Georgia-Pacific Plywood Plant, Springfield, Oregon Run Number Date Run 1 6-9-81 Run 2 6-10-81 Run 3 6-11-81 Average Emission Description: Stack Gas Flow Rate (NM'/Hln)3' b Stack Temperature (°C) Percent Moisture by Volume0 Scrubber Pressure Drop (nun IIjO) Production Rate (1000 meters'/hr)d Total Organic Results8 parts/million, C^ g/USCP, Cj kg/hour, Cj kg/1000 m> , Cj System Collection Efficiency (%) Uncontrolled Controlled Uncontrolled Controlled Uncontrolled Controlled Uncontrolled Controlled 351.2 150.6 31.9 430.2 74.4 37.4« 360.7 3.32 1,027 1,536 0.50 0.76 10.6 19.5 3.20 5.87 <0 339.8 156.7 34.0 495.6 74.4 37.4« 381.0 3.20 1,389 1,286 0.69 0.64 14.0 19.1 4.37 5.95 <0 356.8 153.9 31.8 456.0 75.6 39.3£ 353.1 3.17 1,524 1,030 0.76 0.50 16.2 14.0 5.10 4.42 13.6 348.3 461.6 154 74.8 32.6 JB.UC 365.8 3.22 1,313 0.62 13.6 4.22 1,284 0.63 17.5 5.41 <0 a Standard conditions: 760 mm Hg at 20%. b Outlet volumetric flows (NM'/min) suspected to be slightly biased high due to possible wind interference. 0 Actual measured moisture adjusted to saturated conditions using psychometric equation (controlled emissions only). d 1000 square meters veneer per hour on 9.5-mm basis) includes trim factor) does not account for redry material. 8 Emission results calculated and reported as Cj. Does not include front half results from Method 5X collector, and cannot be compared with Method 5X mass determinations. f Theoretical moisture content at saturation. ------- 2.3.1 Scrubber Inlet A summary of Method 25 condensible and noncondensible organics data collected at the scrubber inlet is presented in Tables 2-6 and 2-8. Table 2-6 shows relevant emission data and presents total organic emissions calculated by TRC, PCS, and NCASI as concentration, mass emission rate, and emission rate per unit production. Table 2-8 presents a breakdown of the total organic emissions into condensible and noncondensible organics as analyzed by the three laboratories. In addition, individual sample train analyses results are shown. The relative standard deviation between the paired sample trains is also presented. Emissions of carbon (C.) from the scrubber as analyzed by TRC, PCS and NCASI showed good overall correlation. The precision of the test data between paired samples (relative standard deviation-RSD) was excellent overall, averaging 19.8 percent RSD for the three laboratories involved. 2.3.2 Scrubber Outlet A summary of Method 25 condensible and noncondensible organics data collected at the scrubber outlet is presented in Tables 2-7 and 2-9. Table 2-7 shows relevant emission data and presents total organic emissions calcula- ted by TRC, PCS and NCASI as concentration, mass emission rate, and emission rate per unit production. Table 2-9 presents a breakdown of the total organic emissions into condensible and noncondensible organics as analyzed by the three laboratories. In addition, individual sample train analyses results are shown. The relative standard deviation between paired sample trains is also presented. Emissions of carbon (C-) from the scrubber as measured by TRC and NCASI showed good correlation. The average emissions calculated by TRC were slightly greater than those calculated by NCASI and PCS. There is no apparent 2-16 ------- TABLE 2-6 SUMMARY METHOD 25 INDIVIDUAL TOTAL ORGANIC MEASUREMENTS FOR GASES ENTERING THE GEORGIA-PACIFIC SCRUBBER SYSTEM Georgia-Pacific Plywood Plant Springfield, Oregon to I Run Number Date Stack Gas Flow Rate (DSCFM)a Stack Temperature (°F) Percent Moisture by Volume Production Rate (1000 ft»/hr)b Laboratory Performing Analysis Total Organic Results*- parts/million, Cj g/DSCF, Cj Ibs/hour, Cj lbs/1000 ft»r, Cj Run 1 6/9/81 12,400 303 31.9 35.7 TRC PCS NCASI 1016 655 1210 0.22 0.19 0.26 23.7 19. B 28.1 0.66 0.55 0.78 Run 2 6/10/81 12,000 314 34.0 34.4 TRC PCS NCASI 1482 1392 1295 0.32 0.30 0.28 33.3 31.2 29.1 0.97 0.91 0.84 Run 3 6/11/81 12,600 309 31.8 34.1 TRC PCS NCASI 2137 1101 1334 0.47 0.24 0.29 50.4 25.9 31.4 1.48 0.76 0.92 Average 12,300 J09 32.6 34.7 TRC PCS 1545 1116 0.34 0.24 35.8 25.6 1.03 0.74 NCASI 1280 0.28 29.5 0.85 a Standard Conditions: 29.92 inches llg at 68°F, NCASI uses O°C. b 1000 square feet veneer per hour on 3/B inch basis; includes trim factor) does not account for redry material. c Emission reaults calculated and reported as Cj . Does not include front half results from Method 5X collection, cannot be compared with Method 5X mass determination. ------- N) I TABLE 2-7 SUMMARY OF METHOD 25 INDIVIDUAL TOTAL ORGANIC MEASUREMENTS FOR GASES EXITING THE GEORGIA-PACIFIC SCRUBBER SYSTEM Run Number Date Stack Gas Flow Rate (DSCFM)a> b Stack Temperature °F Percent Moisture by Volume0 Production Rate (1000 ft«/hr)d Laboratory Performing Analysis Total Organic Results0 parts/million, Cj gr/DSCF, G! Ibs/hour, Cj lbs/1000 ft« , Cj Georgia-Pacific Run 1 10/9/81 15,200 166 37.4 35.7 TRC PCS NCASI 1762 1538 1306 0.38 0.33 0.28 50.1 43.7 37.2 1.40 1.22 1.04 Plywood Plant, Springfield, Oregon Run 2 Run 3 10/10/81 10/11/81 17,500 16,100 166 168 37.4 39.3 34.4 34.1 TRC PCS NCASI TRC PCS NCASI 1694 941.6 1221 1264 903 923 0.37 0.20 0.27 0.28 0.20 0.20 55.5 30.8 40.0 38.1 27.2 27.8 1.61 0.90 1.16 1.12 0.80 0.82 Average 16,300 167 38.0 34.7 TRC PCS NCASI 1573 1127 0.34 0.25 48.0 34.2 1.38 0.99 1150 0.25 35.1 1.01 a Standard Conditional 29.92 Inches Hg at 6B°F, NCASI uses O°C. b Outlet flows suspected to be slightly biased high due to wind Interference. c Theoretical moisture content at saturation. (See Section 5.) d 1000 square feet veneer per hour on 3/8 inch basis; Includes trim factort does not account for redry material 8 Emission results calculated and reported as Cj. Does not include front half results from Method 5X collector, and cannot be compared with Method 5X mass determinations. ------- TABLE 2-8 SUMMARY OF METHOD 25 INDIVIDUAL TOTAL ORGANIC TRAP, TANK MEASUREMENTS FOR GASES ENTERING THE GEORGIA-PACIFIC SCRUBBER SYSTEM Georgia-Pacific Plywood Plant, Springfield, Oregon Condensible Catch Run 1 June 9 Run 1 June 10 Run 3 June 11 Overall Average Stack Gas Flow Rate Sample DSCFM3 I.D. No. 12,400 1-A 1-B 1-C 1-D 9-1 9-2 Average 12,000 2-A 2-B 2-C 2-D 10-1 10-2 Average 12,600 3-A 3-B 3-C 3-D 11-1 11-2 Average 12,300 Lab TRC TRC PCS PCS NCASI NCASI TRC TRC PCS PCS NCASI NCASI TRC TRC PCS PCS NCASI NCASI HjO Ice Trap (ppm) 477 245 509.7 720.1 N.A.b N.A.b 488 483 489 539 549.7 N.A.b N.A.b 515 545 940 469.0 576.7 N.A.b N.A.b 639 547 C02 Ice Trap (ppm) 300 601 267.7 164.1 1231.4 1045.6 601.6 163 212 352.2 367.9 1531.1 863.2 5B2 469 433 550. 8 315.9 1143.1 1155.2 678 621 NonCondensible Catch (ppm) 409 n.d. 25.0 22.5 60.9 82.3 100 804 814 433 541.8 63.3 132.0 465 1102 786 159.2 102.6 161.1 209.4 420 328 Total Catch (ppm) 1186 846 802.4 906.7 1292 1128 1027 1450 1515 1324 1459 1594 995 1389 2116 2159 1206 995.2 1304 1365 1524 1314 Pair Average (ppra) 1016 855 1210 1027 1482 1392 1295 1389 2137 1101 1334 1524 1314 Emission Rate (Ibs/hr) 23.7 19.8 28.1 23.8 33.3 31.2 29.1 31.1 50.4 25.9 31.4 35.8 30.2 Relation Standard Deviation 4.22 11.6 4.0 6.6 32.2 14.6 3.06 16.6 70.3 7.4 31.2 36.3 19.8 a Standard Conditionsi 29.92 inches Hg at 60°F; NCASI uses 0°C. b NCASI does not use an 1)2° ice trap in their train. ------- TABLE 2-9 SUMMARY OF METHOD 25 INDIVIDUAL TOTAL ORGANIC TRAP, TANK MEASUREMENTS FOR GASES EXITING THE GEORGIA-PACIFIC SCRUBBER SYSTEM NJ I N) o Georgia-Pacific Plywood Plant, Springfield, Oregon Condensible Catch Stack Gas Flow Rate Sample DSCFM3 I.D. No. Run 1 June 9 15,200 Run 2 17,500 June 10 Run 3 16,000 June 11 / Overall Average 16,300 1-A 1-B 1-C 1-D 9-1 9-2 Average 2-A 2-B 2-C 2-D 10-1 10-2 Average 3-C 3-D 3-A 3-B 11-1 11-2 Average 11 20 Ice Trap Lab (ppm) TRC TRC PCS PCS NCASI NCASI TRC TRC PCS PCS NCASI NCASI TRC TRC PCS PCS NCASI NCASI 815 758 1180.2 905.0 N/Ab N/Ab 915 465 792 579.9 571.0 N/Ab N/Ab 606 831 441 534.6 414.3 N/Ab N/Ab 555 692 C02 Total Ice Trap Non-Condensible Catch Catch (ppm) (ppm) (ppm) 278 740 405.6 116.6 1242.9 1276.3 676.6 332 303 117.5 102.4 1270.2 1039.9 527 196 193 176.8 641.1 949.4 743.4 483 562 546 388 323.3 145.5 49.2 44.9 249 687 810 316.8 177.5 23.6 108.5 354 443 424 0 39.0 68.6 85.2 177 260 1639 1886 1909 1167 1292 1321 1536 1484 1905 1032 . 850.9 1294 1148 1286 1470 1058 711.4 1094 1018 828.6 1030 1284 Pair Average (ppm) 1762 1538 1307 1536 1694 941.6 1221 1286 1264 903 923 1030 1284 Emission Rate (Ibs/hr) 50.1 43.7 37.2 43.6 55.5 30.8 40.0 42.0 38.1 27.2 27.8 30.9 38.8 Relation Standard Deviation 10 2.9 63 25.3 5.7 7.3 11. a 8.3 4.3 3.3 b.9 4.8 12.8 a Standard condltionsi 29.92 inches llg at 68°F| NCASI uses 0°C. b NCASI does not use an II20 ice trap in their train. ------- explanation for this difference. The precision of the test data between paired samples (relative standard deviation) was excellent overall, averaging 12.8 percent RSD for the three laboratories involved. 2.4 Visible Emissions Visible emissions observations were not conducted during this sampling program as planned. Overcast skies prevented an accurate determination of the scrubber outlet plume opacity. Further details of the decision to abort this phase of the test program are presented in Sections 1.4 and 5.7. 2.5 Scrubber Operational Summary A summary of operational parameters of the G-P scrubber system during the test program is presented in Table 2-10. Pressure drop measurements (AP) across the scrubber system are presented as well as scrubber solution analysis data. Scrubber solution samples were taken from the recirculating tank every 30 minutes and then composited into one sample per test. Sample analyses for total organic carbon (TOC) ranged from 3,010 mg/1 for test 1 to 2,860 mg/1 for test 3, averaging 2,956 mg/1 for the three tests. Pressure drop (AP) measurements across the scrubber were made at 30 minute intervals during the test program. The AP gradually increased during tests 1 and 2, and averaged 14.2 and 15.0 inches water, respectively. During test 3, however, the AP peaked shortly after the start of testing and then gradually declined for the rest of the test, averaging 13.9 inches water. The three test average AP was 14.4 inches water. 2-21 ------- TABLE 2-10 GEORGIA-PACIFIC SCRUBBER SYSTEM OPERATIONAL DATA SUMMARY Georgia-Pacific Plywood Plant, Springfield, Oregon June 1981 Pressure Drop Measurements Run Number Date Time 1 June 9 1400 1430 1500 1530 1600 1630 1700 1730 Average 2 June 10 1300 1400 1430 1500 1530 1600 1630 Average 3 June 11 1145 1215 1245 1315 1345 1415 1445 Average AP (in. H90) 12.5 13.0 14.0 14.8 14.5 14.7 15.1 15.2 14.2 14.7 14.8 15.0 14.8 15.2 15.4 15.3 15.0 13.8 - 14.5 14.0 13.9 13.8 14.2 12.9 13.9 Scrubber Solution Volume Collected TOC (ml) (mg/1) 800 3010 . 700 3000 700 2860 2-22 ------- 2.6 Summary of Fugitive Emissions (Provided by RTI) The temperature and pressure changes that a veneer dryer is subjected to make it very difficult for a dryer to be completely air tight. Door seals and skins, green and dry end baffles, and abort stacks will with time all even- tually develop leaks. Door seals and dryer skins most readily develop them. At Springfield all the dryers had fugitive emissions. The three older dryers leaked more from around the elephant ears than from individual door seals. The jet dryer also experienced door leaks with quality checks showing varying amounts from one day to another. It was impossible to estimate volume of fug- itive gases from any fugitive source. At the abort stacks there were also fugitive emissions. Very little was seen from the jet dryer abort stacks, while opacities (unofficially) up to 30 percent were seen from the other three dryer abort stacks. Cooling section air volumes are large. No bluish haze was seen coming from cooling section exhausts from the three longitudinal dryers. However, the jet dryer cooling stacks showed some bluish opacity. All of the fugitive emissions evaluations were purely qualitative and visual. Tables 2-lla, 2-llb and 2-llc contain fugitive emission data. 2.7 Ambient Air Measurements A summary of ambient temperature and relative humidity measurements by RTI and DGA is presented along with process information in Table 3-1. Ambient temperatures ranged from 54 to 75 F, while relative humidity ranged from 36 percent to 76 percent during the test program. 2-23 ------- TABLE 2-1la FUGITIVE EMISSION SUMMARY Georgia-Pacific Plywood Plant, Springfield, Oregon (Provided by RTI) June 9, 1981 to I CO Abort Dryer I/Time Green End 1 2ilO 2:45 3>40 4:50 Si 30 2 It 50 2i45 3:45 4:50 5:30 3 2:00 2:45 3:45 4:45 5:30 4 1:45 2:45 ' 3:40 4:40 5:27 Building Pans H 11 H H H II II H N N N N N 1 L 25% 30% M H M II II L 10% 10% 30% 30% 10% 20% N N N N N 2 Stacks Dry End L M 30% II 20% II 30% H 30% L L M M 10% M 10% II 20% M 30% M 30% L H 30% N N N N N 3 Cooling N N H N N N N N H N N N H N N 5% 10% 10% 10% 15% 4 Overall1 L H M H II M M M M M H II H M 11 5 100 200 200 250 300 150 200 200 150 150 250 250 250 ISO 300 L - L - L - VL L 6 cfm cfm cfm cfm cfm cfm cfm cfm cfm cfm cfm cfm cfm cfm cfm (Opacity) 2:00 5% 5% - 5% - not Time 2:10 3:25 4:15 5:05 5:40 2:05 3:25 4:15 5:05 5:40 2:05 3:15 4:13 5:00 5:40 2:00 3:18 4:12 5:00 5:40 N » T • VL • L • observed M - 2:45 3:45 4:45 5:30 5% 5% 5% 5t 10% 10% 5% 5% - - - - 5% 5% 5% 10% - - - 5% « • •• M II » VII " Door Seals 2M 2L, 2T 4M 1L 3L 1L H N N N N N N N N 1M, 1L 1L 1H, 1L 111, 1L 2M None Trace Inside Elephant Above Ears Dryers 1M 1M, 1L 1M, 211 4M, III 2M 1L N 1M 1L 1L N N N N N NA NA NA NA NA KEY M M H II M L L M M H H H 11 H H L L L L M 1 cfm estimates are gross Very Light estimates Light Moderate % opacity Heavy on blue Very Heavy plumes ------- TABLE 2-llb FUGITIVE EMISSION SUMMARY Georgia-Pacific Plywood Plant, Springfield, Oregon (Provided by RTI) June 10, 1981) NJ Ul Inside Appearance Dryer 1 2 3 'i 4 I/Time 12:20 1:15 2:20 2:45 3:45 4:15 4:45 12:17 1:15 2:15 2:50 3:45 4:15 4:45 12:15 1:15 2:15 2:50 3:50 4:15 4:40 12:13 1:12 2:15 2:48 3:50 4:15 4:40 Roof Vents 12:20 1:15 2:15 2:45 3:45 4:15 4:40 Abort Green End S 10% 11 H M 10 M 15 L 10 L 10 11 II H L 10% L L 5t L 10 M 20t L M 10% 11 20% M 10 H 20 II 20 N N H N N N N 907 908 5 5 5 5 5 5 5 5 5 5 5 5 Stacks Dry End M 20 M II 20 H 20 11 30 H 30 M 20 M S M 101 M 10% M 10 M 10 M 10 M 20% L M 10% M 20 11 20 II 20 H 20 N N N N N N N 909 N - - - - - - Door Elephant Cooling N N N N N N N N N N N N N N N N N N N N N 10-15 5 10% 10-15 N 5-10 5 910 5 - 5 5 5 5 - Overall1 M 150 cfm M 150 cfm H 250 cfm M 150 cfm M 200 Cfm 11 250 cfm M 200 cfm M 300 cfm M 150 cfm M 200 cfm L 150 cfm L 150 cfm L 100 cfm L 150 cfm M 200 cfm 50 cfm M 100 cfm II 250 cfm H 250 cfm 11 250 cfm II 250 cfm Had at Dry End H H II N It H 911 912 5 5 5 - 5 - - 5 Time 12:10 1:25 2:07 2:40 3:40 4:05 4:38 12:08 1:25 2:06 2:40 3:40 4:05 4:35 12:05 1:20 2:05 2:40 3:35 4:05 4:33 12:03 1:20 2:03 2:38 3:35 4:00 4:33 N - T « VL - L • M * II " VII " Seals 1 VL 2 VL 2 VL 1 S, 1 L 2 L 2 L 2 L 2 M 2 VL N N N N N N N N N N N N 1 L, 1 M 2 M 1 L 1 L 1 L 1 M 1 M, TR KEY None Trace Very Light Light Moderate Heavy Very Heavy Ears Top N L N L N L 1 L M 1 L M 1 L L 1 L M IS L IS M 1 M M 1 L II 1 L M 1 L L 1 L L N M N M N II N II N H N M N H NA M NA M NA L NA L NA L NA M NA L cfm estimates are gross estimates t opacity on blue plumes Comments M S mid fans Mod. around mid fans 11 at dry end None at Green Large leakage around exhaust dry end stk. Cooling stks. from this dryer and roof vents are the major roof-ton emitters. ------- TABLE 2-llc FUGITIVE EMISSION SUMMARY Georgia-Pacific Plywood Plant, Springfield, Oregon (Provided by RTI) June 11, 1981) to I cn Abort Stacks Dryer/Time 1 2 3 4 10:50 Ilt46 ill 20 12:55 1(06 2:28 11:10 11:45 12:20 12:55 1:46 2:28 11:10 11:45 12120 12:55 1:48 2:27 11:10 11:45 12:17 12t55 1:48 2:26 Roof Vents 11:10 11:45 12:19 12:55 1:48 2:27 Green End II II M 5 M 5 M 5 M 10 II H II II II M 10 H 30 II H II 10 II 10 M 20 N N N N N N 907 908 5 5 5 10 5 5 10 10 5 Dry End H II 20 H 10 H 10 H 30 H 20 M H 10 M 10 L M 10 M 10 1! 20 H 20 H 20 H 20 II 20 II 20 N N N N N N 909 - - 5 - - - Cool ing N N N N N N N N H N N N N N N N N N lot 5-10 0-5 5-10 5 5-10 910 5 - - 5 5 5 Ov«rall' 300 H 300 H 250 H 250 II 300 II 250 H 350 H 300 H 250 11 150 M 300 H 200 M 350 H 350 11 350 11 350 11 300 H 250 H 911 912 5 5 - - - 5 Time 10:42 11:32 12:08 12:45 1:26 2:17 2:47 10:39 11:30 12:07 12:45 1:26 2:16 2:47 10>38 11:30 12:06 12:43 1:27 2:15 2:48 10:37 11:28 12:06 12:43 1:27 2:14 2:47 N T VL L H H VH Inside Appearance Door Elephant Seals 3 L 3 L 2 L 2 VL 2 VL 2 L 1 L N N N N N N N N N N N N N N 2 VL, 3 T 2 VL N N N N N KEX • None » Trace « Very Light » Light » Moderate - Heavy » Very Heavy Ears 1 L 1 L 1 M 1 M 1 L 1 L 1 L 1 L 1 L 1 L 1 L 1 L 1 L 1 L N N N N N N N NA NA NA NA NA NA NA Cfm Top L M M L M 11 M L L M L L M M H H 11 H M H L L VL VL VL VL VL estimates Comments M e top center Cans M 9 center fans at top Stack leaks k middle fans hlgn emissions are gross estimates % » opacity on blue plumes ------- 2.8 Clean-Dp Evaluation Results of the clean-up evaluations performed on both Method 5X sampling trains are presented in Table 2-12. Clean-up evaluation rationale and pro- cedures are presented in Section 1.4.6 and Section 6.1. Front half total residue collected was 55.5 mg and 12.1 rag for the inlet and outlet sampling trains, respectively. Back half total residue collected was 39.8 mg and 164.2 mg, respectively. Total residue collected during the clean-up evaluation was 96.3 mg and 176.3 mg for the inlet and outlet trains, respectively. The high blank value of the inlet probe wash is probably due to the fact that the probe was not acid washed before the evaluation. The high value detected as impinger residue was probably due to a large amount of chromium residue (2.50 mg/1) remaining from the pretest chromic acid wash of the glass- ware. Further analysis for chromium was performed on the actual test impinger solutions with only a trace amount being detected in test 1 samples. There- fore, it is believed no test sample interference resulted from chromium con- tamination. Chromium analytical data are presented in Appendix G. 2.9 Possible Test Interferences A possibility exists that components and reagents used in the Method 5X sampling train may cause interferences with the Method 25 samples drawn from the Method 5X train. At the time of this report, a study is being performed by TRC to quantify the possible interfering effects of acetone, silicone vacuum grease, and silicone rubber sealant (RTV) on the Method 25 procedures. The scrubber outlet Method 5X filtration temperature could not be main- o tamed in the planned 350 H^25 F range due to insufficient heater capacity. Temperatures ranged from 310° to 340°F. Although not a factor in the 2-27 ------- TABLE 2-12 METHOD 5X CLEAN-UP EVALUATION RESULTS, JUNE 8, 1981 Georgia-Pacific Plywood Plant, Springfield, Oregon Train Component Sample Fraction Residue Weight (g) Inlet Outlet Front Half Probe Wash (DD H20) Probe Wash (acetone) Front Filter Front Half Total 0.0337* 0.0448 NAt 0.0785 0.0116 0.0113 NAt 0.0229 Back Half Impinger Water Organic Extraction Evaporation Acetone Rinse Back-up Filter Back Half Total Total Sample 0.0014 0.0230 0.0151 NAt 0.0395 0.1180 0.0019 0.1360** 0.0260 NAt 0.1639 0.1868 * Probe not acid washed prior to test program. ** Upon further analysis, it was discovered that Sample 5X-0-0-4 contained 2.50 mg/liter chromium, indicating residue remaining from the chromic acid cleaning solution used in the pretest preparation of the glassware. t Filters not inserted into trains. 2-28 ------- Method 5X sample collection, this may have resulted in a slight low bias for the Method 25 samples. Organic materials which would have passed through the o filter at 350 +25 F may instead have condensed and been collected on the lower temperature filter. 2-29 ------- 3.0 PROCESS DESCRIPTION AND OPERATIONS (Provided by RTI) This section describes the plywood manufacturing process, specifically the veneer drying process and its emission control/ a G-P scrubber system. Pro- duction monitoring as well as process operational conditions during the test program are also discussed. 3.1 Process Equipment The veneer drying operation begins after the veneer has been peeled from the log at the lathe operation. The veneer then proceeds to the drying opera- tion. Here, the veneer is continuously hand-fed onto the dryer feed conveyor and into the dryer. The purpose of the operation is to thermally drive the moisture out of the veneer in preparation for the layup and laminating opera- tions which follow. During the drying operation, organic compounds are also driven out of the veneer. The G-P Springfield plant has four veneer dryers. Each is a steam heated, multideck unit, with the number of drying zones varying between dryers. Dryers 1, 2 and 3 are longitudinal dryers, with 22, 18 and 18 zones, respect- ively. Dryer 4, a new unit, is a 22 zone jet dryer. Each dryer has two exhausts from the heated zones, except dryer 4 which has three exhausts. Atop each exhaust is an abort damper for emergency use only. These are a source of fugitive emissions. The eight exhausts are ducted to a common manifold which carries the exhaust to the Georgia-Pacific scrubber system. 3.2 Emission Control Equipment The Georgia-Pacific scrubber system shown in Figure 1-1 includes a wet spray zone, six wet cyclones, a packed tower, and a mesh pad entrainment separator. As dryer exhaust gases pass through the 35.25-inch inside diameter 3-1 ------- duct, six nozzles inject water countercurrently into the gas stream to saturate and cool the gas stream, thereby condensing the organics. Solids are separated and agglomerated droplets are collected in the six wet cyclones which follow. The remaining moisture laden gases are drawn through an induced draft fan and forced through a packed tower and a mesh pad to rid the effluent of aerosols. All water within the scrubber system is recirculated. The gas stream, at approximately 165 F and 35 percent moisture by volume, then discharges to the atmosphere through a 9-foot i.d. stack. 3.3 Production and Control Equipment Monitoring All production monitoring data collected by RTI and DGA is presented in Table 3-1. Scrubber operational data, collected by TRC, is presented in Table 2-10. 3.4 Process Operating Conditions During Test Program The operation of each dryer is set according to the size, thickness, and kind of wood being dried. The operation of the three longitudinal dryers does not frequently vary, but it appears that the jet dryer makes more frequent changes. During the testing hours of the first day (June 9) there was a change of production, despite efforts on Georgia-Pacific's part to keep the dryer operation steady state. Dryer 4 changed from drying 1/6-inch sapwood to 1/6-inch heartwood. This was not considered to be a major change worthy of cancelling the test run. It is normal for small plugups in the feeding and outloading mechanisms to occur and this did happen during the tests. On the third test day (June 11) there was more of this type upset than usual on Dryer 4, but these were con- sidered minor and insufficient cause for cancelling a test. 3-2 ------- TABLE 3-1 SUMMARY OF OPERATING CONDITIONS (Provided by RTI) Georgia-Pacific Plywood Plant, Springfield, Oregon June 8 June 9 June 10 June 11 I Production (ft2, per hour on 3/8-in. basis) S apwood Heartwood Total 19,935 12,090 31,424 21,604 14,143 35,747 21,509 12,863 34,372 19,847 14,221 34,068 II Redry Rate (%) 9.7 9.0 9.6 11.4 III Steam Use (Ibs per hour) No Evaluation 51,430 50,467 51,450 IV Temperatures No Evaluation steady 325-375°F steady 325-375°F steady 325-375°F V Fugitives 1. abort stacks 2. door leaks 3. above dryers 4. cooling stacks No Evaluation 150-300 CFM* nos. 1&4 nos. 1 & 3 had blue haze 150-300 CFM* 150-300 CFM* nos. 1,2 & 4 less small leaks than on June 9 noticeable above all dryers VI Weather No cloudy, Evaluation showers, mid 60s 53-74% rel. humidity cloudy, small showers, 60°-75° 36-70% rel. humidity morning fog, sunny , 54°-66° 51-76% rel. humidity *Per dryer, except no. 4 3-3 ------- Steam usage, dryer temperatures, and drying times were maintained evenly throughout the three days of tests. Process operating conditions are summarized in Table 3-1. Production figures provided are not the actual square footage of green veneer dried in the steam-heated dryers but rather a figure that accounts for trim and shrinkage. A full green veneer sheet is approximately 54 inches by 101 inches and will eventually be trimmed to 48 inches by 96 inches following shrinkage in the dryer. The amount of shrinkage depends on the original moisture level. As is the case with all western softwoods, Douglas fir sapwood will shrink more than heartwood. An expected shrinkage loss is 5 percent to 7 percent. The production figures reported are, therefore, approx- imately 85 percent of the actual throughput of the dryers. All veneer has been converted to a 3/8-inch basis. 3-4 ------- 4.0 DESCRIPTION OF THE SAMPLING LOCATIONS This section presents a description of each sampling location and a sum- mary of the work performed at each site. Figure 4-1 presents a schematic layout of the veneer dryer exhaust system and identifies all sampling loca- tions. 4.1 Scrubber Inlet The inlet to the scrubber system was sampled employing EPA Methods 1, 2, 4, 5X and 25 in the 35.25-inch inside diameter insulated duct at sampling ports 45 above the horizontal duct axis. These ports were located 30 feet downstream (>8 diameters) and 24 feet upstream (>2 diameters) from the nearest respective flow disturbances. In accordance with EPA Method 1, sam- pling was performed at 12 traverse points. Sampling port and traverse point locations are presented in Figure 4-2. Duct static pressure was also measured at this location. Method 5X tests performed at this location were 60 minutes in duration as were the Method 25 tests performed simultaneously. A total of three Method 5X and 18 Method 25 tests was performed at this location. 4.2 Scrubber Outlet Sampling ports were located 9 feet upstream (1 diameter) from the top of the 9-foot i.d. stack and approximately 6 feet (2/3 diameter) downstream from the mesh pad entrainment separator. In accordance with EPA Method 1, sampling was performed at 48 traverse points. Sampling port and traverse point locations are presented in Figure 4-3. 4-1 ------- Scrubber Outlet _ Sampling Location Scrubber Solution Sampling Location Spray- -J Nozzles Inlet Location PRESSURE MEASUREMENT LOCATIONS Figure 4-1. Veneer Dryer Exhaust and Scrubber System Sampling Locations Georgia-Pacific Plywood Plant Springfield, Oregon 4-2 ------- 351/4"I.D. To Scrubber Traverse Point Number 1 2 3 4 5 6 Traverse Point Location From Inside Duct Wall (Inches) 1.6 5.1 10.4 24.8 30.1 33.7 Figure 4-2. Scrubber Inlet Sampling Port Configuration and Traverse Point Locations, Georgia-Pacific Plywood Plant Springfield, Oregon 4-3 ------- 108" From Scrubber T 108" 108"I.D. Traverse Point Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Traverse Point Location From Inside Duct Wall 1.2 3.5 5.9 8.5 11.3 14.3 17.4 21.0 24.8 29.4 34.9 43.0 65.0 73.1 78.6 83.2 87.0 90.6 93.7 96.7 99.5 102.1 104.5 106.8 Figure 4-3. Scrubber Outlet Sampling Port Configuration and Traverse Point Locations Georgia-Pacific Plywood Plant, Springfield Oregon 4-4 ------- Method 5X tests performed at this location were 96 minutes in duration. Method 25 sampling began 18 minutes into each Method 5X test and ran for 60 minutes, leaving 18 minutes of Method 5X testing after the completion of the Method 25 tests. A total of three Method 5X and 18 Method 25 tests were per- formed at this location. 4.3 Scrubber Operational Measurement Locations Scrubber solution samples were taken from the scrubber recirculation tank during each test. Pressure drop across the scrubber system was calculated from static pressures in the duct measured at the scrubber inlet and at two pressure taps each located 2 feet from the scrubber fan inlet and outlet. These sampling locations are shown in Figure 4-1. 4.4 Fugitive Emissions Fugitive emissions were observed by RTI and DGA around the veneer dryers and their abort dampers. 4-5 ------- 5.0 SAMPLING AND ANALYTICAL METHODS This section presents descriptions of sampling and analysis procedures employed during the emission testing conducted at the Georgia-Pacific plywood facility in Springfield, Oregon during the week of June 8, 1981. EPA Methods 1, 2, 4, 5X*, 9, 22 and 25 were used to measure emissions at the veneer dryer exhaust and from the scrubber outlet. These methods are presented in greater detail in Appendix G. 5.1 EPA Reference Methods Used in This Program The following EPA Reference Methods were used for the testing at the G-P plywood plant. These methods** were taken from CFR 40, July 1, 1980, part 60, "Standards of Performance for New Stationary Sources," Appendix A, pp. 183 ff.; and Federal Register, volume 45, no. 194, Friday, October 3, 1980, pp. 65959 ff. Method 1 - Sample and Velocity Traverses for Stationary Sources This method specifies the number and location of sampling points within a duct, taking into account duct size and shape and local flow disturbances. Method 2 - Determination of Stack Gas Velocity and Volumetric Flow Rate This method specifies the measurement of gas velocity and flow rate using an S-type pitot tube, manometer, and temperature sensor. The physical dimensions of the pitot tube and its spatial relationship to the temper- ature sensor and a sampling probe are also specified. Method 4 - Determination of Moisture Content in Stack Gases This method specifies the procedures by which the water vapor content of a gas stream be can determined. * Method 5X will be assigned a reference letter designation when the NSS regulation is proposed in the Federal Register. This method was derived from EPA Method 5 and ODEQ Method 7. ** With the exception of the Provisional Method 5X, which has yet to be proposed. 5-1 ------- Method 5X - Determination of Particulate and Condensible Organic (Provisional) Emissions from Stationary Sources in the Plywood Industry This method, based upon EPA Method 5 and ODEQ Method 7, describes pro- cedures for measuring emissions in the context of the following defini- tions. Particulate matter is material which condenses at or above fil- tration temperature and is collected by the front half of the sampling train. Condensible organic matter is that material which remains after extraction, filtration, and evaporation of the impinger portion of the train. Method 9 - Visual Determination of the Opacity of Emissions From Sta- tionary Sources This method specifies the procedures by which opacity of emissions are measured. Method 22 - Visual Determination of Fugitive Emissions from Material Processing Sources This method specifies the procedures for visual determination of the pre- sence and total time of occurence of fugitive process emissions. Method 25 - Determination of Total Gaseous Nonmethane Organic Emissions as Carbon This method describes procedures for the sampling and analysis of gaseous nonmethane organic emissions. An emission sample is drawn through a con- densate trap and into an evacuated tank. Trap and tank contents are oxi- dized to carbon dioxide, reduced to methane, and analyzed by a flame ion- ization detector. 5.2 Preliminary Measurements Before the start of emission sampling, each location was tested according to EPA Methods 1, 2 and 4 to determine the preliminary stack gas velocity and moisture content within the ducts. 5.3 Measurements for Particulate, Condensible and Noncondensible Emissions 5.3.1 EPA Method 5X (Provisional) - Particulate and Condensible Organic Compounds This section presents a summary of procedures followed by TRC during par- ticulate and Condensible organic sample collection, recovery and preparation, analysis, and data reduction. Deviations from the specified method are 5-2 ------- explained in this section. Further details of this method are presented in Appendix G. 5.3.1.1 Method 5X - Sample Collection The sampling train was a modified EPA Method 5X train as shown in Figure 5-1. This train was designed and built by TRC. A slipstream was drawn from behind the heated Method 5X filter to quadruplicate TRC and dupli- cate NCASI Method 25 sampling trains. Vacuum grease was used in the assembly of the Method 5X train ahead of the Teflon sample line-impinger train connec- tion for test 1. This may have caused contamination of the total organic compound samples for test 1. No vacuum grease was used at those locations during tests 2 and 3. A minimum amount of grease was used in the impinger train. Leak checks were performed on the complete sampling train (modified 5X train attached to the six Method 25 trains) before and after each test. Field data were recorded on standard EPA Method 5 data sheets which are pre- sented in Appendix C. The Method 5X sampling train is essentially the same as that described by EPA Method 5 with the following modifications. A flexible Teflon sample line was used to connect the outlet of the 4-1/2 inch glass-fiber Gelman Spectro- grade no. 64948 filter to the impinger train. Since the filter was at a higher elevation than the impinger train, condensation in the sample line ran into the first impinger and not back into the filter. The Method 5X impinger train consisted of four impingers and a 2-1/2 inch glass-fiber filter. The first impinger was a modified Greenburg-Smith (impingement plate removed) charged with 100 ml of deionized distilled (D.D.) water. The second impinger was a regular Greenburg-Smith unit also charged with 100 ml D.D. water. The third was another modified Greenburg-Smith and was empty. The fourth was 5-3 ------- if, j STACK WALL- 17 13 LEGEND 1 - NOZZLE 2 - PROBE 3 - FILTER HOLDER 4 - HEATED FILTER BOX 5 - IMPINGER ICE BATH 6 - UMBILICAL CORD 7 - VACUUM GAUGE 8 - MAIN VALVE TO PUMP 9 - PUMP 10 - BYPASS VALVE 11 - DRY GAS METER 12 - ORIFICE AND MANOMETER 13 - PITOT TUBE AND MANOMETER 14 - THERMOCOUPLE READOUT 15 - FLEXIBLE TEFLON SAMPLE LINE 16 - BACK-UP FILTER HOLDER 17 - THERMOCOUPLES I I I II I SLIPSTREAM" TO METHOD 25 TRAINS 12 Figure 5-1. Modified EPA participate and condensible organics sampling train, (August 18, 1977 Federal Register) ------- also a modified Greenburg-Smith type and was charged with 200 grams of silica gel. A 2-1/2 inch glass-fiber filter (similar to the 4-1/2 inch filter) was inserted between the third and fourth impinger to collect any organic material condensed but not collected in the impingers. Prior to initial field use, all glassware was washed with a chromic acid solution and rinsed with D.D. water and acetone according to Method 5X. Sampling train operations were identical to those of EPA Method 5, with several exceptions. In order to prevent condensation of organic materials in the probe and on the 4-1/2 inch glass-fiber filter, the stainless steel probe and the filter were heated to 350° +25°F. Thermocouples were inserted into the probe and the filter outlet gas stream to ensure that proper tem- peratures were maintained. These temperatures were noted on the field data sheet during routine data recording intervals. During sampling at the scrubber outlet it was sometimes impossible to maintain probe and filter outlet temperatures in the range of 350 +25 F because of insufficient heater capacity. Probe and filter outlet tempera- tures ranged between 310 to 340 F at the scrubber outlet throughout the test program and averaged about 320 F. Filter box temperatures were main- tained at 350 ^25 F with no problems. Impinger outlet temperatures were monitored and maintained below 68 F throughout the test program. These temperatures were not, however, recorded on the scrubber outlet field data sheets. Velocity pressure at the scrubber outlet was extremely low as measured by the S-type pitot tube, hovering near the lower detection limit of EPA Method 2. In addition, wind gusting across the top of the stack sometimes caused turbulence within the 9-foot i.d. duct, creating the illusion of negative flows. If this phenomenon was more than momentary, sampling was halted until flows again appeared positive, and then restarted. 5-5 ------- A 2 1/2-inch filter was used in the front half of the inlet sampling train for tests 1 and 2, and the outlet sampling train for test 2. This was done because the 4 1/2-inch filter assembly would not pass the leak check. During outlet test 3, the filter assembly was broken in the process of changing sampling ports. The filter was recovered and replaced with another filter assembly to complete the test. 5.3.1.2 Method 5X - Sample Recovery and Preparation Sample recovery was performed in an improvised laboratory on site. Because this area had a clean, wind-free environment and was well lighted, it was suited for sample recovery and preparation for shipment. Sample recovery was performed in accordance with EPA Methods 5 and 5X as presented in Appendix G. At the conclusion of each test run, separate sample fractions were collected from each Method 5X sampling train by a three-person clean-up crew. The liquid samples were placed in glass sample jars with Teflon-lined lids, and the filters were placed in inert petri dishes and sealed. The sample fractions collected were as follows: Container 1 - 4-1/2 inch glass-fiber filter (2-1/2 inch filters were used during test 1-1, 1-2, and 0-2). Container 2 - D.D. H2O wash of nozzle, probe and front half of the 4-1/2 inch filter holder. Container 3 - Acetone wash of nozzle, probe and front half of the 4-1/2 inch filter holder. Container 4 - Exposed impinger solution from impingers 1, 2 and 3 and D.D. H^ wash of impingers, connectors, Teflon sample line, back half of 4-1/2 inch filter holder and front half of 2-1/2 inch filter holder. Container 5 - Acetone wash of first three impingers, connectors, Teflon sam- ple line, back half of 4-1/2 inch filter holder, and front half of 2-1/2 inch filter holder. Container 6 - 2-1/2 inch glass-fiber filter. 5-6 ------- The probe and nozzle were brushed and rinsed three times with D.D. H-O, which was deposited in container 2. The front half of the 4-1/2 inch filter holder was also rinsed with D.D. H2O, which was deposited in container 2. The probe, nozzle and front half of the 4-1/2 inch filter holder were brushed and rinsed with acetone in the same manner and deposited in container 3. The Teflon sample line was drained into the impinger train. The Teflon sample line was not brushed because the particulate catch in the sample line is generally considered to be insignificant. Impinger contents were weighed to determine moisture catch and deposited in container 4. The Teflon sample line, impingers, connectors and the back half of the 4-1/2 inch filter holder were rinsed three times with D.D. HO into container 4, and then rinsed three times with acetone into container 5. Prior to tests 2 and 3 both the probe and Teflon sample line were washed with D.D. H_0 after the acetone wash to remove any acetone residue which might have contaminated the EPA Method 25 samples. These washes were dis- carded and the components allowed to dry at ambient conditions before being reassembled. A possibility of Method 25 acetone contamination exists for test 1. Both filters were removed from their holders and deposited into their respective petri dishes, containers 1 and 6. Filter residue on the filter holders was scraped and deposited into the same acetone rinse containers as the front halves of their respective filter holders. The stainless steel and glass filter frits used in the filter holders were not rinsed during sample recovery, because any organic material collected on the frits is generally considered to be insignificant. Glass and/or metal particles could become detached and contaminate sample fractions. Silica gel samples were weighed immediately at the conclusion of each test and the weights recorded by the clean-up crew. All Method 5X samples 5-7 ------- were packed in shock-proof containers and driven to the CH_MHill laboratory in Corvalis, Oregon for analysis at the conclusion of the test program. 5.3.1.3 Method 5X - Sample Analysis With the exception of the silica gel samples, all sample fractions were analyzed by CHJlHill. CHJIHill was chosen to perform the analytical *L ft phase of the Method 5X sampling program because of their extensive experience with Oregon DEQ Method 7, from which EPA Method 5X was derived. All analyses were performed in accordance with EPA Method 5X and as approved by EPA/EMB. The sample fractions were analyzed as follows: Container 1 - (4-1/2 inch glass-fiber filter) - desiccate and weigh after 24 hours. Container 2 - (D.D. 1^0 probe rinse) - evaporate, desiccate and weigh after 24 hours. Container 3 - (acetone probe rinse) - evaporate, desiccate and weigh after 24 hours. Container 4 - (impinger water solution and D.D. ^0 rinse) - extract, desiccate and weigh. Container 5 - (impinger acetone wash) - evaporate, desiccate and weigh. Container 6 - (2-1/2 inch glass-fiber filter) - desiccate and weigh after 24 hours. Silica gel samples were weighed on site with a triple-beam balance at the conclusion of each test by the Method 5X sample recovery crew. The weight gain of the silica gel was determined to the nearest 0.5 gram and recorded. All analytical data were recorded on the data sheets as presented in Appendix H. Sample residue remaining after analysis was~ retained for 90 days after the end of the field program after which they were discarded according to EPA instructions. 5-8 ------- 5.3.1.4 Method 5X Data Reduction All Method 5X data reduction was performed in a manner identical to pro- cedures described by EPA Method 5. (See Appendices B and G.) The only variation from these calculations was as follows. Because of the unacceptable super- isokinetic sampling conditions during tests 1 and 2 at the scrubber inlet and tests 1 and 3 at the scrubber outlet, the particulate mass emission rate (MER) for these runs were calculated by two methods: the concentration method (by which calculations are normally done) and the area ratio method.* With the former method, the concentration of particulate matter entering the nozzle is calculated and then multiplied by the volumetric flow rate to obtain the mass emission rate: (m/V) x Q = MER (Ibs/hr) (Eq. 5-1) where m = amount of particulate sampled (Ibs) V = volume of sampled gas (DSCF) Q = volumetric flow rate (DSCF/hr) If the nozzle sampling velocity is greater than the stack gas velocity (superisokinetic sampling conditions), then the calculated mass flow rate will be less than the true MER. This is because the heavier particles will leave their streamlines (gas streamlines diverted into the nozzle) and will not enter the nozzle, as they would under isokinetic conditions. Since the volume of gas sampled is greater than what would be sampled under isokinetic conditions, the concentration (m/V) will be less than that under isokinetic conditions. With the area ratio method, the mass of particulate matter collected is divided by the sampling time and then multiplied by the ratio of the stack area to the nozzle area to obtain the mass emission rate: * Brenchley, D.F., C.D. Turley and R.F. Yarmac. Industrial Source Sampling. Ann Arbor Science Publishers, Inc., 1973, p. 173 ff. 5-9 ------- (m/t) x (A /A ) = HER (Ibs/hr) (Eq. 5-2) s n where m = amount of particulate sample (Ibs) t = sampling time (hrs) As = area of stack (ft2.) An = area of nozzle (ft"2.) Again, if the nozzle sampling velocity is greater than the stack gas velocity, then the MER calculated by this method will be somewhat greater than the true MER. The lighter particles follow the diverted streamlines into the nozzle; the amount of particulate matter sampled in time (t) is therefore assumed to be greater than what should be sampled. The volume of sampled gas is not a factor in this calculation. The average of the two calculated MERs was used as an estimate of the true MER. Gas stream moisture content measured at the scrubber outlet exceeded the saturation values for the duct temperatures measured. Therefore moisture values were recalculated using the following psychometric equation: HO = B where Py = vapor pressure of air at a given temperature PB = barometric pressure 5.3.2 EPA Reference Method 25 - Condensible and Noncondensible Organic Compounds This section presents a summary of procedures followed by TRC during con- densible and noncondensible organic sampling equipment preparation, sample collection, field sample recovery, and sample analysis. The TRC Method 25 sampling train is shown in Figure 5-2. Deviations from the method are also 5-10 ------- SWAGELOK CONNECTORS H2°-l£ ICE W** CONDENSATE TRAPS VACUUM GAUGE FLOW RATE CONTROLLER ON/OFF FLOW VALVE OUICK r±| CONNECT '1 EVACUATED SAMPLE TANK Figure 5-2. Method 25 Sampling Train ------- explained in this section. Further details of Method 25 are presented in Appendix G. NCASI Method 25 procedures are also presented in Appendix G. 5.3.2.1 Method 25 - Sampling Equipment Preparation This procedure is based on and supplements EPA Method 25 / "Determination of Total Gaseous Nonmethane Organic Emissions as Carbon."* Condensate Trap After being checked for any sign of physical damage, each trap was inter- connected to a hydrocarbon (HC)-free air cylinder, flowmeters and C0_ moni- tor (nondispersive infrared detector (NDIR)) and inserted in the furnace as shown in Figure 5-3. The trap was then purged with the HC-free air at a 100 ml/min flow rate with the furnace operating at a temperature of 600 C. A propane torch was used to heat those portions of the trap and probe assembly that extend outside the furnace. The purge was performed until the C02 monitor indicated a concentration of 10 ppra or less. Sample Tank Each sample tank was connected to a cylinder of HC-free air, a vacuum pump, and a mercury manometer as shown in Figure 5-4. The tank was evacuated to 29 inches Hg vacuum after which the three-way valve was switched and the tank pressurized to 10 inches Hg with HC-free air. This cycle was repeated three times. After the third pressurization, the tank was connected to the TGNMO analyzer and a sample analysis was performed. If a nonmethane organic concentration greater than 10 ppm was measured, the tank was again subjected * Federal Register, volume 45, no. 194, October 3, 1980, pp. 65959-73. 5-12 ------- ® FLOW T METER HC __ FREE I "" Figure 5-3 t 3-WAY /— v VALVE T HC FREE AIR co2 \^\ ANALYZER "*" FLOW \ METER -* ^ TUflP iw '»"' | *— J !?5? ^- rURNACE Mm% . Method 25 Trap Preparation nilTTK TANK v^x Figure 5-4. Method 25 Tank Purging and Evacuation ADJUSTMENTVALVE FLOWMETER ON/OFF VALVE FLQW CONTROL ASSEMBLY PRESSURE GAUGE TRAP QUICK CONNECT TANK Figure 5-5. Method 25 Flow Control Assembly Adjustment 5-13 ------- to the evacuation-pressurization analysis procedure until accepted. Each tank was then evacuated and pressurized with dry nitrogen for shipment to the field. Flow Control Assembly The sampling train was assembled as shown in Figure 5-5 and leak checked. The probe end cap was removed and the probe connected to a flow meter as shown. The sample flow shut-off valve was opened and the flow con- trol valve adjusted to achieve a flow rate of 50 +5 ml/minute. The flow con- trol adjustment screw was sealed after the flow rate was achieved. The flow control valve number and calibration data were recorded on forms presented in Appendix E. 5.3.2.2 Method 25 - Sample Collection The sampling train was a modified EPA Method 25 apparatus. The modifi- cation consists of placing an additional condensibles trap, immersed in a water ice bath, ahead of the trap immersed in dry (C0_) ice. (See Figure 5-2.) The additional trap is intended to remove the high moisture content associated with the process emission streams and prevent freezeup in the dry ice trap which leads to premature sample flow stoppage.* The sample tanks were shipped to the site slightly pressurized with dry nitrogen. Immediately prior to each test, tanks were evacuated. The tank vacuum, ambient temperature and barometric pressure were recorded on the field sampling data sheet. (See Appendix C.) * "Method Development for the Plywood/Plywood Veneer Industry," EPA Contract 68-02-3543, Work Assignment 1. TRC - Environmental Consultants, Inc., August 1981. 5-14 ------- With the flow shut-off valve in the closed position, the train was checked again after a minimum period of 10 minutes. If the indicated vacuum had not changed, the portion of the sampling train behind the shut-off valve did not leak and was considered acceptable. Assuring that the probe tip was tightly capped, the front part of the sampling train was leak checked by opening the flow shut-off valve. After a short period to allow pressure stabilization (not more than 2 minutes), the indicated gauge vacuum was formed. The tank vacuum as indicated by the vacuum gauge was recorded and assembled as shown in Figure 5-2. The pretest leak check was then per- formed. After a minimum period of 10 minutes, the indicated vacuum was again noted. The leak check was considered acceptable if no visible change in vacuum occurred. The pretest leak rate (inches Hg/10 minutes) was recorded if observed. At the completion of the leak checks, the sample flow shut-off valve was closed. After the leak check had been performed, the sample tank number and trap numbers for each sampling train were recorded on the field data sheet with the respective test run number and sampling site. Four TRC and two NCASI sampling trains were connected to each Method 5X sampling train at the insulated outlet of their respective hotbox filters. Immediately prior to sampling, the gauge vacuum and clock time were noted. The flow shut-off valve was opened and sampling begun. TRC gauge vacuum readings were recorded every 5 minutes during the sampling period. At the end of the sampling period, the flow shut-off valve was closed, the time and final gauge vacuum recorded. After the Method 5X sampling was completed, the Method 25 probe lines were disconnected from the Method 5X interface and tightly capped. A post-test leak check was performed prior to disassembly of the sampling train. After assuring that the probe had been tightly capped, the flow shut- 5-15 ------- off valve was opened and the gauge vacuum monitored for a minimum of 10 minutes. The leak check was acceptable if no visible change in tank vacuum occurred. The post-test leak rate (inches Hg/10 minutes) was recorded if observed. At the completion of the leak check, the flow shut-off valve was closed. 5.3.2.3 Method 25 - Field Sample Recovery After the post-test leak check was completed, the TRC sampling train com- ponents were disconnected. Both ends of each condensibles trap were tightly sealed. The traps were then packed in dry ice for sample preservation and shipment to the laboratory. 5.3.2.4 Method 25 - Sample Analysis Two inlet sampling trains and two outlet sampling trains from each test were analyzed by TRC. The other two inlet sampling trains and two outlet sampling trains from each test were analyzed by Pollution Control Science, Inc., (PCS). NCASI analyzed the samples collected with their equipment. Additional analyses of the PCS analyzed samples were performed at the TRC laboratory. The purpose of this approach was to identify the cause of the poor paired sample data precision obtained during the method development pro- gram. The analyses were performed in general accordance with the method as pub- lished (Appendix F). Prior to the nonmethane organic analysis of the tank samples, a preliminary analysis was performed with an FID to determine the relative sample concentrations. The purpose of this analysis was to provide additional information for the resolution of any poor precision. If the pre- 5-16 ------- liminary tank sample analyses indicated relatively equal tank concentrations for paired samples/ the nonmethane organic analyses should also produce relatively equal values. These data are presented in Appendix F. Preliminary Sample Tank Analysis The preliminary sample tank analysis was performed using the nonmethane organic analyzer. However, the separation column, oxidation catalyst, and reduction catalyst were bypassed. The sample tank was pressurized to approx- imately 100 mmHg gauge and then connected to the analyzer. The sample loop was purged an charged with sample gas. The sample was then injected from the loop into the analyzer, bypassing the separation column and catalysts, and introduced directly to the flame ionization detector. The detector response was plotted and scaled by the integrator. The responses for paired sample tanks were compared for relative magnitude on a ppm CH (methane) basis and recorded for reference during the nonmethane organic analysis pro- cedure. TRC Analysis Equipment The analyzer was fabricated by TRC using the following base components: Varian Model 2800 gas chromatograph with flame ionization detector; and Hewlett-Packard Model 3390A Reporting Integrator. These components were interconnected to provide an analyzer scheme very sim- ilar to that described in the method. However, TRC has made some changes which improved the ease of operation without affecting analyzer performance. Figure 5-6 is a schematic rendering of the analyzer as assembled. A high- grade, HC-free carrier gas is used which eliminates the necessity for the purification furnace. 5-17 ------- Ul I CD I SEPARATION COLUMN NONMETHANE ORGANIC (BACKFLUSH) CO C02 1 CH4 COLUMN BACKFLUSH VALV OXIDATION CATALYST HEATED CHAMBER VALVE AIR SAMPLE INJECT VALVE REDUCTION CATALYST HEATED CHAMBER I | DATA RECORDER SAMPLE TANK CALIBRATION CYLINDERS MOLECULAR SIEVE FLOW METER CARRIER GAS He Fiaure 5-6. TRC Nonmetharie Organic Analyzer ------- A six-port valve (Carle Model 5521) was substituted for the two four-port valves in the oxidation catalyst flow scheme. One four-port valve was used instead of two four-port valves in the reduction catalyst flow scheme. In effect the latter valving modification precluded hydrogen venting within the laboratory. The exit line from the oxidation furnace to the six-port valve was heat traced to avoid condensation. Additionally, all four switching valves incor- porated in the analyzer were enclosed in a heated, insulated compartment thermostatically controlled to maintain a constant 100°C temperature. The separation column used was prepared by Supelco, Inc. It was a 4-1/2 foot long, 1/8-inch diameter stainless steel tube with two packed sections. The injection side section was 3 feet long and contains 10 percent OV-101 (liquid methyl silicone) on 80/100 mesh Supelcoport. The following section was 1-1/2 feet long packed with 60/80 mesh Poropak Q. The reduction catalyst was a Byron Instruments unit with integral heater. This was mounted within the Varian gas chromatograph oven to ensure constant temperature operation. Although not clearly shown in Figure 5-6, a single combustion air source services both the oxidation catalyst and the flame ionization detector. Individual metering valves are used after the flow splitter to regulate the supply to each device. The condensate recovery and conditioning apparatus equipment was assem- bled by TRC as shown in Figure 5-7 and was essentially the same as the con- figuration detailed by the method. The NDIR incorporated was an Anarad AR 400, with a range of 0 to 10,000 ppm C02> The TRC arrangement did not incorporate the vacuum pump in a direct link with other equipment. Instead it was located remotely. This was done to avoid contamination by the oil mist vented from the vacuum pump. 5-19 ------- A tube furnace was used for volatilization of the condensate trap sample. This provides more even, high temperature heating of the trap. A propane torch was used to heat those parts of the trap, including the probe, which remain outside the furnace during the sample recovery procedure. Valves A, B, C and D in Figure 5-7 and their connecting tubing were enclosed in a thermostatically controlled oven maintained at 180°C to prevent con- densation. An oxygen rich carrier gas passed through the condensate trap during heating and oxidized the organic compounds to C0_ and water vapor. The flow exited the trap, passed through a water trap and NDIR, and entered the intermediate collection vessel. Analyzer Operating Conditions; Gas Regulator Pressure (psig) Flow Rate (cc/min) Helium 42 25 Air 45 30 FID 50 Oxidation Catalyst Hydrogen 20 30 Separation column normal temperature - 0°C Separation column backflush temperature - 100°C Oxidation catalyst temperature - 750°C Reduction catalyst temperature - 100°C (32 VAC) Condensate Recovery Conditions; Gas Regulator Pressure (psig) Flow Rate (cc/min) Oxygen 10 150 Air 15 50 Oxidation catalyst temperature - 850°C Details of the NCASI analyzer and procedures are presented in Appendix H. Nonmethane Organic Analysis Procedure The analysis was performed in accordance with the published procedure. (See Appendix H.) However, the condensate trap carbon dioxide purge (Section A.3.2 of the published procedure) was modified. After briefly purging the trap according to the procedure, the valves were switched so that the trap 5-20 ------- L±3 FLOW METERS FLOW xCONTROL\ V. VALVES^ HEATED SWITCHING VALVES CONNECTORS CATALYST HEATED CHAMBER I I i ' I ' ' Llt-i I SAMPLE CONDENSATE | TRAP I OXIDATION CATALYST I HEATED | _CHAMBER,_ J VENT 1 1 \~* J NDIR ANALYZER _, VALVE A QUICK A CONNECT[Q| HoO TRAP MERCURY INTERMEDIATE MANOMETER Figure 5-7. TRC Condensate Recovery and Conditioning Apparatus 5-21 ------- was bypassed. After the trap had been bypassed, the carrier gas flow con- tinued through the system and into the tank for approximately 5 minutes. It was then vented to the atmosphere through the valve located downstream of the NDIR. (See Figure 5-7.) This time period was sufficient to purge the inter- connecting tubing and NDIR cell volume. Prior to resuming flow through the condensate trap, the valve was switched to introduce again the flow into the sample tank. The trap was removed from the dry ice bath and allowed to warm to room temperature (determined by touch) . The trap was placed back into the dry ice bath and the valves switched to resume carrier gas flow through the trap after frosting appeared on external trap surfaces. The procedure was then completed as described. This modification to the procedure was intended to assure the removal of any CO which may be trapped within the ice crystals present in the trap.* 5.4 Preliminary Moisture Determination Preliminary moisture tests were performed at the scrubber inlet and outlet prior to emission testing. Testing was performed in accordance with EPA Method 4. Data were recorded on field moisture determination forms as presented in Appendix C. 5.5 Preliminary velocity Determination Preliminary velocity measurements were made at the scrubber inlet and outlet prior to emission testing. EPA Methods I and 2 were followed in measuring the velocity of the gas stream. Data were recorded on the field data sheets (Traverse point Location for Circular Ducts and Preliminary Velocity Traverse, Appendix C). * "investigation of Carbon Dioxide Interference with Method 25." EPA Contract 68-02-2814, Work Assignment 41. Midwest Research Institute, April 15, 1981, p. 7. 5-22 ------- 5.6 Visible Emissions Scrubber outlet visible emissions were not monitored as planned because of overcast sky background conditions. The controlled emissions were bluish- white which made it impossible to distinguish the emissions from the overcast sky. Overcast skies were present on June 8, 9 and 10. During the last test day, June 11, the sky began to clear and only scattered clouds were present in the afternoon. However, the final test sequence was nearly completed by this time. Consequently, no visible emission observations were recorded. Although the scrubber outlet stack had an attached steam plume, the Method 9 observations were not cancelled because of this condition. The method provides for attached steam plumes by requiring that observations be made at the point where the condensed water vapor is no longer visible. 5.7 Pressure Drop Measurements The pressure drop across the scrubber system was measured in order to determine if the unit was operating at design conditions and to provide a means of correlating a scrubber operating parameter with collection effi- ciency. O-tube water manometers were used to measure static pressure at three points at 30-minute intervals during the test period. Measurements were made at the inlet sampling location, just before, and just after the induced draft fan. Pressure drop across the scrubber system was then cal- culated from the static pressures. 5.8 Scrubber Solution Samples Scrubber solution samples were taken from the holding tank of the recir- culating system concurrently with the particulate/condensible organics test- ing performed at the scrubber outlet. These samples were taken to determine 5-23 ------- the relationship between the total organic carbon content of the scrubber solution and the actual measured emissions. A 100-ral sample was taken approximately every 30 minutes while Method 5X was being performed at the outlet. These samples were taken by dipping a 100 ml graduated cylinder into the holding tank. Sample numbers and collection times were recorded on the Scrubber Solution Sample Collection form. The 100-ral aliquots collected during a test were composited into one sample jar for that test. The composite samples were placed in shock-proof containers and trans- ported to CH2MHill at the conclusion of the test program. The scrubber solution samples were analyzed for total organic carbon (TOC) following Stan- dard Method 415.1. All analyses were performed within two weeks of sample collection. 5.9 Fugitive Emissions The purpose of these measurements was to visually determine the frequency of occurrence of emissions that are not emitted directly from the process stack or duct. These are generally referred to as fugitive emissions and include such emissions as those: (1) escaping capture by process equipment exhaust hoods; (2) emitted during material transfer; (3) emitted from build- ings housing material processing or handling equipment; (4) emitted directly from process equipment. EPA Method 22 modified guidelines, as presented in Appendix F, were used to determine fugitive emissions from the veneer dryer doors and abort stacks. The method does not require that the opacity of emissions be deter- mined. Instead, the amount of time that any visible emissions occur during the observation period is measured. 5-24 ------- Fugitive emissions from the veneer dryers were monitored by RTI and DGA. Observations were recorded periodically throughout the test program. 5.10 Ambient Temperature and Relative Humidity Outdoor ambient air temperature and relative humidity were measured with a psychrometer at the beginning and end of each test period. Measurements were made by RTI and their subcontractor to determine if a relationship exists between ambient temperature and relative humidity, and the emissions from the veneer dryers. Data was recorded on a form presented in Appendix I. 5-25 ------- 6.0 QUALITY ASSURANCE The TRC quality assurance program is designed to ensure that emission measurement work is performed by qualified people using proper equipment following written procedures in order to provide accurate, defensible data. This program is based upon the EPA Quality Assurance Handbook for Air Pollu- tion Measurement Systems, Volume III (EPA-600/4-7-027b). At the beginning of each day, a meeting was held to orient personnel to the activities scheduled for that day and to discuss results from the pre- vious day, and to determine if any special considerations were appropriate for the day's work. 6.1 Method 5X TRC's measurement devices, pitot tubes, dry gas meters, thermocouples, probes and nozzles are uniquely identified and calibrated with documented procedures and acceptance criteria before and after each field effort. Records of all calibration data are maintained in TRC files. Samples of these calibration forms are presented in Appendix F. All Method 5X sampling shall be 100 +10 percent isokinetic. Probe and hotbox temperatures were maintained at 350° +25°F. Deviations from these criteria were reported to the EPA/EMB task manager to decide whether a test run should be repeated or continued. A single clean-up evaluation test was performed on each initial set (collector train) of glassware prior to collecting field samples. The eval- uation tests (Method 5X) were performed in the field clean-up laboratory and were observed by the EPA task manager. Necessary changes or modifications to the clean-up procedures were specified by the EPA task manager prior to collecting field samples. The sets of glassware, including the probes, 6-1 ------- were prepared and precleaned before conducting the clean-up evaluation tests. The impingers were precharged as specified in the actual test pro- gram. Afterward, the sample collectors, including probes, were cleaned and the blank samples recovered and analyzed as specified in the actual test pro- gram. Results are presented in Section 2 of this report. In summary, the evaluation tests were designed to precondition the sample collectors, to establish blank background values, and to educate the clean-up personnel in specific sample recovery procedures. Acetone was provided by CH_MHill in glass-lined containers. Both the acetone and D.D. water were analyzed by CH MHill prior to field use. Resi- due data from this preliminary analysis was evaluated by the EPA/EMB task manager with respect to the suitability for use during the test program. These data are presented in Appendix H. In addition, three blank samples of D.D. water, acetone, and both 2-1/2 inch and 4-1/2 inch filters were collected for background analysis. All clean-up evaluation and blank samples were analyzed in conjunction with the actual test samples. All sample recovery was performed by a three-person clean-up crew. Appropriate sample recovery data were recorded on the sample identification log, sample handling log, chain-of -custody form, and analytical data forms as presented in Appendix D. Recovered samples were secured in padlocked, shock-proof, steel con- tainers for storage and shipment for analysis. All preparation and analysis of Method 5X samples were performed by 111, which has extensive experience with Oregon DEQ Method 7, from which Method 5X derives. CH MHill adhered to the standards of quality assurance set forth in the Quality Assurance Handbook for Air Pollution 6-2 ------- Measurement Systems, Volume III (EPA-600/4-7-027b) and the Handbook for Analytical Quality Control in Water and Wastewater Laboratories (EPA-600/4- 79-019, March 1979). 6.2 Method 25 Method 25 traps were burned out according to the method prior to testing and spot-checked for contamination. All Method 25 tanks were flushed with nitrogen and checked for contamination prior to field use. Six sampling trains were used to provide a check on data precision. Two trains were analyzed by TRC; two by PCS, and NCASI analyzed the remaining two trains. All tanks and traps have permanently engraved identification numbers. Analyzers were calibrated over the specified ranges using certified cali- bration gases. Certification forms are provided in Appendix F. 6.3 Method 9 The TRC observer had been certified within the past 6 months to perform visible emission evaluations. 6-3 ------- |