Document vV0kyogYnbqaD8gJGM5rMkjn9

APPENDIX B-l VACUUM METHOD From "Test Plan for Control of Fugitive Emissions from the Synthetic Organic Chemical Manufacturing Industry" Prepared for U.S. Environmental Protection Agency by Radian Corporation DON 79-219-006-04-18 October 19, 1986 Pages 29-36 CUH 000001 70s APPENDIX B-l VACUUM METHOD The screening process gives a rough estimate of the leak rate from a source, but more rigorous procedures are required. to accurately determine tha mass emission rata. Several procedures have baaa used successfully in previous saapling efforts, and aaeh has advantages in certain situations. Tha final decision as to vhleh method vlU be used primarily vill not be made until the process selection is finalized. The first step in any of these methods, however, is source enclosure. 5.3.1 Source Enclosure To accurately measure the leak rate from any given fitting, it is accessary to isolate chat fitting from the ambient air. This is accomplished by an enclosure (or tent) of Mylar plastic (polyethylene terephthalace) formed around the leak souree. The thlclaess of the Mylar ean range from 1.5-15 ail depending on the type of source being bagged. Radian has found that Mylar is veil suited to this function as it does not absorb significant amounts of hydrocarbons, it is very tough, and it has a high melting point (230*C). k typical tent la shown in figure 5-1. The enclosures should be kape as small as practical. * This has several beneficial effects. -l-l 00 000 \.7 0*^ f Figura 5-1. Tone construction around tha saal araa of a varticai pop. 8-1-2 CUH ooooomo \ MAONNUO 2t M WI figure 5-2* gangling train for luggable eourcee of hydrocarbon antaaloea ualag a dlaphragn angling gong. tu to o o o o hho a 3.3.2 the time required to rtich equilibria 1a kept to a llBiflU! the time required to construct the enclosure is minimised, o aero effective aaaI results from che reduced seal ATAAa ABd condensation of heavy hydrocarbons Inside tho enclosure 1a minimised or provescod duo to rodoeod residence time osd docroAAod ourfoeo tru tnlZiblt for boot tn&ifir. Vecuua Method Zs che poet, the method proferred for sampling leeks froa beggsble sources hss bees che dllutloa or vseua aethod* The oeapllag trela used la this aeehod Is shown la Figure 5-2. The crela is Bounced on a portable cart, vhieh can be easily pushed around the unit froa source to source. The major equipment Items la the sampling train are the vacuum pump used to draw air through the system* and the dry gat aetsr used to aessure the flow rete of gas through ths train. The vecuua pump is a 4.8 CJM Teflon-ring piston type equipped with a */* horsepower alr-4rlvea no tor. Low pressure air filOO pelg) is available at or near most process units. The dry gas meter is a Rockwell Modal 1733 Test Gas Meter with a Humber 83 Test Index. Other equipment in the train includes Whltey valves, copper and stainless steal cubing* Teflon hose, 100 ce glass airtight syringe, thermometers, mercury and water manometers, a cold trap, and an air-driven diaphragm sapling pap. The leek source Is shown as a valve la Figure 5*2. Bowever, the seas sampling train can b used for all baggable source sampling with the vacuum technique. The alas and shape of the leak source enclosure (teat} la adjusted to fit each par* titular source shape and operating condition. Vhan the full sampling train is connected to a souree, the vecua puap is able to maintain a maximum flow rate of approximately 2k cubic feet per minute. B-l-4 CUH 00000171 The tent Is connected by uas of the bulkhead fitting and Teflon to the sample train. A aaparata Una is connected from tha tane to a apyMhallc. This allows continuous monitoring of the pressure Inside of tha case* If- a significant veeuim exists inside tha tent whan air is being through, a hole la made In tha opposite side of tha tent froa tha outlet to tha stapling train. This allows air to enter tha tent sore easily gag thus reduces the vacua In the enclosure. In practice, it has been found that only a wary slight vacuus (0.1 la. HiO) is prsssnt la tha taat during moot of ths sanpllng, swan la tha absasea of a hole through tha east vsll. Sufficient air enters around tha asale to prevent the development of g significant vacuum In the tent. Seaple bags will be used to collect gee eaaples and transport then to the mobile laboratory for analysas. Several types of bags wars tested by Sadias In tha laboratory and In the field. Most of then. Including Calibrated Instrument Company1 e five-layer "snout" bags, were found to adsorb hydrocarbons,.making thee unsuitable for use. Bags of 2 Mylar and Tedlar plastic have been constructed and found to be satisfactory. A drawing of a typical sample bag is shown in Figure 5-3. A cold bath la plaead la the system to condense water and heavy organics, thus preventing' condensation la downstream lines and equipment. The cold trap is simply s 500 ml flask la sn les bath. It should be placed es close ss possible to the tent. This Ice bath was found to be very effective in preventing condensation la the remainder of the sampling train and la the gas ssmplt bag. Aay organic eendaaaate that collects in the cold trsp will be measured for later use in calculating total leak rates. The uee of such a cold trap is critical; without it, order of magnitude errors are possible. B-l-5 CUH 000001713 Figure 5-3. Mjler plastic staple beg* B-l-6 oQ oo CA' 0^ I  obtain screening value with the OVA, t&doH the fittins Is a tight Mylar shroud, connect ths sampling train to tha "tost," lamerse ths cold trap in an lea bath, sots tha initial reading of the dry gas aster, start tha raeuua piasp and a stopwatch simultaneously, a record the temperature and pressure at the dry gas aeeer, observe the VOC concentration at tha vacuum punp exhaust vlth the OVA, record the temperature, pressure, dry gas aeeer reading, outlet VOC concentration and elapsed time every 2 to 5 minutes, when the outlet VOC concentration stabilises, tha system is at equilibrium and a gas sample bag is filled from the discharge of the Teflon-lined diaphragm pump, another beg Is filled with ambient elr neez the tent erea to detect the beekground VOC concentration,* a final sat of readings is taken and tha vacuum pump stopped, the cold.trap Is removed, sealed, end transported to the laboratory along vlth the tvo beg samples and the data sheet, remove tent, rescreen source vlth the OVA* All of the above data and any pertinent cements are recorded In e permanent laboratory notebook as vail as on the individual data shear. *Thia bag sample may be omitted in favor of the OVA reading if the background VOC concentration is less than 5 percent of the lover limit of the screening range (l.e., <50 ppm for the 1000 to 10,000 ppm elass). C, D Tbs flow rati through tha fyito can b nrid by throttling flow with a control valve immediately upstram of tha vacuum pump. a* tho : flow rats la decraased, tha concentration of hydrocarbon l&crasti in tht gas flowing through tha sampling ayecam. This allows consldarabla flaxl. bUlty in avoiding operations with an explosive mixture of hydrocarbon in ^ air. If an explosive mixture ia present, tha flow rata may either be adjusted upward (to the marl nun pump capacity) or downward. Tha hydrocarbo, concentration is tha gaa stream can be raised above tha upper explosive limit in many cases by reduel&g tha flow rata. Conversely, increasing the flow rate can reduce the hydrocarbon concentration below the lower exploit*. limit. B-i-8 CUH 000001716 APPENDIX B-2 BLOW THROUGH METHOD c D ] CUH oooo1?17 APPENDIX B-2 BLOW THROUCH METHOD BLOW-THROUGH TENTING OF COMPONENTS LEAKING VOC Tenting refers to measuring the mass leak rate from a piece of equipment (i.e., a valve or pump) by enclosing it in an envelope of flexible sheet material (Figure 1). For instance* Industrial grade aluminum foil can be wrapped around the valve yoke and the ends taped on the bonnet and stem or yoke. If the valve temperature is too high for duct tape* aluminum foil strips can be compressed around the bonnet area with a rope noose. Blow-through refers to blowing nitrogen through the cent to create a constant VOC (volatile organic compound) concentration inside the tent. Nitrogen is metered into the tent through one or two polyvinyl chloride tubes. The temperature and oxygen concentrations are measured inside the tent with a platinum-RTD thermocouple and an oxygen/corr.bustible gas monitor. The flow of nitrogen is monitored in a gas rotameter calibrated to nitrogen. The nitrogen passes through activated charcoal and drierlte to remove any organics-and moisture. The pressure in the cent should never exceed 1 psig. Gas samples from the tent can be collected with a portable sampling pump and transported to a lab for chemical speciation and concentration measurement. Alternately* direct field measurements can be used to determine emission rate. This approach* as discussed below for an organic vapor analyzer (OVA) instrument, has not been generally recog nized by regulatory agencies. At least two OVA measurements are made inside the tent. Each measurement should be made at two or more different nitrogen flow rates to increase accuracy. (The OVA must be attached to a dilution probe and calibrated to nitrogen-diluted gases to allow its use In the nitrogen atmosphere in the tent.) The OVA readings are converted to emission rates of VOC via knowledge of the flow rates of the scream passing through the component. The oxygen concentration is also measured to determine the total gas flow rate through the tent. However* oxygen concentration Is also used to race the quality of the tent, and except in a few extremely difficult situations* VOC concentrations are not measured until the oxygen concentration in the tent is reduced below 5%. Air can replace nitrogen as a dilution gas if the hydrocarbon concentration is not expected to be high enough to cause an explosive atmosphere inside the tent. However, calculations based on air cent data muse asuur.e a nominal tent leakage -are, i.e., e::tra flow, through Che tent due to air entering the tent that is not metered through the tubing. B-2-1 CUH 000001718 Safety has been a key factor in the development of this procedure. In one company, over 400 valves have been tented with this technique without incident. Nitrogen is used as a dilution gas instead of air to prevent an explosive atmosphere within the tent. All cf the instruments used are battery-operated, approved for Class I, Division L use. Listed below is a step-by-step procedure for tenting a pump. Valve tenting is similar. Pump Tenting Procedure 1. From the unit operator determine the composition of the material in the pump (in weight or volume X) , and the operating conditions of the pump. 2. Screen the pump by placing a Tygon nozzle on the end of the organic vapor analyzer, holding the end of the nozzle within 1 centimeter of the seal flange/shaft intersection, and recording the highest concentration seen on the OVA readout. 3. Cut a tent from appropriate material, such as heavy gauge aluminum foil, that will easily fit over the pump shaft housing. 4. Plug any holes in the pump base plate with an appropriate material. 5. Connect Cubing from the nearest low pressure nitrogen station to a rotameter stand, which includes a regulator, dessicant, activated charcoal, and a rotameter in series. 6. Run tubing from the rotameter outlet to a "Y" chat splits the nitrogen flow into two pieces of tubing. Insert the tubes into openings located on either side of the tent. 7. Turn on nitrogen at the utilities station and regulate it at the rotameter to approximately 40 liters/minute. 8. Wrap aluminum foil around the pump shaft housing after the nitrogen is flowing so that there Is no avenue for air to enter the tentenclosed volume* Cut-outs may be necessary for auxiliary piping, e.g., oil mist systems. 9. Secure Che tent to the pump with an appropriate material such as duct tape, wire, and/or rope. 10. Put a third hole In the tent roughly equidistant from the two nitrogen feed holes. 11. Measure Che oxygen concentration in the tent by inserting the lead from an On meter into the third hole. Adjust the tent (add addi tional tape, foil, rope, etc.) until the 0, concentration is less than 5%. B-2-2 CUH 00000J. 12. Measure the temperature in the tent with a platinum-RTD thermocouple. 13. Calibrate the OVA to nethane or hexane at a know concentration in nitrogen using the OVA dilution probe. Remember to correct for the dilution. 14. Check the VOC concentration at several points in the tent with the OVA to Insure that the tent contents are at steady state. 15. Measure the hydrocarbon VOC concentration in Che tent with the OVA at three different nitrogen flow rates. Typically the flow rates will be 40, 30 and 20 liters /minute. However, the lowest rate may have to be at 5-10 liters/minute if there is no OVA response at the higher rates (l.e., the mass leak rate from the pump is very low). After each adjustment of the nitrogen flow, check the 0. concen tration to ensure it stays below 5Z. Alternatively, collect samples in Tedlar or aluminized sample bags by drawing a sample out of the bag with a portable sampling pump. 16. Remove the tent and any plugs from the pump and collect any conden sate on the inside of the Cent in a plastic graduated cylinder. Record the amount collected and the elapsed time the tent was on the pump. 17. If there is liquid dripping from the seal area, collect the drips for a timed period to produce enough collected material for accu rate volume measurement. B-2-3 CWH 0000017 Plant Nitrogen (or Air) 01 t*oI . j C X o o o si }J-OL Figure 1 EQUIPMENT REQUIRED FOR TENTING VALVES 1/4" Insulated Teflon Tubing Not To Sctlm APPENDIX B-3 MULTICOMPONET EMISSION TESTING TO VALIDATE DATA SETS C D CUH 000001722 w I MULTICOMPONENT EMISSION TESTING TO VALIDATE DATA SETS EFFECTIVENESS OF FUGITIVE EMISSION CONTROLS jrry M. Schroy Monsanto Co. 800 N. Lindbergh Blvd. St. Louis, Missouri 63167 The emission of specific hazardous pollutants to the workplace pose both workplace and exvorkplace concerns. Understanding the control effectiveness of various fugitive emission control devices leads to use of effective engineering control practices and to acceptable personnel exposure levels. Verification of the emission rates from various fugitive emission sources for specific chemicals is a key to effective controls. K method for checking the fugitive emission rates from an operating plant will be presented along with validation results. B-3-1 CUH 000001723 f INTRODUCTION 87.93.3 The quantification of fugitive amissions is an increasingly important task as more is learned about the acute and chronic toxicity of various chemicals. The issue of workplace exposure and chronic exposure of those that live near chemical process units has become a key to operation of facilities handling hazardous or toxic chemicals, where the initial work on fugitive emissions by the EPA for petroleum refineries centered on the issue of ozone formation, today the issue of concern is the impact of toxic chemicals on people. In the mid 1970's the EPA worked to characterize emissions from fugitive sources (eg., flanges, valve stems, pumps, compressors, sampling systems and open ended lines) in terms of total hydrocarbon as measured by flame ionization, infrared and ultraviolet light detector devices held close to the source. While the concentration was not a direct measure of leakage rate it provided a rapid method of identifying sources that were the most probable sources of high volatile organic chemical (VOC) leakage rates. In fact correlations were defined which related concentration at the source to emission rate and as long as the methodology was applied to identical processing facilities (ie., other refineries) the methodology was reproducible. In the late 1970*s the EPA worked to apply the protocols developed in refineries to SOCMI (synthetic organic chemical manufacturing industries) facilities. premise of their endeavors was: The "A LEAK IS A LEAK." In some plants, those handling highly volatile low toxicity organics (eg., ethane, propane, ethylene, propylene etc) operating large scale equipment the results were found to be similar to refinery fugitive emission rates. This provided confirmation for the EPA that the petroleum refinery fugitive emission data base could be applied to SOCMI facilities. Unfortunately in selectively examining the data base the EPA ignored a substantial amount of data that illustrated that SOCMX facilities were different than refineries and that fugitive emissions in the chemical industry are not represented by the refinery data base. The reasons for differences noted in fugitive emissions rates and leak frequency were the type and size of equipment used in SOCMI plants and the fact that many SOCMI facilities handle highly toxic chemicals and must control emissions to protect workers and those that live near the production facilities. The EPA did adopt the lower leak frequency results found in the the chemical industry units but failed to acknowledge the lower emission rates defined in the same studies. The issues the EPA failed to assess were the impact of the volatility of the chemicals being processed and the size and design of the equipment. A good example of this limited B-3-2 mt. * cuh 0000 97.93.8 picture of fugitive emission leak sources was the omission of agitator seal leakage rates. The petroleum refining industry uses few. if any agitators and this class of equipment was ignored by the CPA. However, agitators are used extensively in the chemical industry and have been identified in several studies as a key source of fugitive emission losses. This is illustrated by SOCMI facilities efforts to control workplace exposure through special vent designs and the use of only double mechanical seals with flush fluid to limit fugitive losses to the workplace. It has been found that single mechanical seals on agitators are not appropriate because of inadequate heat removal and frequent seal failure. Over the years a number of researchers have worked to quantify losses for a variety of valve and seal designs and the work has been published widely. A summary of the literature fugitive emission rates and those from the CPA may be found in the a paper published in tha Annals of Occupational Hygiene (Schroy, 1986) DISCUSSION In an effort to understand when to use the widely varied fugitive emission factors Monsanto found it necessary to validate the fugitive emission factors for its widely diverse chemical processes. The validation work vaa also necessary to help understand the differences between organic chemical leakage and emissions of inorganic gases (e.g., HC1, HF, H2S, etc). Since the emission rates published by the CPA and most other researchers dealt with organics there is no clear answer for dealing with compressed acid gases or solutions of inorganic and or organic chemicals. Fugitive Emissions A comparison of fugitive emission rate factors taken from CPA publications and general literature is given in Table 1. As discussed above, CPA did not address all services (gas, liquid and heavy liquid), for all potential leak sources but data are available for these sources from literature. The impact of equipment size can also be derived from literature. The CPA's definition of light liquid (ie. any organic with a vapor pressure at 25 deg. C of greater than or equal to 2.2 mm Hg) also has led to classification of a lot of heavy liquids as light liquids. The CPA's conservative approach to defining emission rate from sealing devices (eg., using very high emission rates from the petroleum refining industry for chemical industry facilities) may be appropriate when dealing with ozone but it leads to high estimates when characterizing control device effectiveness. This means that application of monitoring and maintenance standards will result in prediction of larger reductions and bigger improvements in ozone than are possible. The failure to achieve predicted ozone levels in the ambient air is one sign of the fallacy of this approach. For hazardous chemicals this use of B-3-3 CUH 000001725 87.93.8 overly conservative emission rates will lead to over prediction of excess chronic impacts around chemical plants where they are used or produced. The performance standards imposed by OSHA were based on the need to control toxic impacts in the workplace but are viewed by EPA as a lax control approach. This regulatory dichotomy results in duplication of standards for industry from the two agencies and redundant efforts by the agencies to further restrict the type of equipment that can be used. The truth lies somewhere between OSHA*s and EPA's view of an industrial facility. The emissions from engineering controls in SOCMX facilities during normal operations is below the level defined by the EPA. The actual level depends on the type of chemicals used, the process conditions, the equipment sizes, and the type of products. The ambient concentration in and around a process unit are not totally a function of emissions from sealing devices during normal operations, as viewed by OSHA in their latest regulatory action on benzene. Upset conditions, routine maintenance (ie. equipment and instruments) and routine process monitoring actions (eg tank gauging) lead to emissions which result in a background ambient level in the workplace. These operations may in fact, be excluded from performance standards with personnel permitted to use respiratory protection. But in most cases EPA forgets that personnel not directly involved in the exempted operation may also be impacted. At present, the fugitive emission control practices in vinyl chloride, acrylonitrile and benzene facilities maximize available engineering controls. Sophisticated sampling devices, use of ball valves, use of mechanically sealed pumps and minimizing use of open equipment all lead to low emissions to the workplace and low ambient VOC concentrations. The State of Illinois EPA, in their latest fugitive emission regulation, recognized the impact of osHA control actions on fugitive emissions. They exempted any operation which uses ball valves, mechanically sealed pumps, etc. from the maintenance and monitoring standards established in accordance with EPA recommendations. The losses from potential leak sources given in Tables 2 to 5 were calculated using EPA emission rate data for several Monsanto facilities. These calculations were then compared to levels found in the exhaust from workplace ventilation systems and a set of emission factors assembled from literature (Schroy, 1981 and Schroy, 1986). While these mass emission comparisons were only possible for enclosed production facilities, the significant deviation demonstrates that the EPA emission rates are not realistic when toxic chemicals are being processed, similar comparisons have been made for outside production facilities, which incorporate area monitors, with similar findings. It should be pointed out that the comparisons given in Tables 2 to 5 do not represent a few hours or days 8-3-4 CWH 000001726 87.93.8 of data. They, in fact, represent years of workplace monitoring record* and employed exposure monitoring record*. In response to any proposed regulatory action, OSHA or EPA driven, or any internally requirement to quantify fugitive emissions it is suggested that emission estimates be determined as follows: 1. using EPA SOCMX factors, 2. using a fugitive emission summary of literature results, and 3 monitoring exhaust gas concsntrations in building axhausts along with exhaust gas flow ratss and calculating mass losses for your specific building. Both the EPA SOCMI amission factors and a summary of fugitive emission factors from litsratura ars glvan in Table 1. The three estimates can be compared and the most appropriats calculated values (method l or 2) can be selected. The most appropriate estimated values would be those which matched most closely to the mass emission found through method 3. It is suggested that for the initial estimates from the literature, values be made assuming good maintenance practices. If the values for the SOCMI estimates are high and the good maintenance estimates are low, select some fraction of good maintenance (eg., 0.9, 0.8, 0.7 etc.) for all components. The fraction of good maintenance is appropriate when the estimate of losses match the measured value for the facilities. For example, the Plant 4 figures represent 90% of good maintenance. Summaries of comparisons made at four plants ars given in Tables 2 and 3. The five production areas tested all had emissions significantly below the estimates based on EPA* a SOCMI factors. Two cases. Tables 4 and 5, are given as examples of how the estimates are made. If poor maintenance practices vers the rule in SOCMI facilities, the fugitive emissions would be a significant loss, and it is likely OSHA exposure criteria may be exceeded. But, as has been demonstrated at most Monsanto facilities, OSHA criteria are being met and fugitive emissions are significantly below EPA forecasts. B-3-5 CUH 00000172?