Document EdpBRyz4e2da9GZY172YBB6Gg

MINUTES VINYL CHLORIDE SAFETY ASSOCIATION MEETING New Orleans, Louisiana October 31, 1974 COMPANY CONFfDENTfAL DISTRIBUTION LIST E. L. Beeler L. A. Bennett W. E. Brodine H. R. Calsing F. R. Carvell A. W. Clements C. B. Cooper L. B. Crider B. A. DiLiddo F. C. Dilley R. L. Ehnle A. M. Fairlie C. R. Flynn D. Giffin J. W. Goetsch W. T. Gunning E. W. Harrington M. N. Johnson T. B. Johnson E. B. Katzenmeyer, Jr. F. E. Krause R. M. Kreager R. A. Krueger P. H. Lawrence T. R. Linak M. Lukity R. S. Mather G. H. Metzger J. L. Nelson D. P. O'Keefe G. Pow j J. R. Render R. N. Rylands M. D. Schrader E. G. Schwaegerle R. D. Scott R. A. Spurlock R. Stillman R. W. Strassburg C. C. Talbott M. D. Tavney J. A. TePas R. L. Toole A. Vittone P. A. Wagner L. S. Wallis H. Waltemate P. J. Weaver A. R. Webber W. J. Wilcox C. L, Woods B. M. G. Zwicker / 3S46001 BFG26692 MINUTES VINYL CHLORIDE SAFETY ASSOCIATION MEETING New Orleans, Louisiana October 31, 1974 GENERAL 1. The following people were elected on next year's committee: Bob Frantz - Union Carbide Amos Dixon - Goodyear Pete Bogart - Tenneco Walt Iliff - Uniroyal Ray Asti - Stauffer CHAIRMAN VICE CHAIRMAN PROGRAM CHAIRMAN SECRETARY TREASURER RobinTech now has their 35,000 gallon Shin-Etsu polys in operation. They have made ten to fifteen batches to date in this equipment. Certain-Teed expects to start up at the end of the year. Georgia Pacific will start up their plant in January. Bob Babbitt will be involved in the start-up of the Georgia Pacific plant. 2. Tenneco uses a horoscope (a fiberoptic device) to check nozzles for build-up on closed polys. 3. General Tire is using THF solvent cleaning to reduce entry frequency. Seven other companies are either currently using this or plan to do so. 4. General Tire hydraulically pressure test their polys at 170 psig after entry or maintenance. 5. Diamond has a new designed pressure measuring device which does not plug up with polymer. 6. General uses their gas holder for emergency manual venting. 7. FRB fans should not be used in ventilation systems and stacks due to the pos sible static charge build-up. 8. Most companies will use personnel monitoring as the primary control and use area monitoring for alarms. zoo 9w e BFG26693 VCI Safety Association October 3i, 1974 - Minutes Page 2 Reactors -- Joe Mudd, General Tire, Moderator General Tire has 18-4,000 gallon glass lined reactors. In 1968 they in stalled a gas holder and since that time have used the water displacement method to force the unreacted vinyl to a gas holder before opening the re actor. The water is discharged through a knock-out tank with a float valve and eventually goes back into the lake. They clean each reactor after 3.7 charges. They now use hot water to absorb less vinyl chloride. Typical readings inside the reactor prior to cleaning are 4,000-7,000 ppm. The VCI vapor does not get into the room since air is exhausted from the reactor. The operator must wear a respirator at all times when cleaning the polys. They are currently opening 270 reactors per month. A Fetterolf valve is being used quite successfully for flushing out the re actors. General has now installed solvent cleaning using THF. They still use a water purge prior to THF cleaning. A vinyl chloride and THF mixture created too many recovery problems since recovery pressures were too high due to the presence of vinyl chloride monomer. General fills the poly rydrostatically full with THF at the start of the program. All gaskets had to be replaced to eliminate leaks. The program must begin with a clean re actor since lumps do not dissolve but will break off and plug lines. They have now operated with one reactor closed up for a period of 20 days. They have Pressure Products sight glasses or this reactor, with solvent cleaning they will reduce from 270 reactor entering per month to 20 per month. They plan to solvent clean after each 3 to 4 charges. General Tire has considerable experience with THF recovery units since they use THF in their fabrication plants. They use THF instead of EDC be cause it will dissolve five times more PVC. Therefore, the net cost is lower despite the higher initial price of ThF. They did not license the THF cleaning process from Monsanto. Stauffer uses EDC to clean polymerizers. They use a spray rinse after cleaning to remove the film and to prevent black specs and fisheyes. The recovery system for EDC is the same as that for THF. Stauffer reported that the EDC did not destroy the bacteria in their waste treatment system. General Tire pressure test their polys with water at 170 psig after they are entered or any maintenance work is done on them; a good idea to prevent rupture disc bursting at less than their rated pressure. Seven companies reported that they are using or contemplating using solvent cleaning and seven companies reported they are now using high pressure water cleaning. Sixteen companies open and clean after every batch. Many people go multiple batches before cleaning. Two companies reported they went more than 5 charges before opening for any type of cleaning. Twelve companies clean without going into the autoclave, and ten companies open at the com pletion of each charge. BFG26694 23646003 VC1 Safety Association October 31, 1974 - Minutes Page 3 Six companies are entering polymerizers without using supplied air masks. Tenneco reported they now use a stand-ay and a mask for all reactor entering. In cases where Tenneco does not enter for cleaning, they use a boro- scope to check, rupture disc and nozzles for build-up. A horoscope is a fiberoptic device and allows this nozzle to be inspected. Fourteen companies pressure test reactors after cleaning before recharging. Thirteen companies have quick opening manheads. Three or four companies reported that they use a double block and bleed valve to blank off lines for reactor entry. One company uses a water purge between the double blocked valve. Tenneco uses a Hamer blind. Control of Runaway Batches - Eli Zinn, Diamond Shamrock. Moderator At their new plant. Diamond will have pressure and agitator alarms. Each reactor will have a shortstop addition system that will be injected to kill the batch automatically on high pressure or failure of the agitator. The pressure impulse is measured with a system of new design, and this prevents the line from plugging up. (Note: It is possible that this type design has something unique, which could be purchased from Diamond.) Generally, the shortstop will get into the reaction such that it will be dispersed before agitation is completely stopped. Shortstop is injected in to the top of the autoclave. They expect to inject the shortstop with nitro gen pressure. The reason they have decided uoon an automatic system is that they want to take this operation completely out of the operators hands, and make it automatic. Some shortstops mentioned at the meeting were quinones, 1-3 butadiene, and isoprene. ICI reported that they found that venting off the reactor into the recovery system will help distribution of shortstop into the batch. Some people re ported they could reactivate a bate, after it was shortstopped while others reported they could not and the resin was off-grade. Some keep a special tank especially for runaway charges. General have their gas holder ready to use for emergency manual venting. Diamond reported that if the ampere load drops below some point it activates an alarm. They have a two-level pressure alarm such that if the pressure goes up to one point, the alarm sounds and if the pressure reaches the second point, the batch would automatically be shortstopped. Several companies reported they do not like to use automatic degassing for emergencies, but would prefer to use manual degassing since it gives the operator more control over the operation. Four companies reported using condensers for primary cooling. General is looking at a condenser and tem perature device to predict runaway batches. W 3646004 BFG26695 VC1 Safety Association October 31, 1974 - Minutes Page 4 Twenty-three companies reported having an evacuation plan and twenty-two companies have gas alarms. Three companies have pressure interlock systems that sound an alarm if the reactor is trying to be opened while it is under pressure. One company re ported a system to prevent a man from opening the wrong polymerizer. In this system there roust be a metal piece removed below the bottom valve before the manhead is opened. If the metal piece is removed while pressure is on the re actor an alarm sounds. This is a variation of our Long Beach lock and key system. Wacker reported they have a four-stage alarm system. The stages consist as follows: First Stage - Stay and Continue to Work Second Stage - Control the Leak Third Stage - Wait and Pray Fourth Stage - Evacuate They have an evacuation training program once per year. Most people at the session indicated they prefer outside installations for PVC plants. Vinyl Chloride Monomer Storage -- W. C. Koch, Allied Chemical. Moderator Mr. Koch indicated that EFA may set up their own accident response program similar to Chemtrec. The following method was used to clean vinyl chloride cars. Depressure car -- either vent vinyl chloride to a flare or recovery system. Pressur ize the car with nitrogen. Break lines and purge with water. A nitrogen purge is required before sending the car to the shop for repair. Railroad shops will not touch cars for service unless they have less than 1 ppm vinyl chloride content. Stauffer is looking for other people to go into a joint venture for car cleaning before sending them into shop for repair and reservicing. They would have approximately 130 cars per year to be cleaned. DOT regulations require tank cars to be checked each 10 years, relief val ves every five years. Most companies check vinyl chloride sphere relief valves each 2 years. There is no reported corrosion in spheres with pure monomer. However, re cycled monomer does create corrosion problems. BFG26696 23646005 VC1 Safety Association October 31, 1974 - Minutes Page 5 Accident Summary Section A. Hooker reported on the vinyl chloride escape in their mass plant and sub sequent ignition. A Hamer blind was not tightened and caused a vinyl chloride leak. Ignition source in this case was either static or light bulbs. The fire melted the plastic tubing to the control valve for the VC1 supply and it would not close. Some companies have initiated policies against the use of poly ethylene tubing for air supplied valves since it will melt and control will be lost. B. British Petroleum reported on the aluminum ladder rubbing the rusty pipe in which case the thermite effect caused the ignition of vinyl chloride. This was reported in the mass review with Rhone-Progil. C. Wacker reported on their fire in their mass plant. One important consider ation that had not been brought out before was the type of blastwall instal led between the polymerization area and the area of the rest of the plant. They had a perpendicular blastwall in which case the explosion debris hit this wall and ricocheted into the area causing damage. If the blastwall had a 45* angle, the debris from the explosion would have been deflected upward causing less problem. > 4. As a result of this fire, the following policies are observed: 200 feet between the PVC plant and other parts of the process, such as the lab, etc. Easy access of the fire brigade to the polymerization area. Have autoclaves accessible from the outside. Cooling water system should stay intact so that it can be used to cool the reaction down in case of fire or explosion. One man should be responsible for determining when a job is completed when a number of crafts are involved. This man should give his permission before the vessel or the equipment is put back into operation. General reported that communication between shifts was the cause of a runaway and vented charge. (Our board of inquiry reports also show this to be a cause of vented charges in B.F.G. plants.) General uses a man to overlap on shift changes and he would work 11:00 p.m. to 7:00 a.m. for example. BFG26691 23646006 VC1 Safety Association October 31, 1974 - Minutes Page 6 D. Air Products reported the ball blew out of a Pfaudler rinse valve allowing 10,000 lbs. of vinyl chloride to be released into the building. They have since changed to Fetterolf valves which are safer. In this building 25 interchanges per hour was normal and they have a big overhead fan for emer gencies, This was turned on and the people left the building. It took approximately one-half hour to clear the area of vinyl chloride vapors. E. Air Products reported an incident where vinyl chloride from the reactor backed up into the catalyst charge pot. At this time they did not have a rupture disc on catalyst charge pots but now have 2" rupture discs. F. One company reported a manhead failure in which case the manhead popped open allowing 1,000 lbs. of vinyl chloride into the building. Since this time all manheads are checked daily to prevent a reoccurrence of this type of accident. A pin holding the yolk together had sheared off allowing the manhead to come open slightly causing the accident. These pins were found to be cracked from repeated opening and closing of the manhead. Since this time all manheads have been checked in the plant. G. PPG reported the accident concerning IPP catalyst in which case the material detonated with explosive violence. A copy of this report is attached to these minutes, and has been circulated to each one of our PVC plants. This is the only time that IPP has detonated to our knowledge. H. I reported our recent accident involving the use of Evertight fittings. About half of the people use these type of fittings while the others use LPG fittings and find that they are safer and more reliable in their opera tion. I. The subject of poly peroxide in recovered vinyl was briefly discussed. It was reported that many people are routinely checking vessels for oxygen con centration in the vapor space to avoid the formation of poly peroxide. It is recommended that we institute this practice in our plants. Federal EPA Proceedings -- R. W. Landry, General Tire, Moderator EPA will probably come up with a performance standard based on the data from all PVC industry. They will allow the industry some time to comply. The following types of information will be typical of what will be required. 1. Vinyl chloride content of resin3 before and after compounding, pro jected future residual monomer content of resin, and your time sched ule for obtaining these goals. The residual vinyl chloride content of finished products. 2. The emission level obtained in melting or solution make-up of PVC resins. 3. Rate of loss of monomer for PVC pellets considering the time and temperature of the material. fw BFG26698 3646007 VC1 Safety Association October 31, 1974 - Minutes Page 7 4. Relative location of polymerization and compounding facilities. 5. All major processes employed in your operations. 6. EPA will want to prepare detailed flow sheets of your process and will want information such as flow rates of each manufacturing pro cess . 7. Who your major customers are and the end products they make with the resin. It is estimated that EPA will come up with a standard sometime within the next six months. Wacker reported that the German standard for monomer emission is a maximum of 6 lbs. per hour for the total plant. No plant in Germany can presently live with this standard. There was considerable discussion about emergencies where rupture discs allow large quantities of vinyl chloride to escape Into the atmosphere. One company located outside the United States reportedly have a condenser to try to con trol this situation. Most plants do occasionally vent charges to the atmos phere. Some reported about five batches per month or less. Goodyear reported very good service with an FWI ball valve which will stop vinyl chloride from leaking through a normally closed valve. Dow reportedly uses an air powered generator for low voltage lights when a light is required for illumination for polymerizer entry. FRB fans should not be used in ventilating polymerization buildings due to possible static charge build-up. QSHA Permanent Standard -- J. T. Barr, Air Products, Moderator Mr. Barr expects the interpretation by OSHA of this standard to come out in late November or early December. It is difficult to tell at this time how many of the sections will actually be interpreted. Some of his interpreta tions are as follows. All people who supply services to the action or regulated area are also covered under the standard. All persons entering an action area or regu lated area are covered under the standard, and their employers are obligated by law to apply all of the standard's provisions. PVC operators will no doubt be liable for all costs and damages caused by concentrations above the "action level" to persons, companies, or contractors in or near the PVC oper ation or plant. Transportation and transfer station problems will be formidable. BFG26699 23646008 VC1 Safety Association October 31, 1974 - Minutes Page 8 General expects to use job rotation to reduce employee exposure levels. For example, 4 hours in a regulated area and 4 hours in another area of 0 concentration. The railroad men will not be allowed in the plant. The 5 ppm over a 15-minute period could be considered a ceiling value. The standard covers all people handling PVC resin. The standard did not specify whether part per million was measured by volume or weight basis. For monitoring it specifies the plants must have area monitoring and per sonnel monitoring. However, OSHA will check the plants using personnel monitoring. For employees working at a 0.5 ppm level or lower, only a training program is required. However, the 0.5 ppm level or lower must be obtained by two values taken at least five days apart. Allied in their vinyl chloride manufacturing plant plan to use fixed moni toring systems with a grid so that they can relate the concentration in the area to the employees concentration. Firestone will use a fixed moni toring system but supplement it with personnel monitoring. Limited access to regulated areas and a daily roster must be maintained. It has been calculated that one pound of vinyl chloride will represent 1 ppm in an area of 6.75 acres to a depth of 20 feet. We may have to build fences around regulated areas since we are responsible for controlling movement of people in and out of these areas. It is the opinion of John Barr that bulletin boards will not be satisfactory for notifying personnel of their exposure level. A system of requiring the employees to sign a sheet indicating they have been given their personnel monitoring data may be required. OSHA feels that any concentration of greater than 100 ppm would be consid ered an emergency situation. They must be informed of all emergency situa tions within 24 hours. You must have a regular plan to deal with such emer gency situations. Air Products have measured the level of residual vinyl chloride in clothing and have not found any problems. No special clothing is required outside the hazardous area. A training program will have to be provided for all employees. They were hoping that SPI would set up a format for such a training program. John would be willing to compare notes on training programs with other companies. Air Products is making medical surveillance mandatory. Each employee after 10 years service will be checked on a six-month frequency. After employees are exposed to 100 ppm, the emergency level, appropriate medical surveillance is required. Most companies reported 50% of the people failed the SMA-12 test the first time, then on retest 85% of the people passed the test. BFG2610 eocm sez VC1 Safety Association October 31, 1974 - Minutes Page 9 It is probably going to be necessary to work out an agreement with outside contractors to comply with the standard. If below the action point of 0.5 ppm, individual surveillance will not be required. All regulated areas will be posted with appropriate signs and labels as re quired. Anything containing vinyl chloride will have to be labeled -- cylinders, etc. In many cases the lab must be considered a regulated area since vinyl chloride is handled there and the concentration can be greater than 1 ppm. A truck containing bags of PVC need not be labeled, but the bags themselves will have to be labeled. Air Products plan to use microfilm in storing their records for 30 years. All monitoring data and the hours of employment will be recorded. The employee must be informed of his monitoring values within 10 days. There is a real question of how to determine what an employee is exposed to if he is wearing a respirator. This type of information is required to put into his medical records. A legal interpretation will have to be used in this case. Nineteen companies have area monitoring while seventeen additional com panies plan to add area monitoring. Only three out of thirty-six companies will use area monitoring only for primary control purposes. All other com panies will use personnel monitoring as their first line of defense and area monitoring will be used for alarms. Five companies use IR analytical equip ment for continuous monitoring. Firestone indicated it may not be valid at 1 ppm or less. Average Continuous Monitoring From All Companies < 1 <5 <10 <25 <50 Polymerization Area 0 3 14 12 Finishing Area 1 19 13 VC1 Storage 2 27 8 Other Areas 30 2 Fifteen companies have separate control rooms and all expect to get below 1 ppm. Job rotation between the polymerization and the finishing areas may be used to reduce personnel exposure. BFG26701 VC1 Safety Association October 31, 1974 - Minutes Page 10 Several companies have designed their heating systems for 45F and this may not meet some state regulations. Nineteen plants have breathing air supplied systems. Two of this nineteen use water sealed compressors and one company uses bottled air for each building. Five companies have car bon filters in their air supply systems and eleven companies have filters to remove particulate matter. Four companies continuously monitor the breathing air for presence of impurities. General reported using a two-stage stripping system for removal of vinyl chloride monomer from PVC. They use temperatures of 170 to 180 for approximately 75 minutes at a 20" vacuum. This gets the residual vinyl chloride air to values under 100 easily. Wacker Reported Data From European PVC Companies on RVCM Content Emulsion PVC Fine Particle Size Emulsion PVC Coarse Particle Size Suspension PVC Mass PVC Max. < 10 2 200 10-1000 00-1000 100 John Barr, Air Products, stated that a 10 ppm RVCM would be low enough. He felt that this represented only 50,000 lbs. max. per year of VCM from all 5 billion lbs. per year of U.S. PVC and this would be an acceptable level (BFG questions this). GDS/cls 11/15/74 Att. 2 G. D. Schaaf / C BFG26702 OCT o D1974 ''QiL(2av\ 0^ IAAju^ ?1 r*g *25Si. INDUSTRIES PPG INDUSTRIES, INC./20325 CENTER RIDGE RD./CLEVELAND, OHIO 44116/AREA 216/331-4100 foh fr'li October 24, 1974 Regional Sales Office Industrial Chemical Division Mr. E. W. Shaw B. F. Goodrich Chemical Company 6100 Oak Tree Boulevard Cleveland, Ohio 44131 Dear Mr. Shaw: SUBJECT: SAFE HANDLING AND STORAGE OF IPP A unique type of decomposition involving a quantity of stored Di-Isopropyl Peroxydicarbonate (IPP) has recently occurred. We wish to re-emphasize to our customers that IPP, as well as our other undiluted Peroxydicarbonates, should always be maintained at 0 degree F or less during storage. As a cautionary measure, we suggest that you review your plant practices to ensure that IPP is always stored at or below this recommended temperature level. As far as is known by PPG, all previous decompositions of stored IPP have followed a classical pattern: 1) Adequate cooling has not been maintained. 2) The temperature of the product has increased, and this has been accompanied by normal exothermic decomposition. 3) As the product temperature has increased, the rate of decomposition has increased, culminating in an autoaccelerative decomposition. 4) The peak of decomposition has frequently been accompanied by fire, started either by an external ignition source or by auto-ignition of the flammable decomposition products. The most recent decomposition involved 650 pounds of IPP that was stored in an insulated shipping chest which utilizes dry ice as the cooling mechanism. The summary of events concerning this decomposition are as follows: 1) The chest was assumed to be empty of product and dry ice was not replenished. Loss of cooling has definitely been established as the primary factor causing this decomposition^, (Continued) <r BFG26703 h PPG INDUSTRIES/INDUSTRIAL CHEMICAL DIVISION/CLEVELAND, OHIO Mr. E. tf. Shaw Page -2October 24, 1974 2) After several days with no dry ice additions to the storage box, decomposition occurred. 3) Fire was observed to accompany the decomposition. 4) After a short period of flames issuing from the box, an explosion took place. 5) Fortunately, there weTe no personal injuries involved in this decomposition. This explosion of stored IPP is totally unprecedented in so far as is known by PPG, but it must now be acknowledged that such an event can occur. The force of the explosion was significant. It completely destroyed the storage chest, and the shed type storage facility was damaged beyond repair. However, an adjacent well-iced chest of IPP (750 Pounds) was extensively damaged, but the contents did not decompose. To reitterate, we do recommend that all IPP consumers again inspect their storage facilities, review plant procedures and practices to ensure that storage temperatures are maintained at 0 degree F or less. Any storage facility that is smoking, burning or obviously contains decomposing Percarbonates, should not be approached until all signs of activity have disappeared. If you have further questions concerning this incident ot would lik to discuss Safe Handling And Storage Procedures, please call on us. Very truly yours, PPG Industries, Inc. Industrial Chemical Division fames E. Kerns Regional SalesManager JEK:eps BFG26704 23646013