Document NGxmOY61GYZwnRwNbwO96LGaw

ESCAMBIA CHEMICAL CORPORATION INTEROFFICE MEMORANDUM To: From: Re; File Date: Copies: A. K. McMillan/G. B. H. Speed Tenneco's PVC Plant Explosion New York Meeting on October 24, 1969 December 10, 1969 The subject meeting was held in New York by Tenneco Chemical to report to the PVC resin producers on the safety and engineering aspects of the fire and explosion which occurred at 3:46 p. m. on August 12, 1969 at their Flemington, New Jersey plant site. Every PVC producer in the U. S. was present. The meeting was opened by Mr. Martin G. Caine, President of Tenneco Chemical, who outlined the personnel and public relations surronding the explosion as follows; 1. The explosion and fire was confined to the reactor building. 2. Four men received extensive burns, two of which died, one was just discharged from the hospital and the fourth Is nearing the point of discharge. 3. Emergency procedures went into effect within a few moments of the explosion. 4. State police were notified (there is no Local police force). 5. Within five minutes, first aid squads were transferring the injured to the Huntington Medical Center in FLemington where there are special facilities. 6. That evening Tenneco got in touch with the St. Vincent Hospital burn center in New York which is highLy rated in this field. 7. Three of the injured men were transported to St. Vincent, New York, the following day. 8. The president of the American Burn Society, Dr. Cornell, was also brought in. AP00036102 File -2- December 10, 1969 9. Tenneco feels they provided the best possible medical care, attention, and facilities available in, the country. 10. The oldest operator in the plant (aged 60) died in 3 days. The other fatality held on for 30 days. 11. Hotel rooms and board were provided for the family in either New York or Fiemington. 12. One man from the company was assigned to each family to help in any way he could or provide entertainment where it was desired. 13. employees. Daily medical bulletins were published to the families , and 14. Tenneco met with the township committee at the plant site the first night to provide available information and show how confined the damage was to combat apparent flagrant misrepresentation of what happened. Press releases were given out but apparently the newspapers chose to ignore the releases and publish their own version of what happened. 15. Tenneco followed up with more formal press releases and they net again with the township committee. The previous meeting was poorly ttended due to a misunderstanding on the meeting time. 16. All operating personnel were kept on the payroll and used for laintenance and clean-up. Mr, Gordan Rozand presented a brief history of the plant, investigajn of the fire and explosion and damage. 1. Mr. Rozand stated that it was a simple accident with grave nsequence. 2. The Fiemington plant was started-up in 1958 with a capacity of MM lbs. /year. It was expanded in 1962 and by 1965 to a capacity of 80 MM ./year. 3. The plant is on a 27 acre site and includes research facilities, 4. The plant is located in a community of about 25, 000, is not .stralized and is about half way between New York and Philadelphia. ''ile. -3- December 10, 1969 5. Tenneco has a second plant in Burlington, New Jersey which oduces primarily co-polymers and dispersion resins. 6. There were definitely two explosions, probably a third and a estionable fourth in the following time pattern: (a) 3:46 p. m. - flash fire (b) 3:47 p. m. - high concussion explosion (c) 3:47 p.m. + 12-13 seconds - high concussion explosion (d) 3:50 p.m. - last explosion very small 7. The flash fire at 3:46 p.m. restilted in flames 5 to 40 feet in ht. 8. A picture was shown of the fire and there was not much intensity. 9. A second picture showed the fire a short while later and there i considerable decrease. The picture also showed the reactors venting gh a vertical stack from each reactor. The vented monomer did not catch 10. Power was lost at time of explosion and water circulation was > the reactors. 11. Damage was confined to the reactor building. An aerial photo plant was shown which confirmed this. 12. Difficult to plan for an explosion. Personnel responded quickly it promptly to their stations. 13. The first emergency rule is to evacuate and there are designated iy points for office and other peopLe not directly involved. 14. The Assistant Plant Manager and the two operators stayed with lem instead of evacuating the plant. They suffered injuries as a mce. Tenneco is concerned on finding out how to get peopLe to respond ished procedures. 15. The Assistant Plant Manager , Plant Superintendent, and reactor were on the (second floor) reactor floor when they heard a hissing 16. Plant Superintendent ran down to first floor to investigate. He aople at No. 11 reactor and monomer pouring out of bottom of reactor. n to boiler room about 100 ft. away to follow procedure which called risking fires. AP00036104 -4- December 10, 1969 \ The Assistant Plant Manager went downstairs and joined :or operators who were there. All three were injured. There are two reactor operators. Each is responsible for a r of reactors on which they performed all operations including umping. There was one departure from normal procedure in that the operator went downstairs to assist the No. 1 operator by ;g a reactor. ie second operator in going to close No. 10 reactor bottom jlipped. To break his fall he reached upward and grabbed ndle on No. 11. The valve handle was approximately 6 ft. floor. The valve handle came off (held on by capscrew), ball type dump valve opened and dumped the batch from the which had been 20 minutes on cycle. The operator was >nomer fumes. The contents of the reactor dumped in 3 elapsed time from the first hissing sound to the flash \g to determine facts after the incident, the operator oped made reference to bottom valve on No. 11 reactor -vsported to the hospital. This was later corroborated dent. There is a connection from the single reactor line which is only connected when dumping the reactor, sel and recovery is directly from the reactor. The ainless steel fitting, Sch, 5, with quick-closing connection was designed for safety reasons to protect tg inside of a reactor. The elbow on the connecting iped on the bottom of the valve. A sketch of the 1, The other end is clamped on the dump line when tnneco management has been aware of sLippery the floor caused by flushing reactors. They have isualiy slippery and have taken several steps to is roughing the floor, using rubber mats, and oom is vented 15 air changes/minute. The vapors the boiler room which is 100 feet away. There away from the reactor building as 235 feet, vapor between the reactors and boiler house. -5- December 10, 1969 re observed as high as 40 feet. There was considerable >aper in the transfer building. In a very short time there tration of vapors. Flames from burning monomer 'S but were very minor {like a flare). They chose to allow as the source was from broken lines at the pure monomer cable to get to valve to shut the monomer supply off. co does not know the source of ignition. No one was allowed fire was secured, Investigation was quite extensive even % the movements of all people. Some suspected sources of a . Boiler Room , Unloading Compressor . Bulk Transfer System Static : testimony were taken from all people who were involved explosion in any way. They attempted to corroborate it took place. They were unable to enter the plant to icial investigation as they had to wait for the flames to 'ere extinguished six nights after the explosion. They were mother explosion after the flames were out but there have a next door neighbor (Visquene) with whom they rt in case of an emergency which requires that they bring re was a lot of HC1 apparent in the area during the six as burning. Seven fire departments were on the scene explosion. They had trouble keeping the firemen from monomer flame, even though they had been previously e. called in an explosives expert, Mr. Glen `Williams, Thompson explosion. He felt that the most probable plastic hose in the bulk transfer system. They normally el hoses. Apparently there are problems with this xperimenting with various types. There was definite -ie plastic hose. The transfer hose is used in transferring a potential and the hose was in use at the time of the ipson reported on the extent of damage and engineering ssion followed Mr. Thompson's report. polymer building will require repair or it. There was steel damage due to longlg. Steel was displaced by the explosion. File -6- December 10, 1969 2. All of the eLectrical equipment in the building is Class I, Group D, Division. I. 3. There are 26 air changes/hour. Blowers operate at 25 ft. / sq. ft. (1 sq. ft./72 sq. ft. of volume). 4. There are concrete block walls to protect the stair wells. 5. The roof is Flexicore concrete, tar and gravel. 6. The sprinklers are .25 gpm/sq.ft, of floor area. A high hazard rating is .35 gpm. 7. Both floors have manual pushbuttons and fire alarms and there are also outside pushbutton stations. 8. Two-thirds of the roof lifted and was destroyed. Vent walls were completely blown out. The stairwell block wall was blown out and the fire wall between the reactor and transfer buildings were also knocked out. 9* The electrical conduit along the fire walls were also knocked out and this shorted the eLectrical line to the fire pumps. 10. The reactor vessels are independently supported from the ground . 11. Electrical wiring was a total loss from acid attack. 12. They were able to maintain water in the plant from four wells operating from another power source. The power was returned to the main fire pump by 8 p. m. 13. The conveying system also needs repairs. 14. Damage to the slurry tank was primarily electrical. The sLurry tanks and bulk bins were carboned up. 15. Photographs were shown of the flames and twisted and damaged piping. 16. Evidence of carbon was seen on tanks at a distance of 50 feet from the reactor building. There was evidence of flame propagation in the unloading area which was 125 ft. FiLe 7- December 10, 1969 from the reactor building. Burned grass was seen 200 ft. from the reactor building. Plastic pipe was damaged a distance of 80 ft. from the reactor buiding. 17. They have open head deluge type sprinklers on the vessels outside the reactor building. They have a gravity flow water tower that contains 50, 000 gallons and was a real help. The sprinkler system remained activated. Heat activated the deluge system. 18. The State Department of Labor and Industry issued instruc tions that there should be no reconstruction prior to their review of the plants. The insurance company also issued similar instructions. The following are changes which Tenneco is planning in reconstruction and a general discussion of the meeting: (a) The quick couplings wilL be replaced with standard flanges. (b) The dump line from the reactors will have double block and bleed valves. The valves will be gear operated with wheel handle s. (c) Letdown vessels will be installed. They are also looking at gas sensing devices which were strongly recommended by the insurance company. These devices will turn on maximum air flpws at 20% lower explosion limits and the sprinkler system at 40% lower explosion limits. The sprinkler system will have a density of . 35 gaW sq.ft, of building and are operated automatically by the gas sensing device or manually as desired. They are looking at air volume change rates of 30, 60 and 120 air changes per hour. There is some concern about where to vent the vapors due to air inversions and possibly poor dissipation conditions. There was no major damage due to any lack of fire protection. Thompson had reported they liked the gas analyzers and had monomer that tripped the deluge system and they felt the water prevented an explosion. Another idea presented was the use of an automatic closing valve on monomer tanks and the valves should be installed as close to the vessel as possible. Tenneco did not have a valve on the day tank to shut off the monomer flow. Tenneco had main storage tanks at distances of 120 ft. and 60 ft. from the reactor building with a remote operating shut-off valve. They are also AP00036108 File -8- December 10, 1969 strongly considering separating the reactor building from present site. Each reactor was separately vented through a vent 20 ft. high above the reactor building. Photographs were shown which showed the monomer venting and it did not ignite at any time. They are considering excess flow valves in the day monomer storage tanks. Some people present use excess flow valves and some avoid the use of day tanks. The relief valves are 6" and B" and are protected by rupture discs. There was no one in the reactor vessels at the time of the explosion and everyone else was going to or in the change room. The off-going shift had charged the vessel. There are normally three people in the reactor building but the operators did not relieve on the job. There were several questions asked, such as what should be considered explosion proof areas. The general opinion was that the reactor building should be Class X, Group D, Division I, but that this rating was not needed for the dryer or transfer buildings. Another question concerned what a safe distance is for the reactor building from other buildings. General Tire said 160 ft. Continental OiL says they separate large monomer concentrations such as tanks and reactor buildings by 250 ft. and separate the reactor building from other major investments by 150 ft. One company says the inside of the dryer and reactor building should be explosion proof but the outside could be TEFC. Continental OiL's setup cuts the vent fans in automatically at maximum rates at 25% of LEL and the sprinkler has 50% LEL. They have a multiple 10 point selector switch program on the gas sensing device which operates at (20) 8 second intervals. Several companies have gas sensing instrumentation. The instrumentation generally is not considered to be a maintenance problem. It was suggested that there would always be static problems and the only way to off-set this sort of thing was to accumulate safety protection methods and devices. Some companies test their sensing systems every 30 days. Emergency procedures were stressed as being needed and that routine reviews were necessary. Several companies reported low order explosions in their pumps and recovery systems due to residual peroxides. Dow in particular reports two such explosions. Dow attempts to compensate peroxide formation by preventing C>2 getting into recovery system. Tenneco routinely caustic flushes their lines for the same resuLt. General Tire has not found any peroxides present. Goodrich has reported peroxides but have had no incidents. In general, no one apparently has an absolute way to eliminate peroxides. There was no concern AP00036109 File 9- December 10, 1969 about dust explosions. One company reports peroxide formations in positive dis placement compressors, but not in Nash type compressors. Monsanto disagreed with this observation. Peroxides seem to occur between stages the recovery compressor and in Letdown vesseL. Some people monitor oxygen in the recovery system. Tenneco is planning to use nitrogen to break acuum on the reactors. There was considerable comment that this was too ostly and beyond the economics of PVC operation. One European company eported that they had an air inlet valve open on the suction of the monomer ecovery compressor and experienced an explosion 40 ft. downstream from .e compressor. Tenneco reported experiences with spontaneous combustion on e inlet to the dryer on co-polymer production. Similar experience was ported by others. Even when inlet temperatures were kept below 400F mneco keeps the inlet at 380F maximum and as low as 350F. These mperature ranges seem to be common to other industries. The resin parently gets back into the inlet chamber when dryer air flows are shut off i temperatures also get high. The filter paper could ignite at 370F. Some pLants have outlawed any baLl valve as a terminal connection insist on a blind flange. SightgLasses have been abolished since the mpson explosion. Emergency procedures should prevent all buildings rr negative pressure to be shut down to help prevent taking in vapor. It a calm humid day when the Tenneco explosion occurred. The group was canvassed concerning continuing meetings of the group to discuss safety. The most favorable suggestion was to tie in Ch meetings with the first meeting scheduled in Atlanta in February, e was not too much enthusiasm for associating with the Synthetic Rubber oil. There was general concern that the meetings not be open to others those in the PVC resin industry and that insurance people in particular ot out. In separate conversations with the Tenneco peopLe, we got the distinct ssion that they still have not resolved what they plan to do in reconstruction r plant. They are apparently inundated with suggestions and recommen3 from their Labor Department and the insurance company, many of I apparently are not practical. It seems fairly certain that they will 1 te the reactor building from the other buildings by 150* to 250'. We also I impression that they have to carefully weigh the economics of all the l 3 and improvements they consider in view of the tight profit margin on I Tenneco is far from being firmed up on what they will do. \ We also talked to the Vice President of Operations from Panasot and l apparently using independent hydraulic drives on each reactor and 1 at the infinite variation in agitator speed is a necessity in quality resin on. These people have bottom entering agitators which they state \ st improvement over the top entering agitators. This has been an | development program. AP00036110 File -10- December 10, 1969 All those present were asked to send letters to Xenneco stating that they would like to have safety meetings and Tenneco will attempt to make necessary arrangements if there are sufficient interested. / fls T. L. Carey (2) Dr. M. Smith Dr. D. Mann D. V. Lowe E, M. Spurlock W. A. Miller L. Plowman J. D. Kramer A. K. McMillan AP00036H 1 EXHIBIT / PIPING BOTTOM NO. If REACTOR AP00036112 EXHIBIT / PiPt'NG bottom mu reactor AP00036113 Job No__ Location. SUBJXCT-- ESCAMBIA CHEMICAL CORP. - * < k ,~t' l fas. oatc. By___ f* i p ** *) A/o /} pi Pc Li *~ Rtac tb*~ W* Tastes hu*-t Be /7 vc /b <_ '\ //6~ ^ pjc*re* on/^ 4 P * ct *-i ej &u'ic.i. AP00036114  Approximately 50 attendees Responses - Safety Meeting October 17, 1969 Addressee Representative Mr. B. D. Buckles Allied Chemical Corp. Morristown, N.J. Sanford Schreiber William T. Gerson Mr. A. N. Wohlwend Airco- Chemical & Plastics New York, N.Y. John Barr Joseph Bourque, Jr. Mr, Sidney Weinstein Atlantic Tubing & Rubber Cranston, R. I. Harold A. Winstead Carl H. Swick NAME TAG TO SHOW; Representative's name Name of Company Title________________ Corporate Manager, Safety & Loss Prevention Production Supervisor Painesville PVC Plant Tech. Mgr. Calvert City -------------------- -------------------- Plan- Assistant Safety Director, Air Reduction Co, Inc. Chairman of the Board Executive Director Dr. Harry C. Wechsler Borden Inc. New York, New York William J. Hollowav Emil Lorfing Mr. Gordon Ferguson Conoco Plastics Wilston, Conn. F. D. Fisackerly Robert S. Allen ^LVI^n*/iTT-ji i J ,, Wi Mr. Harry E. Connors, Jr. Diamond Shamrock Chemical Clevland, Ohio E. A. Rose Harold E. Birr Mr. C-. J. Williams Dow Chemical Co. Midland, Michigan Mr. A. K. McMillan Escambia Chemical Corp. Pensacola, Florida Mr. C. M. Neher Ethyl Corporation New York, N.Y. ` J. C. Weisel D. /,, ttozas CJ r G. B. H, Sneed A. K McMillan C. t, Mebl M e rKn-Mr-K enerrke -Jaene-9' -i i- Srieriey' Onerations Mgr. Chief Eneineer General Manager, Aberde Plar Mer. New Products Plant Manager Plant Manager Delaware Citv PVC Plan Supv., FVC Manui. Vinyl ChLor. Manuf. Sun- Production Mgr. PVC Plant Manager ^Ve-Pten^&ctp^rmeertdetit AP00036115 o Mu, Lee Kuhn Firestone Pla.sti.cs Co. Pottstown, Pa. -2W. W. Madden Mr. George Hackim General Tire & Rubber Akron, Ohio M. Gi Glenn' - Mr. Robert D. Scott B. F. -Goodrich Chemical Cleveland, Ohio P. H. Lawrence D. L. Dowell Mr.V. L. Peterson Goodyear Tire & Rubber Akron, Ohio Malcolm Trowbridge Mr. M. Rosensaft Great American Plastics Fitchburg, Mass. A1 Fuhrman (Dr.) John Floros' Mr. E. E. Winne Hooker Chemical Corp. Hicksville, L. X. Mr. J. B. Keysor Keysor-Century Corp. Saugus, Calif. Wm. J. Wetzel J. R, Mehall Russell W. Peters -- Mr. Francis E. Reese Monsanto Co. Creve Coeur, Mo. Paul Bureau Roger Wojcik Mr. Alan E. Ardis Olin Mathieson Chemical Assonet, Mass. Mr. Oliver L. Niehouse The Pantosote Co. New York. N.Y. Alan E. Ardis Herbert Kaster John E. Ertel Edward Biehl Coordinator of SDecial Products Plant Mgr. PVC Plant, Ashtabula, Oh General Mfg. Manager Plastic Materials Safety Supervisor for Chemical Division General Foreman-Vrnvl Di: Vice President Plant Manager Division Engineering Mgr. Facility Mgr, -HomoDolvmt Plant Manager -- Plant Safetv Supv. Springfield plant Manuf. Tech. Liaison Mai Springfield Plant V. P. Dperati nna Director of Mfg. . Vice President, Production Mgr., Resin Operations AP00036116 -3- Mr. Chester L. Arnold Stauffer Chemical Co. New York, N.Y. Frank Michels H. R. Jepsen Dr. T. F. Hart Union Carbide Corp. New York, N.Y. T-et euhuiu W. R. Manning (Dr.) Mr. Earle S. Ebers I Uniroyal, Inc. New York, N.Y. M. J. Kleinfield J. M. Poynter Mgr, Project Engineers Plant Mgr. Delaware City Production Mgr. Vinyl Resins & Vinyl Chloride Texas City, T. Engineering Mgr. Vinyl Resins-Vinyl Chloride South Charleston, W. Va Development Mgr. Marvinc ----------------------r (PVC)Chwn.Di Factory Manager, Painesville, Ohio PVC Plai TENNECO CHEMICALS Dr. S. R. Sheeran S. J. Spitz, Jr. R. H, Marks J. F. Jacobetz J. P. Lamb TENNECO PLASTICS Martin G. Caine G. I. Rozand C. G. Thompson A. Siegel P. Bogart P. Scarito AP00036117 8H9000dV