Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
29 COMMUNICATING CORPORATE DISASTER The Aldicarb Oxime Release at the Union Carbide Plant at Institute, West Virginia on August 11, 1985 Rob Coppock The escape of a cloud of toxic gas from Union Carbide's Institute, West Virginia plant created far less hazard than did the methyl isocyanate gas leak in Bhopal, India the December before. But occurring as it did in the Bhopal spotlight--and just months after the company declared that a Bhopal-like disaster could not happen at the Institute plant--the accident dealt a heavy blow to the company~s credibility with the public. The release occurred on Sunday morning at 9:24. About 10:00 a.m., the County Office of Emergency Services activated its emergency siren at Institute. Hospital emergency facilities treated 134 people, and 28 were hospitalized. The following Friday, August 16th, was the formal announcement of the creation of the National Institute for Chemical Studies (NICS) in Charleston, state capital of West Virginia and Kanawha County seat. NICS had been created by business, academic, and government leaders to serve as a bridging organization between business and the general public. Its activities have influenced subsequent events. On the day of the accident, and throughout the following week, Union Carbide officials conducted a series of press conferences. The following describes these communications, the context within which they took place, and such information as is available regarding their impact. BACKGROUND 1 The Kanawha valley, named for the Kanawha River which carved it, is one of just a few flat areas of West Virginia. Charleston, the state capital and the county seat of Kanawha County, is located at the confluence of the Elk River and the Kanawha. Before the first white settlers arrived, Native American Indians refined brine to make salt for food preservation at
30 the site, and it was the presence of this and other chemical resources that stimulated settlement By 1807, 52 salt furnaces were in operation along the Kanawha River. The salt, oil, gas, and coal that are still plentiful in the area were attracting interest by the late lX00s as the basis for other chemical production. Shortly after the turn of the century, chlorine and caustic were being produced from brine. When World War I cut off chlorine and alkali supplies from Germany, production in the Kanawha Valley increased in importance. The war also resulted in construction of a naval ordinance plant and a plant devoted to making nitrocellulose for gunpowder. In 1920, Union Carbide constructed a plant about 25 miles north, and moved to its present location in South Charleston in 1925. DuPont also built its plant in Belle in 1925. Since then, chemical production in the valley has continued to expand. Currently, the following facilities involving hazardous chemicals operate in Kanawha Valley: FMC, Nitro; Monsanto; Kincaid; Artel: PB&S Chemical; FMC, Institute: Union Carbide, Institute; Rhone-Polenc; Union Carbide, South Charleston; Olin; FMC, Spring Hill; Union Carbide Technical Center; DuPont; and Occidental. The greater Charleston area has a population of about 250,000. In contrast to much of West Virginia, the area has an unemployment rate that is about equal to the national average. The residents are generally aware that the chemical industry has an excellent safety record, especially compared to the coal industry--another major employer in the state. In 1986, 18 people died in the coal mines and more than 1100 were injured. The other major employers in Charleston are state government and medicine. In principle, the community can be divided into three groups with different needs and fears: people living adjacent to chemical plants for whom emergency response is the dominant concern; residents who are somewhat concerned about chronic effects of long-term exposure to emissions; and those who see the chemical industry principally as the provider of jobs and are supportive of it. However, there is some blurring of the borders separating these categories, especially the latter two.
31 THE ALDICARB OXIME EMERGENCY In December 1984, a Union Carbide production plant in Bhopal, India was the site of a leak of methyl isocyanate gas that resulted in about 2000 deaths and estimates of as many as several hundred thousand injuries. Union Carbide also produced methyl isocyanate, or MIC, at a sister plant in Institute, West Virginia. In response to this event, Union Carbide altered its operations at the plant in Institute. The company stopped production of MIC-based pesticides for five months following the Bhopal incident. It spent about five million dollars improving safety measures in the MIC unit before it resumed operation in May 1985. Previously, Union Carbide had produced MIC in Institute and shipped it to Woodbine, Georgia, where it was combined with aldicarb oxime, purchased from Allied Corporation, to produce aldicarb. Aldicarb is the active ingredient in Temik, a Union Carbide pesticide. In May 1985, Union Carbide shifted aldicarb production to Institute to avoid the need to ship MIC. Aldicarb was now shipped in solution to Woodbine, where it was formulated into Temik.2 Production of aldicarb had begun in a unit originally built for other service, which was scheduled to be replaced by a new, larger system. But on August 11, about 3800 pounds of material were expelled from a reactor being used as a storage vessel. Union Carbide chemists and engineers concluded that an exothermic decomposition of aldicarb oxime was the most likely cause of the eruption. The material expelled included about 650 pounds of methylene chloride solvent, 300 pounds of solid residue in pipes from previous use of the reactor, and only ~ pounds of alciicarb oxime. Other materials expelled were carbon monoxide, carbon dioxide, and sulfur compounds.3 A detailed account of the accident may be found in Appendix A. The release occurred at 9:24 a.m. Within 60 seconds, the plant alarm was activated. At 9:35 a.m., the Kanawha County Office of Emergency Services was advised of the release and that the material was, at that time, not identified. Ambulance service was requested because control room operators had been overcome. Due to an error in the computer telephone log at the Union Carbide plant, this communication was recorded as occurring at 9:44 a.m. This error resulted in confusion and later criticism. At 9:56 a.m., a deputy sheriff arrived and remained
32 at the Plant Emergency Center throughout the emergency. Some individuals recall hearing the County emergency siren at this time, indicating that it was activated before 10:00 a.m. By 10:15 a.m., Union Carbide concluded that the release contained a mixture of aldicarb oxime, dichioromethane, carbon monoxide, carbon dioxide and sulfur compounds and the plant physician had completed contact to all area hospitals.4 The plume of toxic materials drifted beyond the plant boundaries into a nearby residential area and across the river to a shopping center. One hundred thirty-four people were treated at hospital emergency facilities, 28 were hospitalized, 6 of whom were Union Carbide employees, with 2 remaining in the hospital more than 5 days. UNION CARBIDE COMMUNICATIONS The toxic material escaped on Sunday morning, when few supervisors were in the plant. The air was hot and heavy, with little or no wind. Within minutes after the release, the shift supervisor employed a computerized simulation program to predict the likely spread of the plume. Not knowing the exact material released, he used MIC as a surrogate. The simulation program predicted that the plume would not cross the boundaries of the plant, and initial communication to the county emergency office did not indicate the need for evacuation or other community response. About 20 minutes after the plant alarm, the assistant plant manager arrived. Shortly after this, the deputy sheriff arrived and state police began setting up roadblocks. By this time the plume had drifted over the plant boundaries and the public alarm had been sounded. By 1:00 p.m., ABC News, CBS News, NBC News, Reuters, and the Today Show had contacted Union Carbide. At about 2:00 p.m., the plant manager met with the regional director of public affairs. In consultation with the head office in Danbury, Connecticut, a press release was prepared. It is attached as Appendix C. The State Pollution Control Commission was also informed. A second press conference was held the next day in the parking lot of the Institute plant. It was prepared with the intention of addressing the questions of the exact materials that had
33 been released and the time that had elapsed before county officials were contacted. It is attached as Appendix D. By Tuesday, the second day after the emergency, reports were appearing in the national press criticizing another aspect of Union Carbide's actions. The plant physician had told area hospitals that aldicarb oxime was a very minor irritant that would have no long-term effects, but the company had classified the substance in the most toxic of four categories in a memo submitted to a Congressional subcommittee the year before.5 Hazards associates with the materials released are described in Appendix B. That day, Union Carbide's regional director of public affairs conducted a third press conference (see Appendix E). His statement went into greater detail about the properties of aldicarb oxime than had previous ones, and announced that a special committee on Safety, Health and Environmental Affairs, established by the company in January 1985 and chaired by Russell Train, President of the U.S. World Wildlife Fund, would conduct an independent investigation of the emergency. Senator Byrd and U.S. Environmental Protection Agency Administrator Lee Thomas toured the plant. On Friday, August 16, Union Carbide Chairman Warren Anderson conducted a press conference at which he pledged that the company would speed up its warnings of leaks. From now on," he said, owe will pull the cord first, than apologize if it wasn't necessary." (See the press release attached as Appendix F.) The press conference was scheduled in part to remove discussion of the Union Carbide emergency from the press conference announcing the establishment of NICS. On Sunday, August 18, about 350 people attended a meeting at West Virginia State College, whose campus abuts the Union Carbide Institute plant. It was attended by the Union Carbide president, the regional director of public relations, and the plant manager. MAJOR INFLUENCES ON UNION CARBIDE'S COMMUNICATIONS Union Carbide's communications about the aldicarb oxime emergency described above can be divided into roughly three separate concerns: activation of county and other emergency
34 response organizations, informing exposed individuals and reassuring local residents, and maintenance of the reputation of the company and its ability to do business (avoiding lawsuits, for example, or countering lower prices for its stocks). There was tension between three conflicting perspectives--a conflict that probably exists for corporate communications about any emergency. First and foremost were the legal ramifications. After the initial emergency response, when the overriding concern is what to do to contain and stop the emergency, almost every business person immediately wonders who will bring suit. Legal advice is almost always to give out as little information as possible so as to avoid providing ammunition for use in court. This is in almost direct conflict with what communications and community relations experts advise, which is to say everything you know, as quickly as possible, in terms the layperson can easily understand. Somewhere in the middle is the scientific and engineering perspective, which cautions against attributing cause-and- effect before being reasonably certain what happened. They prefer to say as little as possible until being sure about the course of events. Whatever message is finally sent out will probably involve a compromise among these three perspectives. One result is that at Union Carbide, like most large corporations, almost all formal communications are carefully vetted. Pre-approved information sheets provide data on toxicity of various materials, and the contents of the two initial press statements were approved by the corporate headquarters via telephone. By Tuesday, when more detail was presented, communications experts from corporate headquarters and local public relations consultants were involved. A second major influence on the content of Union Carbide's press statements was the treatment of the emergency by the press. This attempt to be responsive to the needs of reporters, however, can backfire. For example, the initial statements indicated that the release included both aldicarb oxime and dichloromethane. The press, however, devoted its attention almost exclusively to aldicarb oxime and the discrepancy between the plant physician's description of the material as a minor irritant and its classification by Union Carbide in its most toxic category. Subsequent statements by Union Carbide went into detail about the
35 toxicity of aldicarb oxime. Sometime later in the week, about Thursday, a reporter accused Union Carbide of trying to cover up the fact that dichloromethane was also released. THE MEDIA A detailed examination of press treatment is beyond the scope of this paper. At least sixty-nine articles related to the emergency appeared during the first week. However, the following opinions seem to be shared by Union Carbide personnel, employees of other chemical companies, and local officials with whom I spoke. Many individuals in the Kanawha Valley feel that Union Carbide has been unfairly treated by the national press as a result of the Bhopal affair. One company employee felt that national reporters came to town to gather support for a story they had already written rather than to find out what happened. Nearly everyone I spoke to characterized local news as having a better understanding of regional economics and the chemical industry. The now-deceased editor of one of the two Charleston-based newspapers had a policy of Sustained outrage" in his paper. It reputedly challenged government and industry consistently about pollution and similar problems. But both papers are viewed as dealing with the emergency more fairly and more knowledgeably than the national press. As has been observed with respect to other issues,6 the local and regional press included more Human interests stories. Interviews of the first residents to hav detected the release, a song about the release written by a local disc jockey, plant workers sharing oxygen masks during the emergency, a canvass showing the sirens failed to warn most residents--these are stories that probably lack lasting appeal beyond the local or regional area. COMMUNITY ATTITUDES In principle, the community can be divided into three groups with different needs and fears: people living adjacent to chemical plants for whom emergency response is a dominant concern; residents who are somewhat concerned about chronic effects of long-term exposure to emissions; and residents who see the chemical industry principally as the provider of jobs
36 and are supportive of it. However, there is some blurring of the borders separating these categories, especially the latter two. In late spring 1985, before the emergency described in this paper, NICS initiated an activity designed to assess the views of residents in the Kanawha Valley on the benefits and environmental risks of living in the Valley. The Public Agenda Foundation, a nonpartisan, nonprofit research and education organization located in New York City, was asked to undertake a two-part project. Phase one consisted of six focused group interviews with residents, chemical plant workers, and community activists held in August and September. The second phase consisted of a survey of public opinion later that fall. The focused group interviews suggested that the August 11 leak and the subsequent events were polarizing public opinion between chemical workers, on the one hand, and accident victims, on the other.7 A series of hypotheses developed from these focused group interviews are presented in Appendix G. Phase two tested the hypotheses developed in the initial phase.8 The principal findings of the survey are: - People in the Valley have a reservoir of good will toward the chemical companies; - People in the Valley want a sound economy and safe environment but currently feel they have neither one, nor are they optimistic that conditions in the Valley will improve in the f uture; - People are willing to make certain trade-offs in order to improve the economy and environment, but only within limits, and not generally at a risk to their health or by rolling back environmental regulations; - While people believe that the industry has helped to improve the environment in the Valley, they question the companies' commitment to reducing health hazards, and they don't entirely trust them as reliable sources of information; and - Although residents are generally well informed, many people have little trust in the information sources available in the Valley.
37 The report on phase two includes the following conclusion: When the Public Agenda began its research in the Kanawha Valley, in August of l9SS, it was by coincidence that the Union Carbide incident occurred at Institute. It was striking that at that time public attitudes in the Valley, at least in the focused group discussions, did not appear to shift dramatically against the chemical industry. Nevertheless, because residents do not entirely trust chemical company spokespersons or other information sources and because of the clear concern for the environment revealed in the survey, one cannot be confident that public attitudes in the Valley will remain where they are today. It is the Public Agenda's sense that another leak or accident--one perhaps not even in the Valley--could turn current public attitudes which appear to be relatively stable into a more volatile, perhaps even hostile, state. The most vocal of those living adjacent to the thirteen chemical plants in the Valley appear to be those living in Institute. This may reflect an underlying racial issue. The residents of Institute are predominantly black, and the State College was originally a Negro teachers college (current enrollment is about 85% white, matching the figure for the West Virginia population as a whole). Unlike other communities bordering chemical plants, residents in Institute cannot count employees of the plant among their numbers. Until only a few years ago, the company did not hire blacks. Not only does the lack of ~neighbor- spokesman~ mean that the firm's position is not represented in the community, but there appears to be a residual feeling of distrust and hostility toward the company's previous policies. On August 18, l9SS, just a week after the aldicarb oxime release, about 350 people attended an emotional meeting at West Virginia State College, whose campus abuts the Union Carbide plant. The meeting was sponsored by People Concerned About MIC," a citizens' organization, and was attended by Robert Kennedy, President of Union Carbide. A year later, about 50 people met at a quieter meeting to remember the event and assess change.9 At the first meeting, Kennedy promised improvements in emissions, and by the next year they had been reduced by 30 percent. At the 1985 meeting, a Union Carbide employee accused the company of manning the plant with undertrained personnel. A year later, the same person believed that a new safety training program had resulted in greater safety. Questions about the extent of injuries that would have occurred had the college been in session, as opposed to
38 the actual event with occurred on a Sunday morning, have yet to be resolved. Although an additional evacuation route has been built, some believe another access ramp to the interstate highway is needed. People who live adjacent to the chemical plants and for whom emergency action is the greatest concern probably number in the hundreds. Those who are primarily concerned about chronic effects probably number in the several thousands. The remaining hundreds of thousand residents of the Valley probably see the chemical industry as the provider of jobs and are supportive of it. RESULTING CHANGES IN UNION CARBIDE POLICY AND PROCEDURES Several changes appear to have been instituted in Union Carbide as a result of this emergency. The company claims to have reduced the degree of judgment allowed personnel in the decision of when to initiate communications with emergency response organizations outside its plants. It aims at being more conservative, to err on the side of rapid and immediate communication. The company also claims to have initiated more intensive audits of safety procedures in its plants. Internal audit teams review the procedures developed by plant managers. In addition, community emergency response organizations employing more sophisticated control centers, such as that in St. Charles Parish, Louisiana, have been examined. Finally, a series of regional meetings with plant managers was organized. These were two day seminars covering topics such as how to deal with print and electronic media, how to identify and use information resources, and how to establish rapport with local media and civic leaders.
39 NOTES 1. The material in this section relies on an anonymous paper titled Chemical Reactions in the Kanawha Valley of West Virginia" distributed to participants at the National Institute for Chemical Studies conference "Living with Chemicals: Communicating the Risks, Benefits, and Choices, March I 1-13, 1987. 2. "Union Carbide: New Accidents Revive Safety Issues," Chemical and Engineering News, August 19, 1985, p.4. 3. Carbide Restructures: Problems Prompt Massive Cutbacks," Chemical and Engineering News, September 2, 1985, p.7. Union Carbide: New Accidents Revive Safety Issues," Chemical and Engineering News August 19, l9X5, p.5. 4. Based on an interview with Thad Epps on May IS, 1987, and on a Union Carbide press statement dated August 12, 1985. 5. Stuart Diamond,"Carbide Blames a Faulty Design for Toxic Leak; Effects Can Be Serious, Company Memo Says," New York Times, August 13, 1985, pp.Al, B8. 6. See, for example, Harold I. Sharlin, MOB: A Case Study in the Communication of Health Risk, report to the Office of Policy Analysis, U.S. Environmental Protection Agency, Washington, D.C., January 9, 1985. 7. The Public Agenda Foundation,~Public Chemistry--A Report on the Feasibility of a Communications Project for the Kanawha Valley," The Public Agenda Foundation, New York, undated. S. The Public Agenda Foundation,~The Kanawha Valley: A Report on Public Attitudes Toward Economic Benefits and Environmental Risks," The Public Agenda Foundation, New York, January 1986. 9. Mike Towle, "Carbide Neighbors View Leak As Catalyst For Change," Charleston Daily Mail, August 8, 1986, p.l-A.
40 APPENDIX A - THE CHRONOLOGY OF EVENTS (based on: "Carbide Restructures: Problems Prompt Massive Cutbacks, Chemical and Engineering News, September 2, 1985, p.7.) As Union Carbide investigators reconstruct the incident, methylene chloride was piped Into the aldicarb reactor on August 1. Aldicarb oxime was introduced next. A malfunction in a flow meter caused too much oxime to be let in. Engineers installed a temporary pipeline to pump the solution into another 5000 gallon, glass-lined reactor that had not been used since 1 984. Unknown to the operators, steam flowed into the jacket of this reactor from August 1 to 7. The heat caused methylene chloride (boiling point 104°F) to boil off from the oxime (boiling point 410°F), raising the concentration of oxime from 38 percent to 81 percent. Methylene chloride was later found to have condensed in two connected vessels called the knock-out pot and the crystallizer. Not realizing that this change had occurred, Union Carbide operators tried to pump the reactor contents back into the aldicarb reactor on August 7. After two and a half hours, the pump appeared not to be moving any more material, and it was shut off. Operators could not use level indicators to determine that the pumping was incomplete because the remaining 500 gallons formed what is called a tree! in the convex bottom of the vessel. Steam continued to flow into the reactor jacket, and the temperature rose. To know that this was happening, operators would have had to call up the display of the reactor's condition on the control room computer. Thinking the reactor empty, they did not do this. From laboratory experiments and computer simulations, Union Carbide investigators conclude that on the morning of August 11 the reactor temperature reached 300 °F. This initiated exothermic decomposition of the oxime. At 9:25 a.m., a differential pressure alarm went off, indicating that a safety valve had burst and that a high flow of gas was flowing
41 through the knock-out pot and scrubber to the flare tower. An operator entering the control room heard a rumbling sound from the reactor. He looked toward the flare tower and saw a . lot of smoke. After four seconds, the differential pressure indicator dropped precipitously, Indicating that other release valves had opened. The plant alarm was sounded at 9:26 a.m. An opaque, white fume engulfed the yard and control room, trapping five operators and their foreman. The men could not see the panels in front of them. At 9:32 a.m. the eruption ended. A minute later the workers were led to safety. At 9:36 a.m. county officials were notified. By 10:00 a.m., a public warning siren had been sounded.
42 APPENDIX B HAZARDS ASSOCIATED WITH THE MATERIALS (based on:~Union Carbide: New Accidents Revive Safety Issues, Chemical and Engineering News, August 19, 19SS, p.5) The principal materials released were methylene chloride solvent, aldicarb oxime, carbon dioxide and monoxide, and sulfur compounds. The principal hazards were associated with methylene chloride, which made up the bulk of the cloud, and aldicarb oxime. Methylene chloride, which boils at 104°F, is nonflammable, and its mixtures with air are not explosive. The LD50 (dose that kills 50% of experimental animals) of methylene chloride . is 1.6 ml per kg in rats. EPA considers methylene chloride to be a potential carcinogen on the basis of one recent study. Aldicarb oxime is a high-boiling liquid that melts at 70°F and boils at 410°F. Its vapor pressure is less than 0.1 tort at 70°F. The oxime is an irritant to eyes, skin, and nasal passages. Union Carbide says its LD50 is one tenth that of MIC, and does not believe exposure to the substance to have any long-term health effects, even though it classifies the substance in its "most hazardous category. Because of the lack of extensive health effect studies on aldicarb oxime, West Virginia Health Department officials will conduct a long-term study of all 135 people hospitalized after the exposure.
43 ~ AGRICULTURAL PRODUCTS COMPANY, INC. P.O. BO X 2 8 3 1, C MA Q L ES T O N , w E S T v I Q G l Al I A 2 5 3 30 TITUTE PLANT APPENDIX C FOR IMMED IATE RELEASE August 11, 1985 Institute, WV-- At 9:35 this morning the Union Carbide plant at Institute experienced a release of a mixture of aldocarb oxime, dichlor- omethane, carbon monxide, carbon dioxide and sulfur compounds. A prelimi- nary investigation indicates a gasket failed because of a pressure buildup in a reactor being used as a storage tank. It is estimated the vessel con- tained approximately 500 gallons of material, the maximum amount which could have been released. However, much of the material was neutralized through venting to a scrubber and flare. An immediate investigation will be made to establish the cause of the accident and the quantity of material released. Exposure to these materials can cause respiratory discomfort, vomiting, nausea and eye irritation. It is not expected that exposure would cause any long term affects. Six Union Carbide employees were hospitilized and are under 24 hour observation as a precautionary measure. The Institute plant physician contacted area hospital emergency rooms within 40 minutes of the inc ident to indicate appropriate treatment for exposure . The plant notif fed the County Of f ice of Emergency Services of the incident within approximately five minutes and recommender initiation, of the f first stages of the emergency response plan .
44 UNION CARBIDE PRESS CONFERENCE STATEMENT AUGUST 12, 1985 August 12, 1985 There are three specific areas I want to cover. APPENDIX D First, the situation with the 28 people who were hospitalized, second the status of our investigation and the fact that no MIC was involved, and third the misconception that the Institute Plant did not respond in quick, timely fashion to this incident. fifteen of the 28 people hospitalized yesterday were released earlier today. More are expected to be released later today or tomorrow. All 13 who remain hospitalized are in satisfactory condition. We have no reason to believe that there pi 11 be any long term health affects . In al 1, 134 people were seen at hospital emergency rooms, according to hospital officials who have spoken with our plant physician. Union Carbide Corporation will assume responsibility for medical expenses for al l 134 people . Second, a team has been formed to fu l ly investigate the leak of material f ram a plant storage tank early Sunday morning . Representatives of several regulatory agencies such as the West Virginia Air Pollution Control Commission, EPA, Department of Natural Resources and OSHA are in the plant. Further inspection of the tank Monday conf i Ems ear! ier reports that there were about 500 gal ions of aid icarb ox ime/d ichioromethane mixture in the tank . It i s now suspected the material overheated when steam entered a jacket on the storage tank resulting in a pressure but ld-up. This caused three gaskets on the tank to fail. In addition, a safety valve on the tank opened and discharged material into an emergency vent system. Further investigation showed that a rupture disc in the emergency system also opened discharging material to the atmosphere . We re-emphas ize there was no MIC re leased in the inc ident . The third thing I want to talk about is Union Carbide's response to this emergency situation. ~ want to emphasize that the Institute Plant followed the emergency response procedures devised by the Institute - Jest Ounbar^-Pinewood Community Sub-Area Planning Council and adopted by the Kanawha Val ley Emergency Planning Counci 1 (KVEPC) . We were efficient in our reaction. We were timely with our response:
45 PRESS STATEMENT Page Two Here's the chronology of events: At 9:24 a.m. there was a release. Within 60 seconds, the plant alarm was activated and our emergency squad responded. At that time, we did not believe the emergency affected the community because the cloud was hovering over the plant. Later, our SAFER System indicated the cloud would move in a southwesterly direction. At 9: 44 a.m., we advised the Kanawha County Office of Emergency Services that there had been a re lease and that we were not able to, at that moment, identify the material. We also advised them at that time that we needed ambu lance as s i stance because contra 1 room operators had been overcome . Later, and, we be ~ ieve, before lO: DO a . m ., the County Off ice of Emergency Serv i ces act ivated its emergency s iren at Institute . At 9: 56 a. m, a Deputy Sheriff arrived at our plant and was escorted immed late ly to our P lent Emerger~cy Center where he remained throughout the eme rgency . By 10: 15 a.m., we determined that the release contained a mixture of aldicarb oxime, dichloromethane, carbon monoxide, carbon dioxide and su lfur compound s and our plant phy s ic fan had completed contact to al 1 area ho s p i ta 1 s to common i cate appropriate treatment f or e x pa su re . 0341r
Update statement 10 a 46 A Union Carbide spokesman stated today that of the 28 people hospitalized because of exposure to the gas release at Union Carbide's plant in Institute, W. Va. yesterday, two have been released this morning according to the hospitals. The others are in satisfactory condition without any signs of permanent injury The spokesman stated that the emergency response system worked according to plan and al ~ government authorities and hospital emergency rooms were notified in a time ly fashion . The spokesman stated that tile unit involved contained aldicarb oxime, not methy ~ i socyanate . An investigation of the cause of the incident is under way by Union Carbide, and Federal and state agencies pi ~ ~ also conduct their investigations . l
47 STATEMENT BY THEO EPPS * AUGUST 1 3, 19 8 5 - UNION CARBIDE INSTITUTE PLANT This morning I would like to cover the following points: hospital status of employees and res idents involved in the accidental release of aldicarb ox ime on Sunday, pug . 11; steps we are taking to help ensure a simi lar incident can be avoided in the future; clarify some aspects of the incident itself; and comments on the community emergency response system as we see it. As of this morning, 15 persons remain hospitalized, one of which is a Union Carbide employee . We expect that at least two additional poople pi ~ ~ be released today including our employee; the remainder are in satisfactory condition and pi ll remain in the hospital a bit longer under observation. APPENDIX E Whi le I be ~ ieve most of you are aware of this, I would ~ ike to repeat that the chemical methyl isocyanate was not involved in this incident. '9ldicarb oxime was. It is a product that is not made from MIC. It is later reacted with MIC to produce aid i carte . The initial production of aid icarb at Institute occurred in May 198S in a small existing unit. Aldicarb solution has been manufactured by Union Carbide at three locations - Woodbine, France, and Brazi 1, for the equivalent of over 20 years of safe plant operation . ~ new and larger aid icarb solution sy stem has been in startup at Institute . The incident occurred in the smal ler system which was in the process of being phased out. In view of the need for a complete understanding of the exact circumstances which led to the incident we have decided to temporarily suspend all aidicarb solution manufacture at Inst i tute . *Thad Epps is the Regional Director of Co~nuni~cy Relat ions for Union Carbide
48 Thi s morning, Warren M. Anderson, Chairman of Union Carbide Corporation, recommended to the Board of Directors and the Board has agreed that a special committee on Safety, Health and Environmental Affairs established in January ·1905 and chaired by Russell Train, will conduct an independent investigation of the incident. It is expected that the investigation pi 11 involve outside consultants and an external peer review. Further, Union Carbide is committed to cooperating with all appropriate local, state and federal authorities. The temporary suspension of operations of the aldicarb units at Institute will affect the routine assignments of approximately 60 employees at Institute and another 240 employees at the company ' ~ Woodbine plant. To clarify some aspects regarding the incident on Sunday, let me explain our s tatement of Monday regard ing what we do know about the source of the emi s s ion: It i ~ now suspected that about 500 gal ions of a Id icarb oxime/dichloromethane mixture in a tank overheated when steam entered a jacket on the storage tank resulting in a pressure build-up. This caused three gaskets on the tank to fail. In addition, a safety valve on the tank opened and discharged material into an emergency vent system. Furtherinvestigation showed that a rupture disc in the emergency system also opened dischargning material to the atmosphere. We re-emphasize there was no MIC released in the incident. Regarding the toxicity of aidicarb oxime, there have been some reports suggesting that it is equal to methyl isocyanate. It is not. Union Carbide's toxicological information shows that MIC's toxicity, based upon LD50 acute tests on oral, dermal, and inhalation applications, is about ten times that of aidicarb oxime. It is true that we rate both as a number four (4) for Union Carbide's rating system of individual chemcial's potential health effects. The number four indicates that all due precaution must be taken. However, within that category there are varying degrees of risk, with aldicarb oxime being on the low side and MIC being on the high side.
49 f inal ly, I would like to discuss the concern raised by some that the amount of time that elapsed between the occurrence of the incident and the activation of the emergency response plan might have been too long. Under the circumstances of this particular incident, we believe the time lapse was understandable. Nonetheless, we believe that there is, indeed, amiguity between public and private authorities as to who make those decisions to activate the emergency response plan that results in evacuation, roadblocks and other responses. I would be remiss if I didn't pant out to everyone that many things were managed well during the incident Sunday . The County emergency directors' organization responded effectively; roadblocks were in place effectively; emergency treatment centers were quickly set up; and the hospital emergency rooms performed most admirably. The good points not withstanding, Union Carbide intends to include as part of its investigation a critical review of this process, for it has implications for the entire chemical industry, not just Union Carbide. ## 0342T
50 APPENDIX F I ,~d~eloEc~^ro~ecom~qarEQ~~c~r~s~~o~=r~~rec-=a,c~g~=cAc _~~eu=~ C~CtOki~ c663 ~ 1~ ~ 1 = ~ i ~ FOR RELEASE IMMEDIATELY CONTACT: Ed Van Den Ameele (203 ~ 794-6985 or Harvey Cobert (203 ~ 794-7027 or Tom Failla (203 ~ 794-6928 ~ E C E ~ ~ E D hUGi9 1305 T. O. EPIC REPORS ON INSTTTU1E INCIDENT WILL BE MADE . NEXT NEED, SAYS UNION CARBIDE CHAl`4N C~RLESION, NV, August 16 -- A Union Carbide report on the causes of the incident at the company 's Institute, West Virginia plant will be made next week, Union Carbide Chaiman liarren t4. Anderson said today at a press briefing here. He emphasized that the company is also cooperating fully with the Occupational, Sat ety and Health Administration; Environmental Protection Agency, and state and county officials in their investigate ens. While co~plimcoting the cooperation and team work of community and plant people during the incident, Mr. Anderson said the company 's response to emergencies ts being eve hated. New plant procedures mill be to err on the side of caution "d early warning. Afros now on, n he said, "we will pull the cord first, then apologize it it wasn ' t necessary . ~ tin Anderson reported that company officials will be holding meetings with community leaders. There has to be an understanding between the community and the company that we know what we 're doing, that we can handle overselves under all circumstances, and that there is a willingness and a desire and a need to work togethe r, ~ he said . 1985 P-1 -6~04-85-070 - MORE -
51 Page 2 of 2 "The chemi Cal industry has had a good track record here and the communities and companies have worked together effectively over a long period of time ,~ said Mr. Anderson. "Carbiders are a part of this community. They live here, they work here, they associate with their friends and neighbors, and they 're proud of what they do. I have every degree of confidence in their professionalism and their capabilities . ~ A separate and independent lovestigation of the incident in being conducted by the Union Carbide Board of Directors' health, safety and environmental affairs committee. - END -
52 APPENDIX G HYPOTHESES DEVELOPED DURING FOCUSED GROUP INTERVIEWS 1. Respondents trust the chemical companies A clear consensus of those interviewed had a favorable impression of the chemical companies and believed that the economic benefits of the industry far outweigh the environmental risks. 2. People agree the Valley needs the chemical companies Most respondents believe that the economy of the Kanawha Valley would "dies if the chemical companies left because they could no longer operate profitably. 3. Communities feel protected against risk Respondents felt reasonably protected from environmental risks because the federal government had forced companies to improve health and safety--even though they had strong concerns about EPAts ability to regulate effectively and fairly. 4. Individuals don't feel personally at risk Because people trust the chemical companies, depend on them for jobs, and feel protected by the government, they do not feel personally at risk from emissions or accidents. 5. Everyone feels the national media distorts news about accidents Most respondents believed that the media had blown such incidents as the August Union Carbide leak at Institute out of proportion and were angry at what they saw as bad publicity for the Valley. 6. Communities want more say Respondents said they want better communications between their communities and the chemical companies and want more say in governmental and company decisions which affect their communities. 7. No one wants to increase environmental risks Residents and activists both opposed relaxing environmental standards to create jobs. 8. People will consider both economic and environmental trade-offs Because people are concerned about both their jobs and their health, they are willing to make economic and environmental trade-offs. 9. The activists' views Even activists, who admitted difficulty talking about the economic benefits of the chemical industry, felt they had to discuss the economic costs of environmental improvements in order to have their arguments considered by other residents.