- Created: Friday, 06 May 2011 13:37
- Last Updated: Monday, 01 May 2017 08:55
- Published: Friday, 06 May 2011 13:37
- Written by Frank Pearce and Steve Tombs
- Hits: 5756
UCC, UCE, UCIL and the Disaster
UCC (Union Carbide Corporation) responded to the disaster by making a series of public, defensive, claims: that the disaster was totally unprecedented and unanticipated so it was not surprising that an “evacuation or safety plan had never been developed”; that they had not located the MIC plant at Bhopal “for reasons of economy or to avoid safety standards”; that they had the same safety standards in their American and overseas operations — “in India or Brazil or someplace else … same equipment, same design, same everything” (Everest 1986: 47 8). UCC claimed that its safety standards were identical to the standards at Institute, West Virginia; the plant had an excellent safety record; the plant’s Standard Operating Procedures (SOPs) - UCC’s responsibility - were basically sound. At the same time they tried to distance themselves from anything that be might found to be problematic at the Bhopal plant.
Hence, they also asserted: that the production of MIC [methyl isocyanate] in India, the siting of the plant and the quality of the materials used, were all the responsibility of UCIL and the Indian State; that exclusively Indian nationals managed the Bhopal plant; that UCIL (Union Carbide India Ltd) was an independent company responsible for its own affairs; that India’s cultural backwardness was responsible for the poor maintenance and management, poor planning procedures and the inadequate enforcement of safety regulations; and that the ‘accident’ was due to sabotage by an Indian national. Even now, in 2011, UCC ‘s representations of the causes, course and consequences of the disaster, its denial of responsibility and attempts to shift blame to Indians and India, and in particular an Indian saboteur remain the same – see the so-called Bhopal Information Center, established by UCC, at www.bhopal.com.
A Brief History of UCC, UCE and UCIL
It is useful at this point to provide some account of the relationship between the American parent and the Indian subsidiary. Its beginning can be traced back to 1905 when the National Carbon Company, which had been trading in the U.S. since 1888, decided to sell its products in India and in order to distribute them created there an Indian company, National Carbon, a wholly owned subsidiary. Over the next ninety years the parent company and the subsidiary had mutated - through name changes, buyouts, amalgamations and restructurings - into different forms of organization and company names, but always sustaining a similar asymmetrical relationship between the American parent company and its Indian subsidiary. There is no space here to go through the whole complex story, but an appropriate place and time to take up the narrative is India in 1956. At that time, the descendant of the American National Carbon Company, the massive conglomerate Union Carbon and Carbide Corporation (UCCC), founded in 1917, bowed to the provisions of India’s 1956 Company Act and accepted that its Indian subsidiary, the private company, National Carbon India Limited, should go public by issuing 800,000 shares (equivalent to 40% of its equity) available for purchase only by Indian investors. The shares were purchased by individual and institutional Indian investors leaving UCCC with 60% of the company’s equity.
In 1957 UCCC reorganized, refocused, and refounded itself as a multinational company, Union Carbide Corporation (UCC). Two years later, its Indian subsidiary was also refounded as Union Carbide India Ltd (UCIL). The latter became not only the major Indian manufacturer of batteries and allied goods, many of them sold as ‘Eveready’ products, but also, as the Carbide Chemicals Company (CHEMCO), a major manufacturer of chemicals. CHEMCO was given a major boost with the opening of the Bombay plant in 1961 and five years later it pioneered India’s petrochemical industry by installing its first Naphtha cracker there (D’Silva 2006).
Also in 1966, UCC created three new large companies, Union Carbide Europe, Union Carbide Pan America and Union Carbide Eastern (UCE), in order yet again to reorder and co-ordinate regionally its vast array of global interests. Each of these reported directly to UCC and was responsible for overseeing its interests and equity in the relevant region. Thus, UCE oversaw UCIL and the latter generally reported directly to UCE. In the same year, UCC created a global Agricultural Products Division and UCIL soon followed suit by creating its own Agricultural Products Division to which it assigned the Bhopal plan.
UCIL had been importing Carbaryl, formulating it and selling it under the brand name Sevin to Indian customers, when, in 1970, Edward Munoz, the General Manager of UCC’s global Agricultural Products Division, wrote to the GOI requesting permission “to erect facilities to manufacture up to 2,000 tons of MIC and 5,000 tons a year of methyl-isocyanate based pesticide” and to set up a “foreign collaboration agreement” (D’Silva 2006: 42). The estimated additional cost of installing and equipping the MIC/Sevin plant was $28 million plus. In its March 1972 response, the GOI approved in principle the annual manufacture of 5,000 tons of MIC-based pesticides but it also made clear that, when covering the cost of the expansion of the Chemical Facility, the company would be required to raise at least a quarter of this investment by new share offerings to Indian investors.
In November 1973 UCC/CE and UCIL signed a Design Transfer Agreement, Article IV of which noted that UCC was providing “the best manufacturing information available,” and a Technical Services Agreement. These would enable UCIL to build a Sevin plant at Bhopal. The design was loosely based on a plant in Institute and while it allowed carbaryl (Sevin) to be produced in small or large batches by using chemical reactions between Alpha-Naphthol (aka 1-Naphthol) and MIC, “for economic reasons”, both the Alpha-Naphthol and MIC would be produced onsite - in the case of the latter, in such large quantities that special, large storage tanks would be required. UCC/UCE and UCIL agreed that the estimated cost of the plant should be revised down by UCIL to $20 million.
In 1974 UCIL applied for a new licence, this time to produce up to 5,250 tons of Sevin in Bhopal annually. In 1975 the GOI agreed to provide an industrial (i.e. manufacturing) licence, but again included the condition that approximately a quarter of any monies raised for the MIC/Sevin plant had to be raised by a new UCIL share issue in India. It estimated that the plant’s total cost would be ‘approximately $26,500,000’ and it authorised the issue of 3,294,500 new shares to Indian residents to raise ‘Rs 527.12 lakhs, approximately 22. 3 per cent of capital financing’ [or $6.2m] which would bring “down UCC’s equity in UCIL to 50.9 per cent” (D’Silva 2006: 51). In order to accommodate the new Alpha-Naphthol, MIC, and Carbaryl plants and the attendant increased number of personnel, transportation etc. UCIL estimated that the Chemical facility would need to increase in size from 5 acres to more than 80 acres. In 1967 there were only 300,000 people in Bhopal; by 1975 there were nearly 500,000 and shanty towns were springing up near the railway terminal and the bus station and thus near the UCIL facility. Nevertheless, UCIL proposed that they should expand the size of their then current site by leasing more land in its immediate vicinity. They were aware that to stay on this site would violate the requirement of the 1975 Bhopal Development Plan, that such “obnoxious industries’ should be in a an industrial zone 25 kms away from the city centre, and that it also meant that they would ignore the order of the commissioner and director of town and country planning for the state that the manufacturing plant should be located well away from the city. UCIL’s response was to lobby the Madhya Pradesh and Federal authorities to overrule the Municipality; their lobbying efforts met with success since the state planning board classified the plant as “general industry” rather than “hazardous industry” in a 1976 review. This decision allowed both ongoing activities and the new construction to go ahead at the existing location (Varma and Varma 2005: 4).
After this decision was made, in 1977/1978, the GOI and UCC agreed that UCC’s equity would decrease from 60% to 50.9 % (D’Silva 2006: 37) which remained a majority ownership; new shares in UCIL were released for purchase by investors who already owned shares, be they individuals or Indian institutions, including different organs of the GOI. Construction of the enlarged plant began in 1979 under the supervision of the Bombay office of the British engineering company, Humphreys and Glasgow. Within a year the plant was completed and a year later it was producing MIC and the pesticide Sevin. Construction of the enlarged plant began in 1979 under the supervision of the Bombay office of the British engineering company, Humphreys and Glasgow. This was a significant increase in the scale and complexity of the chemical plants and should have been accompanied by a greater degree of planning, tighter organization, increased monitoring, regular and timely maintenance and an increase in security. In 1984 UCC still controlled 50.9% of UCIL’s equity.
Revisiting the Bhopal facility
Since 1976, workers and their unions had expressed concern about safety and health hazards at the Bhopal plant. These concerns became more pressing after the MIC and Sevin plants went into production. There were three serious leaks from these plants between 1981 and early 1982, killing one worker and hospitalizing a number of others.
In May 1982, a UCC safety team monitored the plant and found 61 hazards, 30 of them considered major, of which 11 were in the phosgene/MIC unit; areas of concern were described as “procedures training and enforcement together with attention to the equipment and mechanical deficiencies” (Everest 1987: 56). The recommendations of the report were not implemented: production simply carried on as dangerously as before. There were two more serious incidents that year.
Following two decades of huge growth, encouraged by government “subsidies, tax breaks, low-cost loans and lax safety regulations”, the pesticides market in India had, by the end of the 1970’s, become extremely competitive. Indeed, by the beginning of the 1980’s, “pesticide demand in India had collapsed” (Shrivastava 1993: 258-60). This collapse was in part structural, due to the influx of a great deal of agro-chemical capital, leading to intense competition. But it was exacerbated by more contingent factors: agricultural production in India declined severely in 1980, and only recovered mildly in the next three years; weather conditions in 1982 and 1983 caused many farmers to abandon temporarily their use of pesticides (Shrivastava 1992: 30 5). The Bhopal facility broke even in 1981, but after that began to lose money.
In May 1983 the two unions at the Bhopal plant were forced to sign a ‘memorandum of agreement’ with UCIL allowing cutbacks in staffing, and “eliminating such work practices which are not conducive to efficient working of the plant.” (URG 1985a). By 1984, the workforce at the Bhopal facility was roughly half its 1980 size. New operatives were less qualified and less well trained than before, and 150 operatives were taken from their trained details and used as floating labour (ICFTU/ICEF 1985). The number of supervisors was halved - meaning that they often had to manage more than one plant and had less specialist knowledge (Chouhan 1984: 12). The number of field maintenance staff was halved, and, as maintenance became more and more sporadic, many problems were undiagnosed, with defective parts neither repaired nor replaced The chemical plants began to function poorly, making safe, efficient production difficult. Instrumentation became increasingly unreliable, and safety and back up systems were increasingly in disrepair. Since SOPs were designed on the basis of fully functioning systems and adequate numbers of personnel, their absence forced both operators and supervisors to improvise solutions to problems , always then with the potential to create new problems.
Pesticide production: in theory and in practice
The Carbaryl process in use in Bhopal involved a number of processes: a) it began with coke and oxygen producing Carbon Monoxide; b) the Carbon Monoxide interacted with Chlorine to produce Phosgene; c) Purified Phosgene and Methylamine, with Chloroform as a solvent, were reacted together to form MIC; d) Naphtholene was sulfonated to produce Alpha-Naphthol with a toxic side product, Beta-Napthhol, the removal of which proved intractable, hence the process was abandoned and UCIL had to rely on imports from UCC; e) MIC reacted with Alpha-Naphthol to produce Carbaryl; f) Carbaryl was then formulated to produce Sevin. While only a small proportion of any batch of the MIC could be used immediately, the “economics” of this production process favoured its production in large batches. As a result, the design included 2 large 15,000-gallon tanks where the MIC could be stored for future use and a third 15,000-gallon tank, which, while usually used for the temporary storage of “impure” chemicals, could also be used as an overflow tank. Our focus here is primarily on the MIC, the MIC storage tanks, the Carbaryl/Sevin plants and their interconnections.
Before returning to events on the night of the disaster, further contextualising comments are useful. First, between 1980-84, the entire work crew in the MIC unit dropped from 13 to 6 and their training was cut back from 6 months to 15 days. Second, the UCC safety manual instructed operators to make sure that the two MIC storage tanks (E-610, E- 611) were no more than half full, and/or to keep one of the tanks empty at all times, and/or to transfer some of the MIC to the standby tank, E-619. The operators were instructed to keep each ‘tank’s temperature below 15 (degrees) C (59 degrees F) and preferably at about 0o C (32 o F)’, since MIC was significantly less reactive at that temperature and, in part for the same reason, it was necessary to maintain “an atmosphere of dry nitrogen, under slight pressure, [of approximately 2 pounds per square inch (psi)] in the vapor space of the storage tank” (cited in D’Silva 2006: 54-55). This pressure also helped exclude potential contaminants. Increasing the nitrogen pressure was integral to the movement of MIC from the tanks to the Carbaryl unit.
On the night of December 2nd/3rd 1984, there were startling discrepancies between what was mandated by the safety manual and by the SOPs and the actual condition of the MIC plant, the storage tanks, and the Carbaryl plant and their chemicals transferral systems (pipes, valves, and so on). First, Tank E-610 was not half empty but almost full with 42 tons of MIC, Tank E-611 contained 20 tons of MIC and even Tank E-619 was not empty but contained one ton of MIC. All three tanks and the transfer systems were compromised by extensive rust and corrosion. The transfer of the gas from Tank E-60 was made by difficult by the excessively low pressure of the Nitrogen and the fact that it had a defective valve (carbon steel valves were used at the factory, even though they corrode when exposed to acid). These deficiencies also made it dangerously easy for external contaminants to enter the tank, relating the potential for dangerous exothermic chemical reactions. The contents of these tanks had not been analysed since October 26th, the date the last batch of MIC had been produced and stored. According to the operators, the MIC tank pressure gauge had been malfunctioning for roughly a week. Other tanks were used rather than repairing the gauge. The build-up in temperature and pressure is believed to have affected the magnitude of the gas release. The readings of pressure and temperature for these tanks should have been taken every two hours, with any observed deviations being dealt with immediately; in fact, the frequency of readings had been changed from every 2 hours to every 8 hours. Some of the meters were not working properly.
Then, even though it was winter in Bhopal, the lowest expected ambient temperature was only 15o C. The tanks were not being cooled so the internal temperature was around 20o C. This amount of stored MIC was in violation of all UCC safety standards, as was storing it above 0o C. The reason for this high temperature was that after the last batch of MIC had been produced and stored on October 26th, the Works Manager, decided to “shut down the principal safety system... because the factory was no longer active, these systems were no longer needed” and “no accident could occur in a factory that was no longer operating” (Lapierre and Moro 2002: 213). Part of this shutdown included saving minimal costs by no longer refrigerating the MIC tanks - the refrigerant was later pumped out of the system and used elsewhere in the plant. The flare tower was also taken offline for maintenance and repairs, as was the decontamination scrubber. Shrivastava described the plant at this point thus:
"The working environment of the Bhopal plant tolerated negligence and a lack of safety consciousness among workers and managers. Employee morale was low because the plant was losing money and being considered for divestment." (Shrivistava 1987: 41)
Between 7-8 p.m. on the night of December 2nd, a relatively new and under- trained worker who had transferred MIC from the storage tanks into the carbaryl reactor earlier that day was given an additional task, one usually undertaken by maintenance staff: he was to use water to wash chemicals out of the pipes used for the earlier transfer, still attached at one end to the MIC storage tanks. He succeeded in flushing out the chemicals, but was unable to open all the stopcocks to get rid of all the water. Nor had he been told that a key safety measure was to seal pipes by inserting metal slip binds at each end. He reported to his supervisor that he could not rid the pipes of all the water, but the supervisor told him to continue flushing the pipes and keep the water running because the night shift would sort out the problem. The supervisor and the operator left when the night shift crew arrived. The latter also tried to empty the pipes, but failed and gave up. Water had now been running for at least four hours and some of it entered Tank E-610 unhindered by slip binds (not inserted), a valve (defective) or nitrogen pressure (too low to inhibit the movement of water), so starting an exothermic reaction which in turn produced others which ultimately produced the release of gas. UCC disputes the accuracy of the whole of this account; rather, it claims that water had been deliberately added to tank E-610.
Workers reported smelling MIC, but could not locate its source. The MIC tank alarms had not worked for four years. Thus the dangerous practice of the human nose being the technology used to detect MIC – though it is only at 50 times its LTV that its smell can be detected. Workers informed a supervisor of the leak, who was skeptical that it was MIC and postponed investigating its source until after a tea break. Eventually, as the smell of MIC grew ever stronger, they informed their production manager that large quantities of gas were escaping (he ordered them to activate the flare tower), and they also called in the fire service. As the production manager should have known but in any case was promptly reminded, the flare tower was out of commission - it had been dismantled because its pipes were corroded and had not been replaced.
Fire fighters tried to douse the gas but the water pressure available for the firemen’s hoses was insufficient for the water to reach the stack from which the gas was escaping. By then, Tank 610 was rumbling, concrete cracking, and its temperature was about 200 Celsius, the pressure at over 180 psi, 140 psi in excess of the tank’s rupture disk limit. Gases, vapours and liquids burst past the rupture disk, shot through the relief valve vent header, then by the vent gas scrubber, which was on standby and was not in working order, bypassed the inoperative flare tower and into the atmosphere. The operators could not dilute the MIC in tank E 610 since the tank was already overly full, while the emergency dump tank had a defective gauge indicating that it was also 22% full.
Clearly, badly designed and maintained equipment, lack of spare parts, inadequate SOPs and untrained staff all contributed to the incident. Important too were ad hoc modifications to the plant designs, such as a jumper line that may well have been the means by which water entered the MIC tank. An even greater problem was the lack of preparedness for such a disaster, amounting to absolute indifference to the safety of the public. UCC argued that most of these deficiencies were the fault of UCIL, its workers, and the Indian government and, of course, the saboteur.
Click on References for the extensive bibliography on the Bhopal disaster used in this article and the series to follow. The series makes use of ch. 6 of Frank Pearce and Steve Tombs' Toxic Capitalism: Corporate Crime in the Chemical Industry, 1998, Ashgate: Aldershot; paperback version, 1999, Canadian Scholars Press, Toronto. See also: Tombs, S. and Whyte, D. (2007) Safety Crimes. Cullompton: Willan. Both of these books can be bought through the CrimeTalk Shop.
Click on the title to return to Flowers at the altar of profit and power: the continuing disaster at Bhopal.
For part 3 of this series, click on Flowers at the altar of profit and power Part 3: Was the disaster at Bhopal "unforeseeable"?
For part 4, click on Flowers at the altar of profit and power Part 4: the Bhopal 'Settlement"
For Part 5, click on Bhopal: criminal, immoral or the cost of business as usual?
To go to the Pearce and Tombs column: Crimes of the Powerful and Insurgent Resistance
A FREE e-copy of Toxic Capitalism can be downloaded at Frank Pearce's website here.