A quick tour of the major discoveries
Click on the graphics for Carbide documents related to the story below

Unlike Institute, West Virginia where effluent had to be cleaned to a standard sufficient to discharge to a river, the Bhopal factory's toxins would collect in open "solar ponds"

Groundwater contamination was clearly foreseen as a potential problem. The ponds would need impermeable linings.

"Institute has no solar pond" and consequently an environmental impact rating of 0, as opposed to Bhopal.

Napthol emissions from the solar ponds, combined with chloroform emissions from the undersized vent scrubber would give Bhopal a poorer rating of 1.5.

Carbide's Engineering Department warned of danger of groundwater contamination (July 21, 1972). The proposed design risked "danger of polluting subsurface water supplies in the Bhopal area". To avoid this "new ponds will have to be constructed at one to two-year intervals throughout the life of the project". They were not, of course.

One of the solar evaporation ponds was soon leaking. On 25 March 1982 a panicky telex was sent from India to Union Carbide in Danbury. Some days later, the Phase II pond was still leaking, causing great concern, and the emergency pond was also discovered to be leaking. Also in 1982, an audit carried out by US Carbide engineers found major problems with the plant and warned of the danger of a major toxic release. On 2 December 1984, their fears were realised.

In the post-disaster period the factory was shut down and it became necessary to clean up the hundreds of tons of highly toxic waste that lay in drums and in waste pits, plus the wastes in the solar evaporation ponds. Union Carbide India was dependent on UCC Danbury for advice on how to tackle the clean-up. On 15 and 27 May 1986, desperate sounding telexes asked what to do with 15 tons of highly dangerous chlorosulfonic acid sludge.

Four months later Union Carbide Danbury wrote (2 September 1986) referring to advice given by DuPont about the sludge, and saying they had no further ideas to offer.
Incredibly, two and a half years later, the dangerous sludge was still in the tanks, but by now (1 Oct 1987) it is described as "fuming" and of a mud-like appearance. For its dangers, see here. And here.
In case of spills of it is necessary to protect nearby people. Carbide never bothered to alert them.


In September 1988, Carbide in an internal memo (1, 2) admitted it had no idea at all what to do about the problem of Napthol and Sevin tars. They wanted to burn them (no thought given to dangers of toxic smoke to nearby communities).

Burying in a landfill was even more problematical. It would mean monitoring it for several years and by this time the company was looking for a quick cheap fix that would permit it to get rid of the contaminated plant and land.

A meeting on 27-28 June 1989 at Carbide's Charleston plant talked about rehabilitation of the Bhopal site. A chronology (1, 2) listed significant events at Bhopal over the years - problems with products and production - but, amazingly, made no mention at all of the fatal gas leak of 2 December 1984. The solutions proposed were cosmetic (and potentially hazardous to local people, whose welfare was not considered) including mixing with uncontaminated soil and sowing crops and brickmaking (!) (2, 3)

Another document wants to restore the land for use as a light industrial site. The solar evaporation ponds must be got into a state suitable for handing over to the Madhya Pradesh State government. The company is obsessed with the need to recover as much from the sale of assets as possible, including chemicals dredged from the ponds. Decide it will be politically better to form a subsidiary to do the work, thus also allowing Union Carbide Corporation to dissociate itself from Bhopal.


A document of 17 November 1989 reiterates the need to end the company's embarrassing Bhopal connection. It talks of meeting "local clean-up standards" which it knows to be inadequate and these may be preferred over US or WHO standards "if appropriate". If no standards are in place, a "risk-based assessment" is to be used. However, the methods proposed are all cheap and cheerful and can be carried out onsite. In view of the toxicity of chemicals known to exist in the ponds, all of these ideas are no more than a cheap fix designed to get Carbide out of Bhopal as quickly as possible.

At the instance of the Madhya Pradesh state government, Carbide appointed
NEERI (the National Environmental Engineering Research Institute) to make a scoping study and plan for reclaiming the site. NEERI had no relevant experience at all - indeed there was no-one in India with the necessary credentials - so Arthur D Little was appointed to oversee NEERI's work.

Arthur D Little believed itself to be working with UCC in Danbury, but all reference to UCC was to be deleted. ADL was to pretend to be working with UCIL alone.

NEERI's report was available to Carbide early in 1990. On May 15 at its AGM, Carbide's management was embarrassed by activists reporting the discovery of toxic chemicals, particularly lethal dicholorobenzene, in Bhopal's groundwater. The report was carried widely by Indian media.
This hysterical letter from Carbide to the state govenrment asserts that NEERI found no such contamination and asks the state government to take action "to avoid unnecessary agitation by interested parties". Earlier that year an internal Carbide memo accused environmentalists of being communists.


Norman Gaines,
head of Health, Safety & Environmental Research at Carbide's US HQ explicitly warns his colleagues on 22 May 1990 that the issue of contamination and clean-up must be made to appear an Indian-only affair, not in any way involving the parent company.

Gaines said that though the NEERI report "seems to implicity clear the plant" he recommended "caution quoting the NEERI data" as the methods used were unknown.

See document, pp 1 and 2.

Gaines then goes on to suggest a PR strategy
. Meanwhile it is clear that the company cannot go on evading the issue. A decision is taken to conduct an internal study into the contamination at the plant.

On 2 Nov, a plan drawn up by UCIL is sent to Norm Gaines.
It says: "During the monsoon the areas receive rainfall ... part of which percolates into the soil ... water percolating under the surface is a matter of great concern in view of the environmental hazard potential (of) organic contaminants".

To this day, during the monsoon, shallow lagoons and lakes form in low-lying ground and fill with water polluted by the abandoned factory wastes. Local children play in the water and animals drink there. People wash their clothes and cooking utensils with the water.

The company knew this happened because its officials on the ground could see it with their own eyes, but still they issued no warnings. Nor, despite their own misgivings, did they caution people about drinking water from nearby wells.


Carbide's preliminary study found major pollution of soil and water in the factory. Some water samples produced a 100% death rate among fish placed in them. The company was under no illusions about the seriousness of these findings and suggested that further work be carried out, the results of which which could be kept secret, or as the document puts it, "primarily for our own understanding of the situation".

In other words, issuing no warnings had become an official policy.

On 15 November 1991, a cheery memo from Carbide Hong Kong's Mike Buckingham to Norman Gaines refers to the latter's lack of concern about detection of mercury in groundwater "at less than 1ppb" (one part per billion). No action was therefore necessary. (The result: by 2002, mercury was present in the breast milk of women in local communities.)
Mike B is more worried about what the Indian government might do next and acknowledges the revived criminal charges against Carbide - which Dow still pretends do not exist.

On 6 April 1992, an internal Union Carbide memo acknowledged that the condition of the solar ponds was essentially the same as on the night of the gas disaster in 1984.

The company reiterates its desire to be shot of the plant, and return the land to the Madhya Pradesh government from which it had been leased. But Napthol is present in pond 1 and the contaminated plant is supposed to be cleaned up completely under the conditions of the lease before it can be returned.


Sure enough, Carbide opted for the cheapest method, burial. They began burying the contaminated soil from ponds 2 and 3 in a part of pond 3, and covering it with a thin liner. They also proposed to pump all the concentrated liquid waste from pond 1 into this burial, despite a warning from their own expert, who feared that the liner may split.

Survivors groups protested against this burial and warned of dangers to come. Carbide ignored them.

The site had to be certified clean enough to be handed over to the state government. What standards would the state government, which had shown itself to be totally clueless, accept? A meeting held in March 1995 spoke of US-EPA standards, but then pointed out that those standards did not apply to "a closed site like Bhopal".

Who could be found to certify that burial was a safe method? Why none other than old incompetent NEERI.

The meeting decided to proceed with burial of the toxic waste, despite the long-recognised fact that this would pose a real danger to groundwater. They would use NEERI to justify this choice. Arthur D Little would act as a rubberstamp, and if the state government did not like it, Carbide would wash their hands of the whole business. (The Sevin residues by the way are to this day lying in the site, in rotting drums, and in heaps in the open air outside the old Napthol plant.


Union Carbide India Limited had changed its name in 1994 to Eveready Industries (Eveready was a Carbide brand name), but the "poison papers" make it clear that Carbide in Danbury was still calling the shots.

NEERI's new report was ready in 1997. It was a masterpiece of incompetence. Samples had not been analysed for important metals like mercury. Dump materials were not analysed for hazardous volatiles. The list of criticisms ran to three pages. (1, 2, 3)

Arthur D Little too, had major problems with NEERI's report. Field logs were not provided, geological cross-sections and ground water maps were not included. NEERI got crucial things hopelessly wrong. Lindane, a hazardous pesticide, had been detected at levels from 20 to 420 higher than permitted for residential standards. "Therefore," ADL commented acidly, "these concentrations should not be considered as 'trace'."

Arthur Little's cricitisms ran to two pages. (1, 2)

In the most crucial area of all NEERI's report was seriously flawed. Arthur D Little warned that NEERI had grossly underestimated the speed with which the aquifer could be polluted, thus endangering the drinking water of local areas.

"Travel times could be significantly faster than assumed... one can argue that the worst case scenario travel time could be 2 years".

Two years later Greenpeace's study (PDF file) confirmed this prediction - the wells were being poisoned.


Arthur D Little agreed that NEERI's samples showed no contamination, but implied that this did not mean that there was no contamination to be found. NEERI had simply not found it. There were not enough samples, what there were had not been well enough documented and ADL added ominously "it is not known whether contamination migration will impact groundwater in the near future".

Arthur D Little advised against saying that the drinking water was safe. But Carbide continued to deny that there was a problem.

On 9 July 1998 the Madhya Pradesh government gullibly took back the land and plant from Carbide on the basis of the valueless NEERI report. Carbide had managed to con them.

But on finding that the derelict, rotting plant was still crammed with dangerous chemicals, that the land remained toxic and with fears mounting about groundwater, the state government finally lost its temper and demanded that Carbide, which had polluted the site in the first place, should now be compelled to pay for the clean-up.

But Carbide had slipped away.Their resident manager wrote: "The company ceased to be the occupier of the site on and from 9.7.98. The State Government as the rightful occupier of the premises and having full knowledge of the status of the site is expected to do whatever is required to be done in regard to the site... The company has no locus standi in the matter, and is neither in a position nor is required to be involved in the various activities which the State Government as occupier may think it fit to undertake by itself or through any of its agencies."