Bhopal Gas Leak: Why couldn’t the safety arrangements save the city?

Bhopal gas tragedy

Abstract

The 1984 Bhopal gas tragedy occurred at a factory where several safety systems were in place for handling hazardous chemicals. Refrigeration was meant to keep MIC cool, alarms warned of abnormal changes, and the scrubber and flare were meant to handle escaping gas. But having equipment installed and having it work effectively when needed are two different things. That night, some systems were unavailable, monitoring was weak, and no useful warning reached people living outside. Read together, the investigation findings make this accident look like a far-reaching systemic failure, not just the failure of a single machine.

The factory had safety systems, so how did the gas reach the city?

In the previous article, we explained how methyl isocyanate, or MIC, damages the eyes and lungs. Before that, we looked at the reaction in Tank 610. Now the question is about the systems that, somewhere in this chain of events, were supposed to stop the danger.

It is natural to ask whether a factory storing such a poisonous chemical had any safety measures at all. The answer is yes. But the next question is more important: which of those measures was actually in working order that night?

Later engineering reviews identified separate shortcomings in refrigeration, instruments, alarms, gas-treatment systems, and emergency response. Still, saying “every safety device failed” does not give an accurate picture. Some equipment did its limited job, while the systems that followed it failed to keep people safe. 

This distinction helps explain the story of the Bhopal gas leak. Reducing the pressure in a tank, controlling toxic gas, and warning the city were three separate responsibilities. Action in one area did not automatically fulfill the others.

Safety was not the name of a single machine

In the chemical industry, safety is built in several layers. The first goal is to prevent a hazardous situation from arising at all. If one does arise, it should be detected and stopped quickly. If a substance still escapes, its impact should be limited.

It is difficult to think of this as a single arrangement, like a lock on a house. Here, equipment, monitoring, employees, maintenance, and the preparedness of surrounding communities are all connected. If an alarm sounds on time, but it is unclear what to do afterward, the benefit of the warning is reduced.

Prevention and limiting damage are separate tasks

Keeping MIC at a safe temperature was related to reducing the initial risk. Detecting abnormal temperature or pressure offered an opportunity to intervene. If gas was released, handling it through treatment or controlled burning was the next responsibility.

AIChE’s analysis of Bhopal views the incident through this idea of layers of protection. Its central principle is that the layers of safety should be available when needed and should not all become ineffective because of a single weakness. 

The impact of a single decision can be felt in many places

Suppose the normal temperature of a chemical storage area is changed. Now, it is not only the question of cooling that changes. The alarm setting, gauge range, and operator response must also be reviewed again.

This is why it is dangerous to treat a change at a plant as a minor convenience or a separate maintenance decision. HSE guidance says that a formal safety review is also needed when changes are made to operating methods, procedures, and personnel, as well as equipment. The impact of a change can reach parts of the plant located far away.

Refrigeration: Why was the cooling system important?

MIC was not a substance that could be stored in large quantities and left without continuous monitoring. Its reactivity and volatility made storage conditions important. Refrigeration was part of keeping these conditions under control.

HSE considered the shutdown of the refrigeration system a change that contributed to the Bhopal disaster. According to HSE, without cooling, the tank temperature was above the specified design condition.

Cooling was linked to the rate of reaction

In the previous article, we saw that heat-generating reactions began after water entered Tank 610. As the temperature rises, the rate of such reactions can increase. So a low temperature was not merely a storage convenience; it also affected the response time available in an abnormal situation. 

But exaggeration should be avoided here. Claiming that if refrigeration had been running, contamination of any quantity would certainly have been contained would go beyond the available evidence. A more responsible conclusion is that removing cooling weakened an important protective layer.

Why do different reports give different dates for refrigeration?

Some details about refrigeration discuss the initial shutdowns and changes in operating practices, while others focus on later decommissioning or the availability of equipment. So turning one date into the whole story without context can cause confusion.

The central fact for this article is not the difference between the dates. Major technical assessments agree that, at the time of the accident, the system intended to keep the MIC cool was not in effective service.

It is important to understand this distinction: a unit stopping temporarily, remaining shut down for a long time, and continuing to operate on the assumption that its protection is available are not the same circumstances.

Monitoring: Why wasn’t the tank’s deteriorating condition understood in time?

The temperature and pressure indicators were meant to convey the conditions inside to the control room. But monitoring does not mean merely installing a meter. The reading must be reliable, unusual changes must be detected, and the response procedure must be clear.

The French government’s ARIA accident assessment documents shortcomings in Bhopal’s temperature, pressure, and storage-level indicators. The same account also mentions that high-temperature audible alarms had been disconnected

Why was the high-temperature alarm important?

If the temperature in the tank rose abnormally, the alarm would draw the employee’s attention to the change. That would be the moment to intervene. With the alarm removed, the operator would have to detect the same problem through another indicator.

IChemE’s incident summary considers the disconnection of the high-temperature alarm a critical factor. Later technical analysis links the refrigeration being switched off, the changed temperature conditions, and the unavailability of the alarm. icheme.org

This was not a minor administrative change. The device meant to signal an abnormal condition was not there to perform that role when needed.

The difference between a reading and a trend

According to the HSE, at around 11 p.m. on the night of December 2, the tank pressure appeared higher than normal, although it was not yet outside the working pressure range. Later, the pressure rose rapidly.

This illustrates a general principle of monitoring. It is not enough to check whether a number is within limits. How quickly it is changing, what other instruments are showing, and what the actual process conditions are—all of these matter too.

Today’s alarm-management guidance also emphasizes timely assessment, clear response procedures, and operator workload. The purpose of an alarm is not merely to make a sound, but to enable useful action in time. 

Gas scrubber: What could the gas-cleaning device do?

Hearing the name “vent gas scrubber,” it is easy to imagine a filter that would stop every poisonous gas. In reality, its process was based on chemical treatment. The system in Bhopal used sodium hydroxide, or caustic solution.

Its purpose was to treat the harmful substance being released. But effectiveness depends on gas flow, the availability of chemicals, the condition of the equipment, and its design capacity. If a release far exceeds what it is designed for, the equipment’s mere presence is not enough.

What condition was the scrubber in that night?

HSE’s account mentions an attempt to start the scrubber during the emergency and says it was not in operational mode. Other technical accounts describe it as out of service or on standby. It would not be right to collapse these descriptions into a single term without distinction. The result, however, is clear: the scrubber could not effectively control that release. The gas reached the atmosphere.

Being operational and being adequate were separate questions

According to Ronald J. Willey’s AIChE analysis, the scrubber was sized for small releases and would have been overwhelmed by an MIC release of this magnitude. In other words, besides its immediate availability, its design capacity was also a concern.

This is where the claim that “if the scrubber had been operational, all the gas would have been cleaned” falls short. Determining how much an operating device might have reduced the damage requires engineering calculations and scenario assessment. It is easy to promise a definite safeguard using information learned later, but such a claim would not be appropriate in research-based writing. 

In Bhopal, the real concern was twofold—the equipment was not in the expected condition, and its capacity to handle a major accident was also limited.

Flare system: Why was the gas-burning route unavailable?

The purpose of a flare system is to handle certain gases by burning them under controlled conditions. A tall tower alone is not enough. The gas must reach it, the connecting system must be available, and the necessary conditions for combustion must be met—only then will it do its job.

Technical analysis related to Bhopal records that the flare was out of service. According to Willey’s account, the system was unavailable while awaiting the replacement of a section of corroded pipeline.

Maintenance is not inherently a mistake. A damaged pipeline must be repaired. The question is how the associated hazards were managed during that time. When a large stock of a hazardous chemical is present, it is essential to check whether the remaining arrangements are adequate after critical protection has been removed.

Would the flare alone have saved the city?

No definite claim can be made. Given the flow, composition, and temperature of the massive release, the flare’s actual capacity would need to be assessed. The fact that it was unavailable must be distinguished from the claim that it could have handled the entire release under every possible circumstance.

Even so, its unavailability cannot be treated as a minor issue. A designated route for handling the toxic substance before it spread toward the city could not be used when needed.

What did the relief valve do? 

An important fact emerges here. According to Willey’s analysis, the rupture disc and relief valve responded to the pressure. But because the flare was unavailable, the released material was not safely destroyed.

So “all the equipment sat idle without doing anything” is not an accurate description. Pressure relief protected the vessel, but the problem of toxic release remained afterward. Preventing a tank from rupturing was not the same as protecting nearby people from poisonous air.

Why couldn’t the water sprays stop the gas?

The purpose of the emergency water spray was to help reduce the effects of escaping vapours. But such a system is useful only when water reaches the right place, at sufficient pressure and with the necessary coverage.

Technical assessments describe the water spray in Bhopal as ineffective. Later engineering discussions also point out that the sprays did not effectively reach the gas plume.

It is useful to understand that firefighting equipment and toxic-gas mitigation do not have the same requirements. Water coming out of a hose is only the first step. Its performance must also be adequate in relation to the release’s height, direction and volume.

That is why emergency drills matter. Exercises should reveal whether the available equipment will work under demands like those of a real situation. Looking at the equipment inventory and assuming that the response is ready is a false sense of security. 

Where did the system for warning the city fall short?

Identifying a hazard inside a factory and giving an actionable warning to people outside are separate responsibilities. Simply hearing a siren is not enough for a family. They need to know in advance what the signal means and where to get the next update.

According to ARIA’s account, the external public alarm and the internal plant alarm had become separate. The public warning was not triggered in time. The assessment also identifies gaps in nearby residents’ awareness of chemical hazards and in evacuation planning.

The difference between sound and information

Imagine a siren sounding in a neighborhood at midnight. Without prior information, people might think it signals a fire, a machine malfunction, or a drill. A warning is useful only when the community knows what it means.

This example is not a documented incident involving a particular family; it explains the need for emergency communication. During an actual disaster, there is little time to make decisions. It is difficult to explain the chemical’s name and the full process for staying safe for the first time in that moment.

Getting out also requires clear directions

When reading about the Bhopal incident, questions like “Why did people run?” or “Why didn’t they stay put?” can easily arise. But giving advice in hindsight and making decisions that night without any information are different circumstances. 

In community planning, responsibilities among authorities, the plant and emergency services should be clear. Who would announce the decision to shelter or evacuate, how the message would reach people, and who would assist those seeking help—all of this should be practiced in advance. Later reviews of Bhopal document serious shortcomings in community alert and response.

Getting information to hospitals was also part of safety

Even after the gas had crossed the plant boundary, opportunities to reduce the harm had not entirely disappeared. Medical services needed useful information about the chemical involved, possible injuries and exposure.

According to HSE, during the initial response, hospitals did not have adequate information about the gas and its effects. The sudden arrival of large numbers of patients had already placed enormous pressure on emergency services.

This does not mean that doctors waited for a definitive chemical analysis before treating patients. It means that the necessary cooperation between bedside treatment and reliable industrial information was inadequate.

In its account, the company says it shared the toxicology information available to it with authorities and medical personnel. That is its position. Whether information was sent and whether it reached the right hospital in a useful form at the right time are separate questions. 

What had earlier inspections found?

A record of safety inspections and warnings existed before December 1984. In the May 1982 Operational Safety Survey, UCC’s team conducted interviews, reviewed procedures and carried out a physical inspection. The report praised some improved procedures and also documented major concerns. archive.sambhavnabhopal.org

Both parts should be kept in mind when reading this document. It neither declared the entire plant safe in every respect, nor did its summary announce an imminent, certain disaster. Concerns about the possibility of a serious incident were present.

Amnesty’s document-based review mentions concerns related to the possibility of a toxic release, maintenance of safety valves and instruments, and personnel turnover. 

What happened after the audit?

According to a 1985 trade-union fact-finding report, UCIL prepared an action plan and sent progress reports. In the same report, workers described temporary fixes to several problems and said that conditions later deteriorated again.

So “there was never any inspection” is wrong. “An inspection was done, so everything was fine” is also wrong. The significance of an audit lies in sustained action afterward—which problems were fixed, who verified the improvements, and whether the system continued to work later.

Was the whole story simply about cost-cutting?

Cost-cutting decisions are an important part of discussions about safety in Bhopal. But attributing every technical failure to the same economic motive without investigating it separately also oversimplifies the story.

IChemE’s root-cause summary lists shortcomings in maintenance, training, operational oversight, risk assessment and management of change. In other words, both the condition of the equipment and the decision-making process were important.

Management of change has a simple meaning—identifying new hazards before changing an arrangement. After shutting down cooling, changing a procedure or temporarily removing safety equipment, it is necessary to check again whether the plant is as safe as before.

Today’s HSE guidance links this review to equipment, methods and safety conditions. This is a current safety principle; it should not be treated as a direct substitute for every legal obligation in 1984. But it helps explain the technical shortcomings of historical decisions.

Questions that do not have certain answers

There is disagreement over the exact route by which water entered Tank 610. Union Carbide claims sabotage and says the safety systems were not designed to handle a reaction on this scale. This is the company’s position. HSE says the exact point of water entry is uncertain.

But disagreement over the initial contamination does not resolve questions about refrigeration, monitoring, gas treatment and emergency planning. The cause of an incident starting and the reason it spread to the city can be investigated separately.

Similarly, there is no exact answer to how many lives would have been saved if a particular piece of equipment had been operating. What is clear is that several protective measures were in a weakened state. Stating a definite outcome for every hypothetical scenario would go beyond the research. 

The final question: Who was safety for?

The most unsettling part of Bhopal’s safety story is that, despite the equipment being present, nearby families did not receive the protection they were supposed to. The necessary links between managing tank pressure, preventing a release, providing warnings, and preparing a medical response were weak.

The Bhopal gas tragedy forces us to look beyond a checklist of safety measures.

Refrigeration is installed—but is it running? An alarm is present—but will it sound in an emergency? A scrubber is available—but is its capacity sufficient? Will the siren sound—but do people know what it means?

These questions have to be answered not when equipment is purchased, but throughout its entire life cycle. On the night of Bhopal, one system was supposed to be backed up by another after a failure. In many places, that backup was unavailable. The people living outside the factory bore the consequences.

Conclusion: Why did the factory’s safety in Bhopal not reach the city?

On the night of December 2 and 3, 1984, when the pressure in tank 610 increased, the relief valve did its job. It let the pressure escape from the tank. After that, the responsibility was with the systems that were meant to handle the gas that came out. The flare was not available that night, and the scrubber could not control the gas leak. The gas crossed the factory boundary and reached residential areas.

People outside the factory did not receive a timely warning that would have helped them understand the danger and decide what to do. Meanwhile, patients began arriving at hospitals, while doctors had limited information about the gas and its effects. There were also serious gaps in the information and preparation the city needed to deal with the crisis that had begun inside the factory.

That night in Bhopal showed that the responsibility for a factory’s safety does not end at its walls. From identifying the early signs of danger in the tank to informing nearby people in time and getting the right information to hospitals, every step is part of that responsibility.

Disclaimer:

This article is based on government accident assessments, original audit records, and later engineering analyses. Sources differ on the exact operating status of some equipment, shutdown dates, and the outcomes of different possible scenarios. Historical findings, company claims, and general safety principles are presented in separate contexts. This is not personal guidance or an operational manual for chemical emergencies. 

Leave a Reply

Your email address will not be published. Required fields are marked *