by Nargis Kumar
As fires rage in Baghjan, indiscriminately eating up everything that comes in their way, Oil India Limited (OIL) continues to obfuscate the technical and human reasons for the oil and gas blowout. Much remains to be known about exactly what errors, decisions and glitches led to this mishap. But what we do know is (a) the Government of India’s slapdash environmental clearances to OIL for drilling near a forest reserve and human settlements, and OIL’s violations of regulations even before it got clearances; and (b) that the blowout preventer (BOP), a device meant to prevent such accidents, had been removed for maintenance work.
Research done more than a decade ago by geo-scientist Prodip Saikia claimed that this was a condensate well, not a gas well, which alters the way drilling would be approached, as condensate is far more volatile. The administrative and bureaucratic labyrinths through which condensate was turned into gas (possibly to make clearances and technical processes easier) allow us to question the way categories of ‘oil’, ‘gas’, ‘condensate’, etc. are created and dismantled, and the motivations and epistemologies mobilised to construct these ontological categories.
Undoubtedly this is a case of slippages in safety precautions, adherence to regulations, and prior inspections on the part of the oil company, and negligent allowances on the part of the government, in addition to ignoring science. OIL has begun shrouding the details, and will portray this as a one-off incident which is a product of a mistake made by the technicians present at the site. By suspending those engineers it is signalling that responsibility for this mishap lies solely on their shoulders, scapegoating them and not taking any responsibility itself. Accidents, however, are intrinsic to the system and cannot be done away with, no matter what improvements the industry undergoes.
The Imagination and Impossibility of this Industry
Every engineer in the oil industry, who I have spoken to in the course of my research, has pointed out that the most common problem they face is that of technical equipment not working properly. There are constant troubles in the plant from valves, motors, heat exchangers, pipes, etc. “Instrument failure consists of things like tank valves not opening via automation and needing to be opened manually, but this is very critical as we are dealing with oil moving at a high speed and pressure. If the valve does not open, it will burst and oil will spill,” the engineers at a port that imports and exports oil said. Rakesh, a senior officer at a refinery informed me: “Corrosion of equipment, its coating with crude and impurities, valves not shutting properly, overflows and leakages are routine issues we face. The main hazard is the auto-ignition of oil at high temperatures. If it comes out, it will definitely catch fire, and it has happened. Lives have been lost.” Fugitive emissions – minor leakages and vapour release – happen everywhere all the time. Their smell covers all oil sites such as drill-wells, tanks, ports, refineries, distribution terminals or petrol pumps. These fugitives are ordinary hangers on, but can result in extraordinary events.
“When there is a breakdown in a machine, we have to decide whether the damage requires fixing and replacing (and if yes, then how soon), or whether to continue operations (and if yes, then at what level)”, Ashwin, in one of India’s leading refineries, said. This is a decision between productivity and safety. “Our decisions are about how to comply with guidelines and bring safety and efficiency together. Should I run the plant with this leakage so I don’t compromise efficiency, or is it unsafe and also inefficient? How much loss will it cause to run it versus to stop it? Should I fix it online or do I need to shut it for fixing it, or do I need to replace it?” Targets set by the company and government influence the decision, and can put safety at risk. “The management puts pressure to increase production. So we have to keep producing, even during problems. We try to solve them online, and then accidents happen,” Sukant, in another refinery, argued. “Problems arise due to overloading the system, because authorities demand more output. That throws up all kinds of issues. Minor accidents happen all the time,” Naveen explained, who also works in a refinery.
Another engineer gave the reason of the inevitable corrosion of pipes and vessels due to sulphur, which causes leakages and requires frequent maintenance. Vinay, at a port, claimed, “Sometimes all parameters are correct at the control panel, so we are unable to detect the problem in the equipment. The control panel doesn’t always reflect the reality of the plant, like a choked drain or the level of fluid in a vessel, for a variety of reasons that we don’t fully understand. Once we detect this discrepancy, till we are able to fix it, we work with calculated values, and not what the control room is showing us.” None of these problems are avoidable unless production is slowed down. Even that would slow the occurrence of these problems, but not eliminate them. Since these problems are basic to the functioning of the oil industry, so are their consequences, aka accidents.
“Even when our equipment is top quality, problems arise as demands on this industry are such, processes and products are such, that to handle them we don’t have materials on earth. Nothing is fool-proof, leak-proof, corrosion-proof” said Abhishek, in another refinery. Fundamental to the oil industry, thus, is the sheer impossibility of what it crucially relies on for its everyday work: completely leak-proof and blast-proof materials simply do not exist, which engineers recognise, and yet continue depending on it. They say that it is beyond human and technical capacity to control oil and its effects after a point. Spills and accidents are not anomalies, they happen all the time. “They are bound to”, engineers casually remark. This is not just a function of human error, but a combination of the materials and the systems they have developed. The systems are so high pressure and high velocity, that after a point they are destined to break the materials. With time these systems have become so complex, that they have turned extra-terrestrial, not in the metaphorical sense signifying enormity, but literally: earth and its intelligent inhabitants do not provide the means to maintain these systems, and that has consequences for all of us: spills, fires, toxins, on a regular basis. When completely and sustainably leak-proof and blast-proof methods and materials simply do not exist on this planet, extra-terrestriality seems to be embedded in the oil industry, becoming a curious contradiction.
Whether the maze of pipes, cables and vessels at refineries, ships, jetties and pipes at ports, drill-wells at production sites, or the control panels with screens and buttons dominating one’s vision, oil sites inhabit a high level of complexity. Observing these sites makes clear that without this level of complexity in software and hardware, modern day fuels would not be possible. Without computerisation oil is impossible to extract and control at speeds, pressures and volumes that are currently the norm. This level of specification is not possible. This quantity of production is not possible. This refining quality is not possible. It is on this basis that a greater demand and higher quality is made feasible. As quantities and qualities spiral to greater heights, the system itself becomes impossible to maintain. Constant breakdowns then become a regular feature.
Admittedly, the system is built to cope with minor breakdowns. Not all of these are dangerous, but they can be, because of the nature of the material they are dealing with. Here is where the combination of several factors comes into play: the physicality of oil and gas, which is inherently combustible, the system which now functions at proportions and intensities that assume extra-terrestrial tools, and errors, which may or may not be individually situated. The system keeps running, but with constant strain and collapses, making accidents endemic.
Situating a Mistake
Industry officials often blame the staff’s negligence for mishaps, but given how frequent they are, despite all necessary precautions, there is more than negligence at work here. Ankit, in a refinery, explained, “When fatalities happen, everyone from the top to the bottom of the management is responsible. It is not the fault of just one person. It is the system that creates problems, because of which errors happen.” This reflects a broader understanding of their work as a network, in which all agents are interconnected and interdependent. If one agent makes a mistake, the system has led them to do so. Individuals are confined by a system that manoeuvres them. The human factor lies not in individual decisions as much as it does in the collective decisions they make about how their system will operate. Errors need to be seen as intrinsic to, spread over, and products of the system that these individuals are a part of. They rarely act as individuals making personal choices because there is hardly any room for that, given the many scripts they have to follow (made collectively by several people), and the high level of automation (also designed collectively by humans).
Engineers get daily plans and all action is divided into tasks and checklists. The smallest of tasks like taking a sample has a detailed Standard Operating Procedure (SOP). This highly systematised method is to avoid subjective interpretation in decision making, to keep things transparent, and to keep employees disciplined. But engineers assert, “Sometimes these plans don’t match the plant conditions, which change due to maintenance, repair and modifications. New problems also arise, which were not imagined by the plan. Weather conditions are not accounted for in the SOP either. And random unexplained changes and situations happen. This opens up possibilities for accidents.” This reveals that engineers acknowledge that it is not humanly possible to predict everything.
With gradually increasing automation, many human decisions are being taken by algorithms. With artificial intelligence, automation and more and more detailed SOPs, these decisions are being taken out of the hands of engineers and being made by machines, where engineers are expected to simply follow the script. The trade-off between safety, efficiency and productivity is not entirely subjectively decided, but also by scripts and algorithms. “In our decisions, factors we consider are safety, business risk, and resources available. A lot of the decisions and procedures are influenced by government stipulations, the company’s rules, as well as international norms that we are mandated to follow. Our decisions are constricted by many rules” an engineer said. Humans are forced to think like the system and align their priorities to the system’s. This system isn’t just the technical plant, but also the business corporation that they work for.
According to Ahmed, working in a distribution terminal, mistakes often happen due to overwork and fatigue. Ashok, Vipin and Daljeet in a refinery also had this opinion: “One person does the job of two-four people as we have less man power here. And we are underpaid. In the Bhopal gas tragedy, the main reason was low man power, overwork, fatigue. This is why mistakes happen. No one talks about that. Fatigue and overwork cause serious risks.”
Mistakes are inevitable also because the processes they develop to deal with oil can only be based on current knowledge, whereas oil and gas keep throwing up new problems and scenarios for them, under various circumstances. Ranjit, one of the many engineers I interacted with, remarked: “A lot of these things need not be seen as problems. They are challenges which we learn to solve. It is a constant process of learning and improving technology, systems, work styles, etc.” While this is normal to any area of work, as are mistakes, in this industry, this creates a huge cost. The safe functioning of this industry is premised on the total absence of errors, or at least that of big ones. While it has made enormous strides in controlling these errors by instituting numerous checks and balances and a variety of systems (both, administrative and technical), it has not, and cannot, completely eradicate them, thereby always hosting fatalities.
The Unpredictability of Oil
What happened at Baghjan was a ‘kick’ – a forceful rise of oil and gas through the drill-pipe – leading to a blowout. This enormous geological force is fought by meagre means comprising three barriers. The first is drill-mud, which is meant to fill the pipe that the oil and gas are climbing through, to block their rise. This is inherently fragile and cannot face the intense pressure of blowouts. It is suitable for seepages, but not large volumes at high pressures. The second barrier is the BOP which seals the pipe by introducing locks. This relies on spill-proof, unbreakable and fast-acting seals, which again are fragile in the face of the mammoth geological forces they are dealing with. They surely prevent smaller blowouts and their science cannot be dismissed, but high pressure, high velocity forces either rupture these seals or are too quick for them and climb up the pipes before the BOP has time to act. Examples of these breaches across the world show that such systems buckle under the pressures of subterranean forces. In Baghjan’s case, this barrier wasn’t present, in order to carry out maintenance work. The final barrier, in case the first two don’t work, is the Automatic Mode Function, which starts into motion when the technical system detects a blowout. This relies on automation, which depends on electricity, wires, computers, batteries, all of which come under threat and can be knocked out under the circumstances of a blowout. We don’t know whether OIL had installed this barrier in Baghjan.
These barriers are designed to act on predictable scenarios. But blowouts naturally happen in unpredictable ways, buckling pipes and damaging machines in ways that cannot be accounted for in these preventive systems. As Ranjit said, it is a constant process of learning and improving. As more blowouts happen, the industry gains more knowledge of the many ways in which it can happen, and improves its systems accordingly. But each blowout costs lives, if not human, then definitely nonhuman. It is acceptable and natural for things to not be perfect, such as the BOP. But is it acceptable for imperfect systems to go on operating when they come at huge costs? Nobody expects a perfect oil industry. But the imperfect one is severely damaging.
Removal of the BOP has been identified as the primary cause of the blowout in Baghjan. However, since 1970 blowouts have happened almost every year around the world. There are several scenarios possible under which the BOP fails. This is similar to the narrative that many engineers had: sometimes machines fail to engage, other checks also do not kick in, machines don’t respond to commands, or break down, there is an automation failure, etc. All these factors indicate that there could be many reasons for an accident, none of which are extrinsic to the system. Focusing on slipshod observance of precautions implies that it would be okay to drill anywhere had all safety rules been followed diligently. Focusing on callous environmental clearance procedures and violations implies that it would be okay had OIL been drilling elsewhere. But neither drilling in a desert, nor with all possible precautions can guarantee the prevention of a catastrophe. Nonhuman life comes under threat with the smallest of leaks, as does the air we breathe and the water we drink.
Against such an endemic presence of accidents in this industry, the question arises why the clearance procedures for it are so lenient. The answer lies in the relationship between oil and power: how the state is invested in oil production and consumption, how this is tied to economic growth, and how that fuels national development and state power. Furthermore, prioritising safety in this industry implies enormous investments and precautions, significantly reducing profits, which defies the very logic of this industry. When costs of equipment are so high, with what logic do we expect companies to replace machines quickly, or to prioritise safety over productivity? If the system was less intense (thereby exerting less pressure on machines), production would be less, and so would profits. So would, then, be the incentive to invest in better technology. Because of the logic of profit, safety will always be secondary, and will be invested in only up to a certain point. The internal logic of this industry, its relationship with the state, the geological forces it deals with, and its inherent contradictions (functioning at intensities by assuming the availability of materials that they innately defy), combine to expose the industry to extreme risk on the one hand, and provide it with impunity and unaccountability on the other, making the perfect recipe for disaster.
Playing with Fire
I found myself in a town-hall meeting on safety in one of India’s leading oil refineries. Seniors here observed, “Hazardous incidents have not decreased. We need to do something different in our technology, systems, and the precautions we take.” This is telling because it reveals the ubiquity of the ‘accident’ in this industry, and shows that it is not so accidental after all. Questions discussed in this meeting were: when should alarms be raised; what factors to consider in assessing risk; what protocols to follow; who should be authorised to allow work inside the plant; what should the thickness of a pipe be; working after dark; what all needs to be checked; and including the motto of safety, not just productivity, in every worker’s mind. This throws light on the many ways in which the decisions of engineers are conditioned and preordained by higher authorities, spreading human errors across the network. Statements like “No risk, no reward”, “No work can happen without taking risks. Travelling by air or even by car is also risky. So we have to assess the risk by our own judgement” were casually made. These statements clarify that the industry knows that safety breaches may be lurking just around the corner. Thus, accidents are not external to the industry, since risk is integral to it.
The industry demands of itself something that is not possible. Materials, obviously then fail the industry from time to time. Instead of recognising this, authorities continue to place the blame on individuals, portraying accidents as incidental. The oil industry plays with fire, literally, and with gargantuan geological forces, much larger than humans can control. The industry’s officials acknowledge this, as was displayed by an employee of OIL: “In the entire world, no drilling and exploration company has experts to control such a massive blowout. As far as I know, there are only three blowout management firms in the world and one of them is already present here.” Clearly, they recognise their powerlessness over occurrences like this, and make these statements as if it is unreasonable to expect them to have that control. And yet, it is reasonable for them to continue to do this work. If they are playing with fire, they should have the means to deal with it, instead of justifying their absence. And if their absence is justified, then how are they allowed to continue working?
Given the lack of care for ensuring safety and for heeding scientific research, this particular case may have been avoidable. But that does not extinguish the possibility of the next one. OIL will learn from this and improve its technical and administrative systems to prevent such fatalities in the future. It can make its pipes stronger, its reading of early signs of a blowout more astute, its employees better trained, etc. But that is all essentially based on current knowledge. There are plenty of new ways for oil and gas to blow out. How do humans imagine predicting and preparing for these infinite ways?
(Oil sites visited for this research have not been named in the interest of anonymity. Names of individuals cited have been changed.)