The Bhopal Disaster - Essay Example

? The Bhopal Disaster The tragedy at Bhopal occurred on the 3rd of December 1984, when a huge gray cloud of poisonous gases leaked out from Union Carbide’s pesticide plant which was located there. The tragedy was on an unprecedented scale, wherein more than 3780 persons were killed and more than 200,000 people suffered from latent injuries, which continued to manifest themselves over the years (Willey et al, 2006). The legal consequences of this disaster were such that the company paid out over $470 million as well as other considerations as damages. The technical reasons for the occurrence of the Bhopal disaster was the contamination of a storage tank of methyl isocyanare with large amounts of water, which caused a runaway chemical reaction that caused a huge explosion. In essence, an exothermic reaction took place when the water was released into the storage tank of methyl isocyanate(Willey et al, 2006). The temperature of the tank then flared to over 2000C and since the tank was not built to withstand such high pressures upon it, an explosion took place, releasing the cloud of poisonous gases over Bhopal. From a design perspective, the following reasons can be said to have contributed to the disaster, as set out below (Weir, 1987:41-44): (a) The tank design incorporated a vent gas scrubber for the purpose of neutralizing any methyl isocyanate, but this had been shut off. There was only one vent scrubber, whereas the size of the tank mandated four such scrubbers to handle the level of pressure generated during the disaster (Weir, 1987:41-42) (b) The tank design incorporated a flare tower, which was supposed to burn off the methyl isocyanate escaping from the scrubber, but this was also turned off because a corroded piece of pipe inside it had to be replaced. The sole flare tower was sorely inadequate to handle the level of pressure and the tank design should have included more flare towers. Moreover, the water curtain, which could have neutralized any remaining gas, was too short to reach the top of the flare tower. (c) The 30 ton capacity refrigeration unit had been turned off to save money’ had this been operational, the MIC could have been kept at a low temperature of 40C and would not have overheated. (d) No effective warning or alarm systems existed, and existing gauges that were used to measure temperature and pressure in different parts f the unit, were unreliable. (e) The MIC tank was filled beyond capacity, so much that the overflow tank was also full. (f) The malfunctioning blow down value of the MIC tank was malfunctioning because it was in an open position. Consequences of the disaster: As stated earlier, there were 3800 plus deaths and more than 200,000 injured. Apart from the loss of human life, crops, animals and marine life were destroyed. Many of the women developed diseases such as gonorrhoea and pelvic inflammation, excessive menstruation and suppressed lactation. There was a high incidence of still births and among the babies who where born, there was a high rate of babies born with deformities. Many foetuses suffered from respiratory ailments and complications, which also afflicted the mothers. Other symptoms which were produced included impaired audio and visual memory, neuroses and anxiety reactions. Reliability Block Diagram: The reliability of any system is derived in terms of the reliabilities of the individual components of the system. The advantages of using reliability block diagrams are as follows: (a) Integrating the probability of faults occurring into modules (b) Including within each module, the probability of failure, the failure rate, the distribution of time to failure, the steady state and instantaneous availability (c) The ability to organize each module in a structured way, dependent upon the effects of each module’s failure The reliability block diagram can be used to assess probability for failure for the Bhopal disaster as well. The major cause of the failure was the faulty valve which allowed water to pass into the MIC tank. This major fault was only exacerbated by other related factors, such as switched off air-conditioning, and malfunctioning pressure and temperature gauges, which prevented the faulty condition being detected on time. The reliability block diagram for the disaster would then look like this: = Blow-up of MIC tank From the diagram above, it may be noted that the probability of the air-conditioning being switched off will affect the probability of exacerbation of the faulty value and indirectly lead to the blowup; similarly, the probability of faulty pressure and temperature gauges will affect the potential for disaster arising out of undetected changes in pressure and temperature in the tank that could potentially be harmful. For instance if Lambda 1 is assumed to be the failure rate for the monitor and an exponential distribution is assumed for the failures, then Rair-conditioning-off(t) = e -l1• t (where l equals lambda). Similarly, R faulty-valve (t) = e –l2• t and R faulty-temperature-and-pressure-guages = e –l3• t and this can be the basis upon which the statistical probability for disaster developing can be calculated. Fault Tree Analysis: Fault trees have been defined as “a graphic model of the pathways within a system that can lead to a foreseeable, undesirable, loss event.” (www.fault-tree.net). Given below is the fault tree highlighting the major causes of the Bhopal tragedy, in which both first level and second level contributors are shown, duly linked through logic gates. FIRST LEVEL CONTRIBUTORS SECOND LEVEL CONTRIBUTORS Based upon the above diagram, it may be noted that the faulty valve was the primary cause, because it let water into the MIC tank. The realistic probability of faulty valves developing is quite high, i.e, up to 50%; however, the probability of it going undetected or untreated is much lower, i.e, 5%. The probability of faulty valves occurring side by side with other limitations such as the refrigeration units and warning systems being turned off is even lower, i.e, practically nil, especially in a chemical plant where normal safety protocol would mandate regular checks and updating/repair of faulty or malfunctioning devices. The probability of such an event occurring based upon the other first level contributor, i.e, faulty tank design through the use of only one vent scrubber and one flare tower, with the water curtain placed too low would be quite low, because design specifications would have required that adequate safety precautions be inbuilt in the tank. If S= successes and F = failures, then Reliability R is given by R = S/S+F, while the failure probability PF = F/S+F. As a result, reliability plus failure probability, i.e., R + PF would be given by (S/S+F) + F/S+F = 1. On this bases, Lambda or the Fault rate would be given by the formula 1/MTBF (www.fault-tree.net, pp 27). For most system elements, the fault rates would be constant over long periods, occurring randomly rather than regularly. In modelling fault probability, the dependent factor would be a function of the exposure interval, by the exponential. In the case of the probability of the faulty valve going undetected, assuming the interval is brief, i.e, 5%, T < 0.5 MTBF, PF would be approximated within 5% by Lambda T. In the diagram provided earlier, the gates are assigned letters – A, B, C and D. The boxes are numbered, i.e, the topmost event is 1, the two major events are 2 and 3, while the boxes in the last row are numbered 4, 5 and 6. The top event entries are A and B. These two form the top of the matrix, but A would be horizontally replaced by 2, while B would be horizontally replaced by 3. The inputs for these boxes would replace the values vertically, so that 2 would be replaced vertically by 4 and 5 because they are its inputs, while 3 would be replaced vertically by 6. Hence the final matrix would look like this, the minimal cut set as below: Emergency planning: The Bhopal disaster is a typical example of cost cutting measures that actually proved dangerous because they bypassed safety protocols. In the first instance, the design of the tank should have allowed for a much large scale of disaster by introducing more vent gas scrubbers and flare towers. Secondly, the scale of the disaster was caused by the combination of factors, i.e, switching off the airconditioning, carrying out repairs on corroded pipes and allowing the MIC tank to overflow. Leaving an untrained person on shift duty and the lack of effective warning systems all contributed to the disaster. Therefore, the major recommendation in this instance would be to beef up the security protocols and precautions; to ensure trained personnel are on duty and avoid cost cutting measures that leave the plant vulnerable to a repeat of the Bhopal disaster. Bibliography Fault Tree Analysis. Retrieved February 21, 2011 from: www.fault-tree.net/papers/clemens-fta-tutorial.pdf Reliability Block Diagram. Retrieved February 23, 2011 from: http://www.csupomona.edu/~myin/518Lectures/RDBS2009.pdf Weir, David, 1987. “The Bhopal Syndrome”, Random House, Inc. Willey, Ronald H, Hendershot, Dennis C and Berger, Scott, 2006. “The Accident in Bhopal: Observations 20 years later”; retrieved Febryuary 21, 2011 from: http://www.aiche.org/uploadedfiles/ccps/about/bhopal20yearslater.pdf Read More