IE 419 Case Study 3: BP Oil Spill

George Monninger

James Mayeski

Siri Swayampu

Jassim Al Mulla

April 30, 2013

Executive Summary

On April 20, 2010, the disaster known as the BP oil spill occurred in the Gulf of Mexico, which became the largest ocean oil spills in history. This accident is responsible for catastrophic environmental contamination that still persists to this day and the deaths of 126 crew members. Events such as these simply cannot continue to happen, as the lives of employees and the environment are assets far too precious to lose. Altogether, nine incidents were found to be responsible for the occurrence on the BP oil drill rig. Using a variety of online resources and research and fault tree analysis, it was possible to determine the events that caused the disaster and the relationships between them. Using these relationships, a probability analysis was conducted to determine the likelihood of such events occurring, dictating that there is a 0.94% chance of an accident occurring on the BP oil drill rig.

Introduction

According to an article entitled “The eight failures that caused the Gulf oil spill” from the website database , it is believed that eight different events are responsible for the cause of the accident. These eight causes include poor cement quality, a failure to notice the problem, a misinterpretation of the pressure test, two valve failures, an overpowered separator, and no gas alarm or blow out preventer. In addition to these eight causes of failure, an additional cause was discovered and found to be the product of using sea water instead of standard drilling mud, which causes additional stress on the seal of the well. A number of safety solutions are available that can help prevent catastrophic accidents such as the BP oil spill. Through the process of research and analysis, several approaches can be taken in order to decrease the chance of an accident occurring in the future.

Objective

● Determine causes of the oil spill disaster.

● Conduct a fault tree analysis depicting the accident.

● Determine possible solutions to prevent such accidents.

● Conduct a cost/benefit analysis to determine what safety precautions are necessary at what cost.

Methods

Using a series of probabilistic models, it was possible to determine a number of viable solutions to prevent a catastrophe such as the BP oil spill. Through the process of research, it was possible to construct a fault tree diagram illustrating the root causes of the accident. Probabilities were assigned to each of the nine root causes based on likelihood of the event occurring. In reference to the fault tree diagram, it is more likely that an event such as a valve failure occurs versus structural or concrete failure.

A series of probability relationships were then conducted based on the root causes assigned to the fault tree. Probabilities were assigned to each of the nine events according to the likelihood of an occurrence. According to chapter 8 of , an “and” gate refers to the relationship between probabilities in which all inputs must occur before an output event to occur.With an “or” gate, any of the events related in the branch can occur, causing output. Using these relationships, it was possible to calculate the total probability of an accident occurring on the drill rig. The labels on the fault tree diagram indicate the events on the probability calculations table.

Calculations of the total probability were performed in three separate steps. First, the four lower combined probabilities of each “and” and “or” sequence were calculated. Next, the two upper “or” probability relationships were calculated.

After the probability that the event was calculated, a cost-benefit analysis was done. First, the cost of the accident was found. This cost was multiplied by the initial probability to come up with the base criticality of the event.

Several remedies to the root causes were determined. Their probabilities were found, and new values were calculated for the probability that the oil spill would happen. Cost-benefit analyses were done for the recalculated probabilities as well. From this, a

Results

Reasonable probabilities were assigned to each of the root causesin order to compute the probabilities of the systems of events occurring. The result of 0.94% is considerably high based, especially when compared to the accident probabilities found in various examples throughout the textbook, such as Example 8.6 on page 352. Figure 1 shows the fault-tree diagram, and Table 1 shows the assigned probabilities of each event, as well as the probability that the oil spill could have happened. Table 2 shows the base criticality of the oil spill, which was $396.2 million. Both the fault tree diagram and the cost/benefit analysis can be seen in the appendix.

Several fixes to the root causes were identified. Table 3 shows the proposed changes, the events that the changes affect, the new probabilities of the events, and the new values of the probability that the oil spill could have happened. Table 4 shows the cost-benefit analysis of the proposed fixes. According to the cost-benefit equation, the best proposed solution is to inspect the cement. With this solution, the probability of the oil spill drops to 0.16%, and the criticality drops to $67.5 million.

Discussion

The present case study uses concepts found in Chapter 82 of the text to evaluate the BP oil spill that took place on 20th April 2010. Oil spills are considered some of the most lethal accidents since their side-effects are often permanent and especially harmful to the environment, mankind and animal life. Of all the oil related accidents till date, BP oil spill stands out as the most prominent due to its severity of casualties. A statement4 by the then CEO of BP, Tony Hayward, stressed that this accident was caused by a series of faults rather than a single one. Taking in to account his statement, a fault tree diagram was drawn in order to assess the causes of the spill in a deductive manner.

Two main fault events were identified for the spill; namely, cement failure and gas explosion. Each of these fault events were further studied to pinpoint the basic events leading to the accident. The first fault event sealing failure had two basic events like the usage of sea water instead of drilling mud and dodgy cement. According to a report by the National Commission5, it is revealed that BP had been using seawater instead of drilling mud to seal the borehole. The reason behind this practice had been to cut costs. The term dodgy cement refers to the improper cement seal formulation around the borehole.

The second fault event, explosion, had three sub events such as worker error, system failure and safety failure. Worker error consisted of basic events like the crew failing to identify the leak that occurred fifty minutes before the blast, misinterpreting a pressure test and deciding to divert the blowout material on rig. Once the leak started the crew onboard the rig had an option to divert the blowout material away from the rig via pipelines. However, they decided to lead the blowout material on to the rig via a separator instead. The separator is a device that is designed to separate gas trapped in the mud. The use of separator to process large quantities of blowout material overwhelmed it, causing flammable gas to surround the rig. The second sub event, System failures, consisted of two basic events pertaining to the malfunction of mechanical work pieces like failure of valve in the cement seal and valve in the blowout preventer. The third and final fault event was safety failure. This was comprised of the basic event relating to the crash in the rig’s gas alarm. The lack of battery power in the gas alarm was the reason for malfunction. All of the above events were depicted in the fault tree diagram, using a sequential order.

The next step was to identify the probability of each of these causes and possible solutions to prevent these events from recurring. The issue of sealing failures was found to occur with a probability of 0.16. This failure can be avoided by inspecting the cement mixture beforehand to ensure acceptable adhesion properties. This inspection is anticipated to cost BP $20,000 year. Also, the use of drilling mud for sealing should be mandatory, irrespective of the costs. Saving $100,000 dollars by not using drilling mud is not worth risking a possible catastrophic spill.

The worker or human errors such as failure to notice leak and misinterpretation of pressure occur constitute a probability of 0.2. These can be prevented by subjecting workers through a mandatory training workshop. This workshop will cost $10,000 and will be teaching the crew a comprehensive checklist of tasks such as reading pressure pumps, identifying flammable gas concentrations, and implementation of regulations mandated by the Environmental Protection Agency (EPA) and Mineral Management Service (MMS). Furthermore, an inspector with sufficient experience in the study of petroleum and natural gas should be hired to oversee the leak operations and supervision of crew. Without having an inspector onsite, the probability of accidents is 0.89.This additional safety measure would cost BP $100,000 a year.

To increase and ensure safety mechanisms onboard the rig, a complete revamp of safety systems should be executed. This revamp will include multiple methane detectors along the bore pipe along with duplicate valves installed on the cement sealant and Blowout Preventer. This revamping cost a hefty price of a million dollars per year. However, its execution will provide an extra layer of safety mechanisms to the rig, while eliminating the 0.85 probability of a spill. Finally BP should invest in buying and installing multiple gas alarms and Blowout preventers. This adjustment will cost $25,000 a year and has a 0.92 probability of preventing spills.

By taking in to account the individual probabilities of these basic events, the total probability of accident was calculated. The base criticality was then determined by multiplying the cost of the accident (42 Billion) and the accident probability (0.94). The criticality and benefit values were then calculated for each individual prevention measure, followed by the ratio of cost to benefit. The prevention measure with the least value of cost to benefit ratio was chosen to be implemented. In this case, it turned out to be cement inspection.

This result is justified, since carrying out this measure will stop the accident at the neck. Proper inspection of cement sealant will also avert miniature leaks in the ocean floor keeping the ecosystem and aquatic life safe. If there were no restriction on the budget, then all of the above mentioned safety measure should be executed. Oil spills leave a shadow of toxic and long lasting consequences in their wake. As fellow human beings, the decision makers at BP should follow the utmost preventive measures in order to preserve nature's delicate ecosystem. While not all accidents can be prevented, certain steps should be taken to eliminate as much risk as possible.

Conclusion

Analyzing the events leading to the B.P. oil spill made it possible to develop a system of probabilities and cost benefit analysis which can be used to decrease the chances for events such as these to occur. Altogether, nine events were believed to be the root cause of the accident, according to several internet sources. Because of these root causes, there was a 0.94% chance that the oil spill could have occurred. The criticality of the event was $396.2 million. By regularly inspecting the cement that made up part of the oil rig, the probability of the accident could have been reduced to 0.16%. Additionally, the critically would be reduced by $328.7 million to $67.5 million.

Appendix

Figure 1: Fault-Tree Diagram

Table 1: Initial Probability Calculation

Table 2: Initial Cost-Benefit Analysis

Table 3: Proposed Changes and Recalculated Probabilities

Table 4: Final Cost-Benefit Analysis

Works Cited[1][2]

  1. Durando, Jessica. "BP: 'Sequence of Failures' Caused Gulf Oil Spill." USA Today.

Gannett, 08 Sept.

2010. Web. 25 Apr. 2014.

  1. Freivalds, Andris, and Benjamin W. Niebel. "8 Workplace and Systems Safety." Niebel's Methods,

Standards, and Work Wesign. Boston: McGraw-Hill Higher Education, 2014. N. pag. Print.

  1. Mullins, Justin. "The Eight Failures That Caused the Gulf Oil Spill." New Scientist. N.p., 08 Sept.

2010. Web. 24 Apr. 2014.

  1. "BP Chief Tony Hayward's Statement in Full." Theguardian.com. Guardian News and Media, 17 June 2010. Web. 30 Apr. 2014
  1. "Macondo: The Gulf Oil Disaster, Chief Counsel's Report, 2011." Google Books. National Commission on BP Deepwater Horizon Oil Spill and Off-Shore Drilling, Jan. 2011. Web. 30 Apr. 2014.

[1]Make sure Works Cited is in MLA format

[2]Ex: alphabetical order, proper spacing. Use easy bib if nessesary