An international oil company developed a deep water exploratory drilling program scheduled to start in 2011. The company’s safety department required a financial risk analysis of the drilling program to aid in establishing insurance coverage levels. The purpose was to establish a probability cost distribution associated with a lost well control (LWC) incident. A numerical probabilistic simulation approach was deemed necessary since historical LWC cost data were insufficient for statistical analysis.
A ten branch fault tree with up to 38 elements (events) per branch was selected to represent the range of damage conditions that could follow a LWC incident. Probability density functions representing each event’s cost range were selected using a best fit procedure. The simulation procedure randomly selected a tree branch and then passed through the branch selecting costs randomly from each event’s cost-probability distribution. Event costs were summed to provide a total LWC cost for that simulation run. The simulation was repeated multiple times providing a suite of LWC total cost values. These data were then evaluated statistically to establish a LWC cost probability distribution. These results were used by the company’s risk division as an aid in determining the level of insurance to acquire.
