A Comprehensive Model for Oil Spill Liability Estimation

Abstract

Abstract The potential liability associated with oil spills from operations both offshore and onshore is enormous and in some cases can present an existential threat to a company. The cost of this liability varies widely, depending on situational, geographic, and jurisdictional factors. Companies consider these potential liabilities as they evaluate the risks of oil spills from various operations and apply that knowledge to risk management planning. To support such assessments, we developed a Bayesian network model that relates spill characteristics to specified Cost Elements. The oil spill liability estimation model (OSLEM) presented here uses two types of variables to estimate total liability—Cost Elements and Scaling Elements. Cost Elements include, at a minimum: (1) response and cleanup costs, (2) environmental damages and restoration costs, and (3) third-party claim costs (such as property damage, economic loss, and health effects). The values of the Cost Elements in the Bayesian network are predicted from user-specified values for the various Scaling Elements in the model. Scaling Elements include the location of the oil spill, the type of product, the type of operation, the size of the spill, the response actions employed, and the socio-economic and environmental sensitivities of the areas affected, among others. The output of the Bayesian OSLEM is a summation, or single estimate, of the most likely value from each Cost Element for a given scenario (location, spill size, etc.), as well as a characterization of the uncertainty of that estimate (i.e., range and likelihood of possible costs). The biggest challenge in parameterizing the model is acquiring data for the Cost Elements, which come from numerous sources. By modeling different operations, OSLEM can provide a comprehensive estimate of potential oil spill liabilities for a global portfolio. As a result of the modeling exercise, upstream management teams can identify a range of potential costs associated with an oil spill and deepen their understanding of the factors that affect each operation and, ultimately, the complex, multidimensional spill liability regime.