ABSTRACT
This chapter presents the Bayesian Network (BN) for causes, mitigation measures employed to counter those causes, its interrelationships and post-event scenarios for loss of containment (LOC) of hydrocarbon pipeline. Pipelines carry crude oil and products such as hydrocarbon liquids and gas from production centers to consumer points. Huge pipeline networks exist in USA, Europe, UK and Canada. In fact, such pipeline networks are critical infrastructures and therefore need to designed, operated and maintained at the highest level of safety. NoisyOr distribution can be used when there are several possible causes for an event, any of which can cause the event by itself, but only with a certain probability. The BN pipeline model can be tuned to see the situation by conducting a hindsight review of the impact of parent nodes on other parameters. The node states have been given the values that are thought to be the most likely situation before the LOC of the pipeline.
