Third Party Damages of Offshore Pipeline

Risk assessment is established to assist authorities in determining the priority of maintenance using risk which integrates both safety and failure. An efficient pipeline risk assessment should be able to characterize and calculate the risk associated with the pipeline. Unfortunately, the calculation of risk requires knowledge about the probability of failure and the consequence of failure. Both of which are difficult to estimate and in practical, the system under analysis cannot be characterized exactly. Numerical or objective data are often inadequate, highly uncertain and sometimes not available to perform calculations. To deal with this kind of situation effectively and consistently, a rigorous method of quantifying uncertainty using provided data is needed as well as to update existing information when new knowledge and data become available.
In this paper, a probability analysis model of offshore pipeline failure due to third party damages is presented. The interaction between ship anchors, dropped objects and fishing gears are discussed. Bayesian networks model is proposed to determine the probability of third party damages to subsea pipelines. To generate the probabilities of different kind of nodes in a Bayesian network, a systematic probability approach is proposed with emphasis on eliciting the conditional probability tables with multi-parents. The UK PARLOC database and DNV reports were used for the work. The paper concluded that Bayesian Network is a superior technique for risk analysis of pipeline failure. It is envisaged that the proposed approach could serve as a basis for decision making of pipeline maintenance.