Reliability Assessment of Pipeline Third Party Damage

Abstract

Pipeline failure statistics indicate that mechanical damage caused by third-party excavation represents the largest threat to the integrity of onshore oil and gas pipelines in North America. In 1999, PRCI developed a reliability model that quantifies the pipeline probability of failure due to the 3rd party damage threat. The model employs a fault tree approach comprised of four main elements: the probability of excavation occurring on the pipeline alignment, the effectiveness of damage preventive measures, the probability that the excavation depth exceeds the depth of cover, and the probability that the excavator force is sufficient to fully penetrate the pipe wall. The PRCI model has been implemented by numerous operating companies over the past two decades. Despite this large contribution, there has been a gap in quantitative assessment techniques regarding the effectiveness of the methods used to prevent mechanical damage, and the pipelines resistance to the impact loads applied to pipelines by excavation equipment. In 2020 Enbridge applied this model to its 25,000+ km liquid pipeline system. During implementation numerous learnings and areas for improvement were identified. Correspondingly, the model was expanded to improve consideration of four important 3rd party damage threats that are not currently included within the model: agricultural activity, vehicle crossings, pipeline exposures, and mitigation activities. The results of this updated model showed that the probability of failure’s due to 3rd party damage were generally increased at locations with high population density, agricultural land use, and road crossings, that exhibited shallow cover. It is expected that this updated model will assist in prioritizing the mitigation of various locations that are potentially susceptible to the 3rd party damage threat in alignment with operator expectations. This paper discusses the data gathering steps required for implementation, example probability of failure results, and provides the details of the model updates which may be incorporated by other operators.