Bayesian corrosion assessment of unpiggable pipelines

对不可搬运的油气管道进行腐蚀评估是一项具有挑战性的任务，因为这类管道 检查工具无法访问。贝叶斯方法已被用来进行腐蚀评估 基于有限数量的现场调查数据的不可分流的管道。然而，这个问题的复杂性 所涉及的公式和计算技术极大地阻碍了这些方法的应用。 拟议的项目将使用贝叶斯网络(BN)来开发更新的方法和工具 基于有限数量的不可清管管道上腐蚀缺陷的深度、长度、密度和增长 来自类似邻近管道的现场调查数据或检查数据。研究的重点是确保 所开发的方法和工具在理论上是可靠的，在实践中也是可行的 申请。简单的可视化和轻松地结合观察数据和专家判断使 BN特别适合于拟议的研究。这项研究将与 行业合作伙伴，位于新斯科舍省哈利法克斯的劳埃德注册中心，将培训一名博士后研究员一段时间 六个月了。研究成果将为无公害油气的完整性管理提供参考 并提高庞大石油和天然气管网的安全性、可靠性和经济可行性 在加拿大。

Corrosion assessment of unpiggable oil and gas pipelines is a challenging task because such pipelines are inaccessible to inspection tools. Bayesian methodologies have been used to carry out corrosion assessment of unpiggable pipelines based on limited number of field survey data. However, the complexity of the formulation and computational techniques involved greatly hinders the application of these methodologies. The proposed project will employ the Bayesian network (BN) to develop methodologies and tools for updating the depth, length, density and growth of corrosion defects on unpiggable pipelines based on limited number of field survey data or inspection data from similar neighbouring pipelines. The focus of the research is to ensure that the developed methodologies and tools are theoretically sound as well as amenable for practical application. The simple visualization and ease to incorporate both observed data and expert judgements make BN particularly suited for the proposed research. The research will be carried out in collaboration with the industry partner, Lloyd's Register in Halifax, Nova Scotia, and will train one postdoctoral fellow for a duration of six months. The research outcome will facilitate the integrity management of unpiggable oil and gas pipelines and improve the safety, reliability, and economical viability of the vast oil and gas pipeline networks in Canada.