Maximizing Critical Infrastructure Performance with Probabilistic Methods

To enhance the level of understanding of threats to infrastructure and allow operators to predict asset performance, calculate system risk, and maximize safety, research on potential threats and development of probabilistic models is needed to ensure that the current body of knowledge is leveraged and transferred into a useful set of tools.

GTI is using advanced modeling and analysis tools to meet this need. A dedicated team is building, organizing, and maintaining applications that combine sound science with applied models to drive informed decision making. The focus is on better integration of field data and risk assessment methods to provide better results. GTI is working with key stakeholders in helping to support integration of these results into industry guidelines, standards, and regulations.

Results from the OTD-funded “Determination of the Leak-Rupture Boundary” project are providing operators with a better understanding of the manner in which a pipeline could fail (leak or rupture) based on material properties and defect size. This predictive capability, provided as a unique calculator tool, allows operators to model potential consequences and implement appropriate preventative and mitigative measures.

A modeling tool has been developed to provide pipeline operators with a better understanding of superimposed threat interactions. The underlying methodology has been completed and demonstrated as a deliverable of a project on understanding threat interactions. Further refinements will continue through early 2013.

Several projects addressing related topics are in various stages of proposal and execution. The individual projects enhance database structures and increase the understanding of failure mechanisms in pipelines.

One project—Probability of Failure Model for High Risk Pipe Segments—is targeted at identifying specific material microstructures and other attributes that are known to increase likelihood of failure for vintage plastics and cast iron. These material specific attributes will be integrated into GTI’s threat algebra methodology and used to improve probability of failure models.