Risk-Based Data-Driven Optimization of Pipeline Inspection Interval

Abstract

Abstract In-line inspection (ILI) is a standard industry practice in managing the safety of buried transmission pipelines. It is important to develop a planning strategy for each ILI program to ensure pipeline safety while maintaining cost-efficiency. Defaulting to the maximum inspection interval allowed by regulations without analyzing time-dependent threats affecting a pipeline segment could lead to loss of containment. An optimal inspection interval serves as another barrier in the defense-in-depth safety concept. In this paper, a risk-based data-driven analysis approach is presented to optimize the inspection interval in terms of operational costs while managing the failure risks. The objective function is formulated to consist of cost components (in present value) and weighted residual risk of failure. The cost components are associated with the periodic inline inspections, as well as the expected mitigation actions. The annual mitigation cost is incurred by the number of features that are sentenced for excavation, field inspection, and repair every year according to the risk of failure. A shorter inspection interval would increase the total ILI cost; however, this would also curtail the rising cost of mitigation due to potential feature growth, as well as the residual risk of failure that depends on the mitigation program. The risk of failure is estimated in terms of the probability and consequence of failure. The residual risk is calculated based on features that are either unmitigated or potentially undetected by the in-line inspection tool run. To demonstrate the application of this approach, numerical examples based on external corrosion are provided to show how various conditions of a pipeline segment affect the optimality of inspection interval. The results support further application of the analysis approach in an industrial scale system-wide analysis, such that optimal inspection intervals can be dynamically updated at segment level using the latest ILI data as they become available in an ILI program.