Advanced Assessment of Pipeline Stress-Relief

Abstract Pipelines traversing landslides can be subjected to forces due to ground displacement. Stress-relief projects are usually planned before strain in the pipe, or a girth weld reaches one of the strain capacities associated with a limit state event using a certain safety margin. Accurate estimation of the pipe strain growth due to the slide movement is a key input for planning stress-relief projects. Through presenting a case study, this paper introduces the concept of an advanced stress-relief assessment approach. This method uses finite element simulation of pipe-soil interaction, to estimate the evolution of strain before stress-relief, pipe rebound as a result of stress-relief, and extension of the asset life after the backfill is completed. Using this modeling, the effects of additional improvements that can be added to a stress-relief project can also be estimated properly. These measures can enhance the performance of the pipe by increasing its strain capacity or reducing the impact of the slide movement. The primary benefit of completing such assessment is to be able to optimize the timing of stress-relief projects. Given that the strain growth rate usually accelerates as more strain accumulates in the pipe, timely stress-relief can prevent excessive plastic deformations and consequently increase the benefit of stress-relief projects. This is especially important for lines, for which pipe replacement is much more expensive than a simple stress-relief (without cutouts). The next step in this assessment is to use the output of this model for different scenarios of stress-relief in a risk assessment. This risk assessment provided an opportunity to compare the stress-relief option with horizontal directional drilling with a completely different cost/performance combination. The result of this assessment showed that if planned and executed as recommended, the stress-relief option can result in risk reduction below the risk target with significantly lower cost and complexity.