Probabilistic Erosion Assessment of Natural Gas Storage Wells

Abstract US operators are required to perform risk assessments for relevant threats on their underground gas storage wells as per 49 CFR 192.12 [1]. Sand brought up from the reservoir can present an erosion threat to the withdrawal piping attached to the wellhead (well laterals), where a failure would result in an immediate release to the atmosphere. This paper outlines a probabilistic model for assessing the erosion of well laterals that is part of a quantitative risk framework for SoCalGas’ underground storage sites in California. DNV published a recommended practice for managing sand production and erosion (DNVGL-RP-O501)[2], which includes empirical equations to determine the rate of wall thickness loss for piping components such as straight pipe, reducers, target tees, elbows, and production chokes. The inputs are the gas and sand properties, the gas flow rate, and the components’ size and orientation. The probabilistic model uses statistical distributions to account for the uncertainty in predicting erosion, specifically the variability in operating conditions and the amount of sand produced from wells. Using Monte Carlo simulation, the model assesses each lateral component’s failure probability and determines an overall time to failure for the lateral system while considering the potential for inspections and repairs. Erosion was modeled for various lateral configurations and operating conditions to demonstrate the potential range in the probability of failure. The probabilistic model quantified the probability of failure and reported an expected year of failure for the more susceptible components. Further, the probabilistic model was used to demonstrate how certain combinations of operating conditions and lateral design lead to potentially erosive environments and how adding a sacrificial erosion probe and/or scheduling inspections of the laterals can effectively mitigate the erosion probability of failure. The potential for high probabilities of failure demonstrates the benefit of well-specific probabilistic assessments to gauge erosion potential and inform decisions on mitigation actions.