Risk assessment model of sustained casing pressure of gas well based on Bayesian network and its application analysis

Sustained casing pressure gets serious increasingly in the process of gas well production, which increases the risks of well integrity failure. To solve this problem, this paper assessed and studied the risks of sustained casing pressure. The early warning model of sustained casing pressure based on Bayesian network and actual production data was established. By virtue of the model, the influential factors of sustained casing pressure were divided into two units, i.e., tubing/casing leakage and cement sheath failure, and the main risk factors and risk failure probability of each unit were determined. In addition, the quantitative risk assessment indexes of sustained casing pressure were set and the risks were classified correspondingly. Thus, the risk assessment method of sustained casing pressure of gas well is developed. It was applied practically based on the statistical failure probability of 191 wells in No.1 Gas Production Plant of Xinjiang Oilfield Company. The case application indicates that this model can quantitatively calculate the potential risk of sustained casing pressure, deduce the reverse origins of main risk factors and provide decision making basis for sustained casing pressure prevention and control, which is conductive to reduce the occurrence probability of sustained casing pressure of gas well.