Providing a comprehensive approach to oil well blowout risk assessment

Abstract

Drilling is one of the most dangerous activities in the oil and gas process industries. Therefore, a holistic approach was presented to prevent and control risks and reduce the uncertainty of blowouts, kick scenarios, and control layers based on the Fuzzy Bayesian Network (FBN). Four independent protection layers (IPLs) were used to evaluate the blowout outcome, and the failure probabilities of IPL1 and IPL2 were calculated with Fault Tree Analysis (FTA). Then, different events were transferred to the Bayesian Network (BN) in GeNIe software, and updated the probabilities. The Fuzzy Fault Tree (FFT) results reveal that the failure probabilities for IPL1 and IPL2 amount to 8.94×10−4 and 4.97×10−21, respectively. Also, the blowout probability results based on FFT were larger than FBN. According to FBN, the probability of the kick event was equal to 6.60×10−34. Sensitivity analysis showed that X1 (Mud volume/flow change) had the highest degree of importance in the blowout of oil wells. The results of this study can be used in both a preventive and reactive approach. Deductive and inductive reasoning, the dynamic nature and conditional dependencies, and causal relationships between events can make the model more realistic.