Real-time risk analysis method for diagnosis and warning of offshore downhole drilling incident

Downhole drilling incidents in offshore oil and gas fields pose a threat to wellbore safety. Moreover, they are likely to develop into hazardous mishaps owing to delayed detection and false alarm. Real-time risk diagnosis and warning analysis are of substantial importance to gain time for reconstructing the wellbore pressure balance and thereby improve the well-control safety at an early stage. This paper presents a method for comprehensively analyzing the drilling parameters and data to estimate drilling downhole risk in real-time. The relative difference is introduced to capture dynamic characteristics for quantifying the trend evolution of the drilling data. The method involves models for determining the dynamic safety thresholds to diagnose downhole incidents. Further, it imparts the capability for preventing measurement errors and unapparent trends caused by ambiguous data. This approach is based on the Bayesian inference theory and incorporates the Gaussian rules to handle uncertainty issues and reduce false alarm rates. In contrast to previous works, the non-linear dependence of the drilling parameters is considered for the warning probability estimation. Case studies that were focused on lost-circulation incident for an offshore field well are used to illustrate the feasibility of the proposed method and to demonstrate that three alarm-levels based on risk propagation can be triggered rapidly for decision-making.