Application of Artificial Intelligence to Deepwater Offshore Drilling: BOP Control Expert System

DNV has developed an expert system for condition monitoring, fault assessment, and automatic reconfiguration for advanced Blowout Preventer (BOP) control systems for deepwater offshore drilling in collaboration with an offshore drilling company. Expert knowledge of control system architecture, failure modes, and strategies for reconfiguration to respond to component failures was captured through modeling the system architecture and interviewing subsea engineers, drillers, maintenance personnel, and other subject matter experts. Response tree logic diagrams were used to capture knowledge of system architecture and alternative success paths for performing BOP functions. The advanced BOP control system considered for this analysis assumes a three-channel architecture with flexible routing capabilities that can tolerate failures in all three channels while continuing to operate within regulatory guidelines. To take full advantage of the increased fault-tolerance, algorithms were developed enabling continuous monitoring of component availability, selection of success paths for providing hydraulic fluid to the BOP rams to compensate for any combination of component failures, and automatic reconfiguration to implement the success path. A systematic information requirements analysis identified information and instrumentation required to support the assessments. The expert system also includes continuous assessment of compliance with Bureau of Safety and Environmental Enforcement (BSEE) requirements, simplifying communication with regulatory authorities during operation.