Holistic Systems Analysis: A Case Study Demonstrating Simple Models Improving the Reliability of the BOP Control Equipment Ecosystem

摘要

Owing to their complexity and lack of system level data, subsea BOP control system design hasrnhistorically focused on the component level design and analysis. When system level design and analysisrnis considered, it is often performed at assumed steady state conditions and without a detailed knowledgernof end-user operating conditions. The interaction of multi-domain dynamically coupled systems can resultrnin unintended system level behavior, including field performance issues and system failure. The timelyrnresolution of these field performance issues is of critical concern as rig non-productive time (NPT) canrncost operators of floating rigs in excess of USD 15 thousand per hour [1]. The current approach tornresolving field performance issues typically involves replacement of the failed component or overengineeringrnto improve robustness at the expense of cost, weight, and efficiency. However, as the industryrnpushes for cost reduction and increased reliability, solutions must meet more than fit, form, and function;rna holistic approach to solving field performance issues must be developed to ensure original equipmentrnmanufacturers (OEMs) and operators collaborate together to treat not only the u0002symptomsu0002 of equipmentrnfailure, but also address the underlying system "disease". In this work, a case study is presented of arnlow-order mathematical model of a hydraulic control circuit and its utility as a heuristic for rapidly testingrnand optimizing solutions to address field performance failures resulting from pressure transients in a realrnworld hydraulic control circuit.rnA low-order model of hydraulic control circuit was created using a multi-domain lumped parameterrnapproach and replicated the pressure transients detailed in field performance reports, which causedrnrepeated catastrophic failures, e.g. burst inlet/outlet piping, cracked fittings, etc., of the pressure reducingrnvalve. The model was then subjected to a sensitivity analysis to determine dominating input variables, andrnthrough repeated iterations of potential solutions, converged to a high-confidence solution.rnThe high-confidence solution and technical justification from the model were presented to operators forrnfeedback and implementation, and the field verification results are detailed in this paper. The deep systemrnlevel understanding provided by the model aided in consensus building between OEM and operator.rnAdditionally, the results of the model yielded insight to system level communication failures at thernorganizational level between operator and OEM.