Dynamic risk analysis on offshore natural gas hydrate production test based on DBN-GO method

摘要

Preventing secondary hydrate formation caused by low temperature andthe high-pressure environment is one of the key factors for the successof offshore gas hydrate production test. In this paper, a dynamicBayesian network and GO (DBN-GO) model are used to build asuccessful GO method model for offshore gas hydrate productionexperiments and map it to the DBN model. The results show that thesuccess rate of preventing the secondary formation of gas hydrate in thetest of offshore gas hydrate production cannot reach 100% due to theinfluence of artificial operation and equipment degradation. Throughthe maintenance of the equipment, the success rate of preventing thesecondary formation of hydrate can be significantly improved.Compared with unmaintained equipment, it is more conducive to thesuccessful development of offshore gas hydrate. Regular maintenanceof equipment can significantly improve the success rate of preventionof hydrate secondary formation, the higher the maintenance frequency,the greater the success rate of prevention. Through the importanceanalysis, it is found that it is necessary to actively maintain the seabedpump in order to improve the success rate of preventing the secondaryformation of hydrate. Through sensitivity analysis, it is found that thefailure rate has the most obvious influence on the reliability andavailability of subsea pumps.