Research On Experimental Prototype Development Of Deepwater BOP Stacks And Fault Diagnosis Methods

In the rapid development of petroleum industry today, deep sea oil has gradually become another treasure house. ―Hai Yang Shi You 981‖ platform has worked in the South China Sea Area. This indicates that China has formally started in deepwater oil drilling. Study on downhole tools and well control equipments is still in the initial stage in China. And we have made great achievements. But, the three companies in America is still in a monopoly position in the technology and the market. Deepwater BOP is the last barrier to guarantee the safety of deepwater drilling. The reliability and safety of the Deepwater BOP stacks is very important for the safety of national property and oil workers and the protection of envirment. Therefore, it has great significance for providing good technical support and test conditions to study the actual deepwater BOP stacks to develop a experimental prototype of deepwater BOP stacks. Furhermore, there are few researches on the fault diagnosis of deepwater BOP stacks. But it is important for improving the realiability of BOP stacks to develop its fault diagnosis methods and systems. And it can reduce the time of checking and repairing the deepwater BOP stacks. Therefore, it’s necessary to study the fault diagnosis method of deepwater BOP stacks. This paper studies the hydraulic control system of subsea BOP stacks, designs and builds a prototype of deepwater BOP stacks and studies the fault diagnosis method of the hydraulic control system of deepwater BOP stacks based on Bayesian networks.On the basis of analyzing the major components and fuctions of the existing underwater BOP stacks, the deepwater BOP stacks prototype is developed and its control principle is analyzed. This control system consists of high pressure circuit and low voltage circuit, which separately control the operations of BOP and the operations of the locking device. The control functions of the hydraulic control system are carried out by experimental studies. The result achieves the intended purpose of the design effect.Bayesian network approach and the mathematical model of network parameters are stuied, and the mathematical model based on Bayesian network and multi-sensor information fusion is also established. In order to avoid the problem of NP, a approach of establishing the diagnosis Bayesian network by the system function modules is proposed, which can reduce the difficulty and complexity of establishing Bayesian network and network parameters.The fault diagnosis model of deepwater BOP stacks experimental prototype is established. And the feasibility of the method of diagnosis is verified.