Root cause reasoning based on fuzzy fault Petri net for petroleum engineering systems

Abstract

In petroleum industry, engineering systems are becoming increasingly large-scale, high speed, automatic and intelligent. They are closely linked with the exploitation and transportation processes of oil and gas, making up an open complex system. Its failure often causes a chain reaction, which will further lead to serious or even catastrophic accidents with significant economic losses. Fault coupling and cascading are the “the real ultimate causes” of system breakdown, due to the complexity of its structure and behavior, as well as the high correlation and strong interaction among subsystems, system and environment. A new approach for root cause reasoning based on fault fuzzy Petri net is proposed considering the characteristics of cascading faults. The probability of each fault mode can be calculated and the fault propagation paths are ratiocinated. In this way, the proactive maintenance strategy or emergency plan can be made in advance to prevent accident. The effectiveness and accuracy of the proposed method is validated in the application of gearbox of the flue gas turbine used in refining system. The results reported in this paper could be useful for the petroleum engineering system designers and safety management personal.