Study Of Fault Diagnosis Methods In Well Drilling Process

Well drilling is a complex project and because of the complexity and elusive of drilling stratum, the drilling process has a lot of ambiguity, uncertainty and randomness. Generally many complex situations even severe accidents resulted from the poor understanding of the objective situation or subjectively wrong decisions may occur. Therefore, it is significant important to safe operation by using each kind of information in well drilling process, discovering the unusual situations and preventing the occur of drilling accidents.In this paper, Fault Detection and Diagnosis methods are discussed in detail and Grey Relational Analysis and Time Sequence Analysis method adapted to well drilling are concluded. Analyzed characteristic parameter of dynamic signals and developed Mathematical Description for dynamic signals. Analyzed the changes of various parameters in well drilling process from the perspective of signal, and concluded changing characteristics of dynamic signal while drilling tools work abnormally. Based on the investigation of related studies in and abroad, the vibration reasons and failure mechanisms are analyzed and surface vibration testing device are designed using axial and angular accelerators sensors; and also discussed the method of detecting the mode of operation of drilling tools at real time using vibration signals and compound logging parameters and developed the associated diagnostic software.Based on the current compound logging system and drill string vibration measurement system, the On-Line Fault Detection and Diagnosis System using the correlative theory of Fault Detection and Diagnosis has a relatively high recognition precision through experiment test and can assist the operators to detect faults. The Fault Detection and Diagnosis system also has limitations that the surface sensors can not detect torque and vibration signals of mud motor.