Research On Composite Model Of Working Condition Diagnosis For Surface-Driving Progressive Cavity Pump Well

The application of progressive cavity pump lifting technique promotes the rapiddevelopment of the oil industry, but in recent years, the phenomenon of leakage of pump andsucker rod parting limit the further development of this lifting technology. In order tocorrectly identify the working conditions of progressive cavity pump well, guide theadjustment of production parameters and improve their management level, a new diagnosismethod of surface-driving progressive cavity pump well working condition is established inthis paper.Based on analysis and settling of progressive cavity pump well production data tested byfield, the type of working conditions can be determined and subdivided into10categoriessuch as the normal, pump leakage and analyze presentation and its reason of all kind ofworking conditions.8variables which can characterize the different type of workingconditions in progressive cavity pump well as characteristic parameters are selected toestablish artificial neural networks and support vector machines of surface-drivingprogressive cavity pump well working condition diagnosis method. The8variables includesproduction, dynamic liquid level, testing torque, testing axial force, sand production or not,wax containing or not, current fluctuation or not, tubing and casing is connected or not. WithActive X technology, using VB, Matlab and LibSVM mixed programming theory, theworking condition diagnosis software of surface-driving progressive cavity pump well isprogramed to calculate and analysis the Ji7block in Xinjiang oil field as an example.Through diagnosing and analyzing27progressive cavity pump wells of Ji7block inXinjiang oil field, the results show that the accuracy rate of the two methods is88.9%and96.3%respectively. Compared with the artificial neural network, support vector machinemethod has a higher accuracy rate because of its advantage of unique statistics. The softwarecan help field staff to discover potential failures and reduce economic losses due to the lyingwell, and improve management level of progressive cavity pump well.