Algorithms for Diagnostics of Fault Conditions of the Sucker Rod Pumping Units

T. A. Aliev; O. G. Nusratov; G. A. Guluyev; As. H. Rzayev; F. H. Pashayev; M. H. Rezvan; A. B. Kerimov

doi:10.17587/mau.16.314-320

Algorithms for Diagnostics of Fault Conditions of the Sucker Rod Pumping Units

T. A. Aliev, O. G. Nusratov, G. A. Guluyev, As. H. Rzayev, F. H. Pashayev, M. H. Rezvan, A. B. Kerimov

https://doi.org/10.17587/mau.16.314-320

Full Text:

PDF (Rus)

Generate QR code

Abstract

The dynamometer card-based methods for diagnostics of the underground equipment for sucker rod pumping units have the following aspects: diagnostics based directly on the characteristics of a ground dynamometer card; diagnostics based on the secondary characteristics of a dynamometer card (spectral characteristics: variance, correlation and regression of the signal of the force sensor, Fourier coefficients, series expansion for a dynamometer card, etc.); and diagnostics based on the typical characteristics of the shape of a ground dynamometer card. Our studies and operation experience demonstrated that the above methods are not invariant with respect to the pump running depth and the changes in the pumping unit operating conditions. Therefore, development and introduction of new diagnostic methods based on new information technologies are necessary and relevant. The paper deals with the development of the algorithms for an automated diagnostics of the fault conditions of the sucker rod pumping units in the oil extraction industry. The authors propose algorithms for an automated identification of the fault conditions of the sucker rod pumping units with the use of the position-binary technology for analysis of the periodic signals of the force of the rod string on the hanger. The authors solved the problem of precise determination of the beginning and duration of the pumping unit period based on the data on the polished rod stroke with the use of a moving average algorithm. The following algorithms are presented: for retrieval of information about the force per a pumping period, the beginning of which coincides with the beginning of the polished rod stroke upwards from the common array of the data (force, stroke, consumed power) received from the well; normalization and scaling of the parameters of the array of force per a pumping period; the essence of the position-binary identification, which forces the signals' transformation into the position-binary components, in which the informative attributes are their duration in each position, which changes, depending on a signal form. Based on the position-binary technology, the program of the algorithm for an automated identification of the form of curves of the force on the polished rod of the sucker rod pumps was compiled in Borland DELPHI 7 object-oriented programming language and included in Ayna, general control, diagnostics and robust management complex, introduced at the oil field 3 of Shirvan Operating Company Ltd. in Azerbaijan. The authors provide an example of comparison of the position-binary components of the current signal received from the well with the known reference systems.