1. IDENTIFICATION OF A WELL HYDRODYNAMIC MODEL WITH AN ELECTRIC SUBMERSIBLE PUMP USING THE PERTURBED OPERATION MODES CONTROL DATA
- Author
-
Ilya G. Solovyev, Denis A. Govorkov, and Vladimir R. Tsibulsky
- Subjects
well ,electric centrifugal pump ,complicating factors ,decomposition ,regularization ,assessment ,identification ,least squares method ,observers ,state filters ,test disturbances ,Engineering geology. Rock mechanics. Soil mechanics. Underground construction ,TA703-712 - Abstract
The relevance of the research is associated with the problems of creating and maintaining dynamic models of electrical submerged pump-equipped well as a part of operational monitoring and control systems under conditions of the elevator and inflow parameters deviation. Assessment of the root causes of well desired operating mode violation and the forecast of the immediate negative consequences are important for balanced (optimal) decisions on operational modes correction and ensuring stable operating conditions. The main aim of the research is to develop the least-squares algorithm – estimating the drift parameters of the elevator including the electrical submerged pump and the inflow of a dynamic well model and methods for stabilizing these estimates in real-time operations based on noisy measurements data. Methods: numerical integration of nonlinear differential equations, regression least-squares analysis methods, statistical computational tests, test perturbations, methods of regularizing estimation algorithms. Results. The authors have developed the two-stage algorithm for estimating the drift parameters of the elevator (wear, impurities) and inflow (productivity and reservoir parameters) to simplify the presentation of a complex dynamic model «inflow–elevator–electrical submerged pump–wellhead» according to two speed scales for monitoring the operating conditions of the system with inflow dynamics observer under conditions of test frequency disturbances. Summary. The proposed rules for stabilizing the model parameters estimates based on the generation of test perturbations, reducing dimension by decomposing the complex model into composite blocks, filtering noisy measurement signals by approximation, implementing a dynamic inflow observer form the basis for identification algorithm design with reliable estimation of the main group of deformation parameters of pump pressure-head characteristics and inflow. Despite the low accuracy of the elevator hydraulic resistance identification, the error in reproducing the dynamics of the variable states of the system in disturbed operating modes from the observer outputs does not exceed 2 %, which indicates the low sensitivity of the standard operating modes of the elevator to the desired parameter setting within the framework of the model solutions used. The required accuracy of the pipe hydraulic resistance identification should be consistent with the conditions for the periodic cleaning of the lift from sorbents.
- Published
- 2020
- Full Text
- View/download PDF