Известия Томского политехнического университета: Инжиниринг георесурсов (Sep 2022)
ESTIMATION OF ESP OPERATIONAL RESOURCE DYNAMICS USING FACTOR-BASED MODEL
Abstract
The relevance of the study is related to the search of the approaches to increase the control efficiency for the operation processes of wells equipped with electric submersible pumps. Dynamics of pump operational resource usage can be estimated on the basis of factorial model construction and identification according to control data of the power unit during its life run to failure. Unlike the models based on the fundamental physics of insulation aging and mechanical wear, the introduced simplified description of the resource development dynamics reflects the relationship between the operating lifetime and the well-known factors that reduce operational reliability in general for the entire submersible unit. The main scope: formalization of the electric submersible pumps continuous operation resource model in dynamics and formalization of resource estimation and submersible unit accelerated wear factors; development of the parametric estimation algorithm of the resource model operational block with control over the calculation correctness according to the operating modes data information content of a selected group of wells. Object: electric submersible pumps-equipped wells with continuous operating modes. Methods: thermobaric modeling of variable states of well systems and regression estimation of the operational resource consumption dynamics; analysis of the continuous operating modes control data of a selected subgroup of wells equipped with electric submersible pumps with similar run life potentials; structural-parametric estimation of the resource model operational block with information content assessment of the regression variables in the orthogonal decomposition scheme. Research results and summary. Consumption estimation of the accumulated energy potential (resource) initially founded by the submersible unit manufacturer makes it possible to predict its run-life considering the effects of accelerated aging and wear factors. Allocating units structurally similar in design and standard sizes in equal run-life clusters allows focusing the identification problem of the operational block of the regression resource model with further additional detailing of its structural block. The information content control of the wear dynamics factors based on the orthogonal decomposition of regression equation by the method of the leading element makes it possible to correct the factor set dimension to ensure the stability of the required parametric estimates. The example of statistical tests on operation control data for a subgroup of 33 wells shows that the reduction of the uninformative factor «overvoltage» in identification scheme leads to the significant decrease in the parametric estimation error. The obtained results can serve as a theoretical and methodological basis for creating an adaptive system for monitoring the resource usage dynamics of electric submersible pumps-equipped wells with continuous operating modes and optimization of submersible unit selection, operation and maintenance.
Keywords