Statistical Approach for PCP Standardizing

Abstract The selection of a progressing cavity pump heavily relies on the expertise and experience of field personnel. Therefore, the selection of pumps for a large number of wells becomes an energy consuming and cost-effective task where heuristic prevails over an algorithmic procedure. This paper proposes a new approach that attempt to perform such task by combining multiphase calculations with well known pump selection criteria and statistical clustering method. The methodology requires massive information about the well completions, fluid properties and operation conditions that are utilized for predicting the pressure rising required, inlet volumetric gas fraction among other variables for every well in the field. Two approaches can be utilized to identify standardized pump models: hit frequency and K-mean algorithm. The standardized models for a particular well are considered equivalent pumps, which means that any of standardized models can be utilized in the same well with equivalent results. The process also allows identifying in how many wells a particular pump model can be utilized. A case study applying the standardizing methodology is presented. The case is a extra-heavy oil field where standardized pump models for 592 wells are required for future purchase. The methodology proposed is the first step to establish an analytical procedure for identifying standardized pump in a field. The elastomer selection needs to be added to the methodology for sake of completeness; however, it has been left out due to lack of predicting model for the elastomer swelling, thermal expansion and fluid-material chemical interaction.