Summary Fundamental understanding of the flow inside progressing-cavity pumps (PCPs) represents an important step in the optimization of the efficiency of these pumps, which are largely used in artificial-lift processes in the petroleum industry. The computation of the flow inside a PCP is extremely complex because of the transient character of the flow, the moving boundaries, and the difference in length scale of the channel height between the stator and rotor. This complexity makes the use of computational fluid dynamics (CFD) as an engineering tool almost impossible. This work presents an asymptotic model to describe the single-phase flow inside PCPs using lubrication theory. The model was developed for Newtonian fluid, and lubrication theory was used to reduce the 3D Navier-Stokes equations in cylindrical coordinates to a 2D Poisson's equation for the pressure field at each timestep, which is solved numerically by a second-order finite-difference method. The predictions are close to the experimental data and the results obtained by solving the complete 3D, transient Navier-Stokes equations with moving boundaries, available in the literature. Although the accuracy is similar to the complete 3D model, the computing time of the presented model is orders of magnitude smaller. The model was used to study the effect of geometry, fluid properties, and operating parameters in the pump-performance curves and can be used in the design of new pumping processes.

A decision support system was developed based on Failure Root Cause Analysis aiming to reduce the mean time between failures (MTBF) of sucker rod pumping wells. It collects, organizes and integrates data about pulling services, failures, and technical observations and uses a combination of Failures Modes and Effects (FMEA) and Fault tree (FTA) analysis techniques to generate diagnosis. The system is based on an ontology that describes the domain, including: equipments, failures modes and causes. The diagnosis is inferred by a specialist system based on production rules and identifies well severity level based on data aggregation algorithms, in particular, k-means, which clusters the behavior of similar wells given the subsurface reservoir and equipment characteristics, type of material installed and the failures characteristics. The sucker rod pumping failures diagnostic system is currently being implemented at Petrobras. Initial results are promising: even during pilot tests, analysts were able to detect problems with manufactures and a lack of synchronicity between component changes.