Session: 01-03 Advanced Analytics for Condition Monitoring and Asset Reliability

Paper Number: 173634

173634 - Parametric Pump Modeling for Health Diagnosis and Performance Assessment in Sucker Rod Systems 

In this study, we present a parametric physical model for the sucker rod pumping system, crafted to accurately reflect the pump's behavior under a broad spectrum of operating conditions. This model explicitly delineates the dynamic interactions among plunger motion, valve operation, fluid compressibility, and leakage effects, thereby aiding in the identification of critical pump parameters such as plunger clearance, valve efficiency, and pressure dynamics. By deriving these parameters from surface-acquired data, the framework establishes a solid foundation for assessing pump health, diagnosing performance losses, and identifying failure modes like gas interference, fluid pound, or mechanical wear. To enhance the accuracy of diagnostic outputs, we integrate the pump model with a finite element (FEM) model of the rod string, which considers elastic wave propagation, axial stretch, buckling behavior, and wellbore deviation. This integration allows for the generation of synthetic pump cards that more precisely capture downhole conditions, especially in deviated wells where conventional modeling approaches often fail to predict true pump behavior. The integrated model also supports sensitivity analyses to assess how varying operational parameters affect system performance, offering valuable insights for production engineers aiming to optimize lift efficiency and extend equipment lifespan. This comprehensive modeling approach enhances system understanding by linking surface measurements with downhole dynamics, providing a pathway toward advanced condition monitoring, predictive maintenance, and optimization of sucker rod pumping operations. The developed framework lays the groundwork for future digital twin applications in artificial lift systems, merging physics-based modeling with data-driven diagnostics to advance operational efficiency in both conventional and unconventional oil and gas fields.

Presenting Author: Omar Khaled University of Houston

Presenting Author Biography: Omar Khaled received an associate degree in Math and Physics from 'IPEIS - Institut Préparatoire aux Études d'Ingénieur de Sfax' in 2018 in Sfax, Tunisia. He also received a multidisciplinary engineering degree from École Polytechnique de Tunisie in 2021, in Tunis, Tunisia. He is currently pursuing a Ph.D. in Mechanical Engineering at the University of Houston, Houston, TX, USA. He has been working as a research assistant since 2022. His research focuses on linear and nonlinear system identification, diagnostics and prognostics, uncertainty quantification, and machine learning.