Abstract
Many oil and gas companies rely on natural intelligence, resident knowledge, and rules-based logic to optimize production. This is especially true for fields where electric submersible pumps (ESP) make up a considerable proportion of production. The nature of ESP artificial lift systems makes them well suited for greater remote monitoring, enhanced automation, and implementation of machine learning (ML) for autonomous optimization. Extensive use of electrical surface controls integrated with downhole sensors provides an ideal operating environment to implement Artificial Intelligence (AI) through machine learning to achieve autonomous full self-pumping (FSP) operation. However, most operating companies stop short of using automation and machine learning to its full potential.
This paper will present a case study highlighting the demonstration, refinement, and implementation of a machine learning algorithm to optimize multiple ESP wells in the Permian Basin. The paper will also present a case study for the development, demonstration, and refinement of an autonomous ESP optimization solution driven by this machine learning model. The paper will discuss key learnings from each study to assist operators in their digital journey with considerations for effective field implementation.