URTEC-198281-MS Machine Learning for Progressive Cavity Pump Performance Analysis: A Coal Seam Gas Case Study Fahd Saghir, M.E. Gonzalez Perdomo, and Peter Behrenbruch, University of Adelaide Copyright 2019, Unconventional Resources Technology Conference (URTeC) This paper was prepared for presentation at the SPE/AAPG/SEG Asia Pacific Unconventional Resources Technology Conference held in Brisbane, Australia, 18 – 19 November 2019. The URTeC Technical Program Committee accepted this presentation on the basis of information contained in an abstract submitted by the author(s). The contents of this paper have not been reviewed by URTeC and URTeC does not warrant the accuracy, reliability, or timeliness of any information herein. All information is the responsibility of, and, is subject to corrections by the author(s). Any person or entity that relies on any information obtained from this paper does so at their own risk. The information herein does not necessarily reflect any position of URTeC. Any reproduction, distribution, or storage of any part of this paper without the written consent of URTeC is prohibited. Abstract Limited research work and publications are available to examine the performance of Progressive Cavity Pumps (PCP) based on machine learning methods, especially in Coal Seam Gas (CSG) operations. Previous work done in this space either focuses on exception-based surveillance on time-series data [1], or the use of machine learning to optimize completion design [2] and production [3]. This paper will discuss how data approximation and unsupervised machine learning methods can be applied to time-series data-sets, using data gathered from automation systems, to help analyze PCP performance and detect anomalous pump behavior. Introduction The majority of CSG wells operated in Australia are automated with Remote Telemetry Units (RTUs) and Supervisory Control and Data Acquisition (SCADA) systems [1, 2]. These automation systems gather production, mechanical, and electrical time-series data from PCPs, which allow operators to manage day-to- day production operations. However, SCADA systems are not suited to run advanced analytics and machine learning algorithms that can help determine PCP performance. To exploit information from SCADA systems, a time-series based image conversion technique is utilized to aid with a better understanding of PCP performance. Machine learning based image classification techniques are applied to these converted images, where they are clustered based on t-Distributed Stochastic Neighbor Embedding (t-SNE) and k-Means algorithms (unsupervised learning). Results from this study depict how time-series based heatmap conversion, coupled with unsupervised machine learning techniques can provide an innovative method to identify the abnormal behavior of PCPs in CSG wells. The findings discussed in this paper are based on three (3) years' worth of time series PCP data collected from forty-two (42) CSG wells.