Oil and Gas operators now have the possibility to collect and leverage significant amounts of data directly at the extremities of their production networks. Data combined with Industrial Internet of Things (IIoT) architecture is an opportunity to improve maintenance of assets, increase their up-time, reduce safety risks and optimize operational costs. However, to turn data into meaningful insights, Oil and Gas industry needs to fully take benefit of Machine Learning (ML) models which are able to consume real-time data and provide insights in isolated locations with scarce connectivity. These ML models need to be precise, robust and compatible with Edge computing capabilities. This paper presents an analytics solution for rod pumps, capable of automated Dynagraph Card recognition at the wellhead leveraging an ensemble of ML models deployed at the Edge. The proposed solution does not require Internet connectivity to generate alarms and addresses confidentiality requirements of Oil and Gas industry. An overview of the employed ML models as well as the computing and communication infrastructure is given. We believe the given outline is insightful for the petroleum industry on its road to digitization and optimization of Artificial Lift systems.

No abstract

As data processing capabilities improve electronic device performance, it becomes necessary for Oil and Gas operators to understand how computing power can be harnessed at the extremities of their production network by deploying Edge Analytics solutions. This paper will discuss adapting the Industrial Internet of Things (IIoT) in the Upstream Automation domain, specifically in Artificial Lift assisted production, and will shed light on how Edge Analytics can be leveraged to deploy Machine Learning Models (MLMs) directly at the well-head.

Progressive Cavity Pumps (PCPs) are the predominant form of artificial lift method deployed by Australian operators in Coal Seam Gas (CSG) wells. With over five thousand CSG wells [1] operating in Queensland's Bowen and Surat Basins, managing and maintaining PCP supported production becomes a significant challenge for operators. Especially when these pumps face regular failures due to the production of coal fines. It is possible to gauge the holistic production performance of PCPs with the aid of real-time data, as this allows for pro-active and informed management of artificially lifted CSG wells. Based on data obtained from two (2) CSG operators, this paper will discuss in detail how features extracted from time series data can be converted to images, which can then aid in autonomously detecting abnormal PCP behavior.

Assessing real-time performance of Artificial Lift Pumps is a prevalent time-series problem to tackle for natural gas operators in Eastern Australia. Multiple physics, data-driven, and hybrid approaches have been investigated to analyse or predict pump performance. However, these methods present a challenge in running compute-heavy algorithms on streaming time-series data. As there is limited research on novel approaches to tackle multivariate time-series analytics for Artificial Lift systems, this paper introduces a human-in-the-loop approach, where petroleum engineers label clustered time-series data to aid in streaming analytics. We rely on our recently developed novel approach of converting streaming time-series data into heatmap images to assist with real-time pump performance analytics. During this study, we were able to automate the labelling of streaming time-series data, which helped petroleum and well surveillance engineers better manage Artificial Lift Pumps through machine learning supported exception-based surveillance. The streaming analytics system developed as part of this research used historical time-series data from three hundred and fifty-nine (359) coal seam gas wells. The developed method is currently used by two natural gas operators, where the operators can accurately detect ten (10) performance-related events and five (5) anomalous events. This paper serves a two-fold purpose; first, we describe a step-by-step methodology that readers can use to reproduce the clustering method for multivariate time-series data. Second, we demonstrate how a human-in-the-loop approach adds value to the proposed method and achieves real-world results.