Addressing Diverse Petroleum Industry Problems Using Machine Learning Techniques: Literary Methodology─Spotlight on Predicting Well Integrity Failures

Abstract

Artificial intelligence (AI) and machine learning (ML) are transforming industries, where low-cost, big data can utilize computing power to optimize system performance. Oil and gas (O&G) fields are getting mature, where well integrity (WI) problems become more common and field operations are now more challenging. Hence, they are good candidates for transformation due to the low cost of data storage, highlighting the oil market decline, along with dynamic risk posed during operations. This paper is presenting a comprehensive compilation of different ML applications in diverse disciplines of the petroleum industry. The pool of AI and ML with respect to different areas of applications along with publication years has been categorized. The main focus of this study is classifying well integrity failures where the authors found that the potential of AI and ML in predicting well integrity failures has not been efficiently tapped, and there is an explicit gap in the literature. First, the applications of AI, ML, and data analytics in the O&G industry are discussed thoroughly, so this paper can be a comprehensive reference for readers and future researchers. Then data preprocessing is explained. This includes data gathering, cleaning, and feature engineering. Next, the different ML models are compared and discussed. Finally, model performance evaluation and best model selection are described. This study would be a concrete foundation in the design and construction of ML programs that can be deployed for WI risk management. The developed model can be simply used for any well stock, providing quick and easy assessment instead of subjective and tedious assessment. The layout can be simply adjusted to reflect the risk profile of any well type or any field.

Current Status and Prospects of Artificial Intelligence Technology Application in Oil and Gas Field Development

“…Yang et al proposed the construction and application of a comprehensive research digital platform for oil and gas exploration and development, which greatly improves the solving efficiency of artificial intelligence recognition. In recent years, the ensemble Kalman filter (EnKF), ensemble smooth multiple data assimilation (ESMDA), random gradient approximation algorithm, Markov process, and other non gradient methods for historical fitting have been widely applied. − At the same time, breakthroughs in machine learning and deep learning have also brought new ideas to artificial intelligence recognition. − Artificial intelligence uses the EnKF method to assist in automated history fitting, − and by establishing a reservoir model, set the parameters that need to be fitted. − Based on the EnKF method combined with production performance data, reservoir parameter inversion and reservoir simulation optimization are achieved, greatly improving the fitting accuracy. − This method greatly reduces the workload of reservoir engineering personnel and simplifies the history fitting workflow. However, the upstream development of China’s petroleum industry still faces challenges…”

Applications of AI in oil and gas projects towards sustainable development: a systematic literature review

“…Several improvements in the theoretical evidence have been made over the last decade. It is because of the increased production of oil and gas worldwide (Salem et al 2022 ; Khamis et al 2020 ). The modern operational management methods can be entirely related to total quality management or specifically associated with customised operations where the strategies can be significantly diverse from the theoretical research (Ibrahim and Elkatatny 2022 ).…”

Modeling Permeability Using Advanced White-Box Machine Learning Technique: Application to a Heterogeneous Carbonate Reservoir

“…Artificial intelligence (AI), an essential part of the engineering toolkit in recent decades, has been used to solve various environmental and engineering problems. ,− Machine learning (ML) is a subfield of AI that encompasses a variety of data processing techniques, including classification, regression, and clustering . Supervised and unsupervised techniques are the two broad divisions of ML. , Because of the ability of ML algorithms to recognize patterns and provide valuable predictions, the use of data-driven/ML algorithms has gained much attention in the energy, oil, and gas industry. ,,− Data-driven and ML techniques have been used when sufficient core data or, similar to most previous studies, a combination of core data and well-logging/seismic data are available to build predictive models. ,− The principal use of data-driven strategies and ML algorithms in petrophysics is rock typing and permeability predictions. ,− Most studies in the literature used different variations of support vector machine (SVM) and artificial neural network (ANN) to classify the reservoir into homogeneous clusters and then predict the permeability of each cluster.…”