Artificial Intelligence Applications in Reservoir Engineering: A Status Check

Abstract

This article provides a comprehensive review of the state-of-art in the area of artificial intelligence applications to solve reservoir engineering problems. Research works including proxy model development, artificial-intelligence-assisted history-matching, project design, and optimization, etc. are presented to demonstrate the robustness of the intelligence systems. The successes of the developments prove the advantages of the AI approaches in terms of high computational efficacy and strong learning capabilities. Thus, the implementation of intelligence models enables reservoir engineers to accomplish many challenging and time-intensive works more effectively. However, it is not yet astute to completely replace the conventional reservoir engineering models with intelligent systems, since the defects of the technology cannot be ignored. The trend of research and industrial practices of reservoir engineering area would be establishing a hand-shaking protocol between the conventional modeling and the intelligent systems. Taking advantages of both methods, more robust solutions could be obtained with significantly less computational overheads.

Smart Proxy Modeling of a Fractured Reservoir Model for Production Optimization: Implementation of Metaheuristic Algorithm and Probabilistic Application

“…Fortunately, in todayÕs world of digitalization, methods of artificial intelligence and machine learning (AI&ML) have come to the rescue. In this context, Ertekin and Sun (2019) provided a very comprehensive review on the implementation of AI&ML methods in the field of reservoir engineering. They also proposed the use of hand-shaking protocol that would combine the advantages of both traditional and intelligent reservoir modeling to develop more powerful computational protocols.…”

Application of nature-inspired algorithms and artificial neural network in waterflooding well control optimization

“…NRS can be conveniently (and is also frequently) coupled with any mathematical algorithm to optimize waterflooding or any EOR techniques. This has also been one of the most common practices in the industry as highlighted in some literatures (Peaceman 1977;Jansen et al 2009;Ertekin and Sun 2019;Baumann et al 2020). Nonetheless, as perceived, NRS is developed based upon the physics to model the behavior of fluid flow in porous media.…”

“…In this aspect, the word "proxy" denotes "to act on behalf of another." This denotes that proxy models are the replica of numerical reservoir models which can be readily employed for practical applications in the industry (Mohaghegh 2011;Ertekin and Sun 2019).…”

Application of artificial neural network for predicting the performance of CO2 enhanced oil recovery and storage in residual oil zones

“…These issues could be solved by an Artificial Neural Network (ANN) to create the smart proxy model for the predictive purpose. ANN can be employed as an alternative solution for complicated problems in the reservoir engineering 29 . There are many studies to use ANN in petroleum engineering, such as the screening enhance oil recovery “method” 30 , assisted history matching 31 , estimation dew point pressure 32 , drilling engineering 33 ,etc.…”