A Review of Machine Learning Algorithms Applied to Reservoir Exploration and Prediction
Year: 2025
Type: conference-paper
Abstract: Reservoir exploration and prediction are highly complex tasks. They require knowledge from multiple disciplines. Key challenges include interpreting multidimensional geological data and evaluating low-permeability reservoirs. Traditional methods mainly rely on porosity and permeability. However, they often fail to reflect the heterogeneity and true storage capacity of tight reservoirs. In recent years, machine learning has become a promising alternative. Algorithms such as Support Vector Machines (SVM), Random Forests (RF), clustering, and deep learning models like CNNs and RNNs provide new tools for reservoir analysis. These methods are well-suited for handling complex, nonlinear, and high-dimensional data. In this work, we review recent developments in the use of machine learning for reservoir evaluation. It highlights how these techniques improve quantitative assessment and support better decision-making in oil and gas development.
Authors: Bing Lu, Hanqing Wang, Huilan Zhao, Huilan Song, Yan Li
Institution: China National Petroleum Corporation (China)
Language: English
Subfield: Mechanics of Materials
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4413170397