Review of artificial intelligence applications in geothermal energy extraction from hot dry rock

Abstract The geothermal resources in hot dry rock (HDR) are considered the future trend in geothermal energy extraction due to their abundant reserves. However, exploitation of the resources is fraught with complexity and technical challenges arising from their unique characteristics of high temperature, high strength, and low permeability. With the continuous advancement of artificial intelligence (AI) technology, intelligent algorithms such as machine learning and evolutionary algorithms are gradually replacing or assisting traditional research methods, providing new solutions for HDR geothermal resource exploitation. This study first provides an overview of HDR geothermal resource exploitation technologies and AI methods. Then, the latest research progress is systematically reviewed in AI applications in HDR geothermal reservoir characterization, deep drilling, heat production, and operational parameter optimization. Additionally, this study discusses the potential limitations of AI methods in HDR geothermal resource exploitation and highlights the corresponding opportunities for AI's application. Notably, the study proposes the framework of an intelligent HDR exploitation system, offering a valuable reference for future research and practice.