Optimization of Gas Hydrate Well Placement Based on K-Means Cluster Analysis and Genetic Algorithm

Abstract

Natural gas hydrate, also known as "combustible ice", is an efficient and clean backup energy source. The purpose of this study is to conduct a comprehensive quantitative assessment of natural gas hydrate resources in a sea area, and to propose a site selection plan for additional wells. Firstly, this paper obtains the geological data of 14 exploration wells in this sea area, establishes a K-means clustering model, and derives the clustering of well data into three different regions, and the distribution range of natural gas hydrate is obtained by calculating the location of the center of each cluster. Secondly, combined with GIS tools, spatial interpolation analysis was carried out to generate the spatial distribution map of each parameter (effective thickness, porosity, hydrate saturation) in the study area. Finally, this study proposed to add five more wells in the region, and the optimal spatial coordinates of the five well locations were obtained by constructing a multi-objective optimization model using a genetic algorithm. This study creatively proposes a dual-stage optimization model integrating K-means clustering and genetic algorithms, which not only addresses the gap in comprehensive decision-making frameworks but also enhances exploration accuracy in complex marine environments and promotes the commercialization development process.