Author(s)

Xuelian Cheng, Weiliang Li, Xin Yu, Ying Xu, Xianshuo Huo

Corresponding Author

Xuelian Cheng

Abstract

To determine the distribution of natural gas hydrate resources, GIS and spatial analysis tools are utilized to map exploration wells and resource density. Data from 14 well locations are used to create a minimum bounding rectangle, outlining potential resource boundaries. Resource parameters (effective thickness, porosity, and hydrate saturation) are analyzed using descriptive statistics and mixture distribution models to handle data complexity. Estimation combines geological statistical methods and models like the volumetric method, factoring in effective area, thickness, and gas production, along with hydrate saturation and porosity. Voronoi diagram analysis identifies optimal new exploration well locations. Machine learning algorithms process well data, generating probability distribution and spatial variation models, and Kriging interpolation supplements missing data. This enhances the precision of resource distribution maps, aiding scientific decision-making.

Keywords

Natural Gas Hydrate, GIS, Spatial Analysis, Resource Estimation, Machine Learning, Voronoi Diagram, Probability Distribution