Deep learning and comparative study of gas saturation models for tight sandstone reservoirs with strong heterogeneity: a case study of the Upper Triassic Xujiahe Formation in Sichuan Basin, China

Advanced and effective models for gas saturation interpretation in tight reservoirs are mainly based on electricity and acoustics; however, these methods are not always effective, especially in determining the rationality of some undetermined constant methods. Therefore, the purpose of this study is to achieve high-precision prediction of gas saturation in tight reservoirs by comparing the principles and case analysis results of different typical gas saturation models. In this paper, the Archie formula, the three-phase Biot-type equation, and the effective medium symmetrical conduction model were used to evaluate the gas saturation of tight sandstone reservoirs from the perspective of electricity and acoustics. Through the research of this paper, we determined the Archie constant of the target layer. In the application of the three-phase Biot-type equation, an adaptive method is proposed to obtain the value of the formation consolidation parameter α. In addition, a quantitative relationship between resistivity and α was established. The evaluation accuracy of this method is less affected by the shale content, although it will be affected by the true Sg of the formation. The effective medium symmetrical conductivity model has a good predictive ability for Sg in sandstone reservoirs with high shale content. The research results can provide a wide range of references for the quality evaluation of tight sandstone reservoirs. © 2021, Saudi Society for Geosciences.