An analytical model for surrounding rock classification during underground water-sealed caverns construction: a case study from eastern China

Scientific surrounding rock classification (SRC) of underground water-sealed oil reservoir caverns is greatly significant for reducing risks and costs during construction. As to underground water-sealed caverns, not only the stability and deformation of the surrounding rock, but also the requirement of the water-sealed function must be considered. At present, there is no special standard for the SRC of underground water-sealed caverns. Based on the cloud theory and principal component analysis (PCA) method, a novel analytical model was proposed for SRC of underground water-sealed caverns. Five indexes: rock strength, rock integrity, discontinuity surface characteristics, groundwater, and the angle between the horizontal plane and the main discontinuity were selected to establish a multi-index evaluating system. With enough measured samples, the weights of the five indexes were determined based on the PCA method. According to cloud theory, five indexes affecting the SRC were investigated to establish five corresponding single-index normal cloud models. Subsequently, the single-index certainty degrees to every grade of surrounding rock were calculated. Finally, combined with the weights, a multi-index certainty degree was calculated as guidance to classify the surrounding rock. The proposed novel model was applied to a case study, i.e., the first underground water-sealed oil reservoir in China. The evaluating results were in broad agreement with the excavation results. The present findings imply that the proposed novel model is an effective method for the SRC in underground water-sealed caverns and can provide some useful references to similar projects.