Risk assessment of seepage failure in deep excavations based on fuzzy analytic hierarchy process and cloud model

Seepage failures are frequently encountered in deep excavation owing to the strong demand of underground space projects, which triggers serious geo-disasters and numerous casualties. However, fuzziness, uncertainty, and randomness continue to obstruct the prediction and evaluation of seepage risk in this area. This study collects the objective and subjective factors affecting seepage failure to present a seepage risk assessment system for deep excavation. This system consists of a risk assessment methodology, a risk assessment index system, and seepage risk level. Fuzzy analytic hierarchy process and cloud model are employed to establish the risk assessment methodology. Then, the assessment index system is proposed by selecting four first-level indices and seventeen second-level indices considering the fluid-soil interactions and external conditions. The seepage risk level is divided into four grades and determined by the maximum certainty principle. The proposed risk assessment system has been applied to assess the seepage risk in the Taihu tunnel construction. The assessment results show good agreement with other methods and practical excavation situations, which can verify the accuracy and reliability of the proposed model. This study therefore not only highlights the influence of seepage risk but also provides significance for the evaluation of seepage failure in geotechnical engineering practices.