Risk assessment of water inrush disasters of karst tunnels based on variable weight-cloud model: A case study of Zhongliangshan tunnel

In order to solve the uncertainty and complexity of risk factors and the subjectivity of risk assessment of water inrush disasters in karst tunnels, the risk of water inrush disaster has been scientifically assessed. According to the Zhongliangshan karst tunnel project on the middle route of Chengdu-Chongqing, the study constructed a risk assessment model of water inrush disasters in karst tunnels based on variable weight-cloud model. First of all, referring to the research methods of Wu Xin, Liu Dunwen and others, and consulting professors in the field of geological disasters and engineers from tunnel construction and inspection units, a total of 8 experts determined the grade and classification standard of each influencing factor on water inrush disasters of karst tunnels, and clarified the parameter value of each influencing factor of Zhongliangshan tunnel on the middle route of Chengdu-Chongqing. In this study, five influencing factors were selected to construct an index system of risk assessment of the water inrush in karst tunnels. These five factors include formation lithology (calcium carbonate content in strata and rock structure), geologic structures (water-conducting fault structure, water-blocking fault structure and fold structure), surface catchment conditions, tunnel spatial locations and alternating conditions of groundwater circulation. In addition, the grading standards of water inrush disasters were determined, and accordingly the disasters were divided into five risk levels, low, mild, moderate, high and highest.Firstly, the cloud model was used to determine digital characteristics of the risk level of each index. The diagram of membership cloud of each influencing factor was drawn by MATLAB. The single factor membership degree (μj(x)) of each influencing factor was calculated according to parameter values of water inrush disasters in karst tunnels. Secondly, the analytic hierarchy process (AHP) was used to determine the constant weight. In order to avoid the situation that the constant weight does not change with the state value of the index to be evaluated, the punitive variable weight method was used to determine the variable weight vector (W(x)) and the comprehensive membership degree (U). Finally, according to the principle of maximum membership degree, risk levels of water inrush disasters in karst tunnels were calculated, and water inrush disaster situations of 7 sections in Zhongliangshan Tunnel on the middle route of Chengdu-Chongqing were determined. The results show that water inrush disasters in Zhongliangshan tunnel are between level III and level VI, with a high risk. Among them, DK15 + 630-DK15 + 680 and DK16 + 750-DK16 + 78 are the sections with a moderate risk; DK16 + 020-DK16 + 460 are of high risk; DK14 + 720-DK15 + 630, DK15 + 680-DK16 + 020, DK16 + 460-DK16 + 750 and DK16 + 785-DK17 + 380 are the sections with a highest risk. Water inrush disasters of karst tunnels can be attributed to a variety of influencing factors. The parameter value of each influencing factor of the high-risk section is higher than that of the low-risk section, and the risk of water inrush disasters in a transition zone between a karst area and a non-karst area is the highest. With the large porosity, the developed karst, and active groundwater, the soluble rock stratum is a three-medium system of pores, fissures and pipelines, which provides conditions for the occurrence of water inrush disasters and thus increases disaster possibility.The assessment result is in consistency with the actual situation of water inrush and tunneling. The consistency indicates that the risk assessment index and its system are applicable to water inrush assessments in karst tunnel areas. The cloud model intuitively reflects a fuzzy membership of risk; the variable weight theory constructs an equilibrium function, and each index is weighted according to the specific situation. It is a good solution to the problem of mutual neutralization between the indexes in the risk assessment of water inrush in karst tunnels, which is conducive to observing the change range and relative importance of each index. The risk assessment method of water inrush disasters of karst tunnels constructed in this paper can realize the objectivity of risk classification of water inrush disasters in tunnels from a multiple decision-making perspective, which is applicable to the risk assessment of karst tunnels and provides reference for the tunnel quality control and life assessment in the future.