Multivariate probabilistic modelling for seepage risk assessment in tunnel segments

Seepage in the lining structure of a tunnel has become an increasingly significant issue in underground engineering, which impacts the performance and service life of the tunnel structure. This paper presents a generic multivariate probabilistic macro model for reliability assessment of seepage behaviour in tunnel segments consisting of pre-manufactured concrete lining rings. The model is characterised by multiple influencing factors including the tunnel line, rail operating direction, seepage joint location and tunnel operational time. A multivariate statistical approach, proportional hazard modelling technique, is exploited to empirically model the seepage occurrence probability along the tunnel segment length, meanwhile integrating impacts from different influencing factors on the tunnel section seepage. The most commonly used Weibull and log-normal distributions are employed in the development of the proportional hazards modelling. The methodology is illustrated with the seepage data collected from real-world subway tunnels. Seepage occurrence probability is also assessed over the critical factors.