Probabilistische Prognose von Karstwasserzutritten beim Bau von Untertagebauwerken

Abstract

Various methods have been developed in recent decades to predict hazards associated with karst voids in underground construction. Common to all these methods is that the predicted range of water inflow is often insufficient for the purpose of implementing the planned construction works. This is usually due to an incomplete knowledge of the karst conduit system within a project area, making it difficult to predict the position and characteristics of karst voids. The method presented in this paper permits a robust prediction of karst water inflow. It is based on a combination of stochastically generated, pseudo‐genetic karst conduit systems and hydraulic modelling of the hydrogeological conditions using a Monte Carlo approach. This approach facilitates a plausible estimation of the expected range of karst‐induced water inflows and also enables the probability of encountering a karst voids. to be determined. The predictions allow for differentiated treatment of the hazards associated with karst water during the construction and operation phase of underground structures. In concrete terms, this relates to the planning and implementation of exploratory measures and ground‐improvement measures, the design of the dewatering system and its monitoring during the construction and operation phase.

Trends in forecasting groundwater ingresses into underground structures

“…In fact, on the one hand, analytical, semi-analytical, empirical, semi-empirical, and numerical solutions often failed to accurately forecast the ingresses of groundwater in underground structures. 10,11 The main reason is that the dominant parameters linked to hydrogeological and excavations conditions are particularly hypothesized and simplified in the aforesaid solutions. On the other hand, huge relevant data are needed by numerical and machine learning solutions to provide reasonable results.…”