In deep tunnels the knowledge of in-situ rock temperatures is of great importance, e.g. for the design of ventilation and cooling. A finite element model simulation technique for rock temperature prediction is described. The method handles complex thermal/hydraulic characteristics, pronounced topographic relief, deep groundwater circulation, transient coupled heat/fluid transfer, uplift/erosion and thermal history. During tunnel excavation the rock temperatures were measured. The procedure has been verified by comparing measurement and prediction under Alpine conditions, namely in the Gotthard base tunnel, Switzerland and the Koralm Tunnel, Austria: The agreement is well, within ±15%. Reliable simulations call for a solid data base; e.g. surface temperatures, basal heat flow and uplift/erosion patterns must be determined beforehand. The range of geothermal and hydrogeologic properties should be known. Only temperature data along the planned tunnel trace below the highest cover, preferably measured in well positioned boreholes, enable proper model calibration.
The distribution of in-situ rock temperatures in deep tunnelling is of predominant importance in planning, construction and operation, e.g. for the design of ventilation and cooling. The determination of rock temperatures along a planned tunnel trace at depth is especially a demanding task in mountainous terrain. The rock temperature field within a mountain massif is the result of heat transport processes (heat conduction and advection) and depends on several boundary conditions (e.g. surface temperature, basal heat flow) as well as on numerous parameters (e.g. 3-D topography, distribution of geological units/of thermal conductivity, water circulation pattern/distribution of hydraulic conductivity), along with transient processes like uplift/erosion or paleoclimatic changes. The complexity of these parameters calls for a correspondingly flexible and efficient calculation method. Only advanced numerical model simulation can cope with these manifold requirements. The development and application of such a modelling approach is presented for the Gotthard Base Tunnel (GBT), Switzerland, combined with its verification based on actual measurements along the excavated tunnel.