Thermal aspects of groundwater circulation in bedrock and its effect on crustal geothermal modelling in Finland, the central Fennoscandian Shield

Abstract

The factors which influence calculations of temperature and heat-flow density (HFD) in the crust are discussed with special emphasis on groundwater circulation in bedrock. Typical geotherm calculations are sensitive to variations in the surface HFD value, and a 10 mW/m 2 change in the applied surface HFD value produces a 150–200 K change in Moho temperatures in a one-dimensional layer model. Peclet and Rayleigh number analyses were applied to different circulation schemes based on hydrogeological and geochemical data representative for the Finnish part of the shield. Bedrock was approximated as a porous medium, and the flow schemes discussed are (1) forced convection in the uppermost 1 km of bedrock, characterized by fresh meteoric groundwaters, and (2) free thermal/thermohaline convection in deeper bedrock, characterized by saline groundwaters. Hydraulic permeability is the key parameter in determining the possibility of geothermally significant disturbances by forced convection. For local groundwater flow in the uppermost 1 km (hydraulic gradient 0.01 and flow distance about 5 km), values of 10 −15 − 10 −14 m 2 and for regional flow in the uppermost 1 km (gradient 0.001 and flow distance about 100 km), values of 10 −13 − 10 −12 m 2, respectively, would be required to create deviations of more than 10% of the undisturbed conductive HFD value. The onset of free thermal/thermohaline convection in the uppermost crust is efficiently limited by downward increasing salinity of groundwater, but the converse applies if salinity decreases downwards. Rayleigh number analysis also shows that free thermal convection is possible in layers saturated with either fresh- or saline water of homogeneous salinity. A permeability value of about 10 −14 m 2 is the critical limit between conductive and convective systems. In addition to the porous media approximations, forced convective disturbances were also studied using a simplified two-dimensional model of vertical and horizontal fracture zones. It is shown that groundwater flow in major hydraulically connected vertical and horizontal fracture zones having high hydraulic permeabilities and considerable dimensions (several kilometres), can create flow systems which would be very difficult to recognize, at least with the present number of HFD sites in the Fennoscandian Shield. This is due to the fact that very few drill holes are available in the immediate vicinity of major fracture zones of hydrogeological interest. As a result, most holes represent bedrock blocks in which heat transfer is mainly conductive, which may not necessarily be representative of general behaviour in the local bedrock. It is conceivable that we have underestimated HFD in the Shield because of unidentified convective heat transfer in crystalline bedrock.