HEAT ADVECTION VERSUS CONDUCTION AT THE KTB - POSSIBLE REASONS FOR VERTICAL VARIATIONS IN HEAT-FLOW DENSITY

Data from the 4 km deep KTB pilot hole (VB) show a strong vertical var iation in heat-flow density (HFD) by as much as 50 per cent. This may be caused both by heat conduction, by advection, and by transient diff usion. At the moment it is not possible to quantify exactly the contri bution of each of these. However, 2-D simulations help to define the p arameter ranges and structural features required if these processes ar e to be thermally efficient. The main results are: (1) thermal conduct ivity contrasts combined with structural heterogeneities as seen in th e drilled profile give rise to steady-state, lateral refraction of hea t. 2-D simulations of heat conduction indicate that this effect alone is sufficiently strong to account for the observed variation of HFD wi th depth. (2) Vertical Peclet number analyses of T-logs in shallow bor eholes and the KTB-VB indicate a NE-SW flow of meteoric water across t he Franconian Line (FL). However, average Peclet numbers of -0.37 +/- 0.13 in the potential recharge zone east of the FL are compatible with 2-D, steady-state simulations of heat and fluid flow only up to a dis tance of about 10 km east of the FL, and only if a crystalline permeab ility k(c) = 10(-14) m2 is assumed. (3) A permeability this high, howe ver, is not confirmed by a comparison of temperature and HFD from nume rical simulations and data from the KTB boreholes, neither for a model focusing on shallow flow systems nor a deep structural model investig ating potential contributions of convection in the entire upper crust. (4) Alternatively, a joint inversion of T-logs from the same shallow holes yields a ground-temperature history (GTH) that is in remarkably good agreement with long-term meteorological records. (5) It appears, therefore, as if the thermal regime at the KTB was generally dominated by conduction, with additional advective, topography-driven contribut ions mainly at shallow depths. The conductive regime, however, is a co mplicated one, characterized by lateral heat flow due to structural he terogeneity (and possibly anisotropy), and, at least at shallower dept hs, by transient diffusion of paleoclimatic temperature signals into t he subsurface.