Library

Notes: The estimation of thermal conductivity from well logs is of great importance for all studies on thermal evolution of sedimentary basins. The approach proposed here is to first interpret the well logs in terms of mineralogic composition and porosity and then to estimate the in situ conductivity from a mixing formula: the geometric average of individual conductivities weighted by the volumetric proportion of each component. The validity of the model has been widely verified by laboratory measurements on isotropic samples. However there is a need for validation of such an approach for in situ conditions.Temperature data recorded at equilibrium in two wells of the Paris basin offer a unique possibility of testing this approach: in the case of a steady state conductive thermal régime (Fourier's law), the conductivity must be proportional to the inverse of the temperature gradient. The well logs are quantitatively interpreted in terms of mineralogic composition, porosity and then thermal conductivity using a conductivity law calibrated on cuttings and cores laboratory measurements. For three examples of well sections, the computed thermal conductivity compares reasonably well with the inverse of the gradient.For each of the three examples, it is possible to define a depth range where the apparent heat flow is almost constant with depth and where the conductive régime seems to prevail. This allows a refinement of the conductivity model with emphasis on the contribution of clays: assuming that the conductivities of non-clay end terms are known, it is possible to obtain the conductivity of clay giving the best fit to the temperature measurements. In the three cases the conductivity of clay (1.2-1.4 W m−1 K−1) is found to be lower than the assumed one (1.8 W m−1 K−1) which probably indicates an anisotropy effect.Finally an empirical model where the bulk conductivity is expressed directly as a function of well logs is tested. In one case the results obtained are quite compatible with the previous one, but in the other two sections, the stability and the physical meaning of this model are found to be questionable. This could be due to the lack of constraints in the underlying implicit mineralogy model.