Résumé
We propose a model for simulating the changes in porosity and permeability caused by hydrothermal diagenesis in sedimentary aquifer where salinity, temperature and fluid flow vary in space and time. Such modifications of the hydrodynamic properties of the medium are bounded to geochemical reactions and groundwater flow. Fluid velocity is particularly low in deep reservoirs (typically less than 1 m/year). Then, the local equilibrium simplification, which is justified by a set of world-wide data of the chemical composition of groundwater, can be implemented toward straightforward transient calculations. In the model presented here, the coupled processes of fluid flow, temperature and chemical species transport are solved using well established methods. The originality of the model is the development carried on to predict the permeability evolution controlled by the mineral dissolution and precipitation. Usually to simulate permeability changes modelers use the classical porosity-permeability model based on statistical analyses of in situ or laboratory measurements. However, hydraulic conductivity changes are not controlled solely by porosity changes, but also depend on pore-scale structure transformations. Depending on the mineral type, the precipitation or dissolution of the same quantity of volumetric quantity will induce very different changes in the hydraulic conductivity. Principally clay minerals depict a wide range of atypical organisations of different microstructural characteristics of the porous media. The spatial distribution of these characteristics cannot be modelled at basin scale. Away from both too complicated and too unrealistically simplified approach, the model presented here is based on the calculation of the permeability evolution from the change in the mineral fraction due to mineral precipitation and dissolution. To simplify, the minerals are divided into two groups: clay minerals and non-clay minerals. The specific contribution of clay minerals is controlled by a single weighting coefficient. This coefficient is associated to the proportion of poorly connected porosity that characterize clay structure, albeit it is presently impossible to propose any quantitative relationship between the value of this parameter and the microstructural characteristics of the diagenetic clays. The model is tested here to simulate the evolution of the porosity and the permeability in a peculiar zone of the Paris Basin. The study area of several hundred meters large is inside the Dogger aquifer, close to the Bray fault zone where invasion of saline water from Triassic formation takes place. This zone is characterised by high thermal and salinity gradient as well as by the superposition of sub-horizontal regional flow and ascendant fault-controlled flow: it is an ideal case study for examining the importance of taking into account the specific contribution by clay minerals when computing permeability evolution. This study is proposed as a parameter sensibility analysis: - to compare the relative influence of the clay weighting coefficient, the temperature, the salinity, and the cementation exponent on the computed evolution of the permeability, - to discuss the consequences of the introduction of the clay weighting coefficient in comparison to the classical porosity - permeability evolution model, - to simulate various evolution scenarios of past and future thermal and geochemical constraints and their consequences on the evolution of the permeability changes in the Bray fault zone taking into account uncertainties on the value of the clay weighting coefficient and on the cementation exponent. Forty-one simulations of one million years were necessary to cover a large spectrum of the expected variations of each parameter. The results show that: - the local variation of the permeability depends on the time evolution of temperature and of salinity, and on the values of the cementation exponent of the porosity-permeability law and of the clay weighting coefficient. Within reasonable ranges of these four parameters, their influence on the permeability changes is of the same order of magnitude, - the influence of the clay weighting coefficient on the porosity evolution is negligible. Feedback effects of permeability evolution on the porosity evolution, through the change in the flow regime, is minor, - by the use of a classical model without a clay weighting coefficient, permeability and porosity present the same pattern of evolution: they both increase or decrease. By the use of the clay weighting coefficient, in some places the permeability and porosity can show opposite evolution. One increases when the other decreases even for low values of the coefficient, - in the vicinity of the fault, the model predict an increase of permeability independently of potential temperature and salinity modifications and whatever the clay mineral weighting coefficient is: Bray fault sealing is unlikely as long as head gradient is maintained in the fracture zone.