Can equivalent continuum models simulate flow and pollutant transport in fractured aquifers?

Model predictions of flow and pollutant transport in fractured rocks are subject to uncertainties due to imprecise knowledge of the position, orientation, length, aperture and density of the fractures. These properties are difficult to quantify precisely because fractures are located in depth in subsoil and, generally, tectonic and stratigraphic studies may provide only fracture frequency and their orientation. Subsequently, the use of the "equivalent" continuum models might help hydrogeologists to solve flow and pollutant transport problems in fractured aquifers, when fracture properties are unknown. The study presented herein has quantified the limitations of the equivalent continuum model when it is applied to simulate groundwater flow and pollutant transports in limestone aquifers. Both discrete and continuum model outputs have been compared with results derived from a pumping tracer test, carried out at Bari fractured aquifer. The test results have shown a delay of velocity estimated using continuum models, with respect to the discrete model, that decreases by increasing the hydraulic conductivity of the limestone aquifer under consideration. Maximum discrepancies have been noted for conductivity (< 10 -4 m/s) typically associated with non-karst limestone aquifers. The tortuosity has been then included into the codes in order to address flow velocity calculations in numerical codes, such as MT3DMS. Successful simulations of flow and pollutant transport have been carried out at the Bari fractured aquifers by using tortuosity. A «tortuosity/ conductivity» relationship has been also proposed in order to estimate tortuosity in fractured limestone aquifers, when results of tracer tests are not available.