Numerical simulation of the ore forming fluid migration in the sediment-hosted stratiform copper deposit, Zambian Copperbelt

A finite element numerical model was developed and applied to study densitydriven convective flow as a potential ore-forming mechanism for sediment-hosted stratiform copper deposits in the Zambian Copperbelt (ZCB). The numerical model simulates free thermohaline convection in an inhomogeneous, anisotropic, compressible porous medium completely saturated by a compressible fluid bearing dissolved sodium chloride. The model also takes into account possible inner salt and heat sources to simulate rock salt dissolution and heat emission by a magmatic intrusion. The model employs the finite element Petrov-Galerkin method with nonsymmetrical 'upwinding' weighting functions. An implicit time-stepping scheme with 'weightings' involves an iterative procedure within each time step to handle the nonlinearity of the problem. The model validation was carried out by 1) comparison of the numerical results with analytical solutions for simplified problems; 2) application of the model to three problems of free thermal convection (anisotropy effect, convective cell aspect ratio as a function of the Rayleigh number, and free discharge) and comparison with published results of numerical and experimental studies; 3) simulation ofthe groundwater flow in a complex faulted sedimentary section of the McArthur Basin (Northern Territory, Australia) to compare the model output with the results of an independent study (Garven et al., 2001). The model was applied to a 43.5km x 17km numerical domain, representing a theoretical geological section of the ZCB. The section incorporates elements of basement, footwall succession (Mindola Clastic Formation), Ore Shale, hanging wall Upper Roan dolomites, salt layer, Mwashia dolomitic and siliciclastic members and overlying Kundelungu shales. As a layer-cake structure of the section shows no significant topography, density variations due to geothermal gradient and Roan Salt dissolution are considered the main mechanism driving pore fluid circulation. Sixteen potential paleo-hydrologic scenarios were tested to study factors controlling ore-forming fluid flow: fault network configuration, salt layer geometry, rock salt physical characteristics, and piercement structures.The simulations showed notable convective flow through the basement; the result corroborates the hypothesis of a basement source for the metals contained in the Copperbelt orebodies (Sweeney et al., 1991). Supporting the study of fluid inclusion data (Annels, 1989), the modeling confirms ore body formation under lowtemerature and high salinity conditions. It also showed significant lateral flow of the rising basement fluid through the Mindola Clastic Formation beneath the Ore Shale seal; as a consequence, notable amounts of Cu can be precipitated in the footwall clastics even assuming its low depositional efficiency, as supported by geochemical and petrographic data. Low permeability shales and the salt layer form barriers to cross-stratal fluid flow; nevertheless, horizontal flow rates in the Ore Shale are sufficient for formation of ore bodies assuming this unit was an effective chemical trap. Tested scenarios revealed the essential influence of Upper Roan Salt dissolution on fluid flow and, therefore, amount of metal precipitated. In summary, the study suggests that free convection driven by geothermal gradient and dissolution of the massive Roan Salt sheet can be an effective oreforming mechanism in the stratiform ZCB system. The rising basement fluid is redirected laterally beneath the Copperbelt Orebody Member - Mindola Clastic Formation interface. The flow rates of the metal-transporting fluid are shown to be capable of forming ore bodies within the Ore Body Member and underlying footwall. The modeled amount of potentially precipitated Cu agrees with the available geochemical and petrographic data.