Abstract
Inter-well tracer tests have numerous applications in determining the volumetric sweep, levels of heterogeneity, and delineation of flow barriers such as the faults in subsurface in reservoirs. Tracer tests are also performed in the laboratory to determine core samples’ heterogeneity. Simulation of physical dispersion requires refined models, which inevitably increase the numerical dispersion in the obtained results. In this paper, we quantify the numerical dispersion associated with various techniques available for the simulation of tracer flow. Numerical dispersion in the simulations can be quantified by comparing the simulation results with the experimental data. For this purpose, we first reviewed the fundamentals of tracer flow and introduced the Convection-dispersion equation as a basic model to describe tracer flow in porous media. Then we constructed a refined model to simulate a series of tracer flood scenarios. A detailed sensitivity analysis was carried out in a systematic manner to specify the individual impact of physical and numerical dispersion. Finally, we modelled tracer experiments performed on Indiana Limestone (IL) carbonate rocks to examine the accuracy of Fick's and Darcy's equations, and the results are presented. The impact of both numerical (non-physical) and physical dispersion was examined during the core scale simulations. It was concluded that the numerical TVD algorithm (embedded within the commercial ECLIPSE software) can appropriately model the tracer flow in porous media with minimal numerical errors during simulations. It was also shown that physical dispersion significantly affects tracer test results and that it must be considered when simulating tracer flows by defining an appropriate Peclet number. Finally, the results showed that solving the conventional convection-dispersion equation along with the numerical TVD algorithm can perfectly match the experimental data of several tracer flood tests performed on outcrop Indiana limestone core samples.
