UPSCALING SOLUTE TRANSPORT IN SATURATED POROUS MEDIA FROM THE EXPERIMENT SCALE TO FIELD SCALE

Computational models for flow and solute transport in geological media are effective tools in hydrogeological and engineering related applications. However, modeling processes at larger scale (e.g., on the order of hundred meters and above) involves substantial computational burden. This makes it impossible to incorporate smaller scale heterogeneity into site-scale flow and transport models. A common approach to overcome such difficulty is to perform upscaling by creating models that require less intensive computations. Any upscaling method must preserve important properties of the spatial variability of flow and transport parameters such as permeability. Upscaled numerical models are often used with a coarse grid discretization. The generalizations of grid and heterogeneity result in an inaccurate description and loss of detailed information to some degree, prohibiting our ability for having accurate predictions. The main goal of this work is to perform upscaling on related parameters for flow and non-reactive solute transport at the small laboratory scale, and to explore the suitability of the upscaled parameters from experiments for site scale models. We also investigate potential impacts of upscaling on the temporal and spatial evolution of dissolved solute plumes in groundwater. One of our goal is to explore how uncertainty evolve as a result of upscaling. We performed detailed Monte Carlo simulations on flow and non-reactive solute transport in heterogeneous media, while using block-scale effective dispersion tensor for transport simulations in order to represent loss of heterogeneity during upscaling. Our results provide a computationally efficient tool in order to describe solute transport at different temporal and spatial scales.