Preferential flow modelling in a vadose zone using macro 5.0 - Cape Flats porous sands and Mpumalanga highveld clays case studies
MetadataShow full item record
Understanding fluid flow and solute transport within the vadose (unsaturated) zone is an essential prerequisite for protection of groundwater from contaminant sources occurring overland. Preferential flow paths in the vadose zone pose a significant problem because they are potential avenues for rapid transport of chemicals from contamination sources to the water table. The objectives of this study were: i) To review and understand flow and transport processes in unsaturated zones. In this study, particular emphasis is placed on understanding mechanisms that cause non-uniform (preferential) flow for two case studies, namely the Cape Flats sandy environment and the Mpumalanga Highveld fractured rock environment. ii) To evaluate the adequacy of models, in particular MACRO 5.0, in simulating flow and transport in the vadose zone, by making use of two case study sites (Cape Flats and Mpumalanga Highveld). Of particular importance is the evaluation of transfer coefficients to represent fluid and solute exchange between macropores and matrix. iii) To run a sensitivity analysis with MACRO 5.0 in order determine which input model parameters are the most relevant in describing the effects of preferential flow in water and solute transport. Two case studies were investigated, the first at a landfill site overlying sandy unconfined aquifer (Coastal Park, Cape Town), and the second at an industrial site overlying cracking clayey soil and fractured rocks (Mpumalanga Highveld - Secunda, Mpumalanga Province). For the Coastal Park site, simulations of soil water content and leaching of a generic mobile contaminant were compared to monitored soil water contents and chloride concentrations in groundwater. For the Mpumalanga Highveld site, simulations of soil water content and concentrations of boron and fluoride originating from effluent irrigation were compared to soil profile measurements. In both cases, the MACRO 5.0 model predictions agreed with measurements well, provided appropriate input calibration data were used. The sensitivity analysis indicated that soil water properties related to preferential flow (hydraulic conductivity at the boundary between macropores and matrix, soil water content and tension, and diffusion) have influence on simulation results. Similarly, the solute balance is mostly influenced by degradation rate coefficients (both in solid and liquid phases), sorption distribution coefficients and solute concentrations.