On the infiltration process in treated effluents spreading basins

Secondary treated effluents originating from the Dan Region in Israel are sent to tertiary treatment that uses Soil Aquifer Treatment (SAT) for purification within the vadose zone. The SAT is based on intermittent flooding (1-2 days) and drying (2-3 days) cycles in spreading basins constructed at the surface of a 40-m deep vadose zone. The site is located in the natural sand dunes north to the city of Ashdod, above the Israeli Coastal Plain Aquifer. The study aim is to investigate the physical and chemical processes that occur within the upper 2 meters of the spreading basins’ sandy soil profiles during the cyclic SAT operation. We explored two 2-m profiles about 50 m apart. In addition to ponding depth, continuous measurements of volumetric water content (VWC), temperature, electrical conductivity (EC) and oxidation-reduction potential at 8 different depths within the first profile were recorded. Data were collected in 15-min resolution during infiltration events for 3 months. Measurements in the second profile have been collected for a few weeks now and also include air pressure measurements. Additionally, soil samples were taken from both profiles to determine hydraulic parameters. Preliminary results indicate that the infiltration rate in the first profile is about 72 cm day-1, a low rate compared to what would be expected from a sandy profile. The VWC changes along this profile during the flooding stage imply percolation in the form of a double wetting front. First, the wetting front proceeds from the surface downward until effective saturation of 0.55. Second, the wetting front proceeds from 2-m upwards until effective saturation of 0.7 is reached. We assume the presence of a local lower hydraulic conductivity layer or a local perched water table at a depth of 4-5 m (perched above a deeper low hydraulic conductivity layer). This layer may cause the observed double wetting front. This combined with approximately 30% of entrapped air within the pores may be responsible for the low infiltration rate. Quantification of these mechanisms is an on-going effort. Spatial and temporal distribution of redox was identified and results indicate that nitrate and iron reducing conditions dominate the upper 1 m profile during the flooding (Eh -100 to 200 mV). Once the drying process begins atmospheric oxygen penetrates from the surface and re-oxidation occurs (Eh 500mV). Organic matter content in the soil decreases from ~0.7% in the upper 0.5 m to ~0.2% at the 0.5-2 m depth. Simulations of the infiltration process is in progress; however, the lack of complete saturation and the observed double wetting front have proven difficult to simulate. Our results suggest that the incomplete saturation govern the physico-chemical process along the upper 2-m of the spreading basin. Current efforts are focused on better quantification and simulations of the observed processes as well as more accurately identifying the changes in the hydraulic parameters along the entire soil profile.