|M.Sc Student||Ben Moshe Shany|
|Subject||The Effect of Operational Dynamics on Biogeochemical|
Efficiency of a Soil Aquifer Treatment System
|Department||Department of Civil and Environmental Engineering||Supervisors||Professor Alex Furman|
|Professor Noam Veisbrod|
|Full Thesis text|
Sustainable irrigation with treated wastewater (TWW) is a well-established solution for water scarcity in arid and semi-arid regions. Soil aquifer treatment (SAT) provides a solution for both the need for tertiary treatment and seasonal storage of wastewater. Stresses over land use and the need to control the obtained water quality makes the optimization of SAT of great importance. This study looks into the influence of SAT systems' operational dynamics (i.e. flooding and drying periods) as well as some aspects of the inflow biochemical composition on their biogeochemical state (mainly dissolved oxygen (DO) concentrations and oxidation-reduction potential (ORP)) and the ultimate outflow quality. A series of four long-column experiments was conducted, flooding periods (FP) were kept constant at 60 minutes for all experiments while drying periods (DP) were 2.5 and 4 times the duration of the flooding periods. To examine a wider range of DP durations as well as different inflow compositions, a numerical model was developed and calibrated. The model includes both description of the water flow in the system and reactive solute transport, considering the main biogeochemical reactions that are dominant in SAT.
Our experimental results show that the longer DP (240 minutes) had a significant advantage over the shorter periods (150 minutes) in terms of DO concentrations and ORP in the upper parts of the column as well as in the deeper parts, which indicates that larger volumes of the profile were able to maintain aerobic conditions. DO concentrations in the deeper parts of the column stabilized at ~ 3-4 mg/L in the longer DP compared to ~1-2 mg/L in the shorter DP. This advantage was also evident in outflow composition that showed significantly lower concentrations of NH4-N, dissolved organic carbon and total Kjeldahl nitrogen in the outflow for the longer DP (~ 0.03, ~1.65 and ~ 0.62 mg/L respectively) compared to the shorter DP (~ 0.5, ~ 4.4 and ~ 3.8 mg/L, respectively). Comparing experimental ORP values in response to different DP to field measurements obtained in one of the SAT ponds of the SHAFDAN, Israel, we found that despite the major scale differences between the experimental 1D system and the field 3D conditions, ORP trends in response to changes in DP, qualitatively match. The numerical model was successfully calibrated to the experimental data and sensitivity analysis revealed very high sensitivity to soil-water retention parameters. Using the model, we were able to demonstrate increased rates of microbial-mediated kinetic oxidation reactions which verify our experimental observations. A simple optimization process was performed based on simulation results and local optimum for the experimental SAT system was found at a DP duration of ~ 2.8 times the FPs.
We conclude that longer DP not only ensure oxidizing conditions close to the surface, but also enlarge the active (oxidizing) region of the SAT profile. This suggest that SAT should be treated as a pseudo-reactor that to a great extent could be manipulated hydraulically to achieve the desired water quality while increasing the recharge volumes.