The candidate areas reviewed for river bank filtration are scattered over tens of kilometer. To increase efficiency and cost-effectiveness, screening those candidates and selecting the optimum districts must be preceded before implementation. Based on...
The candidate areas reviewed for river bank filtration are scattered over tens of kilometer. To increase efficiency and cost-effectiveness, screening those candidates and selecting the optimum districts must be preceded before implementation. Based on the previous research data and information, a set of standards to conduct electrical specific resistance, boring tests, and pumping test was prepared. These surveys and tests were performed according to the standards. Based on the survey and test results, hydraulic conductivity, aquifer thickness, and the length for bank filtration were calculated. By assuming the use of horizontal collection wells, potential intake volumes were calculated through the primary MODFLOW modeling.
As hydraulic conductivity, thickness of the major aquifer, and length are the most important factors in calculating intake volume, these three elements were given specific weight values to score the candidate districts. Using the scores, the candidate districts were ranked. The ranking was compared with the ranking acquired through the primary MODFLOW modeling to adjust the scoring methodology. From the ranked candidate list, pumping tests were conducted at the top 3 districts again to calculate accurate hydraulic conductivity. Those districts were scored again and the scores were compared with the secondary MODFLOW modeling results.
Using the groudwater modeling program, FEFLOW, a model to estimate intake volume was prepared. Studies on the entire modeling processes from survey, and modeling to calibration, were conducted. The results, including the Definition of a modeling area, mesh and aquifer composition, FEFLOW modeling composition, application of hydraulic constants were analysed. The model was adjusted using the behaviour of static groundwater data before bank filtration; and calibrated using unsteady groundwater flow level to the function of timely changes of water level of the Nak-dong River. Groundwater flow was modeled by assuming 10 radial horizontal collection wells to evaluate intake volume.
Additionally, in terms of water quality, oxidation processes of iron and manganese when bank filtered water with low DO and high iron and manganese is combined with water from the Nam-gang Dam with high content of DO were studied using EPANET-MSX (multi-species extension). Equations for reactions were developed and for the first time in Korea, by building a numeric model, concentrations of iron and manganese were predicted.