Arsenic Removal from a Highly Contaminated Groundwater by a Combined Coagulation-Filtration-Adsorption Process

The aim of the present experimental study was to perform a technical-economic evaluation of a combined treatment system, consisting of coagulation-flocculation or rapid sand filtration as pre-treatment followed by column adsorption, for reducing the arsenic concentration from approximately 1 mg/L to below the limit set for groundwater remediation and drinking water, i.e., 0.01 mg/L, according to the legislation in force. A wide number of operating conditions were experimentally evaluated in the different tests. In the coagulation-flocculation study, it was initially investigated if the iron contained in a mining drainage co-mixed with the groundwater would be able to achieve a better As content reduction by adsorption/precipitation, thus avoiding fresh coagulant addition. Then, different polyelectrolyte dosages were tested varying the mixing ratio. None of the tested conditions allowed to improve the arsenic removal so significantly to warrant the consequent incremental costs. Therefore, the optimal condition was considered any mixing with a different liquid stream and any polyelectrolyte dosage. The iron content naturally present in the groundwater and contact with air was capable alone of reducing As concentration of about 80%. Sand filtration reached approximately the same removal efficiency (about 80%) at the lower surface loading rate among the values tested. Between coagulation and sand filtration, in terms of costs, the latter showed to be more convenient than coagulation-flocculation, at the same removal efficiency: therefore, it was considered the optimal pre-treatment. The following adsorption column plant was capable of further reducing As concentration up to the required value of 0.01 mg/L. Among the two iron-based commercial adsorbents applied in the adsorption column tests, the hybrid media consisting of an exchange resin with iron oxides showed to be preferable under the selected operating conditions: it offered higher adsorption capacity at breakthrough and, after exhaustion, could be regenerated for a number of cycles. The influent pH showed to have a great influence on the duration of the adsorbent media, and values around neutrality were considered preferable. The estimated cost of the full treatment was computed to be about 0.50 €/m3 of purified water. Therefore, the capacity of achieving the required remediation goal, the limited cost, and simplicity of operation make the proposed combined treatment being potentially suitable for real application.