The objective of this project was to develop a novel phosphorus removal system using steel slag filters applicable in decentralized applications and to propose design criteria about maintenance needs. Slag exhaustion functions were measured on 2-3 mm, 3-5 mm, 5-10 mm and 16-23 mm slag. Three steel slag columns with particle size of 2-3 mm, 3-5 mm and 5-10 mm were fed with the effluent of an aerated lagoon during 589 days. A barrel reactor test was fed during 365 days with the effluent of an attached growth aerated biological reactor. The o-PO4 concentration at the effluent of the 2-3 mm and 3-5 mm columns and barrel reactor test was between 0.04 and 0.3 mg P/L. Particulate phosphorus, however, was removed by about 50%. The P-Hydroslag model implemented in PHREEQC was successfully calibrated with data from the column test, and validated with data from the barrel reactor test. The calibrated model was used to simulate long-term operation of a slag barrel reactor with two parallel streams of five replaceable steel slag barrels, with total hydraulic retention time of voids of 15 h. The system longevity was strongly influenced by the influent alkalinity. The simulated longevity was 7 years with an influent alkalinity of 50 mg CaCO3/L and 2 years with an influent of 210 mg CaCO3/L. The alkalinity of the steel slag filter influent was influenced by the type of aquifer supplying drinking water, the presence of nitrification activity and by the CO2 concentration in the enriched air of the upstream biological process. Simulated scenarios with partial barrel replacement (e. g. barrels 1 and 2 out of 5 replaced at frequency of 0.5, 1, 1.5, 2, 2.5, 3, 3.5 or 4 years) increased the system longevity up to 14 years while slightly increasing the number of barrels needed.
Keywords: Alkalinity; P-Hydroslag; PHREEQC; Physicochemical modelling; Reactive filter.
Copyright © 2018 Elsevier Ltd. All rights reserved.