Estimating phosphorus removal by steel slag in a flume experiment: Effects of P concentration and subsurface hydrological condition

Excessive phosphorus (P) in water bodies is one of the most concerning water quality issues in the U.S., impacting public health, the environment and the economy. Managing excessive P is essential to reduce the incidence of environmental quality issues, such as eutrophication and harmful algal blooms. One potential strategy that is being developed for this purpose is the use of P sorption materials (PSMs) to sequester P from water systems, which is the objective of this study. The performance of steel slag, an industrial byproduct with high P sorption potential, was evaluated in a nutrient dynamics study using a laboratory flume, where different experimental conditions were tested. The 10-m flume configuration, designed to simulate transport processes occurring in a drainage ditch, is comprised of four 2.5-m sequential segments: a sediment bed, a slag bed over sediment, a slag dam built over a slag bed, and another sediment bed. In the experiments, all four segments of the flume were set to either saturation or a constant drainage (percolation) of 0.1 L min–1 for each segment. The experiment was conducted with a constant flow of elevated P water (1 or 5 mg L–1) at 7.3 L min–1 for 4 hrs (adsorption run), followed 24 hrs later by a 4-hr run of deionized water (desorption run) at the same inflow rate. The adsorption/desorption cycle was repeated three times with the same sediment and slag materials, to allow testing of the resilience of P sorption under different PSM placement, subsurface hydrologic and P loading conditions. During each 4-hour injection, water quality samples and, when appropriate, leachate samples were collected every 6 minutes and 12 minutes, respectively, and analyzed for P and 16 other metals. pH values were manually recorded every 30 minutes. The pH measurements showed that steel slag considerably increases pH in the deionized water used in the study. The concentration of heavy metals and other introduced elements were not detected in concentrations higher than what is normally found in soils. Water flowing through the PSM persistently resulted in a higher P removal when compared to the flow-over situation, as a result of a longer retention time and more contact surface in section 3, the flow-through section. Also, when comparing drainage and saturation, the latter showed a relatively greater retention in both 1 and 5 mg L–1 injections. Over the entire experiment, the retention in 1 mg L–1 saturation and drainage was respectively 33% and 29% versus 45% and 15% when 5 mg L –1 of P was injected. This contradicts the expectations and raises the need for an extra replicate in drainage conditions in future work to confirm the observed pattern. Finally, the leachate samples showed that the P removal is concentration driven, with more mass being transported by drainage flux for the 5 mg L –1 injection. In a real situation, in order to obtain an effective removal, findings from this study show that steel slag must be placed (1) in locations that receive higher P concentrations in runoff, since the removal is consistently concentration driven, and (2) in flow-through conditions, which allows more contact surface, a determining factor for P removal by steel slag