Phosphorus removal mechanisms in soil and slag media

University of Technology, Sydney. Faculty of Engineering.Excessive phosphorus (P) is one of the major pollutants in natural water that are responsible for algal blooms and eutrophication. P removal by soil and slag is an attractive solution if the P sorption capacity of soil or slag is significant. To design an efficient land treatment facility, basic information on the behaviour of P in the media-water environment is required. In this study, detailed experiments were conducted to study P removal under static and dynamic conditions, and mathematical models were developed to describe these processes. The kinetic studies on P sorption onto a sandy loam soil from North Sydney, Australia, and dust and cake waste products from the BHP steel industry revealed that P sorption is a slow process. More than 90% of the P was adsorbed within 70, 12 and 60 hours in a mixing system for soil, dust and cake respectively while it was within 240, 24 and 120 hours respectively in a static (no stirring) system. Dust adsorbed P the most, compared to the other adsorbents (220-225 times and 4-5 times of the sorptivity of soil and cake respectively). P sorption in the batch experiments was described better by an equation using the Langmuir isotherm than one using the Freundlich isotherm. The kinetics of P sorption were satisfactorily explained by a static, physical, non-equilibrium sorption model (SPNSM). The pH of the P solution played a critical role in the extent of removal and the removal mechanisms of P. Removal was at a minimum at pH 2. The effect of pH on P removal varied depending on the type of adsorbents and the initial P concentration. The dominant removal mechanism of P at pH10. Batch flocculation experiments revealed that the P removal efficiency increased with an increase of adsorbent dose, flocculation (contact) time and mixing rate. P sorption is affected by the presence of NH₄ which competes for available sites on the adsorbents. The amount of P adsorbed by dust and cake in the presence of NH₄ was less than that in a single solute system. The reduction percentage of P for dust ranged from 33 to 57%. Detailed column experiments conducted with soil, dust and cake as media indicated that dust and cake have much higher sorption capacities than soil. The solid phase P concentrations on dust and cake calculated from batch experimental isotherm constants are substantially higher than those estimated from the dynamic column experiments. A program to fit a generalised, non-linear, advective, dispersion model (MCMFIT) was used to estimate the optimal model parameter values. All the columns exhibited characteristic S-shaped or curvilinear breakthrough curves. The simulated results from a dynamic physical non-equilibrium sorption model (DPNSM), based on Freundlich isotherm constants K and N (obtained from column experimental results) satisfactorily matched the corresponding experimental breakthrough data. P removal under dynamic conditions was explained better as a physical non-equilibrium process than an equilibrium one. The mobility of P is restricted by the adsorbing medium and it is proportional to the sorption capacity of the adsorbents. The DPNSM predictions were better than those of the equilibrium sorption model (ESM). The ESM predictions were similar to DPNSM predictions at high influent P concentrations for dust and cake. An increase of pH value occurred at the initial stage of P removal. This is due to the liberation of Mg²⁺ and Ca²⁺ ions into the solution from the dust and cake media. The sorption capacity of P in dust and cake columns decreased in the presence of NH₄ due to the competition for sorption sites. Dust and cake can be applied (i) as an adsorbent in sewage treatment plants, (ii) as a substratum in constructed wetland systems and (iii) as a suppressing material for the release of P from sediments due to their high sorption capacity. The effluent concentrations of heavy metals in the dust and cake column were low enough not to harm any living organisms