Prediction of adequate pH and Mg2+dosage using an empirical MgO solubility model for struvite crystallization

Abstract

The struvite crystallization process can recover struvite crystals as valuable slow-release fertilizers from raw water, including high phosphate and ammonium ions. MgO can be used as a magnesium source and an alkaline chemical simultaneously; however, it is difficult to control the dissolved Mg 2 + concentration because of its insufficient and unpredictable solubility. The purpose of this study was to estimate MgO solubility and determine the optimal operating conditions of pH and Mg 2 + dosage during struvite crystallization. An empirical MgO solubility prediction model was devised to dose an adequate amount of MgO slurry in accordance with the properties of raw water. MgO solubility was determined by H + ions donated by NH 4 + , HCO 3 − , and H 2 PO 4 - as the pH increased in the solution. It was found that the molar ratio of donated H + to dissolved Mg 2 + converged to 2:1. This model was verified by performing jar-tests using digested sludge filtrate. When adjusting to a pH of 8.0–9.3 using only MgO, the predicted Mg 2 + concentrations using the model were compared to the measured Mg 2 + concentrations, and they coincided well within the maximum error of 5.7%. This model was applied to determining the operating conditions of pH and Mg 2 + dosage. As a result, considering phosphate removal efficiency higher than 90%, it was feasible to simultaneously achieve an Mg/P molar ratio of 1.3–1.5 and a pH of 8.2–8.5 by only using MgO. Therefore, MgO is recommendable as an appropriate magnesium source and alkali supplier in the struvite process based on experimental jar-tests and the empirical MgO solubility model. • An empirical MgO solubility prediction model was devised and applied to struvite crystallization process. • MgO solubility was predicted using pH, and concentrations of acidic ions. • The optimal conditions of pH and Mg 2 + dosage were determined by this model.