Model-based analysis of struvite recovery in WRRFs using a low-grade magnesium oxide by-product as an alternative magnesium source

Abstract

Abstract

This research presents a dynamic, mechanistic precipitation model describing struvite formation from the dissolution and reactivity of an industrial LG-MgO by-product, including interactions with competing ions such as calcium. The study addresses a key limitation of existing struvite precipitation models, namely their inability to represent the behaviour of heterogeneous and low-reactivity magnesium sources under realistic operating conditions. Calibration of kinetic parameters for LG MgO based systems is particularly challenging, as no previous models have been reported for such magnesium sources. The results demonstrate that struvite precipitation dynamics can be adequately described using a mechanistic approach with a single kinetic constant for dissolution and another for precipitation. The model was calibrated using batch experiments with anaerobic digestion supernatants and subsequently integrated into a plant-wide model of a full-scale wastewater treatment plant. Beyond reproducing experimental observations, the framework enables the evaluation of LG-MgO-based struvite precipitation under realistic and variable operating conditions, supporting its integration into full-scale WRRF design and optimisation. This work highlights the role of dynamic mechanistic modelling as an enabling tool to bridge the gap between experimental studies and practical implementation. Application to the Arazuri WRRF demonstrated that effluent total phosphorus concentrations can be reduced to 0.6 mg P·L-1, while optimal operation was achieved at Mg:P ratios between 1:1 and 2:1 and pH≈8.0, balancing phosphorus removal and reagent consumption. Compared to conventional industrial standard reagents, LG-MgO enables a 90% cost reduction. Overall, the model provides a basis to assess process performance, economic trade-offs, and system-wide impacts prior to full-scale implementation.