Prediction of enhanced coagulation with varied pre-oxidations for seasonal variations of cyanobacteria-dominated algae-laden water quality

Enhanced coagulation with pre-oxidation is a cost-effective approach for managing seasonal water quality pollution caused by algal outbreaks and deaths. The synergistic application of pre-oxidants and coagulants, coupled with intelligent and precise dosage control, constitutes a prominent research focus in water treatment field. This study evaluates the removal of various pre-oxidants, including potassium permanganate (KMnO₄), KMnO₄ composites (PPC), and potassium ferrate (K₂FeO₄), in combination with coagulants like polyaluminum chloride (PACl) and aluminum sulfate (Al₂(SO₄)₃). A machine learning algorithm, based on the least squares support vector machine (LSSVM), was developed to predict optimal dosages. After 15 months of source water quality monitoring, algal contaminations predominantly driven by cyanobacteria-dominated became worse particularly in the high temperature and algae period and autumn, which were positively correlated with UV₂₅₄ and COD_Mn (p < 0.05). The doses of PACl and Al₂(SO₄)₃ were between 100 μM and 120 μM (calculated as Al) across various periods to efficiently remove organic matter. Under optimal chemical dosages, pre-oxidation facilitated the protein-like substances removal. The removal efficiency of PPC surpassed that of KMnO₄ and K₂FeO₄. The LSSVM model demonstrated superior predictive performance for dosages compared to other models like random forest (RF) and back propagation (BP) neural networks, with feature importance analysis identifying water temperature, UV₂₅₄, and conductivity as the core parameters for real-time dosing systems. This study elucidated dosing strategies alongside algae contaminations removal associated with pre-oxidation enhanced coagulation while proposing a methodology for dynamically adjusting oxidant and coagulant dosages through real-time monitoring of both raw water quality and effluent from coagulation-precipitation processes, thereby providing novel insights into precise real-time dosing for chemicals in water treatment facilities.