Fuzzy Inference System for Modeling the Contribution of the Sweep Mechanism in the Coagulation Process for Water Treatment

Coagulation is a critical step in water treatment, with polyaluminum chloride being a commonly used coagulant. Its main mechanisms are charge neutralization and sweep coagulation. Minimizing sweep coagulation, when appropriate, can reduce chemical use and improve sludge management. A Fuzzy Inference System (FIS) is an effective tool for handling uncertainties and is often used for simplified process modeling. In this study, a FIS was developed to estimate the contribution of the sweep mechanism using zeta potential, insoluble aluminum content, and turbidity removal as inputs. The model used three input variables and nine rules, aiming for low computational demand. The highest estimated sweep contribution occurred at pH slightly basic and Al(s) concentration > 2 mg L⁻1 in a natural low-turbidity water sample. The FIS enabled the integration of frequent measured variables into a single numerical output estimation linked to coagulation mechanisms, supporting decision-making and enabling opportunities for automation and cost reduction.