COMPREHENSIVE ASSESSMENT OF WATER QUALITY AND SLUDGE CHARACTERISTICS AT THE OLD KUFA WATER TREATMENT PLANT USING ARTIFICIAL NEURAL NETWORK MODELING
This study delivers an integrated assessment of water quality and residual sludge characteristics at the Old Kufa drinking-water facility in Iraq. The analysis targets the full treatment train of the municipal plant together with the generated water-treatment residuals. Operational records from 2020–2024 and laboratory measurements collected between June and November 2024 were used to trace the behavior of key indicators—turbidity, total suspended solids, and chemical oxygen demand. We further examine how fluctuations in Euphrates source-water quality and coagulation-flocculation settings condition the specific sludge yield. Measured values are benchmarked against current Iraqi drinking-water and environmental standards, and the operational implications for sustainable sludge management are outlined, in cluding exceedance risk appraisal, reagent-dose optimization planning, and avenues for reuse and resource recovery. The modeling component employs an artificial neural network with inputs pH, Cl–, NO₃–, NH₄+, and temperature, and targets defined as suspended solids and specific sludge production; the model attains high predictive accuracy (R² = 0.991). The novelty lies in coupling field and experimental evidence from an urban WTP with ANN-based forecasting of treated-water indicators and sludge generation under variable river inflow. The outcome is a practical decision-support tool enabling sensitivity analysis to influent attributes and dosing strategies. Turbidity and suspended solids emerge as the dominant determinants of sludge formation; a peak specific yield of 278.6 kg /1,000 m³ was observed, motivating adjustments to chemical conditioning and proactive sludge-handling plans. COD exceedances above the 100 mg/L limit indicate environmental risk and the need to reinforce operational resilience through optimized dosing and development of reuse/recovery pathways.