Artificial intelligence for greywater treatment using electrocoagulation process

Treatment of greywater by electrocoagulation using aluminum electrodes was studied. The effects of current-density, electrolysis-time, and inter-electrode-gap on turbidity-removal and electrical-energy consumption were investigated. Under the optimal conditions (J = 12.5 mA/cm2, t = 30 min, and l = 0.5 cm), pollutants removal were: CODtotal = 52.8%, CODsoluble = 31.4%, BODtotal = 32.8%, BODsoluble = 27.6%, SS = 64.6%, TN = 30.1%, and TP = 13.6%. The consumed electrical-energy recorded 4.1 kWh/m3 with an operating cost 0.25 US $/m3. Artificial intelligence was developed to simulate the influence of variables on the turbidity-removal. A 3–6–1 neural network achieved R-values: 0.99 (training), 0.84 (validation) and 0.89 (testing). An adaptive neuro-fuzzy inference system indicated that current-density is the most influential input.