Abstract:
Present study investigates Ni and Mg removal from electroplating wash-water (EWW)
employing electrocoagulation (EC) method using aluminum electrodes. X-ray diffraction (XRD),
energy dispersive spectroscopy (EDAX), scanning electron microscopy (SEM) and thermo
gravimetric analysis (TG/DTA) have been performed for morphological study of sludge and
electrodes. Concentration of Ni and Mg were found to be 583.5 mg/I, 483 mg/I in sample which
were used as initial concentration for experimentation. The simultaneous effect of various
operating parameters like current density, pH, reaction time and electrolyte concentration were
evaluated and optimized using central composite design (CCD) in response surface methodology
(RSM). Economic analysis was also carried out by incorporating the cost analysis parameter.
The efficacy of EC process was elucidated in terms of Ni and Mg removal, COD removal,
specific energy consumption and specific operating cost. Interaction effects between input
parameters and response variables were evaluated by using 3 dimensional response surface
analysis and Second-order models were validated by ANOVA.
Prediction model for Ni, Mg, COD removal, and specific energy consumption, specific operating
cost were generated applying back propagation training to ANN. For hidden layer neurons, Log-
Sigmoid (LOGSIG) transfer function was used whereas for output layer neurons linear
(PURELIN) transfer function was used. The ANN structure (4 inputs, 20 neurons, 1 hidden
layer, 5 outputs) adequately fitted the experimental and predicted data with a regression
coefficient (R2= 0.99968) and MSE 0.0 1206. ANN model was validated using three new set of
experiments.
For RSM, Predicted yields were in a good agreement with the experimental ones. The optimum
values for the selected independent variables were found to be current density86.23 A/rn2,
pH=7.5, operating time60 min and NaCl dose of 0.5 g/l. At these optimum conditions, Ni
removal of 97% and Mg removal of 90% was achieved with energy consumption 22.45 kWh per
kg Ni removed. The specific operating cost for the whole process was estimated to be Z 227 per
kg Ni removed. The RSM and ANN employed in the study predicted the responses adequately.
The process was thus established as high performing and cost effective for the treatment of
electroplating industrial waste water.
