Optimization of Solidification and Cement Stabilization of Electroplating Sludge by Response Surface Methodology and Artificial Neural Network Methods

Solidification/Stabilization (S/S) is used for treatment of various types of hazardous wastes. A systematic study was conducted for treatment of electroplating sludge by Solidification/Stabilization (S/S) with Ordinary Portland Cement (OPC), with the addition of distilled water, lime and magnesium oxide (MgO). Response Surface Methodology (RSM) and Artificial Neural Network (ANN) in combination with Central Composite Design (CCD) were employed to develop the predictive models for simulation and optimization of S/S process. The independent variables were the amount of MgO, electroplating dried sludge, lime, distilled water, while the compressive strength, concentration of zinc and chromium in Toxicity Characteristic Leaching Procedure (TCLP) leachate of the solidified waste after 28 days of hydration were the response variables of 30 specimens. Both RSM and ANN models have been developed based on the experimental designs. The generalization and predictive capabilities of both RSM and ANN were compared by unseen data. Therefore, ten experiments were designed randomly and the predicted responses obtained by ANN and RSM were compared with the experimental results. The results show that the optimized method by ANN or RSM can be used as a confident method for determination of compressive strength of the solidified wastes and concentration of zinc and chromium in TCLP leachates. However, the data calculated by ANN model are more similar to experiments results than that of RSM predicted results. according to RSM and ANN, optimum combination were found 15.3 g, 100 g, 3.6 g, 19 g and 49.5 mL and 14.9 g, 100 g, 3.4 g, 18.5 g, 46 mL of MgO, waste, lime, cement and distilled water, respectively.