Implementation of statistical response surface methodology with desirability function for ion-exchange-based selective demineralization of municipal wastewater and tap water for drinking purposes

This study has significant implications for water treatment and environmental engineering, demonstrating a successful reduction in the concentration of selective minerals from municipal wastewater and tap water through a demineralization process. The use of multi-objective response surface methodology with a desirability function underscores the importance of these findings. Temperature, resin depth and pH were selected as independent factors, while, hardness, concentrations of cations (calcium, magnesium, manganese), conductivity and total dissolved solids were the dependent variables. Individual quadratic regression models were developed for each dependent variable and water sample, yielding high coefficient of determination and low relative error, mean absolute error and mean squared error values. Using a multi-objective optimization approach, optimal values for demineralization were achieved and validated experimentally, showing good agreement between measured and predicted values. Analysis of variance analysis revealed that all independent variables were significant (p < 0.05) and notably affected all responses. The high combined desirability values for both samples indicate that the set of optimal conditions was effective in minimising all responses. The reasonable coefficient of determination for all models, along with the low values of statistical performance indicators suggest that the laboratory test data fit the predicted response values. Both water samples achieved an optimal removal efficiency greater than 95%, with the maximum values of conductivity, cations, and total dissolve solids. The removal/reduction efficiencies of this work were higher than previous published results. This superior performance can be attributed to the efficient optimization of the treatment process through a combination of mathematical modelling and experimental approaches.