Response surface methodology (RSM) has been used to optimize the effect of circulation rates, ozone supply, cephalexin (CEX) concentration, and granular activated carbon (GAC) dose on removal of COD from solution. According to statistical analysis, all of the input variables exerted significant influence on COD removal, however, the effect of interaction variables was not found to be significant on comparative basis. Further, the developed quadratic regression model based on obtained results emphasized the significance of individual terms and little of interaction terms. The values of r2 (0.959), adjusted r2 (0.902) obtained by analysis of variance (ANOVA) indicates the significance of quadratic model in predicting desired response. The maximum of 70% of COD was removed in these experiments and optimized value according to main effect of variables was 60%.


Javaid Akhtar, Noraishah Saidina Amin and Muhammad Khurram Zahoor