Adsorption is an economical and efficient process commonly used in industry and wastewater plants for separation and purification applications. In the current theoretical study, simultaneous batch adsorption of compounds representing different sources of pollutants was investigated using different grades of activated carbons. The total percentage removal of the solutes from the aqueous stream was found to be significantly influenced not only by mixture compositions and process variables (pH, type of carbon and adsorbent dose) but also by interactions between them. Ridge analysis using an improved algorithm developed in-house, traced the locally optimal percentage removal solutions along the direction of steepest ascent towards the eventual global optimum value. Since it was impossible to meet the utopian condition of simultaneously maximizing percentage removal and adsorbent loading, compromise solutions had to be identified. Typical multi-objective optimization methods such as weighted sum, non-dominated sorting genetic algorithm and pattern search could not identify the Pareto fronts satisfactorily. A new method for approximating the true Pareto front for bi-objective optimization problems is proposed and validated. This method could effectively find the Pareto solutions that were not only closest to the utopian point but were more regularly arrayed with fewer discontinuities in the objective space.