An innovative approach for the optimization of general cyclic adsorption separation processes (PSA and VSA), using the Skarstrom cycle with equalization, is proposed. The main objective is to study the optimal operating conditions for different adsorbent types, the optimal equalization stage configurations, and different underlying assumptions that are usually presumed in process modeling and simulation. The partial differential equations corresponding to the bulk gas-phase mass balances were solved with a recent numerical technique, developed within our group, based on an adaptive multiresolution approach. The partial differential equations for the intraparticle balances were solved using the orthogonal collocation method. The proposed dynamic simulator proved to be very efficient in terms of computational time and memory requirements: stability and accuracy are simultaneously ensured. The optimization was performed with a successive quadratic-programming algorithm that proved to be efficient because the optimum can be reached in a modest number of iterations. The proposed approach was applied to the particular case of oxygen production from air by PSA and VSA, with the aim of studying the performance of two commercially available adsorbents, commonly used in this specific separation, and the best configuration during equalization stage. (c) 2005 American Institute of Chemical Engineers.