The industrial separation of p-xylene was considered by means of a standard SMB (Simulated Moving Bed) system as well as the Varicol mode of operation. Both modes of operation were simulated accounting for adsorbent capacity decline using the TMB (True Moving Bed) approach in the case of standard SMB and by a TMB with a non-integer number of columns per section in the Varicol case. Additionally, the TMB model takes into account variable fluid velocity due to the adsorption/desorption rate of p-xylene and its isomers. Dynamic results were used to study the influence of the adsorbent capacity decline on the SMB unit performances and two different compensating measures were tested for a 10 years period: (i) switching time decrease (solid flowrate increase) and (ii) columns redistribution. Both strategies kept the initial purity requirements and reached higher productivity values than the respective unit working without any compensation measure. The switching time compensating strategy proved to be more efficient than the second one, achieving, for the same purity requirements, higher productivity values.