The work is focused on the development of a rigorous, model-based approach for the selection of primary controlled variables as part of a plant-wide control system design methodology. Controlled variables should be selected for their self-optimizing control performance and controllability while ensuring satisfactory performance in terms of dead-time and closed loop interactions. This work has considered both-self-optimizing and control performance as well as has addressed issues related to loop-interactions and superstructure constraints. The new three-stage approach developed in this work results in a large-scale, constrained, mixed-integer multi-objective optimization problem. For solving this problem, a parallelized, bi-directional branch and bound algorithm with dynamic search strategies has been developedto solve the problem on large computer clusters. The proposed approach is then applied to an acid gas removal unit as part of an integrated gasification combined cycle power plant with CO2 capture. (C) 2014 Elsevier Ltd. All rights reserved.