The optimization of combined cooling heating and power (CCHP) systems involves two major tasks: searching for optimum design parameters and for optimum regular operation variables. This paper proposes a two-stage method to solve both tasks. The operation of large thermal power plants must be altered smoothly, as quick changes in system settings may result in cascade tripping of subsystems, ultimately leading to a complete shutdown. This work uses graphical representation of the operational space of the system, which helps in tracking the operation along its optimum trajectory smoothly. The daily energy demands of a five star hotel, collected over a year, were used to demonstrate the applicability of the proposed method. Using the proposed method reduced the total annual cost over 7% and 13% in Australia and Sri Lanka respectively, compared to the conventional method of following thermal load. (C) 2013 Elsevier Ltd. All rights reserved.