Combined cycles are inherently modular, since they are composed of one or several gas turbines cycles coupled with a steam cycle. Thus, it is often possible to consider repowering of conventional steam power plants by replacing the boiler with a gas turbine and a heat recovery steam generator. It constitutes an interesting option to increase power outputs and efficiencies and reduce specific emissions at moderate costs due to the reuse of existing parts and plant infrastructure. The goal of this study was to determine the best options for such a conversion in the case of one of the two independent 150 MWe turbo-groups of the steam power plant in Chavalon, Switzerland. For this purpose, models of heat recovery steam generators with simple and double evaporating pressure levels and models of steam turbines were developed and associated to a catalog of gas turbines available in the present market. These submodels were grouped into a superconfiguration of possible combined cycles in order to perform a simultaneous optimization of configuration, design and operating parameters. This is a mixed integer non-linear problem (MINL) that was solved with a genetic algorithm. For the real design variables such as power ratings of the main components, the algorithm uses real variables genes instead of a binary mapping to increase accuracy. In order to reduce the optimization time, a pre-assessment applying pinch methodology was completed, in order to reduce the number of alternatives in the superconfiguration.