This study presents a new generation expansion planning (GEP) incorporating the effects of renewables variable generation on thermal power plants efficiency. An hourly unit commitment problem was integrated in the GEP problem with the overall goal of supporting the selection of future mixes of power plants through long term planning. The problem resulted in a binary mixed integer non-linear cost optimization model. The model application was demonstrated for the design of electricity plans for a 10 year planning period under different CO2 assumptions for a thermal, hydroelectric and wind power system. The results were compared with the ones obtained using a traditional GEP model, which assumed average operating conditions for thermal power plants. The scenario analysis shows that the impact of renewables variability on the performance of thermal power plants and on the generation expansion planning is non-negligible. The results suggest that assuming average operating conditions can result on the underestimation of the system costs which highlights the importance of the proposed integrated model to strategic decision making. (C) 2017 Elsevier Ltd. All rights reserved.