This paper presents a method to structure an optimal portfolio of electric self-generation with renewable energy sources (RES), considering the Response Demand Program-Time of Use. An optimization model of demand and of simulation-optimization of the generation costs of the project were proposed, through software tools based on simulation techniques Monte Carlo and Genetic Algorithms. The price electricity and the generation were modeled as stochastic variables. The expandend actual value is defined with Theory of Real Options and the optimal portfolio is calculated by minimizing of the costs. The relevance of the method and the viability of RES in the Colombian commercial sector are demonstrated. A portfolio for a case study of a shopping center where photovoltaic sources constitute the largest contribution was structured.