The new Brazilian Energy Model brought a whole new world of challenges to Distribution Companies (DISCOS). They have to meet the load through longterm contracts (five to fifteen years). Uncertainties are enormous, from load evolution to energy availability and associated prices. Restrictions are tight - for instance overcontracting is punished - only a mere 3% excess may be included in the regulated tariff and thus recovered by consumer´s prices. DISCOS must therefore build a contract portfolio under uncertainties, minimize costs (searching for a competitive price) and risks (searching for long-term guarantees). The objective of this work is to help the agent constructing its optimum portfolio - which corresponds to a large-scale non-linear mixed integer problem, in principle difficult to solve by available models (from economic to optimization algorithms). In this model economic and optimization techniques have been combined into a new model able to both analyze a possible portfolio and/or find a set of optimal solutions. The model may be adapted to accommodate different objective functions, such as minimum risk, minimum cost, VaR (Value- at-risk), etc. This model may offer an efficient and flexible, yet precise tool, able to help finding a fair trade, mitigating risks and minimizing costs, leading to a better use of energy and a better price for whole society.