This thesis studies the consequences of hydrological risk for hydropower generators in Brazil and proposes four models to mitigate this risk. In the first part, a model is developed considering the redesigning of hydro energy contracts so that the generators are allowed to sell part of their energy per availability. This solution allows the financial settlement of the accounts to more closely mimic the operation of the system, where energy sources can be optimally switched between hydro and thermal plants depending on water availability. In the second part, three derivative options, namely swap, collar and collar by difference are developed. In the swap model, the generators define a protection level and from that level they exchange energy with the consumers through a power delivery factor. In the collar model, the generators define a generation floor to protect themselves against financial losses in the market in such way they would be long at a high spot price. Also, the generators define a generation cap from which they would be short at a low price. The collar by difference is a collar formed by two distinct agents through a trader. In this case, the generators set an option with an exercise value, while the customers set an opposite position with that same exercise with the trader. To test the efficiency of these four models, two thousand monthly simulations of energy price and power generation for 2016 were generated. The results indicate that by adding flexibility to the contracting rules, hydrological risk is mitigated. Also, the results suggest that the generators with existing contracts may respectively opt for the swap, collar by difference and collar derivatives, while the generators with new energy would opt to sell a portion of their energy in contracts by availability. In the long-term energy tariffs may decrease in future energy auctions due to the reduction in the risk to hydropower generators under these models.