Commonly in designing hydroelectric dams and in the level of consulting engineers, for determining the hydroelectric potential of an storage, the analyze is done based on an RBS (Reliability Based Simulation) model with an obvious operation policy. In this paper the effect of optimizing release rules instead of using a fixed and know release policy, like what it is adopted in an RBS model, was explored through developing a dynamic programming (DP) optimization model and Karoon 5 dam was used as a case study. DP evaluates energy potential while maximizing total annual energy generated as its objective function. In the case where no reliability constraint is considered, DP presents such results having lower reliability in meeting energy demands and lower vulnerability index with a slightly improved firm and total annual energy, compared to RBS results. The DP model was then completed as a reliability DP. This reliability model can control the reliability level in meeting energy demands like what it is done in RBS model. Comparison between reliability DP and RBS models showed that DP results are better than RBS results in term of firm and total energy generated and installed capacity. However, as the reliability level increases, the out performance of DP over RBS model decreases. This is because by adopting a more strict reliability constraint, the optimization problem will get more and highly constrained and there is little room left in this case for optimizing an objective function.