The current power supply situation in Nepal experiences a surplus of power during the rainy season, from June to September, while from December to May, there is a power deficit. Storage hydropower projects help to minimize the power deficit, especially in the dry season. Cascade hydropower stations can increase the generation of storage projects, but they also increase the complexity of reservoir operation. In this paper, a Genetic Algorithm optimization model with the objective of minimizing power deficit is proposed to determine the optimal hourly releases from the reservoir with cascade hydropower stations. The model was developed considering the Kulekhani reservoir, which feeds Kulekhani-I, Kulekhani-II, and Kulekhani-III hydropower stations located in the Makawanpur district of Bagmati Province in Central Nepal. The results obtained from the optimization model showed a significant increase in energy generation compared to the actual generation during the dry period. The obtained results were also able to suggest the reservoir levels at different stages of the year for optimal operation of the reservoir. Also, this study demonstrates the usefulness of Genetic Algorithm for solving complex reservoir optimization problems.