High renewable energy penetration is a goal for many countries to increase energy security and reduce carbon emissions from conventional power plants. Wind energy is one of leading sources among different renewable resources. However, high wind energy penetration in the system brings new challenges to the electric power system due to its variable and stochastic nature, and non-correlation between wind and load profiles. The term non-correlation is used in this research refers to the fact that wind or any other renewable generation, which is nature driven, does not follow the load like conventional power plants. Wind spill is a challenge to utilities with high wind energy penetration levels. This occurs from situations mentioned above and the fact that wind generation sometimes exceeds the servable load minus must-run generation. In these cases there is no option but to curtail non-usable wind generation. This dissertation presents grid-scale energy storage and demand response options as an optimization problem to minimize spilled wind energy. Even after managing this spilled wind energy, there is still a challenge in a system with high wind energy penetration coming from wind power forecast error. Wind power forecast error is handled by having more back-up energy and spilling the non-usable wind power. This research offers a way to use the grid-scale energy storage units to mitigate impacts of wind power forecast error by. A signal processing method is proposed to decompose the fluctuating wind power forecast error signal, based on the fact that each energy storage or conventional unit is more efficient to operate within specific cycling regimes. Finally, an algorithm is proposed schedule energy storage for mitigating both impacts. / Ph. D.