The unpredictable nature of wind energy makes its integration to the electric grid highly challenging. However, these challenges can be addressed by incorporating storage devices (batteries) in the system. We perform an overall assessment of a single domestic power system with a wind turbine supported by an energy storage device. The aim is to investigate the best operation mode of the storage device such that the occurrence of large power spills can be minimized. For estimating the small probability of large power spills, we use the splitting technique for rare-event simulations. An appropriate Importance Function for splitting is formulated such that it reduces the work-load of the probability estimator as compared to the conventional Crude Monte Carlo probability estimator. Simulation results show that the ramp constraints imposed on the charging/discharging rate of the storage device plays a pivotal role in mitigating large power spills. It is observed that by employing a new charging strategy for the storage device large power spills can be minimized further. There exists a trade-off between reducing the large power spills versus reducing the average power spills.