The extensive integration of renewable energy sources in the future electric distribution system will bring critical operational challenges. The expected rise in uncertainty due to intermittent generation and associated physical complexity will play a crucial role in the energy management of such systems. The underlying reason of this complexity is the uncertain and volatile nature of the problem. In order to cater for the reliability in this scenario additional investment and operational expenses will be required. This highlights the need to revisit the uncertainty management methodologies in order to better trade-off between reliability and cost. In this thesis a stochastic optimization methodology is proposed in order to facilitate the handling of increase in uncertainty level. The approach will provide guideline for the optimal reserve scheduling under uncertainty. Chance constrained optimization will be performed at the generation sub-level to optimize the cost of the generation reserves. The outcome will be the probabilistic guarantees against constraint satisfaction that will be related to a performance envelope around the operational point of the generation unit. The second sub-level of the distributed structure will be the load control. Increasing the share of the controllable load can increase the confidence against uncertainty. A methodology for providing insight in the level of the controllable load and the associated degree of confidence can help in dealing with the uncertainty in the system. In this work the controllable load as reserve shall be investigated for decreasing the operational reserve cost. Microgrid will be used as a benchmark application for the proposed approach for both grid connected and island modes. This work will lead to the in-depth analysis of the capability of the Microgrid and optimal allocation of reserves under various level of uncertainty in the forecast variables.