Abstarct: In this thesis, the energy management problem of a virtual power plant in energy markets cosidering contingency conditions is formulated as a stochastic robust optimization model. The virtual power plant includes thermal and renewable generating units, energy storage units and price-responsive demands interconnected through an electric network. Additionally, the smart grid technology provides a bi-directional communication infrastructure by which the energy management system receives required information near real-time, and sends decisions to the day-ahead and real-time energy markets, as well as to the components in the virtual power plant. To tackle uncertainties, a combination of scenarios and prediction intervals is used. This way, the uncertainties in the availability of the components in the virtual power plant and the market prices are presented by scenarios, while the renewable power production is modeled by prediction intervals. The proposed approach is implemented in 5-bus and 24-bus test systems. The results verify the applicability of the decisions considering contingency conditions.