In recent years, the importance of DC networks has increased with further development of railway systems, urban electricity networks, and high-voltage direct current (HVDC) systems, to name some examples. Breaking energy recovery, energy storage integration, and flexible management of power flows with the existing AC grids represent some challenges in the development of modern railway substations. An attractive solution to address all these problems in a comprehensive way is a DC microgrid. In this type of microgrids, different technologies are interconnected by using a common DC link. This allows additional degrees of freedom for the system operation and minimizes the price of consumed electricity. However, as the complexity of the control system increases, the stability must be guaranteed regardless the number of devices connected to the microgrid. This chapter presents a detailed implementation of the energy management system of a DC microgrid that consists of primary, secondary, and tertiary controllers in an experimental test-rig. The DC test network consists of four nodes that provide an interconnection between the AC network link, the AC railway network supply, the battery system, and the auxiliary AC supply. The primary and secondary controllers deal with the power control, the energy recovery, and the voltage control. Meanwhile, the economic optimization is carried out at the tertiary level by using a model predictive control (MPC) approach. Also, a detailed stability analysis of the DC microgrid is carried out to analyze the stability limits under different load conditions. The results of the experimental validation are used to demonstrate the properties of each control level, and the economic benefits obtained with the proposed MPC are highlighted. Finally, the small-signal models and the stability limits are validated in the laboratory prototype.