Abstract
The integration of renewable energy sources in the power grid results in more distributed power generation relative to the traditional grid structure. The new grid characteristics have caused the concept of microgrids to attract more attention. A central advantage of microgrids is that they can be operated in both stand-alone and grid-connected mode. Since they are able to operate in stand-alone mode, implementation of microgrids is seen as a good approach to enforce electrification in rural areas. To enhance the reliability and flexibility of the microgrids in these areas, where connecting to the utility grid might not be possible, interconnecting several microgrids to form a multigrid configuration is advantageous. Interconnecting microgrids arise several technical difficulties, and a vital challenge is to provide a seamless transition from stand-alone to interconnected operation. Upon interconnection, the microgrids need to be synchronized to avoid high inrush current transients at the interconnection moment. The purpose of this thesis is to propose a synchronization technique to enable a seamless transition from stand-alone to interconnected operation with two microgrids. The proposed technique centers around synchronizing frequency and phase angle of one of the microgrids to match the other microgrid. The synchronization control commands are obtained from grid voltage measurements of both grids. The voltage measurements are used to calculate the phase angle and frequency deviations between the two grids. The deviations are passed through parallel PI controllers and added together to form a frequency offset signal, which is passed on to the synchronizing microgrid. The frequency offset from the synchronization control loops is added to the speed reference of the synchronous generator unit in the microgrid. The proposed synchronization technique is tested in simulations. The simulation model consists of two simple microgrid models, both consisting of a synchronous generator, a converter-based generating unit and a local load. The two modeled microgrids are connected through a switch. The synchronization control loops are located at the switch, using grid voltage measurements at each side of the switch. A simulation was conducted on a reference scenario. This showed that the synchronization loops successfully eliminated the deviations in phase angle and frequency between the two microgrids, and thus prevented high inrush currents at the interconnection moment. To test the robustness of the proposed synchronization technique, several simulations were conducted with different grid conditions. The simulations showed that the synchronization process requires a longer time interval for larger frequency deviations, larger power production from the converter-unit, larger system inertia and variations in load during the synchronization process. Some grid conditions challenged the proposed control strategy and resulted in more oscillations in the phase angle and frequency deviation response. However, the simulations show that the proposed synchronization strategy effectively eliminate deviations in phase angle and frequency between the two microgrids.