Abstract
The Split-pi converter can suitably interface an energy storage system (ESS) with a DC microgrid when galvanic isolation is not needed. Usually, the ESS voltage is lower than the grid-side voltage. However, limitations in terms of the ESS current make the use of a high-voltage ESS unavoidable when high power levels are required. In such cases, the ESS voltage can be higher than the microgrid voltage, especially with low microgrid voltages such as 48 V. Despite its bidirectionality and symmetry, the Split-pi exhibits a completely different dynamic behavior if its input and output ports are exchanged. Thus, the present work aims to model the Split-pi converter operating with an ESS voltage higher than the grid-side voltage in three typical microgrid scenarios where the controlled variable is the converter’s output voltage. The devised state-space model considers the parasitic elements and the correct load model for each scenario. Furthermore, it is shown that the presence of the input LC filter can make the design of the loop controllers more complicated than in the case of a lower ESS voltage than the grid-side voltage. Finally, the study is validated through simulations and experimental tests on a lab prototype, and a robustness analysis is performed.