An Optimal Power-Splitting Strategy for Hybrid Storage Systems

The wide deployment of inverter-based resources is changing the grid characteristics and demanding rapid changes in the control and operation of power grids. Storage is becoming a key component of this transition, as it can provide a wide palette of services, from fast frequency response to congestion management. This implies a broad range of requirements for the storage units, in terms of energy, power, response times, etc. Hybrid storage systems, combining units with different properties, such as supercapacitors, high-energy or high-power batteries, represent a natural choice that can cover many grid services. These services do not only require active power but also reactive power, and ideally they will be provided by the storage system while optimizing its overall efficiency. In this work we formally derive active and reactive power splitting strategies that minimize the operation losses among all inverters of the hybrid storage system. These strategies are computationally light and do not rely on complex optimizers, thereby facilitating their implementation in embedded systems. A realistic example depicts the clear improvement of this approach against simple power splitting strategies. While our focus lies on hybrid storage systems, the formulation can be applied as well to hybrid power plants, since their power losses can be characterized by similar mathematical functions.