Optimal Sizing and Strategy of a Hybrid Energy Storage System for Smoothing Renewable Power Fluctuations

Abstract

Developing renewable energy sources entails new challenges which have not been faced previously by the traditional grid. One of the issues is the variability of renewable resources due to their characteristics of weather. The intermittent nature produces uncertainty in generation output, especially in solar and wind power generation. As penetration of both sources increase, variability will be more difficult to handle. And even as this issue is being resolved, another one, that of the application of an energy storage system has arisen. The energy storage is the current and typical means of smoothing wind- or solar-power generation fluctuations. Conducted research mostly developed an energy storage for only one technology, however, different technologies can be combined and complement each other which significantly improve the storage system’s performance. In this research, a hybrid energy storage (HES) is proposed to mitigate power fluctuations of renewable plant output power.

This thesis aims to give understanding of the effective strategy of a hybrid energy storage for PV power output and wind power output fluctuation suppression. The strategy will stand as the foundation for a broader framework which aims to optimize the size of hybrid energy capacity to enable such application in uncertain condition. Wavelet power sharing method is proposed as the strategy to operate the hybrid energy storage using frequency-based method. The strategy can decompose the frequency component of renewable power quickly and allocate the power to the respective device. In this paper, battery and supercapacitor is adopted to meet the electric grid technical requirements for smoothing renewable power. The results show that the supercapacitor peak fluctuations of battery power are moderated by adding the supercapacitor, providing lower peaks and slower derivative of power fed to/drawn from the battery. The result from the power sharing is suitable in terms of improving the battery lifetime. Battery lasts for 1.3 years in a conventional energy storage system, while it can last up to 6 years in the hybrid energy storage.

After getting the optimal power allocation between storage devices, sizing the capacity of hybrid energy storage also considers uncertainty in power output. This research uses investment storage cost, penalty and life cycle cost as the objective function, while hybrid energy storage’s state of charge and the power fluctuation performance as the constraint. The chance constrained programming is used to make the fluctuation of output power under a certain confidence level, which also meet the electrical quality and economy. The optimal solution of hybrid energy storage capacity is carried out by using genetic algorithm based on stochastic simulation. The results show that, with the confidence level of smoothing requirement up to 90%, system contains two energy storage has a lower total cost than a single-sourced energy storage.