A computational hybrid energy storage system (HESS) model was developed in Python, comprised of a vanadium redox flow battery (VRFB) and a super capacitor (SC). Different sized HESS were applied to the use-case of a 50-vehicle public EV charging depot, to take advantage of dynamic electricity prices during the day and to reduce the cost of electricity incurred by the EV charging service provider (EVCSP). Fifteen sizes of HESS were evaluated for this application, comprised of 10 – 50 kWh VRFBs and 5-15 kW SCs.
First, the VRFB was assessed using a multi-period optimization genetic algorithm (GA) to find an optimum charge and discharge schedule over a given monthly EV charging load profile. The simulations showed that incorporating a VRFB to an EV charging depot resulted in cost savings when compared to a depot with no storage. The cost savings amounted to $1.13 to $6.19 per day, corresponding to $380.72 to $2,068.04 of costs saved annually. A greater cost reduction was observed for a greater VRFB size.
The results of the GA showed signs of optimization, where the residual load was greater compared to the EV charging demand during times of low-cost grid electricity, due to the VRFB charging to take advantage of the lower energy costs. During times of high-cost electricity, the residual load was lower compared to the EV charging demand as the VRFB would discharge to supplement the EV charging demand.
A net present value (NPV) calculation was performed to consider the discounted cashflows and VRFB capital cost over a 20-year period for each VRFB size. This yielded strong positive values ranging between $436,600 - $448,485, and the NPV decreased for increasing VRFB sizes due to their high capital costs. The best NPV was $451,800 for a charging depot with no storage. The high capital costs for the VRFB outweighed financial benefits from energy cost savings, and to obtain a NPV equivalent to the ‘no storage’ option, the VRFB capital cost would have to decrease from $650 /kWh to $360 /kWh.Following the VRFB study, SCs of varying sizes were implemented to understand whether further energy cost savings could be achieved.
The SCs were implemented using a rule-based strategy, where they would charge/discharge depending on the charging schedule of the VRFB, to provide additional arbitrage and to reduce the remaining EV charging load that was not covered by the VRFB. The SCs carried out 2-3 discharge cycles per day, and due to their low energy density, they could not carry out sufficient energy arbitrage to make any meaningful cost reduction. A final NPV calculation was carried out for each VRFB+SC combination, which resulted in lower NPV values than those determined for the VRFB only. This was expected due to the SC’s high capital costs and minimal energy cost reduction.