Originally developed in the 1960’s by NASA and already in the 1980’s proven to last for more than 10,000 charge/discharge cycles, flow batteries gain popularity fast. Along with the continuous growing installed base of renewable energy systems, the need to store (very) large quantities of electrical energy and for longer periods of time, has become inevitable to facilitate the energy transition.
The term ‘flow battery’ actually covers a family of storage systems, applying the same fundamental working principle, while using different combinations of active materials. The heart of a flow battery is a so-called electrochemical cell, which is a multi-layer assembly of an ion-selective membrane, catalyst layers and electrodes.
A complete flow battery system, also referred to as a redox flow battery or RFB, is constructed around such electrochemical cells, where chemical energy is provided by the chemical reaction of two active materials. The active materials are contained within the system, separated by the membrane, and circulate in a closed loop in their own respective space.
When an electrical power source is connected (i.e. charging), a chemical redox reaction starts. Ion exchange occurs through the membrane, resulting in an electric current. During discharge, when applying an electrical load, the reverse chemical reaction takes place.
The voltage of the electrochemical cell is determined by the Nernst equation and ranges in practical applications from 1.0 to 2.2 V, depending on the selected active materials.
In order to increase the total electrical power, individual electrochemical cells are stacked, whereby the cells are electrically interconnected in series. To design systems with (very) large power levels, multiple stack assemblies can be interconnected.
Animation courtesy of Pacific Northwest National Laboratory (PNNL) S&T
The power [kW] of a flow battery system, as depicted above, is determined by the surface area of the ion-selective membrane, while the capacity [kWh] of the system is determined by the volume of the catholyte and anolyte reservoirs.
The fact that the membrane surface area and the reservoir volumes can be dimensioned individually highlights one of the most distinguishing properties of flow batteries, as opposed to traditional electricity storage systems where power [kW] and capacity [kWh] scale simultaneously
Power (kW) and Capacity (kWh) are NOT coupled
The flow battery family
Purely from a scientific perspective, flow batteries can be built around a large variety of different chemistries, by using different active materials (‘redox couples’). Whether a particular chemistry eventually also reaches commercialization phase, depends on several factors, of which the most important ones are:
1) Power density [kW/m2]
The power density determines the power the flow battery can deliver per m2 of membrane surface area. A low power density translates into the need for many m2 of cell materials, more in particular, of membranes. In such case this becomes a significant cost driver for the total flow battery system.
2) Energy density [kWh/m3]
The energy density determines the energy the flow battery can store per m3 of active materials. A low energy density translates into the need for large quantities of active materials, potentially adding up to high costs. Furthermore, a low energy density generally results in systems with large physical dimensions.
3) Cost of active materials [€/kWh]
When designing flow batteries with a large storage capacity [kWh], the costs for the active materials become dominant in the total system costs.
From a cost point of view, the ideal flow battery chemistry shows a high power density, a high energy density and is designed around low cost active materials. Therefore, within the flow battery family, the combination of these 3 factors is to a large extent decisive for the storage cost per kWh (or ‘Levelised Cost of Storage’, see under ‘Cost of Storage’) of a particular chemistry.