Redox Flow Batteries for Long Duration Energy Storage

A group of scientists in the materials science department of MIT has published an interesting paper on the possibility of using redox flow batteries for long duration energy storage. In true redox flow batteries (as distinguished from hybrid flow batteries) all of the chemical reaction occur in pair liquid electrolytes in the two half cells of the batteries. The electrolytes are stored in tanks and pumped past the electrodes and past an ion exchange membrane which allows ions of a desired type to pass between the two half cells to complete desired chemical reaction. The electrodes are either neutral sources and sinks of electrons or are coated with catalyst that promotes the desired chemical reaction in the liquid electrolytes. However, no solid state reaction such as intercalation or plating occurs on the electrodes. The power rating of the battery is determined by the electrodes, the ion exchange membrane and the electrolyte pumping rate. The energy storage capacity is determine by the size of the electrolyte tanks. This separation of energy capacity from power rating could potentially allow flow batteries to be used for very long duration energy storage is the cost of the electrolytes.

This situation if very different from other battery types where power rating and storage capacity are not independent. For instance the MIT group estimates that the chemical cost of energy storage for the high temperature sodium sulfur batteries manufactured by NGK Insulators is US $1.60/kWh. At this price 32 cycles would be sufficient to lower the chemical cost to US $0.05/kWh. However, their basic battery module delivers 33kW of power for 7.2 hours (i.e. 216kWh of total energy). The basic physics/chemistry of the battery design does not allow any substantial increase of the ratio of energy storage to power. Therefore delivering 720 hours (thirty days) of energy requires 100 batteries to discharge in succession. Since the manufacturing costs are dominated by other factors than the chemical cost of the electrode materials these batteries are not practical for long term energy storage. For example if the total cost of manufacturing the battery exclusive of chemical costs is US $360/kWh then 7200 cycles are required to lower this part of the cost to $0.05/kWh. However, if power and storage time could be decoupled then low cost could be driven by long discharge times rather than by a large number of cycles. 720 hours of discharge delivered ten times would accomplish the same thing as 7200 charge/discharge cycles. Of course 720 hours of discharge (as opposed to 7.2hours) would increase the chemical cost by a factor of 100 so that low chemical cost is the key to making long storage times economical for redox flow batteries.

The MIT group has proposed a low chemical cost flow battery based on sodium and sulfur but with very different reaction than occur in NGK Insulators batteries. They have built a prototype which demonstrates the concept, but which does not have adequate performance for practical storage application. The cycle life is short and the round trip efficiency is low.

Jan 7, 2018

rogerkb at energystoragenews dot com