Flow batteries and energy storage— a new market for ceramics

By Richard Clark, Global Lead of Energy Storage

Originally published in ACerS Bulletin, January/February 2022, Vol. 101, Iss. 1. https://ceramics.org. Displayed here with permission.

Demand for energy storage technologies is driving dramatic growth in the redox flow battery market, and with it opportunities for the ceramics community.

Redox flow batteries belong to a large and growing group of devices designed for energy storage applications.

Although their origins trace back to a zinc/chlorine flow battery used to power an airship in 1884 (1), rapid, systematic Growth in the redox flow battery market has only been apparent within the past decade.

To understand the reasons for this growth and the concurrent opportunities for the ceramics community, it is important to understand several related areas: the energy storage market and its segmentation; the technology behind flow batteries and how it compares to alternatives; and the challenges that need to be addressed to make flow batteries cost-competitive in the market segments where they have an advantage over alternatives.

The energy storage market and the opportunity for redox flow batteries
Widely misunderstood, the energy storage market is highly segmented, with the characteristics required for a given application (primarily discharge time and system power required) having a huge impact on the relative importance of cost. Complicating this market further is the fact that the classification and naming of the segments is not uniform, making comparison of information from multiple sources challenging.

At the top level, the primary categories for energy storage are transportation (mainly electric vehicles) and stationary. While conceptually redox flow batteries can be used for both, practically the only application of note is stationary storage (unless airships make an unexpected comeback) because of the low volumetric energy density of redox flow batteries compared to other technologies, especially the ubiquitous lithium-ion batteries. Subdivisions for stationary energy storage are generally described by the services they provide as an alternative to conventional grid supply (2). These subdivisions are (on-grid) regulation, arbitrage, back-up and reserves, black start, investment deferral, and (off-grid) independent power supply. These subdivisions are variously related to generation, transmission, and distribution as well as end-user markets (residential, commercial, and industrial).

Several competing technologies exist for each of the stationary storage market segments, which are categorized as electro-chemical (batteries and supercapacitors), thermal (molten salt), and mechanical (flywheels, compressed air, and pumped storage hydropower), although the specific requirements of each segment preclude all technologies from competing for the entire market. Pumped storage hydropower (PSH) is dominant in the overall space: as of mid-2020, 164 GW of PSH is operational, equivalent to an estimated energy in excess of 17 TWh, with an additional 124 GW in various stages of realization (3). The remaining competition in terms of major market scale is increasingly moving toward lithium-ion and redox flow batteries.

For redox flow batteries, 2021 global sales are projected to reach 350 MWh, with revenue at US$270 million. However, BloombergNEF predicts that redox flow batteries could compete with lithium-ion batteries for up to 46% (69 GWh) of the total capacity (150 GWh) required for grid-related stationary energy storage in 2030 (defined as being the segment which would be otherwise addressed by lithium-ion batteries) (3).

In this analysis, redox flow batteries were assumed to have an average system duration of four hours, which makes them particularly favorable for arbitrage (storing energy when the price of electricity is low and releasing it on the grid when high) and peaking capacity (provision of supply to meet the maximum demand of the system, including back-up and reserves and investment deferral applications). If the system duration can be modified significantly from this value, with appropriate capital and operational economics, then the size of the potential market for redox flow batteries will accordingly increase.

Particularly relevant to redox flow batteries, one of the main drivers behind the growth of the stationary energy storage market is the increasing replacement of fossil fuels with renewable energy sources (most notably wind and solar),which are intermittent and hence not directly coherent with demand. The economics of redox flow batteries are well suited to the relatively long durations required to address this imbalance. They also offer additional benefits such as a high level of safety, long maintenance-free lifecycles, and intrinsic modularity, in that power and energy are decoupled, facilitating customization. In many cases, sustainability issues are also addressed because the chemistries of choice are derived from widely available resources and materials can be readily reused.

A more detailed comparison of redox flow batteries with lithium-ion batteries is provided in Table 1 (4). By necessity, this table is generalized, and there are exceptions in some cases. Sustainability is not included as a category, mainly because the diversity of chemistries used in redox flow batteries preclude generalization.

However, there are advantages with many of the more popular types, and it is likely to be an increasingly important consideration in the future.

The technology behind redox flow batteries
A redox flow battery is an electrochemical device that uses the potential difference between a set of redox couples, typically solution-based, to transform electrical energy into stored chemical energy and vice versa.5 At the most basic level, there are two tanks containing electrolyte connected to a stack of cells where the redox reactions occur For a single-cell system, a cell would include two current collectors, two bipolar plates, two electrodes, and one membrane, although there are variations on this setup. Cells are usually stacked, so a pair of current collectors would be used for the full stack and bipolar plates would be the components between cells. A schematic of a single cell arrangement is shown in Figure 1.

There are a wide variety of systems encapsulated by the general definition, including inorganic aqueous, organic aqueous, and nonaqueous, and the definition is frequently broadened to include membrane-less, metal-air, semi-solid (slurry), and electroplated redox flow batteries. The latter systems are also termed hybrid redox flow batteries because the total energy storage capacity depends on both the stack size and the size of the electrolyte storage reservoirs. So, energy and power are not fully decoupled, in contrast to more “classical” types, where decoupling frequently is portrayed as a defining feature. Table 2 provides a summary and an example of each system.

Readers will appreciate that the technologies are in significantly different states of development, from those currently at laboratory scale, where there are fundamental challenges yet to be overcome, to those which are already launched commercially. Typically, there are multiple chemistries under investigation for each system, and for a given base chemistry, there are alternative systems and approaches. This flexibility offers great opportunities for innovation.

Table 3 provides a summary of the current state of market readiness for the main types of flow battery (7). With the exponential growth rate of the energy storage market, the market readiness of each battery is likely to change, and it is quite possible that a technology currently in its infancy could become the dominant force within several years.

Leading commercial redox flow battery systems
All-vanadium redox flow battery (VRFB)

The invention of the all-vanadium redox flow battery (VRFB) is credited to the work of Maria Skyllas-Kazacos and her research team at the University of New South Wales, Australia, in the 1980s. The VRFB electrolyte tanks contain vanadium at four different oxidation states. On the cathode side, vanadium is at +IV (tetravalent) and +V (pentavalent) states within VO2+ and VO2+, respectively. On the anode side, vanadium is at +II (bivalent) and +III (trivalent) states.

Figure 2 shows a general schematic. Typical materials of construction within the cell include graphite (bipolar plates), graphite felt (electrodes), and membranes based on chemically stabilized perfluorosulfonic acid/polytetrafluoroethylene copolymer acid (PFSA/PTFE), such as Chemour’s Nafion 212.

By using vanadium on both sides of the membrane, contamination issues with cross-over are effectively eliminated, although self-discharge would still occur should cross-over happen. Eliminating the possibility of cross-over contamination helps provide VRFBs with extended cycle life, typically 15,000 to 20,000 cycles (9),far above many other chemistries and battery types.

The early work determined suitable, highly reversible redox reactions for vanadium compounds within an aqueous electrolyte containing an appropriate level of sulfuric acid, providing ionic conductivity, and stabilizing the reactions. More recent work at Pacific Northwest National Laboratory determined that using a mixed acid (sulfuric and hydrochloric) within the electrolyte increases the solubility of the vanadium ions and improves the battery performance, increasing the specific energy and expanding the operating temperature range.

A VRFB offers some significant benefits in sustainability.10 First, it has full lifecycle CO2 emissions lower than other battery technologies—estimated at 27% to 37% lower than standard lithium-ion batteries. Secondly, vanadium is not consumed and does not degrade in a VRFB—so at end of life for one battery, it can be reused in another battery or used for another application such as in steel alloys. Additionally, although China is the leading country for vanadium mining, vanadium is mined in many other countries as well (most notably Russia, South Africa, and Brazil), unlike with lithium-ion batteries, where the raw material supply is dominated by China.

With the experience gained over the past three decades and the other benefits noted, it is not surprising that VRFBs are leading the way for flow batteries commercially. Bushveld Energy reports that there are 25 or more VRFB companies globally and that in China alone they are tracking about 2 GWh of VRFBs under construction.10 The largest facility is 200 MW/800 MWh in Dalian, China, installed by Rongke Power (11). China is expected to install between 30 GWh and 60 GWh of new energy storage capacity by 2030, and a considerable portion of this storage capacity is likely to be VRFB, providing the economics are suitable.

Vanadium is estimated to be 48% of the cost of the manufacturing cost of a VRFB, and spot pricing of the relevant chemicals in 2021 is still about 50% higher than needed for VRFB to have an economic advantage over lithium-ion batteries, although the other benefits such as increased safety may be sufficient.

Zinc-bromine flow battery (ZBFB or ZBB)

Development of the zinc-bromine flow battery (ZBFB) in its current form is attributed to work at Exxon in the early 1970s. Unlike most other redox flow battery types, the ZBFB is a hybrid system, in that the total energy stored is not purely proportional to the volume of electrolyte but instead depends on both the volume of electrolyte and the electrode area because zinc metal is plated onto the anode during charge. Concurrently, bromide ions are oxidized to bromine on the other side of the membrane. During discharge, the zinc metal oxidizes to Zr 2+ and dissolves into the electrolyte, and the bromine is reduced to bromide ions.

Figure 3 shows a general schematic of ZBFBs. The figure does not depict some additional requirements related to bromine. One, bromine has limited solubility in water, so a complexing agent is needed on the cathode side to prevent its release. With bromine’s high toxicity, it is essential to maintain certain conditions such as moderate system temperature to maintain the stability of the complex. Two, with bromine being highly oxidative, system components need to be specially selected, which can add to cost. There are also challenges on the anode side. Repeated plating of the zinc can cause uneven deposition and eventually dendrites of zinc can puncture the membrane. Pulsed discharge during charge may be required.

Despite these challenges, two significant benefits of a ZBFB are the high voltage and high energy density for flow batteries, associated with the two electrons per atom of zinc that are engaged in the charge-discharge process. Typical materials of construction within the cell include carbon-filled plastic (bipolar plates), carbon felt (electrodes), and membranes based on a PFSA/PTFE copolymer/polymer composite.

Conceptually, ZBFBs could be very low cost because of the raw materials, but dealing with the issues noted above frequently offsets this benefit. The potential cost reduction continues todrive work in this space (i.e., to solve or mitigate the issues and realize the benefits of the lower cost raw materials), and there are already several companies selling ZBFBs to end users. One recent development, arising from work at the University of Sydney, Australia, is the use of gel electrolytes replacing the liquid electrolytes.

All-iron flow battery (Fe-RFB)
The earliest work on the all-iron flow battery (Fe-RFB) is attributed to L. W. Hruska and R. F. Savinell in Ohio, with their first major publication on the subject in January 1981 (12). The Fe-RFB follows similar principles to those of the VRFB in that by using the same element with multiple valence states on either side of the membrane, losses associated with cross-over contamination can be eliminated. During charging, Fe2+ oxidizes to Fe3+ on the cathode side and Fe2+ reduces to iron metal on the anode side. During discharging, the reverse is true. Additionally, iron is abundant, inexpensive, and nontoxic.

While these characteristics would appear to make Fe-RFBs the ideal flow batteries, there is a major challenge caused by hydrogen generation, and frequent electrolyte rebalancing is required. Recent work at Case Western Reserve University in Cleveland, Ohio, demonstrated that a sealed Fe-RFB is possible with internal rebalancing, hence facilitating desired maintenance-free operation (13). As well as hydrogen evolution, other challenges include dendrite formation during iron plating and a relatively low cell voltage. Typical materials of construction within the cell include graphite (bipolar plates), carbon or graphite felt (electrodes), and microporous polyethylene (membrane).

In addition to the hybrid version of the Fe-RFB (depicted in Figure 4a), an alternative system uses a slurry at the anode side, in which the iron metal deposits on carbon particles (depicted in Figure 4b). As well as the obvious benefit of fully decoupling energy and power, this system also extends the capacity of the cell because the surface area of 3D carbon particles can be made much larger than the area of the negative electrode. This system has its own challenges, however, in that the slurry rheology must be maintained for extended periods and at different states of charge, and iron must deposit uniformly onto the particles.

There are at least three companies currently engaged in commercializing the Fe-RFB technology.

Opportunities for the ceramics community and conclusion
This period is one of rapid growth for energy storage, and redox flow batteries are likely to play an increasingly significant role. Large energy storage installations are very expensive, and already companies have experienced significant losses caused by lithium-ion battery-related fires in energy storage system (ESS) facilities in Australia, Belgium, China, England, South Korea, and the United States. The fires are difficult and dangerous to extinguish, and in at least one case resulted in loss of life (14). While safety standards for electric vehicles are designed to allow sufficient time for occupants to leave vehicles after initiation of battery thermal events, these standards are not as relevant for ESS installations, where prevention of propagation of thermal runaway is the only acceptable solution. Lithium-ion batteries containing solid-state electrolytes will provide a solution in the future, but there are still technical and commercial hurdles to overcome before these solutions will be widely used in cost-sensitive markets.

With ESS applications, redox flow batteries offer an immediate alternative to lithium-ion batteries, and the benefit of safety far outweighs the downside of the low energy density. Solutions already exist to the main technical challenges, and if it is possible to reduce cost as well (in US$/kWh), redox flow batteries can take the leading position in this market.

Already, according to detailed estimates from the U.S. Department of Energy,15 costs are relatively similar for longer duration systems when comparing higher energy lithium-ion battery installations versus VRFB. For example, for a 10 MW installation and 10-hour duration, total installed cost for an NMC-cathode lithium-ion battery system would be US$387/kWh and for a VRFB it would be US$426/kWh. However, the ongoing reduction of lithium-ion battery costs also needs to be considered, and it may be more difficult to reduce costs of redox flow batteries once a large scale (GWh) is reached because of the increasing relative dominance of the chemicals.

For a given redox flow battery chemistry, there are several ways in which cost can be addressed. The first and most obvious is replacement of current materials of battery construction with lower cost substitutes (lower US$, same kWh). A second way is to replace these materials with alternatives that provide improved system characteristics (same US$, higher kWh). For example, increasing the energy density of the stack will mean the total cost of the other system components will be lower on a system energy basis. A third way is to use materials that allow lower cost system design, for example, by combining components.

Bipolar plates and electrodes are usually made from carbon or graphite in a solid or flexible form. Bipolar plates by necessity must be impermeable. Electrodes can be made of flexible or rigid (porous) material, with some benefits offered by each, most notably cost for flexible materials and more uniform flow and extended life for solid ones. The membrane is highly engineered and can dominate stack component costs, particularly if it is based on a PFSA/PTFE copolymer or an equivalent polymer. Finding an alternative high-performance and lower-cost ion exchange membrane may prove to be the key to ongoing widescale commercial success for redox flow batteries, and hence presents an excellent opportunity for ceramics

The membrane in a redox flow battery has some characteristics not significant for other battery types: it must deal with ion cross-over, i.e., selectively allowing counter-ions to pass through, but not allowing active species the same passage; it must limit water transport; it must have a low areal resistance; it must resist fouling; and it must be stable in whichever chemistry is present and at whatever pH is used to stabilize the system.Despite their attractive properties (high proton conductivity and chemical stability), even PFSA/PTFE copolymer-based membranes are not ideal—aside from the cost, they are also selectively permeable to water and allow some cross-over of active species. One example of a ceramic which may prove suitable as a membrane in flow batteries is NaSICON, as demonstrated in recent work at Sandia National Laboratories by Eric Allcorn and others.16 Other work identified sol-gel as a possible process route for fabrication of low-cost silica based membranes (17).

As we progress in this dynamic growth period for energy storage, many options are conceptually possible. However, commercialization of any new technology and scaling it to GWh level takes many years, typically decades. The two most likely systems to capture the new space are lithium-ion batteries and redox flow batteries, which have each matured sufficiently to be scalable to the required level. Although there are advan tages and disadvantages of each, there is no dispute that redox flow batteries are safer from the perspective of minimizing battery-related fires. The three chemistries of redox flow batteries which are furthest advanced are all-vanadium, zinc-bromine, and all-iron. There is yet no clear winner among these, but in each case, innovation will help to realize the potential. If a suitable membrane can be part of the fulfillment,redox flow batteries will be a major new market for the ceramics industry.

References
1. D. Sigler, “Flow batteries at Stanford and in Lichtenstein,” March 8, 2014. Accessed Nov. 29, 2021. http://sustainableskies.org/flow-batteries-stanford-lichtenstein
2. M. Sanders, “The rechargeable battery market and main trends 2019–2030,” avicenne ENERGY, July 30, 2020, presented at the 37th International Battery Seminar & Exhibit.
3. “Energy storage grand challenge: Energy storage market report,” U.S. Department of Energy Technical Report NREL/TP-5400-78461, DOE/GO-102020-5497, December2020.
4. J. Daggett, “Can flow batteries compete with Li-ion?” Accessed Nov. 29, 2021. https://www.dnv.com/article/can-flow-batteries-compete-with-li-ion--179748
5. M.J. Watt-Smith, R.G.A. Wills and F.C.Walsh, “Secondary batteries – Flow systems, Overview,” Encyclopedia of Electrochemical Power Sources, 2009.
6. I. Iwakiri et al., “Redox flow batteries: Materials, design and prospects,” Energies 2021, 14(18):5643.
7.“Redox flow batteries: Technologies and market overview,” IDTechEx, June 2, 2020.
8. A. Clemente and R. Costa-Castelló, “Redox flow batteries: A literature review oriented to automatic control,” Energies 2020, 13:4514.
9. E. Sánchez-Díez, E. Ventosa, M. Guarnieri,et al., “Redox flow batteries: Status and perspective towards sustainable stationary energy storage,” J. Power Sources 2021, 481:228804.
10. “VRFB technology attributes and applicability to developing countries,” Bushveld Energy presentation, November 18, 2020.
11.“VRFB applications to boost China’s V demand: Correction,” Argus Media, March 22, 2021.
12. L.W. Hruska and R.F. Savinell, “Investigation of factors affecting performance of the iron-redox battery” J.Electrochem. Soc. 1981, 128(1):18–25.
13. S. Selverston et al., “All-iron hybrid flow batteries with in-tank rebalancing,” J.Electrochem. Soc. 2019, 166:A1725.
14. “2 firefighters die in battling Beijing’s power station fire,” China Daily, April 17, 2021.
15. “2020 grid energy storage technology cost and performance assessment,” U.S. Department of Energy Technical Report
Publication No. DOE/PA-0204, December 2020.
16. E. Allcorn et al., “Elimination of active species crossover in a room temperature, neutral pH, aqueous flow battery using a ceramic NaSICON membrane,” J. Power Sources 2018, 378:353–361.
17. G. M. Newbloom, “Flexible ceramic membranes for redox flow batteries,” 2018 ECS Meeting Abstracts MA2018-01 2373.