For the past ten years, Anthony Price has organised the International Flow Battery Forum, an annual conference to promote interest in, and encourage the development and use of, flow batteries. BEST magazine invited Price to discuss flow battery technology and its role in energy storage.
Energy storage can be classified in many ways; for example, we can classify a storage type by the technology, by size (power), by energy content (energy) or by lifetime or suitability for particular applications. Batteries can also be of the primary, (that is single-use) or secondary (rechargeable) type.
If we consider rechargeable batteries, then parameters such as calendar lifetime and cycle lifetime also become important and these parameters can also be used as methods of classification.
Flow batteries are a class of rechargeable batteries with many desirable characteristics. Although invented more than a century ago, it has taken a long time for the benefits of flow batteries to be appreciated and for users of energy storage to consider them as a viable option. But lack of public awareness does not mean that flow batteries should be dismissed. For, encouragingly, there are more new flow battery manufacturers moving into the market, and there is increased optimism that the technology may have a key part to play in the future energy storage world.
Investment in the technology is high, from basic research, materials suppliers, manufacturers, project developers and investors. Many power developers, with interests in both conventional and renewable generation, are taking an interest in flow batteries, and they are sizing up the product offerings on the market. They expect to pounce as soon as the costs fall enough to provide a positive internal rate of return (IRR).
We have seen similar moves in other parts of the power industry, first in wind, and more recently in solar and then lithium-ion batteries. Investment in solar was only for the true optimists until the costs of PV panels fell, and together with guaranteed incomes provided a solid and positive IRR. Lithium-ion batteries entered the market, not for bulk time-shifting of energy, but because selling ancillary services to the network operators was a viable project proposition when the cost per kW became relatively low.
Now power project developers are watching the costs of flow batteries, waiting to see when the IRR changes from negative to neutral or even positive— because the unique features of a flow battery give it an advantage over other battery types. An example of a flow cell is shown in Fig 1.
Fig 1: A single flow cell. Source: IFBF
What is a flow battery?
A flow battery is a rechargeable battery in which electrolyte flows through one or more electrochemical cells from one or more tanks. An individual flow cell contains many similarities in concept to a fuel cell, two electrolytes engage in an electrochemical reaction between two electrodes and the cell can keep supplying power for as long as fuel and the oxidising agent is available. In a simple flow battery, it is straightforward to increase the energy storage capacity by increasing the quantity of electrolytes stored in the tanks. The electrochemical cells can be electrically connected in series or parallel to determine the power of the flow battery system. This decoupling of energy rating and power rating is an important feature of flow battery systems.
The interconversion of energy between electrical and stored chemical energy takes place in the electrochemical cell. This consists of two half cells separated by a porous membrane or by an ion-exchange membrane. As well as permitting ionic conduction, the separator minimises the loss of the generated electroactive species in the half cells and so maintains high coulombic efficiency. The redox reactions during charge and discharge take place at the electrodes of the half cells. In its simplest form, the electrodes themselves, usually carbon felt, are not altered by these electrochemical reactions.
The cell voltage is the difference between the negative electrode reaction and that at the positive electrode. During charging, electrons released at the positive electrode through oxidation of the electroactive species in that half-cell are pushed around the circuit to the negative electrode where reduction of electroactive species in that half-cell takes place. The processes are reversed on discharge. The electroactive materials are redox pairs, that is, chemical compounds that can reversibly undergo reduction and oxidation. Because of the redox couples used as electroactive species in each half-cell, a flow battery is sometimes known as a redox battery or a redox flow battery.
The choice of redox pairs is often used as a description of the type of flow battery. Some well-known redox pairs are:
- Vanadium/vanadium (which uses the four different valency states of vanadium)
Usually, both the electroactive species in the redox pairs are soluble in aqueous acid or alkali solutions. However, in some flow batteries, such as zinc-bromine, one active species (in this case, zinc metal) is deposited on the electrode. These types of batteries are sometimes known as hybrid flow batteries. Other flow battery systems use aqueous solutions of organic redox pairs, such as quinones and TEMPO, instead of metal-based redox couples, and other types operate in totally non-aqueous environments, employing organic and organometallic redox couples. These different chemistries have different characteristics, and this means that we should take care not to consider all varieties of flow batteries as having the same performance, applications, form factor, costs or any other parameter.
The energy storage medium is often known as an electrolyte, but some systems, for example, the hydrogen-bromine flow battery, may use a gas (hydrogen) as the storage medium.
Most flow batteries use bipolar electrodes so that cells can form a cell stack or module. Modules can be connected together in series and parallel, both electrically and for distribution of the electrolytes. The actual configuration will depend on detailed engineering, which is resolved by each individual manufacturer to achieve an optimised solution. A schematic of a flow battery energy storage system is shown in Fig 2.
Because electrochemical cells share a common electrolyte, each cell can be at the same state-of-charge (SoC), simplifying cell balancing and battery operation. The SoC of the whole system can be measured at a single point (or several measurement points can be used to check the correct functioning of the battery system).
A comparison between PbA, Li-ion and flow batteries
The use of lead-acid batteries in many power systems (mostly direct current) from the 1880s until the 1920s and 1930s became obsolete due to the high cost, and relatively low efficiency of converting direct current to alternating current, and because primary generation had become so low-cost that using batteries for night-time load or peak shaving was no longer financially necessary. The introduction of pumped hydroelectricity storage provided a solution— offering low-cost bulk storage, albeit with some severe limitations on location.
After the Second World War, the drive was on to develop improved battery systems. Pumped-hydro storage was popular, and many pumped-hydro stations were constructed as a low-cost solution to the provision of peak-power and frequency-response services. The advent of nuclear power, and the promise of off-peak electricity that would be ‘too cheap to meter’, gave a further incentive to construction of pumped-hydro in North America, Europe and many parts of Asia, as pumped-hydro stations could provide black-start services to conventional power stations.
Of course, the world does not sit still, and many lead-acid battery manufacturers responded with a range of improvements, giving enhanced lifetimes, greater depth-of-discharge (DoD) and low costs. Sealed lead-acid systems resulted in a lowering of operations and maintenance costs, but still left some areas where flow batteries could claim the higher ground.
Several research groups and manufacturers began exploring different types of flow batteries, looking for a low-cost battery system for peak management and frequency-response services. The ambition of flow battery developers was to produce a battery that had a low cost of manufacture, low operational and maintenance cost (for example to avoid the need for topping up cells with distilled water or acid), a long and predictable lifetime and was tolerant of a wide range of ambient conditions, that provided a suitable replacement for the existing lead-acid batteries. Other advantages such as their capability of a near 100% DoD would be a bonus.
In the 1980s and early 1990s zinc-based systems, such as zinc-bromine were being developed and commercialised by groups in North America, Austria and Australia. Research into the possibility of using vanadium electrolytes began in the late 1980s by professor Maria Skyllas-Kazakos of the University of New South Wales. Professor Skyllas-Kazakos and her colleagues have built up considerable expertise in vanadium batteries and their group is still at the forefront of research into vanadium-based flow batteries.
In its simplest form, the all-vanadium flow battery has two electrolytes: on one side the electrolyte is oxidised and reduced between vanadium (II) and vanadium (III) and on the other side of the membrane, the reactions are between vanadium (IV) and vanadium (V). The corresponding cell voltage is relatively low, just over one volt. However, the elegance of this system is that because vanadium is on both sides of the cell (although in different valence states) any crossover between the two electrolytes is benign and can be easily corrected by using a simple electrolyte balancing process.
Other attractive features of the vanadium system are the recyclability of the vanadium electrolyte when the battery is eventually retired and the use of low-cost polymers in the cell stack that gives a low overall system cost. The vanadium salts can be at high molarity in the electrolytes, and this leads to improved energy density. Early vanadium systems were operating at energy densities of around 25Wh/litre, but energy densities of 38Wh/l are now achievable.
Most vanadium systems use sulphuric acid in the electrolytes, but it is also possible to use a mixed electrolyte of sulphuric and hydrochloric acid, which leads to higher energy densities of up to 50Wh/l. The mixed electrolyte is also stable over a wide temperature range (from +40°C to -10°C). Heat management in a flow battery is relatively simple in comparison to some lithium-ion installations, which have a smaller operating temperature range.
A complete flow battery system will need to include electrochemical cell stacks, pumps, cell balancing and battery management systems as well as controls for heating and ventilation, and of course, electrolyte tanks. The tanks can be configured in many ways and it is expected that some very large plants will be hydraulically connected in series and parallel, and the electrolytes pumped in and out of two very large tanks servicing the whole system. Other manufacturers prefer to keep the tanks small and associate a pair of tanks with just one or two cell stacks.
In any configuration, there is a requirement for sufficient bunding around the tanks and cell stacks to contain all the electrolyte in the unlikely event of electrolyte leakage. In general terms, flow batteries offer a safe and benign solution to battery energy storage. There are no flammable materials, the electrolytes are non-flammable, and although they may be acid or alkali, containment in the event of an incident can be managed. The risk of explosion through the release of hydrogen, due to overcharging, can be reduced through SoC management.
Apart from the electrolytes, flow batteries do not usually contain any toxic or restricted materials and are not reliant on minerals from sources with poor human rights records. Manufacturing costs, when production is at scale, are low— materials can be injection moulded and machine-assembled into stacks. Batteries can be shipped dry, or filled with distilled water, and the electrolytes dispatched separately.
Flow batteries have a very wide operating range, often cycling is possible between close to 0% to nearly 100% SoC. Frequent cycling is possible— although the term ‘cycling’ is somewhat meaningless when a full charge can be doubled, simply by doubling the size of the tanks. Many flow batteries have been tested in applications to provide not only energy management services but also frequency response services, (with rapid oscillations between charge and discharge) with no detrimental effects on the battery system. Changes in the duty cycle can be easily managed— the large quantity of electrolyte provides a near-perfect medium for removal of heat from the cell stacks, simplifying thermal management.
The basic design concept of a flow battery means that it has a low, or negligible, self-discharge. If not used for an extended period, the pumps and valves can be closed, isolating the cell stacks from the bulk electrolytes. Any self-discharge is limited merely to the electrolyte contained in the cell stack. This can be rapidly replaced, as soon as the pumps are started and electricity is circulated. Such isolation can be used for specific configurations, such as the provision of black-start services (that is providing power when the local network has failed).
In a recent paper, Simon Davidson Kurland of Chalmers University estimated that 55-65kWh of energy was required to manufacture a kWh of a lithium-ion battery cell. Because of the configuration of a flow battery, with the electrolytes stored in tanks outside the cells, the energy cost for production of the battery per kWh of storage is much lower and would depend significantly on the choice of electrolyte. A calculation (by Jiyun Heo of the Korean Institute of Science and Technology) describes the energy cost to produce 1kWh of vanadium electrolyte to be approximately 1kWh. The energy cost for production of the stacks is low, involving relatively simple manufacturing processes such as injection moulding for plastic components and casting and machine of steel endplates. Although the batteries will have membranes, which may require energy-intensive actions, the stack cost is related solely to the power of the system and not the energy content.
The flow battery industry today
There are about twenty OEM flow battery manufacturers working with vanadium-based systems, and probably a similar number working on other chemistries. Some companies have their origins in the days of the early work at the University of New South Wales including Sumitomo. Other companies’ pedigrees can also be traced back to this time through acquisition mergers and staff transfers, including Avalon, and VRB.
Commercial alliances between companies, including vanadium suppliers such as Bushveld with Avalon and Enerox, UET with Rongke Power and Dalian Bolong New Materials, Schmid Energy Systems and SABIC (Saudi Basic Industries Corporation) show confidence in vanadium batteries. There are other OEMs in the sector, some are relatively small companies, including Volterion and Voltstorage, as well as larger companies who are also investing in flow battery technology, such as Lockheed Martin.
Flow batteries can be configured as large multi-MW installations, with massive tanks, and associated pipework, bearing more similarity to a chemical works than a battery, or they can be packaged in ISO shipping containers or even smaller enclosures, even suitable for small commercial or even domestic use. The zinc-based systems are often packaged in smaller enclosures, Redflow, based in Australia, produces systems for domestic use as well as packaged units for co-locating with telecom towers.
Other manufacturers have produced flow batteries based on totally different chemistries. At last year’s IFBF, Kemiwatt displayed their organic flow battery system, and they are joined by several other companies who are also working on aqueous and non-aqueous organic systems.
The case for integrating large-scale renewables has been well made in many different forums and it is a logical extension that to timeshift renewable energy, to when it can be best used, requires considerably more long-duration storage than we currently have on our power systems. This will only happen with investment, and the investment case relies on; low initial capital cost, low cost of ownership and operation, and low environmental footprint.
There could be some advantages in linking batteries of different types together. For example, operating a flow battery alongside a lithium-ion battery system. RedT (who recently merged with Avalon) supplied a hybrid flow and lithium battery on a microgrid at Monash University in Melbourne, Australia, and it has been announced that they are to supply a 2MW vanadium flow battery to Pivot Power, for deployment on the Energy Superhub, in Oxford. The vanadium battery will operate alongside a 100MW lithium battery so that both batteries provide complementary services.
Flow batteries are an obvious choice for large-scale, high-performance energy storage in modern power systems. Expect to see more of them in the next few years.