The order
Stage Two of the Expression of Interest closed on 20 July 2026; Energy Policy WA is now evaluating proposals and expects to name a preferred proponent in the second half of the year. The asset is a 50-megawatt, 500-megawatt-hour system, about ten hours of storage, sited in the Goldfields well away from the main South West Interconnected System, with A$150 million of State funding behind it. Two clauses in the tender do most of the work: the battery must be manufactured in Western Australia from locally sourced and processed vanadium, and the State has halved the royalty on the metal, from 5 per cent to 2.5. The reported bid pairs Australian Vanadium and its VSUN Energy arm with Sumitomo Electric as the flow-battery provider.
The colour of charge
A vanadium flow battery keeps its energy in two tanks of liquid electrolyte rather than inside its cells. Power and energy are therefore sized separately, the stack setting how fast and the tanks how long, which is why the chemistry stretches comfortably to ten hours where lithium strains past four. The electrolyte is the same element on both sides, held in four oxidation states, and it shifts through violet, green, blue and yellow as it moves between them: the battery shows its state of charge as colour. It is, almost literally, an instrument that can be read by eye, which is a fitting thing for a State to buy to steady a grid.
The cost that should have killed it
CSIRO's GenCost 2025-26 puts Australian lithium-ion at A$385 per kilowatt-hour for four hours of storage, a figure that fell 11 to 16 per cent in a single year. Iron-air, the newcomer, targets under US$20 per kilowatt-hour. Lazard's 2025 levelised-cost work finds lithium's costs have fallen faster than flow's, and that lithium remains the most cost-competitive option overall. By the one number most procurements lead with, dollars per kilowatt-hour today, vanadium loses. That it won regardless is the point of the exercise.
Why the cheap seats were empty
Take the alternatives in turn. Pumped hydro is the cheapest long-duration storage Australia builds, at A$237 per kilowatt-hour over 24 hours on the same CSIRO basis, but it needs height and water, and the Goldfields has neither to spare. Compressed-air storage needs the right geology and years of permitting. Both were ruled out by the map before anyone reached for a spreadsheet.
Iron-air is genuinely cheaper and genuinely unproven: its first utility-scale system, 15 megawatts with Georgia Power in the United States, was only reaching the grid around 2026, and it returns under half the energy it stores. For every ten megawatt-hours in, four to five come back out. A flagship grid-security asset is a poor place to run a first-of-kind experiment.
Lithium-ion is the incumbent and the obvious default, until the duty cycle is read closely. Grid lithium is rated for a few thousand full cycles and must nurse its state of charge to reach ten years of service. A battery cycled hard every day for decades is the job it does least well. The residential half of the same lithium wave is read in the lithium-ion decade.
The number that actually matters
Vanadium's case is not the sticker price; it is the second decade. A flow battery's electrolyte does not degrade the way a solid electrode does. Invinity, which built the flow battery now running in Western Australia's own Kununurra pilot, rates its vanadium systems beyond 20,000 deep cycles with no capacity fade and no ceiling on throughput. Sumitomo Electric, the flow-battery provider on the reported Kalgoorlie bid, has run its cell stacks past 50,000 start-stop cycles without measurable loss. Grid lithium is rated for a few thousand full cycles by comparison. For an asset meant to firm a grid daily into the 2050s, that gap is the argument.
| Technology | Upfront cost | Round-trip | Cycle life | Fit for this duty |
|---|---|---|---|---|
| Vanadium flow | Above lithium per kWh | 65-80% | 20,000+ cycles, no fade | Matches: long, daily, hot |
| Lithium-ion | A$385 per kWh (4 h) | 85-95% | 3,000-7,000 cycles | Built for under four hours |
| Iron-air | Under US$20 per kWh (target) | 40-50% | Unproven at scale | Cheap, but first-of-kind |
| Pumped hydro | A$237 per kWh (24 h) | 70-80% | 40+ years | No height, no water |
| Compressed air | Site-dependent | 70-75% | 40+ years | Wrong geology |
Sources by column: CSIRO GenCost 2025-26 for the lithium-ion and pumped hydro capital costs; Lazard's 2025 levelised cost of storage for the cost direction and lithium cycle life; Form Energy, reported by Energy-Storage.News, for iron-air cost and round-trip; Invinity and Sumitomo Electric datasheets for vanadium cycle life. Cells carry ranges or targets wherever the underlying figure does.
The vanadium underneath
Then the tiebreaker, which was never really about electrochemistry. Western Australia holds some of the world's largest vanadium deposits, it has halved the metal's royalty, and it has written into the tender that the battery be built in the State from locally sourced and processed vanadium. None of lithium, iron-air, hydro or air hands the State a resource it already owns and an industry it can raise around it. The technology decision and the industrial-policy decision were, in all likelihood, one decision.
The proof-of-concept is already in the ground, a couple of thousand kilometres north. Horizon Power's 78-kilowatt, 220-kilowatt-hour flow battery at Kununurra, an Invinity stack filled with Australian Vanadium electrolyte and procured through VSUN Energy, has run since October 2024 to answer precisely the question Kalgoorlie asks: whether the chemistry holds in extreme heat. Its results are not yet public. The stack vendor differs from the Kalgoorlie bid; the vanadium does not.
What it signals
Strip away the local politics and a pattern remains. As grids fill with solar and wind, the scarce product stops being energy and becomes firmness: hours of it, on demand, for decades. Lithium will keep the short shifts it already owns. The long, patient duty is where flow chemistries, iron-air and the rest will be judged, and Kalgoorlie is among the first large, public, side-by-side tests of which is ready. How batteries actually earn against that duty, once built, is measured in the TBx capture index. The preferred proponent is due in the second half of 2026, and it will be worth reading not for who wins, but for what the winning bid reveals about the price of ten firm hours in a hot, dry place.