Stage Two closed at four o'clock

Stage Two of the Expression of Interest closed at 4:00pm AWST on 20 July 2026. The Department of Energy and Economic Diversification is now evaluating proposals and expects to name a preferred proponent later in 2026. The State's commitment is up to A$150 million in grant funding, not a project capital cost: total capex is proponent-provided and has not been published, and bidders were required to show their proposals do not depend on additional State support.

The specification is 50 megawatts and 500 megawatt-hours, ten hours of discharge, sited in the Goldfields on the South West Interconnected System, with a target commercial operation date of end-2029. The department's preferred connection point is the West Kalgoorlie Terminal. Two clauses in the tender do most of the work: the battery must use vanadium flow technology and be delivered in the Kalgoorlie region, and the State has halved the royalty on the metal from 5 to 2.5 per cent. One publicly declared consortium pairs Australian Vanadium's VSUN Energy subsidiary with Sumitomo Electric as flow-battery technology provider; another comprises AVESS, Atlantic Vanadium and Lycopodium.

At announcement in November 2025 the project was reported as the largest planned vanadium redox flow battery in Australia. That framing no longer holds. On 14 July 2026 Idemitsu Australia unveiled a proposed 108-megawatt, 864-megawatt-hour, eight-hour flow battery at the former Muswellbrook coal mine in the Hunter Valley. Both remain unbuilt, and both are early-stage development at the date of writing.

The colour of charge

A vanadium flow battery stores its energy in two tanks of liquid electrolyte outside the cells, so power and energy are sized separately: the stack sets how fast, the tanks how long. That decoupling is what lets the chemistry stretch to ten hours without oversizing power. The electrolyte is a single element on both sides, held in four oxidation states, and it shifts through violet, green, blue and yellow as it moves between them, which makes state of charge visible in the liquid itself.

The cost picture, as far as the public record goes

The direct eight-to-twelve-hour levelised-cost comparison this subject invites cannot be sourced from the two analyses the sector normally reaches for. CSIRO's GenCost 2025-26 states plainly that it provides projections for batteries and pumped hydro only, and batteries there means lithium-ion. Flow batteries and iron-air are not modelled, and it does not publish levelised cost of storage at all. Lazard's published storage work restricts its analysis to one-, two- and four-hour durations, so it carries no ten-hour vanadium benchmark either. Any vanadium-versus-lithium cost claim attributed to those two reports is unsupported by them.

The comparable neutral references at ten-hour duration are the Pacific Northwest National Laboratory's Energy Storage Grand Challenge cost and performance database, updated in 2024, which publishes installed cost and levelised cost by technology across power, energy and duration bands including ten-hour vanadium redox flow, and the National Renewable Energy Laboratory's 2025 utility-scale battery cost projections on the lithium-ion side.

For the Australian basis, GenCost puts a current four-hour lithium-ion system at A$385 per kilowatt-hour and a 24-hour system at A$266, with the lithium-ion cost curve down 11 to 16 per cent in a single year. The Australian vanadium reference points that exist are proponent-authored: VSUN Energy's Project Lumina, at scoping-study accuracy of plus or minus 30 per cent, calculated a levelised cost of A$251 per megawatt-hour for an eight-hour 100-megawatt system in November 2024, revised to A$214 in May 2025. Those are the bidder's own numbers at the bidder's own stated accuracy. They are not an independent benchmark.

Pumped hydro, air and iron

Pumped hydro carries the lowest long-duration cost in the GenCost reference set, A$237 per kilowatt-hour over 24 hours, but it needs elevation and water, and the Goldfields is flat and dry. Compressed-air storage needs the right geology and years of permitting. Neither was available at this site.

Iron-air is a different duration category from the duty specified. Form Energy's target of under US$20 per kilowatt-hour at scale is for a 100-hour system aimed at multi-day and weekly cycling, not daily ten-hour firming, and its round-trip efficiency sits materially below both lithium and vanadium: Form Energy publicly targets 50 to 60 per cent. Its first utility-scale commercial system, 1.5 megawatts and 150 megawatt-hours with Great River Energy in Minnesota, has only recently reached deployment. Comparing a 100-hour cost target with a ten-hour asset tells a reader little that is useful, and a flagship grid-security asset is not the natural place to run a first-of-kind experiment at Australian scale.

Lithium-ion is the incumbent and the obvious default until the duty cycle is read closely. Grid lithium-ion in the CSIRO reference costing is priced on a 20-year battery warranty, and per-kilowatt-hour costs rise as duration extends toward ten hours, which is the duration specified here for an asset intended to cycle daily for decades. (The residential side of the lithium story is a different picture, covered in the lithium-ion decade.)

The second decade

Vanadium's case is not the sticker price; it is the second decade. Because the electrolyte is a single element held in four oxidation states, it does not degrade the way a solid electrode does. Invinity, whose VS3 stack is running Western Australia's own Kununurra pilot, publishes a rated cycle life beyond 20,000 cycles at 100 per cent depth of discharge, annual capacity degradation under 0.5 per cent and annual round-trip-efficiency degradation under 0.1 per cent. Sumitomo Electric, the flow-battery technology provider on the reported VSUN Energy bid, publishes a 30-year service life independent of operating pattern. Utility-scale lithium-ion, for comparison, is warranted at 20 years in the CSIRO reference costing.

Vanadium levelised cost from Australian Vanadium's Project Lumina release and its May 2025 revision, stated at scoping-study accuracy and proponent-authored rather than independent. Cycle and service life from the Invinity VS3-022 datasheet and the Sumitomo Electric redox flow battery catalogue. Lithium-ion and pumped hydro capital costs and the lithium warranty basis from CSIRO GenCost. Iron-air cost target and efficiency from Form Energy. Blank cells are figures the referenced source does not publish.

The vanadium underneath

Western Australia holds major vanadium deposits, the State has halved the metal's royalty from 5 to 2.5 per cent, and the tender requires that the battery use vanadium flow technology and be delivered in the Kalgoorlie region, with the State's own framing at announcement calling for local manufacture using WA vanadium. Of the candidate chemistries, vanadium is the one whose value chain, mine to electrolyte, the State can hold locally. On the published record the technology choice and the industrial-policy choice were made together: the tender mandates the chemistry and Kalgoorlie delivery, and the State halved the royalty alongside.

The proof-of-concept is already in the ground in the State's far north. Horizon Power's 78-kilowatt, 220-kilowatt-hour Invinity VS3 flow battery at Kununurra, filled with Australian Vanadium electrolyte and procured through VSUN Energy, was formally launched on 15 November 2024 and is running a twelve-month pilot to answer the question Kalgoorlie asks: whether the chemistry holds in extreme heat and humidity. Public results have not yet been published. The dates deserve a moment's arithmetic. A twelve-month pilot launched on 15 November 2024 reached the end of its window around November 2025, and Stage Two closed on 20 July 2026 with nothing published in between. The tender wrote the chemistry in before the State's only in-climate trial of it had reported. The Kununurra stack vendor also differs from the reported VSUN-Sumitomo bid for Kalgoorlie, though the vanadium and the electrolyte manufacturer are the same.

The context around West Kalgoorlie

It is tempting to read this asset as the direct replacement for a retiring gas-fired station, and the public record does not support that. Synergy's current listing describes the Kalgoorlie Gas Turbine Station as an operational 57-megawatt open-cycle backup with no published retirement date. The 2018 retirement direction that would have closed West Kalgoorlie units 2 and 3 was reversed under a Ministerial Direction so the plant remains available to Western Power as a network control service. After the January 2024 Goldfields blackout the Premier committed to replacing the ageing Kalgoorlie back-up generators, additional backup was contracted from TransAlta, and Western Power began procuring 150 megawatts of reliability services and 1,500 MVA of system-strength services in the Eastern Goldfields. AEMO's 2025 WEM Electricity Statement of Opportunities identifies Collie, Muja D, Pinjar and Bluewaters as the ageing coal and gas retirements in the 2027-2032 window; West Kalgoorlie is not among them. The battery sits inside a wider Eastern Goldfields reliability program rather than substituting for a single named plant on a fixed date.

The pattern is not local

Beyond the local politics, the same pattern is emerging elsewhere. As grids fill with solar and wind, firm capacity becomes the scarce product. Lithium keeps the short-duration duty it already owns. The long-duration end is still open, which is where vanadium flow and iron-air are being tested, and where Kalgoorlie becomes one of the first public trials of a chemistry at Australian scale with an industrial-policy layer, locally made electrolyte from locally mined vanadium, that the tender is explicitly written around. (How batteries earn against their duty, once built, is measured in the TBx capture index.)

The preferred proponent is expected in the second half of 2026. When the winning bid lands, the three things I will look for are the total capital cost, the stack vendor and the route to first electrolyte. None of the three is on the public record today.