Vanadium Pentoxide Market Report 2026: Supply, Demand, and the Balance That Does Not Add Up

Updated 7 September 2026 · Built from USGS MCS 2026, The Metalnomist June 2026 consumption data, Vanitec, FerroAlloyNet and worldsteel · 15 questions answered at the foot of the page.

The demand mix turned over in four years

Four years ago vanadium was a steel additive with a small chemical business attached. It is now a steel additive with a genuinely large and fast-growing storage business attached, and that changes who sets the marginal price.

Sector volumes for steel, chemicals and storage are as reported by The Metalnomist in June 2026; the residual is our arithmetic, not a published figure, and should be read as such.

What matters about that table is not the percentages but the behaviour of the buyers behind them. A steel buyer purchases weekly, in tender, against a rebar or alloy steel order book, and will not pay above what the steel price supports. A storage buyer purchases in lumps, ahead of a project, and is far less price-sensitive per unit because the vanadium is a small part of a large capital cost and the electrolyte is recoverable at end of life. When the marginal tonne moves from the first buyer to the second, the price stops tracking steel margins. In 2026 that transition is underway but not finished.

Supply: 277,600 tonnes of capacity, 163,900 tonnes of output

A 59% utilisation rate in any industry invites the conclusion that higher prices will bring the rest online quickly. In vanadium they will not, and the reason is worth understanding properly because it drives every supply forecast.

The slag route is an ore problem

Vanadium is not mined for vanadium. In China the dominant chain runs vanadium titanomagnetite → blast furnace → vanadium-bearing hot metal → converter slag → roasting and leaching → V₂O₅. The vanadium is a passenger in a steel operation, which means its volume is governed by ironmaking decisions made for entirely different reasons. When the Panzhihua-area magnetite supply tightened, the reported consequence was a cut of 4,500–5,000 tonnes of V₂O₅ equivalent in 2026 - not because anyone chose to make less vanadium, but because there was less ore to blow.

Against that, Qinhuangdao Baigong completed a 10,000 t/yr V₂O₅ line in early 2026 and is guiding to about 5,000 t for the year. Add one new line, subtract one ore shortfall, and the net is close to zero. That is the character of this supply side: incremental changes that cancel.

Stone coal is the swing route that stopped swinging

Stone coal - a low-grade vanadium-bearing black shale - used to be the price-responsive part of Chinese supply. It is not any more. Output fell to 7,600 t in 2025, and by 2026 a single large producer was still running, at 100–120 tonnes a month of ammonium metavanadate on a V₂O₅-equivalent basis. A Shaanxi operation with 300–350 t/month of capacity has been suspended since early 2026 over safety issues.

The number that should end the debate: when flake prices reached about RMB 110,000/t in 2023, stone coal output reached roughly 11,000 t. That was the supply response to a good price. It is less than 7% of Chinese output. Anyone modelling a robust supply curve for vanadium is modelling something that has not existed for years.

Secondary supply runs on molybdenum

Recovery from spent catalysts, refinery residues and alumina by-product rose to 15,100 t in 2025, up 1,900 t, of which about 6,700 t came from alumina by-product recovery. The increase happened despite weak vanadium prices because the same feeds carry molybdenum and tungsten, and those were strong enough to keep the circuits running. Vanadium came out whether or not vanadium wanted it.

This is the recurring theme and it is worth stating once, plainly: almost nobody in this industry produces vanadium because the vanadium price went up. They produce steel, or alumina, or they recover molybdenum, and vanadium arrives as a consequence. That is why supply is inelastic in both directions.

A cross-check worth doing

Two measurement systems, built by different organisations for different purposes, can be made to talk to each other. It is a useful sanity check on whether the Chinese numbers are broadly credible.

System one. The USGS puts 2025 Chinese mine production at 82,000 t of contained vanadium.

System two. Chinese vanadium output of 163,900 t in 2025, of which 15,100 t was secondary. Primary output is therefore 148,800 t on a V₂O₅-equivalent basis. Converting at 56.01% vanadium by mass: 148,800 × 0.5601 = 83,300 t of contained vanadium.

The two figures differ by 1.6%. One is a geological survey's estimate of mine production; the other is an industry count of oxide produced, minus recycled feed. They were never designed to reconcile, and they land within two percent of each other. When that happens, the underlying data is probably sound, and the disagreements between analysts are about interpretation rather than facts.

A working 2026 balance

We are going to show our arithmetic and then explain why we do not fully believe it. That seems more useful than a confident single number.

The lines labelled "estimate" are ours, assembled from published sector forecasts. The 2025 "actual" column is reported data except where marked with a tilde.

Three caveats, and they matter more than the totals.

First, the two columns may not be in the same units. "Output" in Chinese industry reporting counts vanadium products; "consumption" counts V₂O₅ equivalent. If the output figure includes material that is later converted and counted again, the comparison is distorted. We suspect it is not, but we have not been able to prove it, and we would rather say so.

Second, China exports vanadium, in oxide, FeV, VN and chemical forms. A domestic surplus is not a global surplus. The 38,000 t gap in 2025 is what left the country or went into inventory, and neither number is published at a frequency we would rely on. Verify current export volumes and any licensing requirements against customs data and a broker - this is not an area where last year's figure can be carried forward.

Third, electrolyte inventory is invisible in this table. Chinese electrolyte project capacity was reported above 4.5 million cubic metres a year as of May 2026. Plants at that scale hold working stock that never appears in a consumption statistic until a project is commissioned. If a meaningful part of the 2026 storage number is tank-filling rather than installed capacity, then demand is being pulled forward and the balance is tighter than the table suggests in 2026 and looser in 2027.

The concentration problem, stated without drama

China held 68.8% of global vanadium capacity and 72.4% of global production in 2025. The USGS puts 2025 world mine production at 110,000 t of contained vanadium, down from 118,000 t in 2024, with the decline attributed largely to environmental regulations, energy controls and long project lead times - concentrated in Sichuan's Panzhihua-Xichang belt, which is dominated by a handful of integrated producers.

Within China, the picture narrows again: the slag route runs through a small number of plants, and one of them not producing is a market event. In 2025, Xinjiang Da'an and Yunnan Yukun did not produce at all, removing about 8,000 t combined; Chengsteel, Desheng and Dagang raised output by roughly 15% and offset most of it.

For a buyer outside China this is the whole risk picture. Not geology - the USGS reports world reserves in the tens of millions of tonnes against annual output around 110,000 t, so there is no scarcity of rock. The risk is that a small number of plants, subject to one country's environmental and energy policy, set the price for everyone.

Policy signals that actually moved demand

Two 2026 measures are worth knowing, because they are the kind of thing that shifts tonnage more than price commentary does.

• Generation-side capacity pricing, 30 January 2026. China's National Development and Reform Commission and National Energy Administration issued a notice on improving the generation-side capacity price mechanism. It rewards duration, which is precisely the attribute flow batteries have and lithium does not. This is the single most important policy reason the storage number is where it is.
• Rebar production licence rules, 1 April 2026. Tighter licensing and expanded quality traceability for construction rebar should raise the share of vanadium-nitrogen micro-alloyed hot-rolled product. Combined with the stricter rebar standards introduced in 2025, this is why Chinese vanadium intensity per tonne of crude steel rose to 51 g of vanadium metal equivalent in 2025 from 48 g in 2024 even as rebar output fell.

One counterweight: Chinese rebar output fell 4.5% to 186.3 Mt in 2025, and VN consumption fell with it, down 3.8% to 36,690 t. FeV50-equivalent consumption went the other way, up 10.4% to about 39,985 t, carried by machinery, energy, shipbuilding, automotive and rail - automotive production reached 34.778m units in 2025, up 9.8%, civil shipbuilding 52.295m deadweight tonnes, up 18%, and excavator output 379,643 units, up 17%. The composition of steel demand is changing, not just its volume.

Where we disagree with the consensus

The popular framing is that vanadium is the next lithium. We do not think the tonnage supports it, and we would rather say so than repeat it. Global vanadium mine output is around 110,000 t of contained metal a year; lithium carbonate equivalent production is measured in millions of tonnes. A metal whose entire annual output would fit in a few dozen bulk carriers does not get a lithium-style demand curve. What it can get is a genuinely tight, structurally different market in which a 20% demand segment grows at 30% a year and moves the price a lot. Those are different things, and conflating them is how people end up holding inventory at the wrong moment.

We would also give weight to the sceptical view from outside China. Amy Bennett of Fastmarkets, quoted by Investing News Network in August 2026, said vanadium is "starting to lose out in this conversation" in Western storage markets as LFP chemistry takes the sector, and that flow batteries are no longer discussed as much as they were. If the storage story stays predominantly Chinese, its ability to lift a globally traded oxide price is more limited than the headline growth rates suggest.