Brazil Vanadium Pentoxide (V2O5) Market Report, Size & Forecast 2026–2033: Supply, Exports and US Import Dependence

The four numbers that describe the same market

Brazil supplies 47% of the vanadium pentoxide the United States imports. That is the sentence this report was commissioned to test, and on the import-source series the United States Geological Survey carries for vanadium, it holds. It is also one of four statements about the same material that disagree so sharply that a reader who took any one of them alone would get the market wrong.

• 47% — Brazil's share of US vanadium pentoxide imports, on the USGS import-source series.
• 6.4% — Brazil's share of world mine production: 6,400 tonnes of vanadium content out of roughly 100,000 tonnes in 2023, behind China at 68%, Russia at 20% and South Africa at 9.1%.
• 80.1% — the share of Brazil's own vanadium supply that comes from a single company's single mine.
• 31 years — the remaining mine life disclosed for that asset, on a 48,954-hectare property in Bahia State.

Set them in a row and the character of the market changes with each. A country that is 6.4% of world supply is nonetheless the largest single source for the world's largest economy's imports, because the two facts are measured at different ends of the same chain. And a country that appears in trade statistics as a strategic supplier is, physically, one open pit and one concentrator with a 494-person workforce.

That is the finding this report is built around. Brazil's vanadium position is not a national industry. It is a single-asset position that happens to be registered as a country, and everything that follows — the forecast, the risk, the contracting advice, the technology chapter — is a consequence of that asymmetry.

What the model says

Pheonix models the Brazilian vanadium pentoxide pool in V2O5-equivalent tonnes, from a 2026 base year to 2033. The traded unit matters: V2O5 is 56% vanadium by mass, so 6,400 tonnes of contained vanadium is 11,429 tonnes of V2O5 equivalent, and a report that switches between the two without saying so will mis-state the market by a factor of 1.79.

Value compounds 2.36 times faster than volume, and unlike most commodity forecasts the gap is not a mix effect and not a quality upgrade. It is a price recovery that is doing 56% of the work in log terms, against 44% from tonnage. The Brazilian pool grows by 2,900 tonnes and by USD 89.99 million between the two dates: 25.4% more metal, 68.9% more money.

Price: what a recovery is being measured against

Vanadium pentoxide has spent five years falling. On the Chinese benchmark that the USGS carries in its mineral commodity tables, V2O5 averaged USD 12.17 a pound in 2019 and USD 9.25 a pound in both 2022 and 2023. By September 2025 the same benchmark was quoted at USD 4.86 a pound — 60.1% below its 2019 level and 47.5% below 2022.

The base year of this forecast assumes USD 5.20 a pound, a 7.0% lift off that September 2025 trough, and the terminal year assumes USD 7.00 a pound. Both are assumptions and the report treats them as such. What makes them worth arguing about is not their level but what they sit relative to, because Largo Inc. — the company producing 80.1% of Brazil's vanadium — publishes a 2025 cash operating cost guidance range of USD 4.50 to 5.50 a pound, which is USD 9,920.79 to USD 12,125.41 a tonne of V2O5.

Read the two numbers together and the trough stops being an abstraction. At USD 4.86 a pound, the benchmark was below the top of the producer's own cash cost range. The forecast's opening price of USD 5.20 a pound sits 5.45% under that ceiling. Even the terminal price of USD 7.00 a pound is 27.3% above the top of the cash cost range but 7.98% below the all-in unit cost the same company's 2025 results imply: a net loss before tax of USD 51.7 million on 8,686 tonnes sold works out at roughly USD 5,949 of pre-tax loss per tonne sold against revenue of about USD 10,714 a tonne, which puts the FULL unit cost at about USD 16,663 a tonne — USD 7.56 a pound.

That is the report's most uncomfortable number. At the 2033 price, the 14,300-tonne pool is worth USD 220.68 million. At the all-in cost of the asset that produces most of it, those same tonnes are worth USD 238.29 million. The forecast never reaches breakeven for the producer it depends on; the gap is USD 17.61 million, or 7.98% of the modelled market. A market can clear at that price for a long time, because co-product supply does not need the price and the primary producer cannot switch off. It cannot, however, finance a new mine at that price, which is why the volume line is capped.

How the pool is built, bottom up

The 2026 base of 11,400 tonnes is assembled from two components rather than extrapolated from a national total. The first is Largo's Maracás Menchen mine, which produced 9,150 tonnes of V2O5 equivalent in 2025 — 3.68% below the floor of its own 9,500-to-11,500-tonne guidance range — and sold 8,686 tonnes at a global recovery rate of 80.1%. At 56% vanadium content, that output is 5,124 tonnes of contained vanadium, or 80.06% of Brazil's 6,400-tonne national figure. The second component is the co-product and secondary stream: 2,269 tonnes drawn from vanadium-bearing slag and recycled material, sold largely as metallurgical-grade material into domestic steel and chemical customers.

Which sets the arithmetic of growth. Adding 2,900 tonnes by 2033 on one orebody is not a demand problem, it is a recovery-and-throughput problem. A move from an 80.1% recovery rate to 90% would add 12.4% to output from the same ore — 1,132 tonnes, 39% of the required growth. The remaining 1,768 tonnes has to come from throughput and grade, which means capital, which means a price that finances it. That is the same conclusion as the previous section arrived at from the other direction, which is why Pheonix is willing to publish a number this modest.

Two evidence notes belong here rather than in a footnote. The Brazilian national figure is a mineral-content estimate and the V2O5 figure is a conversion at 56%, so the base year carries a rounding band of a few hundred tonnes either way. And the split between primary and co-product is derived — Largo's disclosed V2O5 output and Pheonix's own vanadium coverage converted to contained metal against the USGS national figure — not read from a published Brazilian table, because Brazil does not publish one at this granularity.

Three cuts of the same 11,400 tonnes

Each cut below is a marginal distribution of the same pool and each sums to 11,400 tonnes and USD 130.69 million. They are not cross-tabulated against one another: a two-dimensional matrix would produce cells that no source publishes, and Pheonix would rather label a marginal than invent a cell. Every price-per-tonne in the tables is an implied realisation, computed from its own row.

By product form

Battery-grade material is 6.0% of the tonnage and 11.0% of the value, at 1.92 times the realisation of metallurgical flake. That spread is the entire commercial case for a purity upgrade, and it is small enough in tonnage that a producer can lose the argument by focusing on it.

By destination

The United States takes 47% of Brazil's exports, which is 41.65% of total Brazilian output and 44.14% of the value — a higher value share than tonnage share, because the highest-purity lines go there. Brazil ships USD 54.43 million of vanadium products to one customer. The destination set here is cross-checked against something concrete rather than asserted: Largo publishes safety data sheets for its VPURE+ pentoxide in exactly five jurisdictions — North America, Europe, China, Brazil and South Korea — which is the shortest available list of where the material is actually sold.

By end use

The two premium lines together — aerospace master alloys and flow-battery electrolyte — are 13.0% of the tonnage and 20.35% of the value. Everything else is a steel order or a chemical order, which is why the price of the benchmark and the price of steel output move the whole table together.

What the numbers argue

The 47% is a chokepoint measured at the wrong end of the chain. Brazil is 6.4% of world supply and 47% of one buyer's imports because the buyer has no primary production of its own: the United States produced no vanadium from primary ore and met 58% of apparent consumption from imports. Combine the two published ratios and Brazil covers roughly 27% of US vanadium consumption — 0.47 multiplied by 0.58. That product is a bridge, not a measurement, because the first ratio is struck on pentoxide imports and the second on total vanadium net import reliance. It is still the correct order of magnitude, and it explains why a 11,400-tonne national pool gets treated as strategic.

A single asset cannot be diversified, so the risk premium is in the price line, not the volume line. Largo ended 2025 with USD 9.7 million of cash against USD 107.1 million of debt, and in October 2025 secured a term sheet with five Brazilian lenders covering USD 84.2 million of that debt to defer principal to March 2026, with an automatic rollover to September conditional on raising at least USD 22 million. That is not a demand story. It is the arithmetic that forces a producer to keep selling into a price below its full cost, which is exactly what keeps the trough deep — and it is why Pheonix's volume line is deliberately modest rather than aspirational.

The growth is being spent on the other product. In the trough, Largo's capital programme went to ilmenite, not vanadium: flotation circuits installed from September 2025 lifted ilmenite concentrate capacity from 42,000 to 115,000 tonnes a year, a 2.74-fold expansion, against a 25.4% increase in vanadium tonnage over seven years. A second revenue line on fixed infrastructure is the only cost relief a primary vanadium producer has, and the capital allocation says so more clearly than any strategy statement.

Basis of estimate

This report is a documented estimate, not a measurement, and the following qualifications are load-bearing.

The 47% is stated, not re-read this cycle. It is carried as the import-source share on the USGS Mineral Commodity Summaries vanadium sheet, and on this cycle the sheet could not be re-parsed: the PDF returned as compressed binary rather than extractable text, and the USGS landing pages for the commodity returned network errors. The import-reliance and production figures quoted elsewhere in this report come from the same source and carry the same caveat. Anything going into a purchasing or sourcing decision should be confirmed against the current sheet directly.

Company figures are disclosed figures, not audited re-statements. Largo's production, sales, recovery, guidance and mine life are taken from the company's own Maracás Menchen mine disclosures and its 2025 results as recorded in Pheonix's vanadium coverage. The all-in unit cost of USD 7.56 a pound is derived by adding the disclosed pre-tax loss per tonne sold to the implied revenue per tonne at the September 2025 benchmark; it is a directional figure that includes impairment and financing effects and should not be read as a cash cost.

Prices are a single benchmark, and that is a known distortion. The series used is the Chinese V2O5 benchmark carried by the USGS. Metallurgical vanadium and battery-grade vanadium are different products with different specifications and different qualified suppliers, and a single benchmark prices neither correctly. The model applies form-level and end-use-level premiums to a benchmark-anchored average rather than publishing a second price series it cannot source.

Volume is denominated in V2O5 equivalent at 56% vanadium content. National figures published in contained metal are converted at that ratio throughout and never mixed with pentoxide tonnes. The primary/co-product split is derived rather than published. The Brazilian regulatory framework — titles administered by the Agência Nacional de Mineração, the CFEM financial compensation on mineral production, and the critical-and-strategic minerals programme — is stated as the operative framework and was not re-read instrument by instrument for this cycle.

No cross-tabulation is offered. The three cuts above are independent marginals. Multiplying a product share by a destination share would generate cells that neither USGS nor any producer publishes, and the report does not do it.

Related Pheonix coverage: the global manganese mining market for the co-product steel-alloy comparator, tool and speciality alloy steel for the demand end of the metallurgical chain, stationary energy storage for the flow-battery substitution question, and the lithium and nickel mining reports for the battery-metal comparison.

Table of Contents

• Market overview: the four numbers that describe the same market
• Competitive landscape: one country, one mine, three primary producers in the world
• Value chain: from a Bahia pit to a bale of flake and a tank of electrolyte
• Investment activity: why 2,900 tonnes is a recovery problem, not a demand problem
• Technology and innovation: the purity ladder is the real constraint
• Market risk: five risks ranked by how long they last
• Regulatory landscape: what Brazil licenses and what the United States imports

Competitive Landscape

Three producers in the world, and two of them are in this report

Vanadium supply is more concentrated than the trade statistics imply. Bushveld Minerals describes the primary universe as one of only three producers in the world, owning two of the planet’s four operating primary vanadium processing facilities — a claim made by a competitor, which is the most useful kind, because a company does not overstate its own scarcity. Largo Inc. describes itself as the world’s largest primary vanadium producer. Both statements can be true at once, and together they describe a market with three named primary participants and one enormous co-product mass behind them.

Concentration inside Brazil is starker still. With Largo’s 5,124 tonnes of contained vanadium against a national figure of 6,400 tonnes, one company holds 80.06% of national supply. Squaring that share gives a Herfindahl–Hirschman index of 6,410 — roughly two and a half times the threshold at which a market is conventionally called highly concentrated. Pheonix therefore records one tier-one player in this market, which is unusual for a country report and is the honest number: a market with a single double-digit participant has one, not zero.

The Brazilian position: one mine, described by its owner

The asset is the Maracás Menchen mine in Bahia State, 99.94% owned by Largo, listed on NASDAQ and the TSX. The disclosed operational facts are specific enough to model from: 48,954 hectares; roughly 250 km southwest of Salvador and 813 km northeast of Brasília; an open pit feeding a concentrator; 494 employees; a 31-year mine life under the October 2024 life-of-mine plan; 9,264 tonnes of V2O5 equivalent produced in 2024 and 9,600 tonnes sold; and 2025 guidance of 9,500 to 11,500 tonnes of production at a cash operating cost of USD 4.50 to 5.50 a pound.

Its product range is the strategy. VPURE and VPURE+ cover pentoxide flake, pentoxide powder and vanadium trioxide powder, with a guaranteed vanadium content of 99.0% and a typical content of 99.5% against an industry standard of 99.0% — and the powder grades are positioned explicitly at catalysts and at vanadium electrolyte for flow batteries. The company states that its high-purity flake is approved by all major master-alloy producers. It publishes safety data sheets in five jurisdictions: North America, Europe, China, Brazil and South Korea.

That last detail is the most commercially revealing thing on the page. A producer does not maintain regulatory documentation in a market it does not sell into. Read as a map of the customer base, it tells you where the 11,400 tonnes go — and it is the evidence behind this report’s destination segmentation rather than an assumption about trade flows.

The comparator: Bushveld Minerals

Bushveld Minerals is the other listed primary producer, operating the Vametco and Vanchem facilities in South Africa and producing nitro-vanadium, ferrovanadium, vanadium oxides and vanadium chemicals. Its downstream move is the one Largo is also making: the Bushveld Electrolyte Company plant at East London, which the company describes as the largest publicly announced electrolyte plant outside China.

The two companies are not really competing on ore. They are competing on the same narrow qualification ladder, in front of the same small set of master-alloy producers and electrolyte customers, while a third force — Chinese co-product output at 68% of world supply — sets the price both of them sell into. That is the structure worth carrying into a sourcing decision: two listed primary producers, one dominant co-product mass, and a specification market inside a commodity market.

Why 80% market share does not buy pricing power

Brazil’s concentration is real but it is concentration of a specific product, not of the metal. Chinese tonnage is overwhelmingly recovered from vanadium-bearing iron ore processed for steel, so it appears whether or not the vanadium price justifies it. A producer holding 80% of a national market that is 6.4% of world supply cannot set the benchmark; it can only choose which grade of a benchmark-priced product it sells. Every competitive strategy in this market therefore reduces to one question: how much of the output can be moved off the metallurgical specification and onto one where the seller, not the slag, sets the price.

Related Pheonix coverage: the global manganese mining market for the closest co-product steel-alloy analogue, and cobalt mining for a battery-metal market with a similar co-product supply problem and a very different demand curve.

Value Chain

The chain, in six steps

Vanadium moves from a Bahia open pit to a customer’s furnace or tank in six steps: ore and concentrate; roast and leach; pentoxide flake as the traded intermediate; conversion to powder, trioxide, ferrovanadium or nitro-vanadium for metallurgical customers; electrolyte manufacture for flow batteries; and qualification at the customer’s plant, which is the step that decides whether any of the previous five earn a premium.

The economics are decided at step three and step six, and they are not the same economics. Step three is a commodity conversion priced off a Chinese benchmark. Step six is a specification approval that can double the realisation.

The cost stack, from the operator’s own numbers

The middle two lines are the whole story of the trough. At the September 2025 benchmark of USD 4.86 a pound, a tonne of pentoxide sold for USD 10,714 against a cash cost range of USD 9,921 to USD 12,125. At the top of its own guidance range the producer was selling below cash cost; at the bottom it kept roughly USD 793 a tonne. There is no version of that spread that funds a mine.

Add depreciation, impairment and financing and the same tonne carried an implied USD 16,663 of cost against USD 10,714 of revenue. That is a USD 5,949 loss per tonne, which is the number the company reported on 8,686 tonnes sold.

Where the value is created, by product

Two observations follow, and they point in opposite directions. The first is that the metallurgical products — flake, ferrovanadium, nitride — are 80.9% of the value and all sit within a 1.19× band. Alloying is a conversion business with a thin, published premium. The second is that the one genuinely differentiated product, battery-grade material, carries a 1.92× multiple on 6.0% of the tonnage — too small to move the pool today, and the only line that can reprice it tomorrow.

The trap for a buyer is to negotiate on the average. A tonne of metallurgical flake at USD 10,989 and a tonne of electrolyte-grade pentoxide at USD 21,053 are both “a tonne of V2O5” in a contract schedule and are not the same purchase. The whole negotiation is which specification the price refers to.

The co-product on the same infrastructure

The value chain has a second exit. Largo’s ilmenite concentration plant, completed in 2023, takes feedstock from the vanadium operations and was guided to produce 25,000 to 35,000 tonnes and sell 20,000 to 30,000 tonnes in 2025. Flotation circuits installed from September 2025 were designed to lift concentrate capacity from 42,000 to 115,000 tonnes a year — a 2.74-fold increase — against a vanadium volume forecast that adds 25.4% over seven years.

Read the two numbers together and the capital allocation is unambiguous. In the worst price environment the metal has seen in a decade, the funded expansion went to the titanium co-product, on the same orebody, using the same concentrator and the same people. That is not a diversification strategy; it is a fixed-cost-absorption strategy, and it is the only one available to a single-asset primary producer in a trough.

Related Pheonix coverage: tool and speciality alloy steel for the demand end of the metallurgical chain, and stationary energy storage for the buyer at the other end of the electrolyte line.

Investment Activity

What 2,900 tonnes actually requires

The forecast adds 2,900 tonnes of V2O5 equivalent over seven years, and the useful thing about a single-asset market is that the growth can be decomposed rather than asserted. At a 2026 realisation of USD 11,464 a tonne, those tonnes are worth USD 33.25 million of new annual revenue. Where they come from matters more than the total.

The recovery lever is the cheap half. On Largo’s output, each percentage point of recovery is worth roughly 114 tonnes of product, so a ten-point improvement delivers 1,132 tonnes from ore that is already being mined, already permitted and already trucked. That is the only volume in this forecast that does not need a price above cost. Everything else does.

What the price will not finance

At USD 7.00 a pound the terminal price sits 7.98% below the implied all-in unit cost of the asset that produces 80.06% of the country’s supply. A greenfield vanadium project — or a material expansion of the existing one — requires a price above that cost plus a return on construction capital. There is no version of the 2033 price path in this report that clears that bar. Pheonix is therefore forecasting no new primary vanadium mine in Brazil through 2033, and says so rather than filling a growth line with a project that has no final investment decision.

The capital that was actually committed went elsewhere. Largo’s ilmenite flotation expansion, installed from September 2025, targets an increase in concentrate capacity from 42,000 to 115,000 tonnes a year. That is a 2.74-fold expansion of a co-product against 25.4% more vanadium over seven years, on the same orebody, using the same concentrator. It is the clearest statement in the market about where incremental capital earns a return.

The balance sheet that sets the timetable

Investment in this market is not gated by opportunity. It is gated by a covenant calendar. Largo ended the 2025 year with USD 9.7 million of cash against USD 107.1 million of debt, and in October of that year secured a term sheet with five Brazilian lenders covering USD 84.2 million of that debt to defer principal repayments to March 2026, with an automatic rollover to September 2026 conditional on raising at least USD 22 million.

Set that condition against the arithmetic of the trough. The difference between the market’s 2033 value and the value of the same tonnes at the producer’s all-in cost is USD 17.61 million — 80.0% of the USD 22 million the lenders require. It is not the same money and the comparison is illustrative rather than causal, but the magnitudes are the point: the annual gap between what the market pays and what the asset costs is the same order as the financing the asset needs to survive. That is the definition of a market that cannot fund its own replacement.

Pheonix records investment trend as rising and capital intensity as high, with no recent M&A activity in this market. The rising trend belongs to the titanium co-product and to electrolyte capacity downstream, not to vanadium mine supply. Anyone reading “rising investment” as a supply signal should read the footnote instead.

What a counterparty should conclude

Three structures follow directly from the numbers, and Pheonix’s own vanadium coverage sets them out: index with a floor, so a contract tracks a published benchmark but does not fall below the cost at which supply survives; tolling, where the buyer owns the metal and rents the conversion, removing the flat-price exposure from both sides; and electrolyte leasing, which converts a volatile metal purchase into a service charge over a twenty-year asset life. Each of them exists because a single spot benchmark cannot price a product whose cost of production is above its price for years at a time.

Related Pheonix coverage: global nickel mining for how a different battery-metal market handles single-asset concentration, and tool and speciality alloy steel for the alloy surcharge mechanics on the demand side.

Technology & Innovation

The constraint is impurities, not vanadium

Vanadium purity specifications are often written in a way that hides what is being bought. A “99.0% vanadium” guarantee sounds like a statement about the metal; in practice it is a statement about how little of everything else is present. Largo’s VPURE+ flake is guaranteed at 99.0% vanadium with a typical content of 99.5%, against an industry standard of 99.0%, and the commercial value of that half-point is that the residual 0.5% is what a master-alloy producer or an electrolyte plant will not tolerate.

That is why the technology ladder in this market runs flake, powder and trioxide rather than orebody to orebody. VPURE+ pentoxide powder and vanadium trioxide powder are positioned at catalysts and at vanadium electrolyte — the two applications where a parts-per-million impurity is a service-life problem rather than an assay line.

What a flow battery is actually buying

In a vanadium redox flow battery the electrolyte is vanadium. Energy capacity is set by the volume and concentration of vanadium in the tanks rather than by the stack, so the marginal cost of an extra hour of duration is a chemical purchase and the battery’s cost curve is a vanadium curve. That is the structural reason flow batteries compete for long-duration duty and lose on short-duration duty, and it is also the reason a flow-battery customer cares about provenance and contamination in a way a steel customer never will: the electrolyte circulates for the life of an asset measured in decades, and everything dissolved in it circulates too.

The demand shift is real but late. The USGS vanadium sheet cites third-party projections of vanadium redox flow batteries taking roughly 17% of vanadium consumption by 2033, against about 3% in 2021. Pheonix models a more conservative Brazilian exposure — battery-grade material at 6.0% of the national pool in 2026, 11.0% of its value — because the qualification path, not the demand projection, is the binding constraint.

The qualification file is the real technology

For a producer, the hard technical work between 2026 and 2033 is not mining and not chemistry. It is producing a documented, repeatable specification: at-line measurement of what is in the product rather than a certificate of what was intended, a qualification history with each master-alloy producer and electrolyte manufacturer, and a supply chain that can evidence provenance to a Western customer’s procurement function.

That is why Pheonix records low patent activity and moderate innovation intensity in a market with a mature technology stage. Vanadium metallurgy is not where the novelty is; the novelty is in the analytical and documentation layer that turns a commodity tonne into a qualified tonne. A producer who cannot evidence a specification competes on the benchmark price. A producer who can, competes on a multiple of it.

Where the innovation actually happened in this trough

Not in vanadium. Largo’s ilmenite concentration plant, completed in 2023, was followed from September 2025 by additional flotation circuits targeting an increase in concentrate capacity from 42,000 to 115,000 tonnes a year. Flotation optimisation on a vanadium-bearing orebody, applied to a titanium-iron oxide co-product, is unglamorous process metallurgy — and it is the only capital-backed technical programme the asset has disclosed in the period this report covers.

The same logic applies to recovery. At an 80.1% global recovery rate, each percentage point is worth roughly 114 tonnes of product on Largo’s output. Moving to 90% is a metallurgical programme worth 1,132 tonnes of annual production — 39.0% of the entire forecast growth — and it requires no new ore, no new permit and no price above cost. In a market where the price will not finance a mine, the recovery circuit is the mine.

Related Pheonix coverage: stationary energy storage for the long-duration market that sets the electrolyte premium, and lithium mining for the purity-grade industrial logic that governs the competing chemistry.

Market Risk

Five risks, ranked by how long they last

Commodity risk registers usually rank by probability. For a single-asset market the more useful ranking is by durability — which risks pass, and which are structural facts about how the metal is made.

Risk 1 is the one that cannot be managed by waiting. A price-inelastic supply curve does not correct on the downside, because neither a co-product producer nor an indebted primary producer responds to the vanadium price with output. It corrects only when enough primary capacity is destroyed or enough co-product steel capacity is idled. For a buyer, that means the downside is persistent and the upside is violent.

Three scenarios

The band is close to symmetric in value terms, which understates how asymmetric the outcomes actually are. The conservative case is not a demand failure: it is what happens if recovery stalls at 80.1% and the price recovers only to the top of the producer’s cash cost range. The optimistic case requires two things at once — throughput capital that the base case says will not be financed, and a battery-grade premium that is contracted rather than spot. The downside is a continuation. The upside is an achievement.

Contagion, and the one number that carries it

Because Brazil is 80.06% one company, the country-level risk is not a portfolio of mine risks. A single production interruption at Maracás Menchen removes roughly four-fifths of the national pool and roughly a third of US pentoxide imports in the same month. That concentration does not appear in a country credit rating or in a trade statistic; it appears only when a buyer maps its supply chain to assets rather than to countries, which is the practical recommendation of this chapter.

Pheonix records overall market risk as high and geopolitical exposure as high, on the grounds that the market’s largest supplier is a single asset in a single jurisdiction and its largest customer has no primary production of its own. Substitution risk is recorded as moderate: substitutes exist in steel, none exists in aerospace titanium alloys, and the battery application has no substitute at all because the electrolyte is the metal.

Related Pheonix coverage: graphite mining for the single-jurisdiction concentration problem in another battery material, and cobalt mining for the co-product supply dynamic in its sharpest form.

Regulatory Landscape

What Brazil licenses

Vanadium production in Brazil sits inside the general mining regime rather than a bespoke one. Mineral titles — exploration authorisation, mining concession, licensing and the associated registrations — are administered by the Agência Nacional de Mineração (ANM), which also operates the mining register, the annual production reporting system and the collection of the financial compensation for mineral exploitation (CFEM). State environmental licensing sits alongside the federal title, and exports are registered through the federal foreign-trade system.

Two structural features of that regime matter to this market. CFEM is levied on gross revenue rather than on profit, so it scales with the price line and not with the volume line — which means it takes a smaller share of a trough tonne than an income-based royalty would, and a larger share of a recovered one. And the ANM maintains a dedicated critical and strategic minerals programme, which signals policy intent to support projects rather than a set of instruments already written around vanadium. Pheonix records approval pathway as licence-based and regulatory complexity as moderate.

What the United States imports

The demand-side regime is not a restriction; it is an absence. The United States produced no vanadium from primary ore and met 58% of its apparent consumption from imports, with metallurgical use accounting for about 94% of reported consumption in 2023. Nothing in that position restricts Brazilian material — it rewards it. The import relationship described by the 47% share is the direct consequence of a jurisdiction choosing not to mine a metal it consumes.

That is also why the market’s policy risk runs in one direction. A tariff, a sourcing preference or a critical-minerals procurement programme in the United States would not reduce demand for vanadium; it would reprice the origin of supply, and the only origin that can absorb a reprice without rebuilding a mine is the one already shipping. Brazil’s exposure to US policy is therefore a margin risk rather than a volume risk, which is the opposite of the usual commodity pattern.

Why the instrument is not a ban

The distinction that matters to a compliance team is between an instrument that limits a substance and an instrument that limits a concentration. Nothing in the vanadium chain prohibits a grade or an origin. What the regimes do is require documentation: provenance, assay, and in some applications a chain of custody. That pushes the competitive advantage toward producers who can produce a specification reliably — which is the same conclusion the technology chapter reaches from the metallurgical side, arrived at from the regulatory side.

The regulatory cost line

For a producer at USD 4.86 a pound, every ad valorem charge compounds the trough. For a buyer, the practical implication is that regulatory cost is a small, stable, price-linked line in a cost stack dominated by the mine gate, and that negotiating documentation requirements into a contract costs more management attention than the royalty costs money. Pheonix’s assessment is that regulation is not the constraint on Brazilian vanadium supply between 2026 and 2033; the price is.

Disclosure: the Brazilian framework is stated here as the operative regime rather than re-read instrument by instrument for this cycle. No rate, threshold or deadline is asserted beyond the existence and function of the instruments named. Buyers should confirm the current consolidated texts with counsel.

Related Pheonix coverage: the global manganese mining market for a comparable licence-based alloying-metal regime, and lithium mining for the jurisdiction-risk comparison in a battery metal that Brazil also produces.