Global Purified Wet-Process Phosphoric Acid Market Size, Share & Forecast 2026–2033

The gate, and the year it was repriced

Every tonne of lithium iron phosphate cathode contains 0.621 tonnes of 100 per cent phosphoric acid. Refiners do not sell it at 100 per cent; they sell 85 per cent product, so the requirement is 0.731 tonnes of purified acid per tonne of LFP. On the 3.77 million tonnes of LFP cathode that shipped in 2025 — the volume published by this segment’s own cathode report — that is 2.75 million tonnes of purified phosphoric acid, and it is the largest single destination for the product anywhere in the world.

Purified phosphoric acid is the only point in the phosphate chain where a producer genuinely chooses between two industrial systems. The thermal route burns yellow phosphorus, made in an electric furnace from rock and coke at roughly 13,000 kilowatt-hours a tonne, and produces an acid that is pure because its feedstock was distilled. The wet route dissolves rock in sulfuric acid and then has to take the impurities back out — fluorine, sulfate, iron, aluminium, magnesium, arsenic, and for battery grades the magnetic foreign matter that ruins a cathode. The thermal route is a bet on electricity. The wet route is a bet on sulfur. For twenty years the second bet was the cheap one, and every purification plant in China was financed on that spread.

On 28 February 2026 the Middle East war closed the Strait of Hormuz, which normally moves 40,000 to 50,000 tonnes of sulfur a day and carries almost half of global seaborne sulfur exports. Platts assessed sulfur at USD 815 to 820 a tonne FOB Middle East and USD 1,200 CFR Brazil on 28 May 2026 — all-time highs, and up nearly 1,100 per cent from around USD 100, in Mosaic’s own description. A tonne of 85 per cent purified acid carries 0.523 tonnes of sulfur. That cost line went from USD 52 a tonne of product to USD 429. The entire raw-material spread the wet route held over the thermal route, measured on 2024 inputs, was RMB 2,838 a tonne. The sulfur move alone is RMB 4,130, at a conversion of 7.18.

So the market grew 28.7 per cent in value in 2026, and the growth was a cost increase. It did that into a surplus: China commissioned 2.48 million tonnes of new purified capacity during 2026 against 4.64 million tonnes of output, a supply surplus of roughly 570,000 tonnes. Cost-push met oversupply, and the difference came out of margin rather than price. That is the year, and it is why the value line of this forecast compounds at 5.49 per cent against 8.17 per cent for tonnage.

Companion analysis. The argument in this report is written up for a non-specialist reader in The Acid That Used to Be the Cheap Route, which follows the same sulfur repricing from the plant gate to the cathode, and shows why both routes to the molecule are now squeezed to single digits for opposite reasons.

Market forecast snapshot 2026 – 2033

Key takeaways

• The growth segment is the cheapest tonne in the market. Battery-bound acid is 2.75 Mt, 49.1 per cent of volume, at a realised USD 1,040 a tonne against USD 1,250 for food grade — so it is 46.4 per cent of value. The pool that grows is the pool that pays least, and it is the pool where the specification is written by the customer rather than the seller.
• The unit consumption is confirmed two ways and they agree to half a per cent. The demand model inside the only primary filing in this market uses 0.178 million tonnes of refined acid per gigawatt-hour of LFP, which against 2,450 tonnes of cathode per GWh is 0.727 tonnes per tonne of cathode. The molar mass of LiFePO₄ says 0.621 tonnes of 100 per cent acid, which at 85 per cent is 0.731. Chemistry and procurement arrive at the same number.
• Sulfur is now the second-largest line in a purified acid plant, and it was the fifth. Rock is unchanged at 3.68 tonnes per tonne of product. Sulfur went from 5.3 per cent of the 2025 product price to 43.7 per cent between April 2025 and May 2026, on the Platts assessments.
• The thermal route has not become cheap; it has merely stopped being the expensive one. Yellow phosphorus at the RMB 26,000 a tonne implied by this segment’s own phosphorous acid report is 87.4 per cent of the RMB 8,000 thermal acid price. Two routes, each now dominated by a single input bill, and the market price sits where the two bills meet.
• The battery-grade premium is a qualification, not a distillation. China’s exported refined acid is quoted at USD 1,335 to 1,380 a tonne FOB, within 2 per cent of food grade — and the export-control threshold that fences the technology is written at the iron phosphate stage, at 10 parts per billion of magnetic foreign matter, which is a specification this acid has to meet and cannot demonstrate on a certificate alone.

Scope: what is counted, and the four things that are not

The perimeter is merchant purified phosphoric acid: acid that has been taken through a purification step and sold as a product at industrial, food, battery or electronic grade. That covers wet-process acid purified by solvent extraction, crystallisation, ion exchange or membrane treatment, and it covers thermal acid sold at food or technical grade, because those two are the same product to the buyer and compete on the same line.

Four exclusions are decisions. Fertiliser-grade phosphoric acid — the 50-million-tonne P₂O₅ pool that never sees a purification step — is a different business priced off a different index. Captive acid consumed inside the fence of the company that made it is not a sale, and a large share of the new Chinese capacity is exactly that. Phosphorous acid and phosphonates are the P(III) branch, sized separately in this segment. Phosphate rock and yellow phosphorus are the two inputs, not the product.

One overlap has to be stated plainly, because it is the easiest double count in the phosphate complex. Purified acid is a component of iron phosphate, and iron phosphate is a component of the LFP invoice published by this segment’s cathode report. At RMB 7,050 a tonne for the acid and RMB 11,650 for iron phosphate, the acid is 46.3 per cent of the precursor price. Add the two markets together and 46 per cent of one is counted twice. This report is a subset of that line, and the two are never summed.

The number, and how it was built

The volume base is a construction and the band around it is roughly plus or minus 12 per cent. It is built from the battery pool upward, because that pool has a measured denominator: 3.77 Mt of LFP cathode in 2025 times 0.731 tonnes of 85 per cent acid gives 2.75 Mt. The remaining 2.85 Mt is food, technical, phosphate-salt and electronic grade, derived from the Chinese output series — 4.64 Mt of purified acid output in 2026, per Mysteel — net of the battery-bound share and grossed up for production outside China.

The price base is the published quote rather than a model. Chinese refined acid at 85 per cent was RMB 7,050 a tonne in April 2025, up 5.49 per cent from the start of that year, per Baiinfo data carried in a listed issuer’s filing. By October 2026 the same specification was quoted at RMB 8,000, range 8,000 to 8,550. In September 2026 food-grade acid was offered at USD 1,350 to 1,380 a tonne FOB China in 35-kilogram drums, and industrial-grade purified acid at USD 1,335 to 1,365 — the two grades within 2 per cent of each other, which is the single most useful fact about the grade ladder. A global realised average of USD 1,100 a tonne sits between the Chinese domestic quote and the export quote and reflects the premium the rest of the world pays for acid it does not make.

Three mechanisms produce the gap between 8.17 per cent volume growth and 5.49 per cent value growth, and only one of them is comfortable. The mix shifts toward the cheapest grade, because battery acid is the only pool growing at double digits. Capacity additions of 2.48 Mt in a single year in China reset the price of the commodity grades. And the sulfur shock is a pass-through that fades: it raises 2026 and then stops raising anything, which is why 2026 is the strongest value year in the forecast and 2033 is priced below 2025.

The cost stack, which is the whole report

Read the last two rows together. Wet-process purification at 2024 input costs was RMB 2,838 a tonne cheaper than the phosphorus bill alone on the thermal route, before the thermal plant paid for oxygen, steam or conversion. At the May 2026 sulfur assessment the advantage is RMB 134. At the CFR Brazil assessment of USD 1,200 it is minus RMB 1,293 — wet-process acid is more expensive than the thermal route’s raw material, which is not a spread, it is a reversal.

The market has already said so. Chinese refined acid rose from RMB 6,683 in January 2025 to RMB 8,000 in October 2026 — 19.7 per cent — in a period when the sulfur line inside it rose by roughly 700 to 1,100 per cent and the industry added 2.48 Mt of capacity. The price did not follow the cost. That is what an oversupplied market does with a cost shock: it passes through a fraction and absorbs the rest, and the absorbed part is the margin of every non-integrated purifier in China.

Five pools, and the inversion inside them

The electronic pool is 2 per cent of the tonnage and 3.4 per cent of the value, and it is the only place in this market where the seller names the price. The battery pool is the opposite in every respect: half the tonnage, the lowest price, and a specification set by a cathode maker who runs six suppliers against each other. Everything that happens to the value line of this forecast for the next eight years happens inside that 49.1 per cent.

Companion analysis: Phosphorous Acid Is Priced by a Power Bill, Not a Rock Price argues the P(III) branch of the phosphorus complex is priced off electricity. This report is the P(V) counterpart, and reaches the same structural conclusion from a different input: the purified acid business is not priced by the molecule either. It is priced by whichever of its two feedstock bills is moving, and in 2026 that stopped being rock.

Related research in this segment: Global Phosphorous Acid & Phosphonate Derivatives Market 2026–2033 — the thermal branch this acid competes with for the same rock and the same power; Global LFP Cathode Active Material Market 2026–2033 — the demand denominator, 3.77 Mt of cathode, that every tonne in the largest pool above is derived from; and Global Sodium-Ion Battery Materials Market 2026–2033 — the polyanion sodium cathode is a phosphate cathode, and it draws on this acid too.

Companion analysis: The Acid That Used to Be the Cheap Route — how the sulfur shock reversed the wet route's cost advantage over the thermal branch, and why a 570,000-tonne surplus stopped the industry passing it on.

Table des matières

• Value chain: rock, sulfur and the two bills that set the price
• Technology and innovation: purification routes and the grade ladder
• Competitive landscape: five Chinese rock owners and a merchant tier that has no rock
• Regulatory landscape: export control written at the precursor, and a thermal route policy has closed
• Investment activity: RMB 4,264 a tonne of capex against a spread that has inverted
• Market risk: sulfur, surplus, and a demand model that runs four years ahead of this one

Paysage concurrentiel

Competitive landscape: five rock owners and a merchant tier that has no rock

This market looks fragmented in a supplier list and is concentrated in a balance sheet. The tonnage is spread across dozens of plants in Guizhou, Yunnan, Hubei and Sichuan, plus a European and North American thermal base; but the tonnage that is both merchant and priceable is dominated by a handful of companies that own phosphate rock, own sulfuric acid, and increasingly own the iron phosphate plant that consumes the acid. The rest of the industry buys all three.

The largest single position

Guizhou Phosphorus & Chemical Group describes itself as the world’s largest supplier of high-quality wet-process purified phosphoric acid, and on the disclosed tonnage that is not a marketing claim. It is executing toward a 3 million tonne PPA position and has since published a 4 million tonne capacity layout. In September 2025 it completed a 72-hour performance test on a post-treatment line that lifted a single complex from 400,000 to 600,000 tonnes a year — and did it on a project launched in June 2024, with mechanical completion in March 2025 and first feed in May. That is a build cadence the rest of the industry cannot match, and it is the reason the Chinese supply curve shifts rather than slopes.

The rest of the tier looks like this. Yunnan Yuntianhua holds rock, wet acid and a purification position in the southwest. Hubei Xingfa combines rock, yellow phosphorus and a downstream phosphate chemical franchise, which makes it the one player that can arbitrage the thermal and wet routes internally rather than buy the spread. Chuanheng and Xinyangfeng hold integrated rock-to-precursor positions. Liuguo Chemical is the merchant case with a rock-adjacent parent: 150,000 tonnes of existing refined capacity built with a university process partner since 2010, plus a 280,000 tonne battery-grade project with Wanhua taking a 45 per cent stake in the project company — the customer buying into the supplier, which is the clearest signal available about where the scarcity is expected to sit.

Outside China: thermal, and deliberately not merchant

The Western position is thermal and largely captive to food and industrial customers. ICL runs the largest non-Chinese purified franchise, integrated to its own rock. Prayon licenses wet-process technology and operates purification in Europe. There is no merchant battery-grade capacity of scale in North America or Europe, which is why the export quote carries a premium and why the region is the fastest-growing in this forecast from a base close to zero.

What the two supply numbers say together

China commissioned 2.48 million tonnes of new purified capacity during 2026 across eleven units, split roughly 1.0 million in the first half and 1.48 million in the second, against forecast output of 4.64 million tonnes and a surplus of about 570,000 tonnes. Set those against a global merchant pool of 5.60 million tonnes and the scale of the addition is the story: the year’s new capacity is 44 per cent of the base market.

That is not a market that is short of acid. It is a market that is short of customers qualified to buy the grade that was built, and the utilisation consequence is uneven rather than general. The integrated, rock-owning, battery-qualified plants run hard because their output is contractually spoken for by a precursor plant next door. The merchant plant selling industrial and food grade into a spot market runs at whatever the sulfur-inflated quote allows, which in 2026 meant producers “holding prices and reluctant to sell” — the classic posture of an industry that believes its cost justifies a price its customers will not pay.

Where pricing power actually sits

Not with the seller of acid. The battery pool — half the tonnage — is bought by a precursor maker who qualifies two or three suppliers per line, prices against a published index and can substitute technical MAP for acid in an ammonium-route plant. The food pool is stickier, because the qualification is an audit and switching costs a customer sign-off. The electronic pool is the only one where the supplier sets terms.

So the competitive structure is a barbell. At one end, six integrated companies with rock, acid and precursor, whose returns are a mining return dressed as a chemical return. At the other, a merchant tier whose cost is set by two commodity markets it does not control and whose price is set by a customer who does. The middle of that barbell — a merchant purifier with battery-grade ambition and no rock — is where the 2026 sulfur shock did the most damage, and where the next round of consolidation will come from.

Chaîne de valeur

Value chain: three bills, and the one that changed hands

A purified acid plant has three inputs and no fourth. It buys phosphate rock, it buys sulfur — as sulfuric acid, or as brimstone and a burner — and it buys the energy and reagents that separate one from the other. Everything else in the cost stack is small enough to be a rounding item, and everything in the commercial structure of this industry follows from which of those three the plant owns.

The revealed intensities

The rock figure is the most reliable number in the chain because it comes from a producer rather than a model: 1 tonne of 85 per cent purified acid consumes 3.68 tonnes of phosphate rock, a ratio disclosed in a listed issuer’s filing and attributed there to the operating experience of the world’s largest wet-purification operator. The sulfur figure follows from the stoichiometry of the attack and from the excess every plant runs: 1.883 tonnes of sulfuric acid per tonne of contained H₃PO₄, which is 0.616 tonnes of sulfur, or 0.523 tonnes per tonne of the 85 per cent product. The third bill — steam, power, extraction solvent, and the neutralizing agent used to drop out fluoride and sulfate — is where the difference between a well-run plant and a marginal one sits, and no one publishes it.

What the numbers mean is simple and it is the centre of this report. On 2024 inputs, rock was USD 526 and sulfur USD 52 per tonne of product against a price of RMB 7,050, or USD 982. Rock was 54 per cent of the price and sulfur 5 per cent. By May 2026 rock had moved by single digits and sulfur had moved by a factor of eight, to USD 429 or 44 per cent of the 2025 price. The chain did not add a cost. It swapped which of its two dominant bills mattered, and the industry’s entire asset base was positioned for the old answer.

Where the acid goes, and the first real overlap

The battery route runs acid into iron phosphate and iron phosphate into cathode. The molar arithmetic is tight: a tonne of FePO₄ contains 0.205 tonnes of phosphorus and needs 0.650 tonnes of 100 per cent acid; a tonne of LiFePO₄ contains 0.956 tonnes of iron phosphate and 0.621 tonnes of acid. Those two routes to the same number differ by rounding, and both land on 0.731 tonnes of 85 per cent product per tonne of cathode.

That means the acid is not a small additive to the cathode. At the RMB 7,050 refined-acid quote and the RMB 11,650 iron phosphate price, the acid is 46.3 per cent of the precursor price and 16.6 per cent of the cathode invoice at the April 2026 build. Any reader who adds the purified-acid market to the iron phosphate market, or to the LFP market, is counting the same tonne twice. This report is a slice, not a sibling, and the slice is almost half of the thing above it.

The three iron phosphate routes, and why they matter here

Iron phosphate is made three ways, and the industry names them after the phosphorus source: the sodium route, the ammonium route and the iron route. Only the iron route starts from purified acid. The ammonium route starts from industrial-grade monoammonium phosphate, which is itself made from purified acid but is also sold into water-soluble fertiliser and flame retardants, and the sodium route starts from sodium phosphate salts. This is not a technicality. It is a substitution channel: a precursor maker facing a sulfur-inflated acid quote can switch phosphorus source without changing the cathode, and several Chinese plants are built with flexible lines specifically to make that switch.

The strategic consequence is that purified acid does not own its largest customer. It shares that customer with technical MAP, and the share it holds is set by the relative price of two products that come out of the same purification plant. The one asset that cannot be substituted is the rock, which is why the value chain’s returns sit at the mine and at the mine’s owner, and why the merchant purifier without a mine is the weakest position in the chain — stronger than the precursor maker, weaker than everyone upstream of it.

Captive, merchant, and the share of the market that is not for sale

A large and growing share of Chinese purified capacity is built inside a rock owner’s fence and consumed by that owner’s own iron phosphate plant. It never clears a market price. The merchant pool — the 5.60 million tonnes this report sizes — is what remains after that internal transfer, and its size is therefore a function of integration strategy rather than of demand. Every new integrated complex in Guizhou or Hubei removes tonnage from the priceable pool and adds tonnage to the unpriceable one, while the downstream plant it feeds becomes harder for a merchant supplier to win. That is the quiet structural trend of the value chain, and it is a bigger threat to a merchant purifier than the sulfur price.

Activité d'investissement

Investment activity: RMB 4,264 a tonne of capex bought against a spread that no longer exists

This is a capital-intensive, mature, process industry, and its investment case has always rested on one comparison: the cost of a tonne of annual capacity against the gross spread that tonne can earn. In 2026 that comparison became the most uncomfortable number in the phosphate chain, and it is possible to state it exactly.

The cleanest data point in the market

Liuguo Chemical is building 280,000 tonnes a year of battery-grade refined phosphoric acid in Yichang for a total investment of RMB 1.194 billion, funded partly by a private placement of up to RMB 800 million. That is RMB 4,264 per tonne of nameplate. The disclosed economics are an after-tax IRR of 19.80 per cent and a 6.55-year payback including construction. The output split is 233,500 tonnes sold as iron phosphate raw material and 50,000 tonnes converted to high-grade flame retardant — a flexible line, in other words, that can follow whichever of the two phosphorus products is worth more.

Set that capex against the 2025 gross spread. At a RMB 7,050 price and RMB 4,153 of 2024-basis rock and sulfur, the spread is RMB 2,902 a tonne. So one tonne of capacity costs 1.47 times one year of the gross spread it was underwritten to earn. That is a normal chemical project: roughly a seven-year payback on gross margin before fixed costs, which is exactly what the issuer’s 6.55-year figure implies. The project is not aggressive. It is arithmetic performed on a set of input prices.

And the input prices moved by more than the spread

The sulfur line alone moved RMB 4,130 a tonne between the filing and May 2026. That is 1.42 times the entire gross spread the project was modelled on. Nothing in the operating plan changed; the rock intensity, the recovery, the grade mix and the customer list are all as filed. One purchased input repriced the project out of the money, and the sensitivity is not subtle: every USD 100 a tonne on sulfur moves the raw-material cost of a tonne of 85 per cent acid by USD 52, or RMB 375.

That is the single most important investment fact in the phosphate complex for the next three years, and it is not specific to one issuer. Every purified-acid project approved between 2021 and 2025 was approved inside a sulfur band the industry had not left since 2008. The 2.48 million tonnes commissioned in China in 2026 were financed under the same assumption.

What the strategic money is doing instead

Two behaviours are visible and both point the same way. The first is vertical integration by the customer: Wanhua, through its battery subsidiary, took a 45 per cent stake in the project company building the 280,000-tonne line and holds a matching shareholding in the parent’s project vehicle. A cathode-adjacent buyer is buying equity in its acid supplier, which is what a customer does when it expects the molecule to be short and wants the option rather than the spot price.

The second is consolidation of the specialty platform. In the adjacent phosphonate chain this segment has already recorded PCBL’s acquisition of Aquapharm at INR 3,800 crore, 1.86 times sales and 10.3 times EBITDA, and Dorf Ketal’s exclusive talks to buy Italmatch from Bain for a reported USD 1.6 billion. Neither buyer was buying capacity. Both were buying a validated position on a customer’s specification — and that, not a plant, is what a purified-acid franchise is worth once the commodity spread has gone.

What to underwrite and what not to

Underwrite rock. A tonne of purified capacity backed by a mine has a different cost curve from one that buys its rock at the marginal price, and the 2026 shock proved the difference. Underwrite sulfur security, whether through a smelter-acid contract, a brimstone offtake with a term, or a process design with lower acid intensity. Underwrite battery-grade qualification, because it is the only grade in the market where the customer cannot switch on price alone.

Do not underwrite merchant purification into a spot industrial and food market in a year when 44 per cent of the base market’s tonnage is being added as new capacity. That business earned a spread against a thermal alternative that is now priced at parity, in a market where the price is set by the buyer and the cost by two commodity indices the seller does not control. The capital intensity is high, the returns are real and the position is the weakest in the chain.

Technologie et innovation

Technology and innovation: the impurity, not the concentration

Purification is not a distillation and it is not a concentration step. It is a sequence of removals, and every commercial grade in this market is defined by which removals have been done rather than by how strong the acid is. That distinction explains why the price ladder is far flatter than the grade names suggest, and why the difficulty in this industry sits in the last stage rather than the first.

Two routes, two kinds of purity

Thermal acid is pure by construction. Yellow phosphorus is made in an electric-arc furnace from rock, coke and silica at roughly 13,000 kilowatt-hours a tonne, distils at 280 degrees Celsius, and leaves the volatile metals and the sulfate behind in the slag. Burning it to P₄O₁₀ and hydrating it gives an acid that meets food-grade specification without a purification step. It is also, in China, effectively closed to new capacity: the route is restricted by policy on energy and emissions grounds, which means the thermal pool is a fixed asset base that depreciates and is not replaced.

Wet-process acid starts at 26 to 30 per cent P₂O₅ carrying everything the rock contained. Purification is a chain: pretreatment to defecate sulfate and drop out fluoride and heavy metals, extraction into an organic phase — tributyl phosphate, or a phosphoric ester, or a synergistic mixture with a higher distribution coefficient — scrubbing of the loaded organic to strip the co-extracted impurities, back-extraction into water to recover a clean acid, and post-treatment to polish the last of the organics and the residual metals. Each stage costs yield, and yield is the number that separates the operators.

The last stage is where the capacity is

This is the single most useful operational fact in the sector. When the largest wet-purification operator in the world brought a 600,000 tonne a year line to full rate in 2025, the milestone that took a 72-hour performance test was not the extractor. It was the post-treatment refining single train, which ran at 1,854 tonnes a day. Extraction is licensed, understood and available. The polishing train is the part that has to be debugged against a specific rock, and it is the part that a new entrant understates when it publishes a nameplate.

The consequence is that announced capacity and qualified capacity are different quantities, and the gap between them is not a commissioning lag but a technical one. A plant that can make industrial-grade acid from its extraction train has to add post-treatment to make battery-grade, and battery-grade is where the growth is. That is why a project can be physically complete and still not be in this market.

The grade ladder has four rungs, and the top two are close together

Industrial. Sulfate and fluoride controlled, metals roughly bounded, sold into metal treatment, water treatment and phosphate salts. This is a commodity and is priced as one.

Food. Arsenic, heavy metals and fluoride to a statutory limit; the specification is a regulation, the test is standardised, and the buyer is a food company auditing a supplier rather than a technical function controlling an impurity.

Battery. The limits that matter are not the ones on the certificate. For an iron-phosphate precursor the sensitivity is to aluminium, magnesium, sulfate and magnetic foreign matter, because those are what degrade compaction density and first-cycle efficiency in the cathode. China’s export-control catalogue writes the precursor limb at below 10 parts per billion of magnetic foreign matter — a specification this acid has to meet, and, as the next chapter argues, a control surface that reaches back into this plant.

Electronic. Parts-per-billion metals across the board, particle counts, and a packaging regime. It is two per cent of volume and 3.4 per cent of value, and it is the only grade in the market where an incumbent’s qualification is worth more than its capacity.

What is actually being innovated

Three things, in order of commercial value. Yield, because phosphorus recovered is phosphorus not bought, and a five-point swing on recovery is worth more than a five per cent price move. Rock tolerance, because Chinese rock is declining in grade and a process that can take 26 per cent P₂O₅ instead of 30 per cent is a process that can still be fed. And sulfuric acid intensity, which nobody discussed before 2026 and everybody discusses now: the stoichiometric requirement is 1.5 tonnes of acid per tonne of contained H₃PO₄ and the industry runs about 1.88, so roughly a fifth of the sulfur bill is excess. Hemihydrate and dihydrate process designs, and the recrystallisation routes that trade a little energy for a lot of sulfuric acid, are the technologies that matter in a USD 1,200 sulfur market and were uneconomic in a USD 100 one.

The patent base is thin and the process knowledge is tacit. This is a mature technology in the strict sense: the last fifteen years have produced debottlenecking and control improvements rather than new chemistry. Which is precisely why the barrier is a trained shift team and a rock-specific operating envelope, and why capacity arrives slower than the announcements imply.

Risque de marché

Market risk: six ways this is wrong, and the one that would break it fastest

The forecast is USD 6,160.0 million in 2025 growing 5.49 per cent to USD 9,450.0 million in 2033, on 8.17 per cent volume growth and a 2.48 per cent annual decline in realised price per tonne. The risks below are ordered by how quickly they would move that path, not by how likely they are.

Risk one: sulfur does not normalise, or normalises and comes back

This is the dominant risk and it is already live. The May 2026 assessments — USD 815 to 820 FOB Middle East, USD 1,200 CFR Brazil — put the raw-material bill of a purified acid plant at 97.3 per cent of the 2025 product price at the FOB assessment and above it at the Brazilian one. Mosaic suspended rock production at two Brazilian mines for 45 days and 30 days respectively and withdrew its 2026 phosphate guidance. Producers in China responded by holding offers rather than cutting them, which is what an industry does when it believes cost justifies a price its customers will not pay.

Two things could soften it, and neither is measured here. China’s sulfuric acid supply carries a large smelter-acid component that is a by-product of non-ferrous smelting and is therefore not priced one-for-one off brimstone; that is a genuine structural cushion and I have not put a number on it because I have not measured the ratio. And a Hormuz reopening would collapse the FOB assessment faster than the CFR one, which would help coastal Chinese buyers before inland ones. The downside case in this report assumes the 2026 price level persists rather than reverts — which, given that a reversion would be a tailwind, is the conservative choice.

Risk two: the only demand model in the market runs four years ahead of this one

The issuer filing behind this report publishes a 2030 battery-grade demand figure of 677.82万吨 — 6.78 million tonnes, split 441.08万吨 for new energy vehicles and 236.74万吨 for storage, built on 0.178 million tonnes of refined acid per gigawatt-hour. At the 0.731 unit consumption this report uses, 6.78 Mt implies 9.28 million tonnes of LFP cathode in 2030, or about 3,800 GWh.

This segment’s own cathode report puts LFP at 3.77 Mt in 2025 growing 12.83 per cent a year, which reaches 6.89 Mt in 2030 and 9.90 Mt in 2033. The issuer’s model therefore arrives in 2030 at what my cathode path reaches in 2032 and a half — a 35 per cent gap in the same year. I have not adopted it, because a forecast built on a customer’s upside case is not a forecast. It is the optimistic path, and the honest statement is that this market’s tonnage depends on an LFP trajectory that two credible constructions disagree about by a third.

Risk three: technical MAP takes the phosphorus share without taking the acid

Iron phosphate is made by three routes and only one of them starts from purified acid. The ammonium route starts from industrial-grade monoammonium phosphate, which is a different product from the same purification plant but a direct substitute from the precursor maker’s point of view. Several new Chinese lines are explicitly built with flexible switching. In a year when acid is sulfur-inflated and MAP is ammonia-linked, that switch is free money for the precursor maker and lost tonnage for the acid seller — and no customs line records it.

Risk four: the surplus is a price event with a two-year lag

2.48 million tonnes of new capacity against a 5.60 million tonne base is a 44 per cent addition in one year, and the 570,000-tonne surplus is only the first year of it. Utilisation, not capacity, is what clears a market, and the integrated plants will run at any price because their acid is a transfer price to their own precursor plant. That is what makes the surplus structurally bearish: the rational response of an integrated producer to a price fall is to run harder, which pushes the merchant producer out first. The forecast’s price line already assumes a 2.48 per cent annual decline; the risk is that it is front-loaded into 2027 and 2028.

Risk five: rock grade, and the 3.68 that is not constant

China’s rock is declining in grade and produced under quota. The 3.68-tonne intensity is an operating figure from a specific mine and a specific plant; as grades fall it rises, at roughly the rate of the grade decline. A move from 3.68 to 4.00 tonnes per tonne of acid is a USD 46 raw-material increase at USD 143 rock — smaller than the sulfur shock, permanent rather than temporary, and it compounds.

Risk six: substitution in the cathode, which is slower and bigger

Sodium-ion polyanion cathodes are phosphate cathodes and draw on this acid, so the sodium transition is not a threat to the phosphate molecule — it is a threat to the volume per kilowatt-hour if the chemistry’s energy density improves, because a cathode that stores more per kilogram needs less acid per kilowatt-hour. And LMFP, which substitutes manganese for part of the iron, changes the precursor chemistry and the acid intensity with it. Both are second-order for 2033 and first-order for anything longer.

Six numbers refused, in print: no global purified-acid capacity total, because only China discloses a capacity series and the rest of the world does not; no per-plant utilisation rate, because no one publishes one and the estimates circulating are reconstructions; no measured merchant-versus-captive split of the Chinese pool, which is the largest single uncertainty in the volume base; no Chinese sulfur contract price, only the international assessments; no electronic-grade tonnage, which is why that pool carries a construction rather than a measurement; and no third-party market size, because syndicated estimates for this category surfaced during research and appear nowhere in this report.

Paysage réglementaire

Regulatory landscape: the fence is drawn at the precursor, and the acid has to clear it

Three regimes bear on this market and they push in different directions. One controls where the technology can go, one controls where the feedstock can come from, and one controls what a plant is allowed to do with its waste. The first two are Chinese and the third is local, and it is the third that has quietly decided how much capacity actually gets built.

The export control is written one step downstream

China’s MOFCOM and MOST Announcement No. 28 of 2025, effective 15 July 2025, added battery cathode material preparation technology to the restricted-export catalogue as item 252604X. It is written as a specification rather than a company list: LFP at at least 2.58 g/cm³ compaction density, at least 160 mAh/g, at least 97 per cent first-cycle efficiency, with LMFP set at 2.38 g/cm³ and 3.85 volts. The limb that matters to this report is the precursor limb, which covers iron phosphate above 1.2 g/cc tap density and below 10 parts per billion of magnetic foreign matter.

Purified phosphoric acid is not itself a listed item. Neither is iron phosphate as a commodity. What is listed is the specification, and the specification is a property of the acid. A plant that can hit 10 parts per billion of magnetic foreign matter can serve the controlled cathode chain; a plant that cannot, cannot. So the control surface sits on the acid’s impurity profile while the acid remains exportable — and that asymmetry is the interesting part, because it means the constraint on Western capacity is a quality gate rather than a shipment licence. This is an inference from the published thresholds and not from any enforcement action, and it is stated as one.

The thermal route is closed to new capacity

China’s thermal phosphoric acid route — yellow phosphorus burned to P₄O₁₀ and hydrated — is restricted for new build on energy and emissions grounds, a constraint stated plainly in a listed issuer’s filing and consistent with the electricity intensity of the furnace. The policy consequence is not that thermal acid disappears; it is that the thermal pool is a fixed and depreciating asset base that no one replaces. That matters more than it appears, because thermal capacity is the industry’s only hedge against a wet-route input shock. The hedge is not being renewed.

Rock is a strategic mineral with a quota and a grade problem

Phosphate rock is listed as a strategic protective mineral in China. Mining is constrained by environmental inspection, by restrictions on extraction and by the Yangtze basin protection programme, and the aggregate effect is a resource that is produced under quota rather than under demand. Roughly 60 per cent of Chinese rock goes to fertiliser, with wet-process acid taking about 12 per cent of the product-side split and rising. Two structural facts follow. Existing mines are declining in grade, which raises the 3.68-tonne rock intensity of a tonne of acid over time; and new mining capacity is being added in named increments — 11.5 million tonnes in 2025 and 8.2 million tonnes in 2026 across a dozen named projects — which is a supply curve managed by permit rather than by price.

The waste rule is the binding local constraint

The wet process produces roughly five tonnes of phosphogypsum per tonne of P₂O₅, and phosphogypsum is a stacked solid waste with residual acidity and fluoride. Chinese rules require utilisation rather than indefinite storage, and utilisation has to find a home in cement retarder, gypsum board or soil amendment — markets that are themselves cyclical and regional. A purification plant without a phosphogypsum offtake does not get its environmental permit, and the permit is the gate on the project rather than the financing. This is the least discussed and most decisive constraint on Western and Chinese capacity alike, and it is why announcements convert to nameplate slower than any financial model assumes.

Food law, and why the food premium is thin

Food-grade acid is governed by statutory limits on arsenic, heavy metals and fluoride under national food-additive standards and the international Codex and FCC monographs. Outside China the qualification includes an FSSC 22000 or equivalent site audit, which is a recurring cost and a real barrier. Inside China the audit is lighter and the specification is a laboratory result, which is exactly why Chinese food-grade acid exports within 2 per cent of industrial grade: the market prices the certificate, not the audit.