Editorial Article
The Acid That Used to Be the Cheap Route
The advantage nobody had to defend
Both routes to phosphoric acid end at the same molecule. The difference is the bill. The thermal route reduces phosphate rock in an electric furnace to yellow phosphorus and oxidises the result. The wet route digests rock with sulfuric acid and purifies what comes out. For most of the industry’s working memory the wet route won on cost so reliably that the comparison stopped being an argument and became a background fact — the sort of thing repeated without being re-derived.
Pheonix’s sizing of the purified wet-process market puts it at USD 6.16 billion in 2025, rising to USD 9.45 billion by 2033 — 5.49% a year on volume that compounds at 8.17%, from 5.60 million tonnes to 10.50 million. Value growing slower than tonnage is not a growth forecast, it is a price forecast: revenue per tonne falls from roughly USD 1,100 to USD 900 by 2033, 2.48% a year. The industry is being asked to more than double its output into a falling price. What follows is about the cost line underneath that plan, because in 2026 it moved, and it did not move down.
A tonne of acid is 3.68 tonnes of rock and 0.523 tonnes of sulfur
That is the whole wet-route recipe at 85% concentration, and for twenty years those two inputs behaved in opposite directions: rock drifted up, sulfur was cheap and volatile around a low mean. On 2024 inputs a tonne of purified product carried USD 526 of rock at USD 143 a tonne and USD 52 of sulfur. Raw material came to USD 578, which was 58.9% of the USD 982 a tonne the product sold for in 2025. Rock was 54% of the selling price and sulfur was 5%. That is a comfortable cost structure for anyone who owns a purification plant and no mine, because the dominant input at least moves slowly.
Then the sulfur market left its band. The Strait of Hormuz closed on 28 February 2026, and by 28 May Platts was assessing sulfur at USD 815 to USD 820 FOB Middle East and USD 1,200 CFR Brazil. On a 0.523-tonne sulfur requirement, that line became USD 429 a tonne of acid:
- Rock, 3.68 tonnes at USD 143 — USD 526, unchanged
- Sulfur, 0.523 tonnes — USD 52 on 2024 inputs, USD 429 at the late-May 2026 assessment
- Raw material — USD 578, then USD 955
- Share of the USD 982 selling price — 58.9%, then 97.3%
The wet route’s cost advantage over the thermal route was never enormous. Measured against the phosphorus bill a thermal producer carries, it was about RMB 2,838 a tonne of 85% acid. The sulfur move alone is RMB 4,130. It did not narrow the advantage; it took it out and left change. On the Middle East assessment the wet route retains roughly RMB 134 a tonne of spread. On the Brazilian import price it is negative — about RMB 1,293 worse than simply buying yellow phosphorus and oxidising it.
The other side of the comparison is not healthy either
It is tempting to read this as a thermal-route victory. It is not. Yellow phosphorus at the RMB 26,000 a tonne implied by this segment’s own phosphorous acid work works out at RMB 6,991 a tonne of 85% acid — 87.4% of the RMB 8,000 that thermal acid sells for. The thermal producer is living on 12.6%, with an electricity bill as its largest input and no ability to pass through a power price that is set by a grid, not a contract.
So both routes to the same molecule are now squeezed to single-digit percentages of spread, for opposite reasons: one by a sulfur shock, the other by a power price. That is not a margin problem in one technology. It is a statement about where the value in this chain has moved — from the acid maker to whoever owns the rock, the sulfur and the energy, ideally in the same balance sheet.
And the price barely moved, because the market is drowning in capacity
Here is where the usual cost-push argument breaks. When input costs rise 700% to 1,100% on a line that is 5% of your price, you expect the product to follow. Chinese refined acid went from RMB 6,683 a tonne in January 2025 to RMB 8,000 in October 2026 — 19.7% over twenty-one months, against a sulfur input that rose roughly sevenfold.
The reason is volume. China commissioned 2.48 million tonnes of new purified capacity in 2026 across eleven units, against output of 4.64 million tonnes. That is a 570,000-tonne surplus, and the addition is equal to 44% of the base market being served. Cost-push into oversupply does not produce price; it produces margin compression, and the margin it compresses belongs to every merchant purifier in China who does not own a mine and did not hedge sulfur.
The strategic consequence is uncomfortable for the growth story. The 2.48 million tonnes were justified by a cathode demand curve. They were built at 2024 sulfur prices. And a plant that cannot earn a return at USD 820 sulfur cannot be un-built at USD 1,200.
The customer is the cathode, and the cathode counts acid twice
Half of this market is one customer. LiFePO4 is 19.63% phosphorus, so a tonne of cathode holds 0.621 tonnes of 100% acid — 0.731 tonnes of the 85% product. The procurement side agrees independently: the demand model inside the only primary filing in this market uses 0.178 万t of refined acid per GWh, which at 2,450 tonnes of cathode per GWh works out to 0.727. Two different methods, chemistry and a contract-derived run rate, land within half a per cent of each other. That is the most reliable number in the chain.
Applied to 3.77 million tonnes of 2025 LFP cathode, it means 2.75 million tonnes of acid — 49.1% of this market’s entire volume. Battery-grade acid is the majority of the tonnage and 46.4% of the value, realising USD 1,040 a tonne against USD 1,250 for food grade. The fastest-growing, most strategically important customer in the market pays the least per tonne for it, because the specification rewards consistency rather than purity headroom, and because the buyer is a cathode maker with a 9% gross spread of its own.
The grade premium is a certificate, not a purification cost
September 2026 export offers make the point better than any cost model. Food-grade acid was quoted at USD 1,350 to USD 1,380 FOB. Industrial grade was quoted at USD 1,335 to USD 1,365. The gap is within 2%, on two products that carry different documentation, different audit trails and different customer qualification. Whatever the premium is paying for, it is largely not the incremental purification step.
That matters for anyone modelling a move up the grade ladder as a margin strategy. The ladder is real but the rungs are close together, and the price difference is defended by certificates and customer approvals rather than by a cost barrier. Certificates can be bought. Which means the premium erodes as soon as enough capacity holds the same certificate — exactly the condition 2.48 million tonnes of new capacity creates.
The capex arithmetic, and the IRR it was built on
Liuguo’s 280,000-tonne battery-grade project cost RMB 1.194 billion, which is RMB 4,264 per tonne of nameplate. Against a 2025 gross spread of RMB 2,902 a tonne, capex equals 1.47 years of its own gross spread. The sulfur move alone is 1.42 years of it. Read that again: the input-cost change of the last twelve months is worth almost exactly the entire capital cost of the plant, expressed in years of margin.
The filing carries a 19.80% IRR and a 6.55-year payback. Both are arithmetic performed on a sulfur band the industry had not left since 2008. They are not wrong as calculations; they are answers to a question that changed after they were written. And the most informative detail in the whole project is who bought into it: Wanhua took 45% of the project company. That is a customer buying equity in its acid supplier — a hedge against exactly the pass-through this article has just described, and a signal about who the project’s economics are actually being built to serve.
What to watch, and what would undo the argument
Everything here turns on one variable. Sulfur at USD 815 to USD 820 FOB is a strait-closure price, not a structural one. If Hormuz reopens and sulfur returns to a few hundred dollars, the wet route’s RMB 2,838 spread comes back and this becomes a story about a bad eighteen months. If it does not reopen cleanly, or if the Middle East’s export logistics stay impaired, then purification economics relocate: toward rock-plus-sulfur integration, toward producers who are net long sulfur rather than net short it, and away from the independent merchant purifier who was always the marginal ton.
Two things would falsify the pessimistic reading and are worth tracking honestly. First, purified capacity running below the announced nameplate — 2.48 million tonnes commissioned is not 2.48 million tonnes produced, and 2026-27 utilisation will tell you whether the surplus is as large as the announcement list implies. Second, battery-grade acid finding a premium it currently does not have, which would happen if cathode makers start paying for consistency rather than treating acid as a commodity input. Neither is visible yet.
One figure I did not adopt, and the reason is worth stating. The issuer’s own model puts 2030 battery-grade demand at 6.78 million tonnes, which at 0.731 implies 9.28 million tonnes of LFP cathode in 2030. Pheonix’s cathode work reaches 6.89 million tonnes in 2030 and 9.90 million in 2033 — it arrives in 2032 and a half where the issuer arrives in 2030, a gap of roughly 35%. That is the optimistic path dressed as a forecast, and this report does not use it.
One boundary, before anyone adds two of our reports together
At RMB 7,050 for the acid and RMB 11,650 for iron phosphate, purified acid is 46.3% of the precursor price and 16.6% of the LFP cathode invoice. This market is therefore a slice of the iron phosphate line inside our cathode report, never a sibling to it. Add the two together and 46% of one is counted twice. The same discipline runs the other way: this report sizes the acid, not the rock it is made from, not the sulfur it is digested with, and not the cathode it ends up inside.
The acid was the cheap route because rock and sulfur were cheap and an electric furnace was not. Two of those three conditions held in 2026. The industry built 44% of its own market on the assumption that all three would.
Companion research: Global Purified Wet-Process Phosphoric Acid Market Size, Share & Forecast 2026–2033, with the thermal branch priced in Global Phosphorous Acid & Phosphonate Derivatives and the cathode customer in Global LFP Cathode Active Material.
Frequently Asked Questions
Why did purified wet-process phosphoric acid lose its cost advantage over the thermal route?
Because the wet route's only real advantage was the price of its two inputs, and one of them left its historical range. A tonne of 85% purified acid is 3.68 tonnes of phosphate rock and 0.523 tonnes of sulfur. On 2024 inputs that was USD 526 of rock and USD 52 of sulfur, or 58.9% of the USD 982 a tonne the product sold for. After the Strait of Hormuz closed on 28 February 2026 and sulfur was assessed at USD 815 to USD 820 FOB Middle East on 28 May, the same sulfur line became USD 429 and raw material reached USD 955 — 97.3% of the 2025 selling price. The wet route's spread over the thermal phosphorus bill was about RMB 2,838 a tonne; the sulfur move alone is RMB 4,130.
Did the price of purified phosphoric acid rise to cover the sulfur cost?
Only marginally. Chinese refined acid went from RMB 6,683 a tonne in January 2025 to RMB 8,000 in October 2026, about 19.7%, against a sulfur input that rose 700% to 1,100%. The reason is capacity: China commissioned 2.48 million tonnes of new purified capacity in 2026 across eleven units against output of 4.64 million tonnes, a surplus of roughly 570,000 tonnes and an addition equal to 44% of the base market. Cost-push into oversupply compresses margin rather than lifting price, and the margin that compresses belongs to merchant purifiers without a mine or a sulfur hedge.
How much purified phosphoric acid goes into a tonne of LFP cathode?
About 0.731 tonnes of 85% product per tonne of cathode, which is 0.621 tonnes of 100% acid. That follows from LiFePO4 being 19.63% phosphorus by weight. It is corroborated independently on the procurement side: the demand model inside the only primary filing in this market uses 0.178 万t of refined acid per GWh, which at 2,450 tonnes of cathode per GWh works out to 0.727 tonnes — a difference of half a per cent. Applied to 3.77 million tonnes of 2025 LFP cathode, that is 2.75 million tonnes of acid, or 49.1% of the entire purified market's volume.
Is battery-grade acid the most valuable part of the purified phosphoric acid market?
No — it is the largest by volume and the cheapest by tonne. Battery-bound acid realises about USD 1,040 a tonne against USD 1,250 for food grade, so it delivers 46.4% of market value from 49.1% of volume. The grade ladder is also shallower than it looks: in September 2026, food-grade export offers were USD 1,350 to USD 1,380 FOB while industrial grade was USD 1,335 to USD 1,365 — within 2% of each other. The premium is defended by certification and customer qualification, not by a step-change in purification cost, which is why it erodes as more capacity holds the same certificates.