Who 'Controls' the World's Phosphate? Reserves and Production Are Different Numbers

Two statements about phosphate rock circulate widely, often in the same article, and they appear to contradict each other:

• China is by far the world's largest phosphate producer.
• Morocco controls around 70% of the world's phosphate.

Both are true. They are not in tension, because one is a statement about annual production and the other is a statement about reserves — and those are different quantities that happen to point at different countries. Treating "controls 70%" as a production or supply figure, which is the common error, produces a picture of the phosphate market that is simply wrong about who ships the tonnage today.

Phosphate is also, for a conveyor manufacturer, one of the most instructive bulk commodities there is: the single longest conveyor system on Earth exists to move it. So the distinction is worth getting right.

The production picture: China leads, Morocco is second

The US Geological Survey publishes the reference dataset every January. Here is the 2025 production estimate, in millions of tonnes of marketable phosphate rock:

Source: USGS, Mineral Commodity Summaries 2026 — Phosphate Rock ↗ (published January 2026). Figures are marketable phosphate rock; "e" estimates in the original. USGS data is public domain.

On production, China is dominant — it mines roughly three times what Morocco does and nearly half the world total. Any claim that Morocco "supplies" or "produces" most of the world's phosphate is contradicted by the primary source by a factor of three.

The reserves picture: Morocco holds most of what is left in the ground

Now the other table from the same USGS page — reserves, in millions of tonnes:

Source: USGS, Mineral Commodity Summaries 2026 — Phosphate Rock ↗.

Here Morocco is overwhelming: about 50 billion of the world's ~73 billion tonnes of reserves, roughly 68%. This — not production — is the source of the "70%" figure. It is a real and important number. It just describes a different thing: what is economically extractable in the ground today, not what came out of a mine last year.

So the reconciliation is straightforward. China mines the most now; Morocco holds the most for later. A country can do either, both, or neither, and the two rankings need not agree.

Reserves, resources, and "how many years are left"

The second confusion that attaches to phosphate is the leap from reserves to a doomsday clock — "peak phosphorus," "we have X years of phosphate left." This misreads what a reserve figure is.

Three terms get conflated, and the USGS keeps them strictly separate:

• Reserves — the part of an identified resource that could be economically extracted at the time of the estimate, given current prices and technology. This is the 73,000 Mt figure. It moves as prices, technology and exploration change.
• Resources — the broader identified and estimated quantity, most of which is not currently economic. USGS states world phosphate resources are "more than 300 billion tons" — over four times reserves.
• Static "years left" — reserves divided by annual production. At 73,000 Mt of reserves and 250 Mt/yr, that ratio is roughly 290 years. Against the >300 billion tonne resource base it is well over a thousand. Neither is a forecast; a static ratio assumes production never changes and reserves never grow, and both assumptions are always false.

The USGS puts the conclusion in one sentence: "There are no imminent shortages of phosphate rock." The genuine phosphate concerns are about concentration and access — one country holding most reserves, and price and trade shocks — not about the mineral running out. It is also worth remembering why this matters at all: "There are no substitutes for phosphorus in agriculture," as the same USGS page notes. You cannot engineer phosphorus out of the food system the way you can substitute one metal for another.

Why a conveyor company is reading the phosphate tables

Because phosphate is the commodity that produced the longest conveyor on the planet.

At Bou Craa in Western Sahara, phosphate rock travels roughly 98 km (about 61 miles) by conveyor from the mine to the coast near Laâyoune — the world's longest conveyor belt system ↗, per NASA's Earth Observatory, moving on the order of 2,000 tonnes per hour. Wind blows fine white ore off the line into a streak that is visible from orbit.

That single installation captures why the reserves-versus-production distinction is an engineering question and not only a market-trivia one:

• Reserves are geographically fixed; mines go where the rock is. When most of the economically extractable phosphate sits in one region, the material handling problem is defined by that region's geography — long overland distances from inland deposits to a port. Long, high-tension overland conveyors are the answer to exactly that problem, which is why the extreme example is a phosphate line and not, say, a gravel pit.
• Tonnage rate, not annual output, sizes the belt. Bou Craa's ~2,000 t/hr is an instantaneous rate. As with coal fleets running on a balancing duty, sizing a handling system from annual tonnage rather than peak hourly rate is a design error. Our belt capacity calculator works from the hourly rate, belt width and speed that actually govern the design.
• Long overland duty drives the belt construction. Distances measured in tens of kilometres put the carcass tension beyond what fabric belts carry, which is the domain of steel-cord belts — chosen for overland lines precisely because of their low elongation over long centres.

The point is not that every phosphate producer runs a 98 km belt. It is that the distribution of reserves — the number people misattribute to production — is what puts the mine far from the port, and that is what puts the conveyor in the picture.

If you are citing phosphate figures

The same discipline the cement emissions data and coal capacity figures needed applies here:

• Say production or reserves. They point at different countries — China (~44%) for production, Morocco (~68%) for reserves. A bare "controls X% of phosphate" is ambiguous until you say which.
• Do not turn reserves into "years left." Reserves are a snapshot at current economics, not a countdown. Quote the resource figure (>300 billion t) alongside if you want to discuss longevity, and quote USGS's "no imminent shortages" rather than implying one.
• Cite the year and the edition. USGS publishes annually and revises prior years; the 2026 edition revised several 2024 country figures.
• Attribute to USGS, not to an article about it. The Mineral Commodity Summaries ↗ are the primary, public-domain source; secondary coverage is where the production/reserves conflation usually creeps in.

What we did not verify

• Exact 2025 figures are estimates. The 2025 production numbers carry USGS's "estimated" flag and will be revised in the 2027 edition. We present them as the current best published estimate, not final data.
• Western Sahara's political status. Bou Craa lies in Western Sahara, a territory whose sovereignty is disputed; USGS and other bodies handle the attribution of its output differently. We have described the conveyor and its geography, and taken no position on the sovereignty question, which is outside our scope.
• Bou Craa's current throughput. The ~2,000 t/hr and ~98 km figures are widely reported and consistent with NASA's account, but we did not obtain a current operating figure from the operator; treat them as the well-established published values, not a 2025 operational reading.
• P₂O₅ vs rock. Reserves and production here are marketable rock tonnage. Fertilizer demand is often quoted as P₂O₅ content (world P₂O₅ fertilizer consumption was ~47.8 Mt in 2025 per USGS) — a different basis that should not be compared directly with rock tonnage.

Sources

• USGS (January 2026). Mineral Commodity Summaries 2026 — Phosphate Rock ↗ — all production and reserves figures. Public domain.
• USGS, Mineral Commodity Summaries 2026 ↗ — full publication and definitions of reserves vs resources (Appendix C).
• NASA Earth Observatory, World's Longest Conveyor Belt System ↗ — Bou Craa length, throughput and satellite visibility.

Figures as published in the USGS 2026 edition covering calendar year 2025. Corrections welcome — send them over and we will update with attribution.

Related engineering reading: Conveyor System Parts labels the components of a belt conveyor with links to specifications, and the belt capacity calculator sizes throughput from width, speed and material.