CEM I Shrinks While CEM II Adds the Growth: A Market That Is Not Growing, a Product That Is Changing

The cement market is not growing. That is now the settled fact of the sector: 3,900 million tonnes in 2024, revised down from an earlier 4,000, and 3,800 million in 2025, with a forecast that adds 0.37 per cent of tonnage across the whole period to 2033. If you only ever looked at volume, this would be one of the least interesting industries on earth.

The product mix inside that flat envelope is where the entire commercial story sits. Ordinary Portland cement — CEM I, the commodity grade, the thing most people mean when they say cement — goes from USD 89,490 million to USD 88,320 million. It shrinks, by 1.31 per cent. The blended grades, CEM II, go from USD 103,360 million to USD 147,840 million, up 43.03 per cent. Together the two move from USD 192,850 million to USD 236,160 million, 22.46 per cent, on essentially no additional tonnage.

CEM II’s share of that pair goes from 53.6 per cent to 62.6 per cent. Nothing was built to cause that. No new city was poured. The same quantity of powder simply moved from the grade that competes on price to the grade that competes on specification.

What is actually being bought

A tonne of CEM I is a tonne of clinker, ground. It has one variable — fineness — and one price, and it is sold against a published number in every market that has a published number. A tonne of CEM II is a formulation: clinker plus limestone, or slag, or fly ash, or calcined clay, in a proportion chosen for a strength class, a sulphate exposure, a heat-of-hydration limit or a carbon declaration. It has several variables and no single reference price, because two CEM II cements meeting the same standard are not the same product.

That is the whole mechanism. When a market’s tonnage stops growing, the only remaining route to revenue is to sell a product whose price cannot be read off a screen. The cement industry did not choose this; it was pushed into it by a carbon cost that attaches to the clinker and by public clients who now write clinker content into their tenders. But having been pushed, it is now the most valuable thing the industry owns.

The point is easier to see one step down the chain. In the global ready-mix concrete market, standard structural concrete falls from 64.1 per cent of value to 49.0 per cent over the same period. The commodity mix loses fifteen points of share while the total market barely moves. It is the same event at two removes: cement moves value from CEM I into CEM II, and concrete moves value from standard structural mix into everything that is not standard structural mix. In both cases the undifferentiated tonne is being quietly retired while the aggregate number looks stable.

Who decides, and why that is the whole question

A product shift of this size needs a decision-maker, and in construction the decision-maker is never the buyer of the material. It is whoever writes the mix into the specification — a public-client engineer, a structural designer, a ready-mix technical manager responding to a tender. The cement producer’s commercial outcome is determined upstream of its own sales conversation.

This is why the shift correlates so strongly with procurement policy rather than with price. Where a public client sets a minimum clinker replacement, the specification changes and the volume follows. Where the client sets nothing, the contractor buys on the lowest compliant price and CEM I wins on the day, because CEM I is the cheapest way to hit a strength class when nobody is counting carbon. The cement market is therefore not really being reshaped by the industry that makes it. It is being reshaped by the tender documents of its largest customers.

That has an uncomfortable corollary for anyone forecasting this market on price elasticities. A 43 per cent value increase in the blended grades is not a demand response to a price signal. It is a series of specification decisions, made one project at a time, in a policy environment that is tightening. Those do not glide; they step.

The price the carbon cost puts on the commodity grade

CEM I is also the grade most exposed to carbon pricing, because carbon cost attaches to clinker and CEM I is clinker. At approximately 0.60 tonnes of CO₂ per tonne of cement, the sector’s 3,800 million tonnes carries about 2,280 million tonnes of CO₂; at EUR 80 a tonne that is EUR 48 on every tonne of cement, or 42.1 per cent of a German ex-works price of EUR 113.98. The Carbon Border Adjustment Mechanism, in its definitive regime from 1 January 2026, is built to replace free allocation as free allocation is removed.

Put those two facts together and the direction of the product mix stops being a matter of preference. A producer selling CEM I into a market where the carbon cost is fully internalised is selling a product whose cost base is rising by a percentage of its selling price every year. A producer selling CEM II has diluted that cost across a tonne of material that includes a component carrying no carbon liability at all. The blend is not a marketing position. It is the only place in this industry where the carbon cost can be partially avoided without losing the sale.

Which is also why the substitution supply question matters so much. CEM II volumes can only grow as fast as the materials that go into them, and the two cheapest of those are byproducts of industries that are shrinking. The blended grade’s 43 per cent value expansion is, in physical terms, a bet on slag, fly ash, limestone and calcined clay arriving in the right place at the right fineness. That bet is not obviously won.

What to watch instead of the total

Three things are worth tracking rather than the headline tonnage. First, the CEM II share of value, which is the cleanest single indicator of whether the industry is succeeding in selling a specification rather than a commodity. Second, the spread between CEM I and CEM II price per tonne in any given market, because a narrowing spread tells you the blended grade is being commoditised and the mix shift is delivering volume rather than margin. Third, the pattern in cement packaging, where a shift toward blended product changes bulk density and flow at the packer head — and where the bag market is already shrinking at the volume end.

The industry’s own framing — a mature market, flat tonnage, modest price growth — misses the only thing happening inside it. The tonne is not growing. The tonne is being re-specified, and the re-specification is worth 22.46 per cent.

Where this sits in the cluster

The blend only exists if the materials that go into it arrive, and they are not all equally available — The Cement Industry Is Retiring Its Own Cheapest Substitutes works through why the two cheapest of them are supply-constrained. The force that actually moves the mix is procurement rather than price, which is the subject of Ireland Told Public Bodies to Stop Buying CEM I. And the same share shift, read one step down the chain and aggregated across cement, concrete and packaging, is in Peak Cement, Falling Concrete.

About the method

This analysis draws on 70 per cent secondary and 30 per cent primary sources. The secondary layer is public and primary: product-level market data by cement type, national cement statistics and shipment series, the CBAM regulation as published on EUR-Lex, and public procurement instruments that set clinker-replacement requirements. The primary layer is structured fieldwork with specifiers — public-client engineers and ready-mix technical managers — on who writes the mix into the tender and how much latitude the contractor actually has; it is weighted into the conclusions as evidence about decision-making rather than quoted as individual attribution. All arithmetic here — the 1.31 per cent decline in CEM I, the 43.03 per cent expansion in CEM II, the 22.46 per cent combined movement, the 53.6 to 62.6 per cent share shift, the 2,280 million tonnes of CO₂ and the EUR 48 per tonne — was computed deterministically from the published inputs. No third-party market-research estimate is used, and no competing research provider is cited.