The world's most-used manufactured material begins by taking a rock apart — and the carbon dioxide that comes out of it is not a by-product of the fire, but of the rock itself.
More of it is used, by weight, than any other manufactured thing on Earth.
Cement making is really two chemical stages separated by a gap. In the kiln, heat drives carbon dioxide out of limestone. Later — in a different place, often months later — water rebuilds the material into something solid. Everything in between is grinding, mixing and waiting.
Cement is not concrete, and the difference matters for everything that follows. Cement is a fine grey powder — the glue. Concrete is what you get when that glue is mixed with sand, gravel and water. Cement is roughly a tenth of concrete by weight, but it carries almost all of its carbon.
Cement is made almost entirely from limestone, because what the kiln actually needs is calcium — and limestone is the cheapest calcium on Earth.
The bulk of the mix, and the reason the industry exists where it does. Chalk and marl do the same job. The kiln wants the calcium locked inside it; the carbon it is bonded to is the problem.
Calcium alone would give quicklime, not cement. Clay supplies the silicon and aluminium that let calcium form the silicate minerals which later react with water.
Quarried rock is never quite the right composition, so small amounts of silica, iron oxide and alumina are added to hit a precise chemical target. The industry calls these corrective materials.
Added after the kiln, not before. Around 4–5% of every cement, it exists solely to slow the setting reaction down — without it, concrete would stiffen before it could be placed.
Getting a tonne of cement means quarrying rather more than a tonne of rock, because roughly a third of the limestone's mass leaves as gas. The raw meal entering the kiln is typically around four parts limestone to one part clay — the exact blend is monitored continuously, because a cement that is chemically slightly wrong sets slowly, weakly, or not at all.
A cement kiln is a slowly rotating steel tube, tilted a few degrees, lined with brick and fired from the low end. Raw meal enters cold at the top and works downhill against the flame. Two distinct things happen to it on the way.
Heating limestone releases carbon dioxide — but that only gets you quicklime. Cement needs a second, hotter step in which the material partly melts and re-forms as new minerals.
Before the meal reaches the kiln proper it falls through a tower of cyclones, meeting the kiln's own exhaust gases travelling the other way. The gases give up their heat to the powder, so nothing is wasted twice. By the bottom of the tower the meal is above 900 °C.
There it enters the precalciner, a combustion chamber where the decisive reaction happens: calcium carbonate breaks into calcium oxide and carbon dioxide. This single step accounts for around two-thirds of a cement plant's carbon dioxide — and in a modern precalciner kiln it also consumes most of the fuel.
Limestone is calcium bonded to carbon and oxygen. Heat breaks that bond, and the carbon leaves as gas. There is no way to obtain the calcium without releasing it — chemists call the reaction calcination.
CaCO3→CaO + CO2Quicklime is not cement. It will not set under water, and mixed with water it simply slakes and eventually reverts to limestone. To make a binder that hardens permanently, the calcium has to be locked into new minerals — and that needs the clay.
In the hottest zone the material reaches around 1,450 °C, and something more than heating occurs: the meal partially melts. A fraction of it turns liquid, and that liquid acts as a solvent, letting calcium and silicon migrate and combine. Solid lumps form in the melt — dark, glassy nodules the size of gravel, called clinker. Engineers call this process sintering: the material fuses without ever becoming fully molten.
Lime plus silica gives the calcium silicates that make cement work. Two dominate. Alite, tricalcium silicate, reacts quickly with water and provides early strength; belite, dicalcium silicate, reacts slowly and contributes strength over months. The balance between them is why some cements harden fast and others keep gaining strength for years.
3 CaO + SiO2 → Ca3SiO5 aliteClinker is quenched deliberately fast. Cooled slowly, the useful minerals would decompose back into less reactive forms, and the cement would be weaker. The rush of air that does the quenching is then drawn into the burner, so the clinker's own heat helps fire the kiln that made it.
Clinker leaves the kiln as hard nodules. Ground to powder with a few percent gypsum, it becomes ordinary Portland cement. But almost no modern cement is pure clinker — and the reason is arithmetic rather than chemistry.
Every tonne of clinker replaced by something else is a tonne of limestone that never goes through the kiln. Blending is the cement industry's single largest lever on its own emissions.
Some materials will react usefully with the by-products of cement setting, contributing strength of their own rather than merely padding out the mix. These are supplementary cementitious materials, and the two workhorses are industrial leftovers: fly ash from coal-fired power stations and ground granulated blast furnace slag from steelmaking. Both react with the calcium hydroxide that Portland cement releases as it hardens, and both improve durability as well as cutting carbon.
The awkwardness is that these are by-products of industries that are themselves shrinking or decarbonising. As coal power closes and steelmaking shifts away from blast furnaces, the supply of the cement industry's favourite carbon-saving ingredients falls with it. The replacements now being scaled up are finely ground limestone — used unburnt, as filler — and calcined clay, which is heated but at far lower temperatures than a cement kiln and releases no carbon from the rock.
| Cement family | What's blended in | Why it's used |
|---|---|---|
| Ordinary Portland | Clinker + gypsum only | The reference material. Highest strength per tonne, highest carbon per tonne. |
| Portland-limestone | Ground limestone, unburnt | The fastest-growing blend. In the United States, 95% of blended cement shipments are now this type. |
| Slag cement | Blast furnace slag | Slower to gain strength, but denser and more resistant to chemical attack. |
| Fly ash blends | Coal combustion ash | Improves workability and long-term durability; supply now declining with coal power. |
| Calcined clay (LC3) | Heat-treated clay + limestone | Clay is abundant almost everywhere and calcining it releases no carbon from the mineral itself. |
The industry tracks a single ratio: how much clinker is in a tonne of finished binder. The global average currently sits at about 0.63, and the cement and concrete industry's own roadmap projects 0.58 by 2030 and 0.52 by 2050. Lower is better — but there is a floor, because clinker is what actually makes the material set.
The binding constraint is not always technical. Availability of suitable blending materials varies sharply by region, and in several developed and emerging markets the industry names client acceptance — specifiers unwilling to depart from familiar mixes — as a current barrier to using this lever fully.
Cement leaves the plant as an inert powder and can sit in a silo for months. It becomes concrete only when someone adds water, and what follows is not evaporation but chemistry: the reverse, in a sense, of what the kiln did.
Concrete sets because a gel of interlocking crystals grows through the mix and knits the stones together. That is why it hardens perfectly well under water, and why leaving it to "dry out" ruins it.
By volume, concrete is mostly rock. Sand and gravel — the aggregate — make up the bulk of it, and they are there because they are cheap, hard and dimensionally stable. The cement paste occupies the spaces between them. A useful way to picture it: concrete is a pile of stones, and cement is what stops them being a pile.
When water meets the ground clinker, the calcium silicates dissolve and immediately begin precipitating something new: a stiff, poorly ordered gel of calcium silicate hydrate. It grows outward from every cement grain as a mass of tiny interlocking needles and sheets, threading through the gaps, gripping the aggregate and gradually filling the space that water once occupied. Cement chemists write it C-S-H, with dashes rather than fixed proportions, because its composition genuinely varies from place to place within the same paste.
The reaction also throws off calcium hydroxide, which contributes little strength directly but matters enormously in two ways: it is what supplementary materials feed on, and it keeps the concrete strongly alkaline — which turns out to be the reason reinforced concrete works at all.
Alite reacts first and fastest, giving concrete its strength in the first week. Belite works slowly, adding strength over months. Both produce the same binding gel, and both release heat — which is why a large pour warms noticeably as it sets.
calcium silicates + H2O → C-S-H gel + Ca(OH)2 + heatOnly a modest fraction of the mixing water is consumed by the reaction — roughly a quarter of the cement's weight is enough to hydrate it fully. Any water beyond that has to go somewhere, and when it eventually leaves it leaves behind capillary pores. Those pores are the single biggest determinant of how strong and how durable the finished concrete is.
This is the trade every mix design negotiates. More water makes concrete easier to place and harder to get wrong on site; less water makes it stronger and longer-lived, but stiff and awkward to work. Chemical admixtures now let concrete flow freely at low water content, which is why modern mixes can be both workable and strong.
It also explains curing. Fresh concrete is kept damp, sometimes for weeks, precisely so the reaction can continue. Concrete that dries too early stops hydrating and never reaches its intended strength.
Concrete is superb in compression and poor in tension. Squeeze it and it resists enormously; pull it apart, or bend it so one face stretches, and it cracks at perhaps a tenth of that load. A concrete beam spanning a gap sags, its underside stretches, and it fails — which is why unreinforced concrete is fine for foundations and dams but useless for floors.
Steel has the opposite profile: excellent in tension, and expensive to use in bulk. Casting steel bars into the tension face of a concrete member gives a composite that is strong both ways, and this is the basis of nearly all modern construction.
Two accidents of nature make the pairing work. First, steel and concrete expand at almost exactly the same rate when heated, so temperature swings do not tear the bond apart. Second, the calcium hydroxide left over from hydration keeps the concrete alkaline enough to form a passivating film on the steel, protecting it from rust inside a material that is full of moisture. When that protection eventually fails — as chlorides penetrate, or as the concrete slowly reacts with atmospheric CO2 and loses alkalinity — the steel corrodes, expands, and cracks the concrete from within. Most reinforced concrete that reaches the end of its life does so this way.
Cement is made almost everywhere, because it is cheap and heavy and does not travel well. That should make it a fragmented industry. One country's construction boom made it anything but.
That dominance is now receding. Chinese output fell by just under 8.5% in 2025, to 1.67 billion tonnes from 1.83 billion the year before, and has been declining since 2020 as the country's property investment contracted. Because China is so large a share of the world total, its slowdown pulls the global figure down with it — world production slipped from an estimated 3.9 to 3.8 billion tonnes across the same year.
The pattern elsewhere is different. India is growing steadily and Vietnam has climbed to third place. The United States, by contrast, imports about a fifth of what it consumes, and around seven in ten tonnes of American cement go straight to ready-mixed concrete producers — a reminder that cement is rarely an end product. It is an ingredient in something poured the same day.
Most heavy industry has a decarbonisation story that begins and ends with energy: find a cleaner way to make the heat. Cement has that problem too, but it is the smaller half of the problem.
Around 60% of a cement plant's direct carbon dioxide comes out of the limestone itself, not the fuel. Perfectly clean energy would leave most of the emissions untouched.
The sector emits in excess of 2.5 billion tonnes of carbon dioxide a year. Of the direct emissions, roughly 60% is released by the heated limestone and roughly 40% by burning fuel in the kiln and elsewhere in the plant. That first portion is often called process emissions, and it is unavoidable in the ordinary sense: it is not inefficiency, or a poor choice of fuel, but the stoichiometry of getting calcium out of calcium carbonate.
This is what makes cement genuinely harder than steel or aluminium, and it shapes every credible plan for the industry. There are only three real strategies: use less clinker per tonne of binder, use less binder per structure, and capture the carbon dioxide that is released anyway.
Blend the cement with limestone, slag, ash or calcined clay so less of it comes from the kiln at all.
Better mix design, industrialised batching, and structures designed with carbon as a parameter alongside cost and speed.
Waste-derived fuels already replace part of the coal, with hydrogen and electrified kilns expected later.
The only lever that touches process emissions directly. Not expected to matter at scale until after 2030.
Concrete slowly reabsorbs carbon dioxide from the air over its lifetime, reversing part of the kiln's chemistry — the calcium works its way back toward carbonate. This recarbonation was long known to engineers as a durability issue, since it is also what eventually strips the alkalinity protecting reinforcing steel. Only recently has it been counted in carbon accounting, appearing in the IPCC's Sixth Assessment Report in 2021.
It is a real effect but a modest one. The cement and concrete industry's own roadmap, using a deliberately conservative method, puts global uptake at around 319 Mt of CO2 in 2020 — set against sector emissions above 2,500 Mt. Recarbonation adjusts the ledger. It does not balance it.
There is also a tension the industry does not resolve, and neither should this page. Cement demand rises with development: housing, sanitation, roads and hospitals are made of it, and the regions where consumption is set to grow are the ones with the least of all four. The levers above reduce carbon per tonne. Whether total emissions fall depends on how many tonnes the world decides it needs.