The Weight of What We Build: Concrete, Steel, and the Emissions No Solar Panel Can Offset

Every conversation about decarbonization eventually arrives at the same image: a wind turbine on a ridgeline, a field of solar panels, a battery bank humming in a substation. What that image never shows is the several hundred tonnes of concrete anchoring the turbine to the ground, or the steel in the tower, or the reinforced foundations under the substation. Cement and steel together account for roughly a seventh of global carbon dioxide emissions — each contributing something in the range of seven to eight percent. They are the two materials that make modern construction possible, and they are the two that clean electricity cannot fix.

The reason is chemistry rather than fuel. Making cement requires heating limestone until it breaks down into lime and carbon dioxide, a reaction called calcination. Somewhere around half to sixty percent of cement's emissions come from this reaction itself, not from the kiln's fuel. You can run the kiln on hydrogen, electrify it entirely, and the calcium carbonate will still surrender its carbon. Primary steelmaking has a parallel problem: the blast furnace uses coke both as fuel and as the chemical agent that strips oxygen from iron ore. Carbon is not incidental to these processes. It is the mechanism.

This creates a strange inversion in green building. As insulation, glazing, and heat pumps drive operational energy use down, the share of a building's lifetime carbon locked into its materials rises. A highly efficient new office block may emit more carbon before anyone occupies it than it will during decades of operation. The industry has a term for this — embodied carbon — but most building codes still regulate only the operational half, which means the more efficient our buildings become, the more thoroughly our regulations measure the wrong thing.

The solutions that exist are real but partial. Clinker substitution replaces some of the cement with materials like calcined clay or industrial byproducts, cutting emissions meaningfully without new technology. Electric arc furnaces running on recycled scrap already produce a large share of steel in some markets, though scrap supply cannot meet total demand. Hydrogen-based direct reduction of iron is genuinely transformative and genuinely expensive, and the first commercial plants are only now proving the economics. Carbon capture on cement kilns remains the option most often promised and least often delivered.

What may matter more than any single technology is the thing nobody markets: building less, and building lighter. Structural engineers routinely overdesign because safety margins are cheap and litigation is not. Buildings are demolished at forty years that could stand for a hundred and forty. Reuse of existing structures, longer design lives, and simple material efficiency could cut demand substantially before a single new process is commercialized — and unlike hydrogen steel, they require no pilot plant.

The uncomfortable conclusion is that the energy transition is itself a construction project of extraordinary scale. Every gigawatt of renewable capacity, every kilometre of new transmission, every seawall built for adaptation draws on the two materials we have not learned to make cleanly. We are, for now, building the low-carbon world out of high-carbon stuff. That is not an argument against building it. It is an argument for treating materials policy as climate policy, rather than as a footnote to it.