Everything difficult about this subject follows from one reaction, and EPA’s greenhouse gas inventory states it in stoichiometric terms rather than rhetorical ones. Calcium carbonate, arriving at the kiln as limestone or as recovered cement kiln dust, is heated through roughly 700 to 1,000 degrees Celsius and splits into lime and carbon dioxide. Each mole of calcium carbonate that enters yields one mole of calcium oxide and one mole of carbon dioxide, and that carbon dioxide is vented to the atmosphere as part of the kiln exhaust. The hotter sintering stage that follows, where lime combines with alumina, iron oxide and silica to form clinker nodules, is exothermic and produces few process emissions of its own.
The consequence is a hard bound on what cleaner energy can accomplish. NIST’s 2023 workshop report on circularity in concrete puts the number on it: fossil fuel heating accounts for only 40 percent of the total carbon dioxide from clinker production, so improving energy efficiency has only a partial impact. Electrify the kiln, fire it on hydrogen, run it on captured waste heat, and roughly three fifths of the emissions are still sitting in the exhaust because they came out of the rock.
What the U.S. figure counts, and what sits in another chapter
For 2022, EPA’s inventory reports carbon dioxide emissions from cement production at 41.9 million metric tons of carbon dioxide equivalent, against clinker production of 80,500 kilotons, up 25 percent on 1990. The inventory describes process carbon dioxide from cement production as the second largest source of industrial carbon dioxide emissions in the United States.
The boundary matters more than the total. That 41.9 million tons is the process figure alone, because emissions from the fuels burned to heat the kiln are accounted for in the inventory’s energy chapter. A company or sector figure that bundles combustion is not comparable to it.
The industry producing those emissions is more concentrated than its significance suggests. The U.S. Geological Survey estimates 2024 portland and blended cement production at 84 million tons, made at 99 plants across 34 states and Puerto Rico, with Texas, Missouri, California and Florida accounting for roughly 43 percent of output. Ninety-nine plants is a short list, which is why a state specification change can move a real share of national output and why one closure tightens a regional market.
NIST organizes the pathways into three mechanisms: improve energy efficiency, lower the cement content in concrete, and capture emissions from cement production for use in concrete. Against the 40 percent bound, the first is capped. The other two carry the load, and each has a supply story attached.
The supplementary material supply question has two credible answers
Lowering cement content means substituting supplementary cementitious materials, overwhelmingly fly ash from coal combustion and ground granulated blast furnace slag from iron making. NIST states the technical limit and the supply risk together: cement replacement by traditional supplementary materials has limits beyond which mechanical or durability properties are compromised, and their availability will likely decrease with the departure from coal-fired power and reduced demand for pig iron.
The industry that supplies the ash tells a different story with its own numbers. The American Coal Ash Association’s survey of the 2024 production year, released in December 2025, reports that use of coal fly ash in concrete rose from 11.9 million tons in 2023 to 14.6 million tons in 2024, even as production of new coal combustion products fell from 66.7 million tons to 63.6 million. Seventy-two percent of coal ash produced in 2024 was recycled, up from 69 percent. The association attributes that to a rapidly growing practice of harvesting previously disposed ash from legacy impoundments, and its executive director’s assessment is that delayed coal plant closures combined with expanded harvesting portend ample supplies for construction markets for many years to come.
Both accounts can be true, and reading them together beats choosing. Fresh production is falling and will keep falling. Ash reaching concrete rose anyway, because a second channel opened and less went to landfill. Note the same survey’s less convenient line: use of coal combustion products in cement production, as distinct from concrete, fell from 6.8 million tons to 5.3 million in the same year. The material is moving toward the higher-value application as it tightens, which is what a market under pressure looks like. A specification writer planning a twenty-year program should treat harvested ash as a qualitatively different input.
Limestone is the substitute with volume behind it
NIST names the alternatives that could supply the tonnage: limestone and calcined clay have the potential to provide sufficient volumes, with the caveat that replacement by them may be limited by performance and compositional constraints, impairment of early strength among them. Early strength is not a laboratory nicety on a paving job: it determines when the lane opens.
Blended cement carrying interground limestone is the version of that idea already in commercial production. The Portland Cement Association, an interested party in the framing, reported in May 2022 that 44 states accepted portland-limestone cements, describing the product as allowing 5 to 15 percent ground, uncalcined limestone to be blended or interground with clinker under ASTM C595, and claiming a carbon dioxide footprint up to 10 percent lower at similar performance. The modesty of that figure is the reason to take it seriously: displacing a tenth of the clinker should cut process emissions by roughly a tenth, and a claim proportional to its mechanism is more credible than a large one. For the composition limits an agency can specify, the authoritative documents are ASTM C595 and AASHTO M 240 in their current editions together with the state’s approved materials list, not any summary of them.
Mineralization is chemically sound and institutionally unproven
Carbonation, the reabsorption of carbon dioxide into hydrated cement phases to form carbonates, is not a novel invention. NIST describes it as taking place naturally at ambient temperature and pressure over many years as curing and cured concrete absorbs atmospheric carbon dioxide. Two features make it attractive as a deliberate process. It is thermodynamically favorable and requires little, if any, extrinsic energy, and the carbonates formed are generally stable enough to outlast the life of the structure, which makes the capture permanent. It also tolerates messy inputs.
What is not settled is everything between the chemistry and a paving contract. The workshop’s list of challenges here is short and pointed: uncertainty about the scalability of carbon dioxide injection technology, which it calls a critical factor for adoption; measurement methods to quantify uptake described as nascent; and difficulty distinguishing natural carbonation from accelerated carbonation. The last is a verification problem with commercial consequences, because a credit claimed for accelerated uptake has to be separable from uptake that would have happened anyway.
State research programs are working the pavement version of the question. Minnesota’s Department of Transportation, through the National Road Research Alliance, has run a project on concrete made with a commercial mineralization technology, testing whether sequestration can reduce cement content without compromising resilience and durability, with staged tasks from laboratory work and constructability evaluation through one-year and two-year field monitoring. Anyone wanting a verdict should read that program’s published final report rather than a summary of its interim tasks, and two-year monitoring is also why answers arrive after procurement decisions.
The federal purchasing rule stops at the plant gate
The mechanism now pulling low-carbon materials into highway projects is procurement, and its scope is narrower than most descriptions of it. Section 60506 of the Inflation Reduction Act of 2022 funds FHWA to reimburse or incentivize construction materials with substantially lower levels of embodied greenhouse gas emissions compared with estimated industry averages, as determined by the Administrator of EPA. That determination issued on December 22, 2022, covering asphalt mixtures, concrete and cement, steel, glass and qualifying assemblies.
The definition is a percentile rather than a target. Substantially lower means a global warming potential, evidenced by an environmental product declaration, in the best performing 20th percentile against similar products. Where nothing in that band is available at the project location, the 40th percentile qualifies; where that is unavailable too, better than the estimated industry average qualifies, with written documentation and approval by both regional management and central office technical staff. The declaration must be facility-specific and product-specific, cradle-to-gate, Type III, compliant with the applicable product category rule and with ISO 14025 and ISO 21930, and accompanied by the upstream plant’s ENERGY STAR score where one exists. The machinery behind those documents is treated in environmental product declarations.
The consequential detail is the boundary. EPA prioritizes the production stage, modules A1 to A3, commonly called cradle-to-gate, in line with the broader federal Buy Clean initiative, and states that use and disposal stages are not within the scope of the current determination. A mixture that lasts sixty years instead of thirty earns nothing under that rule, because the benefit lands in the use stage. A mixture with ten percent less clinker and a shorter service life scores. That is a deliberate simplification with a known distortion attached, and it is why the durability arguments and the embodied carbon arguments across more on pavement materials keep talking past each other. The accounting problem is set out in embodied carbon in highway projects, and the case for building to the long horizon runs through long-life pavement.
The same brief adds a footnote that undercuts a term used loosely across this field. Global warming potential, as the phrase appears in declarations and Buy Clean policy, denotes an emissions intensity per unit of material, and the brief states plainly that this is inconsistent with the definition used by the Intergovernmental Panel on Climate Change and in national reporting, where it is an index of radiative forcing. Two communities are using one phrase for two quantities.
What is holding the transition
NIST’s diagnosis of the obstacles is institutional rather than technical: insufficient policies and standards, no agreed way to compare reduction pathways against each other, unstandardized performance requirements for low-carbon cement, and an absence of firm offtake agreements between owners and producers. Underneath all of it the report identifies liability risk, stating that there is a significant need to prove long-term material performance and durability before new materials will be accepted.
An agency that specifies a novel binder across a hundred lane miles and sees premature cracking in year twelve has created a reconstruction liability far larger than the emissions it avoided. The way out is evidence of the kind that accumulates slowly, on instrumented test sections under real traffic. Agencies that funded those sections a decade ago are the ones with options now, which is an argument for research money rather than for patience.