Carbon uptake dynamics of cement-based materials: Linking market structure, material use, and the carbon cycle Authors: Hessam AzariJafari, Ipek Bensu Manav, Motahareh Rahimi, Elizabeth Moore, Bruno Huet, Christophe Levy, Randolph Kirchain Published: 2025 ABSTRACT SignificanceUsing a bottom–up framework, our research provides the first quantification of national-level carbon uptake from cement-based products. Our national-scale results show that the context significantly alters the fraction of CO2 sequestered by the carbonation process. This is consistent with prior work looking at individual buildings and elements. By explicitly accounting for structural geometry, exposure, and regional material use, the results show that national uptake fractions are considerably different than previous global estimates. While our estimate is smaller than previous analyses, it is clear that the carbon uptake of cement-based products is an important, climate-relevant process. It has been widely cited that cement, or more specifically, the concrete and mortar (i.e., cement-based products without coarse aggregates) into which it is formed, is the most widely used engineering material on the planet (1). Given such scale, it is not surprising that cement production generates a significant amount of greenhouse gases (GHGs) [in excess of 2.5 Gt CO2e per year (2)] equivalent to >6% of global emissions (3). Industry stakeholders recognize that this figure must shrink and have committed at the firm, national, and international levels to realize a carbon-neutral cement-based products (CBPs) industry by 2050 (4). Achieving that goal will require not only investment and action throughout the cement value chain [from clinker to construction to end of life (5)] but also an effective, comprehensive system to measure progress. Measuring the impact of cement, like any material, is complex. First, this is true because materials used in production influence the properties of the resulting products, which often changes the GHG emissions of the product during use, far beyond the production gate. Second, in the case of cement, impact measurement is complex because, irrespective of what is blended and formed into it, the CBP remains chemically active toward CO2, binding it as carbonate materials. This paper provides the first-ever bottom–up, national-scale estimate of the extent of this CO2 binding from CBPs. Our approach uses defined archetypes (e.g., 16 specific building types with 11 building elements) and statistical analysis of market volume data to estimate the country-level carbon uptake. Over the life cycle of CBPs, some of the emitted CO₂ is gradually sequestered through carbonation, or carbon uptake, (i.e., the reaction of atmospheric CO₂ with hydrated cement compounds to form solid carbonates) neutralizing a fraction of the initial process emissions. In their most recent report, the Intergovernmental Panel on Climate Change (IPCC) labels carbon uptake by CBPs as a major carbon sink, which sequesters about half of the process* emissions from current global cement production (6). Similarly, Xi at al. (7) reported a 43% process emission sequestration, which was updated to 55% in a later study (8). The extent of carbon uptake varies dramatically from one application to another, being influenced by the composition of the CBP (9)—especially the extent to which alternative binder materials, such as fly ash, slag, and limestone are used; the climatic conditions surrounding the CBP; the geometry, especially the surface area, of the CBP structure; and the durability and end-of-life practices associated with that structure. Given this context sensitivity, we seek to evaluate the amount of CO2 sequestered by CBPs at a national scale while still accounting for thousands of combinations of mixtures, climate, and structural characteristics. Admittedly, it is not tractable to evaluate every context. Instead, we rely on spatially and temporally heterogeneous archetypes to represent various end-use applications (sectors) of CBPs. Using this detailed but comprehensive model, we estimate the relative contribution of each cement end-use over time to the total carbon uptake from all the cement end-use applications. Because of the extent of data required for modeling, we focus our analysis on the United States, but we also carry out an example analysis of a second national market, Mexico, which differs in operational and end-use characteristics from the United States. This second case strongly reinforces the role of context in establishing the extent of carbon uptake. In particular, our results suggest that the current stock of CBPs neutralizes, in the United States, about 13% of the cement calcination emissions in a given year and about 25% in Mexico. Background and Motivation for the Bottom–Up Approach The extent of carbon uptake is dependent on both the reaction rate and the geometry of the CBP, specifically the available surface area. All previous top–down studies estimate available surface area from known cement mass (and therefore concrete volume) and a representative thickness (See the SI Appendix, section S1). The studies cited previously either use a single representative thickness for all applications (10), a single uniform sampling of a range of thicknesses for all applications (7), or, in two cases, one representative thickness for each element type (11, 12). In all cases, these thicknesses are based on the expert opinion of the authors. Therefore, in all the previous studies, the surface area of newly constructed buildings and infrastructure systems was estimated based on the cement consumption in the specific end use (e.g., residential buildings, commercial buildings, pavements, etc.) and a limited number of element thicknesses. Carbon uptake estimation with a simplification of identical surface-to-volume ratio obscures the real variation in design and construction in different regions. Previous studies do not provide details on the use of alternative binders in the material analyzed. Considering that the carbonation process in mixtures incorporating alternative binders is different (in terms of available calcium content and carbonation rate) from that in those made only from pure Portland cement, it is necessary to carefully describe the use of alternative binders to accurately estimate carbon uptake. Finally, all previous studies assume that all end-of-life (EOL) material, once crushed, is exposed to conditions that allow for carbon uptake. Stripple et al. (13) have shown that uptake is limited by stockpile packing. In fact, they observe that the stockpile core remains uncarbonated (for example, assuming a stockpile with a height of 1 m for 3,000 m3 of crushed concrete, only 60% of the volume would be carbonated). To tackle these gaps, we propose a bottom–up approach that allows for more context-sensitive estimation of uptake. Specifically, this approach allows for high-resolution definition of context, including available surface area, exposure condition, and concrete/mortar mix design. In this sense, we incorporate a model of EOL carbon uptake that accounts for the implications of stockpiling. The results of this study can provide a deeper understanding of carbon uptake strategies in the United States and assist cities and transportation authorities in making informed decisions about the potential savings in greenhouse gas emissions and the costs associated with different maintenance, repair, and landfilling practices for concrete materials. Method The modeling framework applied in this paper aims to provide a context-sensitive approach to estimate the amount of CO2 reabsorbed into CBPs during product use and end-of-life. A fundamental challenge is that the extent of carbon uptake varies with the context, including the prevailing atmospheric conditions and the surface area of the structure. Modeling the specifics of each of the millions of structures in the built environment is currently intractable. To address this challenge, we develop archetypes –designs of typical structures—that are used to represent classes of structures (Fig. 1). For each archetype, characteristics of each cementitious element (e.g., slabs, beams, etc.) are computed and used to estimate uptake over time. Global uptake is then estimated by combining element-archetype uptake with information on archetype prevalence. Details of the methodology are provided in the SI Appendix, sections S3–S9. Overview of the bottom–up approach for estimating the carbon uptake of buildings and infrastructure systems. Analyses of individual elements (Left) are aggregated into structures. Prevalent data on structure types are used to scale results to a regional and national scale. Modeling the Carbon Uptake Process. The total amount of CO2 sequestered, Υyn, within a given year, yn, for a national stock of cementitious products is modeled as the sum of uptake from all sectors Υωyn. Sectoral uptake is estimated from the uptake associated with all archetypes in that sector, a∈Aω, built in years yb30%). Carbon Uptake Variations and Influential Factors. To examine the relative influence of different contextual parameters, we conducted a post hoc sensitivity analysis across 50 US states (SI document, Section S10). Among the variables considered, the construction trend and mortar/concrete ratio quantities exhibited the strongest correlations with uptake fraction (r = −0.46 and +0.57, respectively). The importance of geographical variation for uptake accounting is shown in Fig. 4A. Specifically, Fig. 4 plots the uptake fraction (Eq. 4) for each state in the United States. Within the United States, the uptake fraction can vary by fourfold. While multiple competing parameters influence the uptake fraction, we identified two influential factors that systematically contribute to this variation—prevailing construction trend and the mortar/concrete ratio. First, the construction trend (measured as the change in newly constructed surface area over the preceding 5 y) is plotted against modeled uptake in Fig. 4B. The construction trend strongly influences the trend in state-level cement consumption. A steep descending trend implies that more concrete has recently been added to stock (and therefore is intensively participating in carbon uptake) relative to the level of current production. Therefore, a downward trend is associated with a smaller denominator for uptake fraction (i.e., process emissions in the year of assessment) relative to the numerator (i.e., the total absolute uptake), which is still influenced by recent, more intensive cement consumption. Second, the prevalence of mortar applications can increase the uptake fraction (as shown in Fig. 4C). Therefore, for a given unit of cement consumption, mortar applications are associated with larger uptake fractions. These two important factors provide insights into the potential contribution of carbon uptake in the neutralization of cement and concrete emissions. In countries where infrastructure is rapidly being built (i.e., upward construction trend), the uptake fraction would be less pronounced compared to developed countries where infrastructure is more mature. (A) Uptake fraction (as a function of cement calcination emissions). (B) linear trend in the last 5-y construction. (C) cement usage among mortar/concrete application against uptake fractions at the state level. Two-letter regional codes correspond to standard abbreviations (see SI Appendix, Table S7 for full definitions). Sensitivity of Carbon Uptake Estimation to Context. The previous results have shown that carbon uptake on the national scale can be considerable, but more importantly, the extent of uptake is strongly affected by context. In particular, we see that carbon uptake extent and intensity vary based on the element into which it is incorporated, the mix that is used, the local climate, and the relative intensity of use of mortar to concrete. Given these drivers, the authors developed a case study of the Mexican market for cement and concrete. Mexico was chosen because of both differences in cement application (i.e., more intensive use in buildings and the more intensive use of mortar and concrete masonry units) and concrete operations, especially the more intensive use of onsite mixing from bags rather than the reliance on delivered ready-mix concrete. As is shown in Fig. 5A, while the US market for cement has grown over the analysis period (increased about threefold from 1940–2022), the Mexican market has grown more than 80 times larger. Currently, cement use in Mexico is just below half of that in the United States. (A) Annual cement production in the United States and Mexico in 2022. (B) Carbon uptake by end-use application in the United States (Left) and Mexican (Right) cement market (the uptake associated with alterations and repairs is included in the corresponding end-use applications). Fig. 5B plots the distribution of carbon uptake across the end-use applications tracked in this analysis for both the United States (Right) and Mexico (Left). An interesting observation from this figure is that the annual uptake in Mexico today is nearly the same as the uptake in the United States, around 5 and 6 Mt CO2e per year. This similarity emerges even though cement use in Mexico is about half that of the United States. In Mexico, however, cement is used extensively in mortar and CMUs, especially in residential construction. In Mexico, the modeled carbon uptake fraction is approximately 25% (averaged over the trailing 5 y). In other words, the model results indicate that the current stock of CBPs sequesters around one quarter of the industry’s annual process emissions. Based on the analysis shown in Fig. 4B, the current Mexico uptake fraction may yet increase as the market matures and growth rates slow. In the end, comparing the Mexico and the US results makes it clear that local differences in end-use context can dramatically alter the extent of carbon uptake. Discussion This paper has presented the first national-scale, bottom–up analysis of carbon uptake in CBPs. The analyses of a detailed case study of the United States suggest that uptake is significant, currently removing more than 6.5 Mt CO2 per year. Applying the same approach suggests that uptake in Mexico may be approaching the value in the United States, absorbing more than 5 Mt CO2 per year. Given these magnitudes, measurement standards for products, firms, and industries should consider carbon uptake to yield an accurate picture of the net effects of CBPs. More specifically, guidelines for including carbon uptake at the product level (e.g., life-cycle assessment or product category rules) need to represent the complexity of surface geometry, climate conditions, and regional construction practices. At the firm level, it is critical to collect accurate market data for CBP and end-use applications. Finally, a clear standard with detailed modeling from mathematical and empirical perspectives is needed to estimate the sinks and sources of emissions at the regional (e.g., nationwide) level. Notably, uptake occurs both from in-use CBPs (75% in the US analysis) and the EOL of cementitious materials (25% in the US analysis). The importance of EOL practices on carbon uptake is underappreciated. When EOL CBPs are crushed, the surface area increases by more than two orders of magnitude. If the material is not already carbonated, these fresh surfaces can rapidly sequester an amount of CO2 that, in some cases, rivals in-use absorption. Differences in CO₂ uptake across end-use types (SI Appendix, Fig. S14) have direct implications for both mitigation policy and carbon accounting. Short-lived, surface-intensive structures contribute to greater near-term uptake, while long-lived infrastructure provides slower but prolonged sequestration. Embedding these distinctions in national inventories and the IPCC carbon-cycle framework would improve the accuracy of cement-sector reporting and help design sector-specific strategies that align with real material and temporal dynamics. The intensity of uptake, however, is not uniform either across end-use applications or geography. We observe that uptake can vary fivefold across elements within the same building (Fig. 2A) and five-fold again within the same element (Fig. 2B) due to differences in materials and surface exposure conditions. These variations do not simply average out when results are aggregated. We observe a four-fold variation in the uptake per square area across the fifty states that make up the United States. These differences emerge because of regional differences in the distribution of building types, mix designs, and climate. Most readers will likely have noticed that the numbers reported here differ markedly from the range reported by recent national scale analyses, whose analyses of the fraction of process emissions sequestered by carbon uptake range from 3.8% (21, 22) to 55% (17) of process emissions sequestered. Here, we focus on the most widely cited of those references, Xi et al., who estimate global uptake to sequester 43% of the process emissions. For the United States, they estimate that the uptake in 2013 (the last year of analysis) is to be 22.2 Mt CO2 (reported as 6.04 Mt C emissions), equivalent to 57.4% of the process emissions in that year. The authors of this paper do not have direct access to the model used by Xi et al. Nevertheless, we have attempted to replicate their analysis and presented the information explicitly in SI Appendix, section S12. The implications of context and the method applied for calculating carbon uptake are made clear by the significant differences in other studies and our preliminary analysis of uptake in Mexico. Specifically, model results suggest that the absolute magnitude of uptake in Mexico approaches that of the United States although cement use in Mexico is about half that of the United States. This occurs because the Mexican market makes more intensive use of mortar, concrete masonry units, and bagged cement mixed on-site. Model results suggest that uptake from existing CBPs sequesters about one-quarter of each year’s process-related emissions from cement production in Mexico. Although this analysis focuses on the United States and Mexico, the developed approach can be applied to other contexts. Projections of new construction and end-use demand can draw on integrated assessment models such as IMAGE and the Global Building Stock Model and archetype information can be agency sources [for example, TABULA for Europe (23) and emerging datasets for China (24) and India (25)]. When combined with country-level cement-production statistics, these data sources enable a global bottom–up analysis of carbon uptake. We further applied a dynamic climate-accounting method (26) to evaluate time-dependent CO₂ uptake for infrastructure. The dynamic cumulative uptake curves (SI Appendix, Fig. S17) show smaller absolute values than the static 100-y accounting due to the fate of CO₂ in the atmosphere. The variation across these contexts also points to another possibility—the design of cement and concrete applications to enhance carbon uptake. Care must be taken when conventional steel reinforcements are present, but for many applications, it would be possible to increase uptake without deleterious effect on the structure. Enhancement can occur by ensuring that surfaces are not covered with paint or tile unless necessary, not using a stronger mixture than is needed by the design, and finally maximizing surface area, such as through the use of ceilings (e.g., waffle ceilings), beams, or columns with intentional texture. In the end, the fight against climate change will require changes throughout the construction value chain. Enhancing carbon uptake in cement-based products is a valuable strategy to add to the arsenal. While carbon uptake of in-use and end-of-life CBPs provides a quantifiable neutralization mechanism, its overall contribution remains limited compared with the mitigation potential available at the point of manufacture. Achieving deep decarbonization of the cement value chain will require the combined deployment of clinker substitution, low-carbon fuels, process efficiency, and CO2 capture at production facilities (4, 5). Supplementary Material Appendix 01 (PDF)