The climate benefits from cement carbonation are being overestimated

The study addresses how much real climate benefit arises from cement carbonation when timing of emissions and removals is considered. Cement production is a major source of CO₂ due to kiln fuel combustion and calcination, with concrete responsible for over 7% of global anthropogenic CO₂ emissions and 2–3% of global energy use. While strategies such as alternative fuels and novel materials help, many decarbonization roadmaps also credit natural carbonation of hydrated cement as a counterbalance to production emissions. Traditional GWP methods sum emissions and removals as if they occur simultaneously, potentially overstating carbonation’s benefit because emissions occur as a pulse during production, whereas uptake is delayed over decades. Prior assessments suggested large historical and future CO₂ uptake via carbonation, but often ignored end-of-life demolition emissions and the timing of fluxes. The authors’ objective is to model life-cycle cement emissions and CO₂ uptake globally and regionally, incorporate realistic exposure and material factors affecting carbonation, and apply time-adjusted warming potentials (TAWP) to quantify the true climate effect via cumulative radiative forcing. They also evaluate the sensitivity of decarbonization strategies (e.g., SCM use, demolition management) to identify effective pathways that minimize cumulative radiative forcing by 2050 and beyond.

Previous global assessments estimated substantial cement carbonation uptake, including studies indicating historic carbonation offsetting a large fraction of mineral-derived emissions and projections suggesting ~30% of cement production CO₂ could be reabsorbed by 2100. These largely used conventional GWP accounting that ignores timing and often excluded end-of-life processing emissions. Dynamic warming potential studies at the project or regional level (e.g., Quebec buildings 2018–2050) indicated much smaller net benefits (3–10%) when timing is included. The carbonation rate in concrete depends on environmental conditions (relative humidity, CO₂ concentration, temperature), structural and geometric factors (surface-area-to-volume ratio, thickness, porosity, strength class), and mixture composition, especially SCM content, which generally increases carbonation rates while decreasing portlandite availability. Experimental data show wide carbonation rate constants (0.5–15 mm/yr^0.5) yielding 4.2–83.7 mm after 70 years. SCMs can improve durability and extend service life, but often increase carbonation rates; curing conditions can moderate this. Microstructural evolution during carbonation (e.g., CaCO₃ pore blocking in PC; potential pore coarsening in blended systems) further influences rates. The literature highlights uncertainties and sensitivities that must be integrated into realistic models, along with proper accounting for timing via TAWP or equivalent dynamic methods.

Scope and goal: A cradle-to-grave analysis of cement in concrete and mortar was performed globally and for the United States, quantifying CO₂ emissions and uptake across production, use, demolition, and secondary life, and evaluating time-dependent climate impacts using time-adjusted warming potentials (TAWP). Both per-kg cement and regional/global scales were modeled.

Modeling framework: Carbonation was modeled using a Fick’s law-based approach adapted from Xi and Cao, modified to explicitly handle three phases: useful life, demolition, and secondary use. The carbonation coefficient Ki is the product of modifiers for relative humidity exposure (Bec), atmospheric CO₂ concentration/location (BCO2), coatings (Bcc), and SCM effects (Bad). Carbonation depth in each phase is di = Ki√ti, with volumetric uptake derived from exposed surface area (via mass of concrete, cement content, and member thickness) and carbonation depth. For crushed concrete at EoL and secondary life, particles are assumed spherical, and cumulative carbonation depth over demolition and secondary phases is de = ke√ts + ka√ta, with fractional carbonation Fs computed relative to particle size distributions.

Life cycle inventory and assumptions: Cement production emissions include calcination (stoichiometric, assuming clinker 65% lime; cement modeled as 80% clinker, 5% gypsum, 15% interground mineral additives globally; US: 95% clinker, 5% gypsum) and energy/electricity emissions (kiln thermal energy, quarrying, grinding, cooling). Global and US energy mixes from GNR/IEA and PCA sources were used; starting in 2023, energy-derived emissions per kg cement were assumed to decline 1.4% annually to 2050 (aligned with projected coal decline). Emissions from demolition/crushing were modeled from literature averages (15 data points), scaled with decreasing particle size (e.g., ~3× energy for 5 mm vs 25 mm particles).

Cement content and applications: Average cement contents used: global concrete 302 kg/m³ (weighted ERMCO 2001–2018), US concrete 277 kg/m³, mortar 284 kg/m³. End-use categories with distinct thicknesses and service lives were applied: US uses eight categories with mean service lives ~45–90 years; global uses residential, non-residential, and civil engineering. Phase durations: global average demolition/exposure 0.4 years; secondary life 35 years (buried scenario). A typical building example used 64-year service life. Approximately 74% of cement goes to concrete and 26% to mortar.

Exposure and modifiers: Relative humidity exposure classes (outdoor exposed/sheltered, indoors, wet, buried) and coating effects were included; for baseline US/global modeling, coatings were assumed none due to data gaps, exposure assumed exposed during use/demolition and buried during secondary life. Location-based CO₂ concentrations (urban, rural, seaside, industrial, road, buried) provided BCO2 factors. Strength class influenced porosity and carbonation via the exposure coefficient derivation. Mortar applications (rendering/plastering, masonry—with rendering—repair/maintenance) used adjusted rates.

Supplementary cementitious materials (SCMs): Five SCMs and replacement levels up to 50 wt% were modeled (limestone, fly ash, silica fume, blast furnace slag, natural pozzolans—assumed similar to fly ash). SCM factors altered carbonation rates and reduced production-phase emissions by lowering clinker content. Meta-analysis data informed SCM carbonation effects; no adjustment was made for reduced portlandite from pozzolanic reactions in the uptake capacity model.

Time-dependent climate metrics: TAWP integrates radiative forcing impacts accounting for the year of emission or uptake. For each flux m at year y, TAWP = m ∫₀^{AT−y} RFCO2 dt, with analytical time horizons of 20, 30, 50, and 100 years compared to GWP100. Cumulative radiative forcing (CRF) over time was computed for production emissions and uptake in each phase. Global flows from 1930–2050 were modeled; CRF was projected to 2150 to account for 100-year impacts of 2050 fluxes. Sensitivity analyses examined environmental conditions, coatings, CO₂ concentration, strength (porosity), member thickness, surface-area-to-volume ratios, SCM content, demolition particle size and exposure duration.

Data and tools: Historical and projected cement production/consumption from USGS, ERMCO, GNR, IEA, and literature sources were used; model parameters and coefficients are in Source Data 1 and Supplementary Note 1. Code is available in Dryad (doi:10.5061/dryad.6hdr7sr7n).

• Timing halves perceived benefits: Traditional GWP accounting overestimates carbonation benefits because it ignores the decades-long uptake timeline. For a 1 kg cement case, traditional methods overestimate climate benefits by over 100% compared to TAWP-based cumulative radiative forcing.
• Historic/global uptake vs climate effect: From 1930–2015, about 13.8 Gt CO₂ was re-absorbed globally via carbonation, but the slow rate reduces the effective climate benefit by ~60%. For 1950–2050, modeled cumulative CO₂ uptake is about −46 Gt, or 28% of cement production emissions; when timing is included, the global warming benefit is 67% smaller.
• US case: From 1930–2015, US carbonation uptake amounted to 0.8 Mt CO₂ (17% of cumulative cement production emissions by the study’s accounting). Considering timing reduces the climate benefit by 53%.
• Per-kg life-cycle breakdown (typical urban building, 64-year service life): During use, −0.05 kg CO₂/kg cement is reabsorbed (~12% of calcination emissions). At EoL, crushing to 1–40 mm with ~3.5 months average exposure adds −0.12 kg CO₂/kg cement (~30% of calcination), but demolition energy emissions are ~+0.1 kg CO₂/kg cement, nearly negating demolition-phase uptake. About 91% of EoL concrete is buried; buried crushed concrete can absorb up to −0.18 kg CO₂/kg cement (48% of calcination) within 25 years.
• Environmental and design sensitivities: Under low CO₂, coated seaside exposure with slow service-life carbonation (0.35 mm/yr^0.5), total uptake over 100 years can be ~−0.15 kg CO₂/kg cement. Indoors at higher CO₂ with favorable RH and faster rate (12.8 mm/yr^0.5) can reach ~−0.34 kg/kg over 100 years. High-strength concrete (lower porosity) shows ~80% less use-phase uptake and 28% less total uptake over 100 years than low-strength concrete.
• Time-dependence of EoL strategies: Extending demolition exposure from 1 day to 3 months reduces GWP by 10% but TAWP by only 5%; 1 year reduces GWP by 19% and TAWP by 9% (per kg basis). Crushing to 1–10 mm yields net climate benefit only after ~6 months exposure; crushing to 1–5 mm does not yield benefits even after 1 year given current demolition emissions.
• SCMs dominate TAWP reductions via production-phase savings: Introducing SCMs at ≥25% replacement yields the largest TAWP reductions (24–53%), with 34–78% of total TAWP reduction attributable to reduced production emissions. Example TAWP100 reductions relative to Portland cement: 25% slag (36%), 25% fly ash (41%), 50% slag (59%), 50% fly ash (64%); smaller but meaningful reductions for limestone and silica fume mixes.
• Demolition emissions significant: Over 1950–2050, demolition emissions ~2.5 Gt CO₂, about 86% of the demolition-phase uptake over that period, indicating the importance of managing demolition processes and exposure conditions.

The findings show that evaluating cement carbonation with traditional GWP can substantially overstate its climate mitigation value because carbonation is slow and occurs long after production emissions. By using TAWP and CRF, the study quantifies how delayed uptake diminishes the effective benefit by roughly half or more, depending on context. This directly addresses the research question by clarifying that while carbonation is non-negligible, it cannot be relied upon to offset production emissions at the timescales relevant for meeting mid-century climate targets. End-of-life strategies that increase surface area and exposure can increase uptake but also incur additional emissions from crushing; net benefits are modest when timing is included, especially over 20–50 year horizons pertinent to net-zero by 2050. Consequently, decarbonization pathways should prioritize up-front emission reductions—e.g., clinker substitution with SCMs, energy decarbonization—because early reductions yield outsized CRF benefits. SCMs offer substantial TAWP reductions mainly through production-phase savings, with accelerated carbonation as a secondary effect. The results underscore the need to include demolition emissions and realistic exposure conditions in accounting, to manage EoL practices to maximize safe, low-emission carbonation opportunities, and to avoid double counting or over-crediting carbonation in carbon budgets and industry roadmaps.

This work integrates life-cycle modeling of cement emissions and carbonation with time-adjusted warming potentials to reveal that climate benefits from cement carbonation are systematically overestimated when timing is ignored. Globally, large apparent uptakes translate to much smaller CRF benefits; per-kg analyses show EoL uptake can be nearly negated by demolition emissions. The most impactful strategies to reduce cumulative radiative forcing are those cutting production emissions early, particularly via SCM use, complemented by careful EoL management to enable low-cost, realistic carbonation without excessive additional emissions. Future research should refine parameterization of carbonation kinetics (including portlandite availability changes due to pozzolanic reactions), assess evolving atmospheric CO₂ concentrations, improve data on coatings and exposure distributions, and evaluate regional variations in recycling rates and secondary uses. Identifying alternative SCM supplies consistent with broader decarbonization of source industries, quantifying co-impacts on durability and service life, and incorporating human health considerations of demolition will further inform effective, holistic cement decarbonization strategies.

• Timing and concentration assumptions: The model applies TAWP with fixed analytical horizons and does not simulate future atmospheric CO₂ concentration trajectories; results are derived from sensitivity analyses rather than dynamic CO₂ scenarios.
• SCM carbonation chemistry: The model does not adjust carbonation capacity for reduced portlandite due to pozzolanic reactions, potentially biasing uptake estimates in blended systems.
• Data constraints and averages: Average cement contents and thicknesses are used across applications due to limited data; coating prevalence is assumed zero in baseline models; exposure factors are regionally generalized (primarily Europe-derived coefficients) and adapted via weighting.
• Demolition modeling: Demolition emissions are based on literature averages and scaling for particle size; site-specific machinery, energy mixes, and logistics could vary widely. Crushed particles are assumed spherical.
• Phase durations and flows: Demolition exposure duration and secondary life (35 years, buried) are applied uniformly; many real-world cases differ (e.g., recycling rates and secondary uses vary by country). Additional emissions during secondary life processes are out of scope.
• Energy decarbonization pathway: A uniform 1.4%/year decline in energy-derived emissions to 2050 is assumed; different regional decarbonization trajectories would alter results.
• Regional modeling: US and global models rely on available historical/projection datasets and assumed splits between concrete and mortar; uncertainties in material flow allocations propagate to uptake estimates.