Multi-scale sustainability assessment of modern cement and concrete technologies: a focus on decarbonisation potential

Sustainable innovations in cementitious materials, used for concrete production is considered a crucial lever towards decarbonisation of the built environment. This is motivated by the high carbon footprint associated with conventional Portland cement production, and the enormous volumes of concrete used in construction of civil engineering structures. Despite the continued progress in developing novel ‘low-carbon’ cement and concrete materials, accurate determination of their true sustainability credentials remains a challenge. This issue is linked to the limitations of conducting a life cycle assessment (LCA) for the bottom-up evaluation of the decarbonisation potential of well-known Portland blended cementitious materials when used in civil engineering structures.
In response to this challenge, this PhD research focused on advancing the current methodological applications of LCA for whole-life carbon assessment (WLCA) of concrete structures incorporating blended Portland cement (BC) concrete. The four key areas addressed include the accounting of use phase emissions, determination of the impact allocations when handling multifunctionality of by-product supplementary cementitious materials (SCMs), sensitivities in the dynamics of recarbonation potential for CO2 uptake estimation, and the development of a novel circular economy framework for optimal end-of-life (EOL) management of concrete. Accordingly, this thesis contributes to the state-of-the-art by proposing a novel use-phase emissions accounting method implemented with a bespoke graphic user interface-based MATLAB application known as ‘CarbCrete’, a consequential life cycle inventory (LCI) modelling framework for handling multifunctionality of by-product SCMs, a simplified dynamic characterisation for CO2 uptake estimation, and ‘CirCrete’, a framework for optimal EOL management of concrete structures. These methods were validated through three case studies addressing distinct methodological challenges, highlighting the strengths and opportunities for their applicability in other studies.
Case study 1 explored the effect of concrete durability represented by exposure to carbonation (XC) and chloride (XD & XS) environments, as described in existing British standards (e.g. BS 8500), on cradle-to-gate (modules A1–A3) embodied carbon of 129 commercial concrete mix designs in the United Kingdom. A subset of three cement types (CEM I, CEM IIIA and CEM IIIB, as per European standards (EN 197)) were further examined for use-phase emissions (modules A1–A3, B1 and B3) under four XC exposure classes for two structural geometries (slab and other-type elements like beams/columns) in line with Eurocode 2. Results revealed that prescriptive minimum binder contents of concrete, as well as the composition of the binder type used (e.g. content and type of SCM used) for a given exposure class influence the calculated cradle-to-gate emissions. Thus, there is a trade-off to be accounted for when designing concrete to ensure their longevity with the maximum carbon savings possible. A clear example is the current practice of increasing blast furnace slag (GGBS) content to lower embodied carbon. Ultra-high GGBS content in concrete reduces carbonation resistance compared to a Portland cement concrete, potentially shortening service life of concrete structures. Furthermore, excluding use-phase, underestimates embodied carbon by up to 44% and 64% in slabs and other geometries, respectively. Truncating the system boundary to cradle-to-gate overstated carbon reduction benefits in GGBS-based concretes and misclassified them as ‘lower-carbon’ options compared to Portland cement concrete. Results further highlighted that prescriptive Eurocode 2 covers for carbonation exposures over 50 years may not ensure durability in BC concretes. Benchmarking inaccuracies were also observed when uncertainties in LCI and use-phase emissions were unquantified, reinforcing the pre-eminence of the proposed accounting methodology.
Case study 2 evaluated the sensitivity of LCI modelling framework and recarbonation potential dynamics on cradle-to-grave global warming potential (GWP) impact of a hypothetical office building. The analysis considered three sensitivities including three carbonation rate approaches (Lagerblad’s constants, experimental literature values and an empirical machine-learning model), two LCI modelling frameworks (attributional and consequential) and four CO2 uptake timings (static, original dynamic, simplified dynamic and proposed simplified dynamic model). These were analysed for three concrete technologies incorporating Portland (OPC), GGBS and a limestone calcined clay (LC3) cement. Results indicated that carbonation rate choice influenced cradle-to-grave GWP, this being more noticeable for OPC (20% variation), followed by GGBS (12%) and then LC3 (7%), arising from significant differences in the CO2 uptake estimated during use phase and EOL. The Lagerblad’s constants recommended in EN 16757 underestimated the CO2 uptake capacity of BC concrete underscoring the need for revision. Adopting consequential LCI modelling framework resolved inconsistencies arising from impact allocation methods in attributional modelling. In the short term (5 years), this approach reduced GWP, but emissions increased in the medium term (6–10 years) owing to importation of GGBS and LC3 materials, highlighting the influence of market dynamics and consumption trends. Furthermore, adopting a static LCA approach led to approximately twofold overestimation of CO₂ uptake. The proposed simplified dynamic model achieved higher accuracy (Mean absolute percentage error (MAPE) of 13% and 7–39% for use phase and EOL uptake) than the existing simplified dynamic model (MAPE of 34% and 84–175% for use phase and EOL uptake). These results emphasise the need for robust modelling of sensitivities in recarbonation potential and LCI frameworks to ensure accurate WLCA of BC concrete structures.
In case study 3, the same hypothetical office building was used to demonstrate the application of CirCrete. Eight scenarios involving two new build cases for OPC (S1) and LC3 concrete (S2); complete repair and reuse (S3), complete recycling (S4), complete landfill (S5) and three hybrid scenarios combining repair-reuse, recycling, and landfilling in 50:50 proportions (S3–4, S3–5, S4–5). CirCrete integrated carbon emissions, energy demand, resource use, and life cycle cost into a composite circularity indicator with equal weighting and included uncertainty analysis with 20% variation in LCI data. Among all scenarios, S3 was identified as optimal based on the composite indicator, achieving net-negative carbon and energy values. Furthermore, using a single metric like carbon emission for circularity evaluation results in sub-optimal solutions. A developed circularity taxonomy classified scenarios S1, S2 and S5 as linear, S3–5 and S4–5 as circular and S3, S4 and S3–4 as highly circular. Sensitivity analysis confirmed that the composite circularity indicator was largely insensitive to transport distance but moderately influenced by indicator weighting; however, these factors did not alter the identification of the optimal strategy, confirming the robustness of CirCrete.
Overall, this PhD research contributes useful practical and policy support tools for reliably decarbonising concrete for the built environment.