This review differs from previous syntheses in three respects. First, whereas earlier reviews of off-site construction have concentrated on cataloguing environmental performance indicators (Jin et al. 2020) or on a single lens such as circular economy application (Liu & Loo 2025), this paper deliberately synthesises evidence across all lifecycle stages, embodied, operational, and end-of-life, for the residential sector specifically, where the housing-climate policy interface is most acute. Second, the paper moves beyond descriptive synthesis by explaining variation: it interrogates why reported carbon reductions range from marginal to 45%, tracing this heterogeneity to material selection, grid carbon intensity, transport distance, prefabrication rate, and system boundary definitions. Third, and most significantly, the paper advances a novel conceptual contribution, the Carbon–Context Alignment Framework (CCAF). The specific research gap that the CCAF addresses is the absence of a structured model linking construction-method carbon performance to the national and market context in which that method is deployed; existing sustainability assessment tools evaluate buildings, whereas the CCAF evaluates the alignment between a prefabrication strategy and its deployment context. For policymakers and practitioners, the framework functions as a diagnostic checklist for configuring modular housing programmes—identifying which contextual conditions (material supply chains, grid decarbonisation, regulation, and housing demand aggregation) are in place and which represent vulnerabilities, as demonstrated through an illustrative application in Section 4.5.

This paper is guided by three interconnected research objectives:

The research topic was deconstructed into four conceptual groups to enable a structured and comprehensive search, following a PICO-equivalent decomposition commonly used in systematic review methodology. Table 1 presents the four search groups and their associated keywords.

The four groups were combined with Boolean AND operators so that all four dimensions were present in every retrieved document, while OR operators within each group captured terminological variation across disciplines and regions. Wildcards (*) captured plural and variant forms. The final search string applied to titles, abstracts, and keywords (TITLE-ABS-KEY) was:

TITLE-ABS-KEY ( ( “modular construction” OR “offsite construction” OR “off-site construction” OR “prefabricated construction” OR “prefabricated buildings” OR “industrialized building system*” OR “IBS” OR “modular housing” OR “prefab housing” ) AND ( “climate change mitigation” OR “carbon reduction” OR “carbon emission*” OR “greenhouse gas emission*” OR “GHG emission*” OR “decarbonization” OR “low carbon” OR “net zero” ) AND ( “life cycle assessment” OR “lifecycle assessment” OR “LCA” OR “life cycle analysis” OR “lifecycle analysis” OR “embodied carbon” OR “operational carbon” OR “whole life carbon” OR “carbon footprint” ) AND ( “residential building*” OR housing OR dwelling* OR “housing system*” OR “residential construction” ) ).

Beyond the language and subject-area filters applied within Scopus, explicit inclusion and exclusion criteria were defined a priori and applied at the screening and eligibility stages, as summarised below.

The search string was executed in the Scopus database on 8 November 2025 and re-run on 8 April 2026 to capture newly indexed studies. The initial retrieval yielded 59 documents. Because a single database was searched, no cross-database duplicates arose; Scopus's automatic deduplication confirmed zero duplicate records. Title, abstract, and keyword screening, followed by full-text eligibility assessment, was performed independently by the two authors, with disagreements resolved through discussion until consensus was reached. The application of the subject-area filter (Engineering; Environmental Science; Energy; Computer Science; Earth and Planetary Sciences) excluded two records, while the document-type filter (article, conference paper, review, book chapter) and English-language filter excluded no further records. The final dataset therefore comprised 57 documents. Table 2 summarises the filtering sequence, and Figure 1 presents the corresponding PRISMA flow diagram.

The 57 retained documents span publication years 2012 to 2026 and represent a geographically diverse set of research contexts, including the United Kingdom, mainland China, Hong Kong, Australia, South Korea, Malaysia, Portugal, Canada, and the United States. The dataset comprises 38 journal articles, 13 conference papers, 4 review articles, and 2 book chapters.

Because the dataset spans four document types with differing levels of peer-review rigour, a structured quality appraisal was undertaken rather than treating all sources as equivalent. Each study was appraised against four criteria adapted from established LCA reporting practice:

Journal articles and review papers generally satisfied all four criteria. Conference papers and book chapters, which typically undergo lighter peer review, were retained only where they satisfied at least the first three criteria; their findings are used in this synthesis to corroborate patterns established by journal-published evidence rather than as sole support for any conclusion, and this weighting is made explicit where relevant in Sections 3 and 4.

Two potential sources of bias in the evidence base are acknowledged. First, publication bias may inflate the apparent benefits of prefabrication, since studies reporting positive carbon outcomes are more likely to be published than null or negative results. Second, the geographic concentration of studies in China, Hong Kong, and the United Kingdom, contexts with strong policy mandates for prefabrication, means the evidence base over-represents settings already favourable to modular construction. Both issues are considered further in the limitations (Section 4.9) and inform the cautious, context-conditional framing adopted throughout this paper.

A thematic content analysis was employed. Each study was assessed against a structured extraction framework encompassing:

Extracted data were organised thematically to identify patterns, contradictions, and convergences. Where individual studies report specific carbon figures, those figures are quoted directly and contextualised within the wider synthesis. No attempt was made to statistically pool results into a single averaged estimate. This was a deliberate methodological choice: the included studies differ substantially in calculation methods, system boundaries, functional units, and assumptions about materials, transport, and building lifespan, and pooling across such heterogeneous conditions would risk producing a misleading average that represents no single context accurately. This approach is consistent with established practice in systematic reviews where study heterogeneity is high (Jin et al. 2020). The analysis instead prioritises understanding why findings differ across studies and what those differences reveal about the conditions under which prefabrication performs best from a carbon standpoint.

The 57 included studies reflect markedly growing interest in the intersection of prefabricated construction and lifecycle carbon assessment, with publication volume accelerating noticeably from 2019 onwards. Mainland China and Hong Kong account for a substantial share of the evidence base, reflecting both the scale of prefabrication adoption in those contexts and government-mandated carbon reduction policies in the construction sector. The United Kingdom emerges as the second most represented geography, consistent with national commitments under the Climate Change Act 2008 and the UK government's promotion of MMC as a response to both housing shortfall and decarbonisation imperatives (Narula and Finnegan 2025). Table 3 summarises, and Figure 2 visualises, the geographic distribution of included studies.

A notable finding concerns the unequal distribution of research attention across lifecycle stages. Figure 3 illustrates the lifecycle stages of a residential building, expressed using the EN 15978 modular structure, together with the five system boundaries adopted across the review dataset; Table 4 provides the corresponding breakdown. The majority of studies focus primarily or exclusively on the construction phase, particularly the embodied carbon of material manufacturing and factory or on-site assembly. Fewer studies integrate operational carbon in a whole-life perspective, and end-of-life assessment remains substantially under-represented.

Wen et al. (2024) conducted one of the most comprehensive end-of-life assessments in the dataset, finding that steel-framed modular buildings in Hong Kong hold significant carbon savings potential at disassembly, particularly where modules can be reused rather than recycled or landfilled; reuse scenarios recovered up to 30% of the initial embodied carbon investment. Liu and Loo (2025) similarly evaluated MiC lifecycle assessments within a circular economy framework and found that design-for-disassembly principles substantially improve whole-life carbon performance.

The literature reveals a broadly consistent, though not universal, pattern: prefabricated and modular systems tend to produce lower embodied carbon during the construction phase than conventional cast-in-situ or masonry alternatives. The magnitude of the reduction, however, varies substantially, from marginal improvements to savings approaching 45%, depending on material system, geographic context, prefabrication rate, and system boundary. Table 5 summarises the quantitative reductions reported across key studies.

Tian and Spatari (2022) found that prefabricated concrete structures in China produced 10–20% lower construction-phase emissions than cast-in-situ equivalents, though the advantage narrowed as transport distances extended. Du et al. (2019) corroborated this directional finding in a cradle-to-site analysis, noting that prefabrication rates above 50% were associated with lower CO₂ emissions, but that the relationship was non-linear and subject to diminishing returns above certain thresholds.

For timber-based systems, the evidence is particularly encouraging. Younis and Dodoo (2022) compared CLT, light-frame timber, and reinforced concrete residential buildings and found that CLT achieved substantially lower global warming potential (GWP) across the construction phase, with carbon storage in timber products providing an additional climate benefit not captured within many conventional LCA boundaries. Jeleniewicz et al. (2025) confirmed this in a European context, concluding that modular timber structures outperformed equivalent steel-framed alternatives on both embodied carbon and circularity metrics.

Greer and Horvath (2023) produced one of the more ambitious claims in the dataset, suggesting that modular construction in California's housing sector could reduce embodied carbon by up to 45% relative to site-built equivalents, a figure the authors themselves acknowledge depends heavily on assumptions about factory energy source, transport distance, and structural material. This highlights a recurring methodological challenge: extreme efficiency gains in controlled studies often rest on optimistic assumptions that may not translate uniformly to practice.

While embodied carbon has attracted the most consistent research attention, operational carbon, arising from the energy used to heat, cool, ventilate, and power residential buildings, remains the dominant carbon source over a 50–80 year lifespan in many contexts. The relationship between construction method and operational performance is, however, less direct than embodied carbon comparisons might suggest, because operational outcomes are mediated by factors largely independent of the construction method itself: occupant behaviour and household energy practices; the efficiency of heating, ventilation, and air-conditioning (HVAC) systems installed and their replacement cycles; prevailing climatic conditions and their evolution over the building's life; and the building's retrofit potential, since envelope and services upgrades over a multi-decade lifespan can reshape the operational carbon profile substantially.

Narula and Finnegan (2025) examined LGSF modular housing in the context of the UK social housing crisis, finding that operational carbon outcomes depended critically on insulation specification and on the carbon intensity of the national electricity grid. Their modelling demonstrated that under current UK grid conditions, LGSF housing could achieve net zero operational targets with appropriate thermal envelopes, but that this advantage would narrow in contexts with carbon-intensive generation. Tumminia et al. (2018) reinforced this through their assessment of a prefabricated net zero energy building module in Italy: high-performance envelopes combined with on-site renewable generation rendered the operational footprint negligible, but only where solar irradiation and grid decarbonisation aligned favourably.

Nangir et al. (2026) went further in their UK social housing analysis, evaluating whole-life carbon performance of LGSF modular construction incorporating bio-based insulation and geopolymer concrete alternatives. Material innovation within modular systems, particularly substituting high-embodied-carbon materials such as Portland cement concrete with lower-impact alternatives, offered compounding benefits across both embodied and operational carbon categories. Nonetheless, the operational evidence base remains thinner than modelling studies imply: few included studies incorporate measured post-occupancy data, meaning that performance gaps between designed and actual energy use, driven largely by occupant behaviour and HVAC operation, are rarely captured.

End-of-life carbon is the most consistently neglected lifecycle stage in the prefabricated housing literature, yet it represents both a significant risk and a substantial opportunity. Prefabricated systems, by virtue of their modular and often demountable architecture, are theoretically well positioned to support circular economy strategies including disassembly, module reuse, and material recovery.

Wen et al. (2024) produced the most detailed end-of-life carbon analysis in this review, demonstrating through a case study of steel-framed modular buildings in Hong Kong that end-of-life carbon savings can be considerable where modules are reused rather than demolished, with reuse significantly outperforming recycling. Liu and Loo (2025) similarly found that circular economy thinking applied to MiC in Hong Kong improved whole-life carbon performance, though they cautioned that circular outcomes depend on reverse logistics infrastructure that remains immature in most markets. Rajanayagam et al. (2024) examined flat-pack modular systems and identified design for disassembly as a critical variable, finding that systems specifically designed for disassembly could avoid up to 20% of the demolition-phase emissions typical of conventional buildings.

These theoretical benefits, however, face substantial practical barriers that the literature is only beginning to address. Module reuse in practice is constrained by the absence of established secondary markets for used modules; regulatory challenges, including building code recertification of reused structural elements, warranty and liability uncertainty, and the lack of standardised material documentation or ‘module passport’ systems; logistical challenges of storing, transporting, and refurbishing large volumetric units between uses; and unresolved questions of economic feasibility, since disassembly is more labour-intensive than demolition and its business case depends on residual module value that current markets do not reliably price. Material degradation over a first service life, including connection fatigue in steel systems, moisture-related deterioration in timber, and sealant and gasket ageing, further limits which modules will realistically be fit for reuse. Until these barriers are addressed through policy and industry standardisation, the end-of-life carbon savings modelled in the literature should be read as an upper bound on achievable performance rather than an expected outcome.

The 57 reviewed studies reveal a broad directional consensus: prefabricated and modular construction tends to reduce construction-phase embodied carbon relative to conventional methods. The literature is nonetheless marked by significant heterogeneity in findings, methodologies, and contextual assumptions that cautions against uncritical generalisation. Three dominant sources of variation deserve particular attention: material selection, system boundary definition, and geographic or climatic context. Each is examined below, followed in Section 4.2 by a closer interrogation of why reported carbon reductions vary so widely.

The first source of variation is material selection. Steel-framed modular systems produce markedly different carbon profiles from timber or concrete alternatives. Kechidi and Banks (2023) found that steel-framed modular housing in the UK generated meaningful upfront carbon savings relative to masonry construction, but noted that the high embodied carbon of structural steel makes whole-life comparisons sensitive to assumptions about end-of-life recycling rates and grid decarbonisation trajectories. Timber systems, by contrast, benefit from biogenic carbon sequestration, negative or near-zero embodied carbon for structural elements under some accounting frameworks, and strong alignment with circular economy principles (Premrov and Passer 2020; Jeleniewicz and Szlachetka 2025). This advantage must, however, be weighed against countervailing considerations that the timber-focused literature often understates. Fire safety requirements for mass-timber and light-frame systems, particularly in mid- and high-rise applications, can necessitate additional encapsulation materials and design compensations that erode part of the embodied carbon advantage. Moisture durability is a genuine risk for timber modules exposed during transport, storage, and assembly, with implications for both service life and warranty. Engineered timber supply chains remain geographically concentrated, and regional availability constraints, acute in most of Africa, the Middle East, and parts of Asia, mean that imported CLT can carry significant transport emissions and cost premiums that alter the carbon and economic calculus. A balanced reading of the evidence therefore positions timber as highly promising where regulatory, climatic, and supply chain conditions permit, rather than as a universal solution.

The second source of variation is system boundary definition. Studies adopting cradle-to-gate or cradle-to-site boundaries naturally report lower overall carbon figures than those employing whole-life or cradle-to-grave frameworks, making direct comparison hazardous without careful boundary harmonisation, a methodological challenge that Jin et al. (2020) identified as among the most significant barriers to confident knowledge synthesis in this field. Building Information Modelling (BIM)-integrated LCA approaches, explored by Hao et al. (2020) and Xu et al. (2021), offer promising pathways towards more consistent and granular carbon accounting, but their adoption remains uneven across geographies and industry sectors.

The third source of variation is geographic and climatic context. Studies from temperate European climates produce different operational carbon trajectories from tropical or cold-climate studies, given differential heating and cooling demands. Bukoski et al. (2017) demonstrated this in their assessment of an energy-positive modular residence in Thailand, where high solar irradiation enabled near-zero operational carbon despite relatively modest envelope performance. Shahi et al. (2020) similarly showed that modular approaches to residential tower adaptation in cold Canadian climates produced different cost–carbon trade-offs from equivalent projects in milder climates.

Why do reported embodied carbon reductions range from 5% to 45% across studies of nominally similar interventions? Four interacting mechanisms emerge from the synthesis.

Regional electricity mix. Factory-based production shifts energy demand from diesel-powered site plant to grid electricity. Where the grid is coal-dominant, this shift can erode or reverse the expected benefit; where it is renewable-dominant, factory production compounds the advantage. This mechanism explains much of the gap between, for example, optimistic Californian results (Greer and Horvath 2023) and more modest findings from coal-intensive grids. It also implies that the carbon case for prefabrication will strengthen mechanically as national grids decarbonise, since factory electricity demand is progressively cleaned at no additional cost to the builder.

Transport distance. Module transport is carbon-intensive per tonne-kilometre relative to bulk material delivery, because volumetric modules ship significant empty space. Tian and Spatari (2022) showed the prefabrication advantage narrowing as factory-to-site distances extended; beyond context-specific thresholds, transport emissions can offset factory-stage savings entirely. Reported reductions therefore depend on the (often unstated) assumed distance between factory and site.

Prefabrication rate. Du et al. (2019) demonstrated that carbon savings scale non-linearly with the proportion of the building manufactured off-site: gains accrue steeply up to roughly 50% prefabrication and then flatten, exhibiting diminishing returns. Studies comparing high-prefabrication-rate systems with conventional baselines will consequently report larger reductions than studies of hybrid systems, even within the same market.

System boundary and LCA assumptions. Finally, part of the reported variation is an artefact of measurement rather than of performance. Narrow cradle-to-gate boundaries concentrate attention on the stages where prefabrication performs best, inflating apparent reductions; whole-life boundaries dilute the same absolute saving across a much larger denominator. Choices of functional unit (per square metre versus per dwelling), building lifespan, and end-of-life allocation rules further shift results in ways that are rarely harmonised across studies. Sensitivity analyses, where reported (e.g., Teng & Pan, 2020), confirm that boundary and assumption choices can alter reduction estimates by margins comparable to the differences between construction methods themselves.

Closely related to the above, the review's quantitative findings must be read against the substantial uncertainty embedded in LCA methodology itself. Four sources of uncertainty warrant explicit acknowledgement. First, emission factor uncertainty: embodied carbon estimates depend on background databases whose factors for the same material can differ considerably across databases, regions, and vintages, propagating directly into reported reductions. Second, functional unit variability: studies normalise results per square metre of gross floor area, per dwelling, or per building, and these choices are not interconvertible without information that is often unreported, limiting cross-study comparability. Third, building lifespan assumptions: whole-life results are highly sensitive to the assumed service life (commonly anywhere between 50 and 80 years in the included studies); longer assumed lifespans amortise embodied carbon over more years of operation and shift the embodied–operational balance accordingly. Fourth, future grid decarbonisation scenarios: whole-life studies must assume a trajectory for grid carbon intensity decades into the future, and static-grid assumptions systematically overstate operational carbon relative to scenarios reflecting legislated decarbonisation commitments. Teng and Pan (2020) demonstrated formally that parameter, scenario, and model uncertainties of these kinds can materially change embodied carbon estimates for prefabricated high-rise buildings. The quantitative ranges reported in this review should therefore be interpreted as indicative envelopes rather than precise expectations, and future primary studies should routinely report sensitivity analyses across these four dimensions.

The findings of this synthesis both confirm and extend earlier review work. Jin et al. (2020), in their critical review of the environmental performance of off-site constructed facilities, concluded that off-site methods generally outperform conventional construction on waste and emissions but that methodological inconsistency prevents firm quantification; the present review confirms both conclusions on a more recent and residential-specific evidence base, and adds quantitative structure by mapping the reported reduction ranges (10–45%) to their explanatory conditions (Section 4.2). Liu and Loo (2025), reviewing MiC lifecycle assessments through a circular economy lens, emphasised end-of-life potential and reverse logistics immaturity; this review corroborates their conclusions while situating end-of-life within a whole-life and multi-context framing. Two points of tension with earlier syntheses deserve note. First, whereas some earlier review work implicitly treats the carbon benefit of prefabrication as a property of the method, the evidence assembled here indicates the benefit is conditional on context—the central claim formalised in the CCAF. Second, this review is more cautious about headline reduction figures than some prior syntheses, given the boundary and uncertainty artefacts documented in Sections 4.2 and 4.3. The added contribution of this synthesis is therefore threefold: a residential-sector focus across all lifecycle stages, an explanatory (rather than solely descriptive) treatment of heterogeneity, and a transferable conceptual framework linking method to context.

The central theoretical contribution of this paper is the Carbon–Context Alignment Framework (CCAF), developed inductively through the systematic synthesis of the 57 included studies. The CCAF proposes that the climate efficacy of prefabricated housing cannot be determined by construction method alone, but is the product of four simultaneously operating contextual dimensions:

Figure 4 presents the framework schematically, and Table 6 defines each dimension with its supporting evidence.

The CCAF proposes that maximum climate benefit from prefabrication is achievable only when all four dimensions are favourably aligned, and that alignment is effectively multiplicative rather than additive: weak alignment in any single dimension constrains the benefit obtainable from strength in the others. A country with a decarbonised grid but no whole-life carbon policy will not realise prefabrication's full potential; a nation with progressive regulation but a sector dominated by energy-intensive steel modules produced in coal-powered factories will find prefabrication delivering disappointing results in practice.

The CCAF's Social Housing Scale dimension raises a question of particular consequence for developing economies, where the majority of projected housing growth to 2060 will occur (Tang et al. 2024): can modular construction scale under rapidly urbanising, resource-constrained conditions? Four constraints emerge as critical. First, supply chain readiness: prefabrication presupposes reliable upstream supply of processed materials, components, and logistics services that remain underdeveloped in many low-income countries, where cement, steel, and engineered timber are frequently imported at volatile prices. Second, manufacturing infrastructure: volumetric modular production requires significant capital investment in factories, moulds, and handling equipment, together with dependable electricity supply, itself scarce in many rapidly urbanising regions, creating a chicken-and-egg problem in which factories await demand certainty while programmes await factory capacity. Third, workforce transition: industrialised construction shifts labour demand from abundant low-skilled site labour towards factory technicians, logistics operators, and precision assemblers, requiring training systems that take years to establish and raising distributional questions in labour markets where construction is a major employer of low-income workers. Fourth, affordability: the cost premium of early-stage industrialised construction, before economies of scale are achieved, can conflict directly with the affordability imperative of social housing programmes. These constraints do not negate the CCAF's logic; rather, they specify its sequencing implication, that in low-income, high-growth contexts, investments in grid capacity, material supply chains, workforce development, and demand aggregation are preconditions for, not consequences of, climate-effective modular housing at scale.

A further consideration, largely absent from the included studies, is the possibility of rebound effects accompanying large-scale prefabrication adoption. Modular construction's principal industrial attractions—speed, cost predictability, and a perception of environmental superiority, could plausibly stimulate higher overall construction volumes than would otherwise occur, partially or wholly offsetting per-unit carbon savings through aggregate material consumption. Analogous mechanisms are well documented in energy economics, where efficiency improvements induce additional demand. In housing, faster delivery may also encourage larger dwelling sizes, shorter redevelopment cycles, or speculative overbuilding, each carrying embodied carbon consequences. None of the 57 included studies quantifies such effects, which represents a substantive gap: lifecycle assessments compare buildings one-for-one and are structurally blind to volume effects at market scale. Future research combining LCA with construction-market modelling is needed to establish whether, and under what policy conditions (for example, whole-life carbon budgets applied at portfolio or municipal scale), prefabrication's per-unit savings survive aggregation. Acknowledging this possibility does not weaken the case for modular construction, but it does argue for embedding prefabrication within absolute carbon governance frameworks rather than treating per-unit efficiency as sufficient.

Read through the lens of the CCAF, the synthesised evidence yields three implications for policy and professional practice. First, whole-life carbon accounting must become standard in building regulation and procurement. The current dominance of cradle-to-gate and cradle-to-site LCA systematically undervalues operational carbon and fails to incentivise end-of-life planning. Mandatory whole-life carbon assessment, as being progressively introduced in England through forthcoming Future Homes Standard regulations, is a step in the right direction; comparable mandates are urgently needed in the rapidly urbanising markets of the Global South.

Second, material innovation within prefabricated systems deserves greater policy and investment attention. Several included studies demonstrate that substituting conventional structural materials with lower-carbon alternatives, engineered timber, bio-based insulation, geopolymer concrete, can dramatically improve the carbon performance of otherwise conventionally configured modular systems (Nangir et al. 2026; Premrov and Passer 2020). Procurement frameworks that incentivise or require such substitution within modular supply chains could unlock substantial carbon co-benefits.

Third, circular economy thinking must be embedded in modular housing design from the earliest project stages. The evidence from Wen et al. (2024), Liu and Loo (2025), and Rajanayagam et al. (2024) consistently identifies end-of-life module reuse as the highest-value carbon recovery strategy in prefabricated construction, yet reuse requires design decisions, connection detailing, material documentation, and reverse logistics infrastructure to be in place long before a building reaches end of life. Policy mechanisms that reward design for disassembly and establish module passport systems, analogous to product passports under the EU Circular Economy Action Plan, could catalyse a significant shift in practice.

Several limitations should be acknowledged. First, the review relied on a single database (Scopus) and was restricted to English-language publications, which may have excluded relevant research published in Chinese, Korean, Portuguese, or other languages and in sources indexed elsewhere; this plausibly contributes to the geographic imbalance of the evidence base, which over-represents China, Hong Kong, and the United Kingdom while under-representing Africa, Latin America, and much of South Asia, precisely the regions where future housing growth is concentrated. Second, the heterogeneity of LCA methodologies, system boundaries, and functional units across included studies prevented quantitative meta-analysis; the synthesis is therefore narrative and thematic, and the reported reduction ranges should not be read as pooled estimates. Third, the inclusion of conference papers and book chapters, while broadening coverage, introduces variability in peer-review rigour that the quality appraisal in Section 2.4 mitigates but cannot eliminate. Fourth, publication bias towards positive findings may inflate the apparent benefits of prefabrication. Finally, studies indexed after the updated search date of 8 April 2026 are not captured. Future research employing harmonised LCA protocols, as recommended by Jin et al. (2020), together with multi-database and multi-language searching, would substantially strengthen the evidence base.