This study conducted a systematic review of literature related to the CEs of rural dwellings, with a focus on the calculation methods, results, and carbon reduction strategies across the holistic life-cycle stages. The workflow consists of four main stages: (1) literature identification, (2) screening, (3) eligibility assessment, and (4) study inclusion.

A systematic literature search was performed using the Web of Science Core Collection. This database was selected as the sole source because it indexes a comprehensive range of high-impact, peer-reviewed journals in the fields of environmental science and construction engineering, ensuring the academic quality and reliability of the included studies. This study defined the following search keywords: TS = ((“carbon emission*” OR “carbon footprint*” OR “greenhouse gas*”) AND (“rural*” OR “village*”) AND ("residential" OR "housing" OR "dwelling*" OR "detached*" OR "house" OR "home*") AND (“low-carbon” OR “low-energy” OR “passive”)).

The screening process was conducted in two stages by two independent reviewers to minimize bias. First, all retrieved records were imported into Zotero (a reference management software), and duplicates were removed. Second, the remaining records were screened based on their titles and abstracts. Finally, the full texts of the potentially eligible articles were assessed against the pre-defined eligibility criteria. Any discrepancies between the reviewers were resolved through discussion or consultation with a third reviewer.

Additionally, a snowballing technique was applied to the reference lists of the initially identified articles to retrieve any potentially relevant studies not captured by the database search. In the first stage, we retrieved a total of 3,618 articles.

To further align with the primary research objectives of this review, we screened the titles and abstracts of 3,618 articles, followed by a full-text review of articles. Screening was conducted based on the following criteria:

(1) Lack of specific case studies: Articles that were purely theoretical, conceptual reviews, or lacked empirical data from actual building cases were excluded.

(2) Unspecified building function: Studies where the building function was not explicitly defined as residential or housing were excluded to avoid mixing with commercial or public buildings.

(3) Incompatible building scale: Studies involving high-rise buildings (more than three floors) were excluded, as this review focuses specifically on low-rise rural dwellings.

(4) Absence of quantitative data: Articles that did not provide explicit numerical results for CE calculations were excluded.

After this process, 84 articles that were specifically related to CEs of low-rise rural dwellings were retained. Additionally, 6 more studies were screened using the snowball technique (retrieving new studies through mining the reference lists of the 84 studies). Ultimately, 90 studies were screened [Figure 1]. Subsequent analyses were based on the case data from these 90 studies.

The information extracted from the aforementioned 90 studies is divided into three parts. The first part covers the basic building information, which mainly includes the building's region, climate zone, height or number of stories, structural type, and whether it holds certifications such as green or low-carbon building status. The second part focuses on the LCCE of the building cases, detailing the CEs for each life cycle stage. The third part mainly collects the carbon reduction strategies and carbon reduction effectiveness. It should be noted that since not all studies encompassed a complete life cycle scope, we only extracted the LCCE results as calculated in the respective literature. Furthermore, as the climate zone was not specified in most studies, we inferred the corresponding climate zone based on the specific geographic location of each building case.

The core of this research involves the descriptive statistical analysis of fragmented data from the aforementioned 90 studies. After extracting CE data of each life cycle stage from the selected case studies, descriptive statistics (including median) were calculated to characterize the overall dataset. The median was selected as the primary measure of central tendency due to the non-normal distribution of life cycle assessment (LCA) data. This approach allowed us to present a robust overview of current practices based on our synthesized dataset, rather than relying on potentially skewed averages from individual source papers.

To facilitate a unified analysis and comparison, we standardized the units of CE results from the literature. Regarding embodied carbon emission (ECE), we calculated the total amount and divided it by the gross floor area, resulting in a unit of kgCO₂e/m². As for OCE, given that the assumed service lives vary across different studies, we calculated the total OCE over the life cycle and normalized them to an annual per-square-meter basis; therefore, the unit is expressed as kgCO₂e/m²/year. To maintain the authenticity of the original findings and avoid introducing bias through estimation, life cycle stages that were not reported or calculated in the primary studies were excluded from the corresponding analyses.

To address the variations in system boundaries across the collected studies, we adopted a stage-specific data extraction strategy. Rather than relying on aggregated total values, we disaggregated the CE data into distinct life cycle stages (e.g., product stage, construction, use, and end-of-life). We extracted available data for each specific module as reported in the literature. Consequently, all subsequent statistical analyses were performed separately for each life cycle stage. This approach ensures that comparisons are made between equivalent stages, thereby mitigating the impact of differing system boundaries among the source studies.

The 90 screened studies contain a total of 309 building cases. CE assessment results for 309 building instances were compiled. The geographical location, climate type, structure type, number of floors, floor area, and expected service life of the building cases are summarized in Supplementary Figures 1-6.

According to ISO 21930:2017, the holistic life cycle of a building is structured into five primary stages comprising 17 distinct modules. These include material production (A1-A3), construction (A4-A5), usage (B1-B7), end-of-life (C1-C4), and supplementary benefits (D). Notably, Module D is categorized as voluntary information falling outside the scope of the system boundary. In terms of CE categorization, CEs arising from Modules A1-A5, B1-B5, and C1-C4 are collectively termed ECE. Conversely, emissions from Modules B6-B7 are defined as OCE, with Module D serving as a factor related to both ECE and OCE.

Although ISO 21930 establishes the fundamental framework for defining life cycle stages and system boundaries in building assessments, its strict implementation is not consistently observed in practical applications. As shown in Table 1, the 90 reviewed studies were calculated for production (88, 97.8%), construction (61, 67.8%), use (48, 53.3%), and end-of-life stages (48, 53.3%). The number of studies where B6-B7 was calculated was 60 (66.7%). The number of cases where module D was calculated was only 19 (21.1%). Few studies calculated the specific modules in both the use and end-of-life stages for ECE, with the largest number being 27 (30.0%) for maintenance (B2). Sixty (66.7%) studies calculated the operational energy consumption (B6), while only eight (8.9%) calculated the operational water consumption (B7) for OCE.

ECE values are closely related to the buildings’ structural type^[105]. The reviewed studies provided a total of 255 ECE data sets. Among them, there were 66 of masonry, 83 of timber, 18 of steel, 63 cases of concrete structures, and 26 of composite structures. The median statistical results of the ECE are shown in Figure 4. The median ECE for concrete, masonry, timber, steel, and composite structures were 362.50, 359.92, 172.21, 244.86, and 843.00 kgCO₂e/m², respectively, with timber structures having the lowest value. In addition, three bio-based structural buildings from Europe had a median ECE of 178.31 kgCO₂e/m², which was similar to that of wood structures [Figure 4]. It is important to note the varying sample sizes across different structural typologies. While masonry (n = 66) and timber (n = 83) datasets provide a statistically robust basis for analysis, the results for composite structures (n = 26), steel (n = 18), and particularly bio-based materials (n = 3) should be interpreted with caution. The statistical results for these smaller cohorts may have some limitations and may be more susceptible to outliers.

Due to inconsistencies in basic parameters such as CEFs of energy and building materials, and architectural design codes among different countries and regions, there are also certain variations in building ECE between them. Overall, the median ECE values for the China group and the Europe were 377.42 and 352.20 kgCO₂e/m², respectively, which are higher than those of other regions [Figure 5].

Building OCE includes operational energy and water consumption. However, most studies only considered energy consumption, and only Huang et al.^[26], Davis et al.^[65], Cai et al.^[22], Passer et al.^[44], Petrovic et al.^[67], Norouzi et al.^[77], Cuéllar-Franca et al.^[82], and Quintana-Gallardo et al.^[93] explicitly calculated water consumption. The CEs from stage B7 account for a relatively small proportion. For example: Petrovic et al.^[67] indicated that Stage B7 accounts for 6% of the OCE over a 100-year building life cycle. Meanwhile, calculations by Cai et al.^[22] showed that stage B7 CEs constitute only 0.3% of the building's total LCCE. As shown in Supplementary Table 1, the methods and scopes used for calculating energy consumption in stages B6-B7 varied across different studies.

The reviewed case studies provided 178 sets of CE data from stages B6-B7 [Figure 10], with a median value of 14.8 kgCO₂e/m²/yr, accounting for 66.1% of the LCCE. The median OCE values for masonry, concrete, steel, timber, and composite structures are 20.6, 19.2, 24.6, 4.4, and 10.7 kgCO₂e/m²/yr, respectively. Among these, timber structures have the lowest CE_B6-B7. OCE is closely related to the climate zone in which the building is located^[105]. According to the Köppen climate classification method, the world’s climates can be divided into five major zones: A, B, C, D, and E. A total of 140 sets of data were collected, including 9 from the B climate zone, 96 from the C climate zone, and 37 from the D climate zone (data for the A and E climate zones are lacking). Statistically, the median OCE value for the B climate zone is 1,208.5 kgCO₂e/m², for the C climate zone it is 902.6 kgCO₂e/m², and for the D climate zone is the lowest, at only 254.5 kgCO₂e/m². Additionally, there are certain differences in OCE among different countries or regions. A comparison was performed of the OCE in China, Asia (excluding China), and Europe. The median OCE value in the China group was the highest, at 910.0 kgCO₂e/m², followed by the Asia group with a median of 881.3 kgCO₂e/m², while the Europe group had the lowest median, at 792 kgCO₂e/m².

Electricity is the main component of the energy consumed during building operations. The electricity CEF changes over time, unlike primary energies. However, most studies tend to overlook the trend of the electricity CEF. The electricity CEF continues to annually reduce in many countries; therefore, the calculation method of annual power consumption × calculation period (such as 50 years) could overestimate the CE due to electricity consumption. Another problem arising from this oversight was that the CE offset was overestimated by the use of renewable energy (such as photovoltaics) in some cases. Replacing grid power with renewable energy decreased the CE offset, as the grid power CEF decreased.

The carbon benefit of module D is generated from the recycling and reuse of building materials, components, and energy. Among the 90 studies reviewed, a total of 19 (21.1%) considered this module, of which 16 studies conducted calculations, while the remaining 3 studies used empirical data. Due to the lack of measured data, the recycling or reusing rates of materials or components are usually assumed. For example, Moghayedi et al.^[95] assumed that 20% of bricks are recycled after the building’s service life ends. In another study^[96], researchers assumed a recycling rate of 30% for wood and tiles, and 50% for steel reinforcement in conventional buildings; whereas for prefabricated buildings, the recycling rate of floor slabs was assumed to be 80%.

The 90 reviewed studies collectively provided 25 data sets for module D [Figure 11]. Overall, the median carbon benefit of module D was -118.0 kgCO₂e/m², accounting for -31.5% of the total ECE and -8.7% of the LCCE. The median CE_D values for masonry, concrete, steel, and timber structures are -109.3, -73.3, -175.3, and -148.1 kgCO₂e/m², respectively. Among these, steel structures have the lowest CE_D. However, it should be noted that, due to the limited amount of available data, the statistical results regarding the carbon benefits of module D for buildings with different structures may have certain limitations.

Most of the pre-use stages were calculated through the design and construction plan in these case studies, or data recorded in the actual construction process, which tended to be in line with the actual conditions. However, many previous cases cited data or empirical formulas in the calculation of the in-use and post-use stages owing to the lack of independent research. Therefore, the calculation processes were full of assumptions and lacked credibility. Blengini et al.^[54,55] believed that CEs depend on a complex combination of local building technology, site-specific climatic conditions, local regulations, energy mix, and human behavior patterns in different cases; a single improvement based on oversimplification, generalization, and reliance on friendly tools might be ineffective or even deviate from expectations from a lifecycle perspective. A LCA of rural dwellings in Portugal showed that the use of different assessment tools produces inconsistent comparative results on the same issue, ultimately influencing decision-making^[60]. The resulting CEs for building maintenance varied in the range of 2%-6% of the LCCE in a case study in Spain when different parameters (provided by GaBi and Ecoinvent) were used for calculation^[101].

Regulations for load-bearing structure design are well established for small-scale, low-rise buildings, such as the dwellings discussed in this study. The potential for reducing the amount of building materials is limited by their structural safety. In contrast, the substitution of structural systems with low-carbon building materials has a high potential for CE reduction^[110]. The ECE reduction of the optimization schemes was 11.5%-96.5% among the reviewed case studies [Table 3], compared to the corresponding baseline scheme. The ECE reduction was influenced by the difference in CEF and the amount of substitution of building materials between the optimized and baseline structure schemes.

• Almost all studies involving timber (T) structures have low carbon advantages. The CE from the production stage of wood-frame houses was reduced by 64.5% compared to a RC scheme^[18]. The ECE of wood-frame rural dwellings was 37.5% lower than that of light-steel-frame houses^[28]. A study of a wooden house in Sweden showed a lower ECE for a wood structure than that for a concrete structure^[69]. The use of a novel off-site panelized modular timber frame system resulted in a 34% reduction in the ECE compared with conventional construction methods^[84]. Hafner et al.^[92] demonstrated that the use of wood instead of mineral materials for construction can reduce the ECE.

• Divergent findings were observed in the comparative analysis of concrete and steel structures. Analysis of three structural types for a three-storey residence revealed that concrete frameworks generated 44% and 49% higher CEs during the production phase compared to steel and timber alternatives^[17]. When applied to a rural residence in southern Italy, the CEs associated with the cold-formed steel frame and insulated panel wall system were found to be inferior to those of the traditional RC frame with brick walls throughout the pre-use, operational, and end-of-life stages^[56]. According to a statistical assessment of 45 residential units in Iran, concrete structures demonstrated an average ECE reduction of 24.46% compared to their steel counterparts^[32]. This disparity is even more pronounced in prefabricated prefinished volumetric construction, where concrete designs resulted in 54% fewer emissions than steel designs^[34].

• It is also crucial to investigate the capabilities of alternative low-carbon construction techniques. Research indicates that substituting conventional rural housing with straw-bale variations can drastically lower ECE; specifically, reductions of 96.5%, 75.5%, and 45% were observed for wood-frame, lightweight steel, and steel-frame straw-bale systems, respectively^[15]. In the context of Iran, a study covering four climatic regions demonstrated that straw-bale dwellings utilizing wooden structures outperformed those with concrete or interior brick walls, achieving ECE decreases of 76.00%, 16.89%, and 13.67% relative to traditional housing^[31]. Furthermore, composite materials show promise; for instance, steel-bamboo frames in East China yielded a 17.6% ECE advantage over concrete frames^[16]. Similarly, switching from an RC framework to a prefabricated concrete panel wall system in a Malaysian two-story residence resulted in a 26.27% reduction in emissions^[37].

• ECE reduction can be achieved by replacing some of the materials with high CEF used in the buildings. Research conducted on ten residential units in Malaysia indicates that the adoption of low-carbon components and materials can lower ECE by a range of 16.28% to 41.10%^[35]. Similarly, data from Sweden suggests an inverse relationship between the volume of bio-based materials used and the resulting ECE^[67]. In terms of specific material substitutions, one study on a new residence demonstrated a 24% decrease in ECE by utilizing concrete blended with 30% pulverized fly ash and swapping polyvinyl chloride (PVC) window frames for timber ones^[86]. Furthermore, optimizing wooden rural dwellings—specifically by substituting ordinary glass with tin-plated glass and aluminum frames with PVC—yielded a 15.25% reduction. This figure could be further improved by an additional 11.48% through the replacement of 4-ply asphalt roofing with PVC-cellulose alternatives^[97]. Izaola et al.^[111] found that using a wooden window frame to replace an aluminum window frame resulted in a 9.2% reduction in the ECE.

The implementation of building energy efficiency measures and the optimization of the energy structure to reduce the CEF of the energy supply have great potential. The OCE reduction benefits ranged from 9%-67% among the reviewed case studies [Table 4].

• Existing comprehensive technologies related to building energy conservation reduce OCE. A study of the heating scheme in a PH in northwestern France (including the electric heat pump, wood pellet condensing boiler, wood pellet micro-cogeneration unit, and district heating) showed that the PH showed clear advantages over the reference house during the operating stage, regardless of the heating system. The LCCE of the PH was 70.6% lower than that of the reference house following a comparison using the LCA tool, EQUER^[47,49]. A comparison of the OCE using four nZEB schemes and four PH schemes for a two-storey house in Ireland over a 60-year calculation period showed that buildings with high thermal and airtightness and heating systems with low CE (such as biomass boilers or heat pumps) should be prioritized^[52]. A functionally equivalent energy-efficient house reduces the LCCE by 63% compared to a standard home in Michigan^[98]. The study of a wooden house showed that the building ECE increased by 28.5% by improving the building envelope; however, the OCE decreased by 39.3%, which ultimately reduced the LCCE by 32.7%^[18]. A case study of dynamic thermal simulations showed that heavyweight constructions reduce the OCE by 7% compared with lightweight constructions^[87].

• Decarbonization of the power grid had a significant impact on the reduction of OCE. The OCE of operating a rural dwelling was reduced by 67% when using the Swiss mix compared to the Union for the Coordination of Transmission of Electricity (UCTE)^[98]. The lower CEF of Colombia’s grid resulted in a greatly lower building OCE in Colombia compared with that in Spain; an appropriate energy mix of electricity and natural gas reduces OCE by 25% and 9%, respectively, compared with pure electricity in the Mediterranean and Colombia^[112,113]. The OCE of a rural dwelling in Switzerland was only 13% that of the United States because the Swiss energy source mainly originated from hydropower and nuclear energy^[114]. A case study in Nepal showed that improving the proportion of hydroelectricity from 82.51% to 100% could reduce the LCCE by 20.5%-28.2%^[42].

• The low-carbon transformation of building heating systems was prominent in Europe. Comparison of different building envelopes, heating, and PV systems for a concrete-walled house in Finland showed that the heating system had the greatest impact on the LCCE, and the ground source heat pump was the optimal solution^[18]. A study of a rural dwelling in a cold Norwegian climate showed that an all-electric system, heat pump, high-efficiency insulation system, and PV panels achieve zero OCE; however, it was difficult to achieve net zero CEs (ZEB-OM) involving both OCE and ECE^[87]. Research focusing on timber-framed rural housing in Sweden indicates that OCE is significantly influenced by the specific energy supply infrastructure utilized; the OCE of DH + BIG/CC (district heat + biomass-based integrated gasification combined cycle) was approximately 90% lower than that of the RH + CST (electric resistance heating + coal-based steam turbine); the OCE of HP + NGCC (heat pump + natural gas-based combined cycle) was 50% lower than that of the RH + CST. The substitution of fossil fuels with biomass by-product from the wood production chain in a wooden house in Växjö showed that the total CEs during the building material production and building operation were negative when DH + BIG/CC or HP + BIG/CC were used^[70].

• Strategies showing regional differences. For example, different local power grids have different CEFs, and the benefits generated by saving the same amount of electricity are different. For example, the CEF of electricity production by PVs showed that the CEFs were 72.4 and 54.6 gCO₂e/kW, respectively, when the US fuel mix and European UCPTE fuel mix were used^[29]. A comparison of wood rural dwellings in five different locations in Europe showed that the OCE still accounted for 65%-76% of the total LCCE, even though the building envelope was well insulated. A cooler climate led to higher OCE owing to greater heating demand; however, differences in electricity and energy could sometimes mask the differences caused by climate. For example, the CEFs of electricity in Germany, Slovenia, and Spain were 0.406, 0.248, and 0.276 kgCO₂e/kWh, respectively, making the OCE in Munich (Germany) much higher than that in Ljubljana (Slovenia), although the energy consumption of the two cases was similar^[95].

The CE reduction benefits of module D among the reviewed cases ranged in 39.0%-100% (ECE) or 2%-3% (LCCE) together with other technology strategies [Table 5]. This requires technological support, such as waste recycling and scrapping. The study of a house with a bamboo structure showed that using building materials with renewable content reduces ECE by 18.5%, and the recycling of construction and demolition waste reduces ECE by 69.2%^[29]. A study of Palestinian houses showed that the ECE of traditional houses built with limestone and lime mortar was only 20.3% that of modern houses built with RC and concrete blocks^[43]. A study of three typical types of dwellings showed that recycling building materials at the end of their life reduces the LCCE by 3% for detached and semi-rural dwellings and by 2% for terraced houses^[92]. A case study in Belgium showed that the ECE of steel-frame houses was similar to that of masonry houses; however, when the recycling potential of steel is taken into account, the ECE associated with steel frameworks drops to 12% below that of masonry construction^[100,101].

The cases included in this review exhibit significant heterogeneity across multiple dimensions, which to some extent limits the direct comparability of results across studies. Specifically, this heterogeneity is primarily reflected in the following aspects.

In terms of contextual and technical parameters: The reviewed cases are located in different countries and climate zones. This variation directly affects building operational energy consumption and the CEF of electricity, thereby significantly impacting the comparability of OCE calculation results. For instance: for residential buildings with the same three-story masonry structure, the case investigated by Norouzi et al.^[77] is located in the UK (with an electricity CEF of 0.207 kgCO₂e/kWh), while the case studied by Huang et al.^[26] is in China (0.570 kgCO₂e/kWh). Despite the significant difference in electricity CEF between the two countries, and although the case in Norouzi et al.^[77] has a lower energy consumption than that in Huang et al.^[21], its OCE remain higher than the latter. On the other hand, differences in building types and structural systems fundamentally alter the consumption intensity of various materials, leading to different baselines for ECE. For instance, in this review, the median CE of steel-structured residential buildings during the A1-A3 stages is 305.6 kgCO₂/m², whereas that of timber-structured cases is only 113.0 kgCO₂/m².

In terms of methodology: First, the definition of system boundaries (e.g., whether Module D is included) directly determines the accumulation scope of CEs. Second, the lack of uniform functional units renders the horizontal comparison of absolute values meaningless; thus, units must be standardized before any comparison can be made. Finally, the selection of CEF databases (as previously discussed regarding regional differences and database sources) introduces systematic biases into the foundational data.