1. INTRODUCTION
The provision of secure and sustainable housing is fundamental to the United Nations (2015) Sustainable Development Goals (SDGs), yet relentless urban expansion to meet housing needs can be in direct conflict with climate goals (OECD 2018). With 60% of the world’s population projected to live in cities by 2030 (UN-Habitat 2020), there is an urgent need to find strategies for the provision of low-carbon housing.
A severe housing crisis in Ireland has given rise to government targets for the production of 50,000 new housing units per year (Government of Ireland 2021), which are also to be achieved in the context of significant carbon emissions reductions ranging between 44% and 56% across the construction sector by 2030 as part of the national Climate Action Plan (Government of Ireland 2025). In this context, it is clear that meeting both these targets is infeasible under business-as-usual scenarios where new housing demand is largely satisfied by low-rise construction on greenfield sites in rural clusters or on the edges of towns or cities. As in several European countries, planning policy attempts to limit the urban sprawl that this type of development can generate by promoting compact urban growth. By considering the upfront embodied carbon (EC) emissions of individual buildings in an urban planning context, the research suggests how the sustainability claims for such higher densities might be assessed.
To date, national regulations in the European Union (EU) focus mainly on the operational emissions of new construction, regulated through Energy Performance Certificates (EPCs) that provide indicators of greenhouse gas (GHG) emissions and energy performance. The requirement to assess EC emissions has recently been introduced in the revised Energy Performance in Buildings Directive (EPBD), which will require a revised EPC for all new buildings from 2030 that includes whole life-cycle (WLC) emissions. This situation is mirrored in the academic literature, where the focus to date has largely been on the operational phases of WLC emissions, with research into EC being an emerging field where there can be a wide range of results due to differences in scope or the methodology used across studies (Röck et al. 2022; Simonen et al. 2017).
In addition, the existing literature that includes life-cycle stages of EC to date has largely focused on individual buildings, with elements of siteworks at neighbourhood or urban scales either partially included or omitted entirely. Academic studies that assess EC at larger geographical scales have also been shown to produce varying results due to differences in scope, choice of functional unit (FU) and methodology (Lotteau et al. 2015). While holistic assessments of an urban development should include all elements within the site boundary, achieving this creates practical challenges around data-gathering and the estimation of material quantities. For residential developments containing multiple buildings, the siteworks of new construction can be significant, comprising site preparation and earthworks, hard and soft landscaping, and infrastructural utilities that serve the dwellings.
This paper provides a framework for quantifying the EC associated with new residential developments at an urban scale by analysing dwellings and siteworks separately at different scales. In this study siteworks are broken down into the component categories of external landscaping and service infrastructures, although different terminologies for these elements can be found in the literature. The proposed framework carries out separate EC assessments for different building typologies on a site, together with their private and public external areas and associated service infrastructures, allowing the results to be highlighted at a range of scales. The framework and reporting system provide a basis for comparison with other studies that use a different scope, such as those that consider buildings only, or those that include external landscaped areas but not underground service infrastructures (Heinonen & Junnila 2011; Kayaçetin & Tanyer 2020; Sigurðardóttir et al. 2023). By disaggregating the results by element, the relative EC impacts of factors such as building form, height and density can also be compared.
The framework is tested using a case study housing development of 184 dwellings in Dublin, Ireland. The FU of study is the scale of a small district formed of several urban blocks—a scale sufficiently detailed to allow for bottom-up assessment of building case studies, and large enough for a holistic understanding of their impacts at an urban scale. The framework therefore provides an empirical basis for assessing policies that seek to promote higher densities and compact urban growth. This paper considers upfront EC life-cycle stages A1–A5, although future research could be expanded to include WLC.
The three principal research questions are as follows:
How can a holistic framework be developed for carrying out life-cycle assessments (LCAs) that include elements outside the building envelope, such as external landscaping, and infrastructural elements, including buried services?
What is the relationship between the housing typology, density and related urban form, and the EC impact of new development?
How do EC emissions change depending on the scale of analysis, and how can the systems boundaries of LCAs be defined in order to provide a holistic assessment of the climate impacts of housing development?
2. LITERATURE REVIEW
There is a growing understanding that the choice of spatial settlement patterns impacts GHG emissions. In particular, the high carbon implications of low-density urban sprawl due to the operational energy requirements of buildings has been considered, together with other factors such as car dependency and the difficulties of providing infrastructure and services across larger distances (Baldasano et al. 1999; Carruthers & Ulfarsson 2003; Glaeser & Kahn 2010; Jones & Kammen 2014; OECD 2018).
At the same time there is an increasing recognition of the role of EC in the life-cycle impacts of construction. Analysing the Irish residential sector, Hegarty & Kinnane (2023) note that national climate change mitigation policy to date has focused almost entirely on operational carbon (OC). Presenting a detailed methodology for modelled forecasts, they demonstrate that proposed reductions in OC due to retrofitting existing homes and the decarbonisation of the electricity sector will be more than offset by increases in the embodied emissions required to meet government housing targets. Several studies have also noted that as OC emissions reduce through the development of more energy-efficient buildings and decarbonisation of the grid, the proportion of EC in the overall life-cycle energy becomes greater, highlighting its importance as an object of study (Koezjakov et al. 2018; Mirabella et al. 2018; Ruuska & Häkkinen 2015).
At the scale of individual buildings, Röck et al. (2022) note the difficulty of harmonising results across LCA studies due to the diversity of building types, and a general lack of transparency regarding the scope of study, life-cycle stages considered and methodology. Note that the umbrella term ‘LCA’ is used here due to the range of life-cycle stages covered by different studies in the literature. This lack of harmonisation of scope and methodology is also found in the literature that considers siteworks elements outside the building envelope. At these larger scales a framework is required for geographical system boundaries that allows for the elemental breakdown and allocation of carbon impacts. The academic literature on system boundaries mostly considers energy and material inputs, and how far the scope should extend within up- and downstream processes; however, most of these remain at the building level and do not consider larger scales (Kotaji et al. 2003; Ries & Mahdavi 2001). Dixit et al. (2013) consider the problem of differing system boundary definitions across LCA studies, and define a framework for geographical study boundaries using a series of nested categories: whole building, building and site, building and neighbourhood, and finally building and city. Pan et al. (2018) propose an even more comprehensive framework of 12 categories for system boundary definition, with the increasing geographical boundaries following a similar scalar system. The nested categories of scale proposed by these studies could pose practical difficulties, as they do not account for urban forms with different morphological characteristics, where elements may cross categories of scale. For example, many apartments might share a common courtyard, but individual houses have their own private gardens. The definition of ‘site’ given in these studies is also problematic, as real-world examples show that the definition and extent of the site associated with a building can vary significantly with layout and typology.
More precise definitions are therefore required, so this paper draws form the discipline of urban morphology which provides a hierarchical structuring of elements that links different scales, building typologies and forms (Kropf 2017). Regarding the question of geographical boundaries, Berghauser Pont & Haupt (2023) show how density varies with scale, as moving to successively larger scales adds additional external ‘tare’ spaces, such as gardens, streets and parks. This has clear implications for LCA research, as the scope and resultant carbon emissions will change as additional spaces are added. Urban morphology defines repeating patterns and elements of urban composition, and this makes it an ideal structuring system for LCAs. The principal elements of urban form according to this system are the street, the plot and the building (Conzen 1960). These are expressed in a hierarchy with each element of the system spatially containing the next. Kropf (2014) expands on these basic categories to propose an elemental framework called the ‘multi-level diagram’ which adds to the basic elements to allow for the definition of buildings and neighbourhoods at a range of scales. The multilevel diagram therefore provides a useful framework for reporting carbon emissions which can link buildings with external landscapes and infrastructures. Two examples of this diagram based on case studies are shown in Section 3 below. Precise methods for establishing the physical scope of an LCA are therefore important for further research at larger scales that can support policies promoting density and compact growth.
The literature on LCAs at scales, including elements of siteworks, is summarised in Table 1, which lists the various scopes and methodologies used. Table 1 shows that the majority of these studies rely on modelled urban layouts and archetypal building typologies, rather than on real-world case studies. Some studies consider large urban scales and have a wider scope including factors such as transport emissions (Du et al. 2015; Nichols & Kockelman 2014; Stephan et al. 2013). Other studies take a modular approach, modelling synthetic environments based on real-world data that can be scaled to estimate larger areas (Hack et al. 2025; Pomponi et al. 2021; Trigaux et al. 2014). Pomponi et al. (2021) adopt an approach of high-level parametric modelling to assess the impact of density and urban form on whole-life carbon including embodied life-cycle stages. They discovered that while taller urban environments increase overall GHG emissions, building similarly dense low-rise neighbourhoods can provide savings. In addition to modelling buildings, the study includes streets or residual green spaces but not service infrastructures. Schiller (2007) directly considers the relationship between the environmental impacts of buildings and their infrastructures; however, this study uses metrics of material resource efficiency rather than construction-related GHG emissions. The study quantifies material-use relationships between dwellings and infrastructure for different housing typologies, finding that the use of materials rose exponentially with reduced residential density.
Table 1
Summary of the parameters used in the literature, which includes embodied carbon (EC) within the life-cycle scope and elements of siteworks within the physical scope.
| REFERENCE | FUNCTIONAL UNIT/PHYSICAL SCOPE | BUILDINGS/TYPOLOGIES | LANDSCAPING/EXTERNAL AREAS INCLUDED | SERVICE INFRASTRUCTURES INCLUDED | LIFE-CYCLE STAGES | METHOD | LIFE SPAN (years) |
|---|---|---|---|---|---|---|---|
| Du et al. (2015) | High-rise urban area and low-rise suburban area | Prototypical constructions of typologies | Roadways and parking only | No | Cradle to grave | Hybrid LCA | 50 |
| Butters et al. (2024) | Two neighbourhood case studies | Modelled typologies | Yes | No | Cradle to grave | Hybrid LCA | 50 |
| Hack et al. (2025) | Synthetic urban models | Modelled typologies | Roadways only | Modelled estimates | Cradle to grave | Process LCA | 50 |
| Kayaçetin & Tanyer (2020) | Three neighbourhood-scale case studies | Three residential case studies | Yes | No | Cradle to grave | Hybrid LCA | 50 |
| Nichols & Kockelman (2014) | Five neighbourhood-scale case studies | Modelled typologies | Yes | Yes, estimated based on existing research | Cradle to grave | Hybrid LCA | Per annum |
| Pomponi et al. (2021) | 1 km2 synthetic urban models | Modelled typologies | Yes | No | Cradle to grave | Process LCA | 60 |
| Rankin & Saxe (2024) | Existing neighbourhoods | Existing buildings and infrastructure | Not defined | Estimated using geographical information system (GIS) tools | Cradle to gate (A1–A3) | Process LCA | 60 (baseline case) |
| Troy et al. (2003) | Six neighbourhood-scale case studies | Modelled typologies based on age | Roadways only | Foul drainage and water supply only | Cradle to grave | Input–output LCA | 60 |
| Sigurðardóttir et al. (2023) | Neighbourhood-scale case study | Apartments and ancillary buildings | Yes | No | Cradle to gate | Process LCA | Not stated |
| Sjökvist et al. (2025) | Neighbourhood-scale case study | Archetype buildings | Yes | Yes, proprietary tool used | A1–A5, B4, C3–C4 | Process LCA | Not stated |
| Stephan et al. (2013) | Suburban neighbourhood | Modelled typologies | Estimated | Modelled estimates | Cradle to grave | Input–output LCA | 100 |
| Smith & Gill (2022) | Neighbourhood-scale synthetic models | Five archetypes of residential construction | Yes | Modelled underground services | Cradle to grave | Process LCA | 40 |
| Trigaux et al. (2014) | Four abstracted neighbourhood models | Modelled typologies | Modelled based on archetypes | No | Cradle to grave | Process LCA | 60 |
[i] Note: LCA = life-cycle assessment.
The scope of LCAs at larger scales can vary significantly, with varying FUs and different elements of siteworks included or excluded (Lotteau et al. 2015). Terminology varies across the literature, so careful scrutiny is required. In some studies the term ‘infrastructure’ refers to the landscaped surfaces of transport networks such as roads and tram lines, but underground services are not included (Butters et al. 2024; Du et al. 2015; Stephan et al. 2013). For clarity the present paper defines infrastructure as including utilities and buried services, and roadways are categorised as external landscape elements. In other studies infrastructure also includes buried underground services, although there is a range of scope definition for which utilities are included. Of the studies that include buried services, a variety of methodologies is used to estimate these, including models based on local design standards (Trigaux et al. 2014), estimates from previous research (Rankin & Saxe 2024) and specialised LCA tools for infrastructure (Sjökvist et al. 2025). In most of these studies the FU covers an area too large for detailed LCAs to be feasible, and therefore some other methods of estimating infrastructure or sources of secondary data are required.
Sjökvist et al. (2025) analyse a neighbourhood-scale redevelopment in Copenhagen, using an archetype-based approach for building LCAs based on representative studies from across Denmark. Landscape elements between buildings are based on actual project data for the area. Supplementary information included with the paper confirms that below-ground services are measured using a proprietary LCA tool for measuring infrastructure, however a detailed methodology for this tool is not provided. Trigaux et al. (2014) analyse the WLC carbon of four residential building typologies ranging from detached houses to apartments in a context with their external works. Below-ground infrastructures such as piped services were included in this case using modelled archetypes based on local building codes. The studies by Trigaux et al. and Sjökvist et al. are significant as they include site-preparation works such as excavation and earthworks in the LCA scope. Information on this phase of works can be hard to obtain, as the works are often part of a separate contract; however, some research has suggested that GHG emissions due to earthworks can be significant (Forsythe & Ding 2020).
Nichols & Kockelman (2014) use models to estimate energy demands for five neighbourhoods in Austin, Texas. Existing studies are used to estimate the embodied energy of buildings using EC/m2 of floor area. Water and wastewater pipe materials are estimated with geographical information system (GIS) data for locations. It is not stated whether other elements other than pipework were included, or other additional services such electrical supply or lighting ducts. Smith & Gill (2022) carry out LCAs for a series of abstracted residential typologies based on examples in the US. They include an analysis of the infrastructure, utilities and external surfaces that serve the developments, although only water and drainage pipework are included. Several studies quantify service infrastructures by estimating linear metres of service runs and multiplying these by their cross-sectional area to estimate quantities (Rankin & Saxe 2024; Schiller 2007; Troy et al. 2003). The drawback of this method is that it does not count for service connections, discrete elements such as manholes and inspection chambers which can form a significant proportion of total siteworks EC.
Summarising the literature that includes elements of siteworks in an analysis of GHG emissions, it can be seen that there is a wide range of material, temporal and physical scope, making comparison between studies difficult. There is also a clear predominance of modelled and theoretical analysis, rather than real-world assessments, with many studies relying on estimates from previous research, or modelled archetypes to generate data for either buildings or siteworks, suggesting the need for further research. The complexity of real-world development sites containing multiple buildings and infrastructures presents practical challenges for LCA research. The current paper seeks to bridge this gap by providing a framework for structuring LCAs at increasing scales from buildings to neighbourhoods. Results from siteworks can then be compared with building-level LCAs, thereby linking questions of density and compact growth with research on the WLC impacts of construction.
3. METHODS
The present research develops a framework for separately assessing the various building typologies, landscaping and infrastructural elements that comprise a large residential development. A case study project in Ireland is used to illustrate the application of this framework. Following a brief description of the case study, the LCA calculation methodology for all construction elements is described. The framework for physical systems boundaries at different scales is then defined. Finally, there is a detailed description of the siteworks scope as this is the innovative part of the research.
3.1 DESCRIPTION OF THE CASE STUDY
The FU of the case study development is the area within the red line shown in Figure 1. This a recently built housing development for 184 dwellings on the outskirts of Dublin, whose layout can be considered typical for contemporary housing production on the edge of an Irish town or city. In this type of development, it is common for new roads and infrastructure to be provided by the developer, and for these to be later taken in charge by the local municipal authority, a model characteristic of neoliberal urban policy where private developers initially fund and implement housing infrastructure at a local scale and recoup the cost through the sale of the units (Gallent et al. 2020). The majority of the residential buildings are two-storey houses, arranged either in pairs (semi-detached) or as row housing (terraces). Two other higher density typologies complete the development: 28 stacked duplex units of three storeys, and 44 apartments in separate four-storey blocks of 22 units each. These types represent different densities, ranging from 30 dwellings/ha for the houses, 57 dwellings/ha for the duplexes and 103 dwellings/ha for the apartments. The higher density apartments are provided in order to meet regulatory requirements for higher overall densities. The resulting mix of typologies and densities is characteristic of contemporary development in Ireland. The project is part of a larger master plan for the area, but the road network and site layout is designed and constructed by the developer, as are other public infrastructures, including water, public lighting, electricity and telecoms. Car parking is entirely provided at the surface level. It can be seen from the site plan (Figure 1) that a significant part of the site is given over to the new road network, highlighting the fact that this aspect needs to be accounted for in any comprehensive LCA of new development.
Figure 1
Case study project site plan with the three residential typologies.
3.2 SCOPE OF THE LCA STUDY: A COMPARISON OF EU LEVEL(S), EN 15978 AND RICS
The scope of the study for all construction elements follows that set out in EU Level(s) (European Commission 2024) and the European framework for sustainable buildings and system for LCA. This is based on calculation methods set out in EN 15978 (European Commission 2011) and further described in Section 3.3 below. The study is a process-based or bottom-up EC assessment that quantifies all available construction materials based on the project drawings and specification documents. It can also, therefore, be understood as a material flow analysis of all materials within the defined system boundary, to which carbon factors are then applied. The scope of Level(s) provides a limited description of external siteworks elements that are to be included in an assessment, but does not define a physical boundary for these. EN 15978 is more specific, defining the site as the physical space of land occupied by, and attached to, the building, but it excludes construction works outside of the curtilage of the site, such as infrastructure for energy, water, waste and transportation. This would therefore exclude significant siteworks elements associated with a multi-building development such as the current case study (Figure 1). At an urban block and neighbourhood scale, the scope adopted in this study follows standards from The Royal Institute of Chartered Surveyors’ (RICS) whole-life carbon assessment for the built environment (RICS 2024), which include total project impacts within the red line boundary, including siteworks and infrastructure associated with master plans or multi-asset developments of this type. The RICS standards also provide guidance on measuring individual building types where these are repeated, and their results can be extrapolated to determine total impacts.
3.3 SCOPE AND URBAN MORPHOLOGY
In order to address the problems of physical system boundary definition noted in the literature review, a scalar system adopted from the discipline of urban morphology has been proposed. This adapts the multilevel diagram proposed by Kropf (2014) as a tool for defining the physical boundaries of LCAs. Adopting this framework allows for the separate assessment of EC impacts of different building typologies and their associated siteworks within a site.
In this case, three main levels of analysis are defined, each composed of spatial elements that complete that level in the hierarchy. The terminology used can vary, so for consistency in this case the scales of analysis follow the description of Berghauser Pont & Haupt (2023). The main scales of analysis are illustrated in Figure 2, and are defined as follows:
Building level
This is generally a straightforward definition with a physical scope corresponding to the built footprint. In more complex urban developments, separate buildings may share elements of construction such as basement car parks, and these would need to be accounted for at the larger plot scale.
Plot level
The plot can be an individual plot containing multiple dwellings, such as an apartment building, or a plot series combing multiple individual dwellings and their plots that form an urban block, such as is the case with suburban row housing. Berghauser Pont & Haupt (2023) use the term ‘island’ to describe the aggregate of such individual plots; however, for simplicity it is not necessary to introduce this additional term for the LCA methodology presented here. The boundaries of the plot level are clear in this case study, being the area of ownership around each building.
Fabric level
The fabric or tissue scale includes roads, footpaths and publicly accessible elements. The fabric boundaries generally lie along the centre line of roadways between typologies, or on the edge of the block where it abuts a site boundary. The fabric boundaries also generally describe the main lines of drainage infrastructure.
District level
The district or neighbourhood scale is the next level in the framework; however, it is not included in the case study because it is outside of the scope. This scale is composed of fabric boundaries connected with additional spaces such as parks, public road networks or waterways. The choice of the correct level of study is critical for correctly allocating GHG emissions. The case study shown here relies on a public road to the east of the site boundary; however, this is constructed by the municipal authority, so it is considered to be neighbourhood-level infrastructure whose emissions should be accounted for by the public entity that constructed it.
Figure 2
Physical scope of life-cycle assessment (LCA) organised by levels of urban morphology.
Figure 3 illustrates the division of the case study site into individual typologies and their external areas defined by plot and fabric-level boundaries. Separate EC assessments are carried out for each building typology, and also for their landscape and infrastructural elements at plot and building scale. This allows for separate comparison of the EC impacts of the typologies together with their context. Note that EC impacts due to common areas are assigned to the different typologies using a weighted distribution, and this is described further in Section 3.5 below.
Figure 3
Comparison between scopes of different methodologies.
Examples of the multilevel diagram (Figure 4) show the scope of LCAs defined by levels of urban morphology as applied to two of the residential typologies in the case study. The building-level LCA describes construction within the building footprint or envelope for all typologies. The plot-level LCA is comparable with the Level(s) scope described previously. In the case of the plot of an individual house, it consists of the front and back gardens which are in private ownership. In this case, the total urban block or plot series is calculated for all houses, and this is then divided to find the individual EC for each plot. The apartment building has multiple dwellings on a single urban plot that share outdoor open space, so in this case the plot and the plot series have the same morphological boundary. The fabric or public-level LCA includes all streets and common areas in addition to vehicle and pedestrian circulation. Public elements that serve the entire site are divided between the typologies as described in Section 3.5 below. The advantages of conducting separate EC assessments for these different levels of urban morphology are as follows:
They create a basis for comparison between new-build projects where infrastructure must be provided and projects where the public external areas and infrastructure already exist, such as developments on infill urban sites where building and plot-level LCAs only are counted.
They allow for a comparison between LCA methodologies with different scope such as Level(s) and RICS.
Disaggregating sites into component elements allows for a direct comparison of the EC impact of buildings only against their external areas and infrastructure.
Figure 4
Life-cycle assessment (LCA) scope for low-rise houses (left) and apartment building (right) defined by categories of urban morphology.
Source: Adapted from the multilevel diagram of Kropf (2014).
3.4 LCA METHODOLOGY
The analysis of EC is carried out as part of an LCA, according to the calculation methodology set out in EN 15978. The study analysed life-cycle stages A1–A5, from cradle to practical completion, which include the manufacture of materials, their transport to site and construction works.
Construction-level information was obtained from the project developer for the purposes of carrying out the EC assessment, and additional information was obtained from public databases with the building permit for the development hosted by the local government authority. Information from the bill of quantities was used where possible, and additional take-off drawings were prepared where required for the purposes of estimating the quantities of materials. These quantities were then entered into a bill of quantities in preparation for estimating the EC impacts. Quantities were prepared using different FUs, and these were converted into the total kg of material using density figures from the Inventory of Carbon and Energy (ICE) Database managed by the University of Bath,1 where available. Uncertainties in the construction data are apparent at this point, as materials are often called up using generic specifications, and it is not possible to know exactly which products were used. In cases of doubt, generic material factors were used from the ICE Database.
The tool used to carry out the assessment is an Excel-based spreadsheet tool called Upfront, developed by the Irish Green Building Council (IGBC) as part of INDICATE, an EU-funded project to provide building-level data for whole-life carbon in Europe. The Upfront tool follows the scope set out in EU Levels, and calculates WLC by construction element and life-cycle stages according to standards set out in reference document EN 15978. The assumptions regarding life-cycle stages A1–A5 within the Upfront spreadsheet are as follows:
A1–A3: Product stages
The spreadsheet has common construction materials built-in to choose from when entering quantities. The global warming potential (GWP) of materials is generally calculated using carbon factors taken from the ICE Database. Further carbon factors are provided by LCA consultancies Cambridge Architectural Research (CAR) and Circular Ecology. Environmental Product Declarations (EPDs) are used where specific product information is known in order to estimate the fossil fuel GWP of materials used. Biogenic carbon stored in timber and other bio-based products is not included in this study as it only covers stages A1–A5.
A4–A5: Transport and construction stages
Figures for transport distances are taken from the EPD Ireland Product Category Rules managed by the IGBC. These assume distances of 100 km one-way travel for bulk materials, 200 km for other materials and 1000 km for non-Irish materials arriving by air or sea. Carbon factors for materials arriving by both road and sea are taken from ISTRUCTE (2025). Estimates for construction waste included in stage A5 are taken from the calculation rules for life-cycle stages set out in the EU Level(s) manuals;2 however, energy used in the construction process is not taken into account in these stages.
3.5 LCA SCOPE FOR SITEWORKS
As the case study development was constructed on a greenfield site, external landscaping and infrastructure serving the dwellings were also required, and this amounts to significant additional construction and resulting EC that was measured in separate EC assessments. The methodology, including data sources and carbon factors for these assessments, follows that described for buildings in Section 3.4 above. An analysis of the construction drawings and documents shows multiple service utilities that need to be accounted for in the plot- and fabric-level LCAs (see Figure S1 in the supplemental data online).
The siteworks were divided into two categories:
External landscaped areas and their construction build-ups, including asphalt roads, concrete footpaths, and permeable paved areas and car parking.
The infrastructures that serve the dwellings: foul and surface water drainage, public lighting, water supply, electricity supply, and telecoms.
External landscaped areas included in the siteworks assessment were as follows:
asphalt roads: constructed to national standards
concrete footpaths: for pedestrian and vehicular traffic
paved areas: pedestrian areas around dwellings
car parking: permeable paved areas for vehicular traffic
tree-pits, swales and green areas: sustainable drainage systems (SUDS)
fencing and enclosures: concrete post and timber infill fencing.
Infrastructural utilities included in the siteworks assessment were as follows:
foul water drainage: concrete pipework, manholes and individual connections to dwellings
surface water drainage and SUDS
public lighting: lighting poles and ductwork
water supply: pipework, bulk water meters, boundary boxes and sluice meters
electricity supply: individual connections and ductwork
telecoms: individual connections and ductwork.
The EU Level(s) scope being followed includes the requirement to assess external areas, broken down into utilities and landscaping. The categories of landscaping elements are described by Level(s), however no physical system boundaries are described, so these are set in separate LCAs as either plot level or fabric level as per the description in Section 3 above. Service utilities are not described in detail in the scope of EU Level(s), which only includes categories for connections, diversions, substations and equipment. Other infrastructural works are not referred to in Level(s), although these will be required for a complete LCA, and they are therefore included in this study.
To estimate the infrastructural elements, drawings were prepared that calculated the length of all service runs and connections listed under the relevant utility. Construction materials, specifications and quantities of these elements could be estimated with a degree of certainty as the specification of infrastructure in Ireland is regulated by the relevant national bodies. An illustrative example (see Figure S2 in the supplemental data online) shows the drawings used for the quantification of two of the highest impact infrastructural elements: two attenuation tanks made of polypropylene, each 100 m long; and 84 concrete manholes, each using over 3 t of precast concrete.
For the landscaped areas, drawings were prepared to quantify the surface areas of roadways, footpaths and external landscaping. Typical construction build-ups were then used to calculate stages A1–A5 EC of each element of landscaping (see Figure S3 in the supplemental data online). In order to provide EC figures for the landscaping and infrastructure due to each building typology, these were included separately in the plot- and fabric-level LCAs for each type. Finally, external areas in common between all the types were distributed in the individual public LCAs using a weighted distribution based on gross internal floor area (GIFA) of the dwellings. In this case study, a common area to the west of the site contains two large stormwater attenuation tanks that contain significant quantities of EC (see Figure S2 online), and the EC emissions of this element are distributed across the dwelling types using this method.
4. RESULTS
The first set of results presented here are the LCAs for the three building types analysed: semi-detached house, duplex and apartment buildings. Following this, the breakdowns of landscaping and infrastructural elements are shown. Finally, the two sets of results are combined, showing the share of landscaping and infrastructure for each housing type.
4.1 BUILDING-LEVEL LCAs
The overall results of the building-level LCAs are presented here. The total stages A1–A5 emissions for each typology were divided by the gross internal floor area (GIFA) and gave kg CO2e/m2 of dwelling floor area as follows: house, 316; duplex, 396; and apartment, 437.
As can be seen, the house typologies have a lower EC impact per m2 of floor area, and this figure rises for the denser typologies. Figure 5 shows the detailed stages A1–A5 results for each type of dwelling. Table 2 shows parametric calculations for the four highest impact materials for each typology, showing the weights, carbon factors and sources of data.
Table 2
Parameters of the four highest impact materials for each typology for stages A1–A3.
| DWELLING TYPOLOGY | MATERIAL (t) | A1–A3 GWP (kg CO2e/kg) | CARBON FACTOR SOURCE | EC PER DWELLING (kg CO2e) | EC/m2 (kg CO2e/m2) |
|---|---|---|---|---|---|
| House | |||||
| Concrete in-situ 28/35 Mpa | 37.32 | 0.13 | ICE Database | 4,704 | 40 |
| Concrete blocks: medium density | 29.92 | 0.10 | ICE Database | 2,985 | 25 |
| Insulation: mineral wool | 2.03 | 1.31 | ICE Database | 2,663 | 23 |
| Concrete roof tiles | 5.78 | 0.27 | Circular ecology | 1,543 | 13 |
| Duplex | |||||
| Concrete in-situ 28/35 Mpa | 34.20 | 0.13 | ICE Database | 8,621 | 42 |
| Brick | 11.94 | 0.21 | Cambridge Architectural Research (CAR) | 5,088 | 25 |
| Lightweight concrete blocks | 8.68 | 0.26 | ICE Database | 4,861 | 23 |
| Insulation: mineral wool | 1.65 | 1.31 | ICE Database | 4,332 | 21 |
| Apartment | |||||
| Precast concrete slabs | 42.81 | 0.22 | ICE Database | 9,366 | 101 |
| Concrete in-situ 28/35 Mpa | 30.45 | 0.13 | ICE Database | 3,791 | 41 |
| Stainless steel | 0.81 | 4.41 | National average Environmental Product Declaration (EPD) | 3,558 | 38 |
| Cement mortar (1:3 mix) | 16.60 | 0.18 | ICE Database | 3,041 | 33 |
[i] Note: EC = embodied carbon; GWP = global warming potential; ICE = Inventory of Carbon and Energy; Mpa = megapascal.
Figure 5
Embodied carbon (EC) of each typology by building element per m2: stages A1–A5.
Considering the results at the building level, it can be seen that EC increases with residential density. The low-rise semi-detached houses have the lowest EC impact figure of 316 kg CO2e/m2, largely due to the use of a low-carbon timber frame structure. The results for the semi-detached house show that significant carbon is expended in the ground floor, substructure and external walls. The timber frame structure is inherently low carbon, however the buildings are finished externally in rendered concrete blockwork or brick: high-carbon materials that are visible in the results. The roof is also a significant source of carbon emissions due to the concrete tiles used and the amount of material located in the relatively large volume of the pitched roof structure.
The duplex dwelling has a higher figure of 396 kg CO2e/m2, and this is to be expected as the ground floor unit and first floor slab of this building are built in concrete of various types. This is due to the requirements for fire separation between the stacked units as required by building regulations, resulting in concrete walls and a first-floor slab. Another contribution to the high external wall figure is the increased area of brickwork, a material that has a high carbon factor. The external staircase also appears as a carbon hotspot, being made out of in-situ concrete.
The apartment building has the highest EC impact of the three types studied, at 437 kg CO2e/m2. This is to be expected as the building uses either in-situ concrete or concrete blockwork for the load-bearing walls, and precast concrete slabs for the upper floors and roof structure. The external walls make up a lower percentage of the total emissions for stages A1–A5, showing that the more efficient form factor of the apartment and consequent reduced external surface leads to efficiencies when compared with the low-rise types, and that this partly offsets the higher EC of the concrete structure.
4.2 COMBINED RESULTS INCLUDING LANDSCAPING AND INFRASTRUCTURE
The results in Table 3 show that landscaping and infrastructure form a significant part of the total EC of residential construction. In order to allow a comparison between building typologies of different sizes, the EC per dwelling is first listed, then divided by GIFA to give EC/m2. Table 3 breaks down the results by typology, showing that the low-rise house types have the largest proportion of EC per dwelling due to infrastructure, and this decreases for the duplex and apartment types. The final column summarises this key result. This column assumes the EC due to buildings as 100% for each typology, and then lists the percentage of additional EC due to siteworks. For the low-rise houses, this represents an additional 32% of the building EC, falling to 19% for the duplex dwellings and 12% for the apartments. The two-storey semi-detached houses have a high EC figure for landscaping in both the private (plot) and public (fabric) LCAs due to the large areas of asphalt and concrete given over to vehicular traffic. The higher density typologies demonstrate a more efficient layout of service infrastructure at the level of the urban block. This is illustrated in Figure 6, which compares the foul water infrastructures for a single urban block of low-rise houses with one for apartments. It can be seen that the low-rise typologies require individual carbon-intensive service connections for each dwelling, whereas the apartment building has a single external connection, and then distributes services internally. This pattern is repeated for each individual service utility.
Table 3
Stages A1–A5: combined results including landscaping plus infrastructure.
| DWELLING TYPOLOGY | TOTAL EC PER DWELLING, STAGES A1–A5 (kg CO2e) | GROSS INTERNAL FLOOR AREA (GIFA) (m2) | EC/m2 FLOOR AREA, STAGES A1–A5 (kg CO2e/m2) | EC PER DWELLING, INCLUDING LANDSCAPING AND INFRASTRUCTURE (kg CO2e) | EC/m2 INCLUDING LANDSCAPING AND INFRASTRUCTURE (kg CO2e/m2) | PERCENTAGE EC FOR LANDSCAPING AND INFRASTRUCTURE (%) |
|---|---|---|---|---|---|---|
| House | 37,239 | 118.0 | 316 | 49,546 | 420 | 32% |
| Duplex | 40,941 | 103.5 | 396 | 49,199 | 475 | 19% |
| Apartment | 40,662 | 97.7 | 437 | 45,908 | 494 | 12% |
[i] Note: EC = embodied carbon.
Figure 6
Difference between foul water services layout for 26 low-rise dwellings (left) and 22 unit apartment building (right).
Figure 7 shows the three building typologies and the results of their associated EC emissions for infrastructure and landscaping. A similar pattern of decreasing emissions with increased density of the typologies can be observed for both infrastructures and landscaping. Another clear pattern seen in the results is the concentration of EC impacts associated with vehicle infrastructure due to both the material impacts of roads, footpaths and parking, and the drainage infrastructures that lie underneath these elements. The green areas appearing in the landscaping results are mainly tree pits and swales forming part of the SUDS strategy, and these have a significant impact mainly due to the large quantities of gravels and crushed aggregate required in their construction.
Figure 7
Stages A1–A5 kg CO2e of infrastructures per dwelling typology by service utility (above) and A1–A5 kg CO2e of landscaping per dwelling typology (below) by landscape element.
Figure 8 breaks down the EC of individual service infrastructures for all typologies at the fabric level. It can be seen that surface water drainage (including SUDS) is the most impactful utility, followed by foul water drainage and water supply. It also shows the breakdown of EC results by material for these infrastructures, with plastics responsible for the highest impact, followed by concrete and metals. Significant quantities of plastics are present in the surface water drainage, including two 100-m-long attenuation tanks made from polypropylene buried along the western border of the site. Large volumes of plastic are also found in the runs of polyvinylchloride (PVC) pipework that connect the various underground drainage elements and feeds storm water to the large attenuation tanks.
Figure 8
Stages A1–A5 t CO2e by public infrastructure service utility (above) and material (below).
Returning to the categories of urban morphology described in Section 3 above, the proportions of EC due to each level of morphology are illustrated in Figure 9. Measured across all three categories, the results show similar EC emissions per dwelling, despite the much higher concentrations of concrete and steel in the higher density types. The lower density house types, by contrast, have higher EC impacts at plot and fabric levels, highlighting the importance of expanding the system boundary to include these levels. An inverse relationship is seen between the EC of siteworks and residential density, with the lower density house types having the highest proportion of EC due to siteworks, and the denser apartment types the least.
Figure 9
Stages A1–A5 t CO2e per dwelling by urban morphology and dwelling typology.
4.3 ASSUMPTIONS AND LIMITATIONS
It proved difficult to find sufficient data to carry out detailed LCAs, and certain assumptions were made as a result. It was not possible to obtain mechanical and electrical services drawings or detailed specifications for any of the projects, so estimates based from studies carried out by the Chartered Institution of Building Services Engineers (CIBSE) were used. These estimates use methodology TM65 (Embodied Carbon in Building Services) (CIBSE 2023), which provided EC estimation for stages A1–A5 of residential projects with similar service installations. Earthworks and site preparation phases were not included in the study as there was no information available.
5. DISCUSSION
5.1 THE EC IMPACT OF SITEWORKS IN RELATION TO TYPOLOGY AND DENSITY
The results establish a clear relationship between residential building typology and the EC impact of siteworks, although the case study only considers three building typologies in the construction and regulatory context of one country. The proposed framework could provide a basis for larger studies that could consider a wider range of building typologies and site layouts in order to investigate if the inverse relationship between residential density and siteworks EC found here holds true in other contexts. The results have established the fact that the provision of vehicular infrastructure and associated landscape components is associated with high EC emissions. This difference is due to several related indicators: the total length of the infrastructure network, the number of carbon-intensive service connections such as manholes, and the associated landscape components such as roads, footpaths and parking areas. The framework set out in this paper divides urban areas into building typologies and their fabric boundaries, which correspond to the high carbon vehicle infrastructures identified. Future studies could use detailed case studies such as that shown here to provide material for typical siteworks elements, and the network length of fabric boundaries could provide a useful proxy indicator for overall EC impacts due to infrastructure.
5.2 EARTHWORKS AND SITE PREPARATION
The siteworks section of the LCA case study included a complete inventory and quantification of all known infrastructures, service utilities and landscape elements. As previously noted, it did not include earthworks and site preparation, as information about this work phase was unavailable. The limited literature that considers the GHG emissions of earthworks construction suggests that excavations can be a significant contributor, particularly on sloping sites (Forsythe & Ding 2020). Another study of a dense urban redevelopment found that that it formed one-fifth of EC emissions (Sjökvist et al. 2025), although the development in question featured carbon-intensive engineered elements such as retaining walls and canal excavation. Further research is therefore required on the impact of earthworks for different types and scales of development. The scope of such research could be expanded to include other climate impacts relating to earthworks and ground disturbance, such as the environmental effects of sealed and impermeable surfaces associated with low-density suburban development. These effects include increasing flood risk due to reduced penetration of rainwater and increased storm water run-off (Murata & Kawai 2018), and the capacities of soils to absorb carbon (Yan et al. 2015).
5.3 CARBON SEQUESTRATION AND THE ROLE OF NATURE-BASED SOLUTIONS
While the study has included imported topsoil and lawns in the material inventory, it does not consider the potential for these and other natural elements such as trees and plants to sequester and store carbon. Local data on the carbon sequestration of natural elements for different contexts is hard to come by, although the balancing effects of such carbon sequestration could be significant. For example, Zirkle et al. (2011) found that home lawns in the US will sequester carbon under a range of scenarios.
Some of the highest contributors to EC emissions found in the results are due to SUDS, such as permeable paved retention areas, attenuation tanks, and associated manholes and pipework. In several cases these could be replaced by nature-based solutions (NBS), e.g. concrete storm-water pipework could be replaced by swales, and a retention pond could be used instead of the plastic attenuation tanks. LCAs of various types of urban green infrastructures have shown that these can provide significant environmental benefits (Romanovska et al. 2023). However, NBS for stormwater drainage have also been found to have high EC emissions in excess of their capacity to store biogenic carbon due to their construction and maintenance (Moore & Hunt 2013). A productive area for future research might compare traditional underground drainage solutions such as those studied here with their equivalent NBS equivalents. The literature suggests that the cumulative annual effect of carbon sequestration and storage of natural elements over time can be significant, and it would be useful to compare this with upfront EC emissions during construction.
6. CONCLUSIONS
This paper addressed the lack of clarity regarding system boundaries for carbon quantification outside the building envelope that has been noted in both the literature and standard frameworks for sustainability assessment such as European Union Level(s). It adopted a scalar approach from the discipline of urban morphology that describes a hierarchy of urban elements, starting from the scale of rooms forming individual buildings to their organisation into urban blocks and neighbourhoods. This system allows for the accurate definition of physical boundaries for life-cycle assessments (LCAs) depending on the scale being studied. Separate LCAs are therefore carried out at different levels of urban morphology: building level, plot level (areas under private ownership or control) and fabric level (public areas). This definition of LCA scopes at different scales provides greater certainty around material quantification. In more comprehensive studies, these categories could be scaled upwards to neighbourhood and urban district levels, or downwards to include building elements and materials.
Integrating building LCAs with the study of larger urban areas gives the further advantage of including the embodied carbon (EC) of siteworks and infrastructural elements in a holistic assessment of the climate impacts of new urban development. The results of the case study show the usefulness of this framework, as the different building typologies show different overall proportions of EC impacts due to siteworks and infrastructural systems. These range from an additional 32% for low-rise housing typologies to 19% for duplex dwellings, reducing to 12% for apartment buildings. The reasons for the higher figures associated with low-rise development are visible in the results, and include the large quantities of hard-landscaped areas dedicated to vehicle infrastructure, and the multiple individual runs and connections required for each dwelling and each service utility. Conversely, an infrastructural efficiency due to shared connections used in multi-unit buildings such as apartment blocks can be seen in the results, providing lessons for the layout of low carbon residential areas. These results show an inverse relationship between residential density and the EC of siteworks and infrastructure, although this relationship is counterbalanced by the higher emissions associated with building structure as density increases. The overall finding of this study is supported by the literature where detached, dispersed building types have been shown to have higher EC impacts in several studies (Pomponi et al. 2021; Rankin & Saxe 2024; Stephan et al. 2013).
This research therefore suggests that future policy on the calculation of whole-life carbon for residential development should precisely define the scope of LCAs at a range of scales beyond the dwellings themselves, to include all siteworks, landscaping, and transport and service infrastructures. This approach could form part of a more holistic approach to sustainability in the housing sector, that until now has been largely focused on individual dwellings and the energy they consume. It has been shown here that combining the EC results of buildings, landscaping and infrastructures can provide an evidence-based assessment of policies promoting higher densities and compact urban growth that are proposed as goals of sustainable urban development.
Notes
ACKNOWLEDGEMENTS
This research was carried out in collaboration with the Irish Green Building Council (IGBC).
COMPETING INTERESTS
The authors have no competing interests to declare.
DATA ACCESSIBILITY
The data supporting this study are provided in the supplemental data online. Additional data are available from the corresponding author upon reasonable request.
SUPPLEMENTAL DATA
Supplemental data for this article can be accessed at: https://doi.org/10.5334/bc.668.s1