Introduction

The construction industry globally consumes 40% of natural materials, 40% of the total primary energy, 15% of the world’s freshwater resources, generates 25% of all wastes and emits 40–50% of GHG (Ding 2014); the design team is thus charged to adopt an environmentally responsible approach to their design solutions and construction materials’ specification choices.

The building sector is responsible for 20% of Australia’s total energy demand and 23% greenhouse gas (GHG) emissions (Lawania and Biswas 2016). It is projected that 460,000 new houses will be constructed in Western Australia (WA) by 2030, which will necessarily increase demand for construction materials and impact energy usage (NHSC 2011). Without due consideration of environmentally conscious specification choices, the construction industry in Western Australia will experience significant GHG emissions increases; there will be depletion of finite resources, and landfill over flow. This will result in a challenging situation requiring ongoing federal government and local authorities’ ‘green’ tendering guidelines and not least, requires respective design teams and builders to make informed decisions when specifying materials.

Life cycle assessment (LCA) is a decision-making tool that can assist stakeholders in identifying opportunities to make sustainability gains for built assets by selecting the most environmentally friendly option (Seidel 2016). It quantifies and accesses the inputs and outputs affecting environmental performance associated with a product, process or activity throughout its life cycle (Whyte 2012). Whilst the LCA technique is somewhat commonly available, uptake by industry remains still limited. Case-study examples such as those presented here can be argued to increase the profile of LCA application and, by extension, encourage a more sustainable design process (Crawford et al. 2016).

Roofing accounts for 6% of a low-rise building’s volume and, typically, 7% of a residential building’s GHG emissions (from mining to material production) (Lawania and Biswas 2018). Saiz conducted an LCA of a so-called green roof in Madrid, finding that its low solar absorbance resulted in a reduced level of energy demand for the building (Saiz et al. 2006). Chenani et al. (2015) similarly determined the environmental performance of two lightweight ‘green’ roof systems and found that an environmental impact reduction, through layers’ configurations review, was possible. Another recent LCA study found that the use of vaulted roofs can reduce embodied energy by over 40% relative to flat slabs (Huberman et al. 2015).

However, to the best of our knowledge, no study to date studied how life cycle environmental performance(s) of roofing varies with materials choices. For WA’s housing sector, roofs of three different types are common: sheet metal, concrete tile and clay tiles; these alternatives are compared in this current study to determine the environmental implications of different specification options for West Australian climatic conditions.

The next section outlines the methodology to assess the alternative roofing specifications in terms of carbon footprint and embodied energy demand comparisons. Carbon footprints of roofing options are compared, both with and without recycling factored-in since many areas in WA have no available local recycling facilities.

Methodology

The LCA conducted here follows the guidelines outlined in ISO14040-44 (ISO 2006) which consists of four steps: goal and scope; inventory analysis; impact assessment; and interpretation.

Goal and scope

The goal of this study is to assess the environmental impact involved with the use of roof-covering materials in WA residential houses. Three roof-covering materials are considered: clay tile; concrete tile; and, sheet metal.

The system boundary of the LCA study covers the entire life cycle of the product. This is broken down into several stages including raw material acquisition, manufacturing and processing of construction materials, transportation of these materials to the construction site, construction phase, usage stage and ultimately disposal/recycling residual considerations.

The functional unit of this study is 100 m2 of roof-covering materials and the timber structure framework supporting it; the environmental impacts of respective (typically) timber framework(s) will be analysed with the roof-covering material options. The reason for choosing this functional unit was that approximately 50% of average houses in Perth have this size of roof area (Department of Water and Environmental Regulations 2017). Other dwelling superstructure elements are not analysed. This is a process-based LCA, where energy and chemical inputs of all stages during the life cycle of roof-covering materials, have been utilised in assessing global warming impacts and embodied energy demand (Suh et al. 2014; Majeau-Bettez et al. 2011).

Quantities of the structural timber framework supporting the roof-covering materials are calculated in accordance with typical sections made available by local governments in Western Australia. Each item was classified according to its base material (treated pine, concrete, zinc, and aluminium). Natural gas and electrical energy are included where appropriate in consideration of the manufacture processing and installation of roof coverings (Life Cycle Strategy 2015; BPIC 2014).

Construction locations are urban; local suppliers were contacted to determine respective manufacturers/fabricators factory locations. Some industry representatives who provided materials and transportation related information will not be released in the paper due to requests for confidentiality. Where appropriate (raw) materials are deemed to have been shipped to (Fremantle) port and then road-transported to distributors/site. Timber/structural frameworks and roof-covering material installation is deemed by tradesman in situ; (steel/nail) fixings to install the roof-covering material have been included.

It is beyond the scope of this LCA to consider 100% data directly and indirectly associated with the production and use of these roofing materials. Therefore, the GHG emissions and embodied energy demand values that have been calculated using available data are relative values, and these results were used for comparison purposes (Suh et al. 2014; Majeau-Bettez et al. 2011).

Life cycle inventory

LCIs were conducted for clay tiles, concrete tiles and sheet metal on residential buildings to calculate energy and material consumption in all stages of roof life cycle.

Mining to material production

Composition and percentages of raw materials used in clay tiles, sheet metal and concrete tiles, respectively, were assessed (Table 1). For concrete and clay tiles roof, raw material quantities were calculated; the number of tiles per metre square multiplied by the mass per tile gives the total mass of tiles; the total mass then multiplied by relevant percentages finds raw materials required in 100 m2 (Whyte 2015). The timber frame to support the clay tiles, concrete tiles and sheet metal were sourced from typical sections and drawings from the City of Melville, Western Australia, alongside AS 1684.4 (Whyte 2015). A sample calculation is provided in “Appendix A”.

Table 1 Bill of materials for clay, concrete and sheet-metal tiles

Table 2 shows the amount of natural gas and electricity that was required to produce 100 m2 of clay tile, concrete tile and sheet metal from raw materials (BPIC 2014; Atlas Steel 2014).

Table 2 Manufacturing and transportation information for clay, concrete and sheet-metal tiles

Transportation

Construction sites for roof erections is Western Australian urban/metropolitan; shipping freight transportation is assumed from the (interstate) manufacturer/fabricator to WA (Fremantle) port. In order to estimate the tonnes-km (i.e. t km) travelled by land, and sea Google Earth was used for calculating the distances in kilometres between origins and destinations. Local WA industry representatives (Boral and BlueScope Steel and others) noted that, typically clay tiles and sheet metal are manufactured in New South Wales, with concrete tiles made in Victoria. Shipping distance between NSW and WA (Fremantle) has been calculated as 2195 nautical miles (4065.14 km) and between Victoria Melbourne and Fremantle—1681 nautical miles (3113.2 km). Upon arrival at (Fremantle) port, articulated trucked road transportation of materials from the port to distributor/site is calculated as 22.6 km.

Construction

The construction stage involves the construction of the timber framework supporting the roof-covering materials, and the installing of the roof-covering materials, typically: marking-out (tape/pencil/chalk line) the timber; cutting the timber (drop saws/power saws/hand saws); and, nailing members together (hammers/nail guns, with two galvanised steel nails per tile/length or alternatively, nails by mass of mild steel sheet).

The energy consumed in in situ tradesman installation and tile/sheet connection is deemed nominal; effectively positioning and installing the tile/sheet and nailing the tile/sheet onto the battens is a manual process with nominal electrical equipment used.

Use

The usage stage quantifies the effects of varying solar reflectance of roofing materials (clay tile, concrete tile, sheet metal). This research shows that an effective R factor (thermal resistance of roofing material) affects heat loss and heat gain of each roofing material. Physical data and assumptions are inputs in Eq. 1 (below) to calculate heat lost and heat gain.

$$q = U\left[ {\left( {T_{2} - T_{1} } \right) + \frac{\partial G}{H}} \right]$$
(1)

q rate of heat flow per square metre from roof to the inside, U the overall heat transfer coefficient between the ambient and inside (W/m2 K), 1/U the thermal resistance. T2 annual average ambient temperature °C, T1 required level of temperature that needs to be maintained inside the house °C, H outside transfer coefficient between roof and ambient (W/m2 K), rate of absorption to solar radiation, G solar radiation per unit area (W/m2).

Table 3 shows the cooling and heating loads of each roofing material in terms electricity consumption. Fossil and renewable energy account for 95.5% and 4.5% of the total primary energy sources for electricity generation (Grant 2015). “Appendix B” shows a sample calculation for how the cooling load has been calculated for clay tiles. Since the thermal modelling software was unavailable during the time of the study, the heating and cooling load at hourly levels was not determined (for more accurate analysis) and hence it is considered as a limitation of this analysis (Robati et al. 2016, 2017). Also it should be noted that the usage stage has only been considered to capture variation in cooling and heating energy demand due to use of different roof-covering materials over their life cycles. This variation was found very infinitesimal when comparing with impacts resulting from other life cycle stages. Therefore, the exclusion of the detailed thermal modelling analysis can be argued as not significantly affecting the overall outcomes of this LCA study.

Table 3 Heating and cooling load of 100 m2 clay, concrete and sheet-metal roof covers

The timeframe of use is deemed the life cycle of such construction materials. According to multiple sources (Boral 2014; Bluescope steel 2014), clay tile has an average of lifetime of 65 years, compared to concrete tile and sheet-metal life cycles of 50 years and 45 years, respectively. For the purposes of this comparative assessment, the greatest value is adopted, assuming a total lifetime of 65 years which implies the need for 1.3 and 1.45 times more concrete roof tiles and sheet tiles than clay tiles during this period, respectively.

End of life

A non-recycling approach is adopted if no local facilities exist, resultantly in such a scenario construction materials are disposed directly to landfill.

In the case of demolition of roofs and the transportation of demolition waste to landfill, two major activities were considered such as the use of tools and equipment used for demolition of roofs and then its transportation for disposal to landfill site (Lawania and Biswas 2018).

Alternatively, if recycling facilities do exist, values in Table 4 (below) are developed towards recycling databases for the three different roof types. Table 4 presents the construction & demolition (C&D) materials recovered and disposed in WA for the 2008–09 financial year (Hyder Consulting 2009).

Table 4 Recycling rate of materials in WA for the 2008–2009 financial year

For recycling waste clay tiles, this study has determined that the tile will be crushed for aggregates towards potential replacement with limestone in road construction work. Therefore, energy required to produce 3% of a clay tile has been reduced from the total raw material acquisition to calculate a net amount of energy.

In the case of concrete tiles, 45% of concrete tiles are typically recovered and recycled. Potentially, all fines/sand waste arisings from the concrete tile will be recovered and reused.

Metal is potentially 100% recyclable (Biswas 2014); however, for this study a practicable local recycling rate of 78% is used (as Table 4); thus, energy consumed to produce 78% of metal has been taken away from a raw material acquisition stage to calculate the net energy for sheeting.

Impact assessment

Input and output life cycle inventory data were entered into Simapro LCA software (Pré-Consultants 2015); application requires relevant materials to be linked to Australian libraries to ensure representative WA conditions. Where libraries did not exist, new libraries were developed from similar LCA studies. Table 5 shows that most of the emission databases were sourced from Australian unit process libraries (Stephan and Stephan 2014) except for silica and Iron–nickel–chromium alloy (where Ecoinvent was used) and natural gas (using Pré-Consultants 2015). Data concerning product residual uptake (cradle-to-cradle) were input to the software (i.e. chemicals, energy demand and heating and cooling loss) and linked to relevant libraries, towards the generation of associated impacts. The libraries are emission factor databases which include all upstream emissions and embodied energy demand of these inputs. The Intergovernmental Panel on Climate Change—IPCC2007—global warming potential (GWP)—method was used to calculate the associated environmental impact(s) of the product(s) (IPCC 2007). The cumulative energy demand method was also used to generate the embodied energy of the products.

Table 5 Emission factors of inputs for this LCA analysis

Equation (2) shows the conversion of masses of different greenhouse gases associated with the production and use of material and energy inputs into global warming potential (GWP) (Fatimah and Biswas 2016), which is a single carbon dioxide-equivalent metric (CO2e-) (Stephan and Stephan 2014).

$${\text{GWP}} \left( {{\text{CO}}_{2} {\text{e}} - } \right) = \mathop \sum \limits_{i = 1}^{i = N} \mathop \sum \limits_{j = 1}^{j = M} I_{i} {\text{EF}}_{ij} x {\text{CF}}_{j}$$
(2)

where I is the amount of an input. i:1,2,…N; type of inputs (e.g. cement, concrete, aluminium, electricity, natural gas). EFij emission factor = amount of emission of GHG type ‘j’ per kg of input of type ‘i’. CFj: CF1, CF2, CFM; … characterization factors of GHGs (e.g. CF is 1 for CO2, 28 for CH4, 265 for N2O).

Following cumulative energy demand method (Fatimah and Biswas 2016), all inputs in the life cycle inventory have been multiplied by the corresponding energy demand values to find out the total embodied energy demand of a roof cover

$${\text{EE}}_{\text{total}} = \mathop \sum \limits_{i = 1}^{N} I_{i} \times {\text{EE}}_{i}$$
(3)

where EEi is the embodied energy demand of an input i.

Results and discussion

Carbon footprint of roof-cover materials

Respective carbon footprints of roof-cover materials have been estimated for both traditional (residual landfill) and also recycling approaches to determine carbon saving benefits.

Traditional approach

As can been seen in Fig. 1a–c, the total carbon footprints of one hundred square metre of sheet metals, concrete tiles and clay tiles are 9.85 t CO2e-, 9.33 t CO2e-and 4.39 t CO2e-, respectively. Sheet-metal roofing is the most carbon-intensive roofing material, because it is made of aluminium, which is 16 and 12 times more carbon intensive than concrete and clay blocks, respectively (NZAS 2011; Dunlop 2013; FISIS 2013).

Fig. 1
figure 1

Carbon footprint and embodied energy demand of a sheet-metal tiles b concrete tiles and c clay tiles

Raw material acquisition accounted for a significant portion (96%) of the total carbon footprint of one hundred metre square of sheet-metal roof, followed by manufacture and processing (2.6%), transportation (0.77%), construction (0.18%) usage (0.02%) and demolition and disposal stages (0.19%). Raw material acquisition for sheet metal is very high due to aluminium production typically in Victoria State smelters using electricity generated from brown coal. This stage of the life cycle is a significant source of carbon dioxide emissions and consequently the use of virgin sheet metal is the least environmentally friendly roof material specification for residential buildings in WA, in locations where no recycling facilities exist. Whilst less timber structural framework is required to support a lightweight sheet-metal roof, this does not contribute to the overall reduction of GHG emissions. The use of carbon-intensive sheet-metal roofs outweighs the benefit associated with use of light, structural roof-carcassing materials. Like other studies (Stephan and Stephan 2014), this study also found that the demolition and disposal stage accounts for the very tiny portion (≤ 1%) of the overall impact.

Followed by sheet metal, concrete tile use ranked second in terms of output of carbon foot print 9.33 t of CO2e-. Acquiring the required materials alongside manufacturing to final product is deemed energy intensive. Whilst raw material acquisition accounted for 51%, a significant portion of carbon footprint was produced during the manufacturing stage (43% of the overall carbon footprint). The is due to the use of large amounts of natural gas (1.5 times more than clay tiles and 4 times more than sheet metal) to fire the kilns used to burn the limestone, clay shale and other materials.

Comparing these three options, clay tiles produced the lowest carbon footprint of 4.44 t of CO2e-. A large proportion of which is linked to the manufacturing and processing stage(s) (76%) followed by raw material acquisition (10%), transportation (10%), construction (2%), the usage (0.03%) and demolition and disposal stage (1%). Figure 1c shows that the combustion of natural gas in the furnace for clay tiles production contributed a large portion of the overall carbon footprint. In addition, the LCA analysis has also highlighted that transportation contributed 10% of the overall carbon footprint, which is significantly higher than the emissions from the transportation of sheet metal and concrete roofs. This is due to the heavy mass of clay tiles and long travel distance from the manufacturing factory to the construction site (tonnes × km travelled).

Recycling approach

Where local residual processing facilities exist, a recycling approach for roof material assessment is considered below, towards full environmental burden analysis of these materials.

Sheet metal has the potential to be 100% recovered, reused and recycled. This study, however, considered a more practicable local WA recycling rate of ~ 78% (Hyder Consulting 2009); subsequently this significant amount of carbon footprint offset can be attained. This LCA confirms that there is much potential for reducing GHG emissions from sheet-metal roofing (73%) through a recycling approach that reduces emissions from mining and processing of such energy-intensive metals. It is noted that the raw material acquisition stage of highly recyclable sheet-metal materials such as aluminium and zinc is an environmental ‘hot spot’ for this roof type, and so recycling significantly reduces the carbon footprint of a sheet-metal roof covering.

Concrete and clay tiles after their respective end of life are unlikely to be reused again as roof materials specifically, and therefore, recycling/crushing for alternative infrastructure applications can be considered. In the case of concrete tiles, 45% of the waste generated can be recycled/recovered as fine aggregates (e.g. sand) through concrete tile crushing, separation and grading for reuse as either sub-base file or as aggregate fines in recycled concrete. Consideration of this recycling strategy offsets emissions from the acquisition of the raw material stage. Once this recycling strategy has been considered, the overall GHG emissions from the use of 100 m2 of concrete tiles can be reduced to 8.44 t of CO2e- (i.e. by 10%).

To recycle clay tiles, the waste arising is crushed and graded into aggregates towards (localised) percentage replacement for virgin limestone in road construction. The amount of energy that could be avoided due to crushing clay tiles instead of limestone is only 3%. The material acquisition stage accounts for only 10% of the total GHG emission and so reducing (only) 3% of the total energy demand of raw material acquisition due to this recycling approach does not appear to decrease the overall GHG emission of clay tile roof.

For these three roof-covering materials, GHG emissions associated with the demolition and disposal decreased due to decreases in the amount of demolition wastes going to landfill. The emissions from the transportation of construction materials have been decreased for the sheet-metal roofs only, as recycled sheet metal is used for building roof application, which in turn would avoid the shipping of virgin materials from the eastern state. This is not the case for concrete and clay tiles, as recycled versions of these materials are used for different applications.

Embodied energy demand

When residual processing facilities are not available and a recycling approach is not considered, it is not the sheet-metal roof, but rather the concrete roof that is found to have the highest embodied energy demand (83 GJ), followed by sheet metal (58.6 GJ) and clay tiles (52.7 GJ). The embodied energy impact of a sheet-metal roof is not as great as its carbon footprint because GHG emissions are due to electricity generated from brown coal for aluminium production; on the other hand, as mentioned above the manufacturing energy for concrete is significantly higher than that for sheet metal and clay tiles. Concrete manufacturing involves the use of energy for crushing aggregates and mixing concrete constituents in batching plants. Since higher percentages of sheet metal can be recycled (than concrete tiles) with very high energy intensities (i.e. 240 MJ/kg for sheet metal, as opposed to cement for concrete of 7.5 MJ/kg, and clay of 0.05 MJ/kg), the use of recycling approaches are argued to be able to significantly reduce the embodied energy demand of sheet metal (i.e. 71%). Crushing and recycling of concrete and clay waste arisings reduces insignificant amounts of the embodied energy (concrete by 7% and clay tiles by 0.04%) due to the fact that most of the energy was consumed in the manufacturing stage of these materials.

Conclusions

This paper discusses how life cycle assessment (LCA) tool application can help inform the design process and specification choices of building materials in Western Australia, with particular regard to assessing key environmental impacts such as carbon footprint and embodied energy demand. This research compares three different roof alternatives which can be used as a guide for future study for a quick comparison, to guide choice of alternative specifications for (low carbon impact) roofing materials.

Where waste-processing facilities are not available and recycling strategies are not considered, this LCA analysis confirms that sheet metal is the most carbon-intensive roof-cover material, whilst concrete roof tiling has the highest embodied energy demand. Where residual processing facilities do exist on the other hand, recycling strategies are found to be most effective for sheet-metal roof covering, as 73% of GHG emissions and 71% of embodied energy demand can be reduced by recycling. However, a similar recycling of waste arisings approach makes less significant environmental savings (i.e. ≤ 10%) for concrete tile or clay tiles; albeit respective energy conservation during manufacture (of concrete and clay tiles) can reduce overall environmental impacts where waste recycling facilities exist.

The work here raises an awareness of the use of a simplified (off-the-peg software application) LCA approach, towards ongoing encouragement of designers and building materials specification stakeholders to incorporate environmental assessment into their decision-making process.