A Comparative Study on the Life Cycle Assessment of New Zealand Residential Buildings

: In New Zealand, housing is typically low density, with light timber framing being the dominant form of construction with more than 90% of the market. From 2020, as a result of the global pandemic, there was a shortage of timber in New Zealand, resulting in increased popularity for light steel framing, the main alternative to timber for housing. At the same time, the New Zealand government is committed to sustainability practises through legislation and frameworks, such as the reduction of whole-of-life carbon emissions for the building industry. New Zealand recently announced reducing its net greenhouse gas emissions by 50% within 2030. Life cycle assessment (LCA) is a technique for assessing the environmental aspects associated with a product over its life cycle. Despite the popularity of LCA in the construction industry of New Zealand, prior research results seem varied. There is no uniﬁed NZ context database to perform an LCA for buildings. Therefore, in this paper, a comprehensive study using LCA was conducted to quantify and compare the quantity of carbon emissions from two commonly designed houses in the Auckland region, one built from light timber and the other from light steel, both designed for a lifespan of 90 years. The cradle-to-cradle system boundary was used for the LCA. From the results of this study, it was found that the light steel house had 12.3% more carbon in total (including embodied and operational carbons) when compared to the light timber house, of which the manufacturing of two houses had a difference of 50.4% in terms of carbon emissions. However, when the end-of-life (EOL) analysis was included, it was found that the extra carbon could be offset due to the steel’s recyclability, reducing the amount of embodied carbon in the manufacturing process. Therefore, there was no signiﬁcant difference in carbon emissions between the light steel and the light timber building, with the difference being only 12.3%.


Introduction
As the population increases globally, the demand for housing will also naturally increase [1]. Current housing development predominantly relies on non-environmentfriendly building technologies [2]. Furthermore, building construction generally produces a large amount of carbon, which is the leading cause of climate change [3]. Nevertheless, the consequences of global climate change caused by carbon emissions include global warming, sea-level rises, the disruption of ecosystems, and consequences to human health [4]. The global construction industry accounts for about 40% of primary energy use and one-third of global carbon emissions [5].
In New Zealand, the building and construction sector contributes 20% of the nation's total carbon emissions [6]. In addition, the national gross carbon emissions increased by 26.4% between 1990 and 2019 (17,188.6 to 82,317.9 kt CO 2 eq), which accounts for an annual of concrete, steel, and timber building structures. They found that when using a fixed time horizon metric with a cradle-to-cradle system boundary, the timber materials had a lowest result (130 kg CO 2 eq/m 2 ) compared to steel (269 kg CO 2 eq/m 2 ) and concrete (348 kg CO 2 eq/m 2 ) when the sustainably managed forest was assumed to be applied [50]. Johnston et al. contrasted the CFS and hot-rolled steel (HRS) as the portal frames using a cradle-to-gate system boundary. The study revealed that using CFS as the primary framing material resulted in better embodied carbon (up to 30%) than using HRS for the 18 m and 24 m spans [51]. In addition, Ortiz-Rodríguez et al. conducted a comparative whole LCA study in the context of different geographical conditions. They showed that Colombian houses release fewer carbon emissions (65%) than the Mediterranean house, caused by differences in bioclimatic and electricity consumption behaviours in each location [35].
Looking at the LCA of buildings in New Zealand, Moradibistouni et al. examined carbon emissions from a residential building that used polymers. The study revealed that carbon emissions from the house were 20 kg CO 2 eq/m 2 /year [52]. Chandrakumar et al. analysed the whole carbon emissions from a house in New Zealand with a 90-year building lifetime. The research found the carbon emitted from the house was 16 kg CO 2 eq/m 2 /year. They compared the result with other nations' residential building climate impacts, which ranged between 10 and 90 kg CO 2 eq/m 2 /year [53].
Despite the popularity of LCA in construction industry, prior research results seem varied. There is no unified NZ context database to perform an LCA for buildings. Many factors are analysed which influence the various buildings' carbon emissions data, such as the system boundary, geographical horizon, time horizon, methodology or tool, and materials used. In New Zealand, between 2010 and 2019, timber framing became the most popular material (90% of market share) compared with steel and concrete [10]. The use of each of these materials will also result in a different process or work method. Hence, the energy use and carbon emissions will differ. In addition, when analysing the carbon emissions from a building, the availability of an Environmental Product Declaration (EPD) is crucial. The EPD is used as the basis of carbon emission calculations from materials based on the product manufacturing processes to the EOL scenarios. However, the published EPDs across many countries seem to have varied values, which results in variations of total carbon emissions when they are compared. Rasmussen et al. conducted a comprehensive review of the EPDs of structural woods in which they analysed 81 EPDs across several countries. They found that the investigated EPDs had varied values (up to 90% uncertainty for GWP), which were caused by different specificities such as manufacturer, facility, co-production, time of data representation, and supply chain [54]. Janjua et al. investigated the effect of service life on a building's environmental performance. They discovered that the service life of a building and its components had a direct link with the environmental performance, namely the replacement intervals, and that it was recognised as the third major contributing factor to the overall environmental effects after use and pre-use [55]. Referring to previous building LCA studies [46][47][48][49][50][51][52][53], it seemed that the lifespan of buildings has ranged between 50 and 100 years, resulting in varied assessment outcomes. The condition leads to a knowledge gap in implementing environmental assessment of building products and results in varied assessment results, primarily caused by the various scopes of assessment and the unidentical design of the buildings. Therefore, the total carbon emissions from residential buildings cannot be generalised and applied to all types of houses.
Hence, with a range of previous LCA results for building products, this study examined a house in New Zealand using the local embodied carbon data and the national electricity consumption scenario to determine the carbon emissions from NZ houses in the current condition. The assessment was conducted in the design phase, which was specific to the technical design, where the carbon footprints of the building were quantified based on the designed structure and generic material data. The main objective of this research was to compare the carbon emissions of two types of houses, one built from light timber and the other from light steel. The comparison between the two houses was based on the carbon produced during the whole-building life cycle and the potential carbon offset from materials used in the buildings.

Methodology
An LCA was performed to investigate the environmental impact in terms of carbon emissions from two New Zealand residential buildings, one made from light timber and the other from CFS. The design of the two houses is the most common design used in New Zealand. The carbon emissions were computed at each step of the life cycle of the materials (including production to the EOL stage), and a comparison of the carbon emissions from the two buildings was calculated. The analysis included both the embodied and operational carbon emissions, and considered the carbon offset from the materials used in the buildings. The LCA was performed using the NZ-based LCA tool LCAQuick V3.4.4 [56], which was developed by BRANZ. Figure 1 illustrates the methodology used in this study. The method for performing the LCA followed the LCA framework stated in the ISO 14040 series [57]. According to ISO 14040:2006, there are four steps in conducting LCA: (1) goal and scope definition, (2) inventory analysis (LCI), (3) impact assessment (LCIA), and (4) interpretation of the results. In the first step, the goal and scope were identified to achieve the purpose of the study. Some elements were included in the scope of the study, such as system boundary and functional units, which are crucial in the LCA. The second step was to set up the LCI, where the data were collected. The primary data required for this LCA was the quantity of the building materials. Afterward, the LCIA was performed, in which the impact category and assessment tool were selected, and the LCA calculation was performed. Finally, the interpretation of the results was undertaken to present the difference between the carbon emissions from the two investigated houses, one made from CFS and the other from light timber. In order to check the validity of the results, they were compared with previous studies. was to compare the carbon emissions of two types of houses, one built from light timber and the other from light steel. The comparison between the two houses was based on the carbon produced during the whole-building life cycle and the potential carbon offset from materials used in the buildings.

Methodology
An LCA was performed to investigate the environmental impact in terms of carbon emissions from two New Zealand residential buildings, one made from light timber and the other from CFS. The design of the two houses is the most common design used in New Zealand. The carbon emissions were computed at each step of the life cycle of the materials (including production to the EOL stage), and a comparison of the carbon emissions from the two buildings was calculated. The analysis included both the embodied and operational carbon emissions, and considered the carbon offset from the materials used in the buildings. The LCA was performed using the NZ-based LCA tool LCAQuick V3.4.4 [56], which was developed by BRANZ. Figure 1 illustrates the methodology used in this study. The method for performing the LCA followed the LCA framework stated in the ISO 14040 series [57]. According to ISO 14040:2006, there are four steps in conducting LCA: (1) goal and scope definition, (2) inventory analysis (LCI), (3) impact assessment (LCIA), and (4) interpretation of the results. In the first step, the goal and scope were identified to achieve the purpose of the study. Some elements were included in the scope of the study, such as system boundary and functional units, which are crucial in the LCA. The second step was to set up the LCI, where the data were collected. The primary data required for this LCA was the quantity of the building materials. Afterward, the LCIA was performed, in which the impact category and assessment tool were selected, and the LCA calculation was performed. Finally, the interpretation of the results was undertaken to present the difference between the carbon emissions from the two investigated houses, one made from CFS and the other from light timber. In order to check the validity of the results, they were compared with previous studies.

Goal of the LCA
As mentioned previously in the introduction section, this LCA aimed to study the impact of carbon emissions from light timber and light steel houses. The need for buildings' LCA has been reflected in the past, which was assessed herein. The comparison of carbon The system boundary used in this LCA was cradle-to-cradle. Therefore, the study examined the carbon produced by the houses from the material production stage to the EOL stage and considered the potential carbon offset. Figure 2 shows the building life stages used in the LCA study, which adopted the standard building life cycle [58,59].

Goal of the LCA
As mentioned previously in the introduction section, this LCA aimed to study the impact of carbon emissions from light timber and light steel houses. The need for buildings' LCA has been reflected in the past, which was assessed herein. The comparison of carbon emissions from the selected two houses can provide some basic understanding of decision-making at the governmental level.

System Boundaries
The system boundary used in this LCA was cradle-to-cradle. Therefore, the study examined the carbon produced by the houses from the material production stage to the EOL stage and considered the potential carbon offset. Figure 2 shows the building life stages used in the LCA study, which adopted the standard building life cycle [58,59]. Furthermore, the selected geographical boundary in this study was the Auckland region, New Zealand. Based on the BRANZ database for building service life in NZ [59], 90 years of building lifetime was applied in this study.

Functional Unit
The functional unit in this LCA study was the annual carbon emissions per gross floor area (GFA), kg CO2 eq/m 2 /year. Therefore, by calculating the carbon emission results per GFA, it was possible to compare the impact of light timber and light steel construction in terms of their carbon emissions.

Assumptions
Several assumptions were made in the LCA study. Due to the unavailability of the data for staircase material quantities, they were excluded from the LCA. The quantity of floor coverings (e.g., floor tiles or carpet) and paint were not included in this study because they were assumed to be identical and did not result in any significant changes in terms of impacts [49]. Both houses' annual electricity consumption and water demand scenarios were assumed to be steady over their lifetimes. Therefore, the values were not changed.

Limitations
This study was New Zealand-specific, where the local carbon coefficient and energy consumption were applied. The result of the study cannot be applied directly to other buildings outside the scope of the study because the result is highly dependent on the geographical regions, time horizon, and building material types. However, this study can provide an understanding of the potential building materials available in the NZ's housing market when choosing the most environmentally friendly material. It also shows the relationship between the material volume or weight of the two houses (light timber and Benefit and loads beyond the system boundary  Furthermore, the selected geographical boundary in this study was the Auckland region, New Zealand. Based on the BRANZ database for building service life in NZ [59], 90 years of building lifetime was applied in this study.

Functional Unit
The functional unit in this LCA study was the annual carbon emissions per gross floor area (GFA), kg CO 2 eq/m 2 /year. Therefore, by calculating the carbon emission results per GFA, it was possible to compare the impact of light timber and light steel construction in terms of their carbon emissions.

Assumptions
Several assumptions were made in the LCA study. Due to the unavailability of the data for staircase material quantities, they were excluded from the LCA. The quantity of floor coverings (e.g., floor tiles or carpet) and paint were not included in this study because they were assumed to be identical and did not result in any significant changes in terms of impacts [49]. Both houses' annual electricity consumption and water demand scenarios were assumed to be steady over their lifetimes. Therefore, the values were not changed.

Limitations
This study was New Zealand-specific, where the local carbon coefficient and energy consumption were applied. The result of the study cannot be applied directly to other buildings outside the scope of the study because the result is highly dependent on the geographical regions, time horizon, and building material types. However, this study can provide an understanding of the potential building materials available in the NZ's housing market when choosing the most environmentally friendly material. It also shows the relationship between the material volume or weight of the two houses (light timber and light steel houses) and the carbon emissions from the production stages of the materials to their EOL stages, which can be useful in the planning phase of future housing projects to use more sustainable materials for building houses in New Zealand.

Housings Material Data
The material quantities and architectural drawings of the two investigated buildings were collected from the local contractors and manufacturers of the building materials. The light timber house plan (House A) is illustrated in Figure 3, while the light steel house plan (House B) is shown in Figure 4.
Buildings 2022, 11, x FOR PEER REVIEW 6 of 16 light steel houses) and the carbon emissions from the production stages of the materials to their EOL stages, which can be useful in the planning phase of future housing projects to use more sustainable materials for building houses in New Zealand.

Housings Material Data
The material quantities and architectural drawings of the two investigated buildings were collected from the local contractors and manufacturers of the building materials. The light timber house plan (House A) is illustrated in Figure 3, while the light steel house plan (House B) is shown in Figure 4.  Buildings 2022, 11, x FOR PEER REVIEW 6 of 16 light steel houses) and the carbon emissions from the production stages of the materials to their EOL stages, which can be useful in the planning phase of future housing projects to use more sustainable materials for building houses in New Zealand.

Housings Material Data
The material quantities and architectural drawings of the two investigated buildings were collected from the local contractors and manufacturers of the building materials. The light timber house plan (House A) is illustrated in Figure 3, while the light steel house plan (House B) is shown in Figure 4.  According to the floor plans, both houses were two-story buildings with different floor areas on each level. Table 1 shows the gross floor area (GFA) of the two investigated buildings. Furthermore, when conducting the LCA, the material quantities were input in the LCAQuick software [56]. The type of material and volume were calculated from the data sources (i.e., architectural plan and elevation drawings and material take-off sheets). Table 2 shows the material quantities obtained from both the buildings studied herein.

Operational Energy Data
The operational energy consumptions in both houses were calculated using the recommendations in the LCAQuick database [56] and the guidelines in the 2018 Electricity New Zealand report. House A was assumed to be occupied by three people, while House B had five occupants. The annual electricity consumption for House A was found to be 7000 kWh/year, and this value was calculated based on the per capita electricity demand in New Zealand [60]. For House B, the annual electricity demand was 9618 kWh/year, referring to the BRANZ datasheet available in the LCAQuick [56] for five occupants. The annual electricity consumption for both houses is illustrated in Table 3. Plug loads consumed the most extensive electricity for two houses, with approximately 50% of the total electricity consumption of the houses. In this study, the operational water demand referred to the annual water consumption as per the recommendations given in LCAQuick [56], which is 67.60 kL/person/year. Therefore, due to the different numbers of occupants in two houses, the houses had different water demands. Table 4 shows the annual water demand for both houses during the entire lifespan of the buildings.

Impact Assessment
The LCIA can be classified into two methods: midpoints and endpoints (damagedoriented) [61,62]. Midpoints are known to be ties in the cause-effect chain (an environmental mechanism) of the impact group (e.g., GWP and acidification) [63]. In contrast, endpoints consider the end of the cause-effect chain, such as human health and the ecosystem [64]. Therefore, this study used the midpoints approach to analyse a single environmental problem directly influenced by carbon emissions, namely the GWP. The calculation of the carbon emissions was performed using the NZ-based LCA tool called LCAQuick V3.4.4 [56], which enables the midpoints analysis and contains the local emission database.

Assessment of Results
House A produced 13.72 kg CO 2 eq/m 2 /year for the whole building's life cycle, in which the benefits of recycling, recovery, and reuse of materials were considered. By adding Module D in the LCA process, the materials used in House A received a carbon offset of −0.74 kg CO 2 eq/m 2 /year. Furthermore, House B produced 15.41 kg CO 2 eq/m 2 /year in total during its whole life cycle. The materials used in House B had −1.53 kg CO 2 eq/m 2 /year of carbon offsets after being processed in the EOL stage. Hence, House A produced fewer carbon emissions than House B. Table 5 shows the LCA results of the two investigated houses.

Comparison of the Results Obtained from the Two Houses
The comparison of carbon emissions between the two investigated houses having different materials showed that the selected materials massively contributed to the carbon emissions from the two houses. Figure 5 illustrates the comparative LCA results of two houses in the context of the GWP per building during their life cycles.

Comparison of the Results Obtained from the Two Houses
The comparison of carbon emissions between the two investigated houses having different materials showed that the selected materials massively contributed to the carbon emissions from the two houses. Figure 5 illustrates the comparative LCA results of two houses in the context of the GWP per building during their life cycles. From Figure 5, it can be seen that the difference of carbon emissions obtained from the two houses was 1.69 kg CO2 eq/m 2 /year. House B emitted 12.32% higher carbon than House A. In comparing the carbon emissions obtained from both buildings, during the production stages of the building materials, it was found that House A emitted around 50.43% fewer carbon when compared to the carbon emissions of House B. Cole and Kernan [65] found that timber frames' extraction and manufacturing processes have fewer energy usage when compared to other building materials such as concrete and steel. A comparative study between a mid-rise mass timber building and a steel building also indicated that the timber building produced between 28.99% and 34.08% of fewer embodied carbon (A1-A5) than the steel building [66], which confirms that the results obtained in this study are in line with the literature findings. In addition, carbon emissions from the two buildings during their remaining life stages were shown to be similar. The difference in carbon emissions obtained from the construction stages (A4-A5) of both the houses was 0.18 kg CO2 eq/m 2 /year, where House B released 37% fewer carbon than House A. This  From Figure 5, it can be seen that the difference of carbon emissions obtained from the two houses was 1.69 kg CO 2 eq/m 2 /year. House B emitted 12.32% higher carbon than House A. In comparing the carbon emissions obtained from both buildings, during the production stages of the building materials, it was found that House A emitted around 50.43% fewer carbon when compared to the carbon emissions of House B. Cole and Kernan [65] found that timber frames' extraction and manufacturing processes have fewer energy usage when compared to other building materials such as concrete and steel. A comparative study between a mid-rise mass timber building and a steel building also indicated that the timber building produced between 28.99% and 34.08% of fewer embodied carbon (A1-A5) than the steel building [66], which confirms that the results obtained in this study are in line with the literature findings. In addition, carbon emissions from the two buildings during their remaining life stages were shown to be similar. The difference in carbon emissions obtained from the construction stages (A4-A5) of both the houses was 0.18 kg CO 2 eq/m 2 /year, where House B released 37% fewer carbon than House A. This difference may be attributed to the fewer waste materials generated from light steel than light timber during their construction processes.
The maintenance (B2) and replacement (B4) stages of House B produced more carbon emissions than House A, with 27% higher carbon emissions, while the operational energy usage (B6) of both houses was found to be very similar, with the difference being only 1%. The operational water usage (B7) of House A was 17% fewer than that of House B. This is because of the different numbers of occupants present in the two evaluated houses: House A had only three occupants, while House B had five occupants. House B released 42% fewer carbon than House A during its EOL stage (C1-C4).
The light steel house had almost double the carbon offset when compared with the light timber house. This was because of the higher recyclability of light steel than light timber at their EOL stages. According to the BRANZ EOL module [59] (which is specific to New Zealand applications), the recyclability of steel sheet is around 65% higher than that of timber. Hence, the GWP of two investigated houses was highly affected by the choice of material and the quantity of those materials assigned to the investigated buildings.

Life Cycle Analysis of the Two Houses
From the results of the LCA, it was found that different stages of the buildings' life cycle had different impacts in terms of carbon emissions. Figure 6 shows the buildings' life stages and its corresponding carbon emissions for both House A and House B. difference may be attributed to the fewer waste materials generated from light steel than light timber during their construction processes. The maintenance (B2) and replacement (B4) stages of House B produced more carbon emissions than House A, with 27% higher carbon emissions, while the operational energy usage (B6) of both houses was found to be very similar, with the difference being only 1%. The operational water usage (B7) of House A was 17% fewer than that of House B. This is because of the different numbers of occupants present in the two evaluated houses: House A had only three occupants, while House B had five occupants. House B released 42% fewer carbon than House A during its EOL stage (C1-C4).
The light steel house had almost double the carbon offset when compared with the light timber house. This was because of the higher recyclability of light steel than light timber at their EOL stages. According to the BRANZ EOL module [59] (which is specific to New Zealand applications), the recyclability of steel sheet is around 65% higher than that of timber. Hence, the GWP of two investigated houses was highly affected by the choice of material and the quantity of those materials assigned to the investigated buildings.

Life Cycle Analysis of the Two Houses
From the results of the LCA, it was found that different stages of the buildings' life cycle had different impacts in terms of carbon emissions. Figure 6 shows the buildings' life stages and its corresponding carbon emissions for both House A and House B. From Figure 6, both houses emitted a significant amount of embodied and operational carbon. The operational energy usage (B6 module) was the most significant carbon contributor during the whole building's life cycle. The B6 module of House A had a 57.67% carbon contribution, whereas for House B, this contribution was 49.53%. The production stages (A1-A3) contributed the second largest carbon emissions, with 15.98% of carbon for House A and 27.49% for House B.
In addition, it was identified that the time horizon of the building's life stages could influence the amount of carbon produced. By having a 90-year service life for the two houses, the operational stage was the primary contributor to the total carbon emissions when compared to the other stages. The early phases of the building (production to construction) had less completion time, while the operational stage of the building was dependent on the building's life span. Hence, the operational stage in this study contributed to the majority of the carbon emissions for both houses. This finding was also in line with Petrovic   From Figure 6, both houses emitted a significant amount of embodied and operational carbon. The operational energy usage (B6 module) was the most significant carbon contributor during the whole building's life cycle. The B6 module of House A had a 57.67% carbon contribution, whereas for House B, this contribution was 49.53%. The production stages (A1-A3) contributed the second largest carbon emissions, with 15.98% of carbon for House A and 27.49% for House B.
In addition, it was identified that the time horizon of the building's life stages could influence the amount of carbon produced. By having a 90-year service life for the two houses, the operational stage was the primary contributor to the total carbon emissions when compared to the other stages. The early phases of the building (production to construction) had less completion time, while the operational stage of the building was dependent on the building's life span. Hence, the operational stage in this study contributed to the majority of the carbon emissions for both houses. This finding was also in line with Petrovic et al. (2019), who conducted an LCA of a wooden single-family house. With a 100-year lifespan, the operational stage of the building had an enormous carbon contribution, with about 64% of the total carbon emissions coming from the operation phase of the building [47].

A Comparative Analysis of Carbon Emissions Obtained from Different Building Materials
House A used timber as the main building material for structural elements (e.g., wall framing and roof trusses), whereas House B used CFS. Both buildings also had other materials such as concrete, aluminium, and fibre cement used in other primary and secondary structural and non-structural members. Therefore, Figures 7 and 8 show the amount of carbon emissions obtained from seven different building materials that were installed in both houses. 100-year lifespan, the operational stage of the building had an enormous carbon contribution, with about 64% of the total carbon emissions coming from the operation phase of the building [47].

A Comparative Analysis of Carbon Emissions Obtained from Different Building Materials
House A used timber as the main building material for structural elements (e.g., wall framing and roof trusses), whereas House B used CFS. Both buildings also had other materials such as concrete, aluminium, and fibre cement used in other primary and secondary structural and non-structural members. Therefore, Figures 7 and 8 show the amount of carbon emissions obtained from seven different building materials that were installed in both houses.  Figure 7 shows the amount of carbon emissions from the building materials used in House A. Fibre cement sheet was the highest contributor to the total carbon emissions of the house, with 1.19 kg CO2 eq/m 2 /year, or 30.88% of the total carbon emissions. The maintenance and replacement of fibre cement seemed to have the biggest carbon emissions among the other life stages. It was caused by the service life of fibre cement sheets, which need to be replaced after 50 years [59]. As a result, with the buildings' service life of 90 years, the fibre cement sheets needed to be replaced at least once for House A. Consequently, additional carbon was released from House A during the production and construction processes of the new fibre cement sheets. The second biggest carbon contributor was timber softwood, with 0.86 kg CO2 eq/m 2 /year, or 22.24% of the total carbon emissions coming from the softwood. The timber material absorbed a significant amount of CO2 throughout the growth of the trees from where the timber was cut [67]. In addition, all timber used in this study was assumed to come from non-sustainable forest management practices. Sustainable forest management practises are defined as forestry activities that involve the planting of trees to replace those that have been cut down for exploitation [68]. Hence, using timber with non-sustainable practises increased the amount of carbon emissions produced by House A.
Furthermore, Figure 8 shows the amount of carbon emissions from the building materials used in House B. CFS had the most significant carbon footprint, with 1.63 kg CO2 eq/m 2 /year, or 31.50% of the total carbon emissions. According to the Environmental Product Declaration (EPD) published by the Steel Recycling Institute [69], the CFS production processes consist of several activities, such as the manufacturing of hot-dip galvanised steel, which produce a significant amount of carbon. Therefore, the large quantity of CFS used in House B resulted in the considerable contribution of this material to the total result. However, this study found that CFS had the highest carbon offset after the EOL stage. The high carbon offset was due to the very high recovery rates of steel material for recycling [67]. Afterward, the fibre cement sheet was the second highest carbon contributor, with 0.97 kg CO2 eq/m 2 /year, or 18.77% of the total carbon emissions.

Comparison of LCA Results Obtained from the Current Study with Other Studies Available in the Literature
A comparison between the results of this study and the previous studies reported in the literature was conducted to check the validity of the LCA results. This study identified that House A released 13.72 kg CO2 eq/m 2 /year and House B released 15.41 kg CO2 eq/m 2 /year of total carbon emissions for the whole life cycle of the buildings. In order to check the validity of the LCA results, the findings of Chandrakumar et al. [53] were used. Their research found that the investigated house emitted 16 kg CO2 eq/m 2 /year of embodied and operational carbon, and their results were acceptable to the climate impact of residential buildings from around the world, where the total carbon emissions range between 10-90 kg CO2 eq/m 2 /year [53]. By referring to this previous research, which had an identical geographical condition to the houses studied in the current paper, it could be seen that the results of this study have a close match with the previously reported results in terms of total carbon emissions.  Figure 7 shows the amount of carbon emissions from the building materials used in House A. Fibre cement sheet was the highest contributor to the total carbon emissions of the house, with 1.19 kg CO 2 eq/m 2 /year, or 30.88% of the total carbon emissions. The maintenance and replacement of fibre cement seemed to have the biggest carbon emissions among the other life stages. It was caused by the service life of fibre cement sheets, which need to be replaced after 50 years [59]. As a result, with the buildings' service life of 90 years, the fibre cement sheets needed to be replaced at least once for House A. Consequently, additional carbon was released from House A during the production and construction processes of the new fibre cement sheets. The second biggest carbon contributor was timber softwood, with 0.86 kg CO 2 eq/m 2 /year, or 22.24% of the total carbon emissions coming from the softwood. The timber material absorbed a significant amount of CO 2 throughout the growth of the trees from where the timber was cut [67]. In addition, all timber used in this study was assumed to come from non-sustainable forest management practices. Sustainable forest management practises are defined as forestry activities that involve the planting of trees to replace those that have been cut down for exploitation [68]. Hence, using timber with non-sustainable practises increased the amount of carbon emissions produced by House A.
Furthermore, Figure 8 shows the amount of carbon emissions from the building materials used in House B. CFS had the most significant carbon footprint, with 1.63 kg CO 2 eq/m 2 /year, or 31.50% of the total carbon emissions. According to the Environmental Product Declaration (EPD) published by the Steel Recycling Institute [69], the CFS production processes consist of several activities, such as the manufacturing of hot-dip galvanised steel, which produce a significant amount of carbon. Therefore, the large quantity of CFS used in House B resulted in the considerable contribution of this material to the total result. However, this study found that CFS had the highest carbon offset after the EOL stage. The high carbon offset was due to the very high recovery rates of steel material for recycling [67]. Afterward, the fibre cement sheet was the second highest carbon contributor, with 0.97 kg CO 2 eq/m 2 /year, or 18.77% of the total carbon emissions.

Comparison of LCA Results Obtained from the Current Study with Other Studies Available in the Literature
A comparison between the results of this study and the previous studies reported in the literature was conducted to check the validity of the LCA results. This study identified that House A released 13.72 kg CO 2 eq/m 2 /year and House B released 15.41 kg CO 2 eq/m 2 /year of total carbon emissions for the whole life cycle of the buildings. In order to check the validity of the LCA results, the findings of Chandrakumar et al. [53] were used. Their research found that the investigated house emitted 16 kg CO 2 eq/m 2 /year of embodied and operational carbon, and their results were acceptable to the climate impact of residential buildings from around the world, where the total carbon emissions range between 10-90 kg CO 2 eq/m 2 /year [53]. By referring to this previous research, which had an identical geographical condition to the houses studied in the current paper, it could be seen that the results of this study have a close match with the previously reported results in terms of total carbon emissions.
However, the LCA performed on different buildings showed a wide range of results. This variation can be justified by the fact that the design for each of these buildings is unique [45]. Other studies also revealed that errors in the data collection and quality, as well as the geographical location, can impact the findings [70]. The scope and methodology or LCA tool, while performing an LCA of any building, can highly influence the outcomes. Therefore, these factors should not be used to compare the results of other buildings outside the defined scope.

Recommendations
Some suggestions based on the outcome of this research that can improve the energy efficiency of houses to be built in the future are listed below:

•
The operational stages of the investigated houses contributed the maximum amount of carbon from both buildings. Accordingly, the electricity consumption of houses needs to be minimised, i.e., by controlling the use of plug loads and hot water electricity use, as these can lead to significant energy consumption in houses. • New Zealand's electricity comes mainly from renewable sources, while 17.5% of the total electricity supply in 2019 was from burning fossil fuels [7]. More houses should incorporate photovoltaic (PV) panels to reduce carbon emissions from burning fossil fuels.

•
A strategy to reduce carbon emissions from the production stages is needed due to their significance as being the second largest contributor of carbon emissions. The material selection process also plays a crucial role in achieving better energy efficiency of the houses. Hence, selecting more sustainable materials is suggested in the design phase to reduce the total carbon emissions from the houses.

Conclusions
Carbon emissions from buildings cannot be generalised due to many factors that influence the values, such as the material types, geographical horizon, and time horizon. This study used the LCA to evaluate and compare two different two-story houses (one built from light timber and the other one was constructed from light steel) in Auckland, New Zealand. In this LCA, cradle-to-cradle system boundaries were used. From the results of this study, the following conclusions were drawn:

•
The LCA results confirmed that the light timber house emitted 13.72 kg CO 2 eq/m 2 /year, while the light steel house emitted 15.41 kg CO 2 eq/m 2 /year of total carbon. The two houses had comparable carbon emissions, which the light steel house had around 12% higher total carbon emissions (embodied and operational, including carbon offset) than the light timber house. Therefore, there was no significant difference in carbon emissions between those two houses, considering the carbon offset from the building materials after their EOL stages.

•
The study identified a large difference in carbon emissions in the production stages between the two investigated houses. The light timber house had around 50% fewer carbon emissions than the light steel house during its material production stages. However, the carbon offset from the light steel house was twice as high as the light timber house. The higher carbon offset was caused by the higher recyclability of steel than timber. The recyclability of steel used in House B was around 65% higher than the recyclability of the timber house (House A).

•
The operational energy emissions of the two houses became the most significant carbon contributors (49.5-57.6%) during the whole building's life cycle, followed by the production stages (A1-A3). The reason behind such high operational carbon emissions from both houses is that the design life of the houses was 90 years, requiring operational energy for such a prolonged lifespan when compared with the other stages of the building's life cycle. Therefore, controlling human behaviour in terms of electricity consumption and promoting the use of renewable energy can decrease the amount of carbon emissions from the operational stages of houses.
According to the research outcomes, the LCA was a useful approach for assessing the carbon emissions from the buildings and deriving a reliable comparison of the options based on their environmental performance. Using this LCA approach, the study can conclude that light steel can substitute light timber as the housing material to address the current national timber shortage of New Zealand due to the slight difference in carbon emissions produced by these materials.