Carbon footprint assessment of a wood multi-residential building considering biogenic carbon

Journal of Cleaner Production 404 (2023) 136834 Available online 17 March 2023 0959-6526/© 2023 Published by Elsevier Ltd. Carbon footprint assessment of a wood multi-residential building considering biogenic carbon Claudiane M. Ouellet-Plamondon a , * , Livia Ramseier b , Maria Balouktsi c , i , Laetitia Delem d , Greg Foliente e , Nicolas Francart f , i , Antonio Garcia-Martinez g , Endrit Hoxha h , i , Thomas Lützkendorf c , Freja Nygaard Rasmussen i , Bruno Peuportier j , Jarred Butler k , Harpa Birgisdottir i , David Dowdell k , Manish Kumar Dixit l , Vanessa Gomes m , Maristela Gomes da Silva n , Juan Carlos G´ omez de C´ ozar g , Marianne Kjendseth Wiik o , Carmen Llatas g , Ricardo Mateus p , Lizzie M. Pulgrossi m , Martin R¨ ock h , q , Marcella Ruschi Mendes Saade h , Alexander Passer h , Daniel Satola r , Seongwon Seo e , Bernardette Soust Verdaguer g , h , Jakub Veselka s , Martin Volf s , Xiaojin Zhang t , u , Rolf Frischknecht b a ´ Ecole de Technologie Sup´ erieure, Montreal, Canada b Treeze Ltd, CH- 8610, Uster, Switzerland c Karlsruhe Institute of Technology, 76131, Karlsruhe, Germany d Belgian Building Research Institute, 1000, Bruxelles, Belgium e The University of Melbourne, Parkville, VIC 3010, Australia f KTH - Royal Institute of Technology, 114 28, Stockholm, Sweden g Instituto Universitario de Arquitectura y Ciencias de la Construcci´ on, Escuela T´ ecnica Superior de Arquitectura, Universidad de Sevilla, 41004, Sevilla, Spain h Graz University of Technology, 8010, Graz, Austria i Aalborg Universitet København, Department of the Built Environment, 2450, Denmark j Mines Paris, PSL Research University, CES, 75272, Paris, France k BRANZ, Judgeford, 5381, New Zealand l Texas A&M University, College Station, TX 77843, United States m University of Campinas, 13083-852, Campinas, Brazil n Federal University of Espirito Santo, 29075-910, Vitoria, Brazil o SINTEF, NO-7465, Trondheim, Norway p University of Minho, 4710-057, Braga, Portugal q KU Leuven, 3001, Leuven, Belgium r Norwegian University of Science and Technology, 7491, Trondheim, Norway s University Centre for Energy Efficient Buildings, Czech Technical University in Prague, 273 43, Bustehrad, Czech Republic t Technology Assessment Group, Laboratory for Energy Systems Analysis, Paul Scherrer Institute, Forschungsstrasse 111, 5232, Villigen PSI, Switzerland u ETH Zürich, Institute of Construction and Infrastructure Management (IBI), Chair of Sustainable Construction, Stefano-Franscini-Platz 5, 8093, Zürich, Switzerland ARTICLE INFO Handling Editor: Zhifu Mi Keywords: Biogenic carbon Life cycle assessment Building Construction Wood products ABSTRACT Wood and other bio-based building materials are often perceived as a good choice from a climate mitigation perspective. This article compares the life cycle assessment of the same multi-residential building from the perspective of 16 countries participating in the international project Annex 72 of the International Energy Agency to determine the effects of different datasets and methods of accounting for biogenic carbon in wood construction. Three assessment methods are herein considered: two recognized in the standards (the so-called 0/ 0 method and −1/+1 method) and a variation of the latter (−1/+1* method) used in Australia, Canada, France, and New Zealand. The 0/0 method considers neither fixation in the production stage nor releases of biogenic carbon at the end of a wood product’s life. In contrast, the −1/+1 method accounts for the fixation of biogenic carbon in the production stage and its release in the end-of-life stage, irrespective of the disposal scenario (recycling, incineration or landfill). The −1/+1 method assumes that landfills offer only a temporary * Corresponding author. E-mail address: [email protected] (C.M. Ouellet-Plamondon). Contents lists available at ScienceDirect Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro https://doi.org/10.1016/j.jclepro.2023.136834 Received 5 October 2021; Received in revised form 21 July 2022; Accepted 16 March 2023 Journal of Cleaner Production 404 (2023) 136834 2 sequestration of carbon. In the −1/+1* variation, landfills and recycling are considered a partly permanent sequestration of biogenic carbon and thus fewer emissions are accounted for in the end-of-life stage. We examine the variability of the calculated life cycle-based greenhouse gas emissions calculated for a case study building by each participating country, within the same assessment method and across the methods. The results vary substantially. The main reasons for deviations are whether or not landfills and recycling are considered a partly permanent sequestration of biogenic carbon and a mismatch in the biogenic carbon balance. Our findings support the need for further research and to develop practical guidelines to harmonize life cycle assessment methods of buildings with bio-based materials. 1. Introduction The search for solutions to reduce resource use and the impact of human activities on the local and global environment reached the international scale. Since the operation of buildings accounts for approximately 30% and the manufacture of construction products for approximately 10% of global greenhouse gas (GHG) emissions (UN Environment and International Energy Agency, 2017), the great potential for reducing the impacts in the construction sector becomes clear. At the moment, whether and to what extent the increased use of bio-based products in constructing new buildings and refurbishing existing ones can contribute to resource conservation and environmental relief is increasingly discussed (Ramage et al., 2017). Furthermore, methods and data for objectively assessing wood and other bio-based building materials/construction products, as well as buildings with a predominant use of bio-based products are necessary (Andersen et al., 2021; Takano et al., 2015). This results in a need to check the status of methodological discussions concerning biogenic carbon in building life cycle assessment (LCA). Biogenic carbon is defined as the carbon that is extracted from the atmosphere during biomass growth and is released back to the atmosphere later due to combustion of the biomass and decomposition (Stamford, 2020). Furthermore, the consequences of method- and data-related issues for the environmental performance assessment of wood buildings must be discussed, to develop consensus-based and scientifically recognized assessment procedures in the context of sustainability. 1.1. Context and background The world is in a climate emergency (World Green Building Council, 2019). The latest intergovernmental report released by the IPCC (AR6) states that with the current level of government commitment, the planet is on course for a disastrous 3 ◦C temperature rise (IPCC et al., 2021). For sustainable development of the future, it was agreed at COP21 in Paris that global warming should be limited to well below 2 ◦C and preferably to below 1.5 ◦C. To limit the increase in global temperatures by 1.5 ◦C, mankind must drastically reduce carbon emissions until the net-zero status is reached by 2050 latest (IPCC et al., 2018). The design, construction and operation of buildings and other constructed assets have a major impact on the environment with the reduction of GHG emissions across the life cycle of buildings being especially relevant for effectively mitigating climate change (R¨ ock et al., 2020). Many governments are aware of the extent of the environmental impacts caused by the construction and use of buildings, but also acknowledge that the construction and use of buildings are an indispensable basis for economic and social development. To cope with their responsibility in creating the basis and prerequisites for more sustainable national building stocks (R¨ ock et al., 2021), national sustainability strategies have been developed and implemented (Government of Canada, 2019; Ministry of the Interior and Housing, 2021; The Federal Government, 2016), legislative initiatives (Government of Spain, 2021; Ministry of the Interior and Housing, 2021; Swedish National Board of Housing Building and Planning, 2020), funding programs (Association for the Conservation of Energy, 2013), and exemplary public buildings have been proposed, implemented, and built (World Green Building Council, 2021). In the interests of present and future generations, such activities serve to protect the climate, preserve ecosystems, and conserve natural resources. Most of them have long focused on the building operational energy, by reducing energy demand and increasing the renewable energy share. More recently, the impacts embodied in the manufacturing of building products, as well as the maintenance, deconstruction and recycling of buildings has gained attention not only in the academic literature (Birgisdottir et al., 2017), but also in the political circles (ISO, 2017b; Kuittinen and H¨ akkinen, 2020; Minist` ere de la transition ´ ecologique, 2021; Ministry of Business Innovation & Employment, 2020; Swedish National Board of Housing Building and Planning, 2020). As the emissions due to energy consumption decrease, construction products become greater contributors to the building environmental impacts. Recently, the focus has been placed on the role of bio-based construction materials and products, and especially wood, as a contribution to climate change mitigation (Himes and Busby, 2020). With the development of mass timber technology and expertise in recent decades, which made possible to build mid- and high-rise buildings with wood, many view wood materials as a substitute for the traditional construction materials, such as steel and concrete used in building frames (Chen et al., 2020; Green and Karsh, 2012; Harte, 2017). Therefore, increased use of wood and bio-based materials is often seen as an important contribution to resource conservation and climate protection. Furthermore, these building materials and construction methods can store carbon, i.e. they can serve as temporary carbon reservoirs offering time to develop effective technologies to capture and permanently sequester biogenic carbon contained in buildings (Churkina et al., 2020). 1.2. Wooden products and buildings as part of political strategies and actions When analyzing political actions concerning the use of bio-based materials, trends can be observed. Public authorities at varied levels (local, regional or national) directly promote the use of bio-based materials through, for example, local requirements prescribing or promoting the use of bio-based insulation in existing buildings e.g. in Belgium (Crucke and Bue, 2019; Region de Bruxelles-Capitale, 2011), or wood materials for specific projects or areas (Francart et al., 2019), regulatory incentives (e.g. Swiss Forest Act and Swiss CO 2 Act (Schweizerische Eidgenossenschaft, 2020; Schweizerischen Eidgenossenschaft, 2017), funding programs (City of Munich, 2013), and/or advancement of technical specifications and structural standards to ease the wood uptake (Natural Ressources Canada, 2020; Wood Sector Alliance for the New European Bauhaus, 2021). This situation encourages more policy initiatives, research and demonstration projects focusing on wood or bio-based construction. The latest trend is the discussion or realization of multi-storey buildings with a primary load-bearing timber frame. Such actions are particularly prevalent in countries where forestry plays an important economic role (Food and Agriculture Organization (FAO), 2020). However, a sole reliance on wood and bio-based material encouragement policies and related prescriptive requirements can be problematic for both designers who may feel constrained to use such products in sub-optimal designs and non-bio-based product C.M. Ouellet-Plamondon et al. Journal of Cleaner Production 404 (2023) 136834 3 manufacturers who may perceive unfair procurement practices (Francart et al., 2019; Goodland, 2016). Therefore, some governments feel obliged to adopt a technology-open and material-neutral approach when developing funding programs or legal requirements. With a performance-based approach, the focus is on the required performance and not the specified solution (Foliente, 2000) but its implementation requires an accepted assessment method to verify that the proposed solution, product, design and/or construction method meet the performance criteria. For this purpose, the application of environmental LCA has already established itself in international standardization to provide the basis for assessing the environmental performance of individual buildings (ISO, 2010). Furthermore, LCA has already entered practice and is starting to be used in procurement, certification (e.g. DGNB and BNB, SNBS, Minergie-eco, SIA 2040) and regulation (recent examples from e.g. Sweden, France and Denmark (Ministry of the Interior and Housing, 2021; Swedish Parliament, 2021)). Bio-based materials compete with other building materials and construction methods. Therefore, to secure market share and avoid future climate regulatory risks, concrete, steel as well as brick and tile industries are currently developing strategies to reduce the GHG emissions caused during the production stage, to reduce resource consumption by expanding recycling and the use of secondary materials, and to improve durability (Bataille, 2020; Cerame Unie, 2012; Eurofer, 2019; Habert et al., 2020; HERA, 2017). The discussion on and quantification of the effects of using wood-based materials thus takes on a social dimension and necessitates the provision of scientific advice to policy. Therefore, it becomes crucial to understand, interpret, correctly apply, or advance according to current scientific knowledge the basic possibilities of dealing with wood and/or bio-based products and buildings in the context of LCA. Bio-based materials like wood, straw, hemp wool and other (insulation) materials are light, heat insulators, but do not store as much energy as heavy masonry (Hens, 2016). Therefore, they contribute little to the whole building thermal inertia. This results in a lower storage capacity of solar gains which in turn may increase heating needs during the use stage of buildings in certain climate zones (Kuczy´ nski and Staszczuk, 2020; Mantesi et al., 2015). Thermal mass also has implications for thermal comfort and cooling needs. However, the influence of building materials on the indoor climate and operational energy demand is outside the scope of this article. 1.3. Research questions In the international project Annex 72 of the Energy and Building Communities program of the International Energy Agency (IEA), we investigate harmonization issues in the methods of calculation and assessment of primary energy consumption, GHG emissions and other environmental impacts caused along the life cycle of buildings. In particular, the treatment of GHG emissions of bio-based products indicates a lack of consistency and transparency worldwide. This leads to the following questions. •What assumptions and methods are used in LCA of buildings with wood and bio-based materials in different countries? •What causes the variability within the same overall assessment and between the methods for the selected case? •What are the implications of the biogenic carbon accounting method on the carbon footprint of building components and buildings? •What are the similarities and differences when comparing the approaches applied in different countries? This study aims to analyze and understand the differences in methodological approaches across countries when applied to the same multiresidential wood building (construction method, material inventory, operational energy demand, etc.) and address the above questions to better understand the differences in the national methods. The study is based on the national LCA approach of the participating countries. For countries not having a national standard, the national approach means the most common or frequent practice in that specific country currently, in terms of both method and database used. On purpose, the databases are those used in the practice. The authors are aware that it will bring differences and the intention is to understand the sources of variation in the results in these circumstances. This extends the work of the European thematic network PRESCO (Practical recommendations for sustainable construction), where seven building LCA tools were applied to a wooden frame house (Peuportier et al., 2004). The exercise presented here aims to reach countries members and nonmembers of the European union. The following section presents the two primary approaches used by participating countries to consider and model biogenic carbon. The analysis highlights the differences when the same methodology is applied and across the different methodologies. The results reflect how practitioners would take their decision on. The implications of the LCA methodological choices on the outcome are discussed. This comparative LCA helps to reduce the risks of misinterpretation, guide material selection decisions based on LCA and identify key areas for harmonization of methods. 2. System of perspectives and state of the art in the methodology of wood assessment Three different approaches are distinguished within the LCA literature on the calculation of biogenic carbon and fossil greenhouse gases (GHG) (Hoxha et al., 2020), based on how temporal biogenic carbon fixation and its release to the atmosphere are considered (Fig. 1). First, the 0/0 method considers neither fixation nor releases of biogenic carbon. Secondly, in contrast, the −1/+1 method, recommended by EN 15804+A2 (European Committee for Standardization (CEN), 2019), accounts for the fixation of biogenic carbon in the production stage and its release in the end-of-life (EN 15804+A2). Variants of the −1/+1 approach apply a −1/+1* in the case of recycling or landfill at the end of life (meaning that the fixation of biogenic carbon is considered, but no or not all biogenic carbon is modelled as an emission at the end of life). In France a 0/+1 approach is used if no tree is regrowing (i.e. the forest is transformed to agricultural or built-up land) or if the wood stems from native forests (EN 15802+A2) and the wood is incinerated at the end of life (meaning that no fixation of biogenic carbon is considered, but emissions do happen at the end of life). Thirdly, the so called “dynamic” approach accounts for biogenic carbon fixation during the growth of trees and its releases at the end-of-life based on annual balances (Levasseur et al., 2013). The method applies a fixed temporal system boundary (100 years after construction) to derive time-dependent global warming potentials. However discounting future emissions may induce impact shifting, particularly by reducing the benefit of energy-saving measures, and it may be considered unfair to future generations so that climate scientists discourage it (Brunner, 2022). It is adopted in only a few regulations or design tools with specific conditions (France and Norway). The countries’ dynamics is also very variable and there is not yet harmonization on the methods. The dynamic approach has not been used by any of the countries participating in the exercise, so it is not further discussed here. Currently, biogenic carbon is accounted according to three methods by national assessment: the 0/0 method, the −1/+1, and the −1/+1* (Fig. 1, Table 1). Explanation of the modules in the life cycle assessment of buildings can be found in standards (European Committee for Standardization (CEN), 2019). The −1 means that the carbon is extracted from the atmosphere and enters the system. The +1 means that the carbon leaves the system. An overview of available standardized product-level guidance and the type of approach recommended is provided in Table 2. ISO/TC 59 SC 17 and ISO 21930 (ISO, 2017a) provide an international basis for developing EPD, while CEN TC 350 standards and EN 15804+A2 (European Committee for Standardization (CEN), 2019), in particular, apply to Europe. C.M. Ouellet-Plamondon et al. Journal of Cleaner Production 404 (2023) 136834 4 According to the current version of EN 15804:2012+A2:2019 (European Committee for Standardization (CEN), 2019), which came into effect in 2019, bio-based products shall address the full life cycle (i.e. at least the modules A1-A3, C1–C4, and, strictly kept separate, module D), as required for all construction products. This standard offers particularly relevant guidance concerning GHG emission-related calculation for developing EPDs of bio-based products: ● Global Warming Potential (GWP), broken down into (1) GWP - total, (2) GWP - fossil, (3) GWP – biogenic, (4) GWP - LULUC (land use and land use changes); ● The biogenic carbon content expressed as kg C (as additional information). Fig. 2 illustrates the difference between using the 0/0 method, the −1/+1, and the −1/+1* method (showing the three components of GWP separately) for the GHG emissions associated with a wood product using the example of 1 m 3 of untreated, dried sawn softwood. For the −1/+1 methodology, the carbon content of the wood is calculated according to EN 16449:2014 (European Committee for Standardization (CEN), 2014), assuming a moisture content of 10%. At the end-of-life, the scenario is 75% recycling, 25% incineration, 0% landfill, which is the current Belgian situation where landfill is prohibited. The results show that the choice for the −1/+1 or 0/0 approach should, in theory, only influence the relative contribution of the individual life cycle stages to the total GHG emissions, but not the total itself (sum of modules A to C). Indeed, with the −1/+1 approach a negative contribution to GWP biogenic is recorded for the product stage (A1-A3), but this removal of CO 2 from the atmosphere is counterbalanced by equal emissions of biogenic CO 2 back to the natural environment at the end-of-life, resulting in a net-zero biogenic CO 2 balance over the product’s life cycle. The above is only valid when the −1/+1 approach is based on a coherent inventory of biogenic carbon flows. In practice, one may observe both net positive and net negative balances of GWP biogenic carbon because the wood product characteristics (water content, density, carbon content) and the often-generic waste wood characteristics do not match and call for individual adjustments. Moreover, as an inherent material property, biogenic carbon content shall be allocated between co-products based on physical relationships, but it is not always done this way. Therefore, applying the −1/+1 approach on product level using generic data or cradle-to-grave EPD’s usually requires a separate calculation of GWP biogenic , based on the specific biogenic carbon content of the product, to ensure a net-zero biogenic carbon balance over the life cycle (Rasmussen et al., 2021). The −1/+1* approach may lead to very different GWP total results than the 0/0 and −1/+1 approach as it does not lead to a net zero biogenic carbon balance when the end-of-life scenario includes landfilling and/or recycling. Finally, the contribution of the timber to GWP luluc is, in this case, negligible as the timber is assumed to be from sustainably managed forests. According to EN 15804+A2 rules, the contribution to GWP luluc would be substantial for wood logged from native forests as in that case no biogenic CO 2 uptake would be reported in A1-A3, and emissions from release/export of stored carbon at the end-of-life (EOL) would be reported (under GWP luluc instead of GWP biogenic ). 3. Case study description and national methodologies The object of assessment, the PAL6 multi-residential building for social housing, was constructed in 2016 in Quebec City, Canada, in the Fig. 1. Methods applied on modelling biogenic carbon in the LCA bio-based products. Carbon fixation is assumed to happen either before the construction stage or carbon fixation during the use stage of the building life cycle. Table 1 Summary of the biogenic CO 2 accounting approaches considered in this article. Method Forest system Building system Module A Module B Module C 0/0 – – – – −1/+1 Carbon fixation Carbon fixation (−1) – Carbon emission (+1) −1/ +1* Carbon fixation Carbon fixation (−1) – Carbon emission (+1) Carbon landfilled (variable indicated by the star) a ) a 0.11 (Canada); 0.001 (New Zealand); 0.08–0.09 according to wood type (France); 0.1 (Australia). Table 2 Standard methods to account biogenic carbon. Standardized guidance for product-level data Approach PEFCR (European Commission, 2017) a , SIA 2032 (SIA, 2020) 0/0 PAS 2050 (BSI, 2011), EN 15804+A2; 2019 (CEN, 2019), ISO 14067 ( ISO, 2018) and ISO 21930 (ISO, 2017a) −1/+1 a For cradle to grave assessments of final products with a life time of less than 100 years. C.M. Ouellet-Plamondon et al. Journal of Cleaner Production 404 (2023) 136834 5 humid continental climate zone (Dfb) in the K¨ oppen-Geiger climate classification system in a seismic region (Fig. 3, Fig. S1). The building has 59.4 m in length, 18.3 m in width, and 18.2 m in (above ground) height, comprising 6 floors of 1090 m 2 and 59 social housing units. Its lightweight wood structure is made of (in mass) 52% concrete 25 MPa, 27% wood beams (38 mm ×89 mm (locally known as 2 ×4 inches in nominal dimensions) or 38 mm ×140 mm (locally known as 2 ×6 inches)), 7.6% fire-resistant gypsum panels, and 4.6% regular gypsum panels. In total, around 1500 metric tons of wood were used, with a volumetric breakdown into 1896 m 3 softwood beams, 116 m 3 softwood plywood, 102 m 3 oriented strand board panel, 32 m 3 wood fiber board, and 0.1 m 3 laminated wood flooring. The wood used came from new grown certified forest with no land use change after logging. Detailed specifications are given in Table 3. This study aims at comparing the carbon footprint accounting methods from 16 countries. Details on the field of application of each method are shown in Table S1 in the supplementary information. The object of assessment, the PAL6, is the same for all applied national methods. However, each national team applied the reference study period (RSP), life cycle modules covered (according to the modular life cycle model from CEN TC 350 standards), modelling rules, the geographical scope of inventory data as well as the LCA database according to its specific national method. The first two are presented in Table 4. RSP together with the expected service life of replaced elements (given in Table S2 in the supporting information) define the number of replacements to be accounted for under module B4. The latter is often identified as the most influencing parameter on use-stage embodied GHG emissions, as the higher the number of replacements, the larger the relative contribution of B4 is (Goulouti et al., 2020). Except for France and New Zealand, the RSPs used by different countries do not largely variate. New Zealand did the analysis with a RSP of 90 years based on (Johnstone, 1994), but it was recently changed to 50 years (Ministry of Business Innovation Employment, 2022). The building material’s quantities and technical quality are predefined in the inventory and element composition, thus consistent across the different applied methods. All teams presented the results Fig. 2. Life cycle GWP of 1 m 3 planed, dried (MC =10%) softwood beam (density (wet) =465 kg/m 3 , carbon content 0.494 kg/kg dry wood) according to the life cycle stages using the 0/0, −1/+1 and the −1/+1*methodologies. Fig. 3. PAL6 multi-residential building made of a light wood frame structure. C.M. Ouellet-Plamondon et al. Journal of Cleaner Production 404 (2023) 136834 6 according to the same structure of building components. Austria and Norway could not assign the impacts of all modules to building components. Those impacts are summarized in “others”. For Austria, these are the impacts occurring in life cycle stage A5 and for Norway the impacts in the life cycle stages A4-A5, B2–B5 and C1–C4. The stages B6/B7 were not included in the analysis and not discussed. The inclusion/exclusion of module D is here shown only for information; it is not part of the following analysis. Table 5 shows the methodological approach for accounting biogenic carbon, the software, the database, and the source of biogenic carbon of each national expert teams. Each national assessment applied the database most used in the respective national context on purpose, which brings variability in the results. This has also been shown in other similar exercises (Frischknecht et al., 2019, 2020). While some countries have national databases (e.g. ¨ Okobau.dat) and EPDs in place, most countries use different versions of Ecoinvent adapted to the national tools and context. Spain for example chose to keep the continuity of environmental data over time, instead of a new version of the database and they found the version of the database does not affect significantly their results. The method to account for biogenic CO 2 was either 0/0 or −1/+1, or a variation of the latter. Belgium and Switzerland will change to −1/+1, but the change is not yet made. The methods applied in Brazil, Canada, and New Zealand allow to quantify the biogenic GWP integrated in the LCA tool, while in all other countries GWP biogenic is integrated into the GWP total values. The LCA approach was attributional with a process based LCA database approach in all cases. The scope covers the construction, use-stage embodied GHG emissions, not the operational emissions, and end of life of the PAL6 residential building in a national-specific methodological setting. The software SimaPro was the most often used software for the calculation and many countries had their own tool. In the European standard EN 15804:2012+A2:2019, the +1 shall also apply on recycling and landfill of wood. However, in Canada, France, New Zealand and Australia wood sent to landfill gets a factor close to zero. Wood that exits the system boundary e.g. for reuse, recycling gets a “+1” in NZ, and then the potential benefit of its reuse, recycling is calculated in module D. In the French EQUER method (Table 6), negative biogenic CO 2 emissions are accounted for in the production stage if a new tree is growing which is the case for wood from certified forests. But if the wood stems from non-certified forests, the same amount of carbon is stored in the building as if it were stored in the forest. Therefore, no carbon fixation is considered (“0” instead of “-1”). At the end of life, the quantity of biogenic CO 2 is emitted if the wood is incinerated, but not if the wood is landfilled or recycled (see Table 6). See electronic materials for more details on the approaches. Table 7 shows the share of wood recycled, incinerated and landfilled in the end-of-life scenarios of the countries participating to the case study as well as whether or not biogenic carbon in wood landfilled is considered (partly) permanently sequestered. This information is important for the module C end of life assessment. Landfilling of wood is not allowed in certain countries due to the reactive nature of organic materials, leading to leachate and methane which require additional treatment. In Canada, the given proportion of wood recycled and landfilled is based on the practice and not the legal requirement (Audet, 2020). The transformation of landfilled wood into methane is not assessed in every country, for precaution, many countries prefer forbidding to landfill organic materials including wood. Australia, New Zealand, and USA have experimental data to support the landfilling of wood (Wang et al., 2011; Ximenes et al., 2019). Table 3 Specification of the building PAL6. Description Dimension of PAL6 Unit Materialisation of PAL6 Excavation 3053 m 3 0.1 m–2.1 m of backfill, 0.3 m–3.3 m of sandy silt, 13.4 m–19.6 m of sand and 8.4 m or more of clay. Highly seismic soil Refilling 1032 m 3 // Foundation 1121 m 2 Concrete - 30 MPa (GU, with air), reinforcing steel (5/8′′ , 6/8′′ , 7/8′′ , 8/8′′ and 9/8′′ ), drawbar and shrinkage compensator Slabs 1121 m 2 Concrete - 25 MPa (GU, without air entrained) and reinforcing steel Roofs 1108 m 2 Base coat elastomeric bitumen membrane, gypsum panel 16 mm and type X (16–25 mm), polyisocyanurate insulation (100 mm)/rock wool, self-adhesive vapour barrier, softwood plywood and top coat elastomeric bitumen membrane. Pillars (pile head) 289 pieces Concrete - 25 MPa (GU, without air entrained) and reinforcing steel Outer walls basement 982 m 2 Concrete - 30 MPa (GU, with air) and reinforcing steel Flooring basement 269,6 m 2 Concrete - 25 MPa (GU, without air entrained) and reinforcing steel Stairs flooring basement 292 m 2 Concrete - 25 MPa (GU, without air entrained) and reinforcing steel Outer walls first and upper floors 1409 m 2 Ext Wall M-2: aluminium mesh panel/aluminum cladding, gypsum panel 16–25 mm type X, metal stud, OSB panel covered with self-adhesive membrane, polyethylene vapour barrier, wood beam (2 ×4, 2 ×6) 1082 m 2 Ext Wall M-5: gypsum panel 16–25 mm type X, hot galvanized steel plate, lightweight concrete panel, metal stud, OSB panel covered with self-adhesive membrane, perforated solar panels - 10% steel and 90% polycarbonate glazing, polyethylene vapour barrier, prefabricated lightweight concrete panel, rock wool, wood beam (2 ×4, 2 ×6) 101 m 2 Ext Wall M-7: aluminum cladding, metal stud, self-adhesive membrane, softwood plywood 789 m 2 Floor P-2: extruded polystyrene rigid insulation and polyethylene vapour barrier 5417 m 2 Floor P-4: acoustic membrane, acoustic panel, gypsum panel 16–25 mm type X, polyethylene membrane, rock wool, softwood plywood, sound insulators and metal furrings soundproof drywall ceiling 297 m 2 Floor P-5: Acoustic panel, gypsum panel 16–25 mm type X, rock wool and softwood plywood Windows 415 m 2 PVC - U Double-glazed window Inner separating walls 6143 m 2 Non-structural interior partition wall: gypsum panel 16 mm, metal stud, softwood plywood, and wood beam (2 ×4 or 2 ×6) 731 m 2 Non-structural exterior partition wall: aluminium mesh panel, concrete block 190 mm, gypsum panel 16–25 mm type X, metal stud, rock wool, wood beam (2 ×4 or 2 ×6), wood fibre board Inner bearing walls 771 m 2 Structural common partition wall: gypsum panel 16–25 mm type X, oriented strand board panel, rock wool, wood beam (2 ×4 or 2 ×6), wood fibre board 843 m 2 Structural interior partition wall: gypsum panel 16–25 mm type X, oriented strand board panel, rock wool, wood beam (2 ×4 or 2 ×6) 2339 m 2 Structural exterior partition wall: gypsum panel 16–25 mm type X, oriented strand board panel, rock wool, wood beam (2 ×4 or 2 ×6), wood fibre board Inner/Outer doors 123 m 2 Exterior door and exterior sliding patio door Flooring 4477 m 2 Ceramic flooring and laminated wood flooring Ceiling covering 8362 m 2 Acoustic tile and gypsum panel 16 mm C.M. Ouellet-Plamondon et al. Journal of Cleaner Production 404 (2023) 136834 7 4. Results and discussion 4.1. Assessment on life cycle stages of the superstructure The choice of the approach (0/0, −1/+1, −1/+1*) has a greater impact on the variability of the results of the wooden superstructure, especially at the module A and C stages, than the choice of the database and software. Results of greenhouse gas emissions calculated with both 0/0, −1/+1, and −1/+1* approach for the superstructure according to life cycle stages are presented in Fig. 4. The greenhouse gas emissions calculated with a 0/0 approach are, an average of 1.77 kgCO 2 eq/m 2 a with variation between 1.41 kgCO 2 eq/m 2 a for Portugal to 2.44 kgCO 2 eq/m 2 a for Czech Republic. The average of greenhouse gas emissions calculated with −1/+1 equals to 3.75 kg CO 2 eq/m 2 a with a variation between 2.23 kgCO 2 eq/m 2 a for Denmark to 6.43 kgCO 2 eq/ m 2 a for Spain. With the −1/+1* method, the greenhouse gas emissions averaged 1.05 kgCO 2 eq/m 2 a with a variation from −0.41 kgCO 2 eq/m 2 a for New Zealand to 3.11 kgCO 2 eq/m 2 a for Australia. A more in-depth analysis of the greenhouse gas emissions of building Table 4 Reference study period (RSP) and life cycle modules (according to CEN TC 350 standards) covered in the country’s assessment (Stages B6, B7, and D were not analyzed). Country Code RSP Life cycle modules covered Product stage Use stage EoL stage Additional A1-3 A4 A5 B1 B2 B3 B4 B5 B6 B7 C1 C2 C3 C4 D Australia AU 50 x x x x x x x x x Austria AT 50 x x x x x x x x Belgium BE 60 x x x x x x x x x Brazil BR 50 x x x x x x x x x Canada CA 60 x x x x x x x x x Czechia CZ 50 x x x Denmark DK 50 x x x x x France FR 100 x x x x x x x x x x a Germany DE 50 x x x x x x New Zealand NZ 90 x x x x x x x x x x x x Norway NO 60 x x x Portugal PT 50 x x x Spain ES 50 x x Sweden SE 50 x x x x x x Switzerland CH 60 x x x x x x x USA US 50 x x x x x x x x x x a For France, D is reported separately in the Annex 72 spreadsheet but subtracted from C in the tool as it is used in France. Table 5 Method for biogenic carbon, software and database used, as well as the possibility of showing GWP biogenic values (separately from GWP total values. Country code LCA method Software Database Carbon content source Is GWP biogenic separately given? 0/ 0 −1/ +1 −1/ +1 a AU x a SimaPro 9.0.0.41 AusLCI database (2016) AusLCI database in SimaPro and EPDs Australasia Yes AT x SimaPro 9.1.0.8 Ecoinvent v3.5 n/a No BE x SimaPro 8.5.0 Ecoinvent v3.6 n/a No BR x SimaPro 9.0 Ecoinvent v3.4, 3.5, 3.6 n/a No CA x a Simapro 9.1.0.8 Ecoinvent v3.4 Ecoinvent Yes CZ x Excel-based tool/calculation Ecoinvent v3.3 n/a No DK x LCAbyg 3.2 ¨ Okobaudat v2016 EN 16449:2014 No FR x a,b EQUER Ecoinvent v3.4 Ecoinvent No DE x LEGEP 2018 ¨ Okobaudat v2018 EN 16449:2014 No PT x SimaPro 8.4.0 LCIA database for Portuguese Building Elements and Materials EN 16449:2014 No ES x Excel-based tool/calculation Ecoinvent v2.0 n/a No SE x Byggsektorns Milj¨ ober¨ akningsverktyg (BM), (Swedish Building Sector Environmental Calculation Tool) Database embedded in Swedish Building Sector Environmental Calculation Tool (BM) n/a No NZ x a LCAQuick 3.4.2 EPDs +Ecoinvent v3.1 EPDs Australasia GWP total (biogenic carbon can be derived) NO x ZEB tool Norwegian EPDs +Ecoinvent 3.1 EN 16449:2014 Yes CH x Excel-based tool/calculation KBOB LCA data DQRv2:2016 n/a No USA x Athena IE v5.4 GaBi v9.2 n/a No a Wood sent to recycling and landfill gets a value “>0” but ≪1”. b Differentiation between certified (−1) and non-certified (0) forests. Table 6 Biogenic carbon accounting according to the French Equer method. End of life ->Production Incineration Landfill, recycling or reuse Sustainable forest management (a new tree is growing) −1/+1 −1/>0 Other case (non-certified forest) 0/+1 0/>0 C.M. Ouellet-Plamondon et al. Journal of Cleaner Production 404 (2023) 136834 8 life cycle stages highlights large differences in the results of the modules A1-A3 and C1–C4. Comparisons of the greenhouse gas emissions for the stages A1-A3 and C1 to C4 highlight the main difference between the three approaches. The emissions calculated for A1-A3 with the −1/+1 and −1/+1* approaches are (partly) offset by emissions in the modules C1–C4. It deserves to be highlighted that the quantities of biogenic carbon uptake and release may differ due to non-matching biogenic carbon mass balances. For France and New Zealand, the longer service life of 100 years and 90 years, respectively, has the effect of making the results for these countries in Fig. 4, expressed per m 2 a, smaller, in comparison with most countries that use a 50 year or 60 years’ service life. The assessment of the superstructure made of light frame wood construction highlights the influence of the end-of-life scenarios of the wood products and the choice of the methodology for the biogenic carbon accounting on the GHG emission results. The assessments of countries with a −1/+1* and a landfilling scenario approach report less GHG than those applying the 0/0 and −1/+1 method. New Zealand also shows a net fixation of carbon in the superstructure (Fig. 4) because a large share of the wood is landfilled and because it assumes permanent carbon sequestration. Table 5 shows that 75% of wood is landfilled in New Zealand, 55% in Australia and 40% in Canada. The interpretation of landfills as a permanent or temporary sequestration varies among the participating countries. In Canada, it is interpreted as permanent sequestration. In Australasia, two values of degradable organic carbon fraction (DOCf) for softwood timber are allowed: NZ applied the lower value of 0.1% while AU applied the higher value of 10% (Australian Government, 2016; Wood, 2020), which results in 99.9% and 90% assumed permanent sequestration in NZ and AU, respectively. The implications of this 100-fold difference is described in section 4.4. 4.2. Total building and its building components The analyses of the whole building show variations in the results (Fig. 5) and there is less variability among the approaches. In the 0/ 0 approach, the greenhouse gas emissions per square meter per year have a mean value of 5.96 kgCO 2 eq/m 2 a for the total A-C, with the national methods of BR and PT that have obtained the highest and lowest results with 10.3 kgCO 2 eq/m 2 a and 3.99 kgCO 2 eq/m 2 a, respectively. In the case of the +1/-1 approach, the values obtained had a means of 7.17 kgCO 2 eq/m 2 a, with the national methods of ES and NO that have obtained the highest and lowest results, with 12.9 and 4.80 kgCO 2 eq/m 2 a respectively. In the 0/0 and the −1/+1 approach, the data in the LCA model create variability in the results. With the −1/+1* approaches, the total A-C impact averaged 4.35 kgCO 2 eq/m 2 a with a high value of 8.86 kgCO 2 eq/m 2 a for Australia and 1.87 kgCO 2 eq/m 2 a for New Zealand. Within the four −1/+1* assessments (AU, CA, FR, NZ), the difference in the share of wood being incinerated at end-of-life and the mismatch in carbon contents of wood products and its corresponding end of life treatment datasets are the main reasons for the large variability in greenhouse gas emissions. In most assessments, the superstructure contributes most to the specific greenhouse gas emissions. The foundations and the interior construction are significant construction systems in terms of greenhouse gas emissions and present similar values and deviations in both methods. The impacts of these building components also vary among countries. The foundations represent a mean value of 0.97 kgCO 2 eq/ m 2 a, AU and FR have the highest and lowest values, with 1.7 kgCO 2 eq/ m 2 a and 0.42 kgCO 2 eq/m 2 a respectively. The internal construction represents a mean value of 0.91 kgCO 2 eq/m 2 a, ES and NO have the Table 7 Shares of solid wood recycled, incinerated, landfilled in the countries’ assessments for module C assessment (n/a means not available because end-of-life is not considered in the national methods). Wood recycled % Wood incinerated % Wood landfilled % Permanent sequestration in landfill assumed Yes/ No AU 40 5 55 Yes AT 75 25 No BE 75 25 No BR 100 No CA 60 40 Yes CZ n/a n/a n/a n/a DK 100 No FR 100 Yes (see Table 1) DE 100 No ES 10 80 10 Yes NO 100 No PT n/a n/a n/a n/a NZ 15 10 75 Yes SE 100 No CH 50 50 No US n/a n/a n/a n/a Fig. 4. Greenhouse gas emissions of the superstructure of the PAL 6 building assessed. C.M. Ouellet-Plamondon et al. Journal of Cleaner Production 404 (2023) 136834 9 highest and lowest values, with 2.4 kgCO 2 eq/m 2 a and −0.23 kgCO 2 eq/ m 2 a respectively. NO and NZ are the only countries that obtain negative values in some building components. Norway divided the life cycle stages A1-A3 into the building elements, while the impacts of the other life cycle stages are summarized as “others”. 4.3. Wood products The analysis of the wood products shows major differences in module A and C between the countries due to the approach to assess biogenic carbon. Fig. 6a shows the life cycle greenhouse gas emissions associated with 1 kg of solid wood, as reported from the different assessments based on specific data sources. In total, 1896 m 3 softwood is used in the PAL6 building. The cradle-to-gate results (modules A1-A3) from the 0/ 0 approach vary by up to a factor of 2.8, whereas the results from the −1/+1 approach vary by a factor of 1.3, and a factor of 1.5 with the −1/ +1* approach. With the 0/0 approach, the highest and lowest total impacts without D are for Czech Republic and Brazil having impacts of 0.17 and 0.08 kg CO 2 eq/kg softwood, respectively. With the −1/+1 method, the highest and lowest impacts are for Norway and Spain having impacts of 0.16 and 0.12 kg CO 2 eq/kg softwood, respectively. With the −1/+1* method, the highest and lowest impacts are for Australia and New Zealand having impacts of 0.42 and −0.42 kg CO 2 eq/ kg softwood, respectively. Negative results in A4-A5 reported from the Canadian and France cases relate to the assumption that 5% of the needed material is wasted. Production of this wood surplus, which leads to additional carbon fixation, is accounted for in A5. The value remains very small. Further, the waste is modelled as landfilled, assuming some permanent sequestration of biogenic CO 2 , resulting in an overall negative balance of biogenic CO 2 . In modules C1–C4, the biogenic carbon contained in wood products is modelled as a release of biogenic CO 2 in the −1/+1 approach. The release balances the biogenic CO 2 uptake modelled in the production modules, as seen in the figures for ‘total without D’. In the −1/+1* approach, both the Canadian and the New Zealand modelled the end-of- life route of wood as being landfilled (40% and 75%, respectively). They consider also limited decomposition rates of wood in landfills (NZ: 0.1%, CA: 11%), thereby resulting in an overall negative value of biogenic CO 2 emissions and biogenic greenhouse gas emissions for the total life cycle. Module D is only addressed in four of the national approaches reported in this study. Fig. 6b shows the plots of minimum and maximum values from the current assessments into the schematic overview of balanced approaches first introduced in Fig. 2. Fig. 6b shows that several assessment values are distributed around the balanced numbers (dots in black color, see also Section 2), the minimum (blue dot) and maximum values (orange dot). However, specifically for the −1/+1* approach, the reported values are unevenly distributed due to the methodological specifications concerning landfills and (partly) permanent sequestration as detailed in section 4.1 and Table S1, as well as service life. Fig. 7 shows the life cycle greenhouse gas emissions associated with 1 kg of plywood, as reported from the different assessments. In the case of plywood, the countries with the 0/0 approach reported higher greenhouse gas emissions than the countries with the −1/+1 approach and −1/+1*; their results are also more variable. The total reported greenhouse gas emissions (without module D) are notably higher for the plywood than for softwood, with an average of 0.71, 0.45, 0.40 kg CO 2 eq/kg for the 0/0, −1/+1, −1/+1* approaches respectively. This difference reflects the additional manufacturing processes and fossil sourced raw materials associated with the production of plywood compared to sawn wood. With the 0/0 approach, the highest and lowest impacts are noted for Switzerland and Sweden at 1.17 and 0.20 kg CO 2 eq/kg plywood, respectively. With the −1/+1 approach, the highest and lowest impacts are for Norway and Spain at 0.77 and 0.24 kg CO 2 eq/kg plywood, respectively, especially due to the transportation distance and the energy use in the processes. With the 1/+1* approach, the highest and lowest impacts are for Australia and New Zealand at 1,0 and −0.15 kg CO 2 eq/kg plywood, respectively. Only New Zealand reports net-negative emissions for the total greenhouse gas emissions (without module D), arising from landfill emissions based on a DOCf value of 1.4% reported in the Australasia EPD for plywood (Wood, 2017). If landfills were not considered as a permanent carbon Fig. 5. Multi-residential building assessment of greenhouse gas emissions, grouped according to the biogenic carbon modeling approach and the service life (Modules A-C included, B6 excluded; Norway divided the life cycle stages A1-A3 into the elements, while the impacts of the other life cycle stages are summarized as “others”). C.M. Ouellet-Plamondon et al.