Abstract
The building sector has seen recent growth in the number of published environmental product declarations (EPDs). EPDs share environmental data of construction materials, which can help building practitioners prioritize products with lower environmental impacts. However, EPDs rely on varying assumptions within their life cycle assessment (LCA). This study aimed to evaluate the use of EPDs as a data source instead of a generic data source and its effect on the life cycle impacts of a multi-residential building. This study focused on 19 North American EPDs of structural wood products. The impact assessment results found in the EPDs were compared to the Ecoinvent V3.8 database. The findings of the present study suggest that EPDs can generally be used without distinction compared to the data in the Ecoinvent V3.8 database. However, a few data were found to be outliers. In addition, EPDs of structural wood products only disclosed its manufacturing stage. Other life cycle stages, such as transport of the construction product, can have a significant impact on the building’s LCA. Therefore, using EPDs to assess building impacts is recommended over their direct comparison for practitioners to make more comprehensive decisions towards embodied impacts of buildings.
Download PDF
Full Article
Environmental Product Declaration (EPD) Usage in Early Building Design Stages: Review of Effects on the Environmental Life Cycle of a Multi-Residential Building
Gabrielle Pichette,a,* Pierre Blanchet,a Gatien Geraud Essoua Essoua,b and Charles Breton a
The building sector has seen recent growth in the number of published environmental product declarations (EPDs). EPDs share environmental data of construction materials, which can help building practitioners prioritize products with lower environmental impacts. However, EPDs rely on varying assumptions within their life cycle assessment (LCA). This study aimed to evaluate the use of EPDs as a data source instead of a generic data source and its effect on the life cycle impacts of a multi-residential building. This study focused on 19 North American EPDs of structural wood products. The impact assessment results found in the EPDs were compared to the Ecoinvent V3.8 database. The findings of the present study suggest that EPDs can generally be used without distinction compared to the data in the Ecoinvent V3.8 database. However, a few data were found to be outliers. In addition, EPDs of structural wood products only disclosed its manufacturing stage. Other life cycle stages, such as transport of the construction product, can have a significant impact on the building’s LCA. Therefore, using EPDs to assess building impacts is recommended over their direct comparison for practitioners to make more comprehensive decisions towards embodied impacts of buildings.
DOI: 10.15376/biores.18.4.8134-8150
Keywords: Embodied impact; Product category rules (PCR); Building information modeling (BIM); Whole building life cycle assessment (WBLCA); Wood-based structures
Contact information: a: NSERC Industrial Research Chair on Eco-responsible Wood Construction (CIRCERB), Laval University, Department of Wood and Forest Sciences, 2425 De La Terrasse Street, Québec City, QC G1V 0A6, Canada; b: Vertima, 604 St-Viateur, Quebec City, QC G2L 2K8, Canada;
* Corresponding author: gabrielle.pichette.1@ulaval.ca
INTRODUCTION
The building sector is responsible for 13% of Canada’s greenhouse gases (GHG) emissions or 18% when including electricity-related emissions (Natural Resources Canada 2022). The building sector accounts for the third-largest source of emissions in Canada. This is why decarbonization of this sector is important to achieve Canada’s 2030 climate target, reduce emissions by 37% from 2005 to 2030, and a net-zero economy by 2050 (Natural Ressources Canada 2022). Considering the economic importance of the building industry and its share of environmental impacts, buildings have been subjected to many studies using life cycle impact assessment (LCA). LCA is a scientifically based methodology that quantifies the environmental impacts of any product or service over its life cycle (ISO 2006). Efforts have been made by the industry to reduce operational impacts by increasing energy efficiency in buildings (Azari 2019). However, a superior energy efficiency target may be achieved by increasing the use of materials in building envelopes, which can lead to higher embodied impacts of buildings (Feng et al. 2022; Larivière-Lajoie et al. 2022). Several studies suggest that buildings with improved operational energy performance or a low-impact energy mix can lead to embodied impacts being responsible for a greater share of the total impacts of the building (Thormark 2006; Chastas et al. 2016; Lessard et al. 2018; Röck et al. 2020). This is particularly the case for residential buildings, where operational energy is provided mostly by renewable energy, hydroelectricity, such as the province of Quebec in Canada. Therefore, reaching target emissions within the building sector must be done by acknowledging embodied carbon emissions by addressing whole building life cycle assessment (WBLCA) (Pomponi and Moncaster 2016). This requires the implication of all actors of the building sector, from manufacturers, architects, engineers, LCA practitioners to policy makers.
Structural elements are often responsible for a great part of the environmental impact from the production stage of a building, according to LCA (Chau et al. 2007; Lessard et al. 2018). There is great potential to use wood-based materials. Previous research suggests that substituting structural elements made of concrete or steel by wood-based products can help reduce environmental impacts of buildings (Essoua and Lavoie 2019; Hart et al. 2021; Robertson et al. 2012). However, these conclusions may vary depending on end-of-life scenarios (Hart et al. 2021) and carbon sequestration calculations made within the LCA methodology (Morris et al. 2021). Figure 1 presents three wood-based products used as structural elements in buildings. Glued laminated timber (GLT) is made of wood laminations bonded together parallelly with a moisture-resistant adhesive to form stress-rated engineered wood beams. Cross-laminated timber (CLT) is made of an odd number of layers (generally three, five or seven) of kiln-dried lumber boards bonded together perpendicularly with a structural adhesive to form a solid wood engineered panel. Laminated veneer lumber (LVL) is a type of structural composite lumber (SCL). LVL is made of multiple layers of thin wood veneer, parallel to the long direction, bonded together with a moisture-resistant adhesive to form headers and beams or a component of engineered I-joists.
Fig. 1. Structural wood products evaluated in this study: A) Glued laminated timber (GLT), B) cross laminated timber (CLT) and C) laminated veneer lumber (LVL)
Several standards and certifications schemes are used to characterise environmental performance of building materials. The ISO 14020 series presents three types of environmental labels and declarations: type I, II, and III. Type III declarations, also named environment product declarations (EPD), are based on LCA methodology and are third-party verified. EPDs provide environmental data of construction products. EPDs follow guidelines from specific product category rules (PCRs) to conduct the LCA to enable comparison between products (Ingwersen et al. 2012; Del Borghi 2013; Modahl et al. 2013). PCRs contain information that should be found in the EPDs for a specific product category. Some PCRs are made for products within a same geographical context, while others do not specifically mention it. EPDs have the potential to facilitate comparison between products because of the quantitative information disclosed within (Cobut et al. 2013). Certified environmental claims inform architects about potential benefits of wood-based products.
To characterize a building’s material environmental impact, a large amount of data on processes and materials is required. Published databases such as Ecoinvent are generally used to conduct LCAs. As these databases may not always have processes representing specific building materials, EPD results are being used as data to conduct building LCAs in several certification schemes. Primary data, or specific data, refers to data from a specific process within the supply chain of the manufacturer (European Commission 2016). Secondary data, or generic data, is obtained from third-party life cycle inventory (LCI) database or other sources. Published databases such as Ecoinvent, GaBi, and industry-average data are considered secondary data. EPDs are a specific case: They can rely on both primary and secondary data. For the end user consulting the EPD, it may not always be clear which life cycle stage is based on which type of data or assumption. More in-depth information for each life cycle stages is not disclosed in EPDs, as such information would be for a process in an Ecoinvent database.
The amount of building environmental assessment schemes using EPDs is rising in developed countries (Arvizu-Piña and Cuchí Burgos 2017). Consequently, the construction sector saw an increase in demand and publication of EPDs (Gelowitz and McArthur 2016; Bernardi et al. 2017). The architecture, engineering, and construction (AEC) industry seems to use EPDs more often since its adoption as credits in Green Building Rating Systems (GBRS) such as Leadership in Energy and Environmental Design (LEED) (Burke et al. 2018). A study performed by Gelowitz and McArthur (2018) on Canada’s first LEED v4 platinum commercial project discussed the client’s, designer’s, and contractor’s experience with the use of EPD for obtaining the material disclosure and optimization’s credits. Designers found EPDs to be more appropriate for use in whole building assessment rather than direct comparison between similar products due to the lack of harmonization of PCRs. To avoid potential pitfalls, the use of the Integrative Design Process (IDP) was mentioned to be essential for this project, as it allowed all stakeholders to learn about EPDs and facilitate their insertion into the project.
An EPD can help manufacturers to identify and improve a manufacturer’s most impactful process. On the other hand, EPDs are also used to communicate environmental impacts of products to individuals who might not have LCA knowledge (Bergman and Taylor 2011). However, interpretation of LCA results can be difficult for individuals with less experience in LCA (Modahl et al. 2013; Ibáñez-Forés et al. 2016). In practice, lack of harmonization between developed PCRs are responsible for difficulties to compare EPDs (Del Borghi 2013; Hunsager et al. 2014; Gelowitz and McArthur 2017; Kerr et al. 2022). Attempting to compare different types of products using their EPDs can lead to poor comparison assessment, as their respective PCR mandates different rules such as declared unit or LCA scope (Kerr et al. 2022).
Considering the increasing interest in data sources such as EPDs (Burke et al. 2018) and potential difficulties of misunderstanding EPDs by practitioners in the construction industry, this study aims to observe the impact of using EPD results in a LCA of a multi-residential building with a specific focus on wood structural products. The objective of this study was to analyze the impact of replacing the data from Ecoinvent V3.8 of the main wood structure by data available publicly from EPDs. This study focused on structural wood EPDs from the North American market. The multi-residential building used for this study is a hypothetical mass timber building constructed in Quebec City.
EXPERIMENTAL
The project was carried out into two main steps simultaneously: searching and comparing EPDs for the main structural wood elements (LVL, GLT and CLT), and completing a building’s LCA with a structure made of CLT and LVL (Fig. 2).
Fig. 2. Schematic view of the project’s main methodological steps
Table 1. Data Extraction Criteria from EPDs Collected
Searching and Comparing EPDs
The present study focused on North American EPDs. The EPDs follow ISO 14025 and are based on independently verified LCA data in accordance with the ISO 14040 series standard. The search for EPDs were done on the Transparency Catalog, a website that compiles every EPD from program operators in North America with different filter options (Substainable Minds 2022). A second search was conducted directly on an internet search engine, Google, to ensure that no EPDs were missed with the following keywords: CLT EPD, GLT EPD and LVL EPD. Data were collected before August 2022. EPDs were available in PDF format. Data were extracted manually from those EPDs and entered in an Excel spreadsheet. Data extracted from the EPDs and their description are presented in Table 1. If a data item wasn’t available, a blank entry was left.
Impact assessment results from EPDs are compared to their corresponding dataset from the Ecoinvent v3.8 database. The Ecoinvent database was developed in Europe, but a large number of processes also represent other geographical regions in the world. It includes many construction material processes and has regional data available for the province of Quebec, Canada (Lesage and Samson 2016), where the construction site is located. If no processes from the province of Quebec were available, processes modeled for the rest of the world (RoW) were chosen.
Conducting the Building’s LCA
The residential building used in this work has been created for the purpose of this study to limit subjectivity due to limitations that may occur in a typical project. A hypothetical six-story building, for a total of forty-eight units, had a mass timber structure (Fig. 3) and located in Quebec City, Canada. This building was previously used for another study: Hosseini et al. (2023). It was made by an architect to reflect the context of Quebec City. The main components of the structure are CLT, LVL, and steel stud and furring with a concrete foundation. The building envelope is made of stone wool and extruded polystyrene. The building had an underground parking area made of concrete. The building’s facade was made of clay brick.
Fig. 3. 3D section view of the referenced residential building used for this study and exterior wall section
An attributional LCA was conducted on the residential building. The functional unit was designed to accommodate a six-story residential building of eight units per floor for a total of 6 196 m2 (66 693 sq. ft.) in Quebec City (Canada) with a lifespan of 50 years. Based on a literature review of two decades of peer-review publications of environmental evaluations of low-rise and high-rise buildings, it was found that most studies use a lifespan of 50 years or longer (Bahramian and Yetilmezsoy 2020). The LCA software, LCI database and LCIA method used were SimaPro, Ecoinvent v3.8 and TRACI 2.1 respectively.
The system boundaries for the building assessment were cradle-to-grave, which includes manufacturing, construction, use and end-of-life stages (Fig. 4). Life cycle stages declared in EPD’s of structural wood products were only the manufacturing stage, A1, A2 and A3 (Fig. 4). Details related to assumptions made for each life cycle stages are presented in Table 2.
Fig. 4. Life-cycle stages declared for EPDs of structural wood products and the evaluated residential building. Source: Adapted from EN 15804:2012+A2 (CEN/TC 350 2019)
A bill of materials (BoM) representative of knowledge at early design stages of the modeled building is presented at Table 3. Processes used in Ecoinvent for each of them as well as total quantities per material are presented. These processes represent the manufacturing stages (A1-A2-A3) for the studied building. Based on their weight, concrete, CLT and brick have the highest proportion of construction materials. Based on their volume, CLT, stone wool, and concrete have the highest proportion of construction materials. The structural wood products evaluated in this study, CLT and LVL, represent 26% and 38% of weight and volume respectively. The impacts of the distance of transport were considered later in the analysis (see Table 5) and were analysed.
Table 2. LCI Assumptions for Each of the Building’s Life Cycles Stages
Table 3. Bill of Material and Related Ecoinvent Processes for the Studied Building
RESULTS AND DISCUSSION
A total of four LVL, six CLT, and nine GLT EPDs were found in the geographical scope of North America. To understand the structural wood-based product’s industry and its small amount of data available, the number of manufacturers in North America was evaluated (Forest Economic Advisor 2022). The total number of manufacturers for these products in North America and its proportion that have published an EPD is presented in Table 4. Most manufacturers of mass timber products such as CLT and GLT have published an EPD.
Table 4. Proportion of Manufacturers in North America with an EPD, Type of EPD and the PCR’s Program Operator According to the Type of Product
Every EPD declared a functional unit of 1 m3 of structural wood product. The assessed life-cycle stages of all products were the production stage: A1 (extraction), A2 (transport), and A3 (manufacturing). However, as some EPDs only report total aggregated values (A1-A3), environmental impacts could only be compared for the whole production stage. Table 4 presents the proportion of EPDs according to their type and referenced PCR for each structural wood product. There were two different program operators for the PCR used by these EPDs: UL environment and FPInnovations. The FPInnovations PCR North American Structural and Architectural Wood Products was the first one to be published for this category of products. UL Environment followed in 2019 and published Product Category Rule Guidance for Building-Related Products and Services, Part B: Structural and Architectural Wood Products. As EPDs are valid for 5 years, a few EPDs from different referenced PCRs overlap until 2023. EPDs from different referenced PCR may not be comparable because of various reporting formats requested. Concerning the type of EPDs, 3 industry-wide EPDs and 16 product-specific EPDs were reported.
The name of the database for the life cycle inventory (LCI) and the software used were declared for most EPDs. 58% of EPDs declared the use of three or four databases and 26% of EPDs declared the use of one database as their LCI. The use of multiple databases to conduct EPDs adds an additional variable in play when comparing EPDs between them. Therefore, if a single and harmonized background database were used for EPDs, specific data such as EPDs could be considered as added data that would be more consistent with another generic dataset (Lasvaux et al. 2015). More in-depth information about the elementary flows used are not disclosed. This made it difficult to understand where the differences in impact assessment results between EPDs come from. As for the software used, the use of two different software can lead to significant differences between compared scenarios or products (Emami et al. 2019). Differences between database and software used can add variability and make it more difficult to compare EPDs. These findings were supported by previous work of authors who showed difficulties comparing EPDs (Del Borghi 2013; Hunsager et al. 2014; Gelowitz and McArthur 2017; Kerr et al. 2022). However, this present study aims to evaluate if this uncertainty in EPDs limits their usage in a WBLCA.
Impact Assessment Results from EPDs
The environmental impact results from the EPDs were extracted and compared to data from the Ecoinvent V3.8 database. The impact results presented and chosen for comparison are part of the main categories disclosed in the EPDs. Most of EPDs had reported GWP emissions disaggregated: GWP emissions and sequestration of fossil carbon (GWP-fossil) and biogenic carbon (GWP-biogenic). GWP-fossil results, acidification potential (AP) and eutrophication potential (EP) were compared to generic data from the Ecoinvent V3.8 database (Fig. 5).
Fig. 5. Impact assessment results for a) global warming potential (GWP), b) acidification potential (AP) and c) eutrophication potential (EP) by type of structural wood products
The use of box plots was adapted for datasets greater than 5 (Krzywinski and Altman 2014), which was only the case for the CLT and GLT EPD dataset. The box plots length is defined by the upper quartile and lower quartile of the sample. It is the interquartile interval which represent 50% of the data (Statistics Canada 2021). The median is indicated by a line in the middle of the box plot. Whiskers range from the lower and upper quartile up to 1.5 times the interquartile range. If data points are located outside of this range, they are represented by dots and can be considered potential outliers (Statistics Canada 2021).
When compared to their corresponding generic data from Ecoinvent, GWP reported in EPDs had a more wide-ranging difference than their AP and EP results. LVL EPD products tend to have higher GWP results than its corresponding generic data but do overlap it. On the other hand, CLT and GLT EPD results had lower GWP than the generic data, but the GLT box plot overlapped the corresponding generic dataset. Therefore, for CLT and GLT, GWP values for the generic data were more conservative than EPD results, with the exception of one outlier for GLT products. This finding was also supported by Strazza et al. (2016), who studied the usage of EPDs as a source of data for LCA for a specific case study of water bottles distribution onboard a cruise ship. The results suggested that the use of EPDs could avoid overestimation of potential environmental impacts (Strazza et al. 2016). However, LVL EPDs seemed to overestimate GWP results compared to its generic data. As for the AP and EP indicator, all box plot from the EPD results overlapped with its generic dataset. Therefore, using AP and EP indicators presents less variation of EPD data and the Ecoinvent database than the use of GWP indicator.
Considering the use of different PCR, database and software, direct comparison between EPDs is not recommended or should only be done with caution. Every whisker from EPD results overlapped its Ecoinvent generic dataset, with the exception of the GWP indicator for CLT products. The potential benefit of choosing a manufacturer from another is probably not as important as choosing a different kind of product in the context of a building project.
The use of EPDs as a data source also presented limitations. Access to background data is not disclosed in EPDs, which can limit the comprehension of the practitioners as to why a manufacturer’s product has higher impact results than others. Elementary flows are disclosed in generic dataset and can be modified to better represent the geographical location of the scenario. In addition, environmental impacts were not always disclosed in disaggregated form.
Applying CLT and LVL EPDs to the building’s LCA
A life cycle assessment was conducted on the hypothetical multi-story residential building with mass timber structure. The share of each life cycle stage for the environmental impact GWP is presented in Fig. 6. The three most contributive materials within the manufacturing stage are also presented.