Reimagining low-carbon futures: architectural and ecological tradeoffs of mass timber for durable buildings

Abstract

The urgency to rapidly reduce carbon emissions of the built environment make embodied carbon (EC), and thus material decisions, central to architecture’s most ambitious ecological goal. Structural systems are often the most durable and consequential to upfront EC in new construction. Although durability is critical to reducing EC in buildings in the long term, it may be at odds with the short-term goal to reduce resource consumption. This research closely and systematically examines the trade-offs between lower-carbon structural systems needed in the short-term and the durable systems needed to achieve long-term sustainability, functional adaptability and cultural significance. Specifically, this study evaluates the feasibility of using carbon-sequestering biomass to replace the more carbon-intensive structural materials that are more commonly used in buildings designed with extraordinary requirements of durability. The perceived conflict between durability and sustainability calls for more nuanced methods of analysis that consider the role of a building’s service life in EC reduction, and can augment the capacity of Life Cycle Assessment (LCA) to simultaneously consider the architectural impacts of material decisions. The methodology consists of fully redesigning the structure of an existing building with complex demands of sustainability and durability, and performing LCA for scenarios of equivalent architectural qualities, to retrospectively compare and analyze alternative low carbon futures in a context that only real projects can provide. The findings provide a more nuanced understanding of a near future when taller mass timber structures may leverage requirements for increased fire protection, robustness and durability to simultaneously achieve larger and longer-term carbon reductions.

Introduction

The construction and operation of buildings were responsible for the largest share (38%) of global carbon emissions from any sector of the economy in 2019 [1].The increase in operational energy efficiency of buildings, combined with higher energy use associated with producing materials and equipment to achieve those efficiencies [2], has prompted a shift in focus within the field of sustainability from operational carbon (OC) to embodied carbon (EC). EC is a measure of carbon emissions released during the manufacturing, transportation, construction and end-of-life phases of building materials. In contrast, OC measures carbon emissions associated with the use phase of buildings. EC contributes around 11% of all global emissions but was largely overlooked in the past. [3] The relatively long life of buildings was a reason why the use phase has historically been prioritized [4], but as this research shows, this longevity may also demand evaluating and targeting EC reductions and/or carbon storage. The urgency to reduce upfront emissions of the built environment, combined with EC’s growing share of the industry’s emissions, make material selection central to architecture’s most ambitious ecological goal of decarbonization. In the short term this means evaluating the carbon impacts of building new versus reusing existing buildings, prioritizing reuse of parts or whole buildings whenever appropriate, using less carbon-intensive materials for any new construction, and when possible, using materials that can store carbon. In the long term, designing new buildings now to be long-lasting, use-adaptable, and culturally valuable can ensure that future generations have a better opportunity of reuse to minimize the emissions associated with meeting their future needs.

Structural systems are usually the most durable and likely to remain unchanged during most renovations, but also the most consequential to upfront EC in new construction. Structures not only give form and stability to architecture, but also anchor its most lasting components in a physical and cultural place. [5] Differences in annual EC impacts (measured in units of kgCO₂eq / m²-yr) for buildings of different structural systems are significant early on—at the time of construction—but those differences decrease when analyzed over the life of long-lasting buildings (60–120 years) [6]. That is because the durable building not only replaces less parts, but also amortizes that initial embodied investment over more years (kgCO₂eq / year) and generations of people (kgCO₂eq / occupant of the building), reducing the overall carbon emissions of the built environment per capita per year (kgCO₂eq / (m²-yr-occupant)). Although durability is critical to reducing EC in buildings in the long term, it may be at odds with the short-term goal to reduce resource consumption. For example, while optimization of material use in structural cross sections and configurations is seen as an important path to lower EC [7] the increased attention to climate resilience may demand higher structural robustness and redundancy. Thus, to achieve a lower carbon future, architects and structural engineers must design buildings for durability while also considering the time value of carbon, i.e. the fact that carbon reductions are needed sooner rather than later. [8] That is because curbing carbon emissions now may help keep temperature rise to less than 1.5 °C above pre-industrial levels in time to avoid the worst economic, health and safety effects of climate change. [9] However, one of the most promising ways to reduce EC and sequester carbon using timber construction is challenged by questions about its perceived durability.

This research analyzes low-carbon futures by speculatively redesigning the structure of a high-carbon mid-rise building as various lower-carbon alternatives, including a mass timber version, to achieve equivalent architectural quality and capacity to satisfy uniquely complex physical and spatial demands. To compare and analyze the tradeoffs between durability and sustainability in these lower-carbon alternatives, the research uses Life Cycle Assessment (LCA) with methodological adjustments for various architectural factors that affect long-term performance. Specifically, the author examines an existing and architecturally significant steel building that was originally designed to some of the highest design standards of durability, quality, functionality, use-adaptability and sustainability—a United States federal district courthouse—and reimagines it in different lower-carbon structural systems, by applying emerging code and empirical data to meet the same challenging demands that only the context of a real and complex architectural project can provide.

Research Question: assessing mass timber as a long-term solution for a low-carbon future

The perceived conflict between durability and the time value of carbon calls for more nuanced methods of analysis that consider the role of a building’s service life in EC reduction strategies. This research proposes to examine closely and systematically the trade-offs between lower-carbon structural systems needed in the short-term and the durable systems needed to achieve long-term sustainability, functional adaptability, and cultural significance. Specifically, this study focused on the feasibility of using carbon-sequestering biomass to replace the more carbon-intensive structural materials that are more commonly used in buildings designed with extraordinary requirements of durability. The conceptual question that motivates this research is: could mass timber structures provide a feasible low-carbon alternative to high-carbon steel and concrete structures when projects must balance durability and sustainability, especially when these demands are increasingly pushed to new limits?

Mass timber, that is, engineered wood of large section size, including thick panels and large section glued- or block-laminated linear elements, is proving a viable and economic structural alternative to steel and concrete systems for increasingly larger and complex buildings [10]. As a material sourced from renewable biomass, it is also seen as a lower carbon alternative to help achieve the industry’s decarbonization goals. However, there are remaining challenges to its widespread implementation: the inertia of legislative and industrial practices, the limited availability of empirical data, concerns about adequate supply of timber from sustainably harvested forests, and the perception that steel and concrete are better suited for robustness and durability. Furthermore, many experts suggest that the industry has an overly optimistic view of wood as a carbon sink [11]. For decades, fire protection codes and industry practices favored structural steel and concrete for the larger or more critical structures, especially the scale of buildings and denser occupancies more likely to be found in the increasingly urban built environment. Very recent directions in performance-based design for fire protection [12], building code development [13], and experimental research [14] suggest that mass timber is becoming a more feasible alternative. In light of these challenges and opportunities, new research questions emerge: if timber supply is limited, what project types should be prioritized for sequestration using biomass? And how can designers make more meaningful comparisons between structural systems that balance many forms of architectural performance, taking into consideration environmental, technical, spatial, and functional, criteria?

These are the methodological questions that this study attempted to answer. First, the literature review builds an argument for why buildings designed for extraordinary expectations of durability are conceptually better suited for mass timber, and that the best approach to a limited global supply of timber is a focus on increasing building durability. Second, this argument is tested through a novel method of comparison, a hybrid of qualitative and quantitative research methods, to assess life cycle impacts through the lens of various forms of architectural performance, and more comprehensively evaluate the feasibility of lower carbon alternatives. By applying the same durability criteria (structural robustness, structural and spatial redundancy, future-use adaptability, functionality) to various versions of the same design in four primary structural systems (steel, reinforced concrete, mass timber and brick masonry), this research shows how quantifying their EC can take into consideration architectural tradeoffs (space, daylight, proportions, mass, volume). The findings of this research point towards a near future where taller mass timber structures may leverage requirements for increased fire protection, robustness and durability to simultaneously achieve more rigorous standards of sustainability, creating more volume for carbon sequestration that can reliably remain within a long-life building.

Prior work

Professional practitioners use Life Cycle Assessment (LCA) to evaluate entire projects against industry benchmarks or against various alternative design scenarios, usually with the goal of reducing environmental impacts of construction, operation and end-of-life processes in buildings. LCA accounts for the environmental impacts of all life cycle stages from material extraction to manufacturing, transportation, construction and deconstruction. Isolating and comparing the analysis of the EC of structural systems can provide important insights into buildings’ most significant contributor to environmental impacts and can challenge design teams to pause and question more deeply the relationship between architectural concepts for form and space and the primary structural systems that can support them. However, structural systems are usually selected early on during the design process, based on building size, occupancy, spatial and formal requirements, and budget. These schematic design-level decisions follow rules of thumb and conventional industry practices in a region, fixing certain variables in a project early on, and therefore rarely subjected to an extensive analysis of alternatives for such systems. This makes high carbon materials (steel and concrete) the default in commercial and institutional buildings, whether due to the predictability of established regional practices, code limitations and/or perceptions of higher performance. While use of mass timber for larger structures is growing, it is still considered experimental or groundbreaking, which means that in conventional practice the realities of limited schedules, limited budgets and limited experience may continue to favor the selection of higher carbon materials before lower carbon alternatives can be properly and comprehensively considered. However, experimental mass timber projects, often using performance-based design approaches, have led to the expansion of height and area limitations in emerging codes, and are likely to dramatically change the industry in the next decade. Many practitioners, seeing these upcoming changes and facing the pressure to reduce carbon emissions in their projects, are wondering how to evaluate mass timber in larger and more critical projects, and whether it is as good as it sounds. There is a crucial opportunity for architectural research to generate new methods and knowledge that can quickly inform the more time-constrained, project-specific decisions of conventional practices.

Researchers use LCA to evaluate the environmental impacts of buildings, albeit not necessarily in the service of one idiosyncratic project, but to advance generalizable knowledge about specific materials or industrial processes, identify trends or predictable patterns across many projects or products; and even to evaluate the impact of existing methods of analysis or to test new methods that advance the practice of LCA itself. The need to compare results and the dissemination of LCA work prompted standardization of the process, especially its phases: (1) goal and scope definition, (2) inventory analysis, (3) impact assessment, and (4) interpretation. [15] However, these standard phases still allow significant variation in 4s. For LCA of buildings, some studies focus exclusively on the design of the primary structure, showing, for example, that increases in design loads can cause a fairly consistent percent increase in EC or that EC can increase linearly with building height, if lateral loads are not considered. [16] Whole Building LCA (WBLCA) analyzes the impacts of the whole building, rather than the impacts of only components or systems, even if the researcher is examining the relative impact of one system on the whole, making this method increasingly important in structural design [17]. A recent review found that most studies in the literature confirm some expected rules of thumb about the EC of structural systems, e.g. that wood always produces less carbon emissions than concrete [4]. The same review noted a similar trend for wood vs. steel, citing one rare exception of a steel system of high optimization, durability, and deconstructability that resulted in less material use with the ability to be reused [18].

Research studies focused on structure can vary significantly in both the approach to structural design and to LCA, as summarized in a recent review that called for more transparency and clarity in research reporting to overcome the “stresses placed on LCA by methodological variation, data quality and uncertainty.” [19] Many of these studies generate useful results by comparing multiple structural designs, using parametric design to quickly generate many variations, and observing trends in quantitative results. While useful to identify general relationships between emissions and some material or configurational decisions, this type of computational study cannot consider the more complex interplay of quantitative and qualitative criteria present in real and complex projects, and therefore cannot provide insight into the potential tradeoffs with other forms of architectural performance (functional, spatial, etc.). This is also due, in part, to the more standard approaches to LCA that normalize the data by functional units of gross area, which cannot consider the spatial efficiency of structural systems or the adequacy of structural patterns to meet current and future use requirements in long-life buildings. Therefore, there is a need for new hybrid methods that augment the capacity of LCA with measures that approximate architectural impacts of structural design decisions and result in more holistic approach to compare alternative structural systems in specific projects.

Case study research supported by speculative design work can have the advantage of retrospectively comparing and analyzing alternative futures, i.e. to imagine and examine the advantages and disadvantages of what could have been low carbon alternatives for already completed buildings in full consideration of the many competing and conflicting goals that only the context of real projects can provide. This research starts with an in-depth analysis of a case study— the US District Courthouse in Salt Lake City, Utah, United States (US) (Fig. 1), a steel-frame building designed for high energy performance, future-use adaptability, and durability—including resistance to PC; and reimagines alternative material and tectonic responses to the same architectural concept to assess what would have been their ecological impacts. The case study was originally identified as part of a large data set of over one hundred case studies documented for an earlier and much broader Grounded Theory project, which consisted of analyzing drawings and text to extract new knowledge about designing for future use adaptability. [20] The focus of that early work was not specifically structural, but compared many design attributes, including spatial, circulation and fenestration patterns from plans, sections, and elevations; and explored aspects of the design process for long term buildings based on interviews. Nonetheless, structural systems were found to define many of those patterns and processes—a critical system to durability, place specificity and use adaptability. [21] Those findings inspired new questions that are part of this study and other new lines of ongoing research. Specifically, four of these many case studies were preliminarily analyzed with LCA for an exhibit focused on Durability. [22] The four projects were selected because they had similar aspect ratios and each represented didactic experiments on using one of four primary structural materials (concrete, steel, brick masonry and wood) for exceptional criteria of longevity and adaptability. LCA data for the four projects was normalized in various functional units: gross floor area, net floor area, volume, occupant (number of persons occupying the building), and area-year, (the combination of floor area and the number of years the building is in service) as part of various analytical experiments [23]. In that study, the more standard units of kgCO₂eq per gross floor area and kgCO₂eq per net floor area showed that the masonry and concrete buildings, which had the lowest and second lowest surface-to-volume ratios respectively, had the highest EC; followed by the steel building, then the wood building. While the results were clearly didactic for an educational exhibition, the generalizable knowledge that can be extracted from that comparison was limited by the fact that these four buildings had different functions, different volumetric proportions, and different locations around the world, and thus different performance criteria. With the knowledge acquired over four years about these buildings, there was an opportunity to develop more generalizable data by instead focusing a study on one of those four uniquely long-life buildings; modeling it in four different materials, in the same location, with the same function. That was the genesis of this investigation, and the results will offer different and more nuanced insights than that earlier comparison.

Fig. 1

Exterior view of the southwest corner of the Salt Lake City federal district courthouse (left) and photo of glazing frit pattern and louver system seen from the interior (right). Photographs by Michelle Laboy

Selected case study & design parameters

The steel building that was selected for this new phase of the research, was the highest contributor of EC of the four early case studies when normalizing the resulting kgCO₂eq “per person” (or per occupant, i.e. dividing the total emissions by the total expected of human occupancy of the building). This is due to its civic use as a federal courthouse, which required generous assembly spaces and redundant circulation spaces that result in a low density of occupation. It is also the tallest of the four buildings at 10 stories and 53 m (174 feet) in height; the most representative of common building practices for mid- and high-rise commercial buildings in the US (glass curtain wall on a steel frame with concrete slab over steel decking); and designed for some of the most stringent structural requirements, specifically higher design loads for security and adaptability. The award-winning and anticipatory design for the Salt Lake City courthouse, as explained by Thomas Phifer Partners, had to satisfy a highly specialized program and yet be designed for future-use adaptability. [24] It also had to meet the sustainability standards of the General Services Administration (GSA) during President Obama’s administration (mostly focused on operational energy performance as opposed to embodied energy) while meeting the federal government standards for durability, functionality and security of courthouses. [25] The façade is all glass, which would have had a negative impact in operational energy but was compensated for with a frit pattern on the glazing and a veil of vertical aluminum louvers that cover most of the four elevations to control solar heat gain and glare, and limit UV damage (Fig. 1). As will be seen in the findings, the result of this robust system of enclosure to minimize operational energy was a high EC. While fixed, the angle of the louvers is different in each orientation, according to the predominant solar exposure on each side of the building. The geometry allows the façade to completely protect the structure, with no primary structure projecting outside of the thermal envelope, minimizing condensation potential. Furthermore, most of the functional spaces of the building are set back from the façade, protected by a buffer of circulation space that surrounds perimeter rooms with transom windows tall enough to allow a diffuse glow of natural light to penetrate the interior (Fig. 2). This perimeter circulation ring surrounding the courtrooms, which are located on the four corners of the main floors, is a response to the courthouse program’s unique circulation requirements, demanding four distinct entrances and systems of vertical movement for each of four groups: judges, public and staff, prisoners, and deliveries. The restricted circulation ring for judges happens on an approximately 1 m (3.3ft) wide overhanging space between the perimeter moment steel frame and the edge of the slab carrying the façade. That perimeter moment frame enabled all interior columns to be pinned connected, and all interior partitions of the building to be non-load bearing, including the fire rated cores, to maximize adaptability for future changes in use. The minimum clearance required for this circulation and façade zone became critical to the building function, therefore this research sought to model it consistently in the alternative structural models, and as will be explained further, became an important element in differentiating their carbon emissions.

Fig. 2

Photographs of access to the buffer / circulation zone for judges next to facade (left) and and view of transom windows on the interior facade of courtrooms seen from the judges’ circulation zone (right). Photographs by Michelle Laboy

Due to its government use, the design had to consider security threats, which meant designing the structure for progressive collapse. Progressive collapse (PC) is a rare event, usually triggered by a local failure due to terrorism, but when it happens, the cascading failure of the structure that follows loss of a component can be the reason behind most casualties. The federal US government added resistance to PC to its design guidelines after the 1995 attack on the Oklahoma federal building, and it cites the “recent escalation of domestic and international terrorist threat” as the motivation to continue updating the criteria for PC for buildings taller than three stories. [26] Two direct design approaches include an Alternate Path method (AP), which requires the structure to bridge over a missing structural element, and the Specific Local Resistance method (SLR) which is a hardening method adding strength for members to resist specific types of loads. Both of these can result in larger structural members of additional strength, when compared to a building not designed for resistance to PC. The federal guidelines require AP, a form of SLR or enhanced local resistance (ELR) in the form of flexural and shear resistance (at perimeter columns), and tension ties that allow the floor to transfer the portion of unsupported floor to other members. As will be explained, these requirements for robustness and redundancy were proxies for durability in this analysis.

The architects and the GSA project manager, who were originally interviewed for the earlier grounded theory project, were asked follow-up questions specific to this new phase of the research, including about material selection and structural details to re-model the project with more precision in Revit®. The architect explained that steel was chosen for the primary frame because the local market at the time favored steel over concrete for high rise structures, and due to the requirement to design against PC (he explained the latter could have been achieved with concrete as well). The building used a proprietary rigid steel connection called Sideplate® for the lateral system (perimeter moment frame) instead of traditional welding methods, which has schedule and cost advantages. These three factors related to the structure—moment perimeter frame, resistance to PC, and regional regulations and industry practices, would have been significant challenges to implementing a mass timber structure at the time that the Courthouse was built (2014). Moreover, building codes for buildings of this height (53 m or ~ 175ft) and floor area (2,695 m² or 29,000 ft² per floor) would not have allowed a timber structure to be used for this occupancy classification of assembly space.

Since then, newly proposed building code modifications that would enable taller wood structures were approved in 2018, specifically the 2021 edition of the International Building Code. Even though as of the time of this publication the new code has not been officially adopted in any but one US jurisdiction [27] it is expected to be adopted in many states and urban regions soon, although potentially with local modifications as is standard practice. These new codes would enable another critical design consideration in this study of EC: the exposure of the unprotected primary wood structure (i.e. columns and beams) in higher risk buildings (i.e. larger buildings with more restrictive occupancies), if the cross sections are designed for the required fire resistance rating. This can be achieved with additional protective thickness of timber over what is required for structural purposes on each exposed side, based on applying the well documented rate of advancement of the insulating char layer (approximately 36 mm (1.42 inches) per hour) [28] to provide the minimum hours of fire resistance required. Furthermore, in the last decade advances in building technology research and manufacturing for mass timber structures spurred new experimental studies on design of timber structures for resisting PC. Those experiments suggest that, despite the brittle behavior of wood, certain configurations of composite floor systems of engineered wood beams and Cross Laminated Timber (CLT) can achieve the robustness and the catenary action necessary to allow the high deformations in a design when removing a column, an essential mechanism to resist PC. [29] That experiment included the design of custom aluminum plates to allow the necessary rotation at the beam connection. In larger buildings, these advances require more robust, larger cross sections, or more wood material. This study modeled the alternative timber scenario with the added wood thickness and aluminum connections, to create a design more appropriately equivalent to the steel system.

Method

This research is focused on a methodology of comparison, that is, how to account for differences in carbon emissions between equivalent structural systems, taking into consideration architectural differences in spatial efficiency, quality and functionality. This method is starting from the knowledge that steel is the most space efficient structural material (same spans with smaller cross sections to carry the same service loads), often making it more desirable architecturally. Furthermore, most lower carbon alternatives require larger cross-sections, i.e. more volume and/or weight of materials, that challenge the quest for the most slender aesthetic in much of contemporary architecture. The method assumes that the ultimate goal of an architecture concept is space, which is governed by specific critical and useful dimensions that are carefully and economically calibrated: clear open floor area and ceiling height that define functionality (e.g. access), aesthetics (e.g. lofty proportions) and comfort (e.g. daylight) [30] such that any modification of those clear dimensions to use higher volume materials is considered a loss. In this case study, that meant exploring whether different functional units of normalization (net volume instead of net area or the more common gross area in the denominators), or adjustments to the building design (increase building size to achieve the same volume) can better account for the impacts of more space-consuming structures. This method also considers disruptions to function, e.g. program-specific circulation, in the case of this courthouse it would be the net clearance for the perimeter circulation ring for judges; as well as serviceability, i.e. maintaining the plenum space in between and below deeper structural cross sections as a proxy for undisrupted distribution of building services infrastructure.

The analysis leveraged new empirical data and rules of thumb to model alternative mass timber structures, one that keeps the same size building and accepts the loss in net volume, and one that expands the building size to achieve the same net volume. For the purpose of obtaining more data and validating comparisons, additional material alternatives were modeled and analyzed. The first part of the analysis included five models, all of the same overall gross area and building height, including alternative scenarios using the lowest carbon materials available for each system (Fig. 3).

• 1. Original steel: steel primary structure with conventional reinforced concrete floor slab on steel decking

• 2. Modified steel system: same primary structure as the original, with a modified concrete slab mix that substitutes the maximum 50% of Portland cement with alternative cementitious materials (ACM: 30% slag and 20% fly ash)

• 3. Brick masonry system: brick piers with a hollow reinforcement core filled with cement grout, and same floor system as the modified steel building (steel beams and ACM concrete slab)

• 4. Reinforced concrete primary system, with a flat concrete slab, using 30% slag and 20% fly ash for 50% of the Portland cement content for all concrete.

• 5. Mass timber system: Glu-lam columns and beams with CLT floors, with a non-structural topping slab of lightweight concrete using ACM.

  After comparing net area, net ceiling height, and functional disruptions to circulation, each alternative building model (except the modified steel system which only involved a change in concrete mix) was increased in size accordingly, by either a larger footprint area to maintain the circulation perimeter, and/or a larger floor-to-floor heights to maintain the same floor-to-ceiling heights below the same mechanical distribution plenum as the original building, and to achieve the same daylight penetration and views. That resulted in three additional models to analyze, with not only larger volumes and mass of structure, but also added surface area of enclosure:

• 6. Taller concrete building

• 7. Wider masonry building

• 8. Taller and wider mass timber building

Fig. 3

Axonometric drawings of three structural bays on one typical floor at the corner of the case study building, showing configuration and size of existing and alternative structural scenarios

Structural Modeling

Based on project documentation and interviews with designers and owner representatives, the author modeled the existing structural system (a post and beam frame construction with no shear walls) and equivalent alternative systems listed above, for the same criteria (tributary area, number of floors, centerline-to-centerline spans, load paths, and live loads; but with different dead load for the self-weight of slab, beams and columns). The structural design of the system considered gravity-only loads to understand spatial impacts, assuming criteria for PC resistance (Fig. 4). It is assumed that the perimeter moment frame connections can be achieved in all models with the extraordinary size of columns and beams necessary for PC resistance, by fitting adequate metal connectors, which are modeled in the applicable scenarios.

Fig. 4

Assumed structural floor plan (not an exact representation of the actual building construction, but a likely scenario) indicating column removal scenarios and corresponding tributary areas used for design against progressive collapse

Due to the GSA security restrictions, only architectural drawings were available, which showed scaled graphic representations of column and beams in typical plans and sections; but no construction documents with specific structural sections and calculation procedures were available. Therefore, the author used assumptions based on construction time lapse [31], publicly available guidelines, building occupancy and spans, to reverse engineer the original structure, using spreadsheet models based on a simplified version of a linear static analysis procedure [32] including parameters for PC. When the calculated sizes approximated the graphically scaled size of steel members in drawings, it validated the assumptions and procedure to be applied to concrete, timber and masonry. The design assumptions included:

• 1. Live loads for assembly spaces (100 lb/ft², or 4.79kN/m²) [33], local snow loads (43 lb/ft², or 2.06kN/m²) [34], and dead loads based on densities for different specified materials.

• 2. Risk Category III [35] which requires AP method for specified locations, and ELR at ground floor perimeter columns [36]

• 3. Force controlled, gravity load increase factor Ω_LF = 2 [37]

• 4. Specific to the reinforced concrete design: tension controlled for ductility.

• 5. Specific to the mass timber design: two-spans for CLT, as discussed on preliminary PC experiments [38]

• 6. Slenderness limits for constructability and lateral stability specific to different materials: Steel (kL/r < 200), Timber (L/d < 50), Masonry (h/w < 25), Concrete (L/r < 22).

For models 6–8, where the overhanging edge of the floor plates and the floor-to-floor heights increased, the cross-sectional dimensions of perimeter columns and beams were verified using the resulting increases in enclosure loads and effective lengths (column slenderness).

LCA Modeling

The eight versions of the building were modeled in Autodesk® Revit®, with variations of structural materials and sizes as described above, as well as variation in surface area of enclosure for the taller and/or wider building volumes (as will be explained in the findings section, including summaries in Tables 1 and 3). The Tally® plug-in performed quantity take offs of material volumes for structure and enclosure, assigned impacts based on project location and embodied carbon dioxide coefficients for materials in its databases, and provided outputs per life cycle stage, including product stage (A1-A3), transportation (A4), construction installation (A5), maintenance and replacement (B2-B5), operational energy (B6), end of life module (C2-C4) and Module D (reuse or energy recovery potential outside of the system boundary), as well as totals [39]. For this study, all stages were included except operational energy (B6), since the focus is only on the global warming potential of materials, or EC. Module D was considered but always separated for further analysis. Whenever available, materials with Environmental Product Declarations (EPD) were selected.

Table 1 Summary of LCA results. Embodied carbon emissions normalized by gross floor area (for existing steel building and all other re-design scenarios) for both 60 and 150 years, with and without including the impact credits from Module D. Overall the values increase towards the top right of the chart and decrease towards the bottom left

This method of connecting a BIM model with an LCA plug-in is time consuming but has some advantages, i.e. it can model specific geometries and structural configurations, instead of parametric design of generic models. This is particularly important when considering specialized requirements, such as design against PC, which involves engineering judgment in the removal of a column for analysis. For LCA this has may an advantage over software that assumes general assemblies based on user input, such as Athena [40], although for this study the results from both software options were not compared. The assumption is that software like Athena may not fully capture the complexity of specialized systems and connections that were modeled and accounted for in the BIM model for this study, such as the aluminum plates in all the timber building column to beam connections needed for resisting PC. However, the precision of volume takeoffs in BIM models can still be subject to human modeling error. Significant effort was made to verify that the modeling was accurate and consistent from one model to the other, and that materials were assigned correctly, but minor errors that were not visible to the research team cannot be ruled out.

To make the models truly equivalent, all building versions were designed to achieve the same required fire resistance rating on the primary frame, as required by the construction type and occupancy. The most restrictive construction types, per International Building Code, were determined for the different structural systems based on floor area, occupancy, height, and number of stories, measuring to the roof plane, assuming the tall non-combustible curtain wall parapet that covers the mechanical equipment on the roof will continue to be allowed to be of unlimited height above the roof (as currently stated in the International Building Code). Although the models only included structure and enclosure to limit the study to the more long-lasting components of the building, some non-loadbearing partitions were included in the model if they were required for fire rating (shafts and column enclosures). Other program-specific partitions and finishes were not included, due to the low durability, uncertainty of future program, and the relatively low control architects typically have on those systems in early design. The required fire resistance rating is inherent to masonry and concrete, but in the case of steel, steel columns were modeled in Revit® encapsulated with 2 layers of US standard 15.9 mm (0.625 inches) gypsum board to achieve the code-required 2-h rating in the primary frame for Type I-B construction, per typical details approved in accordance to UL laboratory tests [41]. The steel beams, which are behind the ceiling finish and not visible in the space, were assigned a cementitious spray coating in Tally®. Wood was modeled to be a Type IV-B construction type—a new heavy timber type added in recent building code regulations, which also requires a 2-h resistance of the primary structural frame—by including the additional thickness to create the insulating char layer in the cross section of glulam columns and beams [42].The lightweight concrete topping slab allows CLT ceilings to remain exposed; and was included to make all systems of equivalent architectural performance in terms of fire, vibration and acoustic separation.

The foundations were modeled in Revit® but not accounted for in the LCA, given their site specificity, which is believed to make the LCA results more broadly generalizable. However, the mass of the structure was calculated to discuss potential differences in foundation size, which would impact the amount of concrete and its related emissions. Lastly, the LCA was performed for building service lives at the most common 60-years [43] to compare with benchmarks in the literature; as well as 150 years, to further consider the impacts of more extraordinarily long-life buildings.

Analysis

The results of the original and all redesign scenarios are summarized in Table 1, normalized by the more commonly used functional unit of gross area (kgCO₂eq/m²) in order to compare to material-specific benchmarks used in most studies in the literature. The analysis that follows compares the relative differences between scenarios using this and other units of analysis to overcome limitations of LCA methods in addressing specific architectural performance criteria.

Gross Area

Figure 5 shows the LCA results of the first five models (same building footprint and building height) compared to a recent benchmarking study of buildings of all types, which reported a median of 384 kgCO₂eq/m².[44] Comparing this benchmarking median to the models for the courthouse studied at the commonly used building service life of 60 years (without the less commonly used credits from Module D), to use the more common LCA parameters reported in the literature, only the wood building had a lower EC, at 269 kgCO₂eq/m², followed by a significant jump up to numbers ranging from 475 kgCO₂eq/m² (concrete) to 595 kgCO₂eq/m² (steel). This is a notable difference for durable buildings in the context of the time value of carbon given that, as will be further explained in the following sections, a significant portion of those emissions from the wood building would have been deferred to the end of its service life.

Fig. 5

Embodied carbon emissions normalized by gross floor area for existing steel building and alternative building models with modifications to structural system only (no changes to building size). Each model shows results for building service life of 60 years (right) and 150 years (left). Each bar shows the total without Module D (top) and after subtracting Module D (bottom)

For more material-specific comparisons, the EC of the wood model of the courthouse fell within the range of wood buildings reported in another benchmarking study (100 to 400 kgCO₂eq/m² for timber), but above the median of 200 kgCO₂eq/m² [45]. However, that same study reported medians of between 300–400 kgCO₂eq/m² for concrete, steel and steel/concrete hybrids, respectively, which are lower than the courthouse EC in those same materials. The widest range seen in these benchmarks was for steel projects, which go as high as 1200 kgCO₂eq/m². This may relate to findings in the same benchmarking study, which found cultural buildings and LEED-certified buildings to have higher embodied carbon. These are relevant factors for the courthouse structure, due to the robustness and redundancy necessary to meet more strict performance criteria for durability, security, and operational energy efficiency in this type of public assembly building. This may explain why its emissions are higher than the reported material-specific medians by factors of 1.35 × for wood and 1.7 × for steel Notably, when the building service life is increased to 150 years (without Module D), the range of EC for the courthouse goes from 745 kgCO₂eq/m² (wood) to 1071 kgCO₂eq/m² (steel). When comparing these to recently published benchmarks specifically for public assembly buildings, which are presumably more durable types and have a higher median of 433 kgCO₂eq/m², and a range from 100 to 935 kgCO₂eq/m²; all the models of the courthouse at 150 years, including the wood building, are higher than the median by between 1.72 × and 2.47x. It is again worth noting that for wood nearly half (42%) of those emissions are deferred to the end of its service life, which at 150 years would put it comfortably outside the range of the current timeline for climate action.

Steel has higher emissions than all other materials in all scenarios, even when replacing Portland cement in the concrete slab with ACM. The ranking of materials does not change from 60 to 150 years:

$steel>steel\;with\;ACM\hspace{0.17em}>\hspace{0.17em}masonry\hspace{0.17em}>\hspace{0.17em}concrete\hspace{0.17em}>\hspace{0.17em}wood$  

However, the relative impact of building service life on emissions is more significant in the lower-carbon structures. The emissions of the wood building nearly triple from 60 to 150 years. They double in the concrete building, and nearly double in the masonry and steel buildings in that time. While the durable building may still result in less emissions than demolishing the building and building new, this significant increase in emissions when going from 60 to 150 years demonstrates the relative impact of replacement and repair phase in long-life buildings, highlighting the importance of cradle-to-grave, or cradle-to-cradle, rather than cradle-to-gate analysis.

While the comparison with common published benchmarks is helpful, the disadvantage of normalizing by gross area is that it fails to account for the loss of net usable space when changing to different primary structural materials. This is especially relevant to long-life buildings where large spans, robustness and redundancy enable durability, safety and adaptability, but the spatial impact of replacing steel with materials of lesser strength, such as wood, can be significant.

Net Area vs Net Volume

For this analysis, the net area reduced the gross area by the thickness of enclosure and floor voids (equally in all models) but also by the footprint of columns with their fire rating (differently in each model). The wood columns have additional charring thickness and the steel columns have the additional thickness from the channels and gypsum enclosure (Fig. 6). This eliminates the advantage of the steel I-shape, which reduces less area than its equivalent rectangular footprint. And it reduces the penalty on the more massive but inherently fireproof systems of concrete and masonry. However, normalizing by net area would only capture losses in floor area, not in height, neglecting the full interior volume (spaciousness) of the architecture, and as such the aesthetic experience (daylight, monumentality, proportions). Specifically, there is also a significant added structural depth for a robust floor structure that can resist PC in wood and concrete. When the ceiling height and floor area are reduced, the net volume is reduced. Thus, the LCA comparison for the original five models was normalized by net volume (kgCO₂eq/m³) which has the effect of increasing carbon emissions per unit. It is, in other words, a penalty on the more space-intensive materials.

Fig. 6

Column cross-sections for alternative models compared to original steel building, including required perimeter thickness for fire resistance rating where needed

Figure 7 shows the LCA results for each model by stage, without and without the impacts of module D. Module D accounts for the reuse or energy recovery potential of materials at the end of life, and it is considered outside of the system boundary because these savings are not fully realized within the service life of the building being analyzed, but in the life of another end use. In this analysis, when excluding module D, the ranking of materials changes significantly, with the space-intensity of the masonry structure making it now the highest, and the steel building EC is now closer to the wood building, 30% higher instead of 40% when compared by area:

Fig. 7

Embodied carbon emissions (for a longer-life 150-year building) normalized by net interior volume (kgCO₂eq/m³) for existing steel building and alternative building models with modifications to structural system only (no changes to building size). Each model’s emissions are shown by stage without module D (left) and the net total after subtracting the credit for module D (right)

$masonry>\hspace{0.17em}steel\hspace{0.17em}>\hspace{0.17em}steel\;with\;ACM,\hspace{0.17em}>\hspace{0.17em}concrete\hspace{0.17em}>\hspace{0.17em}wood$  

When considering the impact of reusing or recovering energy in module D, then steel is again the highest. However, masonry which has the same floor system as the steel building with ACM slab, is now between steel and steel with ACM.

$steeel\hspace{0.17em}>\hspace{0.17em}masonry\hspace{0.17em}>\hspace{0.17em}steel\;with\;ACM\hspace{0.17em}>\hspace{0.17em}concrete\hspace{0.17em}>\hspace{0.17em}wood;$  

By having the same floor system as the steel with ACM building, this analysis isolates the relative impact of the vertical masonry system. Nonetheless, the results reveal that the proportional impact of the concrete slab on a hybrid building is significant, going from 13.1% of the steel building’s total EC to 10.5% of the masonry building’s total EC, to 9.75% in the steel building with ACM. Even the lightweight concrete topping slab over the CLT of the mass timber building represented 6.1% of its total EC.

Seeing the impacts by stage in Fig. 7 makes evident that the end of life (EoL) stage in wood is proportionally much higher than the other materials, and that is because when wood comes out of the building it is effectively releasing the carbon it had sequestered. Studies suggest that most glued laminated timber (80%) at EoL goes to landfill [46], a much higher number than steel (2%) or concrete (35%). EoL includes the emissions from processing waste and from decomposition in the landfill. On the other hand, Module D gives credit to the wood building by assuming use of landfill gas or energy recovery, which is believed to be overstated especially because of the difficulty of making assumptions about processes, market demand and activities that will happen more than 100 years in the future, when new technologies may make the benefits of energy recovery much lower.

Normalizing through building volume adjustments

Normalizing by volume may penalize the more space- and carbon-intensive materials (such as brick), but it still creates a false equivalence between models in a comparison of options for the same building. One way to make the buildings more functionally equivalent is to increase the size of the different building models to offset any losses of function or architectural experience. However, indiscriminately expanding the floor area to equalize the net area among all models may be wasteful, increasing structural spans without a clear need. To better consider the loss of functional area, the plans with the alternative column sizes were visually analyzed against the current layout of rooms. This requires judgment that cannot be provided by algorithms or quantitative models. Due to the generosity of public spaces in this civic building, for the most part structural columns fell into zones where the additional thickness had no functional impact on the space, other than visual. The only noticeable impact in some alternative models, other than some reduction in shaft openings, was the reduction in net clearance for circulation, in particular around the perimeter corridor for judges. Elsewhere, any reductions in net clearance could be managed by shifting the column grid minimally, without impact on spans.

In those cases where the larger columns interfered with a minimum clearance for circulation (wood and masonry only) the building was expanded outwardly by the distance necessary to maintain a 1 m (3.3 ft) clear space between the enclosure and the column (Fig. 8). In the LCA model, this resulted in a larger floor material volume and additional curtain wall area. In this case the buildings are not the same gross area, but they are more similar in terms of function. To also account for the ceiling area lost to the additional depth of floor structure, the wood and concrete buildings were also expanded vertically to maintain the same net clearances and floor-to-ceiling height. This maintains the same height for daylight and interior proportions, but also the same space of plenum for mechanical distribution. The difference in elevation, seen in Fig. 8, creates a more dramatic structure. Furthermore, if the aesthetic of a purely white ceiling was essential to the architectural concept, this approach does not compromise that idea. The LCA for these models accounted for the additional curtain wall area for the taller building.

Fig. 8

Floor plans and elevations of the original steel building as designed and built are shown on the left, followed by the three scenarios of the same building in different structural systems, after adjusting the size of the building to achieve the same net floor area and ceiling heights. Dashed lines across references the alignment with the original steel building floor levels and footprint

The results for this analysis (Fig. 9) show again the original ranking of:

Fig. 9

Embodied carbon emissions (kgCO₂eq) of original building and alternate models after building volume adjustments, by life cycle stage, at building service lives of 10, 60 and 150 years

$Steeel\hspace{0.17em}>\hspace{0.17em}masonry\hspace{0.17em}>\hspace{0.17em}concrete\hspace{0.17em}>\hspace{0.17em}wood$

but with notable differences in relative impact across the years. As the building life increases, the difference in emissions between the steel building and the wood building is reduced much more rapidly than between the steel building and the others, from a ratio of 2.12x (10 years) to 1.99x (60 years) to 1.43x (150 years). This is followed by the closing gap between steel and concrete. The closing gap between steel and masonry is fairly insignificant. The main reason for this closing gap are the emissions associated with the B2-B5 stage of replacement and maintenance. To be clear, these models assume the structural components have the same life as the building in all scenarios, i.e. a robust structure that is completely protected by the enclosure and therefore is not replaced. While there are some minor differences in the maintenance of the different structural systems, for example, reapplying coatings in the wood structure, or repointing the mortar in the masonry structure, these are insignificant in contrast to the relative impact of the enclosure. The curtain wall replacement happens every 60 years, thus its impact only affects the LCA for service lives that are longer than that (150 years in this study). For high-carbon enclosure systems with shorter service lives than the structure, e.g. the courthouse’s curtain wall + louver enclosure replaced twice in a 150 year life—the emissions from multiple replacements accumulate enough to surpass the EC of the structure (Table 2).

Table 2 Change in relative contribution of structural and enclosure components to the LCA in longer-life buildings, for a service life of 60 years and 150 years. These numbers are based on the LCA of the existing steel + concrete structure of the courthouse case study, assuming standard life spans for the enclosure components (curtain wall and louvers) but matching the life of the structural systems to the life of the building

After adjustments to building size, the concrete and wood buildings have the two largest areas of enclosure, in that order. As a result, the emissions for the maintenance stage are the first and second highest in these two buildings, respectively. The masonry structure required a much smaller increase in enclosure area, primarily because the floor system was the same as the steel building, and therefore it did not require an increase in height. As a result, the maintenance-related emissions are proportionally much closer to the steel building. The larger upfront investment on the enclosure of this building, and its many replacements simulated over a much longer life of 150 years, also explains why the gap between wood and steel closes more quickly in this method of comparison using adjusted building size, than in the gross area comparison. This method of analysis demonstrates that the interaction between structural system selection and quantity of enclosure-related materials can play a significant role in clarifying the difference in EC between alternative designs. Furthermore, it highlights the importance of moving towards Whole Building LCA aided by qualitative design adjustments, as opposed to analyzing structural systems independently and parametrically.

Findings

The three different methods of comparison illustrate the importance of considering architectural criteria of spatial and functional quality. This research shows that LCA can evaluate emissions of different structural systems while considering the full architectural concept. Changing the size of the building to achieve the same spatial quality and functionality yields a more authentic comparison of scenarios for a particular building. These adjustments are necessarily project-specific, partly qualitative and require architectural judgment. Once the scenarios are made truly equivalent, using units of total carbon emissions instead of normalizing units provides a more conservative comparison. Table 3 shows the summary of results for the eight models using total (non-normalized) units of carbon emissions for all models in this study, including the original size building and the expanded building. In order to track spatial impacts, the table shows the loss of net / gross area, and the necessary increases in footprint and enclosure necessary to achieve equivalent architectural space. Wood requires the highest increase in enclosure area, even higher than concrete, to compensate for the loss of volume, in part due to the additional thickness required for fire resistance rating. Although not the focus of this study, it is important to qualify that the change in size also changes the compactness of the building (surface area needed to achieve the same net volume), and that this change would be expected to have impacts on operational energy (OC).

Table 3 Summary of differences in space, mass and carbon emissions for all eight models

The table also shows the differences in structural mass of the superstructure, to understand the potential magnitude of increase on seismic loads as well as impacts on the substructure, even though these were not part of this initial study. Inertial forces, which usually do not include live loads and consist mostly of the dead loads dominated by building structure [47] are directly proportional to mass, i.e. greater mass generates proportionally greater lateral loads for the same ground acceleration [48]. In fact, historically the earliest methods of design for lateral loads used to be simplified as percentage of the weight of the structure above a particular story, ranging from 2–12.5%, until new methods of dynamic analysis became available, but these static methods are still considered the most important advance, and are still used in some simple regular structures [49]. For the purposes of this analysis of high level tradeoffs, using this simplified form of analysis when looking at the weight data shown in Table 3 suggest that wood construction would result in inertial forces approximately 50% of those on the existing steel building, which simplifies the structural design, and can increase the advantage of wood on carbon emissions if the design accounts for the effects on the size of foundations from both lower seismic base shear and lower total gravity load.

Looking at all eight models as absolute total EC, it is evident that the simple change from steel to concrete in beams and columns (without compensating for open volume) can appear to have significant reductions in carbon emissions (11.3%), but once you expand the building height (including enclosure area) to account for the loss in ceiling height, the reductions are nearly erased (2.0%). These reductions do not account for the significant increase in mass that would very likely increase emissions from the foundation. Furthermore, simply changing the mixture of the concrete slab in the steel building with ACM would achieve the same reductions (2.0%) as a full concrete building, but without the loss of area or height. There is almost no gain from masonry structures in this post-and-beam scenario (1.1% reduction), when compensating for the loss of floor area (no increase in height was needed due to using the same steel floor as the steel building). There would likely be additional emissions from foundations, given masonry is the heaviest building by a factor of 2.26 × the steel mass. On the other hand, wood has the largest reductions in emissions relative to the steel (30.5%) and even after the expansion of the building to account for the loss of volume (larger footprint and height, including the enclosure area), the reductions are significant (23.2%). These reductions in emissions are likely going to be higher when considering the significant reduction in mass that would reduce the amount of concrete in the foundations.

It is also important to qualify that in timber systems, deforestation, reforestation, and forest management practices can cause significant variation in carbon emissions [50] and there is ongoing debate about whether these can be over or underestimated in LCA [51]. Tally® provides some negative values (sequestration) for wood in the early product stage, but the transparency of EPDs in the database does not mean any specific standard is met. For example, timber in the US is generally assumed to be sustainably sourced (based on net increase in total forest cover) but products in the database are not individually certified, and thus the EPDs take a conservative approach, i.e. not taking additional credit from conservation of carbon in forest soil. [52] Currently it is not possible to choose certified wood in the software, e.g. FSC certified wood, even though industry experts suggest this has limitations at the start of the supply chain, in accounting for both environmental and social impacts. [53] For the purposes of this study, the assumption is that the numbers are conservative (no additional credit given for sustainable forestry practices based on what Tally® allows), but that for the numbers to be representative of truly lower carbon alternatives the wood would have to be sourced with sustainable and equitable foresting and harvesting practices. Therefore, when sustainable foresting practices become better integrated into the database of LCA tools, the results may increase the advantage of sustainably harvested timber, and potentially decrease the advantage of less sustainably harvested wood. Given the risks of market-driven illegal harvesting, certification will become more important, and therefore all lower-carbon scenarios will likely still depend on a limited supply that has to be adequately managed, preferably used on the highest impact projects. This study suggests that civic buildings designed for maximum durability and robustness are a good bet on sequestering carbon for the longest time frame.

Lastly, this nuanced analysis had to focus on a narrow set of variables. Some of those variables are proxies for other forms of durability. For example, design for resistance to PC is intended to work only during a catastrophic event that protects human safety but would likely render the building unusable after. However, the added robustness and capacity of the structure makes it a good mechanism to simultaneously design for future changes in use. While civic assembly buildings already have some of the highest load requirements in the code, which combined with the flexibility of post-and-beam construction offers significant future use flexibility, some future scenarios could demand higher loads. Designing for resistance to PC, e.g. removing one column, in this case study meant increasing the loads on a column by as much as 70% (by means of more tributary area). If the building use changes substantially and the risk level of the building is reduced to the point where resistance to PC is relaxed, that added capacity translates into added programmatic adaptability.

Variables that were not considered and could be potential next steps in this line of research, include the impact of a different enclosure systems as part of the structural change, e.g. perimeter structural-thermal mass instead of piers for the massive structures, and investigating tradeoffs with daylighting, views, and operational energy.

Conclusion

This research confirms that buildings designed for extraordinary criteria for longevity, such as large civic or cultural buildings, can have significantly higher EC than the median reported in benchmark studies. Studying the structural feasibility of timber alternatives for larger and longer-lasting building types may be the best approach to using a limited global supply of timber. Prioritizing timber projects with the intention and capacity to be extremely durable and culturally valuable can ensure not only dramatic EC reductions in early stages, but also long-term sequestration of carbon.

Redesigning the case study of the US federal district courthouse shows the potential for using mass timber instead of the more carbon intensive system like steel in highly durable and robust structures, and the LCA highlights the importance of a long service life in delaying the EoL emissions of timber. Clearly, the longer a protected timber structure stays in service (inside of a high-quality enclosure), the longer a significant portion of its emissions remain sequestered. However, this analysis attempted to provide an assessment of wood as a lower-carbon alternative that is more exhaustive and more grounded in the spatial and cultural conditions of architectural practice. As a result, this study introduces a methodology with the potential to mitigate factors that otherwise can lead to overestimating the advantages of timber. Specifically, wood buildings do need to be significantly larger in order to achieve the same spaciousness (net volume) as a steel building (long spans, tall spaces), which increases the volume of wood to sequester carbon into, but also increases the area of enclosure. Taller floor-to-floor heights maintain daylight parameters but are more likely than floor area to put pressure on the size of robust timber projects, especially those driven by real estate speculation. Therefore, the focus of future research and policy development should be on expanding the limits to allowable building height independent of expanding the number of stories, as well as exploring aspects of long-term structural performance. The models also show that the increase cross-sectional area (thickness in plan) required to provide fire resistance rating in timber columns still leaves a structure that is as spatially efficient as massive structures in concrete that are currently considered acceptable alternatives to steel in large institutional buildings. However, the timber building also has significantly less mass (weight), while sequestering significantly more carbon; even when accounting for the higher floor height and related additional surface of thermal envelope necessary for the wood building to achieve similar spatial and functional performance.

A whole building LCA shows the gap between steel and wood closing as the service life increases, in great part due to the larger role of the upkeep, maintenance and replacement of the high-carbon enclosure systems needed to keep a larger and durable structure protected. But the gap is still significantly in favor of wood buildings. Furthermore, the EoL impacts of wood are so much higher and its “beyond-system-boundary” energy recovery (Module D) potentially much lower and uncertain, that extending the life of the building is the only certain way to keep its emissions relatively low, and thus more important than in other alternative systems with higher recyclability potential. Therefore, while sustainably harvested wood for mass timber provides a short-term solution to reducing carbon in the construction industry, the durability of such projects is paramount in the long term. A good way to maximize carbon reductions is for the designer of a durable mass timber building to pay closer attention to the durability and emissions of the enclosure, which will likely be the most carbon-intensive component in a life cycle basis, accumulating emissions when being replaced many more times.

This study shows that comparing only similarly sized building versions with just a different amount of structure (different net volume) falls short of presenting a full picture of tradeoffs between environmental and spatial impacts of different carbon scenarios. A more nuanced qualitative modeling of building alternatives that offsets the direct spatial impacts of different structures, and thus their indirect economic and cultural impacts, can augment the power of LCA studies and potentially prove more useful to current and future practitioners to support early design decisions.