Beyond Operational Energy Efficiency: The Urgent Need for Embodied Carbon Regulation in the U.S

Abstract

As countries ramp up efforts to meet net-zero targets, building sector’s climate policies have prioritized operational energy─focusing on efficiency upgrades and renewable energy deployment. These strategies have significantly reduced operational emissions. However, a growing body of research warns that these gains may be offset by embodied carbon─the greenhouse gas (GHG) emissions associated with producing, transporting, and disposing of construction materials. (1,2) Globally, the building sector accounts for 37% of energy- and process-related CO₂ emissions, with embodied carbon comprising at least 10% activities. (3,4) In net-zero buildings, embodied carbon can represent over 50% of life-cycle emissions. (5) Yet, most national building policies─particularly in the U.S.─fail to address it. While the U.S. lacks a national embodied carbon policy, efforts like the Buy Clean California Act mark early progress. (6) To support national benchmarking, this study presents the most detailed assessment to date of U.S. embodied carbon emissions. This article also proposes a phased, building-type-specific roadmap for reducing embodied carbon in the U.S., aligned with international practices. Without such measures, the carbon embedded in new construction will continue to erode operational energy gains. As electricity grids decarbonize and buildings become more operationally efficient, embodied carbon is emerging as a dominant and often overlooked component of building-related emissions. Unlike operational energy, which benefits directly from clean energy deployment, embodied carbon stems largely from energy-intensive industrial processes─such as cement and steel production─that remain heavily reliant on fossil fuels. Consequently, embodied carbon regulation is not only a matter of building policy but a crucial element of broader energy system decarbonization. Without addressing upstream emissions embedded in construction materials, the energy savings achieved through electrification and efficiency will be increasingly offset by carbon-intensive construction practices. Establishing performance benchmarks is a foundational step toward reducing embodied carbon at scale. Only a handful of nations moving toward formal limits. In Europe, Denmark has led the way by embedding embodied carbon thresholds directly into its national building codes. As of 2023, all new buildings over 1,000 m² must comply with a lifecycle carbon emissions limit of 12 kg CO₂e/m²/year, with this cap set to tighten to 7.5 kg CO₂e/m²/year by 2029. Smaller buildings must also conduct a life cycle assessment (LCA), though they are not yet subject to performance thresholds 2029. (7,8) France’s RE2020 regulation similarly combines operational and embodied carbon targets, setting a cap of 415 kg CO₂e/m² for single-family homes and 490 kg CO₂e/m² for apartment buildings by 2031─a 30% reduction from 2022 levels. (9) Switzerland’s Zurich has taken a more holistic approach, aligning building lifecycle carbon targets with its 2000-W Society vision, which sets a goal of 8.5 kg CO₂e/m²/year for new residential construction by 2050. (10) In North America, Toronto is among the few jurisdictions to establish embodied carbon caps through its Toronto Green Standard (TGS). Version 4 of the TGS imposes a maximum of 350 kg CO₂e/m² on city-owned buildings beginning in 2025, with future code cycles expected to expand coverage to private development. Collectively, these examples point toward an emerging consensus: embodied carbon targets must be measurable, enforceable, and aligned with a building’s expected service life and function. While target values differ based on building typology, local material supply chains, and energy sources, most leading global benchmarks for residential buildings cluster around 400–800 kg CO₂e/m² over a 50-year lifespan. (11,12) Nonresidential buildings exhibit wider variability, with embodied carbon values ranging from 100 to 1,200 kg CO₂e/m² depending on structural systems and use function. Despite these differences, international benchmarks serve as critical reference points for the U.S., where no nationally recognized embodied carbon limits currently exist. The next section will assess U.S. building stock against these international targets, highlighting specific archetypes that significantly exceed global benchmarks. This comparison underscores the urgency for U.S. policymakers to adopt typology-specific embodied carbon caps as part of a broader decarbonization strategy for the built environment. In the United States, the most significant knowledge gap related to embodied carbon in buildings lies at the whole-building level, primarily due to the absence of comprehensive national-scale data and a lack of standardized assessment methodologies. (5) Two primary factors contribute to this research deficit. The first is the outdated perception that embodied carbon is negligible. (13) Early studies estimated the ratio of embodied to operational carbon at approximately 1:10, leading to a predominant focus on reducing operational emissions. (2) However, as operational energy performance has improved in recent years, the relative importance of embodied emissions has grown. Röck and colleagues, analyzing 238 case studies, reported increases in both relative and absolute embodied carbon, largely driven by emissions from the manufacturing and processing of construction materials. Embodied carbon now accounts for an estimated 20% to 90% of total life cycle emissions in buildings. (5) The second contributing factor is the lack of embodied carbon data for existing buildings. The heterogeneity and complexity of the national building stock pose substantial challenges for data collection and generalization. Unlike operational energy, which benefits from extensive, standardized data sets across various building types, embodied energy and carbon assessments remain fragmented and limited in scope. This scarcity of data hinders the development of benchmarks, comparative studies, and meaningful emissions reduction targets. (5) A critical component of embodied carbon assessment is the ability to evaluate and spatially map emissions across the existing building stock. Historically, most studies have employed top-down approaches, which estimate emissions using aggregate data sets such as material flow analyses or national economic input-output tables. (13) While these methods are effective at a macro scale, they lack the resolution needed to capture building-level variation. In contrast, bottom-up approaches analyze individual buildings using detailed information on material composition, construction type, and technological systems. (6) As Mastrucii argued, such disaggregated analysis enables more accurate and context-sensitive assessments of embodied carbon by accounting for the unique attributes of each structure. (14) This study adopts a bottom-up methodology by integrating the National Structure Inventory (NSI) with building envelope data extracted from Google Street View (GSV) using machine learning techniques. The NSI, initially developed by the U.S. Army Corps of Engineers for dam and levee risk screening, is now publicly available through the Federal Emergency Management Agency. It provides key attributes, including building type, foundation type, square footage, number of stories, and geolocation. However, the NSI does not include envelope material data─an omission that significantly limits embodied carbon assessments, given the emissions intensity of building envelopes. Recent advances in remote sensing and computer vision have demonstrated the potential of GSV imagery for filling this data gap. For example, Ghione and colleagues applied convolutional neural networks (CNNs) to detect façade materials and lateral-load-resisting systems. (15) Tutzauer and Haala utilized semantic segmentation techniques on GSV imagery to classify building use, leveraging GSV metadata to improve model accuracy. (16) Nguyen and colleagues extracted urban design features─such as pedestrian presence and visual enclosure─from GSV images to assess neighborhood walkability, linking physical attributes of the built environment to public health outcomes. (17) Building on this body of work, our recent study employs zero-shot learning to classify wall materials directly from GSV imagery, eliminating the need for model fine-tuning on manually labeled data sets. (18) Hence, it is practical to create a large national data set to be used to benchmark the embodied carbon of building stock. Building upon this bottom-up framework, the present study leverages the National Structure Inventory and advanced material classification methods using Google Street View imagery to generate a high-resolution data set of embodied carbon intensities across the U.S. building stock. By incorporating structural systems, envelope materials, and foundation types, the data set captures cradle-to-grave emissions (Modules A–C) and allows for detailed comparison across building typologies. This typology-based approach enables the identification of emission hotspots within specific building categories and supports the development of targeted benchmarks and policy interventions. The resulting data set covers 22 nonresidential and 11 residential archetypes, offering the most comprehensive empirical assessment to date of embodied carbon performance across the U.S. built environment. (13,19) This typology-based approach enables a detailed assessment of embodied carbon intensity and serves as a foundation for performance benchmarking. (20) Figure 1 illustrates substantial variation in embodied carbon intensity across U.S. building types, with values ranging from under 40 to nearly 180 kg CO₂e/m²/year. Residential typologies─particularly single-family homes (RES1) and manufactured housing (RES2)─consistently exceed international benchmarks by two- to 3-fold. Among nonresidential types, construction facilities and professional service offices also report high intensities, often above 70 kg CO₂e/m²/year. In contrast, educational buildings (e.g., schools and universities) and light industrial facilities exhibit lower embodied carbon levels, frequently under 50 kg CO₂e/m²/year. These intertypology disparities reinforce the need for differentiated, performance-based targets rather than uniform thresholds. Figure 1. Embodied carbon intensity across U.S building typologies. Two primary factors account for this disparity. The first is the carbon intensity of the U.S. energy mix used in material production. As of 2020, fossil fuels continued to dominate the U.S. energy sector, with natural gas comprising approximately 40% of total consumption, followed by petroleum and coal. (21) In contrast, the European Union has achieved greater decarbonization of its energy supply, with renewable sources─primarily wind and solar─contributing roughly 38% of the energy mix in the same year. (22) The second factor is material selection. In the U.S., single-family housing is predominantly constructed with wood, a relatively low-carbon material. However, other typologies─particularly industrial, commercial, and multifamily buildings─frequently rely on more carbon-intensive materials such as concrete, masonry, and steel. By contrast, European residential and commercial buildings, especially those built in the 20th century, have historically used concrete and masonry but are now increasingly transitioning to low-carbon alternatives. (23) These findings underscore the inadequacy of one-size-fits-all carbon caps. Uniform thresholds risk overburdening low-impact buildings or under-regulating high-impact ones. A differentiated approach─tailored to building function, scale, and material intensity─is essential for effective mitigation. For example, single-family homes and construction facilities warrant more stringent targets, while lower-intensity categories such as educational and civic buildings may serve as exemplars for low-carbon design. Several international policies have already adopted typology-specific benchmarks. France’s RE2020 regulation distinguishes between single-family and multifamily residential buildings, applying different embodied carbon thresholds to each. Zurich’s 2000-W Society sets a limit of 8.5 kg CO₂e/m²/year for residential buildings as part of a broader energy policy framework. Denmark’s building code outlines a phased reduction in allowable emissions─from 12 kg CO₂e/m²/year to 7.5 kg CO₂e/m²/year by 2029. (7) These precedents illustrate the feasibility and benefits of tailored carbon regulation. Establishing typology-specific embodied carbon benchmarks is a critical first step, but translating these benchmarks into actionable policy requires a structured and phased implementation strategy. For the United States, such a roadmap should begin with defining national baselines by building typology, drawing on publicly available data sets such as the Commercial Buildings Energy Consumption Survey (CBECS), Residential Energy Consumption Survey (RECS), and the National Structure Inventory (NSI). Voluntary programs like LEED, the AIA 2030 Commitment, and the Carbon Leadership Forum’s EC3 tool can serve as incubators for benchmark development and pilot initiatives. Figure 2 illustrates a phased policy roadmap using the office building typology as an example. Between 2024 and 2026, a voluntary sustainability standard could be introduced, mandating life cycle assessment (LCA) reporting as a first step toward accountability and transparency. During this pilot phase, embodied carbon targets would remain voluntary, with proposed thresholds such as 14 kg CO₂e/m²/year for buildings over 1,000 m² and an aspirational target of 9 kg CO₂e/m²/year for smaller buildings. This initial stage mirrors Denmark’s 2024 roadmap and allows time for industry alignment and data infrastructure development. Figure 2. Road map for reduction. From 2026 to 2030, the strategy would transition into a regulatory phase, with progressively stricter limits─dropping to 10.5 kg CO₂e/m²/year by 2028 and 9 kg CO₂e/m²/year by 2030. In the final phase, spanning 2032 to 2035, allowable embodied carbon levels would be further reduced to 7.5 and ultimately 5 kg CO₂e/m²/year. This phased and evidence-based framework provides predictability for stakeholders, enabling industry adaptation while ensuring continuous progress toward national and global decarbonization targets. The success of this roadmap depends on more than regulation. Complementary investments are required in LCA infrastructure, workforce training, and standardized data protocols to enable consistent and scalable assessment practices. Embedding embodied carbon thresholds into public procurement policies, zoning codes, and financial incentives─such as carbon tax credits, expedited permitting, and density bonuses─can amplify impact and encourage broader market transformation. Reducing embodied carbon is not solely a technical endeavor; it is a societal imperative. A phased, typology-specific strategy rooted in empirical data and supported by institutional capacity can enable the United States to lead in this domain. By aligning performance targets, regulatory levers, and incentive structures, the built environment can become a key driver of national climate goals─building the transition to a low-carbon future, quite literally, from the ground up. This article demonstrates that many U.S. building typologies─especially single-family homes, construction facilities, and manufactured housing─consistently exceed international embodied carbon benchmarks, often by wide margins. These elevated emissions are not intrinsic to building function but stem from systemic inefficiencies in material selection, energy sources, and entrenched design and construction practices. The typology-specific benchmarking framework introduced here offers a scalable and data-driven foundation for addressing these challenges. By integrating bottom-up building stock data with machine learning-based material identification, this study enables high-resolution life cycle assessments (LCAs) across a broad spectrum of U.S. building types. These assessments support the development of differentiated performance thresholds and identify priority targets for intervention. (24) To translate these insights into action, a phased policy roadmap is proposed, beginning with voluntary LCA reporting and progressing toward mandatory embodied carbon limits. This strategy balances regulatory ambition with industry readiness, drawing on successful precedents from Denmark, France, and Switzerland. It also accounts for variation across building types, aligning carbon targets with function, scale, and material intensity. Looking ahead, three strategic pathways will be essential to accelerate embodied carbon policy in the U.S.: Standardize embodied carbon benchmarks within federal and state building codes, grounded in consistent LCA methodologies; Embed embodied carbon thresholds into procurement policies, zoning regulations, and certification systems to create market demand and policy alignment; and Support a national low-carbon construction ecosystem through investment in data infrastructure, material innovation, and workforce development. Reducing embodied carbon is not solely a technical endeavor; it is a societal and energy systems imperative. As operational emissions decline due to electrification and renewable energy adoption, embodied carbon will constitute a growing share of life-cycle emissions in buildings. These emissions are directly linked to fossil fuel use in upstream industrial sectors, making their reduction essential for achieving energy transition goals. A phased, typology-specific strategy rooted in empirical data and supported by institutional capacity can enable the United States to lead in this domain. By aligning embodied carbon performance targets with regulatory levers, procurement policies, and incentives, the built environment can become a key driver─not a drag─on national energy and climate ambitions. This alignment will ensure that the benefits of decarbonized energy supply are fully realized across the building life cycle. This article references 24 other publications. This article has not yet been cited by other publications.