Carbon footprint of electricity: a systematic methodological and quantitative review

Carbon footprint of electricity: a systematic methodological and quantitative review | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Carbon footprint of electricity: a systematic methodological and quantitative review Camilla Citterio, Nicolò Golinucci, Lorenzo Rinaldi, Matteo Vincenzo Rocco This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9337261/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 9 You are reading this latest preprint version Abstract The carbon footprint of electricity is a critical metric for climate policy and corporate sustainability reporting, yet its quantification is hampered by a lack of methodological consensus and a fragmented evidence base. This fragmentation reflects a central debate in industrial ecology regarding the consistency and comparability of life-cycle–based emission factors across different modelling paradigms. This study addresses this gap by performing a systematic, quantitative comparison of electricity emission factors from a wide range of authoritative sources, including process-LCA databases (Ecoinvent), environmentally extended input-output tables (EXIOBASE, EMERGING, EORA, GLORIA, GTAP), primary agency data (IEA) and others. We harmonize technology taxonomies and GHG characterization factors to compare technology-level and national-level emission factors for the EU and US. Our results reveal, for technology-specific emission factors, that while there is a moderate consensus on direct, operational emissions (carbon intensity), we find wide and systematic divergences for full life-cycle emissions (carbon footprints). These discrepancies propagate to national-level estimates, where a country's calculated footprint can significantly vary, solely based on the data source chosen. This paper provides a useful reference to deconstruct and understand such divergences. We translate these findings into a set of practical selection criteria to support analysts and decision-makers making transparent and context-appropriate choices. We conclude that a more transparent and critical approach to sourcing emission data is fundamental to the credibility of carbon accounting and the broader energy transition. carbon footprint electricity emission factors systematic review corporate sustainability accounting industrial ecology Figures Figure 1 Figure 2 1. INTRODUCTION The generation of electricity constitutes a substantial portion of global greenhouse gases (GHGs) emissions. Around one quarter of worldwide GHGs emissions can be attributed to electricity and heat production, as stated by (Lamb et al., 2021 ) and confirmed by observing this metric from environmentally extended input-output (EEIO) tables: in 2015, they accounted for 26% according to EXIOBASE (Stadler et al., 2021) and for 31% according to FIGARO-E3 (EC JRC, 2024 ). From EEIO databases, it is also possible to state that the contribution of power production over the total carbon footprint of any marketed commodity ranges between 5% and 30% (see Online Resources, OR). Concurrently, the pace of electrification across economic sectors is accelerating rapidly (EMBER, 2025c ). In this context, electricity plays a pivotal role in climate policies and emissions disclosure frameworks, spanning both voluntary schemes (e.g., ESRS Voluntary Sustainability Reporting Standards (ESRS Voluntary Sustainability Reporting Standards for SMEs under the European Sustainability Reporting Standards Framework, 2025)) and binding regulations, such as EU’s Corporate Sustainability Reporting Directive (CSRD, 2022), Carbon Border Adjustment Mechanism (European Commission, 2025 ) or the California’s Senate Bill 253 in the US (Climate Corporate Data Accountability Act, 2023). This evolving context is driving an increasing number of companies to measure and report emissions associated with their operations. In particular, emerging regulations require firms to assess their carbon footprint and to report emissions related to purchased electricity (Scope 2 emissions, i.e. emission associated with companies’ energy supply), emphasizing the need for accurate and reliable emission factors for consumed electricity, typically following standardized approaches, like the GHG Protocol (World Resource Institute, 2011 ). Moreover, implicit assumptions about the quantification of national electricity mixes add uncertainty and significantly influence the calculation of Scope 3 emissions (e.g. those emissions occurring across the value chains). Indeed, as corporations reporting total GHG inventories must evaluate both Scope 2 and Scope 3, they frequently introduce systemic inconsistencies by employing disparate data sources and conflicting hypotheses for each scope, despite electricity's pervasive and dominant impact across the entire value chain. (Davis et al., 2025) show that 25.8% of 10'867 reporting companies in 2023 used spend-based approaches, typically relying on input-output sources. Among those specifying the method, 75% adopted single-region EEIO models, implicitly assuming domestic technologies. This leads to applying domestic electricity and emission intensities to imported goods, thereby underestimating the often higher carbon intensity of global supply chains. Despite the critical importance of electricity in climate mitigation strategies, there is currently no universally agreed methodology, nor a universally accepted reference source, for estimating the carbon footprint per kWh of electricity. Instead, several authoritative sources – including the Intergovernmental Panel on Climate Change (IPCC) assessment reports (ARs), life cycle assessment (LCA) databases (Ecoinvent, 2025 ), international agencies such as the International Energy Agency (IEA, 2024b ), widely cited commercial solutions by scientifically reputable sources, such as Electricity Maps (Electricity Maps, 2025 ) and input-output tables such as EXIOBASE (Stadler et al., 2018 ) or CEDA (Watershed, 2025 ), widely utilized both in the academic and corporate sustainability contexts, often report misaligned values for the environmental impact of electricity production,. Previous works highlighted how the absence of harmonized protocols can lead to divergent estimates: differences in standards and reporting conventions generate substantial discrepancies in CO₂ estimates, resulting in notable variation among published grid emission factors, despite considering the same technology and/or the same geographical boundary (Hawkes, 2010 ; Pehl et al., 2017 ; Turconi et al., 2013 ). Moreover, in many cases, even high-resolution databases yield misleading results as they often embed outdated electricity supply mix data. Electricity data in the most commonly utilized life cycle databases typically suffer from a significant time lag - averaging between 4 and 8 years - in accurately representing the current state of national power mixes, although the frequency of publication and updates has increased in recent years (Olindo et al., 2021). Such variability arises from several scientific and methodological factors, including the definition of system boundaries (e.g., direct combustion emissions only versus broader upstream and downstream coverage through life-cycle approaches (Asdrubali, Baldinelli, D’Alessandro, & Scrucca, 2015 ; Barros et al., 2020 ; “Ecoinvent,” 2025; National Renewable Energy Laboratory, 2021)), assumptions about electricity mix composition and plant efficiencies, and choices regarding marginality (i.e., consequential versus attributional approaches (Beltrami et al., 2025 ; Ryan et al., 2016 ). Additional drivers are the selection and resolution of input data, encompassing not only geographic and technological detail but also temporal resolution, as power grids become increasingly heterogeneous with high penetration of renewable generation (Blizniukova et al. 2023 ; Khan, 2019 ). Examples range from physically grounded but narrow-boundary approaches, such as process-based LCA, and monetary input–output approaches – or hybrid datasets like EXIOBASE Hybrid (Merciai & Schmidt, 2021)–that are more internally consistent from a modeling standpoint but less granular. 1.1 Objective and contribution Our study maps the main families of sources, used both at academic and corporate level, that estimate the carbon intensity and carbon footprint of electricity, undertaking a systematic, methodologically transparent comparison focusing on EU27 countries and the US. To ensure commensurability, all series are harmonized and aligned to a shared technology taxonomy. Results are presented at technology and national‑mix level under both direct and indirect emissions perspectives. While previous research has explored discrepancies within subsets of electricity carbon footprint studies, such as process-LCA meta-reviews or EEIO database comparisons, no prior work has provided an integrated, cross-paradigm assessment that aligns EEIO, P-LCA and agency-based datasets under a harmonized taxonomy and GWP framework. The performed cross-sources comparison is also intended to provide a practical compass for analysts and decision-makers in navigating a fragmented evidence base, clarifying what each source measures, when it is most appropriate, and how boundary choices and methodological assumptions shape results. Comparative findings are translated into selection criteria and reporting conventions that support deliberate, defensible choices (i.e. whether for policy appraisal, procurement, or corporate disclosure) so that users can select sources consciously instead of arbitrarily. 2. LITERATURE REVIEW We analyzed 39 studies published after 2013 to identify the main methodological challenges in quantifying emissions from electricity generation, mainly distinguished on three key levels of complexity: (i) the quantification of emissions factors for specific generation technologies; (ii) the aggregation of these into grid-level intensities; and (iii) the comparison of results from fundamentally different modeling paradigms, such as process-based LCA (P-LCA) and EEIO models. Before presenting the literature review, we introduce the terminology adopted throughout the article. 2.1 Terminology A primary obstacle to comparing emission data is inconsistent terminology. To ensure clarity, we distinguish between two key emissions factors based on system boundaries. We define “carbon intensity” (CI) as the direct, operational emissions from generating 1 kWh of electricity (equivalent to Scope 1 for a power producer). In contrast, we use “carbon footprint” (CF) as a generic term for any emission factor that goes beyond CI. This broader definition is needed because different sources draw system boundaries in different ways, sometimes including upstream processes (e.g., fuel extraction, plant construction) and downstream impacts (e.g. transmission and distribution losses). The generic term “emission factor” (EF) will be used to refer to both CI and CF when the distinction is not critical. Furthermore, a crucial methodological divide, central to the following review, is the distinction between attributional and consequential (or marginal) modeling. Attributional EFs allocate the total emissions of a grid across all consumed electricity, answering the question: "What is the average footprint of the electricity mix?". Consequential EFs, instead, estimate the change in system-wide emissions resulting from a change in demand, answering: "Which power plant will operate to meet an additional kWh of consumption?". For this article we focus on national-scale attributional EFs, since they represent the common ground across multiple applications, including corporate carbon footprint. Consequential approaches, on the other hand, are more suitable when the objective is to provide behavioral or price signals for short-term applications, such as daily or weekly operations (e.g., smart charging), requiring dedicated databases and tools (Merciai et al, 2025 ). 2.2 Technology specific emission factors The foundational challenge in electricity carbon accounting lies in the wide availability and heterogeneity of published EFs for individual generation technologies. Comprehensive reviews by (Khan, 2019 ) and (Barros et al., 2020 ) systematically documented this issue, identifying the primary drivers of divergence: inconsistent system boundaries, methodological choices, and different geographical or technological assumptions. Earlier works, such as (Turconi et al., 2013 ) and (Asdrubali et al., 2015 ), conducted thorough comparisons of LCA analyses’ results for different technologies demonstrating how even subtle analytical choices can yield significantly different results (e.g. variations exceeding a factor of 10 for renewable technologies in Turconi et al., while harmonized data in Asdrubali et al. show maximum differences of 500%). A vast body of literature exists performing LCAs of specific electricity technologies, such as the parametric analysis of nuclear power by (Gibon & Hahn Menacho, 2023 ) and the work of (Raadal et al., 2016 ) on offshore wind turbines. Other studies, like (Hertwich et al., 2015 ), perform environmental impact analysis for various technologies through the same methodology, in this case scenario-based hybrid LCA. The lack of comparability created a clear need for standardized benchmarks. In response, major institutional efforts emerged to harmonize these values. The National Renewable Energy Laboratory's (NREL) work on establishing statistical ranges (Asdrubali et al., 2015 ) and the IPCC’s provision of harmonized CF ranges (Schlömer et al., 2014 ) have become cornerstone references in the field (Clauß et al., 2019 ; Unnewehr et al., 2022 ). These IPCC values, alongside its widely used combustion factors and methodological guidelines (Blizniukova et al., 2023 ; Scarlat et al., 2022 ), provide a crucial, authoritative starting point for analysis. However, they represent technology-level potentials, and the next layer of complexity arises when aggregating them to model real-world, interconnected electricity grids. 2.3 Grid-level emission factors Moving from individual plants to national grids introduces a further layer of methodological complexity. While some studies employ top-down approaches using aggregated national statistics on emissions and generation (Unnewehr et al., 2022 ), the prevalent methods are "bottom-up," building upon technology-specific factors. Even within this dominant category, significant variability exists:(Bertolini et al., 2025 ) compared six different EF-based methods for Italian electricity zones, finding a 12% spread between minimum and maximum average EFs in the North zone alone, and underscoring that no single approach is universally applicable. (Ryan et al., 2016 ), systematically mapped 32 methodological approaches, classifying them into "Empirical Data and Relationship Models" (including simple EFs, commercial exchanges and statistical relationship models) and "Power System Optimization Models". Their US case study showed that the choice of method alone could alter the CI for vehicle charging by up to 68%, a variation often larger than geographic location. Their subsequent recommendations (Ryan et al., 2018 ) highlight the pivotal distinction between attributional and marginal EFs: studies on cross-border electricity trade (Beltrami et al., 2025 ) show that marginal generation technologies in highly interconnected European markets often diverge substantially from the average mix, with spatially resolved marginal factors (Sarhan et al., 2023 ) enabling more granular locational analysis. A more computationally intensive class of methods addresses cross-border carbon accounting through flow-tracing algorithms. (Tranberg et al., 2019 ) and (Schäfer et al., 2020 ) applied these techniques across the EU, revealing significant discrepancies between a country's production-based and consumption-based CFs. For instance, nations with large low-carbon generation shares, such as Slovakia and Austria, see their consumption footprints increase due to electricity imports from higher-emission neighbors like Poland and Czech Republic (Tranberg et al., 2019 ). By incorporating high temporal resolution and explicit trade modeling, these studies demonstrate substantial sub-monthly variations in grid CF. Such time-resolved, flow-tracing approaches enable a more accurate allocation of emissions from producers to final consumers, providing valuable data for actors who can control their electrical loads (Clauß et al., 2019 ; Messagie et al., 2014 ). 2.4 Divergences in foundational databases A final and fundamental source of variation stems from the databases used to model background economies and supply chains. Practitioners must often choose between two distinct paradigms: bottom-up P-LCA databases and top-down, economy-wide EEIO tables. The fundamental differences between these approaches have long been recognized in the debates on completeness and truncation errors: (Suh et al., 2004 ) demonstrated that system boundary selection in LCA significantly affects results, with IO-based approaches capturing a broader economic scope that process-based inventories systematically truncate. (Majeau-Bettez et al., 2011 ) further quantified these truncation and aggregation issues, showing that P-LCA databases tend to underestimate impacts due to incomplete upstream coverage. A comparative work by (Steubing et al., 2022 ) directly addressed this issue by matching product CFs between Ecoinvent and EXIOBASE. Despite both sources being used for similar purposes, the results showed significant discrepancies, with over half the matched products exhibiting CF differences greater than a factor of two. While concordance was better for fossil energy and manufacturing, the discrepancies for electricity from renewable sources were particularly pronounced, driven by differing regional and technological disaggregation. The comparison highlighted fundamental trade-offs: the comprehensive economic coverage of EEIO tables appears to be counterbalanced by the more detailed inclusion of capital goods and multi-year timeframes in P-LCA. This tension is further analysed by (Henriques & Sousa, 2023 ), who show for electricity generation that EEIO-based assessments tend to provide systematically higher impacts than P-LCA, advocating for hybrid options to leverage the strengths of both paradigms. This claim has been strongly supported by (Hagenaars et al., 2025 ), while (Perkins & Suh, 2019 ) provide empirical evidence that hybridization can improve accuracy by combining the completeness of IO boundaries with the detail of process data. However, results can vary even within the same paradigm. A comparison of four major multi-regional input-output databases by (Moran & Wood, 2014 ) found that while the underlying economic structures were broadly similar, divergences in CFs were primarily driven by the satellite environmental accounts used for GHG emissions. Crucially, they demonstrated that harmonizing these satellite accounts alone could reduce the divergence between major economies' footprints to under 10%. This finding underscores that both high-level modeling choices (P-LCA vs. EEIO) and low-level data compilation decisions contribute significantly to the fragmented evidence base that practitioners currently face. 2.5 Research gap identification Ultimately, the body of analysed studies leaves the practitioner facing a series of critical, yet unresolved, methodological forks in the road. In the context of corporate carbon footprint assessment, established protocols such as the GHG Protocol prescribe an annual perspective. Within this temporal scope, the attributional approach represents the standard methodological choice. Nevertheless, an analyst must choose not only a set of technology-specific EFs, but also an aggregation method, and, at an even more fundamental level, a foundational database paradigm (P-LCA or EEIO) adopting the most accurate EF for a defined region and year. Our literature review demonstrates that each of these choices is highly influential yet offers no clear guidance on how to navigate them in concert. While individual studies have illuminated specific points of friction, the cumulative impact of these discrepancies remains unquantified in a single, harmonized framework. Addressing this requires moving from isolated comparisons to a systematic, multi-source confrontation. 3. METHODS Our analysis performs a comprehensive comparison of EFs drawn from a curated selection of widely used data sources, grouped into four main categories: EEIO tables, P-LCA databases, primary data from international agencies, and secondary data from literature harmonization projects. A detailed summary of these sources, including their reference years, scope, and key methodological characteristics, is provided in Table 1 . EEIO databases The EEIO databases considered in this study are: EXIOBASE, in two distinct versions, the hybrid Supply-Use Table (SUT) format (v3.3.18, data year 2011) (Merciai & Schmidt, 2021) and the more recent (v3.10.1) monetary Input-Output Table (IOT) format (Stadler et al., 2025 ); GLORIA (v0.60, SUT format) (ielab, 2026); EMERGING (v199.82) (Huo et al., 2026); EORA26 (v199.82) (Lenzen et al., 2013) and the WIOD format of GTAP (v11) (Aguiar et al., 2022). The GTAP project also provides electricity price data for each region, as detailed in Section 3.3 . P-LCA databases Ecoinvent, as the most utilized process-based database in industrial ecology, is a critical component of our analysis. Due to licensing constraints on publishing direct results, we leverage the data made available by the comparative study of (Steubing et al., 2022 ), which reports carbon footprints from Ecoinvent v3.4 (cut-off system model) with regional information removed. Agencies’ primary data We include the authoritative International Energy Agency (IEA) emission factors for combustion technologies. The IEA provides country-specific CIs for 1990–2023. As this data is licensed, our figures will not show these series. Two versions of IEA CIs are provided, differentiated by assumptions regarding the allocation of emissions from combined heat-power (CHP) plants: “Electricity only” assumes all CHP plant emissions are attributed to electricity generation, while “CHP” assumes a portion of the emissions are allocated to heat, resulting in a lower CI (IEA, 2024a ). One peculiarity of IEA factors is that they do not exclude CO₂ emissions of biogenic origin, as suggested by the IPCC guidelines (Eggleston et al., 2006 ). Secondary and harmonized data To include influential reference values, we incorporate several sources: the IPCC, we use the benchmark technology-level carbon footprints reported by Working Group III in AR5 from 2014 (IPCC, 2014 ); the harmonized LCA emission estimates from (NREL, 2021 ) release. Finally, we consider Electricity Maps, a widely used platform providing real-time electricity flow data (Electricity Maps, 2025 ). The values considered are their declared, static emission factors for each region (Electricity Maps, 2024 ), which are themselves derived from a combination of other key sources, including IPCC AR5 and (UNECE, 2021 ) reports. We also include sources that provide ready-to-use, country-level EFs. From the European Commission's Joint Research Centre (JRC), we include their dataset providing both direct "emissions intensity" (CI) and "lifecycle emissions intensity" (CF) for EU27 countries (Bastos et al., 2024 ). Since JRC data covers EU countries only, the CI for US is obtained from (eGrid, 2023 ). From EMBER, we use their "Carbon emissions intensity" data, which represent full life-cycle CFs (EMBER, 2025b , 2025a ). Table 1 Data sources considered in this study with versioning or publication year, detailing the acronym adopted henceforth, the data reference year, availability of region-specific data for EU and US, the type of emission factors analyzed and on which GHG substance, methodological treatment of co-production processes, data type and source references. Extended version in OR. Data source and publication year (or version) Acronym Data year Detail on US and EU Emission factors analyzed GHGs included Handling co-products Data type and unit Reference EXIOBASE Hybrid SUT v3.3.18 EXIO Hybrid 2011 ˅ CI, CF CO 2 , N 2 O, CH 4 , SF 6 , HFCs Industry-based EEIOT, physical (Merciai & Schmidt, 2021) Ecoinvent v3.4 via Steubing et al. Ecoinvent CF CO 2 , N 2 O, CH 4 , SF 6 , HFCs Cut-off P-LCA, physical (Ecoinvent, 2025 ; Steubing et al., 2022 ) IEA emission factors per plant 2024 IEA (CHP) IEA (ele) 1990–2023 ˅ CI CO 2 , N 2 O, CH 4 Electricity and heat CHP/ Electricity output only Primary data, physical (IEA, 2024b ) Electricity Maps 2025 Electricity Maps ˅ CI, CF Generic CO 2−eq Secondary data, physical (Electricity Maps, 2024 ) NREL harmonization 2021 NREL x CF Generic CO 2−eq Secondary data, physical (NREL, 2021 ) IPCC AR5 WG3 (Annex III) IPCC x CI, CF Generic CO 2−eq Secondary data, physical (Schlömer et al., 2014 ) EXIOBASE IOT ixi/pxp v3.10.1 EXIO ixi EXIO pxp 1995–2022 ˅ CI, CF CO 2 , N 2 O, CH 4 , SF 6 , HFCs Industry-based technology assumption EEIOT, monetary (Stadler et al., 2025 ) EMERGING v2.1 EMERGING 2015–2023 ˅ CI, CF Generic CO 2−eq Industry-based technology as. EEIOT, monetary (Huo et al., 2026) EORA26 v199.82 EORA 1990–2023 (open access − 2017) ˅ CI, CF CO 2 , N 2 O, CH 4 Industry-based technology as. EEIOT, monetary (EORA, 2026; Lenzen et al., 2013) GLORIA v0.60 GLORIA 1990–2025 ˅ CI, CF CO 2 , N 2 O, CH 4 ,CO, SF 6 , HFCs Industry-based technology as. EEIOT, monetary (ielab, 2026) GTAP 11 GTAP 2004, 2007, 2011, 2014, 2017 ˅ CI, CF CO 2 , N 2 O, CH 4 ,CO, SF 6 , HFCs Industry-based technology as. EEIOT, monetary (Aguiar et al., 2022) EMBER 2025 EMBER 2000–2024 ˅ CF Generic CO 2−eq Secondary data, physical (EMBER, 2025b ) JRC CoM JRC 1990–2021 EU only CI, CF CO 2 , N 2 O, CH 4 Secondary data, physical (JRC, 2024 ) eGRID summary tables eGRID 2018–2023 US only CI CO 2 , N 2 O, CH 4 Secondary data, physical (eGrid, 2023 ) 3.1 Data harmonization and processing To enable a robust and fair comparison across the selected sources, a multi-step harmonization protocol was applied to align technology definitions, GHG characterization factors and analytical scope. First, a common technological taxonomy was established. The diverse electricity generation categories reported in each source were mapped and consolidated to a common level of aggregation; details are documented in the OR. Second, all emission factors were standardized to a consistent climate metric. We express all results in terms of Global Warming Potential (GWP) over a 100-year horizon, employing the characterization factors from IPCC AR5, following what is done by similar works (Steubing et al., 2022 ) and since EFs from the same version of the report are utilized in our analysis (IPCC, 2014 ). The processing of EEIO databases was conducted using MARIO (Tahavori et al., 2023 ), an open-source Python package for input-output analysis. Within this framework, a critical distinction was made between activity-based and commodity-based footprints. CI is defined only at the activity level, as environmental accounts are linked to industry outputs. The CF of the electricity commodity, however, is calculated following the industry-based technology assumption (United Nations, 2018 ), by weighting the footprint of each electricity production activity, including those that are not power plant technologies (e.g. industries producing from PV rooftops), on their market share. This modeling approach is particularly relevant for capturing the effects of cogeneration, as detailed in the OR. 3.2 Derivation of grid electricity emission factors To assess the impact of different data sources at a national level, we calculated grid-level emission factors for each EU country and the US as the weighted average of the technology-specific EFs from a given data source, with the weights determined by that country's annual electricity generation mix. The electricity supply mix data was obtained from the ENTSO-E Transparency Platform for EU countries (ENTSO-E, 2025 ) and from the eGRID database for the US (eGrid, 2023 ). Consistently with the comparative objective of the analysis, we limited our calculations to EFs from electricity generated within national borders as, according to Table 6.2 of the GHG Protocol Scope 2 guidelines, national production-based factors are an accepted and adopted option for corporate carbon footprint (World Resource Institute, 2023 ). Moreover, production-based factors represent the necessary foundation for deriving accurate consumption-based factors. It should be noted that, since EXIOBASE Hybrid only provides data for 2011, we did not compute the CF as the table’s electricity commodity, but rather as the weighted average CF of electricity generation activities, as described in previous paragraph. This was done in order to avoid including the impact of any other activities that produce electricity as a co-product, to ensure that electricity supply mixes were the same as for IPCC and NREL results. 3.3 Emission factors from economic databases and conversion to physical units In practice, compiling a corporate carbon footprint compliant with frameworks like the GHG Protocol typically relies on economic databases for Scope 3.1 emission factors, given the scarcity of primary data (David et al., 2023). While Scope 2 emissions are generally assessed using physical sources, electricity often drives a significant portion of the overall impact, necessitating consistent integration across all footprint elements. Consequently, this comparison includes monetary databases to evaluate their viability as a single, unified data source for comprehensive corporate carbon accounting. As mentioned, we analyzed EFs from EXIOBASE, available in monetary IOT format in its latest 3.10.1 version, from which we derived the CF of electricity, by aggregating all electricity production sectors into one, therefore weighting the EFs according to the generation shares included in the IOT. We followed the same procedure both for the product-by-product (pxp) and industry-by-industry (ixi) tables. Other EEIO tables directly provide one unique sector for electricity production: we extracted CI and CF from EMERGING, EORA, GLORIA, GTAP (see Table 1 for more details). The values obtained are emissions per monetary unit, in basic prices. The comparison with EFs in physical units is possible through the conversion with the regional price of electricity provided by GTAP; since those prices are available, for recent years, only for 2017 and 2023, those are the years selected for the comparative analysis in Section 4.3 . 3.4 Benchmark definition Regarding the CIs in physical units, IEA data were used as the reference benchmark, given their primary-data status and its widespread institutional authority. Given that the CF of a product depends on a wide range of factors that we have explicitly addressed, it is not meaningful to define a single benchmark. Instead, we compare the various sources by examining their relative deviations. 4. RESULTS 4.1 Technology-specific emission factors comparison Our cross-source comparison highlights significant variability in the estimation of electricity EFs. Even for direct Carbon Intensities (CIs), where methodological alignment might be expected, the sources exhibit substantial divergences rather than a clear consensus. Moreover, the picture for CFs shows wide-ranging disagreement and methodological heterogeneity, especially for renewable technologies. The analysis of CI for fossil fuel generation (Fig. 1 a, where IEA values are not reported due to license limitations) demonstrates a higher consensus than the corresponding CF. For coal-fired power, the spread between medians, meaning the relative difference between minimum and maximum median, is 36%; for oil the spread is 52% and for natural gas it is 116%. For CFs instead the defined spread ranges from 101% to 631%. This suggests that, when the analytical scope is restricted to direct combustion emissions, estimates from different providers tend to be more comparable, although non-trivial discrepancies may persist.. However, a critical exception immediately highlights a primary source of confusion: the treatment of biogenic CO 2 in biomass-based electricity. The IEA reports CI values for biomass that can exceed 2000 gCO₂eq/kWh, in stark contrast to other sources that report near-zero direct, non-biogenic emissions. This enormous discrepancy stems from a single accounting choice: IEA includes direct CO₂ emissions in its factors even if they are of biogenic origin (IEA, 2024a ), a practice contrary to IPCC guidelines (Eggleston et al., 2006 ), which consider such emissions to be carbon-neutral within the energy system's carbon cycle. This example is a stark illustration of how a differing interpretation of accounting rules, even for direct emissions, can lead to a result that is orders of magnitude different and potentially misleading for policy or reporting purposes. In contrast to the CI analysis, the comparison of CFs reveals larger disagreements across nearly all technologies (Fig. 1 b and Fig. 1 c). For renewables, where the footprint is dominated not by operational but by manufacturing and decommissioning emissions, the choice of data source becomes critical. For example, the reported CF for hydroelectric power in Ecoinvent, and the maximum of the range reported by IPCC, are substantially higher than in any other source, a result (Steubing et al., 2022 ) attribute to the inclusion of diverse plant types and fugitive reservoir emissions. Similarly, for nuclear power, the CF from Electricity Maps is less than half the median reported by the IPCC and falls below the minimum range of the NREL harmonization project. These discrepancies underscore that once the analytical boundary expands beyond the smokestack into the complexities of the full supply chain, any semblance of consensus vanishes, leaving the user in a landscape of highly variable data. 4.2 Systematic differences between EEIO and P-LCA Databases Across the wide variability of CF data, a clear and actionable pattern emerges: the choice between a P-LCA and a EEIO database is not neutral. Our results show that the EEIO database EXIOBASE Hybrid consistently yields the highest or among the highest CF values for nearly every technology analyzed. This effect is highly visible for conventional generation technologies (Fig. 1 b). For natural gas, the medians for IPCC and NREL align on 490 gCO₂eq/kWh, Electricity Maps’ range is tightly concentrated around 529 (coefficient of variation, CV, of only 10%), Ecoinvent has a broader span with median 598 gCO₂eq/kWh. The median CF from EXIOBASE Hybrid, instead, reaches 780 gCO₂eq/kWh. This relatively high value can be attributed to the use of the hybrid EEIO framework, which captures the full range of transactions along the value chain, including services typically excluded from process-based LCA approaches. In addition, the industry-based technology assumption combined with the treatment of blast furnace gas as an input to the power sector indirectly activates the steel industry, thereby inflating the footprint. Despite these methodological effects, process-based studies in the recent literature analyzing the liquefied natural gas supply chain report estimates of a similar magnitude (Howarth, 2024 ). The pattern is even more pronounced for coal: the entire distribution of CF values from EXIOBASE Hybrid lies above the third quartile of NREL, IPCC and Electricity Maps, effectively creating a separate, higher-impact class of results. Even Ecoinvent provides 26% of values below EXIOBASE Hybrid’s lower whisker and 60% of values below its first quartile. This effect persists, and in some cases is amplified, for renewable technologies (Fig. 1 c). For wind power, all medians fall within the narrow range of 11–23 gCO₂eq/kWh, EXIOBASE Hybrid's distribution is significantly higher and wider, with outliers surpassing 150 gCO₂eq/kWh. Similarly, for geothermal energy, EXIOBASE Hybrid's median CF is 53% higher than the value reported by the IPCC and Electricity Maps. This systematic difference is not an error, but a direct consequence of the fundamentally different analytical boundaries of the two paradigms, as explored in the literature review. Across the sources compared here, EEIO-based CF estimates are generally higher than process-LCA-based estimates within our matched sample of electricity technologies and national mixes. We do not interpret this pattern as evidence that EEIO results are necessarily closer to ground truth. Rather, the gap reflects differences in modelling architecture, including broader system boundaries and reduced truncation of upstream services and capital in EEIO, versus greater technological specificity but potentially narrower boundaries in P-LCA. The magnitude and even direction of the gap may also depend on aggregation, multifunctionality assumptions, data vintage, and the geographical representation of supply chains. 4.3 National grid emission factors The methodological divergences identified at the technology level directly propagate to the calculation of national grid emission factors, leading to a wide range of potential carbon intensity and footprint values for any given country. Analyzing the computed CIs, the difference between minimum and maximum estimates, relative to the IEA reference value, ranges between 72% and 730% for 2017 (Fig. 2 a), and between 38% and 612% for 2023 (Fig. 2 c). This confirms that even for direct emissions, the methodological landscape is far from settled. When grouping databases by original unit (monetary vs. physical) a consistent pattern emerges: monetary-based sources tend to yield lower CIs estimates, a trend that is particularly pronounced for the year 2017, in which more than 90% of the minimum values across technologies are attributable to a monetary source. It should be noted, however, that monetary EFs are inherently mediated by electricity prices: a higher price assumption would allocate the same quantity of GHG emissions to a smaller electricity output, thereby resulting in a higher CI estimate. In examining the CF, it is observed that the spread (difference between minimum and maximum) of monetary-derived data is systematically wider than that of physical sources: on average, the range is approximately 150% larger for monetary than for physical databases in 2017. Furthermore, for the 2017 CF values, monetary sources fall below the country-level median in 60–100% of cases (with exception of GTAP 11, which does so in only 12% of countries), whereas physical sources exceed the median in 64–92% of countries (with the exception of IPCC-based values, only 32%). The hybrid EXIOBASE estimates, computed here in combination with respectively the 2017 and 2023 ENTSO-E electricity mix, stand above the country median in 96% of cases; moreover, for 20 out of 28 countries this source reports the highest CF value in 2017, and for 13 out of 28 in 2023. 5. DISCUSSION This study set out to provide a practical compass for navigating the fragmented evidence on electricity's carbon footprint. Our results move beyond simply confirming the existence of discrepancies; by systematically examining them, we can translate our findings into a set of guiding principles. The core conclusion is that an optimal data source does not exist; the "best" choice is contingent on the user's specific application and analytical goals. The first critical decision point for any analyst concerns the analytical scope: reporting direct emissions (CI) versus full life-cycle impacts (CF). Our analysis reveals a moderate consensus for the CI of fossil fuels, suggesting that for straightforward applications like basic Scope 2 reporting, relying on primary data from sources like the IEA provides a widely accepted foundation. However, even within this consensus, our findings on biogenic carbon accounting warn that differing guideline interpretations can yield very different results, demanding user vigilance. When the scope extends to CFs, the choice of modeling paradigm becomes the central strategic consideration. We consolidated a systematic yet typical "paradigm effect," where EEIO databases tend to yield higher CFs than P-LCA sources. This pattern should not be interpreted as evidence of closer agreement with ground truth, but rather because of broader system boundaries and reduced truncation in EEIO models. This is not an error, but a feature that can be leveraged. If an analyst's objective is a conservative, upper-bound estimate that captures the broadest economic context (e.g., in a corporate risk assessment), then the use of an EEIO database is a defensible choice. Moreover, the EEIO framework is, in our view, the only approach that enables a comprehensive assessment of a company's carbon footprint while allowing for a systematic and coherent allocation of Scope 3 emissions. If, conversely, the goal is comparability with standard product LCAs or Environmental Product Declarations (EPDs), which predominantly use process-based data, selecting an LCA-based source is necessary to ensure methodological consistency. A final, crucial distinction lies between physical and monetary units. Our analysis of monetary emission factors revealed a higher spread of values for monetary databases, and we highlight that the obtained EFs are inherently mediated by the price of electricity. This renders their uncritical use for formal reporting highly problematic. While convenient for high-level screening where accountability data is readily available, their application must be approached with extreme caution. Whenever accuracy is a priority, physical factors based on actual consumption (in MWh) should be the default and preferred method. However, for corporate carbon accounting the key question is not which paradigm is “truer”, but which data infrastructure enables coherent, repeatable annual inventories. In practice, mixing Scope 3.1 estimates based on monetary tables that are several years out of date with electricity factors drawn from heterogeneous physical sources is hard to justify especially when electricity can explain a large share of the footprint. This motivates the advantages of a centralized, integrated, regularly updated framework that uses physical data where feasible and monetary proxies where necessary, while preserving internal consistency across scopes and categories, and is complemented by primary data whenever available. To make these discrepancies concrete, consider a 10 MWh electricity bill corresponding to 1’600 EUR (basic price, 2023) in Italy. Utilizing the presented monetary and physical factors, the maximum total GHG emissions are 3 times the minimum value, with a difference of 4 tCO2eq in the considered year, a spread that underscores how methodological choice dominates results and why transparent reporting is non-negotiable. Ultimately, all these decisions are bound by one unbreakable rule: the imperative of radical transparency. Given that a country's calculated footprint can vary dramatically based on the data source, any reported EF must be accompanied by a clear citation of the underlying source and version. This is the minimum requirement for enabling reproducibility and allowing the audience to understand the context and limitations of the presented figures. These principles are summarized in Table 2 as a practical guide for practitioners. Table 2 Decision framework for selecting electricity emission factor sources. Use cases Recommended data type Rationale and key considerations Territorial electricity benchmarking / producer-side analysis Production-based CI or CF in physical units from primary or official sources Use when the question concerns electricity generated within a territory. Suitable for producer-side benchmarking and comparability with generation statistics. Not ideal for end-use footprints where imports/exports materially alter the delivered mix. Scope 2 location-based accounting Consumption-/delivery-oriented CI in physical units; production-based factors only as a proxy when better consumption-based data are unavailable Prefer in highly interconnected or import-dependent grids. Always disclose whether the factor is production- or consumption-based, together with its data year and treatment of imports/exports. Scope 2 market-based accounting Supplier-specific or residual-mix CI in physical units Use only for contractual claims that meet Scope 2 quality criteria. This should be explicitly flagged as outside the main quantitative comparison in the present paper. Scope 3 category 3 (fuel- and energy-related activities, incl. upstream electricity and T&D losses) Physical CF of electricity (kgCO2e/kWh), using average-data, supplier-specific, or hybrid factors Align with the chosen Scope 2 basis. Explicitly state whether transmission and distribution losses are included. Monetary factors are not appropriate once kWh/MWh are known. Corporate Sustainability Disclosure, Hot spotting EEIO databases (e.g., EXIOBASE) Appropriate when expenditure data dominate and many categories or suppliers must be covered. Prefer multiregional models when supply chains are international.(Davis et al., 2025) Use to identify hotspots, then refine material categories with hybrid or supplier-specific/physical data. Comparability with product LCAs or EPDs P-LCA databases (e.g. Ecoinvent) Ensures methodological consistency with the dominant approach in product-level studies. Disclose cut-off rules, multifunctionality assumptions, capital goods treatment, geography, and data vintage. Rapidly evolving grids / time-sensitive applications Freshest year-matched physical factors; where relevant, subannual/hourly or near-real-time datasets If the data lag exceeds ~ 2–3 years, or the grid mix is changing quickly, run a sensitivity analysis or triangulate with a more recent source. Temporal and spatial representativeness should weigh as much as source authority. 5.1 Conclusions The carbon footprint of electricity is not a single, fixed value, but a context-dependent figure shaped by deliberate methodological choices. This study has systematically dissected the impact of those choices, mapping the landscape of divergence from technology-level data to national averages and monetary factors. By transforming these quantitative findings into a practical set of guiding principles, we have provided the tools for analysts, companies, and policymakers to move from arbitrary data selection to conscious, evidence-based choices. These conclusions, however, must be viewed considering the study's methodological boundaries. Our analysis was intentionally focused on a production-based perspective and limited to attributional models to ensure a fair comparison of the underlying data sources. This highlights two critical avenues for future research: first, integrating flow-tracing algorithms to extend this comparative analysis to consumption-based footprints, and second, performing an equally systematic comparison of consequential (marginal) EFs, which is essential for guiding policy on demand-side management and grid decarbonization. It is important to note that the GHG Protocol is already moving toward an update of the Scope 2 Guidance, with increasing emphasis on the temporal and geographical granularity underpinning low-carbon electricity claims. While this represents a relevant advancement, particularly for market-based accounting, it only partially addresses the challenges highlighted in this study, which primarily concern the consistency and comparability of emission factors across data sources and methodologies.. In parallel, the joint effort between GHG Protocol and ISO to establish a harmonized standard represents a significant step forward in enhancing comparability. Nonetheless, these initiatives on their own may prove insufficient. Without more reliable primary data, clear and definitive guidance on temporal and spatial granularity (including an explicit and unique indication on the treatment of imports and exports), robust residual-mix methodologies where applicable, a more precise delineation between location- and market-based accounting, and systematic documentation of sources and versions, significant discrepancies are likely to persist. In essence, while stronger rules are essential, the ultimate quality of Scope 2 reporting will continue to depend on rigorous data and uncompromising transparency. Ultimately, in a rapidly decarbonizing context - and amid polarized views on sustainability regulation - credibility in climate policy and corporate reporting depends on transparent data sourcing and clear communication of analytical boundaries. A practical way to support this credibility is to anchor inventories in a centralized and regularly updated reference dataset, applied consistently over time and complemented by integrated primary data to progressively enable more robust hybrid assessments. Declarations Conflict of Interest Statement: The authors declare no conflict of interest. FUNDING INFORMATION The authors received no external funding. Author Contribution CC: Conceptualization, Data curation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing NG: Conceptualization, Data curation, Methodology, Writing – original draft, Writing – review & editing LR: Conceptualization, Data curation, Methodology, Software, Writing – original draft, Writing – review & editing. MVR: Conceptualization, Methodology, Writing – review & editing. Acknowledgement The authors acknowledge support from the Horizon Europe Innovation Action program project ENTICE (Grant No. 101184775). The content of this paper does not necessarily reflect the opinions of the European Commission and the responsibility for it lies solely with its authors. Data Availability The data that supports the findings of this study are available in the online resources of this article. Restrictions apply to the availability of IEA data, which were used under license for this study. Data are available at [https://www.iea.org/data-and-statistics/data-product/emissions-factors-2024](https:/www.iea.org/data-and-statistics/data-product/emissions-factors-2024) . Ecoinvent data used for the study were not directly extracted from Ecoinvent database but from the supporting materials of Steubing et al , published as open access material under CC-BY-NC 4.0 license. The utilization and dissemination of data from the GTAP database is permitted: the application in fact encompasses a single commodity and 28 geographical regions. Supporting Information Supporting information is linked to this article on the JIE website: References Asdrubali, F., Baldinelli, G., D’Alessandro, F., & Scrucca, F. (2015). Life cycle assessment of electricity production from renewable energies: Review and results harmonization. Renewable and Sustainable Energy Reviews , Vol. 42, pp. 1113–1122. Elsevier Ltd. https://doi.org/10.1016/j.rser.2014.10.082 Barros, M. V., Salvador, R., Piekarski, C. M., de Francisco, A. C., & Freire, F. M. C. S. (2020, January 1). Life cycle assessment of electricity generation: a review of the characteristics of existing literature. International Journal of Life Cycle Assessment , Vol. 25, pp. 36–54. Springer. https://doi.org/10.1007/s11367-019-01652-4 Bastos, J., Monforti-Ferrario, F., & Melica, G. (2024). GHG Emission Factors for Electricity Consumption . European Commission, Joint Research Centre (JRC). Retrieved from http://data.europa.eu/89h/919df040-0252-4e4e-ad82-c054896e1641 Beltrami, F., Grossi, L., & Liebensteiner, M. (2025). Beyond borders: Estimating the marginal emission factor of electricity trade. Energy Economics , 148 . https://doi.org/10.1016/j.eneco.2025.108588 Bertolini, M., Duttilo, P., & Lisi, F. (2025). Accounting carbon emissions from electricity generation: A review and comparison of emission factor-based methods. Applied Energy , 392 . https://doi.org/10.1016/j.apenergy.2025.125992 Blizniukova, D., Holzapfel, P., Unnewehr, J. F., Bach, V., & Finkbeiner, M. (2023). Increasing temporal resolution in greenhouse gas accounting of electricity consumption divided into Scopes 2 and 3: case study of Germany. International Journal of Life Cycle Assessment , 28 (12), 1622–1639. https://doi.org/10.1007/s11367-023-02240-3 CA Senate. Climate Corporate Data Accountability Act . , (2023). California Senate. Clauß, J., Stinner, S., Solli, C., Lindberg, K. B., Madsen, H., & Georges, L. (2019). Evaluation method for the hourly average CO2eq. Intensity of the electricity mix and its application to the demand response of residential heating. Energies , 12 (7). https://doi.org/10.3390/en12071345 CSRD . , (2022). European Parliament and Council. EC JRC. (2024). FIGARO-E3 dataset. Retrieved September 9, 2025, from https://data.jrc.ec.europa.eu/collection/id-00403 ecoinvent. (2025). Retrieved August 27, 2025, from https://ecoinvent.org/ Eggleston, H. S., Buendia, L., Miwa, K., Ngara, T., & Tanabe, K. (2006). 2006 IPCC guidelines for national greenhouse gas inventories (Vol. 2). Institute for Global Environmental Strategies. eGrid. (2023). Summary Tables 2021 . Electricity Maps. (2024). Electricity Maps emission factors methodology . Retrieved from https://github.com/electricitymaps/electricitymaps-contrib/wiki/Emission-factors Electricity Maps. (2025). Regional emission factors. Retrieved February 2, 2025, from https://ww2.electricitymaps.com/methodology#carbon-intensity-and-emission-factors EMBER. (2025a). Data methodology . Retrieved from https://storage.googleapis.com/emb-prod-bkt-publicdata/public-downloads/ember_electricity_data_methodology.pdf EMBER. (2025b). Electricity Data Explorer - Carbon emissions intensity. Retrieved September 5, 2025, from https://ember-energy.org/data/electricity-data-explorer/?entity=World&chart=single_year&data=co2_intensity&fuel=total EMBER. (2025c, June 25). The electrification imperative. Retrieved August 27, 2025, from https://ember-energy.org/latest-insights/the-electrification-imperative/ ENTSO-E. (2025). ENTSO-E Trasparency Platform - Actual Generation per Production Type. Retrieved July 29, 2025, from https://transparency.entsoe.eu/generation/r2/actualGenerationPerProductionType European Commission. (2025a). Carbon Border Adjustment Mechanism. Retrieved August 27, 2025, from https://taxation-customs.ec.europa.eu/carbon-border-adjustment-mechanism_en European Commission. ESRS Voluntary Sustainability Reporting Standards for SMEs under the European Sustainability Reporting Standards framework . , (2025). Gibon, T., & Hahn Menacho, Á. (2023). Parametric Life Cycle Assessment of Nuclear Power for Simplified Models. Environmental Science and Technology , 57 (38), 14194–14205. https://doi.org/10.1021/acs.est.3c03190 Hagenaars, R. H., Heijungs, R., Tukker, A., & Wang, R. (2025, April 1). Hybrid LCA for sustainable transitions: principles, applications, and prospects. Renewable and Sustainable Energy Reviews , Vol. 212. Elsevier Ltd. https://doi.org/10.1016/j.rser.2025.115443 Hawkes, A. D. (2010). Estimating marginal CO2 emissions rates for national electricity systems. Energy Policy , 38 (10), 5977–5987. https://doi.org/10.1016/j.enpol.2010.05.053 Henriques, C. O., & Sousa, S. (2023, March 1). A Review on Economic Input-Output Analysis in the Environmental Assessment of Electricity Generation. Energies , Vol. 16. MDPI. https://doi.org/10.3390/en16062930 Hertwich, E. G., Gibon, T., Bouman, E. A., Arvesen, A., Suh, S., Heath, G. A., … Shi, L. (2015). Integrated life-cycle assessment of electricity-supply scenarios confirms global environmental benefit of low-carbon technologies. Proceedings of the National Academy of Sciences of the United States of America , 112 (20), 6277–6282. https://doi.org/10.1073/pnas.1312753111 Howarth, R. W. (2024). The greenhouse gas footprint of liquefied natural gas (LNG) exported from the United States. Energy Science and Engineering , 12 (11), 4843–4859. https://doi.org/10.1002/ese3.1934 IEA. (2024a). Emission factors 2024 - database documentation . Retrieved from https://iea.blob.core.windows.net/assets/adcb9ea4-fc85-4379-826f-bbdd57401fa5/IEA_Methodology_Emission_Factors_2024.pdf IEA. (2024b). Emissions Factors 2024. Retrieved August 27, 2025, from https://www.iea.org/data-and-statistics/data-product/emissions-factors-2024 IPCC. (2014). Mitigation of Climate Change - Contribution of Working Group III to the Fifth Assessment Report of the IPCC . JRC. (2024). GHG Emission Factors for Electricity Consumption. Retrieved September 25, 2025, from https://data.jrc.ec.europa.eu/dataset/919df040-0252-4e4e-ad82-c054896e1641 JRC. (2025a). Environmental footprints based on FIGARO - Greenhouse gas emissions (in CO2 equivalent). Retrieved September 19, 2025, from https://ec.europa.eu/eurostat/web/environment/database#Greenhouse%20gas%20emissions%20(in%20CO2%20equivalent) JRC. (2025b). FIGARO tables, Annual EU inter-country supply, use and input-output tables. Retrieved September 19, 2025, from https://circabc.europa.eu/ui/group/cec66924-a924-4f91-a0ef-600a0531e3ba/library/d74cb7c1-bf93-4414-868c-89ab25e3e676 Khan, I. (2019, March 1). Greenhouse gas emission accounting approaches in electricity generation systems: A review. Atmospheric Environment , Vol. 200, pp. 131–141. Elsevier Ltd. https://doi.org/10.1016/j.atmosenv.2018.12.005 Lamb, W. F., Wiedmann, T., Pongratz, J., Andrew, R., Crippa, M., Olivier, J. G. J., … Minx, J. (2021). A review of trends and drivers of greenhouse gas emissions by sector from 1990 to 2018. Environmental Research Letters , 16 (7), 073005. https://doi.org/10.1088/1748-9326/abee4e Majeau-Bettez, G., Strømman, A. H., & Hertwich, E. G. (2011). Evaluation of Process- and Input–Output-based Life Cycle Inventory Data with Regard to Truncation and Aggregation Issues. Environmental Science & Technology , 45 (23), 10170–10177. https://doi.org/10.1021/es201308x Merciai et al. (2025). EXIOBASE hybrid - Consequential System. Retrieved September 24, 2025, from https://zenodo.org/records/15421526 Messagie, M., Mertens, J., Oliveira, L., Rangaraju, S., Sanfelix, J., Coosemans, T., … Macharis, C. (2014). The hourly life cycle carbon footprint of electricity generation in Belgium, bringing a temporal resolution in life cycle assessment. Applied Energy , 134 , 469–476. https://doi.org/10.1016/j.apenergy.2014.08.071 Moran, D., & Wood, R. (2014). Convergence between the EORA, WIOD, EXIOBASE, and OPENEU’s consumption-based carbon accounts. Economic Systems Research , 26 (3), 245–261. https://doi.org/10.1080/09535314.2014.935298 NREL. (2021). Life Cycle Greenhouse Gas Emissions from Electricity Generation: Update Life Cycle Assessment of Energy Systems (National Renewable Energy Laboratory) . Retrieved from https://data.nrel.gov/submissions/171 Pehl, M., Arvesen, A., Humpenöder, F., Popp, A., Hertwich, E. G., & Luderer, G. (2017). Understanding future emissions from low-carbon power systems by integration of life-cycle assessment and integrated energy modelling. Nature Energy , 2 (12), 939–945. https://doi.org/10.1038/s41560-017-0032-9 Perkins, J., & Suh, S. (2019). Uncertainty Implications of Hybrid Approach in LCA: Precision versus Accuracy. Environmental Science & Technology , 53 (7), 3681–3688. https://doi.org/10.1021/acs.est.9b00084 Raadal, H. L., Vold, B. I., Myhr, A., & Nygaard, T. A. (2016). GHG emissions and energy performance of offshore wind power . Retrieved from https://www.sciencedirect.com/science/article/pii/S0960148113006654 Ryan, N. A., Johnson, J. X., & Keoleian, G. A. (2016). Comparative Assessment of Models and Methods to Calculate Grid Electricity Emissions. Environmental Science and Technology , 50 (17), 8937–8953. https://doi.org/10.1021/acs.est.5b05216 Ryan, N. A., Johnson, J. X., Keoleian, G. A., & Lewis, G. M. (2018). Decision Support Algorithm for Evaluating Carbon Dioxide Emissions from Electricity Generation in the United States. Journal of Industrial Ecology , 22 (6), 1318–1330. https://doi.org/10.1111/jiec.12708 Sarhan, A., Ramachandaramurthy, V. K., Sin, T. C., Walker, S. L., Salman, B., & Padmanaban, S. (2023). Assessment of Spatial and Temporal Modeling on Greenhouse Gas Emissions From Electricity Generation. IEEE Access , 11 , 97478–97492. https://doi.org/10.1109/ACCESS.2023.3258923 Scarlat, N., Prussi, M., & Padella, M. (2022). Quantification of the carbon intensity of electricity produced and used in Europe. Applied Energy , 305 . https://doi.org/10.1016/j.apenergy.2021.117901 Schäfer, M., Tranberg, B., Jones, D., & Weidlich, A. (2020). Tracing carbon dioxide emissions in the European electricity markets. 17th International Conference on the European Energy Market (EEM) . IEEE. Schlömer, S., Bruckner, T., Fulton, L., Hertwich Austria, E., McKinnon, A. U., Perczyk, D., … Minx, J. (2014). ANNEX III Technology-specific Cost and Performance Parameters (Fifth Assessment Report of the Intergovernmental Panel on Climate Change) . Stadler et al. (2021a). EXIOBASE v3.3.18 Hybrid Multi Regional Supply and Use Tables. Retrieved September 9, 2025, from https://zenodo.org/records/10148587 Stadler et al. (2021b). EXIOBASE v3.8.2 Monetary Multi Regional Supply and Use Tables. Retrieved September 9, 2025, from https://zenodo.org/records/5589597 Stadler et al. (2025). EXIOBASE v3.9.5. Retrieved August 27, 2025, from https://zenodo.org/records/14869924 Stadler, K., Wood, R., Bulavskaya, T., Södersten, C. J., Simas, M., Schmidt, S., … Tukker, A. (2018). EXIOBASE 3: Developing a Time Series of Detailed Environmentally Extended Multi-Regional Input-Output Tables. Journal of Industrial Ecology , 22 (3), 502–515. https://doi.org/10.1111/jiec.12715 Steubing, B., de Koning, A., Merciai, S., & Tukker, A. (2022). How do carbon footprints from LCA and EEIOA databases compare? A comparison of ecoinvent and EXIOBASE. Journal of Industrial Ecology , 26 (4), 1406–1422. https://doi.org/10.1111/jiec.13271 Suh, S., Lenzen, M., Treloar, G. J., Hondo, H., Horvath, A., Huppes, G., … Norris, G. (2004). System Boundary Selection in Life-Cycle Inventories Using Hybrid Approaches. Environmental Science & Technology , 38 (3), 657–664. https://doi.org/10.1021/es0263745 Tahavori, M. A., Golinucci, N., Rinaldi, L., Rocco, M. V., & Colombo, E. (2023). MARIO: A Versatile and User-Friendly Software for Building Input-Output Models. Journal of Open Research Software , 11 . https://doi.org/10.5334/JORS.473 Tranberg, B., Corradi, O., Lajoie, B., Gibon, T., Staffell, I., & Andresen, G. B. (2019). Real-time carbon accounting method for the European electricity markets. Energy Strategy Reviews , 26 . https://doi.org/10.1016/j.esr.2019.100367 Turconi, R., Boldrin, A., & Astrup, T. (2013). Life cycle assessment (LCA) of electricity generation technologies: Overview, comparability and limitations. Renewable and Sustainable Energy Reviews , Vol. 28, pp. 555–565. Elsevier Ltd. https://doi.org/10.1016/j.rser.2013.08.013 UNECE. (2021). Carbon Neutrality in the UNECE Region: Integrated Life-cycle Assessment of Electricity Sources . United Nations. (2018). Handbook on Supply and Use Tables and Input Output-Tables with Extensions and Applications . Unnewehr, J. F., Weidlich, A., Gfüllner, L., & Schäfer, M. (2022). Open-data based carbon emission intensity signals for electricity generation in European countries – top down vs. bottom up approach. Cleaner Energy Systems , 3 . https://doi.org/10.1016/j.cles.2022.100018 Watershed. (2025). Open CEDA by Watershed. Retrieved August 27, 2025, from https://registry.opendata.aws/open-ceda World Resource Institute. (2011). Corporate Value Chain (Scope 3) Standard . Retrieved from https://ghgprotocol.org/corporate-value-chain-scope-3-standard World Resource Institute. (2023). GHG Protocol Scope 2 Guidance . Retrieved from https://ghgprotocol.org/scope-2-guidance Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial1.docx Supporting Information S1: This supporting information provides extended data, methods, and examples underlying the analysis of electricity carbon footprints [docx] SupplementaryMaterial2.xlsx Supporting Information S2: This supporting information provides the data points of carbon intensity and footprint from the considered sources, and a statistical analysis of them, both for technology-level data and for national data; an extended version of Table 1 is also included. [xlsx] Graphicalabstract.tif Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 06 May, 2026 Reviews received at journal 05 May, 2026 Reviews received at journal 25 Apr, 2026 Reviewers agreed at journal 21 Apr, 2026 Reviewers agreed at journal 14 Apr, 2026 Reviewers invited by journal 12 Apr, 2026 Editor assigned by journal 08 Apr, 2026 Submission checks completed at journal 08 Apr, 2026 First submitted to journal 06 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9337261","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":627653305,"identity":"d4c36745-3ea7-46b6-945e-499ae804b5d0","order_by":0,"name":"Camilla Citterio","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/0lEQVRIiWNgGAWjYJCCAw8KgCQzmC3BwM/AwHggAchgw6clwQCkhZmxAaRFsgEkAtKCTw9YCwNYCwODwQGgFrAwDi387acTgbbYARn8xx983GMhZ3y7+cCBhzkWDHzyDVi1SJzJ3QDUkswgcZiZsXHGMwljszvHEg4kbsPtMAMGsBZmBgOgX5p5DkgkbruRY4BfC/9bkJZ6iJY/QC2bZxDSIgG25TBECwNQywYJAlokboBtOc4D9IvhzJ4DEsYSUL/wsLElYA+x/tzNHz5UVMvx9x988OHHgTo5/tnNBx/+3FYnJ998ALs1UMCDZDGGCEEgQYLaUTAKRsEoGBEAAGD1W3L4syz3AAAAAElFTkSuQmCC","orcid":"","institution":"Politecnico di Milano","correspondingAuthor":true,"prefix":"","firstName":"Camilla","middleName":"","lastName":"Citterio","suffix":""},{"id":627653306,"identity":"d64bcae4-b721-4b65-9b7b-55bd86479999","order_by":1,"name":"Nicolò Golinucci","email":"","orcid":"","institution":"eNextGen S.r.l","correspondingAuthor":false,"prefix":"","firstName":"Nicolò","middleName":"","lastName":"Golinucci","suffix":""},{"id":627653309,"identity":"e22bd825-b3ce-4741-b8b8-91d4a77ab13a","order_by":2,"name":"Lorenzo Rinaldi","email":"","orcid":"","institution":"Politecnico di Milano","correspondingAuthor":false,"prefix":"","firstName":"Lorenzo","middleName":"","lastName":"Rinaldi","suffix":""},{"id":627653310,"identity":"405d7d74-d7aa-41f3-9085-f65528cc831b","order_by":3,"name":"Matteo Vincenzo Rocco","email":"","orcid":"","institution":"Politecnico di Milano","correspondingAuthor":false,"prefix":"","firstName":"Matteo","middleName":"Vincenzo","lastName":"Rocco","suffix":""}],"badges":[],"createdAt":"2026-04-06 21:09:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9337261/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9337261/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107716558,"identity":"1cff41e0-8387-46cf-a0d7-11541013a19d","added_by":"auto","created_at":"2026-04-24 10:07:03","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":595055,"visible":true,"origin":"","legend":"\u003cp\u003ePhysical-unit emission factors of electricity generation technologies by source for EU27 and US (gCO2eq/kWh). From top to bottom: a) carbon intensities; b) carbon footprints of all technologies; c) focus on carbon footprints of clean technologies. Each box spans the interquartile range (IQR), with the central line and dashed line indicating the median and mean, respectively; whiskers extend to the most extreme values within 1.5 × IQR, with quartiles computed using linear interpolation.\u003c/p\u003e\n\u003ch3\u003e\u003cbr\u003e\u003c/h3\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-9337261/v1/ae19029f46e96e564ba6c2d3.png"},{"id":107716560,"identity":"291a0b9b-0fd9-46c1-9b8d-ac0f5c24a63b","added_by":"auto","created_at":"2026-04-24 10:07:03","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":547300,"visible":true,"origin":"","legend":"\u003cp\u003ePhysical-unit carbon intensity (a,c) and footprint (b,d) of electricity generation at national level by country (gCO\u003csub\u003e2\u003c/sub\u003eeq/kWh) for 2017 (a,b) and 2023 (c,d). Different colors identify the original unit of measure in which data was provided: physical (mass of emissions per physical unit of energy), hybrid (hybrid SUT table), monetary (mass of emissions per monetary unit of energy, converted using GTAP prices (Aguilar et al., 2022)). The countries Cyprus, Malta and Luxembourg were not represented due to absence of data for numerous sources.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-9337261/v1/5b4be75cd45d98b56edb4e82.png"},{"id":108181076,"identity":"041014c5-497d-43f9-8574-0bd8e4c91deb","added_by":"auto","created_at":"2026-04-30 08:56:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1419627,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9337261/v1/455ae10b-edc4-4ab5-8eae-4fa832bd3acc.pdf"},{"id":107716559,"identity":"f96495d3-9cf9-47c8-8a3e-202969af089c","added_by":"auto","created_at":"2026-04-24 10:07:03","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":786202,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupporting Information S1: \u003c/strong\u003eThis supporting information provides extended data, methods, and examples underlying the analysis of electricity carbon footprints [docx]\u003c/p\u003e","description":"","filename":"SupplementaryMaterial1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9337261/v1/b2fef7177c7933332895cf4c.docx"},{"id":107869760,"identity":"21012295-8f3d-4153-9c93-a8030249d989","added_by":"auto","created_at":"2026-04-27 07:38:06","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":233372,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupporting Information S2: \u003c/strong\u003eThis supporting information provides the data points of carbon intensity and footprint from the considered sources, and a statistical analysis of them, both for technology-level data and for national data; an extended version of Table 1 is also included. [xlsx]\u003c/p\u003e","description":"","filename":"SupplementaryMaterial2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9337261/v1/f72f20e21ab1d716f96cab58.xlsx"},{"id":108005980,"identity":"efd29533-ca57-404f-b79e-20a1964bf968","added_by":"auto","created_at":"2026-04-28 12:51:25","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":158770,"visible":true,"origin":"","legend":"","description":"","filename":"Graphicalabstract.tif","url":"https://assets-eu.researchsquare.com/files/rs-9337261/v1/486ffea56034d0e57da03623.tif"}],"financialInterests":"No competing interests reported.","formattedTitle":"Carbon footprint of electricity: a systematic methodological and quantitative review","fulltext":[{"header":"1. INTRODUCTION","content":"\u003cp\u003eThe generation of electricity constitutes a substantial portion of global greenhouse gases (GHGs) emissions. Around one quarter of worldwide GHGs emissions can be attributed to electricity and heat production, as stated by (Lamb et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and confirmed by observing this metric from environmentally extended input-output (EEIO) tables: in 2015, they accounted for 26% according to EXIOBASE (Stadler et al., 2021) and for 31% according to FIGARO-E3 (EC JRC, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). From EEIO databases, it is also possible to state that the contribution of power production over the total carbon footprint of any marketed commodity ranges between 5% and 30% (see Online Resources, OR). Concurrently, the pace of electrification across economic sectors is accelerating rapidly (EMBER, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2025c\u003c/span\u003e). In this context, electricity plays a pivotal role in climate policies and emissions disclosure frameworks, spanning both voluntary schemes (e.g., ESRS Voluntary Sustainability Reporting Standards (ESRS Voluntary Sustainability Reporting Standards for SMEs under the European Sustainability Reporting Standards Framework, 2025)) and binding regulations, such as EU\u0026rsquo;s Corporate Sustainability Reporting Directive (CSRD, 2022), Carbon Border Adjustment Mechanism (European Commission, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) or the California\u0026rsquo;s Senate Bill 253 in the US (Climate Corporate Data Accountability Act, 2023).\u003c/p\u003e \u003cp\u003eThis evolving context is driving an increasing number of companies to measure and report emissions associated with their operations. In particular, emerging regulations require firms to assess their carbon footprint and to report emissions related to purchased electricity (Scope 2 emissions, i.e. emission associated with companies\u0026rsquo; energy supply), emphasizing the need for accurate and reliable emission factors for consumed electricity, typically following standardized approaches, like the GHG Protocol (World Resource Institute, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Moreover, implicit assumptions about the quantification of national electricity mixes add uncertainty and significantly influence the calculation of Scope 3 emissions (e.g. those emissions occurring across the value chains). Indeed, as corporations reporting total GHG inventories must evaluate both Scope 2 and Scope 3, they frequently introduce systemic inconsistencies by employing disparate data sources and conflicting hypotheses for each scope, despite electricity's pervasive and dominant impact across the entire value chain.\u003c/p\u003e \u003cp\u003e(Davis et al., 2025) show that 25.8% of 10'867 reporting companies in 2023 used spend-based approaches, typically relying on input-output sources. Among those specifying the method, 75% adopted single-region EEIO models, implicitly assuming domestic technologies. This leads to applying domestic electricity and emission intensities to imported goods, thereby underestimating the often higher carbon intensity of global supply chains.\u003c/p\u003e \u003cp\u003eDespite the critical importance of electricity in climate mitigation strategies, there is currently no universally agreed methodology, nor a universally accepted reference source, for estimating the carbon footprint per kWh of electricity. Instead, several authoritative sources \u0026ndash; including the Intergovernmental Panel on Climate Change (IPCC) assessment reports (ARs), life cycle assessment (LCA) databases (Ecoinvent, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), international agencies such as the International Energy Agency (IEA, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024b\u003c/span\u003e), widely cited commercial solutions by scientifically reputable sources, such as Electricity Maps (Electricity Maps, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and input-output tables such as EXIOBASE (Stadler et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) or CEDA (Watershed, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), widely utilized both in the academic and corporate sustainability contexts, often report misaligned values for the environmental impact of electricity production,. Previous works highlighted how the absence of harmonized protocols can lead to divergent estimates: differences in standards and reporting conventions generate substantial discrepancies in CO₂ estimates, resulting in notable variation among published grid emission factors, despite considering the same technology and/or the same geographical boundary (Hawkes, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Pehl et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Turconi et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Moreover, in many cases, even high-resolution databases yield misleading results as they often embed outdated electricity supply mix data. Electricity data in the most commonly utilized life cycle databases typically suffer from a significant time lag - averaging between 4 and 8 years - in accurately representing the current state of national power mixes, although the frequency of publication and updates has increased in recent years (Olindo et al., 2021). Such variability arises from several scientific and methodological factors, including the definition of system boundaries (e.g., direct combustion emissions only versus broader upstream and downstream coverage through life-cycle approaches (Asdrubali, Baldinelli, D\u0026rsquo;Alessandro, \u0026amp; Scrucca, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Barros et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; \u0026ldquo;Ecoinvent,\u0026rdquo; 2025; National Renewable Energy Laboratory, 2021)), assumptions about electricity mix composition and plant efficiencies, and choices regarding marginality (i.e., consequential versus attributional approaches (Beltrami et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Ryan et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Additional drivers are the selection and resolution of input data, encompassing not only geographic and technological detail but also temporal resolution, as power grids become increasingly heterogeneous with high penetration of renewable generation (Blizniukova et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Khan, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Examples range from physically grounded but narrow-boundary approaches, such as process-based LCA, and monetary input\u0026ndash;output approaches \u0026ndash; or hybrid datasets like EXIOBASE Hybrid (Merciai \u0026amp; Schmidt, 2021)\u0026ndash;that are more internally consistent from a modeling standpoint but less granular.\u003c/p\u003e \u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003e1.1 Objective and contribution\u003c/h2\u003e \u003cp\u003eOur study maps the main families of sources, used both at academic and corporate level, that estimate the carbon intensity and carbon footprint of electricity, undertaking a systematic, methodologically transparent comparison focusing on EU27 countries and the US.\u003c/p\u003e \u003cp\u003eTo ensure commensurability, all series are harmonized and aligned to a shared technology taxonomy. Results are presented at technology and national‑mix level under both direct and indirect emissions perspectives.\u003c/p\u003e \u003cp\u003eWhile previous research has explored discrepancies within subsets of electricity carbon footprint studies, such as process-LCA meta-reviews or EEIO database comparisons, no prior work has provided an integrated, cross-paradigm assessment that aligns EEIO, P-LCA and agency-based datasets under a harmonized taxonomy and GWP framework.\u003c/p\u003e \u003cp\u003eThe performed cross-sources comparison is also intended to provide a practical compass for analysts and decision-makers in navigating a fragmented evidence base, clarifying what each source measures, when it is most appropriate, and how boundary choices and methodological assumptions shape results. Comparative findings are translated into selection criteria and reporting conventions that support deliberate, defensible choices (i.e. whether for policy appraisal, procurement, or corporate disclosure) so that users can select sources consciously instead of arbitrarily.\u003c/p\u003e \u003c/div\u003e"},{"header":"2. LITERATURE REVIEW","content":"\u003cp\u003eWe analyzed 39 studies published after 2013 to identify the main methodological challenges in quantifying emissions from electricity generation, mainly distinguished on three key levels of complexity: (i) the quantification of emissions factors for specific generation technologies; (ii) the aggregation of these into grid-level intensities; and (iii) the comparison of results from fundamentally different modeling paradigms, such as process-based LCA (P-LCA) and EEIO models. Before presenting the literature review, we introduce the terminology adopted throughout the article.\u003c/p\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Terminology\u003c/h2\u003e \u003cp\u003eA primary obstacle to comparing emission data is inconsistent terminology. To ensure clarity, we distinguish between two key emissions factors based on system boundaries. We define \u0026ldquo;carbon intensity\u0026rdquo; (CI) as the direct, operational emissions from generating 1 kWh of electricity (equivalent to Scope 1 for a power producer). In contrast, we use \u0026ldquo;carbon footprint\u0026rdquo; (CF) as a generic term for any emission factor that goes beyond CI. This broader definition is needed because different sources draw system boundaries in different ways, sometimes including upstream processes (e.g., fuel extraction, plant construction) and downstream impacts (e.g. transmission and distribution losses). The generic term \u0026ldquo;emission factor\u0026rdquo; (EF) will be used to refer to both CI and CF when the distinction is not critical. Furthermore, a crucial methodological divide, central to the following review, is the distinction between attributional and consequential (or marginal) modeling. Attributional EFs allocate the total emissions of a grid across all consumed electricity, answering the question: \"What is the average footprint of the electricity mix?\". Consequential EFs, instead, estimate the change in system-wide emissions resulting from a change in demand, answering: \"Which power plant will operate to meet an additional kWh of consumption?\". For this article we focus on national-scale attributional EFs, since they represent the common ground across multiple applications, including corporate carbon footprint. Consequential approaches, on the other hand, are more suitable when the objective is to provide behavioral or price signals for short-term applications, such as daily or weekly operations (e.g., smart charging), requiring dedicated databases and tools (Merciai et al, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Technology specific emission factors\u003c/h2\u003e \u003cp\u003eThe foundational challenge in electricity carbon accounting lies in the wide availability and heterogeneity of published EFs for individual generation technologies. Comprehensive reviews by (Khan, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) and (Barros et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) systematically documented this issue, identifying the primary drivers of divergence: inconsistent system boundaries, methodological choices, and different geographical or technological assumptions. Earlier works, such as (Turconi et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) and (Asdrubali et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), conducted thorough comparisons of LCA analyses\u0026rsquo; results for different technologies demonstrating how even subtle analytical choices can yield significantly different results (e.g. variations exceeding a factor of 10 for renewable technologies in Turconi et al., while harmonized data in Asdrubali et al. show maximum differences of 500%). A vast body of literature exists performing LCAs of specific electricity technologies, such as the parametric analysis of nuclear power by (Gibon \u0026amp; Hahn Menacho, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and the work of (Raadal et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) on offshore wind turbines. Other studies, like (Hertwich et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), perform environmental impact analysis for various technologies through the same methodology, in this case scenario-based hybrid LCA.\u003c/p\u003e \u003cp\u003eThe lack of comparability created a clear need for standardized benchmarks. In response, major institutional efforts emerged to harmonize these values. The National Renewable Energy Laboratory's (NREL) work on establishing statistical ranges (Asdrubali et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and the IPCC\u0026rsquo;s provision of harmonized CF ranges (Schl\u0026ouml;mer et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) have become cornerstone references in the field (Clau\u0026szlig; et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Unnewehr et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). These IPCC values, alongside its widely used combustion factors and methodological guidelines (Blizniukova et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Scarlat et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), provide a crucial, authoritative starting point for analysis. However, they represent technology-level potentials, and the next layer of complexity arises when aggregating them to model real-world, interconnected electricity grids.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Grid-level emission factors\u003c/h2\u003e \u003cp\u003eMoving from individual plants to national grids introduces a further layer of methodological complexity. While some studies employ top-down approaches using aggregated national statistics on emissions and generation (Unnewehr et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), the prevalent methods are \"bottom-up,\" building upon technology-specific factors. Even within this dominant category, significant variability exists:(Bertolini et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) compared six different EF-based methods for Italian electricity zones, finding a 12% spread between minimum and maximum average EFs in the North zone alone, and underscoring that no single approach is universally applicable.\u003c/p\u003e \u003cp\u003e(Ryan et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), systematically mapped 32 methodological approaches, classifying them into \"Empirical Data and Relationship Models\" (including simple EFs, commercial exchanges and statistical relationship models) and \"Power System Optimization Models\". Their US case study showed that the choice of method alone could alter the CI for vehicle charging by up to 68%, a variation often larger than geographic location. Their subsequent recommendations (Ryan et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) highlight the pivotal distinction between attributional and marginal EFs: studies on cross-border electricity trade (Beltrami et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) show that marginal generation technologies in highly interconnected European markets often diverge substantially from the average mix, with spatially resolved marginal factors (Sarhan et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) enabling more granular locational analysis.\u003c/p\u003e \u003cp\u003eA more computationally intensive class of methods addresses cross-border carbon accounting through flow-tracing algorithms. (Tranberg et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) and (Sch\u0026auml;fer et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) applied these techniques across the EU, revealing significant discrepancies between a country's production-based and consumption-based CFs. For instance, nations with large low-carbon generation shares, such as Slovakia and Austria, see their consumption footprints increase due to electricity imports from higher-emission neighbors like Poland and Czech Republic (Tranberg et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). By incorporating high temporal resolution and explicit trade modeling, these studies demonstrate substantial sub-monthly variations in grid CF. Such time-resolved, flow-tracing approaches enable a more accurate allocation of emissions from producers to final consumers, providing valuable data for actors who can control their electrical loads (Clau\u0026szlig; et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Messagie et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Divergences in foundational databases\u003c/h2\u003e \u003cp\u003eA final and fundamental source of variation stems from the databases used to model background economies and supply chains. Practitioners must often choose between two distinct paradigms: bottom-up P-LCA databases and top-down, economy-wide EEIO tables. The fundamental differences between these approaches have long been recognized in the debates on completeness and truncation errors: (Suh et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) demonstrated that system boundary selection in LCA significantly affects results, with IO-based approaches capturing a broader economic scope that process-based inventories systematically truncate. (Majeau-Bettez et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) further quantified these truncation and aggregation issues, showing that P-LCA databases tend to underestimate impacts due to incomplete upstream coverage. A comparative work by (Steubing et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) directly addressed this issue by matching product CFs between Ecoinvent and EXIOBASE. Despite both sources being used for similar purposes, the results showed significant discrepancies, with over half the matched products exhibiting CF differences greater than a factor of two. While concordance was better for fossil energy and manufacturing, the discrepancies for electricity from renewable sources were particularly pronounced, driven by differing regional and technological disaggregation.\u003c/p\u003e \u003cp\u003eThe comparison highlighted fundamental trade-offs: the comprehensive economic coverage of EEIO tables appears to be counterbalanced by the more detailed inclusion of capital goods and multi-year timeframes in P-LCA. This tension is further analysed by (Henriques \u0026amp; Sousa, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), who show for electricity generation that EEIO-based assessments tend to provide systematically higher impacts than P-LCA, advocating for hybrid options to leverage the strengths of both paradigms. This claim has been strongly supported by (Hagenaars et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), while (Perkins \u0026amp; Suh, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) provide empirical evidence that hybridization can improve accuracy by combining the completeness of IO boundaries with the detail of process data.\u003c/p\u003e \u003cp\u003eHowever, results can vary even within the same paradigm. A comparison of four major multi-regional input-output databases by (Moran \u0026amp; Wood, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) found that while the underlying economic structures were broadly similar, divergences in CFs were primarily driven by the satellite environmental accounts used for GHG emissions. Crucially, they demonstrated that harmonizing these satellite accounts alone could reduce the divergence between major economies' footprints to under 10%. This finding underscores that both high-level modeling choices (P-LCA vs. EEIO) and low-level data compilation decisions contribute significantly to the fragmented evidence base that practitioners currently face.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Research gap identification\u003c/h2\u003e \u003cp\u003eUltimately, the body of analysed studies leaves the practitioner facing a series of critical, yet unresolved, methodological forks in the road. In the context of corporate carbon footprint assessment, established protocols such as the GHG Protocol prescribe an annual perspective. Within this temporal scope, the attributional approach represents the standard methodological choice. Nevertheless, an analyst must choose not only a set of technology-specific EFs, but also an aggregation method, and, at an even more fundamental level, a foundational database paradigm (P-LCA or EEIO) adopting the most accurate EF for a defined region and year. Our literature review demonstrates that each of these choices is highly influential yet offers no clear guidance on how to navigate them in concert. While individual studies have illuminated specific points of friction, the cumulative impact of these discrepancies remains unquantified in a single, harmonized framework. Addressing this requires moving from isolated comparisons to a systematic, multi-source confrontation.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. METHODS","content":"\u003cp\u003eOur analysis performs a comprehensive comparison of EFs drawn from a curated selection of widely used data sources, grouped into four main categories: EEIO tables, P-LCA databases, primary data from international agencies, and secondary data from literature harmonization projects. A detailed summary of these sources, including their reference years, scope, and key methodological characteristics, is provided in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cem\u003eEEIO databases\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe EEIO databases considered in this study are: EXIOBASE, in two distinct versions, the hybrid Supply-Use Table (SUT) format (v3.3.18, data year 2011) (Merciai \u0026amp; Schmidt, 2021) and the more recent (v3.10.1) monetary Input-Output Table (IOT) format (Stadler et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2025\u003c/span\u003e); GLORIA (v0.60, SUT format) (ielab, 2026); EMERGING (v199.82) (Huo et al., 2026); EORA26 (v199.82) (Lenzen et al., 2013) and the WIOD format of GTAP (v11) (Aguiar et al., 2022). The GTAP project also provides electricity price data for each region, as detailed in Section \u003cspan refid=\"Sec12\" class=\"InternalRef\"\u003e3.3\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cem\u003eP-LCA databases\u003c/em\u003e \u003c/p\u003e \u003cp\u003eEcoinvent, as the most utilized process-based database in industrial ecology, is a critical component of our analysis. Due to licensing constraints on publishing direct results, we leverage the data made available by the comparative study of (Steubing et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), which reports carbon footprints from Ecoinvent v3.4 (cut-off system model) with regional information removed.\u003c/p\u003e \u003cp\u003e \u003cem\u003eAgencies\u0026rsquo; primary data\u003c/em\u003e \u003c/p\u003e \u003cp\u003eWe include the authoritative International Energy Agency (IEA) emission factors for combustion technologies. The IEA provides country-specific CIs for 1990\u0026ndash;2023. As this data is licensed, our figures will not show these series. Two versions of IEA CIs are provided, differentiated by assumptions regarding the allocation of emissions from combined heat-power (CHP) plants: \u0026ldquo;Electricity only\u0026rdquo; assumes all CHP plant emissions are attributed to electricity generation, while \u0026ldquo;CHP\u0026rdquo; assumes a portion of the emissions are allocated to heat, resulting in a lower CI (IEA, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024a\u003c/span\u003e). One peculiarity of IEA factors is that they do not exclude CO₂ emissions of biogenic origin, as suggested by the IPCC guidelines (Eggleston et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2006\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cem\u003eSecondary and harmonized data\u003c/em\u003e \u003c/p\u003e \u003cp\u003eTo include influential reference values, we incorporate several sources: the IPCC, we use the benchmark technology-level carbon footprints reported by Working Group III in AR5 from 2014 (IPCC, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2014\u003c/span\u003e); the harmonized LCA emission estimates from (NREL, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) release. Finally, we consider Electricity Maps, a widely used platform providing real-time electricity flow data (Electricity Maps, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). The values considered are their declared, static emission factors for each region (Electricity Maps, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), which are themselves derived from a combination of other key sources, including IPCC AR5 and (UNECE, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) reports. We also include sources that provide ready-to-use, country-level EFs. From the European Commission's Joint Research Centre (JRC), we include their dataset providing both direct \"emissions intensity\" (CI) and \"lifecycle emissions intensity\" (CF) for EU27 countries (Bastos et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Since JRC data covers EU countries only, the CI for US is obtained from (eGrid, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). From EMBER, we use their \"Carbon emissions intensity\" data, which represent full life-cycle CFs (EMBER, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2025b\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025a\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eData sources considered in this study with versioning or publication year, detailing the acronym adopted henceforth, the data reference year, availability of region-specific data for EU and US, the type of emission factors analyzed and on which GHG substance, methodological treatment of co-production processes, data type and source references. Extended version in OR.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eData source and publication year (or version)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAcronym\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eData year\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDetail on US and EU\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eEmission factors analyzed\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGHGs included\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eHandling co-products\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eData type and unit\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eReference\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEXIOBASE Hybrid SUT v3.3.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEXIO Hybrid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e, SF\u003csub\u003e6\u003c/sub\u003e, HFCs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIndustry-based\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eEEIOT, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(Merciai \u0026amp; Schmidt, 2021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEcoinvent v3.4 via Steubing et al.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEcoinvent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e, SF\u003csub\u003e6\u003c/sub\u003e, HFCs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eCut-off\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eP-LCA, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(Ecoinvent, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Steubing et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIEA emission factors per plant 2024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIEA (CHP)\u003c/p\u003e \u003cp\u003eIEA (ele)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1990\u0026ndash;2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eElectricity and heat CHP/\u003c/p\u003e \u003cp\u003eElectricity output only\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003ePrimary data, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(IEA, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024b\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eElectricity Maps \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2025\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eElectricity Maps\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGeneric CO\u003csub\u003e2\u0026minus;eq\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSecondary data, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(Electricity Maps, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNREL harmonization 2021\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNREL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGeneric CO\u003csub\u003e2\u0026minus;eq\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSecondary data, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(NREL, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2021\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIPCC AR5 WG3 (Annex III)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIPCC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGeneric CO\u003csub\u003e2\u0026minus;eq\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSecondary data, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(Schl\u0026ouml;mer et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2014\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEXIOBASE IOT ixi/pxp v3.10.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEXIO ixi\u003c/p\u003e \u003cp\u003eEXIO pxp\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1995\u0026ndash;2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e, SF\u003csub\u003e6\u003c/sub\u003e, HFCs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIndustry-based technology assumption\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eEEIOT, monetary\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(Stadler et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEMERGING v2.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEMERGING\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2015\u0026ndash;2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGeneric CO\u003csub\u003e2\u0026minus;eq\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIndustry-based technology as.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eEEIOT, monetary\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(Huo et al., 2026)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEORA26 v199.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEORA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1990\u0026ndash;2023 (open access\u0026thinsp;\u0026minus;\u0026thinsp;2017)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIndustry-based technology as.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eEEIOT, monetary\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(EORA, 2026; Lenzen et al., 2013)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGLORIA v0.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGLORIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1990\u0026ndash;2025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e,CO, SF\u003csub\u003e6\u003c/sub\u003e, HFCs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIndustry-based technology as.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eEEIOT, monetary\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(ielab, 2026)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGTAP 11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGTAP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2004, 2007, 2011, 2014, 2017\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e,CO, SF\u003csub\u003e6\u003c/sub\u003e, HFCs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIndustry-based technology as.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eEEIOT, monetary\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(Aguiar et al., 2022)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEMBER 2025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEMBER\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2000\u0026ndash;2024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e˅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGeneric CO\u003csub\u003e2\u0026minus;eq\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSecondary data, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(EMBER, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2025b\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJRC CoM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eJRC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1990\u0026ndash;2021\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEU only\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI, CF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSecondary data, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(JRC, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eeGRID summary tables\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eeGRID\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2018\u0026ndash;2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eUS only\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003eO, CH\u003csub\u003e4\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSecondary data, physical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e(eGrid, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Data harmonization and processing\u003c/h2\u003e \u003cp\u003eTo enable a robust and fair comparison across the selected sources, a multi-step harmonization protocol was applied to align technology definitions, GHG characterization factors and analytical scope.\u003c/p\u003e \u003cp\u003eFirst, a common technological taxonomy was established. The diverse electricity generation categories reported in each source were mapped and consolidated to a common level of aggregation; details are documented in the OR.\u003c/p\u003e \u003cp\u003eSecond, all emission factors were standardized to a consistent climate metric. We express all results in terms of Global Warming Potential (GWP) over a 100-year horizon, employing the characterization factors from IPCC AR5, following what is done by similar works (Steubing et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and since EFs from the same version of the report are utilized in our analysis (IPCC, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe processing of EEIO databases was conducted using MARIO (Tahavori et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), an open-source Python package for input-output analysis. Within this framework, a critical distinction was made between activity-based and commodity-based footprints. CI is defined only at the activity level, as environmental accounts are linked to industry outputs. The CF of the electricity commodity, however, is calculated following the industry-based technology assumption (United Nations, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), by weighting the footprint of each electricity production activity, including those that are not power plant technologies (e.g. industries producing from PV rooftops), on their market share. This modeling approach is particularly relevant for capturing the effects of cogeneration, as detailed in the OR.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Derivation of grid electricity emission factors\u003c/h2\u003e \u003cp\u003eTo assess the impact of different data sources at a national level, we calculated grid-level emission factors for each EU country and the US as the weighted average of the technology-specific EFs from a given data source, with the weights determined by that country's annual electricity generation mix. The electricity supply mix data was obtained from the ENTSO-E Transparency Platform for EU countries (ENTSO-E, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and from the eGRID database for the US (eGrid, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Consistently with the comparative objective of the analysis, we limited our calculations to EFs from electricity generated within national borders as, according to Table\u0026nbsp;6.2 of the GHG Protocol Scope 2 guidelines, national production-based factors are an accepted and adopted option for corporate carbon footprint (World Resource Institute, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Moreover, production-based factors represent the necessary foundation for deriving accurate consumption-based factors.\u003c/p\u003e \u003cp\u003eIt should be noted that, since EXIOBASE Hybrid only provides data for 2011, we did not compute the CF as the table\u0026rsquo;s electricity commodity, but rather as the weighted average CF of electricity generation activities, as described in previous paragraph. This was done in order to avoid including the impact of any other activities that produce electricity as a co-product, to ensure that electricity supply mixes were the same as for IPCC and NREL results.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Emission factors from economic databases and conversion to physical units\u003c/h2\u003e \u003cp\u003eIn practice, compiling a corporate carbon footprint compliant with frameworks like the GHG Protocol typically relies on economic databases for Scope 3.1 emission factors, given the scarcity of primary data (David et al., 2023). While Scope 2 emissions are generally assessed using physical sources, electricity often drives a significant portion of the overall impact, necessitating consistent integration across all footprint elements. Consequently, this comparison includes monetary databases to evaluate their viability as a single, unified data source for comprehensive corporate carbon accounting. As mentioned, we analyzed EFs from EXIOBASE, available in monetary IOT format in its latest 3.10.1 version, from which we derived the CF of electricity, by aggregating all electricity production sectors into one, therefore weighting the EFs according to the generation shares included in the IOT. We followed the same procedure both for the product-by-product (pxp) and industry-by-industry (ixi) tables. Other EEIO tables directly provide one unique sector for electricity production: we extracted CI and CF from EMERGING, EORA, GLORIA, GTAP (see Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e for more details). The values obtained are emissions per monetary unit, in basic prices. The comparison with EFs in physical units is possible through the conversion with the regional price of electricity provided by GTAP; since those prices are available, for recent years, only for 2017 and 2023, those are the years selected for the comparative analysis in Section \u003cspan refid=\"Sec17\" class=\"InternalRef\"\u003e4.3\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Benchmark definition\u003c/h2\u003e \u003cp\u003eRegarding the CIs in physical units, IEA data were used as the reference benchmark, given their primary-data status and its widespread institutional authority. Given that the CF of a product depends on a wide range of factors that we have explicitly addressed, it is not meaningful to define a single benchmark. Instead, we compare the various sources by examining their relative deviations.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. RESULTS","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Technology-specific emission factors comparison\u003c/h2\u003e \u003cp\u003eOur cross-source comparison highlights significant variability in the estimation of electricity EFs. Even for direct Carbon Intensities (CIs), where methodological alignment might be expected, the sources exhibit substantial divergences rather than a clear consensus. Moreover, the picture for CFs shows wide-ranging disagreement and methodological heterogeneity, especially for renewable technologies.\u003c/p\u003e \u003cp\u003eThe analysis of CI for fossil fuel generation (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, where IEA values are not reported due to license limitations) demonstrates a higher consensus than the corresponding CF. For coal-fired power, the spread between medians, meaning the relative difference between minimum and maximum median, is 36%; for oil the spread is 52% and for natural gas it is 116%. For CFs instead the defined spread ranges from 101% to 631%. This suggests that, when the analytical scope is restricted to direct combustion emissions, estimates from different providers tend to be more comparable, although non-trivial discrepancies may persist..\u003c/p\u003e \u003cp\u003eHowever, a critical exception immediately highlights a primary source of confusion: the\u003c/p\u003e \u003cp\u003etreatment of biogenic CO\u003csub\u003e2\u003c/sub\u003e in biomass-based electricity. The IEA reports CI values for biomass that can exceed 2000 gCO₂eq/kWh, in stark contrast to other sources that report near-zero direct, non-biogenic emissions. This enormous discrepancy stems from a single accounting choice: IEA includes direct CO₂ emissions in its factors even if they are of biogenic origin (IEA, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024a\u003c/span\u003e), a practice contrary to IPCC guidelines (Eggleston et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), which consider such emissions to be carbon-neutral within the energy system's carbon cycle. This example is a stark illustration of how a differing interpretation of accounting rules, even for direct emissions, can lead to a result that is orders of magnitude different and potentially misleading for policy or reporting purposes.\u003c/p\u003e \u003cp\u003eIn contrast to the CI analysis, the comparison of CFs reveals larger disagreements across nearly all technologies (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb and Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). For renewables, where the footprint is dominated not by operational but by manufacturing and decommissioning emissions, the choice of data source becomes critical. For example, the reported CF for hydroelectric power in Ecoinvent, and the maximum of the range reported by IPCC, are substantially higher than in any other source, a result (Steubing et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) attribute to the inclusion of diverse plant types and fugitive reservoir emissions. Similarly, for nuclear power, the CF from Electricity Maps is less than half the median reported by the IPCC and falls below the minimum range of the NREL harmonization project. These discrepancies underscore that once the analytical boundary expands beyond the smokestack into the complexities of the full supply chain, any semblance of consensus vanishes, leaving the user in a landscape of highly variable data.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Systematic differences between EEIO and P-LCA Databases\u003c/h2\u003e \u003cp\u003eAcross the wide variability of CF data, a clear and actionable pattern emerges: the choice between a P-LCA and a EEIO database is not neutral. Our results show that the EEIO database EXIOBASE Hybrid consistently yields the highest or among the highest CF values for nearly every technology analyzed.\u003c/p\u003e \u003cp\u003eThis effect is highly visible for conventional generation technologies (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). For natural gas, the medians for IPCC and NREL align on 490 gCO₂eq/kWh, Electricity Maps\u0026rsquo; range is tightly concentrated around 529 (coefficient of variation, CV, of only 10%), Ecoinvent has a broader span with median 598 gCO₂eq/kWh. The median CF from EXIOBASE Hybrid, instead, reaches 780 gCO₂eq/kWh. This relatively high value can be attributed to the use of the hybrid EEIO framework, which captures the full range of transactions along the value chain, including services typically excluded from process-based LCA approaches. In addition, the industry-based technology assumption combined with the treatment of blast furnace gas as an input to the power sector indirectly activates the steel industry, thereby inflating the footprint. Despite these methodological effects, process-based studies in the recent literature analyzing the liquefied natural gas supply chain report estimates of a similar magnitude (Howarth, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The pattern is even more pronounced for coal: the entire distribution of CF values from EXIOBASE Hybrid lies above the third quartile of NREL, IPCC and Electricity Maps, effectively creating a separate, higher-impact class of results. Even Ecoinvent provides 26% of values below EXIOBASE Hybrid\u0026rsquo;s lower whisker and 60% of values below its first quartile.\u003c/p\u003e \u003cp\u003eThis effect persists, and in some cases is amplified, for renewable technologies (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). For wind power, all medians fall within the narrow range of 11\u0026ndash;23 gCO₂eq/kWh, EXIOBASE Hybrid's distribution is significantly higher and wider, with outliers surpassing 150 gCO₂eq/kWh. Similarly, for geothermal energy, EXIOBASE Hybrid's median CF is 53% higher than the value reported by the IPCC and Electricity Maps.\u003c/p\u003e \u003cp\u003eThis systematic difference is not an error, but a direct consequence of the fundamentally different analytical boundaries of the two paradigms, as explored in the literature review.\u003c/p\u003e \u003cp\u003eAcross the sources compared here, EEIO-based CF estimates are generally higher than process-LCA-based estimates within our matched sample of electricity technologies and national mixes. We do not interpret this pattern as evidence that EEIO results are necessarily closer to ground truth. Rather, the gap reflects differences in modelling architecture, including broader system boundaries and reduced truncation of upstream services and capital in EEIO, versus greater technological specificity but potentially narrower boundaries in P-LCA. The magnitude and even direction of the gap may also depend on aggregation, multifunctionality assumptions, data vintage, and the geographical representation of supply chains.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.3 National grid emission factors\u003c/h2\u003e \u003cp\u003eThe methodological divergences identified at the technology level directly propagate to the calculation of national grid emission factors, leading to a wide range of potential carbon intensity and footprint values for any given country. Analyzing the computed CIs, the difference between minimum and maximum estimates, relative to the IEA reference value, ranges between 72% and 730% for 2017 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea), and between 38% and 612% for 2023 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). This confirms that even for direct emissions, the methodological landscape is far from settled. When grouping databases by original unit (monetary vs. physical) a consistent pattern emerges: monetary-based sources tend to yield lower CIs estimates, a trend that is particularly pronounced for the year 2017, in which more than 90% of the minimum values across technologies are attributable to a monetary source. It should be noted, however, that monetary EFs are inherently mediated by electricity prices: a higher price assumption would allocate the same quantity of GHG emissions to a smaller electricity output, thereby resulting in a higher CI estimate.\u003c/p\u003e \u003cp\u003eIn examining the CF, it is observed that the spread (difference between minimum and maximum) of monetary-derived data is systematically wider than that of physical sources: on average, the range is approximately 150% larger for monetary than for physical databases in 2017. Furthermore, for the 2017 CF values, monetary sources fall below the country-level median in 60\u0026ndash;100% of cases (with exception of GTAP 11, which does so in only 12% of countries), whereas physical sources exceed the median in 64\u0026ndash;92% of countries (with the exception of IPCC-based values, only 32%). The hybrid EXIOBASE estimates, computed here in combination with respectively the 2017 and 2023 ENTSO-E electricity mix, stand above the country median in 96% of cases; moreover, for 20 out of 28 countries this source reports the highest CF value in 2017, and for 13 out of 28 in 2023.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"5. DISCUSSION","content":"\u003cp\u003eThis study set out to provide a practical compass for navigating the fragmented evidence on electricity's carbon footprint. Our results move beyond simply confirming the existence of discrepancies; by systematically examining them, we can translate our findings into a set of guiding principles. The core conclusion is that an optimal data source does not exist; the \"best\" choice is contingent on the user's specific application and analytical goals.\u003c/p\u003e \u003cp\u003eThe first critical decision point for any analyst concerns the analytical scope: reporting direct emissions (CI) versus full life-cycle impacts (CF). Our analysis reveals a moderate consensus for the CI of fossil fuels, suggesting that for straightforward applications like basic Scope 2 reporting, relying on primary data from sources like the IEA provides a widely accepted foundation. However, even within this consensus, our findings on biogenic carbon accounting warn that differing guideline interpretations can yield very different results, demanding user vigilance.\u003c/p\u003e \u003cp\u003eWhen the scope extends to CFs, the choice of modeling paradigm becomes the central strategic consideration. We consolidated a systematic yet typical \"paradigm effect,\" where EEIO databases tend to yield higher CFs than P-LCA sources. This pattern should not be interpreted as evidence of closer agreement with ground truth, but rather because of broader system boundaries and reduced truncation in EEIO models. This is not an error, but a feature that can be leveraged. If an analyst's objective is a conservative, upper-bound estimate that captures the broadest economic context (e.g., in a corporate risk assessment), then the use of an EEIO database is a defensible choice. Moreover, the EEIO framework is, in our view, the only approach that enables a comprehensive assessment of a company's carbon footprint while allowing for a systematic and coherent allocation of Scope 3 emissions. If, conversely, the goal is comparability with standard product LCAs or Environmental Product Declarations (EPDs), which predominantly use process-based data, selecting an LCA-based source is necessary to ensure methodological consistency.\u003c/p\u003e \u003cp\u003eA final, crucial distinction lies between physical and monetary units. Our analysis of monetary emission factors revealed a higher spread of values for monetary databases, and we highlight that the obtained EFs are inherently mediated by the price of electricity. This renders their uncritical use for formal reporting highly problematic. While convenient for high-level screening where accountability data is readily available, their application must be approached with extreme caution. Whenever accuracy is a priority, physical factors based on actual consumption (in MWh) should be the default and preferred method. However, for corporate carbon accounting the key question is not which paradigm is \u0026ldquo;truer\u0026rdquo;, but which data infrastructure enables coherent, repeatable annual inventories. In practice, mixing Scope 3.1 estimates based on monetary tables that are several years out of date with electricity factors drawn from heterogeneous physical sources is hard to justify especially when electricity can explain a large share of the footprint. This motivates the advantages of a centralized, integrated, regularly updated framework that uses physical data where feasible and monetary proxies where necessary, while preserving internal consistency across scopes and categories, and is complemented by primary data whenever available.\u003c/p\u003e \u003cp\u003eTo make these discrepancies concrete, consider a 10 MWh electricity bill corresponding to 1\u0026rsquo;600 EUR (basic price, 2023) in Italy. Utilizing the presented monetary and physical factors, the maximum total GHG emissions are 3 times the minimum value, with a difference of 4 tCO2eq in the considered year, a spread that underscores how methodological choice dominates results and why transparent reporting is non-negotiable.\u003c/p\u003e \u003cp\u003eUltimately, all these decisions are bound by one unbreakable rule: the imperative of radical transparency. Given that a country's calculated footprint can vary dramatically based on the data source, any reported EF must be accompanied by a clear citation of the underlying source and version. This is the minimum requirement for enabling reproducibility and allowing the audience to understand the context and limitations of the presented figures. These principles are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e as a practical guide for practitioners.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDecision framework for selecting electricity emission factor sources.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUse cases\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRecommended data type\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRationale and key considerations\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTerritorial electricity benchmarking / producer-side analysis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProduction-based CI or CF in physical units from primary or official sources\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eUse when the question concerns electricity generated within a territory. Suitable for producer-side benchmarking and comparability with generation statistics. Not ideal for end-use footprints where imports/exports materially alter the delivered mix.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScope 2 location-based accounting\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eConsumption-/delivery-oriented CI in physical units; production-based factors only as a proxy when better consumption-based data are unavailable\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePrefer in highly interconnected or import-dependent grids. Always disclose whether the factor is production- or consumption-based, together with its data year and treatment of imports/exports.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScope 2 market-based accounting\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSupplier-specific or residual-mix CI in physical units\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eUse only for contractual claims that meet Scope 2 quality criteria. This should be explicitly flagged as outside the main quantitative comparison in the present paper.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScope 3 category 3 (fuel- and energy-related activities, incl. upstream electricity and T\u0026amp;D losses)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePhysical CF of electricity (kgCO2e/kWh), using average-data, supplier-specific, or hybrid factors\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAlign with the chosen Scope 2 basis. Explicitly state whether transmission and distribution losses are included. Monetary factors are not appropriate once kWh/MWh are known.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCorporate Sustainability Disclosure, Hot spotting\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEEIO databases (e.g., EXIOBASE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAppropriate when expenditure data dominate and many categories or suppliers must be covered. Prefer multiregional models when supply chains are international.(Davis et al., 2025) Use to identify hotspots, then refine material categories with hybrid or supplier-specific/physical data.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eComparability with product LCAs or EPDs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eP-LCA databases (e.g. Ecoinvent)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEnsures methodological consistency with the dominant approach in product-level studies. Disclose cut-off rules, multifunctionality assumptions, capital goods treatment, geography, and data vintage.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRapidly evolving grids / time-sensitive applications\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFreshest year-matched physical factors; where relevant, subannual/hourly or near-real-time datasets\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIf the data lag exceeds\u0026thinsp;~\u0026thinsp;2\u0026ndash;3 years, or the grid mix is changing quickly, run a sensitivity analysis or triangulate with a more recent source. Temporal and spatial representativeness should weigh as much as source authority.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Conclusions\u003c/h2\u003e \u003cp\u003eThe carbon footprint of electricity is not a single, fixed value, but a context-dependent figure shaped by deliberate methodological choices. This study has systematically dissected the impact of those choices, mapping the landscape of divergence from technology-level data to national averages and monetary factors. By transforming these quantitative findings into a practical set of guiding principles, we have provided the tools for analysts, companies, and policymakers to move from arbitrary data selection to conscious, evidence-based choices. These conclusions, however, must be viewed considering the study's methodological boundaries. Our analysis was intentionally focused on a production-based perspective and limited to attributional models to ensure a fair comparison of the underlying data sources. This highlights two critical avenues for future research: first, integrating flow-tracing algorithms to extend this comparative analysis to consumption-based footprints, and second, performing an equally systematic comparison of consequential (marginal) EFs, which is essential for guiding policy on demand-side management and grid decarbonization. It is important to note that the GHG Protocol is already moving toward an update of the Scope 2 Guidance, with increasing emphasis on the temporal and geographical granularity underpinning low-carbon electricity claims. While this represents a relevant advancement, particularly for market-based accounting, it only partially addresses the challenges highlighted in this study, which primarily concern the consistency and comparability of emission factors across data sources and methodologies.. In parallel, the joint effort between GHG Protocol and ISO to establish a harmonized standard represents a significant step forward in enhancing comparability. Nonetheless, these initiatives on their own may prove insufficient. Without more reliable primary data, clear and definitive guidance on temporal and spatial granularity (including an explicit and unique indication on the treatment of imports and exports), robust residual-mix methodologies where applicable, a more precise delineation between location- and market-based accounting, and systematic documentation of sources and versions, significant discrepancies are likely to persist. In essence, while stronger rules are essential, the ultimate quality of Scope 2 reporting will continue to depend on rigorous data and uncompromising transparency.\u003c/p\u003e \u003cp\u003eUltimately, in a rapidly decarbonizing context - and amid polarized views on sustainability regulation \u003cb\u003e-\u003c/b\u003e credibility in climate policy and corporate reporting depends on transparent data sourcing and clear communication of analytical boundaries. A practical way to support this credibility is to anchor inventories in a centralized and regularly updated reference dataset, applied consistently over time and complemented by integrated primary data to progressively enable more robust hybrid assessments.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003e\u003cstrong\u003eConflict of Interest Statement:\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe authors declare no conflict of interest.\u003c/p\u003e\n\u003ch2\u003eFUNDING INFORMATION\u003c/h2\u003e\n\u003cp\u003eThe authors received no external funding.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eCC: Conceptualization, Data curation, Methodology, Software, Visualization, Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing NG: Conceptualization, Data curation, Methodology, Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing LR: Conceptualization, Data curation, Methodology, Software, Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing. MVR: Conceptualization, Methodology, Writing \u0026ndash; review \u0026amp; editing.\u003c/p\u003e\n\u003ch2\u003eAcknowledgement\u003c/h2\u003e\n\u003cp\u003eThe authors acknowledge support from the Horizon Europe Innovation Action program project ENTICE (Grant No. 101184775). The content of this paper does not necessarily reflect the opinions of the European Commission and the responsibility for it lies solely with its authors.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eThe data that supports the findings of this study are available in the online resources of this article.\u0026nbsp;Restrictions apply to the availability of IEA data, which were used under license for this study. Data are available at [https://www.iea.org/data-and-statistics/data-product/emissions-factors-2024](https:/www.iea.org/data-and-statistics/data-product/emissions-factors-2024) . Ecoinvent data used for the study were not directly extracted from Ecoinvent database but from the supporting materials of Steubing et al , published as open access material under CC-BY-NC 4.0 license. The utilization and dissemination of data from the GTAP database is permitted: the application in fact encompasses a single commodity and 28 geographical regions.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eSupporting Information\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupporting information is linked to this article on the \u003cem\u003eJIE\u003c/em\u003e website:\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAsdrubali, F., Baldinelli, G., D\u0026rsquo;Alessandro, F., \u0026amp; Scrucca, F. (2015). Life cycle assessment of electricity production from renewable energies: Review and results harmonization. \u003cem\u003eRenewable and Sustainable Energy Reviews\u003c/em\u003e, Vol. 42, pp. 1113\u0026ndash;1122. Elsevier Ltd. https://doi.org/10.1016/j.rser.2014.10.082\u003c/li\u003e\n\u003cli\u003eBarros, M. V., Salvador, R., Piekarski, C. M., de Francisco, A. C., \u0026amp; Freire, F. M. C. S. (2020, January 1). Life cycle assessment of electricity generation: a review of the characteristics of existing literature. \u003cem\u003eInternational Journal of Life Cycle Assessment\u003c/em\u003e, Vol. 25, pp. 36\u0026ndash;54. Springer. https://doi.org/10.1007/s11367-019-01652-4\u003c/li\u003e\n\u003cli\u003eBastos, J., Monforti-Ferrario, F., \u0026amp; Melica, G. (2024). \u003cem\u003eGHG Emission Factors for Electricity Consumption\u003c/em\u003e. European Commission, Joint Research Centre (JRC). Retrieved from http://data.europa.eu/89h/919df040-0252-4e4e-ad82-c054896e1641\u003c/li\u003e\n\u003cli\u003eBeltrami, F., Grossi, L., \u0026amp; Liebensteiner, M. (2025). Beyond borders: Estimating the marginal emission factor of electricity trade. \u003cem\u003eEnergy Economics\u003c/em\u003e, \u003cem\u003e148\u003c/em\u003e. https://doi.org/10.1016/j.eneco.2025.108588\u003c/li\u003e\n\u003cli\u003eBertolini, M., Duttilo, P., \u0026amp; Lisi, F. (2025). Accounting carbon emissions from electricity generation: A review and comparison of emission factor-based methods. \u003cem\u003eApplied Energy\u003c/em\u003e, \u003cem\u003e392\u003c/em\u003e. https://doi.org/10.1016/j.apenergy.2025.125992\u003c/li\u003e\n\u003cli\u003eBlizniukova, D., Holzapfel, P., Unnewehr, J. F., Bach, V., \u0026amp; Finkbeiner, M. (2023). Increasing temporal resolution in greenhouse gas accounting of electricity consumption divided into Scopes 2 and 3: case study of Germany. \u003cem\u003eInternational Journal of Life Cycle Assessment\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(12), 1622\u0026ndash;1639. https://doi.org/10.1007/s11367-023-02240-3\u003c/li\u003e\n\u003cli\u003eCA Senate. \u003cem\u003eClimate Corporate Data Accountability Act\u003c/em\u003e. , (2023). California Senate.\u003c/li\u003e\n\u003cli\u003eClau\u0026szlig;, J., Stinner, S., Solli, C., Lindberg, K. B., Madsen, H., \u0026amp; Georges, L. (2019). Evaluation method for the hourly average CO2eq. Intensity of the electricity mix and its application to the demand response of residential heating. \u003cem\u003eEnergies\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(7). https://doi.org/10.3390/en12071345\u003c/li\u003e\n\u003cli\u003e\u003cem\u003eCSRD\u003c/em\u003e. , (2022). European Parliament and Council.\u003c/li\u003e\n\u003cli\u003eEC JRC. (2024). FIGARO-E3 dataset. Retrieved September 9, 2025, from https://data.jrc.ec.europa.eu/collection/id-00403\u003c/li\u003e\n\u003cli\u003eecoinvent. (2025). Retrieved August 27, 2025, from https://ecoinvent.org/\u003c/li\u003e\n\u003cli\u003eEggleston, H. S., Buendia, L., Miwa, K., Ngara, T., \u0026amp; Tanabe, K. (2006). \u003cem\u003e2006 IPCC guidelines for national greenhouse gas inventories\u003c/em\u003e (Vol. 2). Institute for Global Environmental Strategies.\u003c/li\u003e\n\u003cli\u003eeGrid. (2023). \u003cem\u003eSummary Tables 2021\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eElectricity Maps. (2024). \u003cem\u003eElectricity Maps emission factors methodology\u003c/em\u003e. Retrieved from https://github.com/electricitymaps/electricitymaps-contrib/wiki/Emission-factors\u003c/li\u003e\n\u003cli\u003eElectricity Maps. (2025). Regional emission factors. Retrieved February 2, 2025, from https://ww2.electricitymaps.com/methodology#carbon-intensity-and-emission-factors\u003c/li\u003e\n\u003cli\u003eEMBER. (2025a). \u003cem\u003eData methodology\u003c/em\u003e. Retrieved from https://storage.googleapis.com/emb-prod-bkt-publicdata/public-downloads/ember_electricity_data_methodology.pdf\u003c/li\u003e\n\u003cli\u003eEMBER. (2025b). Electricity Data Explorer - Carbon emissions intensity. Retrieved September 5, 2025, from https://ember-energy.org/data/electricity-data-explorer/?entity=World\u0026amp;chart=single_year\u0026amp;data=co2_intensity\u0026amp;fuel=total\u003c/li\u003e\n\u003cli\u003eEMBER. (2025c, June 25). The electrification imperative. Retrieved August 27, 2025, from https://ember-energy.org/latest-insights/the-electrification-imperative/\u003c/li\u003e\n\u003cli\u003eENTSO-E. (2025). ENTSO-E Trasparency Platform - Actual Generation per Production Type. Retrieved July 29, 2025, from https://transparency.entsoe.eu/generation/r2/actualGenerationPerProductionType\u003c/li\u003e\n\u003cli\u003eEuropean Commission. (2025a). Carbon Border Adjustment Mechanism. Retrieved August 27, 2025, from https://taxation-customs.ec.europa.eu/carbon-border-adjustment-mechanism_en\u003c/li\u003e\n\u003cli\u003eEuropean Commission. \u003cem\u003eESRS Voluntary Sustainability Reporting Standards for SMEs under the European Sustainability Reporting Standards framework\u003c/em\u003e. , (2025).\u003c/li\u003e\n\u003cli\u003eGibon, T., \u0026amp; Hahn Menacho, \u0026Aacute;. (2023). Parametric Life Cycle Assessment of Nuclear Power for Simplified Models. \u003cem\u003eEnvironmental Science and Technology\u003c/em\u003e, \u003cem\u003e57\u003c/em\u003e(38), 14194\u0026ndash;14205. https://doi.org/10.1021/acs.est.3c03190\u003c/li\u003e\n\u003cli\u003eHagenaars, R. H., Heijungs, R., Tukker, A., \u0026amp; Wang, R. (2025, April 1). Hybrid LCA for sustainable transitions: principles, applications, and prospects. \u003cem\u003eRenewable and Sustainable Energy Reviews\u003c/em\u003e, Vol. 212. Elsevier Ltd. https://doi.org/10.1016/j.rser.2025.115443\u003c/li\u003e\n\u003cli\u003eHawkes, A. D. (2010). Estimating marginal CO2 emissions rates for national electricity systems. \u003cem\u003eEnergy Policy\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(10), 5977\u0026ndash;5987. https://doi.org/10.1016/j.enpol.2010.05.053\u003c/li\u003e\n\u003cli\u003eHenriques, C. O., \u0026amp; Sousa, S. (2023, March 1). A Review on Economic Input-Output Analysis in the Environmental Assessment of Electricity Generation. \u003cem\u003eEnergies\u003c/em\u003e, Vol. 16. MDPI. https://doi.org/10.3390/en16062930\u003c/li\u003e\n\u003cli\u003eHertwich, E. G., Gibon, T., Bouman, E. A., Arvesen, A., Suh, S., Heath, G. A., \u0026hellip; Shi, L. (2015). Integrated life-cycle assessment of electricity-supply scenarios confirms global environmental benefit of low-carbon technologies. \u003cem\u003eProceedings of the National Academy of Sciences of the United States of America\u003c/em\u003e, \u003cem\u003e112\u003c/em\u003e(20), 6277\u0026ndash;6282. https://doi.org/10.1073/pnas.1312753111\u003c/li\u003e\n\u003cli\u003eHowarth, R. W. (2024). The greenhouse gas footprint of liquefied natural gas (LNG) exported from the United States. \u003cem\u003eEnergy Science and Engineering\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(11), 4843\u0026ndash;4859. https://doi.org/10.1002/ese3.1934\u003c/li\u003e\n\u003cli\u003eIEA. (2024a). \u003cem\u003eEmission factors 2024 - database documentation\u003c/em\u003e. Retrieved from https://iea.blob.core.windows.net/assets/adcb9ea4-fc85-4379-826f-bbdd57401fa5/IEA_Methodology_Emission_Factors_2024.pdf\u003c/li\u003e\n\u003cli\u003eIEA. (2024b). Emissions Factors 2024. Retrieved August 27, 2025, from https://www.iea.org/data-and-statistics/data-product/emissions-factors-2024\u003c/li\u003e\n\u003cli\u003eIPCC. (2014). \u003cem\u003eMitigation of Climate Change - Contribution of Working Group III to the Fifth Assessment Report of the IPCC\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eJRC. (2024). GHG Emission Factors for Electricity Consumption. Retrieved September 25, 2025, from https://data.jrc.ec.europa.eu/dataset/919df040-0252-4e4e-ad82-c054896e1641\u003c/li\u003e\n\u003cli\u003eJRC. (2025a). Environmental footprints based on FIGARO - Greenhouse gas emissions (in CO2 equivalent). Retrieved September 19, 2025, from https://ec.europa.eu/eurostat/web/environment/database#Greenhouse%20gas%20emissions%20(in%20CO2%20equivalent)\u003c/li\u003e\n\u003cli\u003eJRC. (2025b). FIGARO tables, Annual EU inter-country supply, use and input-output tables. Retrieved September 19, 2025, from https://circabc.europa.eu/ui/group/cec66924-a924-4f91-a0ef-600a0531e3ba/library/d74cb7c1-bf93-4414-868c-89ab25e3e676\u003c/li\u003e\n\u003cli\u003eKhan, I. (2019, March 1). Greenhouse gas emission accounting approaches in electricity generation systems: A review. \u003cem\u003eAtmospheric Environment\u003c/em\u003e, Vol. 200, pp. 131\u0026ndash;141. Elsevier Ltd. https://doi.org/10.1016/j.atmosenv.2018.12.005\u003c/li\u003e\n\u003cli\u003eLamb, W. F., Wiedmann, T., Pongratz, J., Andrew, R., Crippa, M., Olivier, J. G. J., \u0026hellip; Minx, J. (2021). A review of trends and drivers of greenhouse gas emissions by sector from 1990 to 2018. \u003cem\u003eEnvironmental Research Letters\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(7), 073005. https://doi.org/10.1088/1748-9326/abee4e\u003c/li\u003e\n\u003cli\u003eMajeau-Bettez, G., Str\u0026oslash;mman, A. H., \u0026amp; Hertwich, E. G. (2011). Evaluation of Process- and Input\u0026ndash;Output-based Life Cycle Inventory Data with Regard to Truncation and Aggregation Issues. \u003cem\u003eEnvironmental Science \u0026amp; Technology\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(23), 10170\u0026ndash;10177. https://doi.org/10.1021/es201308x\u003c/li\u003e\n\u003cli\u003eMerciai et al. (2025). EXIOBASE hybrid - Consequential System. Retrieved September 24, 2025, from https://zenodo.org/records/15421526\u003c/li\u003e\n\u003cli\u003eMessagie, M., Mertens, J., Oliveira, L., Rangaraju, S., Sanfelix, J., Coosemans, T., \u0026hellip; Macharis, C. (2014). The hourly life cycle carbon footprint of electricity generation in Belgium, bringing a temporal resolution in life cycle assessment. \u003cem\u003eApplied Energy\u003c/em\u003e, \u003cem\u003e134\u003c/em\u003e, 469\u0026ndash;476. https://doi.org/10.1016/j.apenergy.2014.08.071\u003c/li\u003e\n\u003cli\u003eMoran, D., \u0026amp; Wood, R. (2014). Convergence between the EORA, WIOD, EXIOBASE, and OPENEU\u0026rsquo;s consumption-based carbon accounts. \u003cem\u003eEconomic Systems Research\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e(3), 245\u0026ndash;261. https://doi.org/10.1080/09535314.2014.935298\u003c/li\u003e\n\u003cli\u003eNREL. (2021). \u003cem\u003eLife Cycle Greenhouse Gas Emissions from Electricity Generation: Update Life Cycle Assessment of Energy Systems (National Renewable Energy Laboratory)\u003c/em\u003e. Retrieved from https://data.nrel.gov/submissions/171\u003c/li\u003e\n\u003cli\u003ePehl, M., Arvesen, A., Humpen\u0026ouml;der, F., Popp, A., Hertwich, E. G., \u0026amp; Luderer, G. (2017). Understanding future emissions from low-carbon power systems by integration of life-cycle assessment and integrated energy modelling. \u003cem\u003eNature Energy\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e(12), 939\u0026ndash;945. https://doi.org/10.1038/s41560-017-0032-9\u003c/li\u003e\n\u003cli\u003ePerkins, J., \u0026amp; Suh, S. (2019). Uncertainty Implications of Hybrid Approach in LCA: Precision versus Accuracy. \u003cem\u003eEnvironmental Science \u0026amp; Technology\u003c/em\u003e, \u003cem\u003e53\u003c/em\u003e(7), 3681\u0026ndash;3688. https://doi.org/10.1021/acs.est.9b00084\u003c/li\u003e\n\u003cli\u003eRaadal, H. L., Vold, B. I., Myhr, A., \u0026amp; Nygaard, T. A. (2016). \u003cem\u003eGHG emissions and energy performance of offshore wind power\u003c/em\u003e. Retrieved from https://www.sciencedirect.com/science/article/pii/S0960148113006654\u003c/li\u003e\n\u003cli\u003eRyan, N. A., Johnson, J. X., \u0026amp; Keoleian, G. A. (2016). Comparative Assessment of Models and Methods to Calculate Grid Electricity Emissions. \u003cem\u003eEnvironmental Science and Technology\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e(17), 8937\u0026ndash;8953. https://doi.org/10.1021/acs.est.5b05216\u003c/li\u003e\n\u003cli\u003eRyan, N. A., Johnson, J. X., Keoleian, G. A., \u0026amp; Lewis, G. M. (2018). Decision Support Algorithm for Evaluating Carbon Dioxide Emissions from Electricity Generation in the United States. \u003cem\u003eJournal of Industrial Ecology\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(6), 1318\u0026ndash;1330. https://doi.org/10.1111/jiec.12708\u003c/li\u003e\n\u003cli\u003eSarhan, A., Ramachandaramurthy, V. K., Sin, T. C., Walker, S. L., Salman, B., \u0026amp; Padmanaban, S. (2023). Assessment of Spatial and Temporal Modeling on Greenhouse Gas Emissions From Electricity Generation. \u003cem\u003eIEEE Access\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e, 97478\u0026ndash;97492. https://doi.org/10.1109/ACCESS.2023.3258923\u003c/li\u003e\n\u003cli\u003eScarlat, N., Prussi, M., \u0026amp; Padella, M. (2022). Quantification of the carbon intensity of electricity produced and used in Europe. \u003cem\u003eApplied Energy\u003c/em\u003e, \u003cem\u003e305\u003c/em\u003e. https://doi.org/10.1016/j.apenergy.2021.117901\u003c/li\u003e\n\u003cli\u003eSch\u0026auml;fer, M., Tranberg, B., Jones, D., \u0026amp; Weidlich, A. (2020). Tracing carbon dioxide emissions in the European electricity markets. \u003cem\u003e17th International Conference on the European Energy Market (EEM)\u003c/em\u003e. IEEE.\u003c/li\u003e\n\u003cli\u003eSchl\u0026ouml;mer, S., Bruckner, T., Fulton, L., Hertwich Austria, E., McKinnon, A. U., Perczyk, D., \u0026hellip; Minx, J. (2014). \u003cem\u003eANNEX III Technology-specific Cost and Performance Parameters (Fifth Assessment Report of the Intergovernmental Panel on Climate Change)\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eStadler et al. (2021a). EXIOBASE v3.3.18 Hybrid Multi Regional Supply and Use Tables. Retrieved September 9, 2025, from https://zenodo.org/records/10148587\u003c/li\u003e\n\u003cli\u003eStadler et al. (2021b). EXIOBASE v3.8.2 Monetary Multi Regional Supply and Use Tables. Retrieved September 9, 2025, from https://zenodo.org/records/5589597\u003c/li\u003e\n\u003cli\u003eStadler et al. (2025). EXIOBASE v3.9.5. Retrieved August 27, 2025, from https://zenodo.org/records/14869924\u003c/li\u003e\n\u003cli\u003eStadler, K., Wood, R., Bulavskaya, T., S\u0026ouml;dersten, C. J., Simas, M., Schmidt, S., \u0026hellip; Tukker, A. (2018). EXIOBASE 3: Developing a Time Series of Detailed Environmentally Extended Multi-Regional Input-Output Tables. \u003cem\u003eJournal of Industrial Ecology\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(3), 502\u0026ndash;515. https://doi.org/10.1111/jiec.12715\u003c/li\u003e\n\u003cli\u003eSteubing, B., de Koning, A., Merciai, S., \u0026amp; Tukker, A. (2022). How do carbon footprints from LCA and EEIOA databases compare? A comparison of ecoinvent and EXIOBASE. \u003cem\u003eJournal of Industrial Ecology\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e(4), 1406\u0026ndash;1422. https://doi.org/10.1111/jiec.13271\u003c/li\u003e\n\u003cli\u003eSuh, S., Lenzen, M., Treloar, G. J., Hondo, H., Horvath, A., Huppes, G., \u0026hellip; Norris, G. (2004). System Boundary Selection in Life-Cycle Inventories Using Hybrid Approaches. \u003cem\u003eEnvironmental Science \u0026amp; Technology\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(3), 657\u0026ndash;664. https://doi.org/10.1021/es0263745\u003c/li\u003e\n\u003cli\u003eTahavori, M. A., Golinucci, N., Rinaldi, L., Rocco, M. V., \u0026amp; Colombo, E. (2023). MARIO: A Versatile and User-Friendly Software for Building Input-Output Models. \u003cem\u003eJournal of Open Research Software\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e. https://doi.org/10.5334/JORS.473\u003c/li\u003e\n\u003cli\u003eTranberg, B., Corradi, O., Lajoie, B., Gibon, T., Staffell, I., \u0026amp; Andresen, G. B. (2019). Real-time carbon accounting method for the European electricity markets. \u003cem\u003eEnergy Strategy Reviews\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e. https://doi.org/10.1016/j.esr.2019.100367\u003c/li\u003e\n\u003cli\u003eTurconi, R., Boldrin, A., \u0026amp; Astrup, T. (2013). Life cycle assessment (LCA) of electricity generation technologies: Overview, comparability and limitations. \u003cem\u003eRenewable and Sustainable Energy Reviews\u003c/em\u003e, Vol. 28, pp. 555\u0026ndash;565. Elsevier Ltd. https://doi.org/10.1016/j.rser.2013.08.013\u003c/li\u003e\n\u003cli\u003eUNECE. (2021). \u003cem\u003eCarbon Neutrality in the UNECE Region: Integrated Life-cycle Assessment of Electricity Sources\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eUnited Nations. (2018). \u003cem\u003eHandbook on Supply and Use Tables and Input Output-Tables with Extensions and Applications\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eUnnewehr, J. F., Weidlich, A., Gf\u0026uuml;llner, L., \u0026amp; Sch\u0026auml;fer, M. (2022). Open-data based carbon emission intensity signals for electricity generation in European countries \u0026ndash; top down vs. bottom up approach. \u003cem\u003eCleaner Energy Systems\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e. https://doi.org/10.1016/j.cles.2022.100018\u003c/li\u003e\n\u003cli\u003eWatershed. (2025). Open CEDA by Watershed. Retrieved August 27, 2025, from https://registry.opendata.aws/open-ceda\u003c/li\u003e\n\u003cli\u003eWorld Resource Institute. (2011). \u003cem\u003eCorporate Value Chain (Scope 3) Standard\u003c/em\u003e. Retrieved from https://ghgprotocol.org/corporate-value-chain-scope-3-standard\u003c/li\u003e\n\u003cli\u003eWorld Resource Institute. (2023). \u003cem\u003eGHG Protocol Scope 2 Guidance\u003c/em\u003e. Retrieved from https://ghgprotocol.org/scope-2-guidance\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-industrial-ecology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"44498","submissionUrl":"https://submission.springernature.com/new-submission/44498/3","title":"Journal of Industrial Ecology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"carbon footprint, electricity, emission factors, systematic review, corporate sustainability accounting, industrial ecology","lastPublishedDoi":"10.21203/rs.3.rs-9337261/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9337261/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe carbon footprint of electricity is a critical metric for climate policy and corporate sustainability reporting, yet its quantification is hampered by a lack of methodological consensus and a fragmented evidence base. This fragmentation reflects a central debate in industrial ecology regarding the consistency and comparability of life-cycle\u0026ndash;based emission factors across different modelling paradigms. This study addresses this gap by performing a systematic, quantitative comparison of electricity emission factors from a wide range of authoritative sources, including process-LCA databases (Ecoinvent), environmentally extended input-output tables (EXIOBASE, EMERGING, EORA, GLORIA, GTAP), primary agency data (IEA) and others. We harmonize technology taxonomies and GHG characterization factors to compare technology-level and national-level emission factors for the EU and US.\u003c/p\u003e \u003cp\u003eOur results reveal, for technology-specific emission factors, that while there is a moderate consensus on direct, operational emissions (carbon intensity), we find wide and systematic divergences for full life-cycle emissions (carbon footprints). These discrepancies propagate to national-level estimates, where a country's calculated footprint can significantly vary, solely based on the data source chosen. This paper provides a useful reference to deconstruct and understand such divergences. We translate these findings into a set of practical selection criteria to support analysts and decision-makers making transparent and context-appropriate choices. We conclude that a more transparent and critical approach to sourcing emission data is fundamental to the credibility of carbon accounting and the broader energy transition.\u003c/p\u003e","manuscriptTitle":"Carbon footprint of electricity: a systematic methodological and quantitative review","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-24 10:06:47","doi":"10.21203/rs.3.rs-9337261/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-06T09:36:33+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-05T20:53:14+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-25T05:39:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"147476527793640558438337026542963773496","date":"2026-04-21T07:19:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"150529854887431632033900016534837929964","date":"2026-04-14T10:43:57+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-12T09:58:44+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-08T04:40:21+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-08T04:39:49+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Industrial Ecology","date":"2026-04-06T20:54:28+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-industrial-ecology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"44498","submissionUrl":"https://submission.springernature.com/new-submission/44498/3","title":"Journal of Industrial Ecology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"5f188656-0eb1-43b2-ac4f-69c83742328b","owner":[],"postedDate":"April 24th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-06T09:36:33+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-05T20:53:14+00:00","index":12,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-06T09:41:21+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-24 10:06:47","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9337261","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9337261","identity":"rs-9337261","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}