Carbon and Greenhouse Gas Emissions from Electricity Consumption in European Union Buildings

Abstract

The buildings sector is the single most important end-user of final energy in the European Union and a significant contributor to carbon and greenhouse gas emissions. This work focuses on a review of available data that are used to calculate the annual emissions from electricity generation in the European Union and quantify indirect emissions from the use of electricity in the buildings sector. Historical data since 1990 are used to derive simple empirical correlations for the time evolution of emissions factors related to electricity production in each Member State of the European Union. More recent trajectories using data from the last ten years are also presented. The derived correlations can be easily integrated in building stock modeling and national studies to facilitate forward-looking projections of emissions from electricity use in buildings. The EU-27 averages 0.2883 kgCO₂-eq/kWh_el, ranging from 0.0456 kgCO₂-eq/kWh_el in Sweden up to 1.0595 kgCO₂-eq/kWh_el in Poland. As a case study, the derived coefficients are then used to quantify the indirect emissions from the electricity consumption attributed to the building stock in each EU Member State. The calculated total EU-27 GHG indirect emissions attributed to electricity consumption amounted to 215 MtCO₂-eq for residential buildings and 201 MtCO₂-eq for non-residential buildings. In addition, the proposed correlations are used to demonstrate how they can be used for more realistic future projections of emissions towards the national targets in Greece and Poland.

1. Introduction

The European Union (EU) is making significant efforts to reduce greenhouse gas (GHG) emissions and has committed to be a decarbonized continent by 2050 in support to the United Nations efforts to address climate change [1]. Under the Paris Agreement, the EU contribution was initially set to a reduction of GHG emissions by 40% by 2030 compared with 1990 levels and with the recent Climate Law, the target was set to at least −55% by 2030 [2]. This ambitious target mandates significant efforts across all sectors and coordinated progress in higher energy efficiency and exploitation of renewable energy sources.

Greenhouse gases include a group of seven gases that contribute to global warming and climate change according to the environmental agreement known as the Kyoto Protocol. They consist of non-fluorinated gases (i.e., carbon dioxide (CO₂), methane (CH₄) and nitrous oxide (N₂O) and fluorinated gases (i.e., hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SF₆) and nitrogen trifluoride (NF₃)). They are all expressed in carbon dioxide equivalents (CO₂-eq) in order to compare their individual contributions and quantify their total contributions to global warming.

The GHG emissions associated with the buildings sector include direct emissions from onsite combustion for heating and indirect emissions from power plants to generate electricity using solid, liquid, and gas fossil fuels, as well as gas flaring. Carbon dioxide is generally recognized as the most significant contributor to GHGs associated with human activities as a result of its high global warming potential and atmospheric lifetime.

Overall, the energy consumption in European buildings is the most energy demanding sector in the EU-27, reaching 391.1 million tonnes of oil equivalent (Mtoe) in 2021 or 44.2% of the total final energy use [3]. This total energy consumption includes all different energy carriers used for the operation of buildings, like fossil fuels (e.g., onsite combustion of natural gas, oil or solid fuels for space heating) and/or electricity for all building operations. Direct emissions from EU-27 buildings typically account for about a tenth of the total emissions and are reported separately in [3]. However, the operation of buildings has additional environmental impacts as a result of the indirect emissions that are associated with the generation of electricity, the majority of which is finally used in buildings. In this case, the emissions are collectively reported in the official annual datasheets [3] for the energy sector as a whole, without differentiating for the end-use sectors. To quantify the indirect emissions from buildings, one can use the electricity consumption of the buildings sector and convert it by multiplying with an up-to-date electricity emissions factor. This conversion factor is variable from year to year and depends on the mix of the different energy carriers used for electricity generation.

The historic trends provide some practical insights. Compared to 1990, there has been a notable shift away from the use of fossil fuels in final energy use, which resulted in significantly lower emissions from buildings. Specifically, the GHG emissions from fuel combustion in residential buildings represented 12% of the total in 2021 [4]. Since 1990, the trend of the total GHG emissions was −28% as a result of using less solid fossil fuels (–78%) and liquid fuels (–57%), part of which was counterbalanced by an increase in emissions (57%) using more gaseous fuels. Non-residential buildings accounted for 5% of the total GHG emissions from fuel combustion in 2021, which is mainly dominated by CO₂ emissions from fossil fuels [4]. Emissions from non-residential buildings decreased by 24% during 1990–2021 mainly as a result of using less fossil fuels while using more natural gas for heating [4].

The largest key category for CO₂ and GHG emissions in EU-27 is from public electricity and heat production that annually contributes 717 million tonnes of carbon dioxide equivalent (MtCO₂-eq), which represents 20% of the total GHG emissions in 2021, while the specific CO₂ emissions are estimated at 709 million tons of carbon dioxide (MtCO₂) or 25% of the total [3]. Between 1990 and 2021, emissions from this sector decreased by −47% during the past 30 years as a result of the favorable restricting of the energy mix, for example, shifting away from solid fuels and oil to less-carbon-intensive natural gas and more efficient power generation equipment. For example, during the period of 1990–2021, the use of fossil fuels in thermal power plants decreased by 53% and 85% for solid and liquid fuels, respectively, while the use of natural gas increased by 76% [4]. At the same time, there has been a significant growth in renewables to generate electricity that increased by almost four times, achieving an increase on an annual basis of 782 terawatt-hours (TWh_el) in 2021 compared to 1990 [3]. This is among the top EU policies, aiming to accelerate the use of renewables in the overall energy mix to reach at least 45% by 2030 [5], which may be further increased in order to deliver the target of −55% on the emissions by 2030 towards a decarbonized economy by 2050.

Figure 1 reveals the trends for the total GHG emissions per capita in the EU-27 Member States. Specifically, the EU-27 average is currently at 7.9 tCO₂-eq/capita, achieving −33% since 1990 [3]. Luxembourg has the highest value of the indicator at 12.7 tCO₂-eq/capita, followed by Ireland at 12.7 tCO₂-eq/capita and Czechia at 11.2 tCO₂-eq/capita. The most notable change was observed in Estonia with −63% since 1990, compared with only −7% in Cyprus and Portugal.

In absolute numbers, the total GHG emissions during the past three decades decreased by −28%, from 4921.1 MtCO₂-eq in 1990 down to 3541.4 MtCO₂-eq in 2021 [3]. The data do not account for the natural removal of emissions from the atmosphere by land use, land use change and forestry (LULUCF) activities. The top five EU-27 Member States that have the highest contribution to the total GHG emissions balance include Germany (778.7 MtCO₂-eq or 22% of the total emissions), France (498.8 MtCO₂-eq or 14%), Italy (422.6 MtCO₂-eq or 12%), Poland (402.4 MtCO₂-eq or 11%) and Spain (297.2 MtCO₂-eq or 8%).

GHG emissions in 2021 rebounded by 5.1% compared to 2020 (Figure 1), following a −10.2% in 2020 compared to 2019 as a result of the lower EU-27 final energy use due to the COVID-19 pandemic and the slowdown of the European economy [3]. The energy and emissions drop in 2020 can be used as an example to simulate expected future savings and scenarios of implementing major restrictions on energy use. For example, energy consumption was −12.8% in the transport sector, −5.6% in non-residential buildings and −3.4% in industry compared to the 2019 levels [3]. Interesting enough, energy use remained the same in the residential buildings sector despite the fact that people spend more time living and working from home.

Given the EU policies and commitments to decarbonize the power grid and the building sector, among others, the overall downward trend of reducing emissions is notable. However, it is also evident that EU-27 Member States have made different progress. Among others, this is attributed to different power generation technologies, conversion efficiencies and the energy carriers used in each country, the electricity transmission and distribution (T&D) losses, and even national electricity imports and exports. On a positive note, similar progress and dropping trends are observed in other parts of the world, for example, in China, where emissions decreased by 28.5% during the period from 2005 to 2020 [6], and in the United States, with a notable drop in emissions in 2021 by 35.5% below 2005 levels [7]. However, on a global scale, there is a notable growth in emissions from energy systems that is closely linked to the growing trend of gross domestic product (GDP) per capita [8]. In particular, some countries and regions stand out, like the South-east Asian developed countries and Australia, Japan and New Zealand, which average an annual increase in carbon intensity by about 1.8%.

To gain a better insight on these trends, researchers have investigated the relationships between GHG emissions and the energy consumption of fossil fuels with population and GDP among other economic indicators [9]. The multiple regression analysis derived national correlations that can be used to forecast GHG emissions from various independent variables like GDP and population, resulting in an overall good statistical confidence.

The conversion of energy used in a specific sector or a given activity is performed by multiplying the energy consumption with the appropriate emissions factors for specific energy carriers like fossil fuels and electricity. These conversion coefficients quantify the emissions per unit of energy consumed (e.g., kgCO₂-eq per unit energy natural gas or heating oil) or per unit of energy generated (e.g., kgCO₂-eq per unit of generated electricity).

Several studies have been performed to assess the GHG emissions related to the energy use in cities and compile a GHG balance of their territory [10]. The calculations of the emissions factors usually follow a standard assessment or a life-cycle assessment (LCA), although there are several other GHG accounting methods that are also available with different levels of complexity [11]. The emissions factors using the standard method are in line with the calculation principles introduced by the Intergovernmental Panel on Climate Change (IPCC) and are directly related to operational energy use. In this case, the emissions are related to on-site fuel combustion, like in the case of burning fossils fuels for the space heating of buildings and indirectly from the fuel combustion and related emissions that are associated with electricity generation. The LCA approach considers the overall life cycle of a specific energy carrier. This is a more comprehensive and complex calculation approach that goes beyond the direct emissions from combustion and includes upstream emissions from the extraction, transport and processing of the various fuels.

Default emissions factors are periodically published by the Joint Research Centre (JRC) of the European Commission for the use of various fuels and from electricity consumption [12]. In China, emissions factors for electricity generation, supply, and use have been reported at 0.599, 0.599, and 0.622 kgCO₂-eq/kWh_el, respectively [6].

Detailed LCA calculations of GHG emissions generated from electricity production and use, including upstream, operational and use related emissions, have also revealed significant variations between the EU-27 Member States. Emissions associated with gross electricity generation averaged 0.296 kgCO₂-eq/kWh_el, of which 0.251 kgCO₂-eq/kWh_el is related to fuel combustion, 0.036 kgCO₂-eq/kWh_el is related to upstream fuel supply and 0.009 kgCO₂-eq/kWh_el is related to the construction and decommissioning of electricity plants [13]. However, the necessary input data and detailed information for the various fuels used, the conversion efficiencies for the power generation plants and even the associated emissions for each energy carrier mandate the use of different data resources and increases the complexity for simple routine calculations.

On the other hand, for buildings that involve demand-side management, it may not even be sufficient to use temporally and spatially averaged data [14]. In this case, it is better to use marginal emissions factors to control load shifting and thus minimize the GHG emissions of the entire system. Various calculation methodologies are available and reviewed in [15]. However, access to detailed electricity generation data is limited and not readily available to third parties for routine calculations. Furthermore, the analysis is usually time-demanding, depending on complex market dynamics and bidding behavior of power generation. An alternative approach is to use time-dependent average grid emissions factors that are less sensitive to these variations [15]. In the United States, average and marginal GHG emissions factors are calculated and published on an annual basis for all the different electric power generated in the United States [16].

This work aims to review and document the historic progress of emissions from electricity generation at national and EU levels. The data are then used to derive emissions factors for quantifying the amount of emissions per unit of electricity, present their historical evolution and provide simple linear correlations for future projections. These conversion factors can be used to facilitate more realistic forward-looking calculations and assess national energy efficiency, climate change or decarbonization plans in EU-27 Member States. The analysis is based on information from a comprehensive data resource with official annual statistics so that the work can be regularly and easily updated. The case studies that are also elaborated at the end provide practical insight on how to utilize the results and the proposed correlations for future projections of the emissions in two countries.

2. Methods

The emissions factors are calculated using available data that are easily accessible and consistently reported on an annual basis in official EU publications [3]. Using a single data resource is critical for enabling stakeholders to easily carry out and replicate the work in the future. Although this an important characteristic of the method and overall approach, the database [3] prescribes the available input and thus introduces some inherent limitations that should be taken into consideration as later discussed.

The results reported in this work are based on two simple definitions, using the following equations to calculate the ratio of the amount of emissions divided by the amount of electricity that is available for final end-uses. This way, the calculation approach accounts for the annual electricity transmission and distribution (T&D) losses. In the EU-27, the annual T&D losses reached 279.7 TWh_el [17].

The first equation refers to carbon emissions factors (CEFs), which are calculated taking the annual total CO₂ emissions (kgCO₂) without LULUCF that are allocated to public electricity and heat production and divide them by the final electricity consumption (FEC_el) that is delivered for all end-use sectors (kWh_el) [3]. The final electricity consumption excludes T&D losses, self-consumption, or other statistical differences.

$$CEF=\frac{C{O}_{2}emissions}{{FEC}_{el}}$$

(1)

The second equation refers to GHG emissions factors (GEFs), which are calculated in a similar manner using this time the annual total GHG emissions (kgCO₂-eq) without LULUCF that are allocated to public electricity and heat production and divide them by the FEC_el (kWh_el) that is delivered for all end-use sectors [3].

$$GEF=\frac{GHGemissions}{{FEC}_{el}}$$

(2)

The calculation approach accounts for imported electricity that is included in the FEC_el. However, the calculations do not account for the related emissions of the imported (or exported) electricity in the interconnected power grids that ensure security of supply. In some Member States, this may represent a significant percentage of the final electrical energy use or may change from year to year. For example, France has been historically a main exporter of electricity in Europe with net exports reaching 42 TWh in 2021, of which over 33% was to the UK, 14% was to Spain and 13% was to Germany [18]. Yet, as a result of a hydro and nuclear deficit in 2022, France was transformed to a net importer of electricity in 2022 with 17 TWh from Germany, Spain and the UK [18]. As a result, there was also a notable increase in coal generation in Spain and Germany, resulting in an increase in related emissions as a result of the increased exports of electricity to France. Detailed and dynamic models of the interconnected power grids could potentially monitor electricity flows and trace them to the source for estimating the related emissions and allocating them accordingly. Nonetheless, this is beyond the scope of this work, which focuses on using a single resource with official statistical data [3] that are published on an annual basis and are readily available in order to easily calculate and regularly update the emissions factors.

The proposed equations are straightforward, simple to implement and use data that are annually published in the national energy balances [3]. Accordingly, the calculation approach can be easily replicated in order to consistently update the factors on an annual basis without elaborate calculations. In order to assess the overall trends among the Member States, the annual emissions factors are used to derive a simple linear correlation using all the historic data that reflects a conservative forward looking path. A similar correlation is also developed for the data from the last decade. The 10-year correlation will more realistically reflect recent trends on the use of renewables, energy conservation policies and power generation characteristics for projecting the evolution of the emissions factors. In addition, it will also be possible to adjust and self-adapt projections by repeating calculations on an annual basis that take into account the most recent actual data from the last available year.

The proposed approach exploits the readily available data for public electricity and heat production [3]. Combined heat and power (CHP) plants generate about 12% of the total electricity generated in the EU-27. However, historic data for the generated electricity and heat are only available after 2009 [3], since the collection of CHP data was defined in the implementation guidelines defined in EU decision (2008/952) and the EU Energy Efficiency Directive (2012/27/EU and 2018/2002/EU). According to the available data for 2021, the highest percentage of CHP in total electricity generation is reported in Latvia, with 36% or 2.1 TWh_el, and 26% in Finland and The Netherlands, with 19 TWh_el and 32 TWh_el, respectively [3]. Accordingly, in these cases, emissions per unit electricity from specific power plants will exhibit variations in the event that the emissions for the generated electricity are also allocated to the useful heat from the CHP plants. However, considering that historic data for the amount of generated CHP electricity and heat are not readily available in the utilized data resource [3] until after 2009, along with the uncertainties involved in the calculations for the actual efficiencies of heat and of power production that are usually replaced with assumed (default) values, the overall complexities of the accounting procedures of this work do not directly allocate the emissions to the generated CHP heat and electricity. Consequently, the current calculation approach leads to higher values in some countries where CHP electricity is significant. This is flagged as a limitation so that it may be taken into account, if necessary, as it is elaborated in Section 6. In this case, the allocation of the total emissions (CEF or GEF) can be based on the energy content of the CHP electricity and heat output streams, along with the breakdown of the fuel inputs into CHP that are available in [3] for the periods after 2009. This is further elaborated in Section 6.1 to demonstrate how this calculation can be implemented.

3. Results

The annual emissions factors using Equation (1) are presented for carbon dioxide (CO₂) emissions in Figure 2 and using Equation (2) for GHG emissions expressed in CO₂-equivalents per unit of final electricity consumption in Figure 3 for all Member States and for the EU-27. The results for CO₂ and GHG emissions are similar since the CO₂ emissions for electricity generation dominate all other greenhouse gases. The available raw data [3] and the calculated emissions factors are presented without accounting for LULUCF activities.

Until recently, Estonia had the highest emissions factor with significant variations over the years but a clear dropping trend. The use of oil shale for electricity generation was dominant in Estonia and had a dramatic drop in 2019 as a result of the high cost of CO₂ emissions [19]. The best three performers with the lowest emissions factors were Sweden, Austria and France since they have a high share of low-carbon electricity sources like renewables and nuclear. For example, Sweden generates its gross electricity from 67% renewables and 31% nuclear, Austria with 80% renewables, and France with 68% nuclear and 23% renewables [3].

3.1. Emissions Factors

The emissions factors calculated using Equations (1) and (2) are illustrated in Figure 4. Like the GHG and carbon emissions, the corresponding factors reflect similar relative values since carbon dioxide is the most significant contributor to GHG emissions associated with electricity generation and onsite fossil fuel combustion in the buildings sector. However, both are provided since some national energy performance assessment methods and national regulations are based on carbon emissions, while other studies like the energy and climate plans quantify the results based on GHG emissions.

Historically, Poland has been among the countries with the highest emissions intensities and currently is the worst performer with the highest emissions factor. This is the result of using solid fossil fuels for over 70% of gross electricity generation and only 17% from renewables and 10% from natural gas in 2021 [3]. The EU-27 gross electricity is generated from 38% renewables (13% wind, 13% hydro, 6% solar, 4% solid biofuels), 25% nuclear, 20% natural gas, 15% solid fossil fuels and 2% oil and petroleum products [3]. Greece has made a notable progress to reduce the carbon intensity. However, it remains about 40% higher than the EU average, since the energy mix in Greece is 59% from fossil fuels (41% natural gas, 10% lignite, 8% oil) and 41% from renewables (19% wind, 11% hydro, 10% solar) [3].

3.2. Trends and Correlations

The trends of the annual data for each Member State are illustrated in Appendix A and the best fit correlations are summarized in

Table 1. Linear regressions for carbon dioxide (CO₂) emissions factors for electricity generation (kgCO₂/kWh_el) and goodness-of-fit measure (R²) for the available national data (Figure A1) using all the historical data and last 10-year data.

EU Member State	All Historical Data	Percentage Change (%)	Last 10-Year Data (10 y)	Percentage Change (%)
Austria	y = −0.0046x + 9.4614 R2 = 0.7803	−66.2%	y = −0.0061x + 12.443 R2 = 0.7318	−48.9%
Belgium	y = −0.0074x + 15.134 R2 = 0.9579	−60.7%	y = −0.0045x + 9.3403 R2 = 0.8097	−27.2%
Bulgaria	y = −0.0063x + 13.661 R2 = 0.1866	−31.3%	y = −0.056x + 113.7 R2 = 0.9034	−44.7%
Croatia	y = −0.0057x + 11.729 R2 = 0.4275	−40.8%	y = −0.0104x + 21.255 R2 = 0.8278	−37.0%
Cyprus	y = −0.0099x + 20.656 R2 = 0.8834	−29.3%	y = −0.0112x + 23.383 R2 = 0.803	−15.8%
Czechia	y = −0.0166x + 34.379 R2 = 0.8998	−42.5%	y = −0.034x + 69.349 R2 = 0.9558	−35.2%
Denmark	y = −0.0296x + 60.083 R2 = 0.8615	−77.5%	y = −0.0378x + 76.617 R2 = 0.9192	−64.4%
Estonia	y = −0.0804x + 163.36 R2 = 0.84	−83.5%	y = −0.1484x + 300.74 R2 = 0.7843	−67.8%
Finland	y = −0.0051x + 10.513 R2 = 0.4581	−49.3%	y = −0.0109x + 22.199 R2 = 0.8715	−47.4%
France	y = −0.0013x + 2.7698 R2 = 0.3913	−42.0%	y = 0.0008x − 1.4449 R2 = 0.1235	6.3%
Germany	y = −0.0079x + 16.486 R2 = 0.7566	−43.6%	y = −0.025x + 50.868 R2 = 0.8266	−29.5%
Greece	y = −0.0286x + 58.278 R2 = 0.93	−71.4%	y = −0.059x + 119.7 R2 = 0.9288	−58.1%
Hungary	y = −0.0177x + 36.022 R2 = 0.7971	−59.7%	y = −0.0194x + 39.527 R2 = 0.9247	−48.6%
Ireland	y = −0.0219x + 44.621 R2 = 0.9675	−64.2%	y = −0.0185x + 37.754 R2 = 0.7878	−28.6%
Italy	y = −0.0085x + 17.473 R2 = 0.9417	−56.2%	y = −0.0089x + 18.22 R2 = 0.8647	−28.6%
Latvia	y = −0.0213x + 43.192 R2 = 0.8602	−72.9%	y = −0.0085x + 17.412 R2 = 0.5629	−38.8%
Lithuania	y = −0.0325x + 65.699 R2 = 0.9514	−87.0%	y = −0.024x + 48.622 R2 = 0.777	−61.0%
Luxemburg	y = 0.0069x − 13.743 R2 = 0.6665	1852.1%	y = 0.0058x − 11.568 R2 = 0.1807	265.2%
Malta	y = −0.0222x + 45.52 R2 = 0.6382	−61.7%	y = −0.0354x + 72.361 R2 = 0.8792	−29.0%
Poland	y = −0.0451x + 92.074 R2 = 0.94	−55.3%	y = −0.0384x + 78.496 R2 = 0.9222	−22.8%
Portugal	y = −0.0138x + 28.153 R2 = 0.8115	−78.1%	y = −0.0129x + 26.3 R2 = 0.3045	−54.5%
Romania	y = −0.0355x + 72.145 R2 = 0.9249	−72.5%	y = −0.0442x + 89.679 R2 = 0.9598	−57.9%
Slovakia	y = −0.0123x + 24.958 R2 = 0.8832	−72.9%	y = −0.009x + 18.278 R2 = 0.8598	−34.1%
Slovenia	y = −0.0102x + 20.867 R2 = 0.8616	−53.0%	y = −0.0175x + 35.615 R2 = 0.7648	−38.8%
Spain	y = −0.0119x + 24.196 R2 = 0.8637	−74.0%	y = −0.0172x + 35.009 R2 = 0.726	−55.4%
Sweden	y = −0.001x + 2.0595 R2 = 0.5069	−31.2%	y = −0.0018x + 3.7064 R2 = 0.782	−31.5%
The Netherlands	y = −0.0057x + 11.93 R2 = 0.6487	−41.7%	y = −0.0137x + 28.09 R2 = 0.428	−29.9%
EU-27	y = −0.0104x + 21.347 R2 = 0.9664	−56.2%	y = −0.0163x + 33.241 R2 = 0.9274	−34.0%

for CO₂ emissions factors. Similar results are presented in Appendix B and

Table 2. Linear regressions for greenhouse gas (GHG) emissions factors for electricity generation (kgCO₂-eq/kWh_el) and goodness-of-fit measure (R²) for the available national data (Figure A2) using all the historical data and last 10-year data.

EU Member State	All Historical Data	Percentage Change (%)	Last 10-Year Data (10 y)	Percentage Change (%)
Austria	y = −0.0046x + 9.3806 R2 = 0.7765	−65.7%	y = −0.0061x + 12.51 R2 = 0.7335	−48.5%
Belgium	y = −0.0074x + 15.099 R2 = 0.9575	−60.5%	y = −0.0045x + 9.3559 R2 = 0.8076	−27.1%
Bulgaria	y = −0.0063x + 13.644 R2 = 0.1856	−31.1%	y = −0.056x + 113.8 R2 = 0.9029	−44.6%
Croatia	y = −0.0057x + 11.699 R2 = 0.425	−40.3%	y = −0.0103x + 21.065 R2 = 0.8226	−36.5%
Cyprus	y = −0.0099x + 20.724 R2 = 0.8835	−29.3%	y = −0.0113x + 23.454 R2 = 0.8031	−15.8%
Czechia	y = −0.0167x + 34.465 R2 = 0.8995	−42.4%	y = −0.0341x + 69.596 R2 = 0.9557	−35.2%
Denmark	y = −0.0298x + 60.371 R2 = 0.8585	−76.9%	y = −0.0381x + 77.25 R2 = 0.919	−64.0%
Estonia	y = −0.0802x + 163.14 R2 = 0.8392	−83.3%	y = −0.1483x + 300.45 R2 = 0.7841	−67.6%
Finland	y = −0.0051x + 10.441 R2 = 0.4509	−48.5%	y = −0.011x + 22.299 R2 = 0.8709	−47.0%
France	y = −0.0013x + 2.8042 R2 = 0.3955	−42.1%	y = 0.0007x − 1.3627 R2 = 0.1094	5.8%
Germany	y = −0.0078x + 16.176 R2 = 0.747	−42.9%	y = −0.0249x + 50.813 R2 = 0.8244	−28.9%
Greece	y = −0.0287x + 58.503 R2 = 0.9301	−71.4%	y = −0.0592x + 120.14 R2 = 0.9288	−58.1%
Hungary	y = −0.0177x + 36.055 R2 = 0.797	−59.5%	y = −0.0195x + 39.666 R2 = 0.9256	−48.5%
Ireland	y = −0.022x + 44.667 R2 = 0.9679	−64.1%	y = −0.0186x + 37.922 R2 = 0.7886	−28.6%
Italy	y = −0.0085x + 17.502 R2 = 0.9419	−56.2%	y = −0.0089x + 18.303 R2 = 0.8664	−28.6%
Latvia	y = −0.0212x + 42.985 R2 = 0.8564	−72.1%	y = −0.008x + 16.425 R2 = 0.5304	−37.2%
Lithuania	y = −0.0324x + 65.442 R2 = 0.9505	−86.3%	y = −0.0237x + 47.921 R2 = 0.7741	−59.4%
Luxemburg	y = 0.0069x − 13.754 R2 = 0.6662	1731%	y = 0.0058x − 11.587 R2 = 0.1809	263%
Malta	y = −0.0264x + 53.956 R2 = 0.6177	−61.8%	y = −0.0356x + 72.59 R2 = 0.8792	−29.0%
Poland	y = −0.0453x + 92.409 R2 = 0.9398	−55.3%	y = −0.0385x + 78.858 R2 = 0.9223	−27.8%
Portugal	y = −0.0138x + 28.094 R2 = 0.8094	−77.7%	y = −0.0129x + 26.297 R2 = 0.3021	−54.2%
Romania	y = −0.0356x + 72.329 R2 = 0.9247	−72.4%	y = −0.0444x + 90.018 R2 = 0.9598	−57.9%
Slovakia	y = −0.0123x + 24.965 R2 = 0.8824	−72.8%	y = −0.009x + 18.276 R2 = 0.8623	−33.8%
Slovenia	y = −0.0102x + 20.928 R2 = 0.8614	−52.9%	y = −0.0175x + 35.728 R2 = 0.7643	−38.7%
Spain	y = −0.0119x + 24.222 R2 = 0.8628	−73.8%	y = −0.0173x + 35.11 R2 = 0.7253	−55.1%
Sweden	y = −0.001x + 1.986 R2 = 0.4811	−29.4%	y = −0.0018x + 3.775 R2 = 0.7831	−30.9%
The Netherlands	y = −0.0057x + 11.871 R2 = 0.6453	−41.3%	y = −0.0137x + 28.041 R2 = 0.4264	−29.6%
EU-27	y = −0.0104x + 21.318 R2 = 0.9659	−55.9%	y = −0.0163x + 33.313 R2 = 0.9271	−33.8%

for GHG emissions factors. The correlations along with the coefficient of determination (R²) and the percentage change in the two tables and the Appendixes are provided for the CO₂ emissions over the entire period of 1990–2021 (All Historical Data) and over the last decade (Last 10-year Data (10 y)). These simple correlations provide a practical and easy to implement method for a first assessment of future emissions from indirect emissions of building operations, with engaging in time demanding and complex future energy balance models.

4. Discussion

In the context of the EU and Global Covenant of Mayors for Climate and Energy, the published value for the emissions factors for EU-27 in 2020 was 0.254 kgCO₂/kWh_el and 0.255 kgCO₂-eq/kWh_el using the Intergovernmental Panel on Climate Change approach and 0.293 kgCO₂-eq/kWh_el using the life-cycle assessment to the energy carriers used to generate electricity derived as the ratio of the total emissions from electricity generation that is generated from all input energy carriers and the total final electricity consumption [12]. These factors compare well with the calculated value derived in this work at 0.270 kgCO₂/kWh_el and 0.274 kgCO₂-eq/kWh_el for the EU-27 in 2020.

On the other hand, the values the emissions intensity reported by the European Environment Agency are calculated by dividing the emissions from public electricity production (i.e., excluding emissions resulting from heat production) by the gross electricity production and is reported at 0.238 kgCO₂-eq/kWh_el for 2021 [20]. For comparison, the factor calculated in this work was 0.288 kgCO₂-eq/kWh_el for 2021. Overall, the emissions factors presented in this work may be slightly overestimated since they do not exclude the emissions resulting from heat production in order to follow a method that is based on data from a single source, which is consistently published by Eurostat [3] and needs minimum processing or other elaboration. This way, it is possible to easily replicate and update by national or even regional authorities, which is more relevant in view of the growing use of renewables in electricity generation. However, the differences may be significant in countries where CHP electricity generation dominates the energy sector. This issue is addressed in Section 6.1.

Until 2010, the decreasing trend in emissions factors was attributed to higher transformation efficiencies from fossil fuels to electricity in an effort to also comply with emissions limit values that were introduced by the relevant industrial emissions legislation like the 2001/80/EC European Directive on Large Combustion Plants [21]. After 2010, the progressive decrease in the emissions factors is mainly driven by the integration of renewables in electricity generation and the transition away from fossil fuels, with prices for emissions allowances under the EU Emissions Trading Scheme increasing in relevance, especially since 2019 [21].

As documented by the calculated data, the percentage change for the CO₂ emissions over the period of 1990–2021 is −56.2% for the EU-27 and over the last decade is −34.0% (Table 1). A similar trend is also observed for GHG emissions, with −55.9% and −33.8%, respectively (Table 2). Luxembourg exhibits a very high percentage increase, although the values of the emissions factor remain relatively low (e.g., reaching 0.1582 kgCO₂-eq/kWh_el in 2021). Luxembourg experienced strong economic and population growth, resulting in a relatively significant increase in energy use and emissions from 2001 (1.62 TWh_el and 0.0062 kgCO₂-eq/kWh_el) that peaked in 2010 (4.59 TWh_el and 0.0179 kgCO₂-eq/kWh_el) since the main energy source for generating electricity was natural gas (over 70% in early 2000s, reaching 63% in 2010) [3].

Overall, 2021 was more GHG intensive by 5.4% compared to 2020 (from 0.2736 kgCO₂-eq/kWh_el to 0.2883 kgCO₂-eq/kWh_el), while electricity generation increased by 4.4% in the EU-27 [3]. This GHG emissions intensity increase occurred despite the small 1.3% increase in renewable generation in 2021, since the electricity generated from solid fossil fuels increased by 18.8% [3] as a result of the high prices for natural gas. At the national level, the largest increase in the 2021 emissions factors was 18.8% in Estonia, which reached 0.6908 kgCO₂-eq/kWh_el, followed by an increase in 15.6% in Bulgaria (0.6908 kgCO₂-eq/kWh_el) and 15.2% in Ireland (0.3314 kgCO₂-eq/kWh_el). On the other hand, the best performers in terms of improved national emissions intensity in 2021 compared to the previous year were 23.9% in Portugal (0.1362 kgCO₂-eq/kWh_el), −19.6% in Luxembourg (0.1582 kgCO₂-eq/kWh_el) and −15.3% in France (0.0951 kgCO₂-eq/kWh_el).

Estonia exhibited a notable advancement in decarbonizing its electricity production over the 1990–2021 period (Table 2), mostly as a result of the closure of oil shale units and moving towards renewables for about half of the electricity mix (mainly from solid biofuels and renewable wastes, and wind). The other top five best performing countries include Lithuania, as a result of increasing electricity generation from renewables (mainly wind and hydropower); Denmark, by restructuring electricity generation from onshore and offshore wind and solar photovoltaics while phasing-out coal-fired cogeneration plants; Portugal, by showing a rapid phase out of solid fossil fuels and oil for electricity generation and transition to hydro and wind generation, along with growing use of natural gas; and Spain, with similar phase-out trends moving towards renewables, natural gas and nuclear. In Sweden, practically all electricity comes from renewables (67%) and nuclear (31%), and hence, the GHG emissions factor was only 0.0456 kgCO₂-eq/kWh_el in 2021.

As illustrated in Figure 3, in 2021, the five countries with the highest electricity generation carbon intensity were Poland (1.0595 kgCO₂-eq/kWh_el), Malta (0.7418 kgCO₂-eq/kWh_el), Estonia (0.6908 kgCO₂-eq/kWh_el), Bulgaria (0.6900 kgCO₂-eq/kWh_el) and Cyprus (0.6640 kgCO₂-eq/kWh_el). This is mainly attributed to the national electricity mix, for example, high use of solid fossil fuels, low integration of renewables, and limited or no nuclear. Seven more EU027 Member States have a carbon intensity that is higher than the EU average (i.e., Czechia, Germany, Greece, Romania, Ireland, The Netherlands and Slovenia). The five best performers with the lowest emissions factors were Sweden, Austria, France, Portugal and Lithuania as a result of their high share of low-carbon electricity sources (i.e., renewables and nuclear).

Emissions from the European Building Stock

The calculated indirect emissions associated with the use of electricity from the European residential and non-residential buildings are summarized in

Table 3. Electricity consumption, allocated carbon dioxide (CO₂) emissions (kgCO₂/kWh_el) and greenhouse gas (GHG) emissions (kgCO₂-eq/kWh_el) from European residential and non-residential buildings in 2021.

EU Member State	Electricity Consumption (GWhel) [17]		Calculated Emissions Factors		Calculated Emissions
	Residential	Non-Residential	CO2	GHG	Residential		Non-Residential
			(kgCO2/kWhel)	(kgCO2-eq/kWhel)	CO2 (MtCO2)	GHG (MtCO2-eq)	CO2 (MtCO2)	GHG (MtCO2-eq)
Austria	20,283	11,863	0.0874	0.0891	1.77	1.81	1.04	1.06
Belgium	19,267	20,377	0.1574	0.1586	3.03	3.06	3.21	3.23
Bulgaria	11,956	8560	0.6856	0.6900	8.20	8.25	5.87	5.91
Croatia	6594	5687	0.1659	0.1681	1.09	1.11	0.94	0.96
Cyprus	1814	2035	0.6619	0.6640	1.20	1.20	1.35	1.35
Czechia	17,259	15,259	0.6517	0.6553	11.25	11.31	9.94	10.00
Denmark	10,839	9699	0.1956	0.2018	2.12	2.19	1.90	1.96
Estonia	2233	3256	0.6852	0.6908	1.53	1.54	2.23	2.25
Finland	24,260	17,131	0.1416	0.1447	3.43	3.51	2.43	2.48
France	169,775	131,977	0.0943	0.0951	16.02	16.14	12.45	12.55
Germany	138,467	121,568	0.4192	0.4276	58.05	59.21	50.96	51.99
Greece	17,719	16,549	0.4085	0.4095	7.24	7.26	6.76	6.78
Hungary	13,026	8164	0.2280	0.2297	2.97	2.99	1.86	1.88
Ireland	8804	13,851	0.3278	0.3314	2.89	2.92	4.54	4.59
Italy	67,059	79,933	0.2218	0.2228	14.87	14.94	17.73	17.81
Latvia	1791	2710	0.1987	0.2052	0.36	0.37	0.54	0.56
Lithuania	3408	3524	0.1304	0.1371	0.44	0.47	0.46	0.48
Luxemburg	942	2221	0.1575	0.1582	0.15	0.15	0.35	0.35
Malta	1012	1023	0.7394	0.7418	0.75	0.75	0.76	0.76
Poland	30,587	52,684	1.0547	1.0595	32.26	32.41	55.57	55.82
Portugal	14,282	14,712	0.1338	0.1362	1.91	1.94	1.97	2.00
Romania	14,247	8839	0.3399	0.3414	4.84	4.86	3.00	3.02
Slovakia	5966	7164	0.1699	0.1715	1.01	1.02	1.22	1.23
Slovenia	3803	3326	0.3103	0.3119	1.18	1.19	1.03	1.04
Spain	73,153	69,140	0.1354	0.1372	9.90	10.04	9.36	9.49
Sweden	46,206	29,971	0.0438	0.0456	2.02	2.11	1.31	1.37
The Netherlands	22,760	35,995	0.3270	0.3304	7.44	7.52	11.77	11.89
EU-27	747,497	697,208	0.2850	0.2883	213.04	215.48	198.71	200.98

. The data on the electricity delivered to residential and non-residential buildings sectors, along with the total electricity consumed by all sectors in 2021, are available from the detailed annual energy balance flows [17]. The data are then multiplied by the calculated emissions factor for 2021 using the proposed approach that is also listed in Table 3 to obtain the corresponding calculated GHG emissions for both buildings sectors. These indirect operational emissions from electricity consumption should be added to the commonly reported emissions from the buildings sector, which are limited to direct (onsite) combustion of fossil fuels for heating and other end-uses, like cooking using natural gas. These values are not readily reported and can help quantify the total environmental impacts of the European building stock.

The calculated indirect emissions for residential and non-residential buildings from electricity consumption reveal some interesting insights. The EU-27 emissions between the two sectors differ by only 6.7%. The smaller building stock of non-residential buildings is outweighed by their higher use of electricity compared to the residential buildings sector, which is dominated by use of thermal energy consumption for space heating, DHW and cooking. For comparison, the EU-27 direct emissions from onsite combustion in 2021 are estimated at 305.88 MtCO₂ from residential and 128.50 MtCO₂ from non-residential buildings and for GHG emissions at 324.74 MtCO₂-eq from residential and 129.90 MtCO₂-eq from non-residential buildings [3]. Accordingly, the total carbon emissions of residential buildings reached in 2021 a total of 518.92 MtCO₂, which are about 58% higher than the non-residential buildings sector. Similarly, the total GHG emissions of residential buildings reached in 2021 a total of 540.22 MtCO₂-eq, which are about 63% higher than the non-residential buildings sector. All together, the direct and indirect operational emissions from the EU-27 buildings sector in 2021 reached 29.3% of the total carbon emissions, of which 18.0% were from residential and 11.3% were for non-residential buildings. For the GHG emissions, the available data from the direct and indirect operational emissions translate to 24.6% of the total carbon emissions in 2021, of which 15.2% were from residential and 9.3% were for non-residential buildings.

5. Case Studies on Future Projections

According to the results, it is evident that the GHG intensity of electricity production differs significantly among EU Member States and progressively from year to year. However, several EU countries either use the default European value or do not regularly update their national emissions factors used in energy performance certification schemes or other climate change and decarbonization plans for assessing progress toward short- and long-term targets. Similar issues have also been encountered with other conversion factors like the one used to translate final to primary energy use [22]. This can make a notable difference in the results in Member States where renewables have a significant share in the energy mix for electricity generation (such as in the case of Greece), while it may be less important in others where electricity generation is dominated by the use of fossil fuels (such as in the case of Poland). The following two case studies demonstrate some practical aspects and applications of this work for future projections of emissions savings from energy efficiency measures in Greece and Poland.

5.1. Hellenic Buildings

According to the Hellenic National Energy and Climate Plan (NECP), which performed detailed energy simulations and analysis of the national energy balances [23], the projected electricity generation and carbon emissions for the period 2020–2030 provide values that are very close with the estimated factors derived in this work (Figure 5). The NECP value for 2020 was a projection using the detailed energy model with a reference year of 2017, which was the most recent year for which there was a complete official national energy balance at the time of preparing the plan. The results give confidence in the overall approach. The NECP national official projections from the detailed complex national energy simulation studies are very close or bounded by the estimated values using the proposed correlations. The proposed correlations can be easily updated on an annual basis, acting like auto-corrections as time progresses and more recent annual data become available. In particular, the 10-year trend line will be most directly affected since it is more sensitive every time it is adjusted using more recent data. Apparently, the negative values have no physical meaning and for longer time projections, it may be best to include an end-point for 2050 that will correspond to a future renewable-dominated electrical grid.

The default values for the emissions factors were introduced in Greece by the national regulation on the energy performance of buildings, which was published in 2010 and has not been updated since then [24]. These factors have been used extensively for building certification and the assessment of different energy efficiency measures of the building stock. Specifically, for electricity, the national average was set at 0.989 kgCO₂/kWh_el, differentiating among the mainland power network at 0.850 kgCO₂/kWh_el and the isolated local island networks at 1.0625 kgCO₂/kWh_el. Apparently, the emissions factors have changed significantly considering that the use of renewables in the Hellenic electricity mix was only 13.7% in 2009 and reached 40.6% in 2021 [3]. For comparison, the emissions factors for Greece were calculated in this work at 0.9263 kgCO₂/kWh_el and 0.9293 kgCO₂-eq/kWh_el in 2009 and have dropped to 0.4085 kgCO₂/kWh_el and 0.4095 kgCO₂-eq/kWh_el in 2021. Another data resource reports the 2021 emissions factor in Greece at 0.438 kgCO₂-eq/kWh_el [25].

The indirect emissions in 2021 amounted to 7.24 MtCO₂ and 7.26 MtCO₂-eq from residential buildings and 6.76 MtCO₂ and 6.78 MtCO₂-eq from non-residential buildings (Table 3). In addition, the direct (combustion) emissions are also reported at 4.55 MtCO₂ and 4.81 MtCO₂-eq for residential and 0.62 MtCO₂ and 0.63 MtCO₂-eq for non-residential buildings [3].

Depending on the building type, the average energy use intensity of residential buildings is about 23–47 kWh/m² per year for residential buildings [26] and ranges from 17 to 145 kWh/m² for non-residential buildings [27], depending on the building use (e.g., schools on the lower end and hospitals on the upper end), building age and location as it relates to the prevailing weather conditions. The potential savings from different energy efficiency measures can lower the electricity demand by up to 1.32 TWh per year for residential buildings [26] and by up to 1.41 TWh per year for non-residential buildings [27]. Accordingly, using the constant average national default emissions factor of 0.989 kgCO₂/kWh_el, this could translate to potentially curbing emissions from all building types by (2.73 × 10⁹ kWh_el × 0.989 kgCO₂/kWh_el) or 1.84 MtCO₂ per year. This is consistent with Greece’s NECP, which has set a target to reach by 2030 a reduction of –55% for all GHG emissions and specifically –35.4% for the GHG emissions in the effort sharing sectors (e.g., that related to buildings is limited to only direct combustion) compared to 2005 [23]. Considering that the total CO₂ emissions in 2030 are projected to reach 41.3 MtCO₂ [23] from 60.1 MtCO₂ in 2020 [3], a policymaker may interpret the results that solely the reduced emissions from the electricity savings in the buildings sector may be sufficient to meet this target. Using the default emissions factor the annual savings will accumulate to a total of 18.4MtCO₂ during 2021–2030.

However, this may be an overestimate of the actual savings and the progress towards the national and EU targets. For example, using the linear correlation derived from all historical data (refer to Table 1) it is possible to estimate the annual emissions factors during this decade (as illustrated in Figure 5) and convert the annual electricity savings of 2.73 × 10⁹ kWh_el to emissions savings. Accordingly, this will translate to an abatement of (2.73 × 10⁹ kWh_el × 0.477 kgCO₂/kWh_el) or 1.3 MtCO₂ in 2021, (2.73 × 10⁹ kWh_el × 0.363 kgCO₂/kWh_el) or 0.99 MtCO₂ in 2025 and (2.73 × 10⁹ kWh_el × 0.220 kgCO₂/kWh_el) or 0.60 MtCO₂ in 2030, for a total of 9.52 MtCO₂ during the decade (Figure 6). This is about 48.3% lower than the estimate using the default emissions factor.

Similar results are also obtained using the emissions factor derived from the last 10-year data (10 y) and illustrated in Figure 6. As previously discussed, using the ambitious projections with the 10 y correlation, the negative values are not included in Figure 6 for the last two years. Overall, the cumulative CO₂ emissions savings and consequently the abatement of GHG emissions may fall short of the national target and even distort the effectiveness of certain energy efficiency measures and policies in the buildings sector.

5.2. Polish Buildings

In Poland, the GHG emissions are calculated and published annually by the National Centre for Emissions Management (KOBiZE) according to the national regulation [28], which is a very good practice in order to keep these factors updated. Accordingly, the relevant calculations for the energy performance certification of buildings in Poland are performed using the emissions factors that are published on an annual basis by KOBiZE, unless more accurate data are provided by a specific supplier of electricity.

The use of solid fuels (hard coal and lignite) dominates the fuel mix for electricity generation, although it follows a slowly dropping trend. For example, in 1990 it covered 95.6% of the gross electricity generated on Poland, compared to 71.1% in 2021 [3]. During the 1990s, Poland exhibited a significant decrease in emissions (Figure 4) attributed to the major changes that occurred during the political transformation and the implementation of various policies and energy efficiency measures in power generation and other heavy industries [29]. While emissions factors continued to decline in recent years, there was an increase in emissions from public electricity and heat production in 2021 by 15.5% compared to 2020 that exceeded 20 million tonnes as a result of using more fossil fuels for electricity generation, like solid fossil fuels by 18.8% to generate 21.6 TWh of electricity or 13.7% compared to 2020 [3]. Furthermore, the currently adopted solutions in Poland do not consider a total coal phase-out in the following decades [30]. Accordingly, the dropping trend of the emissions factors will flatten out and will need to be accounted for in the projections by consistently updating with the most recent data.

The total electricity use in residential buildings represents about 12.8% of the final energy consumption [31] that reached 22.1 Mtoe in 2021 [3]. This translates to only 32.9 TWh_el of annual electricity consumption in residential buildings, since electricity is only used for plug loads, lighting and cooking, while electrical heating is limited to 5.1% [31]. According to the Polish NECP, the main economically viable energy conservation measures to reduce GHG emissions in the buildings sector include higher energy performance standards, improved thermal envelope protection, increased use of renewables and alternative heating and electricity supply options for buildings [23]. Specifically, the target for the cumulative energy savings as a result of improving the energy performance of residential buildings is estimated at 43.4 TWh over the period 2021–2030 [23]. On an annual basis, this corresponds to about 4.3 TWh per year or 1.2% of the total final energy consumption in the buildings sector that reached 356.3 TWh in 2021 [3]. Assuming an equivalent breakdown of electricity consumption and annual energy savings, the allocated electricity savings for residential buildings can reach 394.8 GWh_el per year. The annual electricity savings in the non-residential buildings sector are more significant and average 4.2 TWh_el per year [32]. Accordingly, the total electricity savings for the buildings sector average 4.7 TWh_el per year.

The most recent national emissions factor in Poland for 2021 is reported at 0.7608 kgCO₂/kWh_el with LULUCF, which is usually a more conservative value, and at 0.8148 kgCO₂/kWh_el and 0.8013 kgCO₂-eq/kWh_el without LULUCF [29]. Another data resource reports the 2021 emissions factor in Poland at 0.857 kgCO₂-eq/kWh_el [33]. Using the recent national emissions factor, the potential reduction in CO₂ emissions is estimated at (4.7 TWh_el × 0.8148 kgCO₂/kWh_el) or 3.83 MtCO₂ per year, and, similarly, the abatement of GHG emissions is estimated at (4.7 TWh_el × 0.8013 kgCO₂/kWh_el) or 3.77 MtCO₂-eq per year. For comparison, the emissions factors for Poland were calculated in this work (without LULUCF) at 1.0547 kgCO₂/kWh_el and 1.0595 kgCO₂-eq/kWh_el in 2021.

The calculated annual emissions factors using the corresponding national linear projection with the correlation derived from all historical data (Table 1) can be used to convert the annual electricity savings of 4.7 × 10⁹ kWh_el. Accordingly, this will translate to an abatement of (4.7 × 10⁹ kWh_el × 0.927 kgCO₂/kWh_el) or 4.4 MtCO₂ in 2021, (4.7 × 10⁹ kWh_el × 0.746 kgCO₂/kWh_el) or 3.5 MtCO₂ in 2025 and (4.7 × 10⁹ kWh_el × 0.521 kgCO₂/kWh_el) or 2.4 MtCO₂ in 2030, for a total of 34.03 MtCO₂ during the decade (Figure 7). This is about 11.1% lower than the estimate using the most recent national emissions factor.

Figure 7 also illustrates the results from using the correlation derived from the last 10-year data (10 y). The total emissions savings amount to 33.69 MtCO₂ during the decade (Figure 7), which is 12.0% lower than the estimate using the most recent national emissions factor. Overall, while there is a good practice in Poland to calculate and update the national emissions factor, the future projections will still need to utilize the most recent national value, which may lead to deviations and over ambitious estimates.

6. Limitations

The simple calculations presented in this work account for the integration of renewables in the energy mix and the efficiency of the transmission and distribution losses. They are easy to update on an annual basis using officially reported data from a single source [3]. However, they do not account for the emissions associated with the imported electricity. This will impact the calculated factors in some Member States that have high electricity imports. For example, the top net electricity importers in 2021 as a percentage of the final electricity consumption [3] include Luxembourg (89.5%), which imports most of its electricity from Germany and other neighboring countries, like France and Belgium; Lithuania (75.7%), since domestic electricity generation decreased significantly after 2010 following the shutdown of its nuclear reactors; and Estonia (30.4%), which mainly imports from Nordic countries, like Finland and Sweden, since it is often at a lower cost. Overall, Sweden and Bulgaria are the top electricity exporters, while the Baltic countries are mainly top importers. However, detailed tracking of electricity exchanges is challenging and mandates complex analysis to properly account for the allocated emissions. This is very relevant when considering diverse power generation infrastructures from the coal-based production in Bulgaria to the low-cost and low-carbon electricity available in Sweden.

Considering future projections, one needs to exercise caution, especially in countries where there has been a notable growth in electricity generation from renewables. In this case, the most recent 10-year correlations should be updated on an annual basis since they are more sensitive to the recent growth trends of wind and solar generation plants, which may not be sustained. Adapting the corresponding correlation trends will provide more realistic projections.

6.1. Accounting for CHP Heat Bonus

The method used in this work for the calculation of the emissions factors utilizes the available data from [3]. The allocation of the emissions from public electricity and heat production are inclusive of useful heat from CHP plants. As noted in Section 2, the current calculation approach and the results do not account for a heat allocation on the CHP electricity generation since there is no historic data included in [3] before 2009 and no specific information that can be used directly. However, in countries where CHP in total electricity generation is high, the current calculation approach for the emissions factors will result in higher values. In this case, it is necessary to allocate the total CHP input fuel and the associated emissions to the different electricity and heat outputs.

This information is not directly available from [3], and therefore accounting for the CHP heat will involve some additional calculations and assumptions. To start, the available annual data after 2009 include the amount of generated electricity and heat, the total fuel input into CHP and the percentage breakdown of the fuel mix. The latter data can be used to estimate the associated combustion emissions considering default values for the various fuels, for example, 56.2 gCO₂-eq/MJ for natural gas, 73.2 gCO₂-eq/MJ for oil, 96.1 gCO₂-eq/MJ for solid fossil fuel, 54.7 gCO₂-eq/MJ for renewables and waste (biogas).

Different standardized methodologies can be used to allocate the fuels to the two outputs, like the power station displacement method and the boiler displacement method [33]. A commonly used simple method that is adopted from the Digest of UK Energy Statistics (DUKES) assumes that the production of electricity is half as efficient as the production of heat. This derived from the fact that the standalone electricity production plants average an efficiency of about 40%, while the efficiency for separate heat production averages 89% [34]. Therefore, the impact is that the CHP plant will use twice as much fuel to generate one unit of electricity compared to what is required to generate one unit of heat.

As an example, consider Latvia, which has the highest percentage of CHP exploitation in the EU-27 with 36% of the total electricity generation coming from CHP, which also represents 40% of the final electricity consumption in 2020 [3]. The corresponding useful energy contained in the CHP-generated electricity was 7.6 PJ, while the CHP heat production was 12.1 PJ [3]. Latvia has shifted away from the use of fossil fuels in the CHP plants, which was 90% natural gas, 4% solid fossil fuels and 1% oil in 2009. Currently, the share of fuel input into CHP was natural gas (44%) and the remaining from renewables and waste (mainly biogas) for a total fuel input to CHP of 24.2 PJ in 2020 [3]. The allocated fuels for CHP heat generation are then estimated at 3.53 PJ for natural gas and 4.5 PJ for biogas, while the allocated emissions reach 0.198 MtCO₂-eq and 0.246 MtCO₂-eq, respectively. For Latvia, this represents about 34% of the total public electricity and heat production emissions in 2020. Accordingly, accounting for the CHP heat allocation, the corrected GHG emissions factor becomes 0.1340 kgCO₂-eq/kWh_el compared to the estimated values of 0.1610 kgCO₂-eq/kWh_el and 0.2650 kgCO₂-eq/kWh_el using the derived historic and 10 y correlations (Table 2). The differences range from 17% with respect to the estimate using all the historical data up to 49% using the 10-year data. This is the worst-case scenario since this corresponds to a case where CHP plays a significant role in power generation.

Considering that the fuel mix changes from year to year, the relevant calculations have to be updated annually. A similar analysis is performed for all EU-27 Member States for the emissions coefficients in 2020, which is the most recent year with available data. The emissions factors following the correction for the CHP heat allocation are summarized in Figure 8 along with the initially estimated values using the two derived correlations. For comparison, the figure also includes the European Environment Agency (EEA) emissions intensities for 2020 [20]. As illustrated, the highest GHG emissions factors in 2020 occurred in Cyprus, Estonia and Poland. This was the result of using fossil fuels, along with relatively low contributions in the energy generation mixture from renewables and limited or no nuclear.

7. Conclusions

The review of the emissions factors for electricity generation in all EU-27 Member States revealed significant differences as a result of the different energy carriers and the integration of renewables. The EU-27 averaged 0.2883 kgCO₂-eq/kWh_el in 2021, with clear dropping trajectories for most of the countries. Over the past three decades, GHG emissions factors for electricity were 42% less than in 1990, which is similar to the drop of 56% reported from the calculations over the 1990–2020 period in [35]. This decreasing trend is expected to continue by exploiting more renewables in the energy sector. According to the results of this work, the EU-27 projected emissions in 2030 using all historical data and the more ambitious 10 y will reach about 0.2350 and 0.1520 kgCO₂-eq/kWh_el, respectively. Provided that the recent trends continue, the more ambitious projections indicate that it may be possible to reach the indicative intensity levels that would be consistent with the EU’s climate targets of 0.110 to 0.118 kgCO₂-eq/kWh_el by 2030 [20].

The calculated indirect GHG emissions from the electricity consumption in European residential and non-residential buildings revealed some practical insights that can complement the direct (combustion) emissions. The highest annual indirect emissions from residential and non-residential buildings in 2021 (Table 3) occur in Germany and Poland, which together make about 54.4% and 53.6% of the total EU-27 GHG emissions, respectively.

The contribution of the emissions from all types of buildings depend on the size of the building stock, the emissions factors for electricity generation and the different energy carriers used by the building combustion systems for heating. For example, the direct (combustion) emissions [3] are higher in Germany for both the residential buildings (83.54 MtCO₂-eq) and the non-residential buildings (33.31 MtCO₂-eq), compared with the buildings in Poland (i.e., 35.25 MtCO₂-eq from residential buildings and 7.74 MtCO₂-eq from non-residential buildings). Considering the total indirect and direct emissions, the German building stock is responsible for 6.4% of the total EU-27 GHG emissions, followed by Poland with 3.7% of the total EU-27 GHG emissions.

This work also provided results that are not readily reported by European and international organizations to help quantify the total environmental impacts of the European building stock. The lowest annual indirect emissions occur in Luxembourg, as a result of the small size building stock. GHG emissions reach 0.149 MtCO₂-eq from residential buildings (Table 3), and 0.351 MtCO₂-eq from non-residential buildings. Accordingly, the contribution of the total GHG emissions from the building stock in Luxembourg is 1.87 MtCO₂-eq or only 0.05% of the total EU-27 GHG emissions.

The total indirect GHG emissions from the EU-27 residential building stock in 2021 reached a total of 215.48 MtCO₂-eq, which is about two-thirds of the direct emissions at 324.74 MtCO₂-eq. For non-residential buildings, electricity is the most common energy source and as a result the indirect emissions dominate reaching 200.98 MtCO₂-eq, which is over half of the direct emissions at 129.90 MtCO₂-eq.

The operation of the EU-27 buildings contributed a total of 846.13 MtCO₂ and 871.10 MtCO₂-eq in 2021 or 29.3% and 24.6% of the total EU-27 carbon and GHG emissions. The percentage for the GHG emissions is relatively lower since the contributions of other greenhouse gases from the operation of buildings is primarily related to carbon dioxide, while other GHGs like methane primarily originate from various other anthropogenic and natural sources. Accounting for the emissions associated with the construction industry will also add another 10% from manufacturing building products and materials, which is associated with the emerging decarbonization issue of embodied carbon.

The work proposed a consistent, replicable and easy to implement method for estimating and projecting emissions factors for electricity. The calculation approach exploits readily accessible official data from a single source [3], which is annually updated so that the results may be adapted and revised with time. The proposed correlations can be used for simple projections that can be easily integrated in preliminary NECPs before completing the time demanding detailed energy balance studies and future models. In addition, they can be exploited in building energy performance assessments and bottom-up building stock models instead of constant emissions factors in simple life-cycle analysis.

Future Work

For accounting purposes, the use of average values for the emissions factors is sufficient. However, for calculating future carbon offsets and quantifying actual environmental impacts as a result of taking different measures on the grid will mandate the use and forecast of marginal emissions factors. This will play an even more decisive role as a result of the increasing share of using intermittent renewables for electricity generation. For example, marginal GHG emissions can be about 150% higher than the emissions calculated using an average mix [36]. Apparently, as European buildings embark on the era of electrification, we will need this type of detailed information for deciding when to use certain appliances or when to store electricity, among others. However, marginal emissions factors that vary by location and time are not readily available yet. In this direction, emerging marginal emissions models that utilize hourly electricity generation and emissions data will eventually provide the necessary tools.