Assessment of City-Scale Rooftop Photovoltaic Integration and Urban Energy Autonomy Across Europe

Abstract

This study suggests a newly developed model for estimating city-scale photovoltaic rooftop energy potential. This model aims to provide reasonable universal calculations regarding a city’s available space for mounting rooftop photovoltaic systems and their corresponding annual electricity production capacity. For the development of the model, a thorough literature review has been conducted, which compiles and presents mathematical expressions and performance coefficients. Necessary geographic and meteorological data have been obtained from European statistical repositories and the PVGIS tool, respectively. The main inputs refer to a city’s basic geographical data, population, total actual area, geographical coordinates, and, by extension, the optimum PV unit installation angle. This analysis presents a simple and accurate model applicable to European cities for assessing rooftop photovoltaic energy potential and suitable rooftop space for PV units. The findings can aid in advancing PV development in urban areas and contribute to creating environmentally neutral cities in the future. The methodology is verified with data retrieved from the Google Environmental Insights Explorer tool, which shows a deviation of 9.72%. According to the computational analysis for 40 European countries, the photovoltaic energy potential is between 12.31 GWh and 8200 GWh. These values correspond to a net available PV space between 0.03 km² and 31.86 km². The greatest photovoltaic coverage potential is equal to 117.4% for Patras, Greece, while the lowest is 7.27% for Oslo, Norway. Regarding the avoided greenhouse gas emissions, they are found to vary from 5.8 ktons of CO_2-equivalent for Valletta, Malta, and 8109.8 ktons for the city of London, United Kingdom. Finally, the final results of 86 additional cities located on the European continent are given.

1. Introduction

The imperative transition towards sustainable and energy-efficient built environments is particularly crucial due to the escalating worldwide energy demands and climate change impacts. The implementation of transformative energy efficiency and renewable energy solutions is necessary for the transformation of the energy-demanding building sector into an energy-efficient and autonomous sector that supports on-site energy production, self-consumption, and community energy distribution. The combination of passive energy-efficient renovation technologies [1] and the incorporation of renewable photovoltaic systems is an effective strategy for the creation of efficient zero-energy buildings [2]. This strategy is in line with the European Union’s (EU) initial towards fully electrified buildings, the decarbonization of its energy mix [3], and the increase in the renewable contribution to electricity generation [4]. The integration of rooftop photovoltaic (RTPV) systems is a means to increase the renewable energy capacity within the building sector and an important step towards the EU’s holistic approach to transforming its energy landscape and achieving the EU’s 2030 and 2050 climate goals [5].

Based on a study by D’Agostino et al. [6], the use of thermal insulation and the appropriate installation of PV modules, exploiting the shading effect for cooling purposes, results in an almost 58% decrease in total primary energy. Ziozas et al. [7] investigated the retrofitting of an old historic Italian building, and proposed an energy renovation strategy of various passive thermal envelope technologies, in combination with the RTPV system, and calculated annual final energy demand savings of 37%. Kitsopoulou et al. [8] systematically investigated the zero energy balance potential of residential buildings, and calculated the maximum possible number of stories for the four distinct climatic categories of Greece. Except for studies that concentrate on building-scale renewable integration, studies and meta-analyses that take into account diverse geographical and environmental conditions that influence solar availability have played a pivotal role in assessing the RTPV potential across Europe. More specifically, an analysis by Kapsalis et al. [9] incorporates various parameters, namely population density, electricity production, and multiple environmental factors such as shading and performance ratio, to calculate RTPV systems’ contribution potential for various climates. Molnár et al. [10] predicted that RTPV electricity production potential in the EU could escalate to 3.56 PWh by 2060, underlying the major contribution of this technology to meeting renewable energy targets.

Estimating PV potential is a complex, multi-dimensional process with significant variability [11]. There are two main approaches: a constant value method, which uses existing building data, for instance, space, population density, and regional PV potential, to identify relationships, and a GIS-based approach, which is more detailed, utilizing layered maps to assess space availability and PV potential [12]. Regarding the constant value approach, in a study by Gernaat et al. [13] the growing role of rooftop PV systems in the global energy market is examined using the integrated assessment model IMAGE [14]. Socioeconomic factors like income and electricity prices impact rooftop PV deployment more than solar irradiation. Moreover, Castellanos et al. [15] examine methodologies for assessing RTPV potential in urban areas, focusing on scalability and resolution. The study compares high-resolution, city-specific methods with generalized, low-resolution approaches, highlighting significant discrepancies in potential estimations. The findings emphasize the need for more accurate, scalable tools to guide policy development and maximize rooftop PV adoption. The study by Mainzer et al. [16], presents an in-depth analysis of the technical potential for residential PV installations across Germany. This research employs a statistical approach, leveraging data on the available roof area which is, and presents estimates of the PV potential on the municipality level by combining socio-economic data with parameters such as building types, roof inclinations, and orientations. This method allows for calculating the technical potential for PV across different municipalities with high resolution. Finally, Defaix et al. [17] estimate the usable area for PV systems in EU countries based on population density and building floors, projecting electricity generation capacity from PV across the European Union. By using a harmonized approach to assess the suitability of building surfaces for PV installation, their analysis highlights significant potential for PV to support the EU’s renewable energy goals.

Regarding the GIS-based method, Molnár et al. [10] investigate the electricity production capability of RTPV systems across the EU using a GIS-based approach. Using high-resolution geospatial techniques to assess the spatial and temporal factors affecting rooftop PV production, they calculated that the current technical potential could reach about 2.7 PWh, which is closely aligned with the EU’s total power consumption. Germany, France, Italy, and Poland stand out as having the highest PV potential, expected to increase by 30% by 2060. Moreover, Bódis et al. [18] utilize a geospatial methodology with satellite data, statistics, and machine learning to assess the technical and economic viability of rooftop PV across Europe with a spatial resolution of 100 m. Their findings suggest that EU rooftops could cover one-fourth of present electricity consumption, with two-thirds of this potential available at costs lower than ongoing residential rates. However, the study also identifies significant challenges in countries where electricity prices are low, but investment rates are high, underscoring the need for policy adjustments and better integration of rooftop PV assessments into grid management and planning. Additionally, in a study by Rodriguez et al. [19] applied to Germany, it was found that utilizing all available rooftop areas could cover 77% of the territory’s electricity needs, with economic potential at 56% for high irradiance roofs, and some municipalities could exceed 100% coverage under certain scenarios. Meanwhile, Izquierdo et al. [20] estimated that solar hot water systems and PV systems in Spain could meet 70% of service hot water demand, with PV alone projected to provide 10 TWh annually, covering 4.0% of total electricity need. In Sweden, Yang et al. [21] developed a framework to analyze roof-mounted solar PV potential, focusing on the city of Västerås, Sweden, estimating a total rooftop space of 5.74 km² and a potential installed capacity of 727–956 MW, yielding 626–801 GWh annually.

In the study by Assouline et al. [22], a novel GIS method combined with machine learning techniques is used to calculate at a country level the RTPV electricity potential. According to their calculations, for east- or west-oriented rooftops, the annual electricity production potential is equal to 16.3 TWh, which corresponds to one-fourth of the country’s total electricity need for 2017. On the other hand, by combining a GIS method with a statistical model, Phillips et al. [23] calculated that the installation of RTPV systems on residential buildings can result in a 25% contribution to the USA’s electricity market. Lastly, Gassar and Cha [24] conducted a comprehensive review study regarding GIS-based RTPV potential calculation methods in cities. They underlined the combined use of light detection and ranging (LiDAR) technology and GIS as an accurate method for RTPV electrical potential.

The present literature review provides valuable insights into the existing methodologies regarding city-scale photovoltaic electricity production and city energy autonomy. It also aids in optimizing the integration of RTPV, offering valuable perspectives and guidelines for local and national authorities, urban planners, and renewable energy supporters. Building on the insights derived from existing studies, this paper introduces a novel model that broadens the scope of RTPV potential evaluation by incorporating advanced methodologies and open-access statistical data from national, European, and international sources. The newly developed model adjusts the global phenomenological correlational parameters to a regional scale, thereby more accurately estimating RTPV potential for major cities in Europe and Greece. The proposed methodology is an important advancement in comprehending RTPV’s role in building electrification and climate neutrality objectives. This analysis’s novelty and significance stem from the developed model’s accuracy and simplicity, which can be applied across all European cities to provide reasonable estimates of PV energy potentials and the cities’ current PV rooftop availability space. The findings of this study can be used for further photovoltaic implementation in urban areas and to help create future cities that are environmentally sustainable.

2. Materials and Methods

In this chapter, the methodology and the developed calculation model of the RTPV potential in multiple major European is presented in detail.

2.1. Followed Methodology

In this study, a simple model for the calculation of a city’s RTPV electricity production potential is developed. This model aims to provide reasonable universal estimations regarding a city’s available space for PV installation and annual PV electricity production potential. For this reason, basic parameters linked to a city’s geographical data are needed as inputs. These parameters regard a city’s population and total actual area, its latitude angle, and by extension the optimum PV module tilt angle. To first deduct a city’s actual PV available space the utilization factor is defined based on the existing literature. The utilization factor numerates the architectural suitability, solar potential, and shading conditions of a city. Also, an equation correlating the utilization factor with a city’s population density is formed. Following that, for the accurate calculation of the PV installation suitable space, geometric parameters regarding rooftop PV space planning are taken into consideration. These parameters concern corridors between PV clusters and the distances between PV arrays, and information retrieved from existing literature is retrieved for the necessary assumptions. Additionally, a function of the city-scale average packing factor in terms of a city’s latitude is formed. Furthermore, suitable coefficients that take into account operation and shading losses are thoroughly investigated and retrieved from the existing literature. The annual on-plane solar irradiation at the optimum tilt angle is retrieved from the PVGIS tool [25], while a typical PV efficiency is considered. The aforementioned calculation process results in the estimation of a city’s photovoltaic energy potential and net area available for PV installation.

2.2. Mathematical Formulation

For every examined city, information concerning its total actual area, A in [km²] and population, P, are retrieved from Refs. [26,27], and given in detail in Appendix A. According to these numbers, a city’s population density, (ρ) in [m²/capita] as given below:

$$\rho ={\displaystyle \frac{\mathrm{A}}{\mathrm{P}}}$$

(1)

Following, for every city, the available surface area per capita for PVRT installation, denoted as (A_capita) in [m²/capita] is calculated as [15]:

$${\mathrm{A}}_{\mathrm{c}\mathrm{a}\mathrm{p}\mathrm{i}\mathrm{t}\mathrm{a}}=\mathrm{a}·{\rho }^{-\mathrm{c}}$$

(2)

where (a) and (c) are constants with values equal to 172.3 and 0.352, respectively, according to Ref. [15]. According to this, a city’s available surface area for PVRT installation, denoted as (A_city) in [km²] is calculated as:

$${\mathrm{A}}_{\mathrm{c}\mathrm{i}\mathrm{t}\mathrm{y}}={\mathrm{A}}_{\mathrm{c}\mathrm{a}\mathrm{p}\mathrm{i}\mathrm{t}\mathrm{a}}·\mathrm{P}$$

(3)

Additionally, it is important to define the value of building area density [28], denoted with (D) in [m²/capita], which is calculated according to the following equation:

$$\mathrm{D}={\displaystyle \frac{{\mathrm{A}}_{\mathrm{b}\mathrm{u}\mathrm{i}\mathrm{l}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{g}}}{\mathrm{P}}}$$

(4)

where (A_building) in [m²], denotes a city’s building footprint, equivalent to a city’s area covered with buildings. The density of the building area of every examined city is given in Appendix A. The average value of building area density among the examined cities is denoted with (D_average) and is calculated at 46 m²/capita, according to Table A1 in Appendix A.

For the calculation of the actual available surface area for PVRT installation per city, the determination of the utilization factor, (U_t) in [%], is necessary [29]. The utilization factor expresses a city’s architectural and solar suitability. More specifically, the term architectural suitability expresses the suitable part of the building for the installation of PV systems, taking into account neighboring constructions, shading, and historical elements. The term solar suitability expresses the percentage of the surface resulting from architectural suitability that is capable of receiving a satisfactory intensity of incident solar irradiation, namely above 80% of the maximum annual solar irradiation. Their average value is around 60% and 55% for rooftops, respectively [30]. According to Gutschner et al. [30], for Western Europe, the mean value of the utilization factor is around 0.4, but it ranges between 0.25 and 0.75, depending on a city’s population density. A city with a higher population density has a higher utilization factor.

In Appendix A, the calculated population density for every examined city is given in Table A1, according to which 60% of the examined European and Greek cities demonstrate a population density between 100 and 300 m²/capita. For these cities, a utilization factor in the range of 30 to 50% is assumed. According to the aforementioned, Figure 1 is constructed, from which the following logarithmic equation for the calculation of (Ut) is extracted, with an R-squared coefficient of 0.9617.

$${\mathrm{U}}_{\mathrm{t}}=19.466·\mathrm{l}\mathrm{n}(\rho )-62.53$$

(5)

Figure 1. Utilization factor per city per population density.

The actual suitable PV available surface area for each city, (A_S) in [km²], is calculated according to the following equation:

$${\mathrm{A}}_{\mathrm{S}}=\left[\left({\mathrm{A}}_{\mathrm{c}\mathrm{i}\mathrm{t}\mathrm{y}}·{\mathrm{U}}_{\mathrm{t}}\right)-\left|{\mathrm{D}}_{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}-\mathrm{D}\right|·\mathrm{W}·\mathrm{P}\right]$$

(6)

A weigh factor (W) is considered to apply to the correction factor $\left({\mathrm{D}}_{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}-\mathrm{D}\right)$. The weight factor is determined by a city’s population and is calculated as a function of its building area density (D). That is because in smaller cities, which are characterized by a population less than 500,000 capita, the most common building topology is one-story buildings, instead of multistorey as in the case of larger cities. The weight factor is calculated according to the following piecewise function:

$$\mathrm{W}=\left\{\begin{array}{c}{10}^{-4}·{\left|{\mathrm{D}}_{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}-\mathrm{D}\right|}^2-3.6·{10}^{-3}·\left|{\mathrm{D}}_{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}-\mathrm{D}\right|+0.12 , \mathrm{P}\ge 5·{10}^5\\ {2·10}^{-4}·{\left|{\mathrm{D}}_{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}-\mathrm{D}\right|}^2-2.9·{10}^{-3}·\left|{\mathrm{D}}_{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}-\mathrm{D}\right|+0.21 , \mathrm{P}<5·{10}^5\end{array}\right.$$

(7)

The packing factor (PF) is defined as the ratio of the total area covered by the PV arrays, (A_PV,Net) to the total land area they occupy (A_PV,Gross). According to the study of Martín-Chivelet [31], the packing factor can be computed based on the following equation:

$$\mathrm{P}\mathrm{F}={\left(\mathrm{c}\mathrm{o}\mathrm{s}\beta +{\displaystyle \frac{\mathrm{s}\mathrm{i}\mathrm{n}\beta }{\mathrm{t}\mathrm{a}\mathrm{n}(66.5-\phi )}}\right)}^{-1}$$

(8)

where (β) is the tilt angle of RTPV, and φ is the city’s latitude. Figure 2 illustrates the geometrical dimensions of the RTPVs space planning, in which (L) denotes the PV module length, (d) the distance between PV arrays, and (γ) the sun azimuth angle. The packing factor is geometrically expressed as the ratio of (L) over (d) [31].

Figure 2. Visual explanation of packing factor.

For sloped roofs, the packing factor is considered equal to 1.0, because PVs follow the slope of roofs, which is close to the optimum tilt angle. The slope of the roofs is defined mostly according to a location’s latitude, and therefore PV modules are considered to be installed in the best way possible regarding spatial planning, which is translated to a PF equal to 1.0. Northern European countries tend to have more steeply sloped roofs compared to those in southern Europe because of the snowfall. This architectural feature predominantly stems from the climate differences between the two regions. For this reason, it is assumed that for latitudes below 42° (φ < 42°), for the calculation of the packing factor is estimated that RTPVs are installed with the optimum tilt angle on flat roofs. Conversely, for latitudes above or equal to 42° (φ ≥ 42°), the packing factor is estimated as the average between sloped and flat roofs. According to these, Figure 3a,b are plotted, depicting the packing factor in terms of the latitude angle, namely for latitude, φ < 42°, and φ ≥ 42°, respectively. Specifically, for φ < 42° and according to Figure 3a the R-squared is found equal to 0.9715, while for φ ≥ 42° and according to Figure 3b the R-squared is equal to 0.9962, which are both very high coefficients of determination for the two-part linear regression equation of the PF. The calculated PFs for every examined city are summarized in Table A2 in Appendix A.

Figure 3. Packing factor in terms of latitude (a) for latitude (φ < 42°) and (b) for latitude (φ ≥ 42°).

Based on the following plots, (PF) can be estimated as a function of latitude according to the following equations:

$$\mathrm{P}\mathrm{F}=\left\{\begin{array}{c}-0.0168·\phi + 1.1814, \phi <42°\\ \\ -0.0095·\phi + 1.1458, \phi \ge 42°\end{array}\right\}$$

(9)

The efficiency of the photovoltaic panels, η_PV, including inverter losses, is considered equal to 25% [32], a consideration that takes into account the technological advancement in the field of photovoltaics. The annual on-plane irradiation for every city, (G_T) in [kWh/m²], is retrieved from the PVGIS tool [25] for the optimum tilt angle. These values are summarized in Table A2 in Appendix A.

The performance ratio measures the final yield of a photovoltaic (PV) system against its reference yield, comparing the actual energy generated to the energy that would be produced under optimal conditions—specifically, under the same solar irradiation but without any losses [33]. Typically, losses in PV systems arise from factors such as array temperature, solar spectrum variations, angular reflection, surface dirt accumulation, wiring issues, and inefficiencies or failures of other PV components during operation. The annual average PR is typically around 0.8 [34]. In this study, the PR values for each examined city are retrieved from the PVGIS tool [25] and are summarized in Table A2 in Appendix A.

The net area available for PV installation, (A_PV,Net) is calculated as an expression of a city’s total area (A), population (P) and latitude (φ), as given below:

$${\mathrm{A}}_{\mathrm{P}\mathrm{V},\mathrm{N}\mathrm{e}\mathrm{t}}=\left[\mathrm{a}·{\left({\displaystyle \frac{\mathrm{A}}{\mathrm{P}}}\right)}^{-\mathrm{c}}·{\mathrm{U}}_{\mathrm{t}}-\left|{\mathrm{D}}_{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}-\mathrm{D}\right|·\mathrm{W}\right]·\mathrm{P}·\mathrm{P}\mathrm{F}$$

(10)

The overall energy potential by the rooftop PVs, (PVEP) can be calculated accordingly:

$$\mathrm{P}\mathrm{V}\mathrm{E}\mathrm{P}={\mathsf{A}}_{\mathrm{P}\mathrm{V},\mathrm{N}\mathrm{e}\mathrm{t}}·{\mathsf{\eta }}_{\mathrm{P}\mathrm{V}}·{\mathrm{G}}_{\mathrm{T}}·\mathrm{P}\mathrm{R}$$

(11)

Finally, for the calculation of the avoided greenhouse gas emissions, (GHG_emissions) in [kton], due to the electricity production from rooftop PVs the following equation is used:

$$\mathrm{G}\mathrm{H}{\mathrm{G}}_{\mathrm{e}\mathrm{m}\mathrm{i}\mathrm{s}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}}=\mathrm{P}\mathrm{V}\mathrm{E}\mathrm{P}·\mathrm{C}{\mathrm{O}}_{\text{2-equivalent}}$$

(12)

where CO_2-equivalent in [kg/GWh], denotes the electricity conversion factor to GHG emissions retrieved from Ref. [35] for every examined city.

3. Results

This section presents the results retrieved by the presently developed method for multiple European-examined cities. Two validation methodologies are used, highlighting the developed methodology’s satisfactory accuracy. Specifically, the Google Environmental Insights Explorer tool and graphical validation using geospatial data confirm the validity and credibility of the present work.

3.1. PV Energy Potential Results

This subsection presents the necessary calculations for the calculation of a city’s PV energy potential and actual PV available space. Firstly, Figure 4 illustrates the calculated total actual PV-occupied area in terms of a city’s population. The almost linear correlation between A_PV,Net and population shows that a city’s total actual PV-occupied area is directly linked to its population. More specifically, the greater the population, the larger the total actual PV-occupied area of a city. According to the present model, London is characterized as the city with the greatest RTPV potential, with a total actual PV-occupied area of 31.86 km² and a population of 5.527 million. Moreover, Figure 5 depicts the total actual PV-occupied area per capita in terms of the city’s population density (ρ). It is noticed that the total actual PV-occupied area per capita is sharply increasing from very low to normal population densities, while it remains steadily high for very high densities or sparsely populated cities until it decreases again for the maximum population densities. This phenomenon can be explained by how population density relates to building types and available rooftop space. In densely populated cities, most people live in multi-story buildings, which provide limited rooftop area suitable for installing PV systems. In contrast, less densely populated cities usually have smaller buildings that house fewer people but offer more rooftop space for PV installations. In our study, Ljubljana, Slovenia, is characterized by the highest total actual PV-occupied area per capita equal to 22.62 m²/capita, and a population density of 554.3 m²/capita.

Figure 4. Total actual PV occupied area in terms of the city’s population.

Figure 5. Total actual PV occupied area per capita in terms of the city’s population density.

Following that, Figure 6 illustrates the photovoltaic energy potential in terms of a city’s population. As in the case of calculating the total actual PV-occupied area, PVEP is strongly, almost linearly linked to a city’s population. The highest PVEP is calculated to be 8200 GWh for the city of London. Additionally, Figure 7 depicts the rooftop PV energy potential in terms of a city’s total PV installed area. According to this figure, the city of London and Berlin present the high PVEP and greatest total PV installation area, which is calculated at 31.86 m² and 27.08 m², respectively.

Figure 6. Rooftop PV energy potential in terms of the city’s population.

Figure 7. Rooftop PV energy potential in terms of the city’s total PV installed area.

In Figure 8 the PV energy potential per capita is given for all examined cities in descending order. More specifically, according to the calculations based on the newly developed PVEP method, the highest PVEP is calculated for the city of Ljubljana, Slovenia, at 6389.6 kWh/capita, while the lowest is calculated for the city of Paris, France, at 832.6 kWh/capita.

Figure 8. Rooftop PV energy potential per capita for every examined city.

Given the fact that the calculated A_PV,Net is the maximum possible PV-covered area of a city, the following calculated PV-coverage corresponds to a city’s maximum possible PV-covered electricity demand. Based on the estimated energy production of RTPV and the total annual electricity consumption per city, retrieved from the EU energy statistics [36], the maximum possible PV-covered electricity demand for every examined city has been calculated and illustrated in Figure 9. Particularly, Patras, Greece, presents the highest possible PV coverage percentage, equal to 117.4%, while Oslo presents the lowest, equal to 7.27%. The highest PV electricity potential is presented for the southern European cities, while the top three PVEPs are calculated for three Greek cities. More information and the calculations for all examined cities are presented in Table A4 in Appendix B.

Figure 9. Maximum electricity demand PV coverage per city.

3.2. Verification of the Followed Methodology

The present developed methodology is validated through the collection of the available geospatial data retrieved from the Google Environmental Insights Explorer tool [37] which uses GIS data. Specifically, Figure 10 and Table 1 summarize the photovoltaic energy potential calculated by the newly developed methodology, the PVEP retrieved from the online tool, and the calculated deviation between them. This data concerns 21 of the investigated cities. The maximum deviation is found for the cities of Ioannina and Stockholm at 19.0% and the lowest, practically zero deviation for the city of Barcelona. The mean absolute deviation is calculated at 9.72%. This result validates that the developed methodology provides satisfactorily accurate calculations, provided that this is a generic model suitable to be applied to any European city. Additionally, in Appendix C, a comparison is performed between the developed methodology and geospatial data processed in CAD. The geospatial analysis conducted for the four cities of Athens, Barcelona, Munich and Stockholm confirms the high accuracy of the methodology in calculating the available rooftop area suitable for PV installation.

Figure 10. Deviation in PVEP between the present methodology and Google Insights.

Table 1. Comparison of PV energy potential calculations of the present methodology and the Google Insights tool.

	PVEP [GWh]
City	Present Methodology	Google Insights	Deviation [%]
Athens	1046	979	−6.8%
Patras	752	710	−5.9%
Ioannina	156	192	19.0%
Rome	4798	5210	7.9%
Barcelona	1759	1760	0.0%
Madrid	5600	4770	−17.4%
Paris	1751	1780	1.6%
London	8200	7440	−10.2%
Bruxelles	2189	2530	13.5%
Berlin	7196	8100	11.2%
Munich	3300	3930	16.0%
Stockholm	1458	1800	19.0%
Copenhangen	988	1190	16.9%
Warsaw	3528	3800	7.1%
Bucharest	4136	3680	−12.4%
Zurich	1074	1110	3.2%
Amsterdam	1506	1540	2.2%
Vienna	4489	4980	9.9%
Zagreb	1644	1620	−1.5%
Helsinki	1318	1680	21.5%
Luxemburg	415	412	−0.8%

3.3. Avoided Greenhouse Gas Emissions

The installation of rooftop PV systems will lead to the mitigation of greenhouse gas (GHG) emissions. If the available city area to be equipped is considered to be the maximum one, A_PV,Net, the maximum possible GHG emissions are calculated. Specifically, Figure 11 illustrates the maximum avoided GHG emissions due to the maximum PV installation capacity per city. The maximum value of avoided GHG emissions is calculated for the city of London, United Kingdom, and is equal to 8109.8 kton per year, while the minimum value of avoided GHG is calculated at 1.2 kton for the city of Valletta, Malta. For the calculation of these values, the necessary electricity to CO_2-equivalent conversion factors, according to the life-cycle approach, are retrieved from Ref. [35]. This information and the respective calculations for all the examined cities are also given in Table A5 in Appendix B.

Figure 11. Maximum avoided GHG emissions per city.

3.4. Expansion of the Methodology in Extra European Cities

At this point, the presently developed PVEP method is applied to numerous European cities spread around the continent. Specifically, Table 2 summarizes the calculations regarding the total actual PV occupied area and PV energy potential for 86 additional cities located in the European continent, corresponding to 33 countries. These findings are important for the electricity production potential mapping in Europe from renewables, and the identification of decarbonization capability in the European area.

Table 2. Total actual PV occupied area and PV energy potential.

Country	Cities	APV,Net [km2]	PVEP [GWh]
Albania	Elbasan	0.436	180
Austria	Klagenfurt	2.213	660
Belgium	Antwerp	6.330	1651
	La Louvière	1.713	454
	Leuven	2.137	567
Bosnia and Herzegovina	Sarajevo	1.725	598
Bulgaria	Gabrovo	0.195	70
Cyprus	Limassol	0.768	385
Czech Republic	Liberec	1.421	395
Denmark	Aarhus	1.354	331
	Sønderborg	0.181	45
Estonia	Tartu	0.548	133
Finland	Espoo	3.474	897
	Lahti	2.035	467
	Lappeenranta	0.358	84
	Tampere	2.408	544
	Turku	2.063	553
France	Angers Loire Metropole	1.540	476
	Bordeaux Metropole	3.087	1032
	Dijon Metropole	1.791	550
	Dunkerque	1.833	507
	Grenoble-Alpes Metropole	0.978	310
	Lyon	3.149	992
	Marseille	7.613	3056
	Nantes Metropole	2.699	831
Germany	Aachen	3.298	905
	Dortmund	5.530	1460
	Dresden	5.064	1408
	Frankfurt/Main	6.795	1886
	Heidelberg	1.782	501
	Leipzig	5.446	1509
	Mannheim	6.595	1896
	Münster	4.561	1210
Greece	Kozani	0.170	71
	Trikala	0.271	111
Hungary	Miskolc	2.696	824
	Pécs	1.781	574
Iceland	Reykjavík	0.547	146
Ireland	Cork	2.881	782
Italy	Bergamo	2.505	859
	Bologna	4.545	1517
	Florence	4.932	1673
	Milan	9.371	3128
	Padova	4.619	1550
	Parma	3.498	1163
	Prato	4.221	1460
	Turin	7.157	2380
Latvia	Liepāja	0.492	137
Lithuania	Taurage	0.131	35
Luxembourg	Differdange	0.214	57
Malta	Gozo	0.187	91
Montenegro	Podgorica	1.749	697
The Netherlands	Eindhoven & Helmond	1.565	403
	Groningen	1.738	432
	Rotterdam	3.890	1009
	The Hague	3.413	902
	Utrecht	2.920	756
	Oslo	3.064	849
Norway	Stavanger	1.127	344
	Trondheim	0.390	102
Poland	Krakow	5.549	1520
	Łódź	5.354	1481
	Rzeszow	2.136	589
	Wrocław	5.245	1476
Portugal	Guimarães	5.753	815
	Porto	1.870	834
Romania	Cluj-Napoca	2.618	918
	Suceava	0.572	195
Slovakia	Košice	1.235	367
Slovenia	Kranj	0.241	81
	Velenje	0.156	55
Spain	Seville	2.778	1345
	Valencia	2.717	1263
	Valladolid	1.282	583
	Vitoria-Gasteiz	1.356	503
	Zaragoza	1.385	648
Sweden	Gävle	1.282	298
	Gothenburg	4.592	1128
	Helsingborg	1.777	447
	Lund	0.979	250
	Malmö	3.179	834
	Umeå	1.422	359
Turkey	Istanbul	32.479	13,410
	Izmir	0.415	196
United Kingdom	Bristol	3.573	1117
	Glasgow	3.601	959

3.5. Discussion on Methodological Sensitivity

The methodology developed in this work relies heavily on the availability and quality of input data. The key datasets required for city-scale assessments of rooftop PV potential and energy autonomy include population statistics, geographical information, and detailed descriptions of the built environment. However, a city’s built environment is constantly evolving, altering its architecture solar accessibility and therefore it is essential that the methodology be regularly updated. This is particularly evident across Europe, where urban environments are continuously undergoing transformation pursuing climate-resilient [38] and adopting the New European Bauhaus policy [39] to promote green transition and attractive urban spaces. To ensure reliable estimations of a city’s photovoltaic energy potential, it is of primary essence to regularly update inputs such as the utilization factor, which depends on population dynamics and stems from available literature data on the built environment. Furthermore, the escalating trend towards deep building renovation and transition to zero energy buildings is predicted to significantly change urban landscapes in the following years, creating expanded opportunities for renewable energy integration to enhance energy autonomy and support decarbonization.

Additionally, the aging demographic profile of Europe needs to be considered when estimating the availability of rooftop areas. In general, older populations tend to occupy smaller apartments or buildings compared to younger households with families, who typically require larger living spaces. Given that Europe is experiencing a declining birth rate and, consequently, a gradual decrease in urban populations, it is essential to regularly update parameters such as the weight factors applied to the correction factor (D_average − D) to ensure more accurate estimations of the rooftop area available for PV installations. In the present study, the latest population statistics, geographical information and literature-based data on built environment are considered, providing a reliable methodology for city-scale assessments of the available rooftop area for PV installations in Europe.

4. Conclusions

To summarize, this study provides a comprehensive analysis of the feasibility of rooftop photovoltaic systems to cover the electricity demand of multiple major European cities. For this reason, a newly developed methodology is presented that can lead to universally reasonable estimations regarding a city’s net available space for PV installation and its annual PV energy potential. Using basic inputs that concern a city’s total area, population, and latitude, this methodology is applied to multiple European cities, leading to useful numerical results. More specifically, analytical calculations are provided for 40 European countries, according to which the PV energy potential ranges between 12.3 GWh and 8200 GWh for corresponding net available PV spaces that range between 0.03 km² and 31.86 km². The minimum calculated values concern the city of Valletta, Malta, and the maximum values are the city of London, United Kingdom. Regarding the PV energy potential per capita, Ljubljana, Slovenia, is characterized by the highest value, equal to 6.39 MWh/capita, while Paris, France, has the minimum, equal to 0.83 MWh/capita.

Calculations that concern a city’s electricity demand coverage by rooftop PV-installed systems are executed. Among the 40 analytically examined cities, Patras, Greece, showcases the highest PV coverage potential, equal to 117.4%, and Oslo, Norway, the lowest, equal to 7.27%. Electricity production from renewable energy sources contributes to decarbonizing a city’s energy mix. According to the present analysis results, 8109.8 kton of greenhouse gas emissions are avoided in the case of London, United Kingdom, which has the highest calculated value. In contrast, only 1.22 ktons of CO_2-equivalent emissions are avoided for Valletta, Malta, the lowest calculated value. Finally, except for the 40 major European cities, the final results of 86 additional cities located on the European continent are given. Lastly, the model is validated with data retrieved from the environmental Google insights tool, and an average deviation of 9.10% is found. Future research should focus on combining rooftop photovoltaic systems with various energy storage systems or innovative cooling techniques to enhance RTPV electrical performance. This research field will contribute to transforming cities into climate-neutral and energy-autonomous cities.