Solar potential for social benefit: Maps to sustainably address energy poverty utilizing open spatial data in data poor settings

Access to affordable sustainable energy is a significant challenge for many lowincome countries experiencing


Introduction
To decrease the negative consequences of climate change, it is pivotal to reduce carbon emissions and reliance on fossil fuels globally.Sustainable Development Goal (SDG) number 7, as defined and agreed upon by the United Nations (UN), signals the importance of access to affordable, reliable, sustainable, and modern energy to reduce the current anthropogenic causes of greenhouse gas (GHG) emissions that destabilize the global climate (Intergovernmental Panel on Climate Change (IPCC), 2021).A major part of the global emissions come from fossil fuel combustion (Intergovernmental Panel on Climate Change (IPCC), 2021) and in the building sector, about 30 % comes from the built environment during operations (International Energy Agency, 2023) (Adams, Burrows, & Richardson, 2019).Targeting climate change mitigation measures within the energy sector is essential.In response to an actionable plan to address the problem, European Union (EU) countries are placing importance on decarbonising the energy system.
In 2019, the Green Deal was introduced by the EU, setting the goal that by.
2030 EU members to cut GHG emissions by 55 % and strive to make the EU the first climate-neutral continent by 2050 (Lee-Makiyama, 2021).In 2020, the EU acknowledged the need to also address energy poverty within the Green Deal (Massera, 2020).Geopolitical implications of how the Green Deal is structured have been venomously critiqued for its focus on profitable opportunities for Western Europe, under the justification that it is doing moral good, while neglecting contexts outside of Western Europe (Vela Almeida et al., 2021).For the Green Deal goals to be met, fundamental structural changes need to occur in the energy system to reduce GHG emissions, requiring costly investments out of reach for many individual households.Existing political incentives for low-carbon transitions often overlook the needs of low-income populations who face energy poverty (Sovacool, Martiskainen, Hook, & Baker, 2019).
Energy poverty occurs when households struggle to afford adequate energy services in their homes (Nierop, 2014).Fifty million households in the EU, one in four EU citizens, have inadequate access to energy services, and the recent gas price increase is worsening this (Trading Economics, 2022).Increases in fossil energy prices cause civil unrest.For example, in 2018, citizens in Bulgaria protested against the government for the higher fuel prices and taxes implemented on polluting cars (Massera, 2020).Bulgaria is one of the countries in the EU suffering the most from energy poverty.
International politics beyond the EU have also started to acknowledge the need to make the transition to clean energy.In an impressive diplomatic feat, the UN has identified a series of 17 SDGs to address wicked challenges facing our world, for example: the UN SDG 7. Associated with each goal is a series of targets to be achieved by the year 2030, and 241 indicators or datasets which each country is encouraged to self-report to measure how close they are to meeting each goal.
What is currently lacking in SDG indicator data reporting are requirements or suggestions on how to collect and share spatial data that could be used to make maps.Maps are pivotal for illuminating localized sustainable development solutions to achieve a sustainable world (Kraak, Ricker, & Engelhardt, 2018;Kraak, Roth, Ricker, Kagawa, & Le Sourd, 2020;Ricker, Kraak, & Engelhardt, 2020).Defining spatially explicit indicators could help identify the best places to make interventions to meet the targets associated with the SDG 7 s (see Table 1 for all targets and goals associated with Goal 7).More appropriate methods to make maps to help identify solutions to meet the SDGs are needed.As seen in Table 1, SDG 7 targets and indicators are listed.Presently, indicator data are aggregated at the country level as proportions of the total population and do not require spatial data collection.To map these indicators associated with SDG 7, energy data from private energy agencies are needed as well as comprehensive census data.However, these are not readily available in many of the regions facing energy poverty, thus proxy data are needed.(See Table 2.) Spatial data and maps could certainly help with localized decision-making to meet these goals and targets.Creative and innovative mapping techniques outside the official UN SDG indicator metadata standards could be useful to identify where local interventions could be put in place to meet the SDGs and associated targets.These new mapping practices could inform the next set of SDG indicators in the future.
Understanding the link between energy management and socioeconomic disadvantage is beneficial for developing effective mitigation strategies.Careful considerations are needed in decarbonisation policies to avoid redistributive consequences that negatively affect vulnerable groups (Vela Almeida et al., 2021).When these groups are not considered in policy design, only the wealthy enjoy the benefits of clean energy, while low-income sections of society do not have the means to purchase these new low emissions technologies (Korsnes & Throndsen, 2021;Schwanen, 2021).
The building sector contributes one-third of the EU's total emissions and 75 % of the buildings are considered energy inefficient (Massera, 2020).While pursuing decarbonization, one strategy that can be implemented in the built environment is through renewable energy implementation.
Solar technology, in particular photovoltaic (PV), is a technically viable solution to meet future energy demands while decarbonizing the energy system (IEA, 2021;Wiginton, Nguyen, & Pearce, 2010).PV technology could offer financial benefits for the public by offering immediate financial relief to consumers by reducing a household's energy dependency from the grid which currently runs on fossil fuels (Breyer et al., 2017;Coughlin & Cory, 2009).While it is known that other renewable energy sources have similar merits, solar PV is most suitable to implement in the built environment and at the household level, as other sources require community efforts or are significantly more costly for individual households.
Solar technology does come with a high initial cost that renders it an injudicious investment if placed in an area with limited sunlight.The solar potential on a building should be calculated to determine whether a PV investment is viable.Solar potential mapping helps determine optimal PV placement but is also essential for resource planning, grid management, formulating financial schemes, and developing future energy policies (Choi, Suh, & Kim, 2019;Litjens, Kausika, Worrell, & van Sark, 2018;Wiginton et al., 2010).
Solar potential mapping originally focused on technical aspects, such as statistical estimations of building footprints that are sometimes tied to the geographical environment.Improvements have been made to incorporate geographical aspects to account for the building's orientation and roof tilt angle needed to optimize PV system energy yield, as well as with other objects in proximity to the building for shadowing effects.The availability of highly detailed spatial data is often limited to countries that can afford to collect and manage these data regularly, and are willing to make them publicly available.The focus of solar potential mapping has often been on the technical and related economic feasibility of PV, while overlooking the broader societal implications.This oversight is particularly significant when studying countries that grapple with energy poverty.PV technology could be used to decarbonize the building sector while simultaneously reducing energy costs, and alleviating some of the causes of energy poverty.Solar potential mapping, including social economic variables, is required to support effective policy design and localized planning to help reach SDG 7. The addition of new indicators that reflect whether a community is selfsufficient is needed.
Here we offer an approach to map solar and social benefit potential together.We aim here to identify the best places for PV panels, not only to decarbonize the building sector but also to address social economic challenges causing energy poverty.We do this using open spatial data and off-the-self tools readily available to most government agencies.There is value in open data for energy mapping in that the output derived from it could increase the quality of energy policies and planning (Fremouw, Bagaini, & De Pascali, 2020).Using open data is pivotal for replicating these methodologies where data availability is scarce.Geographic open data is useful for solar potential mapping for countries facing energy poverty where localized data required for these calculations are typically unavailable.Adequate solar potential mapping that includes socio-economic factors can help inform policymakers where the benefits of PV can positively impact vulnerable communities that experience energy poverty.Thus, the main research question is: How can solar PV potential be combined with socio-economic metrics to help policymakers achieve SDG7 in data-scarce regions?
The following sub-questions are formulated to help answer the overarching research question, where we will use the city of Plovdiv in Bulgaria as a case study: 1. What is the average solar radiation and potential electricity generation by PV on building rooftops?2. What is the annual electricity yield potential in comparison to the energy demand and potential monetary savings for the citizens?3. Considering the socio-economic context, where are the most strategic places to place solar panels to meet the energy needs of the population (SDG7.1)and alleviate energy poverty?
The remainder of the paper is structured as follows: in Section 0 we describe mapping in data poor regions.Energy poverty is discussed and how it is addressed in policy and mapping.Solutions for solving energy poverty through policy and solar technologies are discussed, and then the solar potential mapping is explained.Section 0 describes the study area of Plovdiv.In Section 0 we outline the methodological approach used for socio-economic solar potential mapping using open data.We present the results obtained in Section 0. In Section 0 we discuss the interpretations and implications of the results and recommendations.Finally, in Section 0 we conclude with a summary of our research findings and policy recommendations.

Mapping in data poor regions with open data
Cartography and maps can aid in the appropriate planning and placement of specific interventions.Maps can be a tool to meet SDGs (Kraak et al., 2018;Kraak et al., 2020;Ricker, Kraak, & Engelhardt, 2020).To make maps, data are required.Several different SDG indicators rely on census data for their calculation.While different UN agencies offer "metadata standards" or methodologies to collect each of the specific SDG indicator data, they are only guidelines that countries are at liberty to modify as a form of data sovereignty.Several of the SDG indicators rely on census data for official calculations.Maps have been used to spatially display the distribution of poverty using census data (Encinas et al., n.d.;Fabbri & Gaspari, 2021).However, not all countries collect census data or make their census data publicly available.Often regions experiencing poverty and inequality also experience it with data as well.It is well known that data quality is not uniform and varies dramatically, so improving data quality will lead to more accurate maps (Encinas et al., n.d.;Fonte, Bastin, See, Foody, & Lupia, 2015;Goodchild & Glennon, 2010).For this reason, creative ways to garner relevant and representative localized data are needed for data poor regions (Miller, 2010).This can be done by extracting data from open global datasets.
Open data has been heralded for its increased transparency, reusability, and citizen engagement, and is seen as a form of democratizing data access (Huijboom & Van den Broek, 2011).Open data is linked to information justice as it can be used to support social justice movements for challenging practice and structures (Johnson, 2014).Examples of how open data can be used and improved for social good can be seen in Cape Town South Africa, the first city in Africa to release open data.Here Non-Government Organisations use spatial open data to create a bi-directional flow of data to augment missing data, and to generate information products for social causes (Ricker, Cinnamon, & Dierwechter, 2020).Similar initiatives could be useful to address SDG 7, to illuminate what data are available, and to assess what is still needed to address access to affordable renewable energy.However, there are limitations in existing open data to realize them for energy poverty mapping (Fremouw et al., 2020;López-Vargas, Ledezma-Espino, & Sanchis-de Miguel, 2022).

Energy poverty: Policy and mapping
Energy poverty is a concept that refers to households that cannot pay for energy supply to meet their energy demands.It refers to a household that is unable to obtain a socially and materially needed degree of energy services in their home (Bouzarovski, Thomas, & Cornelis, n.d.;Bouzarovski, 2014).Energy poverty is a socially shaped and measured complex phenomenon (Sareen et al., 2020).Energy poverty cannot be described by a single metric (Bouzarovski et al., n.d.).There are different ways to calculate and different conceptualizations that offer unique and useful perspectives on the problem.One example of an early attempt to quantitatively define energy poverty is through Boardman's definition: "a household is in energy poverty if it needs to spend more than 10% of income on basic energy needs" (Boardman, 1991).Energy poverty is a residential sector problem with three overarching aspects: high energy bills, low income, and poor energy efficiency (Pye et al., 2015).Insufficient energy efficiency can be assessed using indicators of energy use patterns, such as the type of heating systems used, the share of central heating, and energy consumption type.The housing pattern indicator can be measured using tenure systems and dwelling features to quantify low-income and poor energy efficiency.High energy bills and low income can be assessed using indicators of energy affordability, such as income, energy prices, and energy consumption levels.Energy poverty has direct effects on well-being and health, which are factors often not taken into account during measurement (Castaño-Rosa, Solís-Guzmán, Rubio-Bellido, & Marrero, 2019;Fabbri & Gaspari, 2021).Again, there is no official consensus on the definition (Massera, 2020), which makes it difficult to understand, measure, and map, making it an underresearched topic (Jessel, Sawyer, & Hernández, n.d.).
One limiting factor of energy poverty is that it is researched mostly from the econometric epistemology, and misses the medical, social, and energy aspects of the challenge (Fabbri & Gaspari, 2021;Sareen et al., 2020).Energy poverty is geographically variable and locally dependent (Encinas et al., n.d.;Bouzarovski, 2014) and can be mapped as a representation of socio-spatial income inequality (Encinas et al., n.d.).
As no measure of energy poverty is all-encompassing, no policy action is either, but policies can combat energy poverty incrementally (Bouzarovski et al., n.d.;Sareen et al., 2020).Energy poverty is not directly measured in the SDGs, but is indirectly addressed by SDG target "7.1 By 2030, ensure universal access to affordable, reliable and modern energy services" (U.Nations, 2018).These indicators and statistical indices of energy poverty do not match explicitly with policy commitments (Bouzarovski et al., n.d.), but they do imply the same goal.Easyto-comprehend measures of energy poverty are needed for quick, efficient policy-making (and implementation) and urban planning purposes.
The EU is making efforts to solve energy poverty by implementing different energy efficiency strategies.The primary strategy (65 %) is applied to building retrofit measures that increase buildings' energy efficiency (Pye et al., 2015).Others are through appliance grants (12 %), social housing improvements (8 %), legislation.
(7 %), energy efficiency advice (6 %), and rental property improvements (2 %) (Pye et al., 2015).Countries that face energy poverty are in an energy poverty feedback loop where the energy policies do not produce the desired results, and social policies are too costly to administer (European Union, 2019).Designed to reduce energy poverty, schemes have been created to obligate energy companies to improve energy savings by developing systems to identify their energy-poor consumers and implement energy-efficient strategies.This includes promoting renewable energy installations and new financing payments, similar to the SocialWatt project (European Commission, 2021).Specific suggestions include redistribution of financial resources associated with the Green Deal and access to renewable energy to place more emphasis on benefiting communities of migrants, indigenous peoples, and people of colour (Vela Almeida et al., 2021).
Accurate and up-to-date data are needed to effectively map energy poverty, and then to assign policies that are accountable, transparent, and responsible.Many creative ways of mapping energy poverty have been explored, linking socioeconomic variables to environmental variables such as land surface temperature (Castaño-Rosa et al., 2019;Pérez-Fargallo, Bienvenido-Huertas, Rubio-Bellido, & Trebilcock, 2020).Overly complex methods have been conducted to map energy poverty against independent variables such as land price, housing material quality, and winter indoor temperatures, using a geographically weighted regression model by census tract units, professionalsfound energy poverty to be an unequally distributed factor in the city of Santiago Chile (Encinas et al., n.d.).This method, while comprehensive, innovative, and complex, illustrates the many facets of energy poverty and its spatially uneven nature.An innovative Social Weakness Index was developed based on families with low income, building age, and energy consumption, and then mapped in Bologna, Italy as a form of Energy Poverty Mapping (Fabbri & Gaspari, 2021).However, for a policy planner trying to achieve SDG target 7.1, these examples are likely confusing and do not offer direct suggestions on how to achieve this target.

Solar PV technology potential mapping
Solar photovoltaic (PV) energy is a renewable, modern, clean energy source.Solar PV technology converts solar radiation into usable energy in the form of electricity or heat (Andrews, Jelley, & Jelley, 2017).The technology consists of PV panels that typically comprise of 60-72 solar cells of silicon material in which a p-n junction is made (Reinders, Verlinden, van Sark, & Freundlich, 2017).Photons are absorbed in the cells, and charge carriers are generated and separated due to an electric field, which results from the presence of the p-n junction.Power is collected via external contacts of the cells.PV technology comes in the form of PV panels which are connected to inverters to convert DC to AC power.PV systems are usually placed on rooftops, while large multi-MW systems are typically land-based.The advantage of solar PV technology is the capacity to serve as an emission-less and renewable substitute for conventional fossil fuel sources in electricity generation.By harnessing sunlight, PV panels offer a sustainable alternative that can ultimately reduce carbon emissions in the energy supply.
It is an energy source that is rarely available to those suffering from energy poverty due to the high initial investment.Solar PV technology costs are dropping, making them more popular for residential homes in the past decade (Gernaat, de Boer, Dammeier, & van Vuuren, 2020).PV technology is popular among the renewable energy options for the residential sector since it is noiseless, has non-toxic emissions, has simple operation and maintenance, and is cheaper than retail electricity (known as grid parity) (Gospodinova, Milanov, & Dineff, 2020), however, this can vary depending on the value of the retail electricity price set in the country.Depending on the retail electricity price within the country, solar energy can be considered the most optimal renewable energy source for the building sector, it can be a potential lever to help reduce energy poverty.
Solar potential mapping involves quantifying the solar potential of an area and then displaying the spatial distribution information on maps.This type of visualization can be valuable to inform the decisionmaking process for investment and urban infrastructure planning.This will help identify and communicate optimal PV placement, which will contribute to efforts to meet SDG 7. Common solar potential mapping practices include the calculation of the physical potential (amount of sunlight received in a location) and the technical potential (possibility of electricity to be generated based on the physical solar potential) (Hong, Lee, Koo, Jeong, & Kim, 2017).
There are several approaches for estimating solar potential: the most common and valuable one is finding places to install PV based purely on solar potential, often targeting rooftops in urban environments.Numerous methods exist for calculating this potential, and access to data required to make these calculations varies.For this reason, currently, there are significant differences in approaches for estimating solar potential.There are mixtures of high-level, geographical analyses and lowlevel, statistical analyses to determine solar potential over various countries, cities, and areas, based largely on spatial data availability (Castellanos, Sunter, & Kammen, 2017).A three-step methodology, including the calculation of physical potential and technical potential, is regarded as useful (Hong et al., 2017).The high variation between approaches shows that no one-size-fits-all method makes it possible to tailor the approach for specific policy design decisions.
To make these spatial calculations requires the use of GIS and spatial data.The data required include elevation data for buildings (and ideally trees and other sun obstructions), the orientation and tilt of the rooftop, and georeferenced rooftop areas to determine the average solar radiation the rooftop receives.The ideal dataset to use for elevation for building a solar potential dataset is a high-resolution light detection and ranging (LiDAR) digital surface model (DSM).
Cadastral digital building footprint datasets are also valuable when calculating solar potential (Alhammami & An, 2021;Kausika et al., 2015), which helps to form the georeferenced rooftop area measurements.With a DSM model and building footprint, the solar potential can be calculated using existing spatial analysis tools.An example of one is the Solar Analyst by the ESRI software, which has been used in several studies to date (Buffat et al., 2018;Fogl & Moudrý, 2016;Freitas et al., 2015;Fu & Rich, 1999;Gassar & Cha, 2021;Kausika et al., 2015;Suomalainen et al., 2017), justifying the credibility of the tool usage here.After calculating the solar potential, it is possible to determine the technical and socio-economic potential of installing and capturing solar energy.
The technical potential refers to the electricity generated from the radiation received over a specific area through a technology (Berjawi, Najem, Faour, Abdallah, & Ahmad, 2017;Hong et al., 2017).Economic potential includes the economic potential generated from technical potential.It can be calculated through various economic measures such as levelized cost of electricity (LCOE) (Bódis, Kougias, Jäger-Waldau, Taylor, & Szabó, 2019), net present value (NPV) (Alhammami & An, 2021), or cost-benefit analysis (Berjawi et al., 2017).LCOE is a measure of the net present cost of electricity generation over the lifetime of the project (Blok & Nieuwlaar, 2020).NPV is a measure to determine an investment's economic viability (Blok & Nieuwlaar, 2020).A costbenefit analysis is a measure to determine when a project's total benefits exceed the costs over a time preference of money (Blok & Nieuwlaar, 2020).
The challenge for city planners and policymakers is then, where should solar panels be placed for optimal cost-benefit and social benefit.For example, it could be helpful to illuminate the social relevance of solar potential per rooftop over the population density per building (Wiginton et al., 2010), or to determine the PV panel penetration per area and compare it with factors such as race/ethnicity, dwelling age, and internet access within the area (Reames, 2020).Moreover, analysis of the solar potential versus the adoption rate of PVs could be useful (Kraaijvanger, Verma, Doorn, & Goncalves, 2023).Investigation of the distribution of accessibility to residential solar photovoltaic (PV) systems in urban environments displays how it intersects with rooftop solar potential using the Solar Analyst tool and an agent-based modeling approach (Kraaijvanger et al., 2023).
The values of the local government can be inserted into the local decisionmaking process through tailored solar potential mapping practices.At present, these solar mapping practices do not include socioeconomic spatial variables.Combining social potential mapping and poverty mapping could in theory be easily incorporated into a single map.Sustainable energy planning could help overcome or even hopefully bypass energy poverty (Szabó, Bódis, Huld, & Moner-Girona, 2013).
The EU offers an online open-access PV-GIS web service, which offers calculations of solar irradiance and PV potential.Maps for cities, countries, and regions can be downloaded (E.Commission, 2022;Huld, Müller, & Gambardella, 2012).However, solar panels are often placed on building rooftops in the residential sector, which is not viewable with these types of maps.They do offer infographics for specific locations.Developing PV rooftop area detection tools is essential for energy policy design, but they need to be modified to meet the goals of the policy.

Policy strategies
Policy strategies to alleviate the economic burden of energy poverty through pricing are needed.Policies that do not emphasize changing the energy price level, but instead liberalizing and privatizing the energy sector, may lead to energy prices increasing at a higher rate than household incomes (Primc & Slabe-Erker, 2020).
Energy poverty in low-income households is exacerbated by an aging dwelling that requires capital to improve the insulation (Primc & Slabe-Erker, 2020).Further, limited fuelswitching choices and energy lock-in effects take hold for households, due to a nation's geographical constraints on energy resources.This is the case for Bulgaria, which produces electricity through domestic coal and imports natural gas (I.T. Administration, 2009).Thus, developing energy policy strategies that include incentives for housing retrofitting and geographical-related resources in the design is a viable approach.Current energy policies in Bulgaria do not address these challenges.
There are renewable energy (RE) implementation strategies in place that attempt to alleviate the energy poverty problem.Often, renewable energy projects involving distributed rather than centralized technologies are considered to have positive effects since these projects avoid the use of ecologically or culturally significant areas and usually incorporate community involvement (Banerjee, Prehoda, Sidortsov, & Schelly, 2017).A standard distributed energy system for households could use solar PV panels to reduce energy costs (O.US EPA, 2015).
Yet, RE projects can cause inequality between distributed societal groups, and the impact varies per group (Banerjee et al., 2017).Electricity generated from renewable energy sources is not always economical and may even increase electricity rates, thus limiting access to inexpensive energy (Banerjee et al., 2017).The burdensome costs of RE projects are often disproportionately distributed among consumers in the residential, commercial, and small industrial sectors (Banerjee et al., 2017).Also, transitioning to renewable energy projects requires a substantial land area which often conflicts with local land rights.Locals frequently have close ties to their land in terms of culture, economy, and the environment.The benefits of renewable energy projects can lead to unintended marginalization due to land ownership (Banerjee et al., 2017).Such effects may unfairly prevent persons from obtaining the products and services to which they are legally entitled.Renewable energy projects create new socio-physical realities and can hopefully reduce energy poverty, depending on the project's access and affordability capabilities (Banerjee et al., 2017).Policies cannot encompass or address all aspects of energy poverty (Bouzarovski et al., n.d.;Sareen et al., 2020).

Study Area: Plovdiv and Stolipinovo
Bulgaria is seen as a developing state within Europe (Ionescu et al., n. d.) and suffers the most from energy poverty (European Union, 2019): 27.5 % of the Bulgaria population could not keep their homes adequately warm in 2020 (E.Commission, 2022).Bulgaria has one of the lowest energy prices in Europe in 2018, yet Bulgarian households spend 11.5 % of their income on energy bills (European Union, 2019), which is above the standard of Boardman's definition of household energy poverty (Boardman, 1991).
The household energy costs in Bulgaria have increased over the years, with electricity costing 7.9€ct/kWh in 2008 and 9.9€ct/kWh in 2018 (European Union, 2019).This makes many Bulgarians unable to keep their homes adequately warm and cool, which is caused by high energy prices, low income, and poor energy efficiency in the homes.This rise in cost has led to many households burning coal and wood for heating, increasing GHG emissions and decreasing the air and living quality (Gospodinova et al., 2020).Coal has the highest carbon content of the fossil fuel source, emitting the most GHG emissions (Blok & Nieuwlaar, 2020).
Despite the energy prices in Bulgaria being lower than the EU average, a study suggests that 444,000 households are highly vulnerable to price increases (Gospodinova et al., 2020;Hajdinjak & Asenova, 2019), due to low-income levels.For reference, there is a low-income level in Bulgaria, which was on average around €527 a month estimated from Census data in 2019 (National Statistical Institute Bulgaria, 2021).
More heat needs and more costs for families are caused by inefficient energy use and poor insulation.A study from 2012 showed that 65 % of housing units were built before 1990, and at least 800,000 were prefabricated buildings (Hajdinjak & Asenova, 2019).Bulgaria is addressing energy poverty through social policy design (European Union, 2019; Pye et al., 2015), but this can be costly to administer (Primc & Slabe-Erker, 2020), and thus an interdisciplinary approach combining social policy and technical (PV) solutions is required.
The most densely populated city in Bulgaria is Plovdiv, which has a population size of 342,048 (N. S. I. Bulgaria, 2021) and a population density of 3365/km2 (E.Commission & N. S. I. Bulgaria, n.d.).The total land area of Plovdiv is estimated to be 82 km2, which is 0.07 % of the total land area of Bulgaria.The city of Plovdiv has a large Roma population that has been largely stigmatized and perceived as invisible or intrusive (Asenov, 2020).10 % of the Roma population across Europe have no access to electricity at all (Lakeman, 2023).
Stolipinovo is one of four prominent Roma neighborhoods within Plovdiv and is known as the largest Roma "ghetto" in Europe (Asenov, 2020).Stolipinovo is an impoverished and densely populated neighborhood located on the outskirts of the city with an estimated population of 40,000 (Babourkova, 2016).80 % of the inhabitants live in 34 buildings, with 2430 apartments, and the rest live in self-made homes, which often lack building permits or access to grid infrastructure (Babourkova, 2016).Some of the Roma population practice seasonal nomadism (Asenov, 2020) meaning it is uncertain if they are included in these population estimates or if they have access to energy.Many Roma do not have access to a national grid and instead burn trash to keep warm in the winter or are forced to pay illegal operators for electricity (Lakeman, 2023).Stolipinovo has historically struggled with its energy management.In 2007, citizens accumulated debt from not paying their energy bills and were cut off from the grid in response to the debt (Babourkova, 2016).This event showcases the extreme consequences of energy poverty not being handled properly.Therefore, Stolipinovo is an area that can benefit from a distributed solar energy grid.Stolipinovo is featured in the inset maps of the results.

Materials and methods
Evolving from a solar potential mapping methodology developed by The Bulgarian digital cadastre started in the early 1990s, and currently, the major cities in Bulgaria have a digital cadastre, but smaller towns are still in the process of digitization (Kitsakis et al., 2018).Only 20 % of the country has an operational digital 2D cadastre system, and none has 3D models (Kitsakis et al., 2018).Access to a building footprint from the country cadastre was not publicly available.The open-source and crowd-generated OpenStreetMap (OSM) 2021 dataset was used for the building footprints (O.contributors, Bulgaria, 2021).The OSM dataset available on 10 June 2021 total building count was 436,198.Upon inspection of the OSM dataset in Plovdiv by comparing it to a satellite image available as ESRI satellite basemap, it was noticed that several buildings were missing and not digitized in the dataset, especially in Stolipinovo (see Fig. 3).Despite this drawback, these are the best data available and will still produce meaningful and valuable estimates for solar potential in this area.For the social-economic data, census data (N.S. I. Bulgaria, 2021;National Statistical Institute Bulgaria, 2021;National Statistical Institute Bulgaria, 2022) from the Bulgarian National Statistical Institute (NSL) were used, together with governmental reports for social indicators.
The radiation potential on all the rooftops in Plovdiv was spatially analyzed using the software ESRI ArcGIS Pro 2.7 (ESRI, n.d.) and specifically the Solar Analyst tool.The building footprint file and the clipped Digital Elevation Model (DEM) were used to calculate the solar potential based on available rooftops.This information was then combined with the social demographic characteristics of the population in the study area.The complete workflow analysis conducted is depicted in Fig. 1 and described further in the following sections.

Physical potential
Physical potential refers to the usable solar radiation.The solar irradiance amount is determined by location, time, and atmospheric conditions.All three spatial components determine the radiation intensity and were included in the Solar Analyst tool provided by the ESRI software.
The Solar Analyst tool uses a comprehensive mathematical model for computing the solar radiation over a landscape, which consists of the hemispherical viewshed algorithm and sunmap and skymap calculations (Fu & Rich, 1999).
Hemispherical Viewshed procedure determines the viewsheds for every cell in an input DEM.A viewshed is an angular distribution of sky visibility relative to obstacle; it is comparable to the information found in upward-facing hemispherical (fisheye) photos.The algorithm includes calculating the horizon angles surrounding a point in certain directions, calculating interpolations for these angles in unknown directions, and transforming the horizon angles into a hemispherical coordinate system, which yields a grid that shows which directions of the sky are visible and which are obscured (Fu & Rich, 1999).
Sunmap and skymap calculations are done by the tool by creating skymaps that show the partition of the entire sky into sectors and sunmaps that show the position of the sun throughout the day and year.These maps are used to determine the direct and diffuse solar radiation for each sector by first calculating the gap fraction, or the percentage of clear sky.
Solar potential analysis is complex and the tool does have several assumptions that lead to limitations related to the solar analysis.The tool uses oversimplified atmospheric condition models that do not dynamically incorporate variations in humidity, pollution, or cloud cover (Fu & Rich, 1999;Fu & Rich, 2000).This could result in inaccurate estimations of solar radiation.The tool ignores differences in reflectivity, vegetation, and land use that affect radiation absorption and reflection in favor of assuming homogeneity in surface qualities across Fig. 1.Overview of the workflow starting with the radiation analysis by using local data fed into the Solar Analyst tool provided by ArcGIS Pro.The technical analysis is executed by calculating the electricity generation using the PV panel specifications.Then this is followed by the socio-economic analysis where the potential was divided by consumption and multiplied by energy price.the area (Fu & Rich, 1999;Fu & Rich, 2000).The model is based on static DEMs and ignores changes in vegetation or topography over time, which can have a big impact solar radiation (Fu & Rich, 1999;Fu & Rich, 2000).The main factors taken into account are the direct and diffuse components of solar radiation; the reflected radiation from the ground and other surfaces is frequently disregarded, which could lead to an underestimation of the total solar energy (Fu & Rich, 1999;Fu & Rich, 2000).The accuracy of solar path and radiation distribution models may be impacted by distortions introduced by the projection of the sky dome onto two-dimensional sunmaps and skymaps (Fu & Rich, 1999;Fu & Rich, 2000).The resolution of input DEMs and the selected sky size have a direct impact on how accurate solar radiation estimates are; coarse resolutions result in oversimplified models (Fu & Rich, 2000).The tool's suitability for large-scale analysis or real-time applications lacking significant processing capacity is limited by the high-resolution modeling and extensive study areas that need significant computational resources (Fu & Rich, 2000).The lack of historical or real-time atmospheric data in the tool limits its capacity to accurately reflect actual conditions and produces calculations of solar radiation that are less precise (Fu & Rich, 2000).However, despite these limitations, the tool can provide valid solar estimates to provide an idea of what the potential values might be.
For our analysis, the tool incorporates location through the input DEM and determines the longitude and latitude through the coordinates of the DEM.The tool also offers the ability to include building footprint as the calculation extent for generating the estimates.The Solar Analyst tool incorporates the sun's position over time, e.g. during the day and for a whole year and which year.The tool was set to calculate the radiation for a whole year, in our case the year 2021.The tool also allows for atmospheric conditions to be set through transmissivity and diffusivity factors.The default values of 0.5 and 0.3, respectively, refer to generally clear sky conditions.The tool generates an output as a raster layer with the solar radiation in Wh/m2 and our case Wh/m2/year.
The radiation intensity depends on the orientation and the tilt of the rooftop.The most optimal orientation for radiation in Europe is southfacing surfaces ( Šúri, Huld, & Dunlop, 2005).
Tilt refers to the inclination angle of the roof/PV panel.Tilt angles range from 0 • (horizontal) to 90 • (vertical) and the most optimal tilt concerning annual energy yield depends on the latitude of the installed PV system (Louwen, Schropp, van Sark, & Faaij, 2017;Šúri et al., 2005).For Europe it varies between 25 • and 40 • .The building footprint included data for the entirety of Bulgaria, which was clipped to the study area, Plovdiv.To improve spatial accuracy, the global DSM's cell sizes were resampled from their original resolution of 30x30m to 1x1m, using the Project Raster and Resample tool.For comparison, the original resolution to a smaller cell size is depicted through Fig. 2. Fig. 2 demonstrates what occurred when the datasets with 30 m resolution were run through the tool.The tool calculated the values in a pixel size of 30x30m 2 which was often larger than the building footprint area, and led to a rooftop utilization of over 120 %.It is better to use a dataset with a smaller cell size when using the Solar Analyst for regional analysis (Kausika & Van Sark, 2021).

Technical potential
The technical potential is measured by the type of solar electricity generation by PV panels that specifically are available in the Bulgarian region.Here.
polycrystalline panels are used since they are typically less costly and are better suited to hotter climates than monocrystalline panels (Sendy, 2021).A panel typically used in the region was selected for the study has a unit model size of 1650 by 990 mm and peak power of 245 Wp and a panel efficiency of 15 % (ENF Ltd, 2021).The performance ratio of the PV system (PR) can range from 0.5 to 0.9 (Reich et al., n.d.;US EPA, O, 2018) and depends on various factors.Due to shading loss in the built environment (Moraitis, Kausika, Nortier, & van Sark, 2018), a typical value of 0.75 is used in this analysis.

Calculations for annual electrical energy potential
The PV capacity installed can be calculated by multiplying total usable area with PV panel efficiency. Where: • A is the total usable area in m 2 • η is the PV efficiency: 0.15 The annual electricity generated was calculated based on the annual radiation received on the suitable rooftop area (note that we use a suitability factor of 50 % (Yang, Campana, Stridh, & Yan, 2020)), and multiplied by the PV panel efficiency and the performance ratio: Where: • G is annual irradiation in kWh/m 2 • PR is the performance ratio: 0.75 • E g is annual electric energy generated in kWh

Socio-economic potential
A factor in measuring energy poverty is tracking energy affordability (Pye et al., 2015).Energy affordability can be measured using energy consumption and energy prices (Pye et al., 2015).The Bulgarian electricity consumption was 35.5 TWh in 2019 (IEA, 2020), and with a population of 6,916,548 in 2020 (N. S. I. Bulgaria, 2021), the average electricity consumption per person in Bulgaria is assumed to be 5133 kWh per capita per year.This includes all electricity demand required for industry, offices/hospitals/schools, infrastructure, and residential areas.Plovdiv's total electricity consumption demand thus can be calculated to be 1369.19GWh/year.The national Bulgarian household electricity consumption was found to be 11.36 TWh/year (IEA, 2020).Assuming that the national ratio of 11.36 TWh/year /35.5 TWh/year = 32 % for electricity consumption also holds for Plovdiv, the actual electricity consumption demand per capita in the residential areas (in households) is 1281 kWh per capita per year.Note that total household energy demand is about 2.5 times larger due to heating demands.As for energy prices in Bulgaria, the cost for electricity is 0.09 €/kWh in 2018 (European Union, 2019).Current prices have increased slightly to 0.11 €/kWh (National Statistical Institute Bulgaria, 2022).
To understand the magnitude of the solar potential in Bulgaria, the potential was compared to consumption patterns and monetary savings of PV implementation.Therefore, two measures were created: electricity generated over electricity consumption (Eq.3), also referred to as selfsufficiency ratio (Litjens et al., 2018), and potential monetary savings from utility bills after panel installment (Eq.4): r = Eg Ec Where : (3) • E g is electric energy generated in kWh/year • E c is electricity consumption in kWh/year, which was 1281 kWh per capita per year • r is the ratio of electricity generated over electricity consumed Where: • E g is electric energy generated in kWh/year • PE is the price of electricity in €/kWh, which was 0.11 €/kWh • CE is money saved from the electricity generated by electricity prices in €

Results
The results indicated that money could be saved and steps towards energy poverty alleviation could be met by adding PV panels throughout the city.In this section, we report on the physical, technical, and socio economic potential results of each method described above.Maps showing the spatial distribution of the results from the solar radiation potential building (Fig. 4), electricity potential (Fig. 5), the monetary savings per building were shared (Fig. 6), and self-sufficiency potential (Fig. 7).Table 3 provides an overview of all results on the potentials for all the buildings.

Physical potential results
The radiation available on the buildings in Plovdiv was found to be between.700-1300 kWh/m2 yearly (see Fig. 4), with an average of 1125 kWh/m2 yearly (see Table 3), which is somewhat lower than the daily average range of 4 kWh/m2 (1460kWh/m2 yearly) reported in another paper (Gospodinova et al., 2020).The solar radiation is not normally distributed, with the majority of the radiation being right skewed.The average solar radiation was computed by calculating the average over each building's annual mean solar radiation per square meter value using: The x values represent each data building entry over the entire study area.The total solar radiation received on all buildings in Plovdiv over the year sums to 3542 GWh/m2/year (see Table 3).
Fig. 4 displays the spatial distribution of solar potential across the study area, including an inset map for the area suffering significantly from energy poverty due to its historical background.

Technical potential results
On average, PV would be able to supply nearly 160 kWh/m 2 /year, calculated using Eq.(2) noting that the annual average irradiance in Bulgaria is reported to be 1400 kWh/m 2 /year (Koleva & Mladenov, 2014).
Table 3 shows that the electric potential can consists up to 398.5 GWh/year.
(see Table 3).The spatial distribution of the annual electric generation potential per building can be seen in Fig. 5.

Socio-economic potential results
Using the socio-economic variables and methods, it was found that a PV system of 6 panels (1.470 kWp) would deliver 1241 kWh per year, which equals about the annual household electricity demand based on the per capita demand of 1281 kWh/year.This assumes no physical solar potential losses leading to technical potential loss due to tree shade or nearby buildings, which can be substantial in a dense urban area.Note, using the demand of 5133 kWh per capita per year based on total electricity demand, the PV systems will be 2.85 times larger.
The more detailed approach including social and economic factors presented here yields different results.The required energy use in buildings is between 200 and 300 kWh/m2 yearly (European Commission, 2013).
Maps such as Fig. 6 and Fig. 7 were created to interpret the spatial extent of technical potential with socio-economic variables related to energy poverty indicators.
On average, PV would be able to supply about 130 kWh/m2/year, calculated using the annual average irradiance from Table 3 and Eq. ( 2), the self-sufficiency ratio r would be between 0.65 and 0.43.
While the values shared in the previous section alone convey value and saving potential, it is not clear where the most technical potential is located to support decision-making on where to place PV panels based on poverty needs.
The technical potential is chosen to be demonstrated in terms of the selfsufficiency of buildings and is shown in Fig. 7.For most of the buildings, self-sufficiency is within the r = 0.4-21.09range, while some buildings have even r > 2184.69.The average self-sufficiency ratio is 14.6 per year (see Table 3).
The electricity generation was multiplied by the electricity costs 0.11€/kWh to see how much money is saved from paying the energy bill if the solar panels were placed.The majority of the buildings will save between 74.65 and about 3812.26 €/year, while also very high savings of 394,838.42 €/year are found.
Total savings amount to about 43.84 M€/year (see Table 3).The utility bill in Plovdiv is around 81,49 per household (Numbeo, 2024).We calculate 43.84 M Euros savings./319,612 population = 137.17euros per person.Installing solar PV systems over the entire city would save up to 168 % in people's energy bill.
Natural breaks distribution was used for the data classification for all datasets per map category.Natural breaks distribution clusters data into groups that minimize differences within a category and maximise the between group variance (Centers for Disease Control and Prevention, 2022).This means that each data point in each class falls close to each other.

Interpretation of the results and implications
The integration of solar potential mapping, including sociodemographic information associated with community needs, offers the opportunity to locally customize solar potential mapping based on variables relevant to the social environment, not only the physical environment.This represents a pivotal step in the pursuit of sustainable energy planning, and achieving SDG 7.This study aimed to develop a solar potential mapping approach that includes socioeconomic considerations by using open spatial data, and solar potential mapping to incorporate energy poverty indicators to help inform energy poverty policy design.
The results of this research show that there is a high solar potential in Plovdiv, especially located in high commercial buildings such as the football field.The city's electricity needs are met by 29 % if all the buildings from the OSM dataset have solar panels.Since this dataset was incomplete, the potential cost savings is likely higher.A building needs around 200-300 kWh/m2 yearly (European Commission, 2013) and based on the results, the solar potential on each building rooftop in Bulgaria exceeds the threshold, including areas that struggle with energy poverty, such.
as Stolipinovo.The findings listed in Table 3 imply that if a distributed solar system is in place for Plovdiv, about 92 % of all household electricity needs can be met (using per capacity electricity demand of 1281 kWh/capita/year), and about 29 % of all of Plovdiv's electricity demand (using per capacity electricity demand of 5133 kWh/capita/ year).The estimated money saved from paying utility bills is expected to be around M€43.84 per year, which is about 70,000 times the people's average monthly wages in Plovdiv.The socio-economic results help relate the potential in energy poverty indicators, demonstrating how a PV system helps alleviate energy poverty, which is helpful for policy making.The maps spatially pinpoint which buildings have the highest potential, making them the most strategic buildings to target for the PV placement first if a distributed system is aimed to be placed.
Based on these findings, we suggest that defining spatial explicit SDG The methods presented here are replicable in other cities, as these were all openly available global datasets used.This shows the possibility of using open data and off-the-shelf GIS tools for spatial analysis, and to modify them slightly to reflect the values and needs of the local demographics of the region, offering localized solutions.These methods offer ballpark estimates needed for local policy recommendations and spatial planning.
Our method clearly shows the strengths (calculations and visualizations) and the limitations (accuracy, limited spatial data, limited breadth of variables) in the use of open data for energy poverty mapping (Fremouw et al., 2020;López-Vargas et al., 2022).
More complete and accurate spatial data are needed and could be useful to address SDG 7, to illuminate and address how access to affordable renewable energy could be achieved.Even without these data, it is clear to see the cost benefit and reduction in emissions that could be gained by PV placement across any city.
The findings show that the magnitude of solar energy potential in Plovdiv is significant, which leads to the recommendation of setting up a distributed solar system throughout the city.The installation of a distributed solar system requires funding and political support, and several levers could be implemented to encourage the development.

Policy recommendations based on these results
The global community needs to address the simultaneous need for GHG emission mitigation, and a socially just and affordable energy transition.Policy strategies, such as the Green Deal in the EU and efforts such as those outlined in SDG 7, are the first steps to curb fossil fuel emissions and stimulate renewable energy use.Public financial streams can be deployed through redistributing financial fossil fuel subsidies to reduce government expenditures add value to buildings, and alleviate energy stress to the residents.Bulgaria claims not to have fossil fuel subsidies (van der Burg, Trilling, & Gençsü, 2019) but was found to have been providing illegal aid to fossil fuel power plants (Tsolova, 2017), so instead of subsidy redistribution, an aid redistribution could be useful.
The distributed energy system could be financially supported through its tax regulation mechanism combined with the Bulgarian Energy Efficiency Fund.The tax regulation mechanisms are used as incentive mechanisms for building owners to incorporate renewable energy in buildings (Naydenova, 2012).The Bulgarian Energy Efficiency Fund is offered to initiatives that improve the efficiency standard of buildings and includes the implementation of renewable energy sources (Naydenova, 2012).The electricity generated from the PV panels could also "fuel" electric heating systems, which would provide an alternative solution to conventional heating systems, thus providing benefits to the building owners (and in turn the renters).
In the Netherlands, social housing corporations are required to remove buildings with the energy label EFG from their portfolios by 2028 vanstijn2022, through demolition or (deep) renovation which will lower the energy bills for renters; similar initiatives could be beneficial in Bulgaria.The European Performance of Buildings Directive will be imperative in bringing buildings to be climate neutral (European Commission, n.d.).
Encouragement to set up a distributed solar system throughout the city should not be limited to public sector funding.Solar PV can be an interesting business case for the private sector.For example, a possible business case can be argued through a rooftop leasing system.A collaboration between the energy companies and regions that are vulnerable to energy poverty can be set up through a rooftop leasing system.An area with high energy poverty can consume the EU building average energy consumption of 200-300kWh/m2 free of charge; the excess energy generated can be delivered to the energy providers in exchange for the installation costs of the project.Dutch Companies such as Zonnepanelen.netand GroenLeven offer homeowners with high potential to lease their rooftops to energy providers in exchange for monthly monetary compensation (S.BV, 2021;GroenLeven, 2020).This research shows that there is enough potential in each building to meet the 200-300kWh/m2 mark and still profit from the remaining energy.Additionally, the private sector can develop another business case by selling the excess CO2-free energy generation as carbon credits.Lastly, the private sector should participate in the funding, by setting up a development fund that allows homeowners to take out a loan to install PV panels for their homes in return for a small interest.
Employing useful policy instruments is necessary for the development.Studies in Argentina showed that a distributed system in the residential sector is not viable with a feed-in tariff (Alhammami & An, 2021).Fortunately, Bulgaria is suitable for such a system since the country recently ended its feed-in tariff policy in 2018 (IEA, 2018).
Studies in Germany show that distributed solar energy systems work optimally with a battery storage system (Alhammami & An, 2021).Together with the electric vehicles (EV) subsidies in Bulgaria (IEA, 2019), this can encourage the beginnings of a vehicle-to-grid infrastructure.Vehicle-to-grid technology is the energy transfer between EV and the electricity grid (Guille & Gross, 2009).EVs function as storage units and feed into the grid at peak demand.Integrated solutions can have benefits if appropriately designed and economically competitive in Eastern European countries.
More data about the nature of the energy access problem for the Roma population is necessary, particularly in informal settlements (Lakeman, 2023).Special attention to populations eligible for energy benefits but often denied need (Lakeman, 2023).In Bulgaria, PV systems are considered not economically competitive because the LCOE of PV lies around 0.15€/kWh for 20GWh/year (Bódis et al., 2019).PV systems deliver electricity at a higher cost than tariffs in Bulgaria of 0.11€/kWh.Therefore, to make the distributed system succeed, Bulgaria needs to focus on removing PV barriers.The solution lies in the pricing of electricity.Renewable energy investments can become attractive if retail electricity prices increase.This is not a desirable policy suggestion in the short term since energy prices are already a burden for the consumers in Bulgaria, in the long run, it does make sense and is a logical step for countries that rely on exported energy carriers such as Bulgaria.
Future policies could also include participatory approaches such as renewable energy communities (Koltunov et al., 2023).Reaching SDG 7 in Bulgaria requires attention, support, and involvement from stakeholders (Ionescu et al., n.d.).
Finally, maps can help identify where efforts should first be made in terms of investment in green energy.Presently, mapping the indicators for SDG 7 is impossible without spatial data reporting.We recommend that spatial data be incorporated into the SDG indicators metadata to make it possible to pinpoint where interventions should be placed.Until then, proxy measures like those offered here are only possible in most data poor regions that need renewable energy the most.

Limitations of this study
There are several limitations of this study.First, having a more complete building footprint dataset is necessary for better accuracy of the solar potential of buildings.The OSM dataset missed a lot of building footprints in Stolipinovo, where energy poverty is present, as illustrated in Fig. 3.The exclusion has to do with using an open-sourced dataset that depends on volunteers to record information, but it can be due to the lack of legal standing in these homes that may prevent them from being included (Chakraborty, Wilson, Sarraf, & Jana, 2015;Rice et al., 2013).Incorporating community involvement in the OSM process could improve the building footprints and the data in the area (Chakraborty et al., 2015).To improve rooftop footprints without taking too much time, automated methods to generate more comprehensive and up-todate building footprints could improve the data completion (Lee, 2019;Touzani & Granderson, 2021).
The global DSM datasets come with limitations.The datasets were generated in large cell sizes, reducing the quality of elevation information for buildings and vegetation necessary for potential mapping, and overestimating the capacity without modifications.More accurate census data would be useful since the electricity consumption estimate had to be calculated based on the total country consumption and population size.Actual city consumption estimates were not available, and rarely are available.This distorts the advice of a distributed PV system based on the findings, since the calculations were made based on a country average, which can vary per city.These rough estimates are still valuable and show how this method could be with more accurate datasets that are available internally.Data quality and availability influence the accuracy of the maps for policymakers, who then have to make their decisions based on low accuracy estimates, or have to invest in high-resolution spatial data for research (Castellanos et al., 2017).More research is needed in comparing existing data with highresolution, and high-quality data collection for these techniques (Castellanos et al., 2017).
Additionally, none of the researchers involved were from Bulgaria or speak Bulgarian, this limited the data we were able to search for and find, influencing the results.Future research should include input from Bulgarians.

Future research
Future research could investigate the poverty rates of each neighborhood, so that policymakers can first have an idea of where the vulnerable neighborhoods lie and which buildings could benefit most from PV.The intensity of energy poverty is not measured.There are several ways to measure and map energy poverty, and future research should test new methods.
We recommend that cities who wish to improve the completion rate of building footprints to first apply machine learning techniques such as those reported in (Lee, 2019;Touzani & Granderson, 2021) These data can then be uploaded to OSM to improve future research.
This research was limited to open-sourced global DSM.Stereophotogrammetry can develop a DSM using stereo pair images (Altmaier & Kany, 2002), and it is a more affordable alternative than developing a DSM from LiDAR data or derived from drone imagery and structure from motion software.Ideally, the age of dwellings or energy labels to investigate the housing patterns could be included in mapping.
Future research could include qualitative research related to learning what users gleaned from the maps.What decisions would they make with this information?What information could be useful in policymaking and in advancing and achieving SDG 7? Usability studies of interactive maps could also be insightful.
Finally, social science research to ensure policy initiatives that aim to foster a just energy transition and do not cause further inequalities (Vela Almeida et al., 2021) are needed.

Conclusion
Policymakers need maps that are easy to understand so that they can quickly make locally appropriate decisions to meet the SDGs.Mapping solar potential in areas that struggle with energy poverty using publicly available open spatial data is a useful first step in achieving SDG 7. Using this method, mapping physical solar potential, and technical energy potential with socio-economic needs identifies the best locations for PV installation to address energy poverty.This type of mapping can be a valuable tool for reaching Sustainable Development.(Kraak et al., 2020).This locally customizable mapping method, incorporating socioeconomic factors related to energy poverty indicators, is useful for informing policymakers in decarbonization policy strategies that consider vulnerable groups of society.The results provided maps that demonstrate the optimal PV placement for the case study Plovdiv, Bulgaria.The total potential can supply 92 % of the city's household electricity demand and save about M€43.84 in energy bills.It shows that incorporating socio-economic factors in mapping is beneficial for visualizing the benefits of a solar system for future policy strategies in energy poverty.This spatial planning method could be useful and replicated in other areas of the world suffering from energy poverty and data poverty.
Here we offer new representations of energy poverty to inform policy We offer policy recommendations to include renters and social justice energy transition to alleviate data poverty and energy poverty, using offthe-shelf GIS methods.The series of maps that were made using this method can inform localized decision-making about where to best place PV to be most beneficial to cover costs while meeting energy needs.
Effectively addressing energy poverty through policy and other means requires representative and accurate data (Sareen et al., 2020).Spatial tools to support planning are needed to reach universal clean energy access (Szabó et al., 2013).This methodology, incorporating socio-economic variables and open spatial data, can be replicated in other areas of the world suffering the most from energy poverty.This tool can help achieve SDG 7 and specifically SDG Target 7.1: by 2030, and ensure universal access to affordable, reliable, and modern energy services.Despite the limitations of available open data, important findings were generated which can be very helpful to inform policy and effective investment planning.This method for informed spatial planning could help policymakers make decisions that will alleviate the burden of energy poverty for the people who suffer the most.These methods of mapping can be replicated in any city, to map for a sustainable world.

Declaration of competing interest
None.

G
.Koster et al.Hong et al.(Hong et al., 2017) which included the calculation of physical potential and technical potential, we expanded this to include the socio-economic potential analysis.The physical potential in this research refers to usable solar radiation.The technical potential refers to electricity generation, and the socio-economic potential is based on the technical potential related to the energy poverty indicators: consumption and electricity price.The socio-economic potential analysis is calculated through an energy affordability metric using the energy poverty indicators: consumption and electricity price(Pye et al., 2015).To calculate each, spatial data were needed.The data quality and availability determine what is possible in terms of analysis and accuracy and precision of results.Data required included an elevation model, building footprints, and census data: retail energy prices and city-level energy consumption per citizen or household.Openly available GIS data in Bulgaria are limited.As a consequence, global GIS datasets were used.The Global Administrative Areas (GADM) 2011 was used as the administrative boundary dataset for Plovdiv (G. A. A. (GADM), 2011).The dataset came as a vector polygon with an area of 82 km2 in the coordinate system WGS 1984, which is the most universally used coordinate system on the web.For the determination of the elevation, the global open data Copernicus Digital Elevation Model (DEM) (E. S. Agency, 2021) was used.

Fig. 2 .
Fig. 2. Solar Analyst results of radiation estimates using the 30 m cell size on the left and 5 m cell size on the right.A rough measurement of the pixel size was made and is presented in the measurement content panel.

Fig. 3 .
Fig. 3.The OSM (O.contributors, Bulgaria, 2021) building footprint dataset outlined in red is projected over a satellite image of ESRI to highlight that many of the buildings are not included in the dataset in the neighborhood Stolipinovo, Plovdiv.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4 .
Fig. 4. Average annual solar radiation potential values for rooftops in Plovdiv in megawatthours.Beige indicates the lowest potential and dark red is the highest potential.Below is the legend.On the top right, the inset map features Stolipinovo, an area that struggles with energy poverty.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5 .
Fig. 5. Average yearly electricity generation potential values for rooftops in megawatt-hours.Dark blue indicates the highest potential and light yellow the lowest potential.On the right, a close up of Stolipinovo is provided, an area that struggles with energy poverty.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6 .
Fig. 6.Legend presents a ratio of the number of times the building's electricity generation can supply the city's electricity consumption per capita which is 1281 kWh per person per year.Purple indicates the lowest potential and green the highest potential.On the top right, an inset map of Stolipinovo is provided, an area that struggles with energy poverty.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 7 .
Fig. 7. Money saved avoiding paying for electricity bills in € per year for the entire city using the electricity prices 0.11€/kWh.Light green indicates the lowest potential and dark green the highest potential.On the top right, an inset map of Stolipinovo is provided, an area that struggles with energy poverty.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 2
Summary of characteristics of energy poverty in Bulgaria.

Table 3
Total potential in physical (radiation), technical (electricity), social-economic (demand and monetary savings) for Plovdiv.