*2.1. EU Policy for RES*

The European Union, by developing numerous binding policies, is the biggest promoter of renewable energy investments. Within the European Union policies, RES and its support are strongly anchored in the form of strategies, development goals, priorities and current legislation [16]. The principal initiative to promote renewable energy at the European Union level was launched in 1997, when the European Council and the European Parliament adopted the "White Paper for a Community Strategy and Action Plan", aiming to increase the share of renewable energy, which at that time accounted only 6% of gross energy consumption [17].

EU policy standards for the use of RES have been acquired especially in the last two decades. Firstly, the EU Sustainable Development Strategy [18] stipulated that 12% of energy consumption and 21% of electricity consumption should be covered by RES by 2010, with an increase to 15% by 2015. The Europe 2020 Strategy [19], on the other hand, has indicated policy standards towards increase of the RES usage and to promote energy efficiency and greater energy security.

Directive 2009/28/EC of the European Parliament and of the Council (2009) [20] set a mandatory target by 2020 at the level of a 20% RES energy share. The Directive also sets out various mechanisms that Member States can implement to achieve their goals (joint projects, support schemes, cooperation between Member States and other countries). Moreover, it also defined national renewable energy targets for each country, taking into account its starting point and overall renewable energy potential. These goals range from a low of 10% in Malta to a high of 49% in Sweden. The EU countries define how they plan to meet these objectives and all their renewable energy policies in their National Plans. The results in achieving the national goals are measured every two years when EU countries publish national renewable energy progress reports [21].

The Article 3 of Directive 2018/2001 (2018) [22] on the promotion of the use of energy from renewable sources established a new and binding overall EU target for the total share of energy from renewable sources in the Union's gross final energy consumption for 2030 of at least 32%.

The Green Deal for Europe, proposed at the end of 2019, creates new conditions for a very ambitious climate protection agenda. The vision is for Europe to become the world's first climate-neutral continent by 2050. The package of measures in the European Green Deal should enable European citizens and businesses to reap the benefits of a sustainable green transition. The measures, together with an initial 'roadmap' of key policies, cover tasks such as reducing emissions, investing in cutting-edge research and innovation, together with protecting Europe's environment [23]. The implementation of the Green Deal for Europe idea will let EU countries achieve energy independence, which may have a significant impact on the energy market as well as on their regional policies. It is also assumed that the achieved climate neutrality will contribute to the dynamic development of the economy and improve its competitiveness [24].

The support accompanying implementation the abovementioned policies in EU countries resulted in increased the renewable energy use throughout the Community. In the EU-28 final energy consumption, renewable energy sources accounted for 3.8% in 1990, 5.1%, while in 2019 this percentage increased by a further 5.1 p.p. to 10.2%. A relatively high share of RES consumption is characterized by the household sector, where the share of RES in household final energy consumption in 1990 was 8.5%, in 2004—10.8%, and 16 years later by 7.4 p.p. more. [7].

The construction sector in Europe accounts for 40% of energy consumption and 36% of CO2 emissions [25]. Due to the estimated, high energy saving potential of the housing construction sector, the European Union has established a policy framework focused on reducing energy consumption in buildings, consisting of different policy actions, i.e., Energy Performance of Buildings Directive (EPBD) [26], Energy Efficiency Directive (EED) [27], Ecodesign Directive [28], Energy Labelling Regulation [29] and the mentioned Renewable Energy Directive (RED) [20].

In summary, the EU countries' policies, since the 1990s, have systematically increased the importance and role of RES in the energy sector structure s [30]. All plans for the presumed targets are presented in the National renewable energy action plans 2020 [31]. National RES energy targets for individual countries are established at various levels and the process of reaching them is different depending on the country [24], which also affects the use of RES in the household sector.

### *2.2. The Role of Authorities in Promoting New Energy Solutions*

The EU countries are trying to influence and propose different solutions to improve the situation regarding energy consumption. One of the more effective ways is to exploit renewable energy obtained from natural resources such as wind, sunlight, geothermal heat, etc. instead of non-renewable energy [32].

The consumption of renewable energy is growing in importance and there is an increasing need to encourage households to cooperate. For the development of renewable energy, it is important to ensure an active role of governmen<sup>t</sup> and other authorities at all levels of general interest. It is even suggested [33] that public institutions, both state and municipal, should be legally obliged to install solar systems on the roofs of the buildings where they hold office. Also, electricity buyers can have an important influence on the way energy is generated in public procurement, where RES energy should be given priority.

However, not all countries have a good appreciation of governmen<sup>t</sup> policies to promote RES. For example, in Croatia it is found that citizens are not actively encouraged to participate in investments that would largely benefit the environment [33]. A study in Malta [34] concluded that future programs promoted by the governmen<sup>t</sup> should consider the role of pro-(and anti-)government sentiment in predicting their adoption in the initial stages. Strong pro-governmen<sup>t</sup> sentiment can strengthen citizen initiatives to install new RES technologies. Delegating RES promotion to municipalities or even commercial entities may also result in increased citizen interest in such installations. The financial constraints faced by low-income households are also pointed out [34]. RES technologies are often expensive and the way in which support schemes are implemented require upfront investment, leaving households unable to afford to pay (in cash) for any investments. It is suggested that subsidies and programs offered by the government, should offer staggered payments for the initial investment in order to provide an incentive. Furthermore, the experience in Malta [34] emerged that requiring consumers to pay for net rather than gross value would enable more households to benefit from the program. Support schemes should also pay attention to vulnerable groups of society. Helping elderly households (for example, through preferential feed-in tariffs or targeted communication) could unlock further potential by encouraging older people to engage in investment. Similarly, significant scope appears to encourage investment in rental housing. Some programs can be designed to promote investment agreements between landowners and tenants [34].

Local authorities should assume a central role and responsibility in the task of solarizing their territories. They have the autonomy to regulate the situation on-the-spot, in particular through their well-known water and wastewater utilities, which can also provide other RES for local energy production (biogas from bio-waste and sewage sludge, energy stored in water). Such companies would therefore integrate power generation into their regular activities and could provide installation and maintenance services for power generation systems installed in their area of competence [33].

### *2.3. Willingness to Involving in the RES Use by the Household Sector*

Energy resources have always played a key role in human life. Sufficient energy resources influence economic and social development. A kind of interdependence is observed between technological development, energy consumption and world population growth [35]. Providing adequate energy resources for entities such as households is also about meeting basic social needs. Energy is one of the main categories of consumer expenditure in households [36]. Maintaining adequate thermal comfort affects the consumer life quality [37,38]. On the other hand, improving this kind of comfort is associated with an increase in the consumption of fossil energy carriers, which in turn is associated with an increase in environmental pollution [39,40]. Thus, in order to reconcile social and environmental objectives, it is important to widely involve households in initiatives for the use of energy from renewable energy sources.

An important issue is the problem of consumers' attitude and propensity to make decisions on the use of renewable energy technologies. As indicated by Ropuszy ´nska-Surma and W˛eglarz [41], social acceptance of RES technologies is important for their development and should be taken into account when shaping policies for sustainable development in the region.

The results of research by A. Jacksohn, P. Grösche, K. Rehdanz and C. Schröder [42] sugges<sup>t</sup> that households tend to act fairly rationally in the sense that investors consider the costs and benefits of their decision. Since economic factors influence the decision to invest in a renewable energy system, policy makers can provide reasonable financial incentives to steer households in the desired direction.

A number of studies show that financial incentives become a strong motivation for households to switch to renewables. A study of Italian and Austrian households investing in solar PV found that higher financial support was more likely to attract younger and less educated people, as well as those with an anthropocentric attitude towards nature [43].

In contrast, Wasi and Carson [44] investigated how households' decisions to switch to more environmentally friendly water heaters changed with the introduction of a rebate scheme for hot water systems. They concluded that the likelihood of households choosing to use the aforementioned renewable system for hot water, or a heat pump, increased significantly after the introduction of a scheme to financially support these initiatives. Furthermore, the impact of this rebate policy varied with household income, education, access to the gas grid, hot water consumption and expectations of future electricity prices.

In Germany, research on household investment in RES found that the propensity to adopt them was influenced by housing characteristics, household energy consumption and geographical factors, while most socio-demographic variables were found to be insignificant [45]. Other studies [46,47] have considered the motives for adopting an innovative residential heating system based on renewable energy. It was revealed [46] that the influence of socio-demographic, housing and spatial characteristics was more significant for households replacing a heating system in an existing house than for households choosing a heating system for a newly built house.

The likelihood of investing in RES also increases with environmental concern, income, number of children and solar radiation intensity [48]. Men and well-educated people were more willing to engage in RES installations than older people.

In Poland, social acceptance of RES varies, and depends on age, gender, education, income and type of building inhabited. The groups that showed the highest acceptance for RES installation were men, people aged 30–49 years, having secondary technical education, low income and people living in a single-family house. The rationale for installing RES was the expected long-term savings, while the biggest barrier was the lack of financial resources. The financial aspect is crucial for the installation of RES in Polish households. In the case of prosumers, besides the financial aspect, were also pointed out technical possibilities, unclear regulations, complicated grid connection process and lack of knowledge. Thus, besides financial support, additional support in the form of consumer education, promotion of RES development, technical and legal support of potential prosumers seems to be necessary [49].

D. Štreimikiene and A. Baležentis [ ˙ 50] studying households in Lithuania paid attention to employment status and income level that have a significant impact on the willingness to purchase renewable energy sources in households of this country. The self-employed showed the highest willingness to purchase RES. Private sector employees and, surprisingly, pensioners also showed more willingness to invest in RES compared to other social groups. In terms of education level, only respondents with higher education showed a higher willingness to buy RES than the rest. Thus, the willingness to pay for RES in Lithuanian households was determined by factors such as awareness of their existence, education level and income. Other studies showed that the willingness to pay for RES was influenced by age, gender, education, income, price, geographical place of residence. In contrast, membership in environmental organizations, race, political views and perceived health effects had less influence on household usage of renewable energy technologies [51].

A study conducted in Malta [32] revealed that factors associated with the use of RES energy (in this case, energy from photovoltaic devices) were the age of those forming households (the younger the individuals, the higher the involvement in RES) and unemployment (if present, there were fewer opportunities to invest in RES), both of which sugges<sup>t</sup> that financial motives and constraints are crucial to household uptake of RES initiatives.

In contrast, a study by Luttenberger [33] for Croatia, showed that people support for RES projects increased together with knowledge about them. Positive attitudes towards new clean technologies prevail in this country. However, there is a general lack of solid information and understanding of the concepts that are necessary for the possible benefits for end-users in a household. There is also a lack of professional and trained human resources for renewable energy issues, there are no relevant courses at universities and colleges, no systematic research, a lack of experience of local companies in organizing projects and the volume of theoretical knowledge about RES and practical capabilities involved is limited [52]. Even though household purchasing power is modest, their owners are at least declaratively putting aside more money for RES, but only after sufficient additional information has been provided. Croatians also mention obstacles such as national solar quotas, administrative barriers and complexity of the procedure [53].

Summarizing the abovementioned studies, it can be pointed out that the financial motive is an essential motive for household members to undertake RES installations. On the other hand, educational level and age are indicated among the important socio-economic characteristics of those willing to accept RES. Providing consumers with extensive information and educating them on new technologies for the overall social and environmental good is also an important determinant of consumers' commitment to RES application. So, the impacts of socioeconomic factors provide substantial policy implications for the design of green electricity programs [54].

### *2.4. Application of Cluster Analysis in Comparative Research on RES Use in Different Countries*

Cluster analysis was first introduced in the work of R.C. Tryon [55]. It is a useful tool for exploratory data analysis that aims to arrange individual objects into groups so as to acquire objects within the same group most similar to each other and the objects between other groups most dissimilar to each other [56,57]. This analysis, using several different classification algorithms, detects the data structures without explaining why they occur.

European Union member states vary in terms the exploited both of total energy sources [58] and also of RES. The cluster analysis method may be useful for searching the similarities between individual member states. This method is often represented by the simple hierarchical method. The common feature of the stepwise algorithms used in this method is the clustering by combining smaller clusters, created in the previous steps of the algorithm. The basis of all algorithms of this method is the appropriate determination of the measure for object dissimilarity [59].

Below are presented a few selected studies reporting on the results produced by cluster analysis performed for country classifications regarding RES market. Therefore, for example, the study by Bluszcz, Manowska [58] is applying the agglomeration procedure, which results in the division of European Union member states into clusters according to their similarity, with regard their energy markets. The research results constitute an interesting study that could potentially provide a model for the creation of so-called regional energy markets in the transitional integration phase. The above-mentioned authors chose the following diagnostic variables for the analysis: consumption of electric energy which is generated from renewables per capita (TWH per person), consumption of hard coal (million ton per person), emissions of greenhouse gas per capita, available for final consumption (Gigawatt-hour per person), final energy consumption (thousand ton of oil equivalent (TOE) per person), petroleum available for final consumption (Gigawatthour), natural gas (Terajoule gross calorific value—GCV) per person, energy intensity of GDP (kilograms of oil equivalent (KGOE) per thousand euro), import dependency (%). The paper distinguishes six clusters, consisting of countries with similar levels of energy system development. The cluster formed by Finland and Sweden presented renewable energy production at the highest level in the EU. Finland had slightly higher greenhouse gas emissions per capita in the energy mix and a higher energy consumption factor than Sweden, as solid fuel use accounted for 9%. Luxembourg formed a one-element cluster. This country had the highest level of electricity consumption per capita compared to other UE member states. Greenhouse gas (GHG) emissions per capita and energy dependency levels were also the highest there. Another cluster, comprising France, Slovenia, United

Kingdom, Latvia and Romania had levels of electricity consumption per capita below the EU average. Energy dependency levels were relatively low in this cluster, ranging from 50% in Slovenia to 24% in Romania. The countries grouped in the cluster comprising Greece, Lithuania, Croatia, Hungary, Spain, Portugal and Italy were characterized by high levels of energy dependency, ranging from 52% for Croatia and 58% for Hungary to over 70% for all other countries in this cluster. In these countries, the contribution of solid fuels to electricity production was significantly higher than in the first and second clusters. The cluster including countries such as the Czech Republic, Slovakia, Poland, Bulgaria and Estonia was characterized by a level of greenhouse gas emissions per capita close to the average and thus exceeded the desired GHG emission factors. For this cluster, renewable energy consumption per capita was one of the lowest in the EU. The level of energy dependency was low, less than 1% in Estonia, 36% in Bulgaria and Czech Republic and 44% in Poland (with the exception of Slovenia, at 63%, the highest in this cluster). The last cluster containing the following countries: Netherlands, Germany, Belgium, Ireland, Denmark and Austria had the lowest level of energy intensity, while the consumption of energy from renewable sources was high [58].

Other researchers [60] analyzed in their paper the renewable energy sector in European Union countries. The k-means clustering method was used for grouping of countries. This method is widely used in various areas of science for the data analysis. The advantage of this method is the intuitiveness and simplicity of the basic calculation idea. In the k-means method, the distances between objects are determined by the Euclidean distance or its square (the peculiarity of the algorithm makes the results in both cases the same). The k-means algorithm can be described in three points: 1. The starting point is the division of a given set of objects into k subsets (usually generated by assigning each element to the "closest" preselected representative of the k groups). 2. For each group, the centers of gravity in the space of diagnostic variables are determined. 3. Each element is assigned to the nearest center of gravity, and then it is necessary to return to step 2, if at least one element has been moved to another group [61]. The algorithm of the k-means method can be regarded as a kind of "inverse" of the analysis of variance. It is helpful in finding a division of the studied community into k groups, so as to maximize the intergroup variance and, consequently, the F-statistic [61,62].

In the abovementioned paper, Parobek and colleagues [60] created nine clusters and the following diagnostic variables were selected forest cover, roundwood production, primary energy consumption, primary production of energy from renewable resources, share of renewable in gross final energy consumption, greenhouse gas emissions, employment, gross value added, GDP growth rate, expenditures on R&D, price of electricity, energy dependence. The 1st cluster was formed by the leaders in the use of renewable energy sources and the average production of primary energy from biomass, i.e., Austria, Portugal, Sweden, Finland. However, these countries were below the EU average in terms of employment rate and energy dependence. The 2nd cluster, consisting of Hungary and Belgium, had the use of renewable energy sources below the average. The 3rd cluster that contained Greece and Romania had a utilisation of renewable energy sources slightly above the EU average. These countries had significant production of wind, solar and hydropower energy. The 4th cluster, consisting of Bulgaria, Lithuania, Czech Rep., Denmark, Ireland and Slovakia had a medium level of renewable energy use. The 5th cluster was formed by Cyprus, Estonia, Latvia, Slovenia, Luxembourg and Malta. In these countries the share of renewable energy sources remained above average, but on the other hand they their production of primary energy from renewable resources was insignificant. Germany formed a one-element cluster (the 6th cluster) with the highest production of primary energy from biomass and RES consumption below average. It was also noted that this country is one of the largest producers and users of wood. Another one-piece cluster formed France (the 8th cluster), which (just as Germany) was a leader in the production of energy from wood and had a relatively low level of greenhouse gas emissions compared to the countries in the other clusters. The 7th cluster included Netherlands, Poland and United Kingdom, i.e.,

the countries having lower primary energy production from renewable resources and the lowest share of use of renewable resources. Italy and Spain formed the last, the 9th cluster. Use of renewable energy resources in these countries was below the average, but primary energy production from renewable resources—above the average [60].

In turn, K. Chudy-Laskowska and co-authors [63] in their article distinguished clusters on the basis of similarity such EU countries regarding the current level of the wind energy development. The research applied Ward's analysis and Wroclaw taxonomic methods. The Wroclaw method was described and employed in the mentioned article and in other studies [64,65]. Seven clusters containing from 1 to 7 countries were obtained. Denmark, forming one-element cluster, turned out to be the wind energy leader in the EU. This country had not only very profitable use of wind energy but also the highest rates of wind turbine electricity generation, and also the share of wind energy in gross inland energy consumption. The following diagnostic variables were adopted: wind farms per 100 thous. People, number of turbines per one wind farm, renewable (wind offshore) electricity capacity (MW) per thous. people, renewable (wind onshore) electricity capacity (MW) per thous. people, wind cumulative capacity growth rate, renewable (wind) electricity generation (GWh) per thous. people, share of renewables in gross inland energy consumption of which: wind power. The cluster comprising Spain, Portugal, Ireland, Sweden and Germany also had excellent wind energy technologies. These countries have been leaders in the introduction and development of wind energy for many years and have had a significant number of wind farms (above the European average) and these farms were quite large. Finland is another one-element cluster. This country invests and develops the mentioned branch of renewable energy. The index on the number of wind farms was above the world average, so there were more these farms than the EU average, but they were not large. Three countries: the United Kingdom, the Netherlands and Belgium, forming another cluster, were relatively strong in terms of both wind energy potential and the level of its development. There was also an upward trend in wind energy resources in this group. In the countries forming cluster that included Cyprus, Romania, Greece, Italy and Estonia there were not many wind farms, but the existing were relatively large. In addition, all wind energy indices were below the EU average. Another cluster includes six countries: Austria, Luxembourg, Lithuania, Croatia, Poland and France. In these countries there was a chance to change their position for the better by possible investing in wind energy. However, during the analyzed period, the level of wind energy development was still below the world average there. The last and worst group in terms of both wind energy potential and development level was a cluster including seven countries: Slovakia, Slovenia, Malta, Czech Republic, Hungary, Latvia and Bulgaria. This cluster also included Malta, which did not have a single wind farm, and the other countries in this group had almost no wind farms. These countries invested in other renewable energy sources. Slovakia, Latvia, Hungary and Bulgaria invested mainly in hydropower energy, while Hungary and Malta in solar energy [63].

Overall, the methods k-means and Ward were the most frequently applied clustering techniques. Other studies using cluster analysis in research on the use of renewable energy are [66–69].
