*Article* **Use of Renewable Energy Sources in the European Union and the Visegrad Group Countries—Results of Cluster Analysis**

**Elzbieta Kacperska ˙ 1 , Katarzyna Łukasiewicz <sup>2</sup> and Piotr Pietrzak 2, \***

1


**Abstract:** Increasing the use of renewable energy sources is one of the strategic objectives of the European Union. In this regard, it seems necessary to answer the question: which of the member countries are the most effective in its implementation? Therefore, the main goal was to distinguish groups of European Union countries, including the Visegrad Group, differing in the use of renewable energy sources in transport, electricity, heating and cooling (based on cluster analysis). All members of the EU were determinedly selected for research on 1 February 2020 (27 countries). The research period embraced the years 2009–2019. The sources of materials were the literature on the topic and data from Eurostat. Descriptive, tabular, graphical methods and cluster analysis were used in the presentation and analysis of materials. In 2019 wind and hydro power accounted for two-thirds of the total electricity generated from renewable sources. In 2019, renewable energy sources made up 34% of gross electricity consumption in the EU-27. Wind and hydro power accounted for two-thirds of the total electricity generated from renewable sources (35% each). Moreover, it was determined that there were 5 clusters that differed in their use of renewable energy sources. The highest average renewable energy consumption in transport, heating and cooling in 2019 was characterized by a cluster consisting of Sweden and Finland. In contrast, the highest average renewable energy consumption in electricity was characterized by a cluster consisting of countries such as: Austria, Croatia, Denmark, Latvia and Portugal. Finally, in a group that included countries such as Belgium, France, Luxembourg, Malta, the Netherlands and the entire VG (Hungary, Czechia, Slovakia and Poland), renewable energy consumption rates (in transport, electricity, heating and cooling) were lower than the EU average (27 countries).

**Keywords:** sustainability; renewable energy sources; European Union; Visegrad Group; cluster analysis

#### **1. Introduction**

During the last three decades, the fashionable concept in environmental discourse has been "sustainable development" (SD). "It has spawned a vast literature and has strengthened the arm of empire builders in many research institutes, Universities, national and international bureaucracies and statistical offices" [1] (p. 191). SD is also a fundamental and overarching objective of the European Union (EU), enshrined in Article 3 of the Treaty on EU. Since 2005 Eurostat has regularly produced biennial monitoring reports of the EU Sustainable Development Strategy (EU SDS), based on the EU set of Sustainable Development Indicators (SDIs).

The concept of SD has also been constantly criticized, mostly due to the inconsistency of mixing economic expansion and natural system preservation in one concept [2]. It was also mentioned that "there is no agreement on a comprehensive sustainable development theory, there are different contested theoretical approaches and definitions" [3] (p. 468). Nonetheless, the scientific community has agreed that SD is governed by a dynamic

**Citation:** Kacperska, E.; Łukasiewicz, K.; Pietrzak, P. Use of Renewable Energy Sources in the European Union and the Visegrad Group Countries—Results of Cluster Analysis. *Energies* **2021**, *14*, 5680. https://doi.org/10.3390/en14185680

Academic Editor: Eduardo Álvarez Álvarez

Received: 4 August 2021 Accepted: 7 September 2021 Published: 9 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

balance between the three pillars of civilization's progress: (1) economic, (2) social, and (3) environmental [4]. '

Nowadays, there is a growing emphasis on the importance of applying the concept of SD to the energy sector [5]. Therefore, the term sustainable energy development (SED) is increasingly used in the literature [6]. SED is defined by the International Atomic Energy Agency (IAEA) as "the provision of adequate energy services at affordable cost in a secure and environmentally benign manner, in conformity with social and economic development needs" [7]. Figure 1 depicts the relationship between the three dimensions SD and energy as illustrated by the IEA/IAEA [7]. ergy Agency (IAEA) as "the provision of adequate energy services at affordable cost in a velopment needs" [7]. Figure 1 depicts the relationship between the three dimensions SD

**Figure 1.** Interrelationship among sustainability dimensions of the energy sector.

this area is the "Clean Energy for All Europeans" [8]. In May 2019, the EU completed the pletion of the Energy Union. The package includes "documents on energy efficiency (…) ergy and climate goals" [9]. SED is also one of the priorities of the EU. One of the most important initiatives in this area is the "Clean Energy for All Europeans" [8]. In May 2019, the EU completed the final legislative acts of this package, thus reaching an important stage towards the completion of the Energy Union. The package includes "documents on energy efficiency ( . . . ) new energy and climate laws, consumer rights, energy security, electricity market efficiency, and cooperation between the EU and Member States to achieve the ambitious energy and climate goals" [9].

According to the package, the EU is to become a world leader in the use of renewable energy sources (e.g., biomass energy, hydropower, geothermal power, wind energy, and solar energy). Thus, it seems necessary to answer the question of which member countries are the most efficient in the use of renewable energy sources? That is why the main goal of this article is to distinguish groups of EU countries, including the VG, differing in the use of renewable energy sources in transport, electricity, heating and cooling based on cluster analyses. Through its implementation, it will be possible to identify the countries that are most committed to the use of renewable energy sources and, thus, the countries that most effectively implement the concept of SED. The following set of research tasks was adopted for its implementation: conduct a critical review of the literature on SD; show the changes in the use of renewable energy in transport, electricity heating and cooling in EU member states (including VG countries) from 2009 to 2019; show the structure of utilization of renewable energy sources in EU member states; identify leaders among EU member states in the development of the renewable energy sector.

The remainder of the article is structured as follows. The next section provides a brief description of the methodological approach and is followed by the literature review. The article ends with discussion and some concluding remarks.

#### **2. Materials and Methods**

All members of the EU were selected for research on 1 February 2020 (27 countries). The research period covers the years 2009–2019. In 2009, the European Parliament adopted the Directive 2009/28/EC [10]. It established a common framework for the use of energy from renewable sources in order to limit greenhouse gas emissions and to promote cleaner transport. The last year in which there were complete data needed to carry out the research using the assumed research methods at the time of the research was 2019. The sources of materials were the literature on the subject and also data from Eurostat (share of renewable energy in transport, share of renewable energy sources in electricity, share of renewable energy sources in heating and cooling). The use of Eurostat data made it possible to compare all EU countries.

Descriptive, tabular and graphical methods and cluster analysis were used for the presentation and analysis of materials.

In the first stage of the research, the changes in the use of renewable energy sources in the EU and VG countries were presented. The analysis includes the shares of renewable energy in transport, electricity, heating and cooling.

In the second stage, based on 2019 data, the cluster analysis was conducted. The term "cluster analysis" was coined by Tryon [11] and then further developed by Cattell [12], and the use of cluster methods has increased significantly over the past 30 years [13]. Cluster analysis is the set of multivariate techniques whose main aim is to aggregate items, objects or individuals (here: EU and VG countries) based on their characteristics [14]. The basic criteria used to group objects is their similarities. In this manner, objects belonging to the same cluster are similar to each other concerning the variables that were measured in them, and the elements of distinct clusters are dissimilar for these same variables [15].

Clustering techniques are classified into two types: agglomerative and divisive. In this research the authors used Ward's method, which is one of the most frequently employed agglomerative clustering method. The characteristic feature of this method is the use of a variance analysis for the purpose of determining the distance between clusters. The distance between one cluster composed of objects and another one cannot be directly expressed by way of the distance between the objects belonging to these clusters [16]. Hence, "the method aims to minimize the sum of squared deviations of any two clusters which can be formed at any stage" [17] (p. 54). Therefore, clusters that "ensure the minimum sum of squared distances from the centre of mass of a new cluster, which they create" are merged [18] (p. 74). The literature points out that this kind of agglomerative method is cognitively effective; however, it yields small and yet most natural clusters. In this paper, the measure of similarity used was the squared Euclidean distance.

#### **3. Literature Review**

#### *3.1. The Concept of Sustainable Development*

The concept of SD has been developed in response to serious concerns over the potential of the Earth's global ecosystem to sustain the impact of anthropo-pressure. It has been aimed at the preventive elimination or at least reduction of the imbalance between economic growth and social development as well as socio-economic development and the natural environment [19]. The concept of SD was introduced to the globally used terminology by the United Nations (UN) agencies [20,21]. This term was used extensively for the first time at the UN Conference on the Human Environment in 1972. It stemmed from the original concept of sustainable management of natural resources. It was defined as a strategy aiming at development based on the rational utilization of local resources and knowledge gained by farmers to satisfy the needs of remote rural areas in Third World countries [22].

The concept of SD is defined as an interdisciplinary approach, which covers in its scope the environmental (the natural capital), social (social capital) and economic spheres (the economic capital). It is an idea and at the same time a concept of actions leading to changes in the life of the human population in the 21st century to ensure adequate living conditions for the present and future generations, as well as the potential to satisfy their needs [23].

It may be assumed that the idea of SD is a certain compromise between the concepts for several component capitals of the natural, social and economic development. It needs to be indicated here that the term SD in terms of economic sciences stems from the development economics, which comprises both neoclassical theories (theories based on the linear model of economic growth, based on the two-sector and bipolar character of global economy) and theories which stress the problem of responsibility in the context of planned and realized economic development [24,25]

The concept of SD is mainly considered within the framework of three approaches [26]: (1) the socio-philosophical concept (assuming the need for changes in the system of human values), (2) a modern direction of economic development (assuming new economic organization and management methods), (3) a newly developed discipline of science.

Such studies as those by Górka [27,28] have attempted to standardize the terminology related to the discussed concept. It should be noted that sometimes, wrongly, the term sustainable is replaced by balanced. However, the state of lasting balance is not consistent with the essence of this concept. This may lead to economic stabilization or even retrogression [27].

Pirages [29] was of an opinion that SD refers to economic growth, which is sustained by the natural and social environment. In turn, Goodland and Ledec [30] stressed that SD is a process of economic transformation consisting in the optimization of current economic and social benefits without jeopardizing the potential to attain these benefits in the future. Turner [31] presented an opinion that SD requires maximization of net benefits of economic growth in order to maintain accessibility of environmental services and the quality of natural resources over time.

It should be noted that Pearce et al. [32] were of an opinion that SD includes the formation of the socio-economic system, which sustains the following objectives: growth of real income, improvement of educational standards, health and the quality of life. In turn, Górka et al. [33] defined SD as such a course of economic development, which does not significantly or irreversibly disturb the living environment for humans, while respecting the laws of nature and economics.

In the opinion of Runowski [34] SD consists in efforts to attain balance between various goals of socio-economic development, without which sustainability of the system may be difficult to attain. The primary aim is to ensure lasting development in terms of its stability and continuity. SD provides guidelines for sustainability as a goal to be reached. In turn, Giovannini and Linster [35] stated that the concept of SD refers both to the quality and volume of economic growth and combines three dimensions of welfare: economic, social and natural. Borys [36] defined SD as an integrated order, i.e., a certain game of limitations in the use of all capitals.

Holger [37] was of an opinion that SD strives to define such management conditions which might guarantee sufficiently high ecological, economic and socio-cultural standards to the entire presently living human population and to the future generations while observing tolerance of nature and realizing inter- and intragenerational justice. In turn, Stanny and Czarnecki [38] expressed an opinion that SD is a compromise between environmental, social and economic goals determining the welfare of future generations. The economic aspect refers not only to the satisfaction of the present-day needs, but also preservation of resources required to meet the needs of future generations. The social aspect is connected

with education and the potential to attain the capacity to solve major social problems and to participate in development process of the entire system. Finally, the ecological aspect refers to the identification of absolute limits to human activity.

SD is a concept fully referring to the entire scope of human activity and the resulting interactions with the environment. It may be considered to be a certain type of socioeconomic development, which in view of the changes occurring on Earth needs to be constantly monitored and analyzed.

One of the main principles of SD is the use of renewable energy sources. Therefore, it is to them that the next part of the article will be devoted.

#### *3.2. The Development of Renewable Energy Sources in the Entire EU and the VG Countries*

Energy generated from renewable sources constitutes an important element in the strategy for SD of the EU member countries, including the VG. Public authorities in the EU have adopted the assumptions of SD for the power industry sector, defining them as an efficient use of energy, human, economic and natural resources [39]. This results from the rapid economic development, a continuous increase in energy demand as well as the awareness that global traditional energy resources are limited [40–42]. The concept of SD emphasizes the importance of environmental protection and repletion of renewable resources, which is particularly essential under new conditions observed globally [43]. In view of the above, SD is such an activity, which sustains the natural environment and may not be conducted at the expense of future generations [44,45]. The concept of SD is based on humans as subjects having an impact on the environment, our planet as an area (object) of human impact and partnership as a method of integrated activity [8]. The global actions towards SD need to ensure welfare and peace worldwide. Such foundations were also presented in the UN Resolution "Transforming our world: the 2030 Agenda for Sustainable Development", adopted in September 2015 [46]. The global initiative for SD points to climate change and problems of renewable energy [47]. The 17 global sustainable development goals (SDGs) include energy issues, e.g., SDG7 indicates access to cheap, clean, reliable, technologically advanced and sustainable energy for all people by 2030 [8,48]. This is to be attained by [46]:


The EU has also played a significant role in the development of the 2030 Agenda, which is fully consistent with the European vision and constitutes a global program for actions for SD on the global scale, based on the SDGs. For many years, the EU has been undertaking actions for SD in the power sector. Since the beginning, the energy sector has been the most important aspect of the integration processes in Europe. The establishment of the European Coal and Steel Community (CSC) and the European Atomic Energy Community (EAEC) aimed at controlling this sector and ensure the energy security for the member countries [49,50].

In 1987 the "Single European Act" introduced the environmental protection policy and a year later the "European Commission Working Document on Internal Energy Market" presented goals in the energy policy. Since 1992 the EU has been working on the establishment of the single energy market, which comprises three stages. The next step in the development of cooperation in the energy sector was connected with the adoption in 2010

"A strategy for competitive, sustainable and secure energy", specifying priorities of the EU in the energy policy by 2020. The EU identified these priorities as ensuring competitiveness of prices and energy supply security as well as enhancing the technological advantage in this sector [51]. The assumptions of the "Europe 2020" strategy assumed and increase in energy efficiency by 20%, an increase in the share of energy from renewable energy sources to 20% total energy consumption, reduction of greenhouse gas emissions to the level of 20% in 1990 [8,51]. In the next "2030 Energy Strategy" the EU defined the goals for climate and energy, within which the member countries declared by 2030 to reduce by min. 40% their greenhouse gas emissions compared to the levels of 1990, to increase the share of renewable energy sources to 32% energy consumed in the entire EU, to improve energy efficiency by 30% and to ensure potential transfer of 15% electricity generated in the EU to other EU countries within the framework of interconnection systems [8,51]. Reduction of greenhouse gas emissions by 80–90% compared to that in 1990 is another strategic goal of the EU specified in the "2050 Energy Roadmap" of 2011 [52].

The next step was connected with the adoption of the European Council in 2014 of the "European Energy Security Strategy", assuming short-term actions in case of gas supply stoppages or disruptions in its imports to the EU. The framework for the energy policy for the years 2020–2030 was updated by the European Commission in 2016 in the "Clean Energy for all Europeans" package, which indicated the ambitious goal of energy efficiency increased by 32% [9].

In accordance with the climate strategy assumptions referred to as the European Green Deal [53], presented at the 2019 Climate Summit in Spain, the EU declared to reduce greenhouse gas emissions. Initially, it was assumed to reduce it by 80–95% by 2050 compared to the levels of 1990; finally, it was decided that the EU countries are to become zero-emitters (climate neutral) by 2050. In turn, during the European Council summit on 10–11 December 2020, the UE-27 leaders agreed to increase the CO<sup>2</sup> reduction goal to 55% by 2030 [54].

All the actions undertaken by the EU, including the VG, related to energy and climate are consistent with the "2030 Agenda" assumptions [46]. These ambitious EU goals to attain new climate and energy goals focus primarily on the SD of the energy sector. These actions concentrate, e.g., on increased use of alternative energy sources, including renewable sources, in the energy balance [55]. A growing body of evidence on the negative effect of fossil fuels used to produce energy on the natural environment, human life and health are primary reasons for the growing interest in renewable energy sources [42]. The main goal of the sustainable energy policy is to limit the consequences of the negative impact of the energy sector on the atmosphere [56]. Governments worldwide are promoting the use of renewable energy sources [57,58].

Energy should be produced and consumed solely when generated from clean energy sources, i.e., mainly renewable energy [59–62]. Renewable energy sources include biomass energy, solar energy, hydropower, tidal power, wind and geothermal power [63–65]. In view of the above, SED in individual countries is required for the further existence of the energy sector, and it is key for the development not only of renewable energy sources, but also the economy, the environment and society [66]. "Increased importance of renewable energy in the global fuel and energy balance may contribute to savings in the consumption of energy raw materials and improve the condition of the natural environment thanks to reduced air and water pollution levels and decreased amounts of generated waste. For this reason support for the development of renewable energy sources is rapidly becoming a major direction in politics, which has to be considered when planning the energy policies of many countries worldwide" [42].

The use of renewable energy sources has been investigated in many studies and scientific analyses. They concern mainly the development of renewable energy in the context of SD [67], the potential to use solar energy from photovoltaic systems, their efficiency and environmental impact [68,69], potential to use wind energy [70,71], hydropower [72], tidal energy [73], geothermal energy [74] and biomass energy [75]. Many publications

are devoted to the economic efficiency of investments in renewable energy sources in the EU [5,76–78], and the VG [79–81].

The primary indications for the growth and development of the renewable energy sector include the fact that these sources emit considerable lower amounts of greenhouse gases and other pollutants [82] and contribute to reduced greenhouse gas emissions [83]; renewable energy has a minimal environmental impact [84]; it does not require a specialized infrastructure and may contribute to an increase in employment rates [60] and provide economic benefits, particularly in rural areas, while its production is cheaper compared to conventional sources [61].

However, there are some barriers hindering rapid implementation of renewable energy sources. The main barrier is connected with the high initial cost of renewable technologies (e.g., photovoltaic panels or wind turbines), lack of data and information on resources, lack of storage facilities, insufficient capacity to construct the systems and monitor efficiency of renewable energy sources, challenges related to the integration of conventional and renewable energy technologies, the effect on agricultural land use, lack of potential for the enforcement of respective policies or design and implementation of renewable energy programs [85].

Dependencies between sustainable development and renewable energy indicated in literature on the subject include the role of renewable energy in economic development. Humanity since the very beginning was based on renewable energy. Biomass, water energy or solar energy were the only available energy sources. However, in the course of development, industrial countries started to exploit new energy sources, including also nuclear energy. At present, in many countries, energy is perceived as a right and governments are expected to meet this need. Consumers of energy services mainly want them to be abundant, reliable and accessible. However, many renewable energy sources are dependent on the nature forces and the environment, as is the case with, e.g., wind or solar energy. Thus, abundance or reliability of many energy sources varies depending on the region. Shortages or disruption in energy supply may also be experienced. For small settlements or remote communities, energy may be sufficient, but when considering large agglomerations or industrial areas with high energy demand, the use of renewable energy sources has to be adequately designed. Costs of renewable energy are also crucial. In many cases, the use of renewable energy is being promoted based on the prospective reduction in its cost. The EU policy concentrated on the support for policies and enterprises of its member states to use environmentally friendly energy from renewable sources [56,67].

At present, in the EU, including the VG, it is promoted to use solar energy in households thanks to subsidies for the purchase of photovoltaic panels, replacement of coal-fired furnaces and thermal retrofitting of family housing. Incentives are also introduced for the purchase of electric cars.

An essential aspect discussed in literature is also connected with the energy security as an aspect of sustainable development [85]. This concerns the reliability and availability of energy services, particularly in industrialized countries, where energy supply disruptions generate costs. In turn, the threat of fluctuations in energy prices may influence the economy and in extreme cases lead to an economic crisis. An important role in this respect is played by the state and its energy security policy. The EU, to promote energy security, has formed the single energy market, where a diversification of energy sources is being implemented. The EU is trying to become independent of external energy supplies; thus, diversification is observed in the forms of energy generation aiming at the increased use of renewable energy [9].

In terms of the EU energy policy, including that of the VG, a priority is to maintain a balance between security, satisfaction of social needs, economic competitiveness and environmental protection [67]. The strategy to develop the renewable energy sector indicates rational use of renewable energy sources, which will contribute to improved efficiency in the use and conservation of energy material resources and improve the condition of the natural environment [67].

Within the last 30 years the EU countries have recorded a considerable increase in the production and consumption of energy generated from renewable sources. In the years 1990–2019 greenhouse gas emissions decreased by 24%, while GDP increased by 60% [8].

The EU is the largest world source of public funds allocated to countering climate change. In 2019 they amounted to 21.9 billion euro. The EU finances sustainable transformation to meet the assumption of the European Green Deal. The countries of the VG diversified energy supplies, but in each of these countries, the structure of energy sources was different. Renewable energy sources were also introduced gradually and systematically. Their level is still low, but an upward trend was visible [79].

One of its goals is to co-finance renewable energy production. Although renewable energy in the electricity generation sector has been developing rapidly, an accelerated progress is also needed in transport, heating and cooling [86]. Within the last few years, globally, access to electricity has increased greatly; the use of renewable energy in the power engineering sector has increased, and energy efficiently has improved. However, due to the COVID-19 pandemic, millions of people are losing access to electricity [87]. Progress in the realization of the "2030 Agenda" SDG 7 seems to be too slow to promise the global energy goals are reached by 2030, with the pandemic additionally slowing it down or even reversing the progress [86,87].

#### **4. Results and Discussion**

Tables 1–3 show the changes in renewable energy consumption in transport, electricity, heating and cooling from 2009 to 2019. It is easy to see the increase in the use of renewable energy EU countries. In 2019, countries such as Sweden, Finland and the Netherlands had the largest share of renewable energy use in transport (30.31%, 21.29%, and 12.51%, respectively). For renewable energy use in electricity, countries such as Austria, Sweden and Denmark led the way. When it comes to renewable energy use in heating and cooling, countries such as Sweden, Latvia and Finland were the leaders: 66.12%, 57.76%, and 57.49%.

In contrast, the lowest renewable energy consumption occurred in countries such as:


During the period under review, the biggest changes in renewable energy consumption took place in countries such as: Malta (in 2009, the shares of renewable energy consumption especially in transport, electricity were 0.00, while in 2019, they were already close to the EU average), Estonia (the share of renewable energy consumption in transport has increased more than tenfold), Cyprus (the share of renewable energy consumption in electricity has increased more than 16-fold), and Slovakia (the share of renewable energy consumption in heating and cooling has more than doubled). However, as noted earlier, despite the significant increase in renewable energy consumption, most of the countries mentioned are still characterized by the lowest percentage of renewable energy use.

It should be mentioned that in 2019, renewable energy sources made up 34% of gross electricity consumption in the EU-27, slightly up from 32% in 2018. Wind and hydro power accounted for two-thirds of the total electricity generated from renewable sources (35% each). The remaining one-third of electricity generated was from solar power (13%), solid biofuels (8%) and other renewable sources (9%).

In 2019, hydro power use dominated the renewable energy mix in countries such as: Austria (76%), Bulgaria (48%), Croatia (74%), Finland (43%), France (53%), Italy (41%), Latvia (73%), Romania (65%), Slovakia (65%) and Sweden (66%)—Figure 2. Wind energy, on the other hand, dominated the structure of renewable energy sources in countries such as: Belgium (48%), Cyprus (45%), Denmark (69%), Germany (50%), Greece (42%), Ireland (86%), Lithuania (55%), Netherlands (49%), Poland (57%), Portugal (43%) and Spain (52%)—Figure 2.


**Table 1.** Share of renewable energy in transport in years 2009–2019 (Note: countries were ordered by 2019 index value, from highest to lowest).

**Table 2.** Share of renewable energy in electricity in years 2009–2019 (Note: countries were ordered by 2019 index value, from highest to lowest).



**Figure 2.** Structure of renewable energy use in EU countries in 2019.

In the next step, a cluster analysis was carried out, but before starting the cluster analysis, we standardized all three variables. As a first step in the cluster analysis, we analyzed correlations among the clustering variables (x1: share of renewable energy in transport in 2019, x2: share of renewable energy sources in electricity in 2019, x3: share of renewable energy sources in heating and cooling in 2019): strong correlation leads to an overrepresentation of the variables in the final clustering solution [88]. All bivariate correlations fell well below the 0.9 threshold, indicating no potential collinearity issues.

The clustering was performed based on the method of Ward. The results are given in Figures 3 and 4. The tree diagram (Figure 3) is the first and the simplest result of the cluster analysis, and it is closely related to the second result, the graph of amalgamation schedule (Figure 4). The algorithm first calculates all the Euclidean distances between the countries (and puts them in the tree diagram), and only after arranging the distances in an ascending scale, it shows the amalgamation schedule.

**Figure 3.** Tree diagram: hierarchical cluster analysis of renewable energy consumption in European countries in 2019.

The key to interpreting a hierarchical cluster analysis is to look at the point at which any given pair of countries "join together" in the tree diagram. Countries that join together sooner are more similar to each other than those that join together later. For example, the pair of countries with the lowest (shortest) distance (Spain and Italy; Slovakia and France, distance = 0.45) join together first in the tree diagram.

To find the optimal number of clusters, use the graph of amalgamation schedule. One could observe that at 23rd step, Euclidean distance rises sharply at value 3.9 (indicated by red line). Determining 2.5 as a cutoff point (as suggested by the amalgamation schedule in Figure 4) results in five distinct clusters of EU countries (Figure 3).

Based on the cluster analysis results, it is perceived that the first cluster includes: Lithuania, Estonia, Cyprus, Greece, Slovenia, and Bulgaria. This is the group of countries that is characterized by the lowest share of renewable energy use in transport compared to other clusters. The average for these countries is 5.41% (in 2019).

any given pair of countries "join together" in the tree diagram. Countries that join together

**Figure 4.** Graph of amalgamation schedule.

any given pair of countries "join together" in the tree diagram. Countries that join together The next, third cluster includes nine EU countries: Slovakia, France, Malta, Hungary, Czechia, the Netherlands, Poland, Luxembourg and Belgium. Thus, it is the most diverse cluster. This group includes for example the entire VG (Poland, Slovakia, Czechia and Hungary). This is the cluster with the lowest share of renewable energy in electricity, heating and cooling. In 2019, on average, these shares were: 15.63%, and 16.52%. It is worth noting that in this group all the indicators used in the analysis were below the average for the whole European Union (27 countries).

The fourth cluster includes only two EU countries: Sweden and Finland. In 2019, in these Nordic countries, electricity production was in one half renewable (in average 54.62%). Within it, the largest share was hydro power followed by biomass (from forestry) and wind power (like it was mentioned before). Importantly, in 2017, Finland adopted a "National Energy and Climate Strategy" [89]. A specific target for overall renewable energy share was not defined in this policy, but it had exceeded already in 2014 the 67.5% target set for 2020 in the NREAP (National Renewable Energy Action Plan) [90]. In turn, Sweden had an energy commission in place, which submitted its final report in January 2017 [90]. The commission proposed a 100% renewable energy target for 2040.

The last, fifth, cluster includes five countries: Latvia, Denmark, Portugal, Croatia, and Austria. It is worth noting that this is the group of countries with the largest share of renewable energy consumption in electricity. In 2019, the average for countries was 59.50%. For example, Denmark has the highest share of wind power in the world.

Using the hierarchical cluster analysis method, we can group the EU Countries according to the characteristics of the analyzed three variables, revealing the existing structures as well as the way in which the analyzed countries are linked in hierarchical structures. Thus, by tackling these clusters as a whole, it is possible to improve efficiency and more effectively focus public policies and financial support instruments for renewable energy sources, resulting in effects in the countries that are part of the same cluster.

According to the results of the study, there is a serious gap in 2019 regarding the differences in the use of renewable energy sources in EU countries. In the case of transport, the gap recorded between the lowest share (Cyprus, 3.32%), and the highest (Sweden, 30.31%) was about 9.2 times larger. In the case of electricity, the gap recorded between the lowest share (Malta, 8.04%) and the highest (Austria, 75.14%) was about 9.3 times

larger. Finally, in the case of heating and cooling, the gap recorded between the lowest share (Ireland, 6.32%) and the highest (Sweden, 66.12%) was about 10.5 times larger. These unfavorable differences in the use of renewable energy sources will obviously have an impact in the medium to long term on the ability of individual countries to achieve their sustainable development goals.

The results obtained are consistent with those obtained by Włodarczyk et al. [91]. The cluster analysis conducted by the researchers made it possible to distinguish 5 groups of EU countries differing in their effectiveness in achieving sustainable development goals. The group of countries that were characterized by "highest average value of share of renewable energy in transport (15.97%, exceeding the EU average with 81.7%), highest average value of share of renewable energy in electricity (57.01%, representing an increase of 75.7% compared to the EU average), highest average value of share of renewable energy in heating and cooling (57.35%, exceeding the EU average with a remarkable 91.6%), next to the lowest average value of greenhouse gas emissions intensity (71.77%, representing a decrease of 13.4% compared to the EU average)" [91] (p. 10) included: Denmark, Finland, Latvia and Sweden. In contrast, the group of countries that do not perform as well in these areas included: Belgium, Cyprus, Lithuania, Luxembourg and Malta [91].

Finally, we want to emphasize that the importance of renewable energy in the energy mix is increasingly reflected in specific activities and regulations at the international level. In practice, the environmental benefits of adopting renewable energy sources are undeniable today, and they are increasingly explored and analyzed in the literature. Research in this area has been carried out not only at the European Union level [91,92] but also at the level of individual countries, e.g., Germany [93,94], Hungary [95,96], France [97], Greece [98] or Spain [99].

#### **5. Conclusions**

The use of renewable energy sources is becoming one of the priorities of the EU. This is a consequence of the growing importance of the concept of SD and SED. Thus, more and more often biomass energy, solar energy, hydropower, tidal power, wind and geothermal power are used in cooling, heating, electricity and transport.

Our paper makes several contributions. Firstly, our study contributes to the SD and SED literature by offering a comprehensive grasp of its underpinnings in light of recent advances. Secondly, on the basis of the conducted research, the following can be noted: (1) In 2019, renewable energy sources made up 34% of gross electricity consumption in the EU-27; wind and hydro power accounted for two-thirds of the total electricity generated from renewable sources. (2) Between 2009 and 2019 there was an increase in the use of renewable energy sources in transport, electricity, cooling and heating (the biggest changes in renewable energy consumption took place in countries such as Malta, Estonia, Cyprus and Slovakia). (3) Five groups of EU member states have been identified, which differ in terms of renewable energy consumption. (4) The undisputed leader in the European Union in terms of the development of the renewable energy sector is Sweden, which had the largest share of renewable energy consumption in transport, heating and cooling during the period under review. (5) The entire VG (and also France, Malta, the Netherlands, Luxembourg and Belgium) in comparison with other EU countries is characterized by the lowest share of renewable energy in electricity, heating and cooling.

Despite these contributions, our study is not without limitations. Firstly, the literature review section does not include all possible studies on the discussed concepts. In the selection of literature, the authors were guided by its diversity, availability and timeliness. Secondly, cluster analysis was performed on three indicators only. Such indicators were omitted, e.g., greenhouse gas emissions intensity of energy consumption or final energy consumption in households per capita. Thirdly, the use of Ward's method resulted in low abundance clusters (e.g., one of the clusters includes only two EU countries: Sweden and Finland).

The results provide an interesting starting point for future research. The methodology used in this article can be reproduced with other indicators both quantitative and qualitative. Another suggestion would be to perform a cluster analysis based on indicators showing changes in consumption of renewable energy sources over several years (dynamic approach). Finally, in cluster analysis, other methods could be used in addition to Ward's Method (possibility of comparing results).

**Author Contributions:** Conceptualization, E.K., K.Ł. and P.P.; methodology, E.K., K.Ł. and P.P.; writing—review and editing, E.K., K.Ł. and P.P.; visualization, E.K., K.Ł. and P.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Tomasz Rokicki 1 , Grzegorz Koszela 1 , Luiza Ochnio 1, \* , Kamil Wojtczuk 1 , Marcin Ratajczak 2 , Hubert Szczepaniuk 2 , Konrad Michalski 2 , Piotr Bórawski <sup>3</sup> and Aneta Bełdycka-Bórawska 3**


**Abstract:** The main purpose of this paper is to present the differences in the volume of energy consumption in transport in the EU (European Union) countries. The specific objectives aim to determine the directions of changes and the degree of concentration in the volume of energy utilized by the transport sector in EU states, showing various models in this area, to establish the association between energy absorption and the parameters of the economy and in the field of transport. All EU countries were selected for research by the use of the purposeful selection method as of 31 December 2018. The analyzed period covered the years 2004–2018. For the examination of data, grading data analysis was used as one of the methods of multivariate data analysis. Descriptive, tabular and graphic methods were used to present the results. Findings reveal that there is a general tendency to reduce total energy consumption in the EU countries. The same is the case of energy in transport. Only in 2016–2018 was there an increase in energy absorption in transport. The reason was the better economic situation in this period. Road conveyance is the most important factor in energy utilization (over 90%). The share of other modes of transport was very small. Economically developing countries were the fastest in increasing energy absorption in transport per capita. In turn, highly developed states recorded slight growth and were stable in this aspect. There was a close relationship between energy utilization in transport per capita and GDP per capita. The reduction of energy consumption in transport depends on changes in road haulage, e.g., the pace of introducing innovative energy-saving technologies in automotive transport.

**Keywords:** energy in transport; energetic efficiency; energy sources; economic growth; developing and developed countries

#### **1. Introduction**

#### *1.1. The Importance of Transport in Energy Consumption*

The transport sector is one of the industries with the highest energy absorption in the world. In 2018, this form of business accounted for about 64% of global oil consumption and around 29% of total final energy absorption [1]. Both passenger and freight conveying exploit energy. As a rule, it is not possible to split energy consumption solely into any of these modes of transport [2]. Dingil et al. [3] found that a significant increase in transport energy intensity occurs in cities with a low population density. The main reason for the high energy absorption of transport in such municipalities was the high share of private means of transportation. Brownstone and Golob [4] achieved similar results. More fuel was used in sparsely populated areas. Schippl and Arnold [5] demonstrated that political

**Citation:** Rokicki, T.; Koszela, G.; Ochnio, L.; Wojtczuk, K.; Ratajczak, M.; Szczepaniuk, H.; Michalski, K.; Bórawski, P.; Bełdycka-Bórawska, A. Diversity and Changes in Energy Consumption by Transport in EU Countries. *Energies* **2021**, *14*, 5414. https://doi.org/10.3390/en 14175414

Academic Editor: Sergio Ulgiati

Received: 22 June 2021 Accepted: 27 August 2021 Published: 31 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

measures that limit automotive car mobility would also be needed to achieve a full-scale transition towards multimodal urban mobility. Newman and Kenworthy [6] argued for an energy compromise in transport. In downtowns, energy efficiency was lower than in suburban areas. However, the total fuel depletion is smaller in these areas. The conclusions from the presented research are similar. Urbanization and patterns of settlement used in a given country have a significant impact on the energy efficiency of passenger conveyance.

Thus, the energy consumption of transport may be related to the economic situation and the mobility of the society. Enhanced energy absorption in transport is associated with increased total energy utilization in the economy [7–11]. Such relationships were found in the Banister and Stead studies. They showed that the strong relationship between economic activity and transport demand significantly increases energy consumption and, consequently, carbon dioxide emissions [12]. Thus, higher economic growth leads to increased energy utilization. Conversely, the use of large volumes of energy reflects high levels of economic growth. There is a great number of research confirming such relationships [13–25]. Ozturk and Acaravci [26] presented the feedback between energy absorption and economic growth in Hungary. Belloumi conducted similar research in Tunisia [27].

The increase in transport output causes an enhancement in energy consumption. Innovations are needed to reduce transport energy absorption and to reduce air pollution. For example, advances in vehicle technology can decrease the energy intensity of the transport sector and improve the energy efficiency of haulage activities. As a result, the positive importance of transport in global economic activity will increase. Solutions such as electric drive, hybrid plug-ins, and hydrogen are implemented to reduce energy consumption. Cars also use other innovative technologies that facilitate driving and reduce energy consumption [28–34]. Another way is to maximize the use of the load capacity of the means of transport for the movement of goods and the number of seats for the conveyance of people [35]. When transporting people in cities, no car traffic zones are introduced to force urbanites to use public transport [36–40].

Change in transport is heading towards ecological and economic balancing (sustainability) [41]. Thus, various forms of transportation are used. In general, there is a tendency towards intermodal transport to utilize the best properties of individual means of transport. Scientific research has focused mainly on energy efficiency in road transport [42–44]. Many studies also refer to the efficiency of urban transport, which uses various types of transport [45–47].

#### *1.2. Selected Ways to Improve Energy Efficiency*

Technological progress allows for benefits in terms of productivity and technology of energy utilization, which contribute to the reduction of greenhouse gas emissions. A move towards renewable energy production by emphasizing cleaner energy carriers (such as electricity and hydrogen) would improve urban air quality [48,49]. Efficient use of energy is a very attractive way of reducing the impact of energy on the environment and health. Achieving the same outcomes with less energy should theoretically reduce costs and emissions of local pollutants and greenhouse gases [50–52].

The improvement of global energy efficiency is indicated by the ratio of energy consumption to gross national income (GNI). Historically, total energy absorption per person has steadily increased. This was because the surged energy efficiency coexisted with economic growth, rising expectations, social changes and population growth. Therefore, people must reduce energy-related emissions of greenhouse gases and other pollutants [48,53–55].

#### *1.3. Relationship between Sustainable Transport Development and Economic Growth*

Many studies emphasize the two-way symbiosis between transport and economic growth, which influence each other through feedback. Transport is important to the development of a sustainable economy that aims to provide new services. Transport should enable the movement of goods and people, and at the same time, contribute to environmental protection and ensure safety [56–60]. Sustainable development requires an efficient and safe transport system powered by clean, low-emission, secure and inexpensive energy. Energy used in transport enables social and economic development. Therefore, energy policy in transport should result from the program of sustainable development of the economy [61–63].

The relationship between energy consumption by transport and pollutant emissions (mainly CO<sup>2</sup> and other harmful compounds) is known. The environmental Kuznets curve is empirically tested in many countries and regions using various indicators of environmental degradation and many econometric techniques of cross-sectional and panel data. The Kuznets curve shows the relationship between GDP per capita and inverted U-measures of environmental degradation [64]. Industrialization increases the negative environmental influence of economic activity up to a point where the impact decreases with continued economic growth. Individual EU states differ in terms of economic development, which means that they may be at different stages of evolution. The environmental impact of these countries may also vary. Obtained results and relationships can be related to transport, which absorbs a lot of energy and emits many pollutants at the same time. Energy consumption was used as a variable in many studies [65–76]. Some researchers also negate the assumptions of the existence of the Kuznets curve. It all depends on the type of contamination [77–80]. As a result, in each state or group of countries, it is possible to obtain different outcomes confirming or negating the existence of the Kuznets curve. The results received will largely depend on the level of energy efficiency of the country and region. In general, in states with high GDP, the Kuznets curve was most often used. Examples are France [81,82], Canada [83], Spain [84] and the United States of America [85]. Patterns are also confirmed in countries with average GDP levels, such as Malaysia [86], China [87], Turkey [74], Romania [88], Tunisia [89] and Latin America and the Caribbean [90]. Many studies have confirmed that the use of fossil fuel energy increases air pollution. An example is the use of crude oil to power internal combustion engines [91–93].

It should also be mentioned that there is an interaction between economic growth, energy consumption and environmental quality. These relationships are the subject of energy economics research [94–96]. Environmental quality can generate positive or negative externalities. Consequently, it stimulates economic growth by focusing on human health, which is potentially affected by emissions. The link between energy variables, progress and environmental quality has been the subject of conflicting and paradoxical goals set by policymakers. This relationship is the basis for creating a sound economic policy consistent with its environmental and energy policy objectives. Empirical work on the tripartite causality link between energy, economic growth and the environment can be broken down into three lines of research. The first concerns the relationship between energy variables and economic growth. According to the assumptions, very good economic results require a high level of energy absorption, and effective energy utilization requires large economic growth [97–104].

The sustainable development of the transport sector can be divided into three main sections: society, economy and environment. The evolution of transport requires sustainability to achieve the minimum expectations in these three sectors. Increasing the role of transport in sustainable development is realized by promoting public transport, demand management, improved road management, pricing policy, improved vehicle technology, using clean fuels and transport planning [105,106]. From their current structure to one that is compatible with sustainable development, transforming global transport systems is likely to be a long-term process involving continuous changes in several physical, technological and institutional systems [107].

#### *1.4. Justification, Aims and Structure of the Article*

The subject matter of the article is important and up-to-date. Transport is a significant energy consumer. Many research papers are describing the relationship between energy absorption in transport and the parameters of the economy. A novelty is the application of multidimensional analysis using the Gradestat software. As a result, it was

possible to investigate the situation in individual countries regarding energy consumption in transport and GDP (Gross domestic product). Data per capita were also calculated in the research, which enabled an accurate comparison of countries with different levels of economic development.

There is a research gap that this article can fill. The literature review shows no previous studies on the relationship between energy consumption and economic development. For instance, we found only one publication that reported the relationship between energy absorption in transport per capita and GDP per capita. In addition, our research will cover the area of the EU, which is still quite diverse. In addition, the quite rare method, the GCA algorithm (grid-based clustering algorithm), was used in the study. The above aspects make the research necessary and original.

The main goal of the article is to present the differences and changes in the volume of energy utilization by transport in the EU states. The specific objectives are:


One research hypothesis was formulated in the paper:

Hypotheses: the rate of changes in energy absorption in transport per capita is closely related to the level of economic development of the country.

The organization of this paper is as follows: Section 1 provides an introduction to the subject. The importance of transport in energy consumption, ways of improving energy efficiency, the tripartite relationship between transport energy use, environmental pollution and economic growth are presented. This section also contains the justification and aims of the article. Section 2 proposes methods to identify differences and changes in the volume of energy absorbed by transport in the EU states. In Section 3, the research findings are presented. In Section 4, the reference is made to other research results that dealt with the relationships tested. Finally, Section 5 concludes this paper.

#### **2. Materials and Methods**

#### *2.1. Data Collection, Processing and Limitations*

All EU countries were selected for this research using the purposeful selection method as of 31 December 2018. In total, 28 EU states were examined. When presenting the results in tables and graphs, the abbreviations of country names were used. Acronyms of the country name were used in work in accordance with ISO 3166-1 alfa-2. They are as follows: Austria (AT), Belgium (BE), Bulgaria (BG), Cyprus (CY), Czechia (CZ), Germany (DE), Great Britain (GB), Denmark (DK), Estonia (EE), Spain (ES), Finland (FI), France (FR), Greece (GR), Croatia (HR), Hungary (HU), Ireland (IE), Italy (IT), Lithuania (LT), Luxembourg (LU), Latvia (LV), Malta (MT), Netherlands (NL), Poland (PO), Portugal (PT), Romania (RO), Sweden (SE), Slovenia (SI), Slovakia (SK).

The research period covered the years 2004–2018. This is because 2004 saw a significant expansion of the EU with 10 new states, and 2018 was the last year when complete research data were available.

The data used in the study come from Eurostat for the 15 years 2004–2018. To ensure the stability and transparency of the obtained results, this period was most often divided into 3-year sub-periods. Data collection is limited by the lack of detailed and timely information on energy in transport. Additionally, these data are aggregated at the country level, so there is a problem with performing analyses at the regional level.

Energy absorption was measured in the toe. The ton of oil equivalent (toe) is a unit of energy defined as the amount of energy released by burning 1 ton of crude oil. It is approximately 42 gigajoules or 11.630 megawatt-hours, although as different crude oils have different calorific values, the exact value is defined by convention [1].

The study is a result of the authors' previous research on transport. Quite recently, the field of the writers' interest has been power engineering. These two areas are closely connected because without energy, transport is impossible. The vast majority of authors are economists. Therefore, the aspect related to economics was raised. Additionally, it was noted that there are no current academic studies on the relationship between energy consumption in transport and economic development.

#### *2.2. Applied Methods*

The research was divided into stages. Figure 1 shows a diagram of the conducted research.

The first stage of the research shows the concentration of energy consumption in total and transport in individual EU states. The data concern 2018, the last year of the analyzed period. As a result, it was possible to compare the degree of concentration of energy absorption in total with energy intended for transport. All EU countries were examined. Furthermore, all 3-year periods in 2004–2018 were investigated. As a result, it was possible to notice the current trends in the EU in terms of total energy utilization and in transport.

In the second stage of the research, the structure of energy consumption in transport with the division into modes of transport was shown. The purpose of this section was to identify the importance of different modes of transport in the EU. Additionally, trends in individual modes of transport were presented.

The third stage of the research presents changes in GDP per capita in individual EU states. For data examination, the grading data analysis method was implemented as one of the multivariate data analysis techniques that can be used to graphically present the dynamics of phenomena or differences between objects in the form of overrepresentation maps.

The GCA algorithm (grid-based clustering algorithm) also allows creating groups, but it generates them in a way that allows creating, in this case, 3 objects that are characterized by the greatest possible differentiation among themselves. These clusters are formed as a result of combining objects that ensure such differentiation, and for this purpose, a certain independence index, Ro or Tau, is optimized [108].

There are many proposals in the literature for the construction of structure dissimilarity indicators. Distances are often used for this purpose, e.g., Minkowski metric [109].

$$d(\mathbf{x}, \mathbf{y}) = \left(\sum\_{i=1}^{n} |\mathbf{x}\_i - \mathbf{y}\_i|^p\right)^{\frac{1}{p}} \tag{1}$$

If we have two structures: *x* and *y*, where:

$$\mathbf{x}\_{i} \ge \mathbf{0} \qquad \sum\_{i=1}^{n} \mathbf{x}\_{i} = 1 \qquad y\_{i} \ge \mathbf{0} \qquad \sum\_{i=1}^{n} y\_{i} = 1 \tag{2}$$

this measure certainly meets two conditions:


$$\Lambda\_{\underline{n}\geq\underline{k}>j>i\geq 1}d(\mathfrak{x},\mathfrak{x}\_{\mathrm{ij},\varepsilon}) \leq d(\mathfrak{x},\mathfrak{x}\_{\mathrm{ik},\varepsilon})\tag{3}$$

However, one can have some doubts as to the correctness of the fulfillment of the third condition by the dissimilarity index:

3. The distance measure changes according to the "transfer sensitivity" adopted in the concentration indices. The increase in the value of the dissimilarity index at a constant transfer value is the greater the "richness" of the object to which the transfer was made.

$$\mathbf{x} = (\mathbf{x}\_1, \dots, \mathbf{x}\_{\mathbf{i}}, \dots, \mathbf{x}\_{\mathbf{j}}, \dots, \mathbf{x}\_{\mathbf{j}}, \dots, \mathbf{x}\_{\mathbf{k}}, \dots, \mathbf{x}\_{\mathbf{n}}) \tag{4}$$

$$\mathbf{x}\_{i\bar{j},\varepsilon} = (\mathbf{x}\_1, \dots, \mathbf{x}\_{\bar{i}} - \varepsilon, \dots, \mathbf{x}\_{\bar{j}} + \varepsilon, \dots, \mathbf{x}\_{\bar{k}}, \dots, \mathbf{x}\_n) \tag{5}$$

$$\mathbf{x}\_{i\mathbf{k},\varepsilon} = (\mathbf{x}\_1, \dots, \mathbf{x}\_i - \varepsilon, \dots, \mathbf{x}\_j, \dots, \mathbf{x}\_k + \varepsilon, \dots, \mathbf{x}\_n) \tag{6}$$

In the case of this study of energy consumption from transport, it is about shifting the value of energy absorption between years (the more years shifted, the greater the value of the dissimilarity index should be) because we are interested in which countries have experienced faster growth in energy consumption. The construction of the dissimilarity index of structures meeting condition three can then be based on the concentration index (Gini coefficient) and the Lorentz curve.

By analogy with the Lorentz curve, the dissimilarity of the *Y* structure to the *X* structure can be presented as a broken line connecting certain points, the coordinates of which in this case are successive cumulative structures, and the measure of the dissimilarity of the *Y* structure to the *X* structure also by analogy—this time with the Gini coefficient—is the measure "*ar*".

$$ar(y:x) = ar\left(\mathbb{C}\_{[y:x]}\right) = 1 - 2\int \mathbb{C}\_{[y:x]}(t)dt\tag{7}$$

where *C*[*y*:*x*] : [0, 1] → [0, 1] belongs to the group of continuous functions.

By measuring the distances between structures (in the case of our study, e.g., dynamics of changes in the GDP per capita of the European Union countries) using the ar measure, we can spot subtleties to which Minkowski's metric is insensitive. Visualization of the structures was made with the use of overrepresentation maps. Overrepresentation, in this case, is the ratio of the component structures (in this case, the structures for individual countries in periods) to the average value. Thus, as an average, we understand the ratio

of the sum of the quantities, e.g., energy consumption in individual periods, to energy consumption in the entire period under examination for the entire EU.

After determining the average values, we can calculate the so-called "overrepresentation indicators". The overrepresentation indicator shows how far the observed value differs from that which would be expected given the ideal proportionality of the distribution. For an ideal representation, the indicator will take the value 1. Those determined overrepresentation coefficients allow drawing the "map of overrepresentation" where, with appropriate values of the indicators, different shades of gray are encoded (the overrepresentation map for the proportional distribution would be uniformly gray without any shades). The map of overrepresentation is a square with sides equal to 1, wherein in this case, the rows are EU countries, and the columns are energy consumption in particular periods. Colors show overrepresentation (extreme black) or underrepresentation (extreme white). The map has rows and columns of varying heights and widths:


The concepts discussed: the concentration curve, the ar index and the overrepresentation map are closely related to the Grade Data Analysis (GDA). As part of the Grade Data Analysis, some quite complex operations are performed. The main issue in GDA is studying the diversity of rows and columns and striving to arrange them in the data matrix in such a way as to achieve the maximum contrast between the outermost rows and columns. This goal is implemented by the GCA (Grade Correspondence Analysis) algorithm. It rearranges the rows and columns of the data matrix to maximize a certain dependency ratio. In this case, only the rows are rearranged as the columns are in chronological order. This dependency index is the rho-Spearman or Kendall's tau correlation coefficient and depends mainly on the dissimilarity index ar. Based on these indicators, clusters are built in such a way as to maximize the differentiation between them. In contrast, the differentiation between two clusters is understood as the differentiation between two objects formed from these groups as the sum of the objects included in them.

The number of clusters, in this case, depends on the number of observations (there are only 28). Therefore, it is a subjective choice of the authors.

The third stage of the research also shows the relationship between energy consumption and GDP in individual EU countries. The aim was to determine whether such a correlation exists and whether it concerns all EU states or a group of countries. Looking for a linear relationship between two rankings, it was decided to perform a procedure that allowed to reconcile the classic approach of the Spearman rank correlation coefficient r<sup>s</sup> with Pearson's linear correlation coefficient r [110,111].

Descriptive, tabular and graphic methods were also used to present some of the findings.

#### **3. Results**

#### *3.1. Energy Consumption in the EU Countries*

As an introduction to the study, it seems reasonable to define energy consumption in EU countries compared to the entire EU. For this purpose 2018, was taken into account. It should come as no surprise that the EU states in terms of energy absorption in the analyzed period were dominated by the countries with the largest population, i.e., Germany, France, Great Britain and Italy, which together consumed more than 50% of energy for the entire EU (Figure 2a). Figure 2a shows the share of individual countries in energy consumption compared to the EU as a whole for 2018.

48.35% 39.49% 37.52% 37.44%

15% 20% 25% 30% 35% 40% 45% 50%

**Figure 2.** Final energy consumption in EU in 2018. (**a**) Total energy consumption. (**b**) Transport energy consumption.

Bearing in mind that in the scale of the entire EU, approx. 30% of energy consumption resulted from transport, Figure 2b, which, in this case, is a certain supplement to the list in Figure 2a. On the EU scale, in terms of energy absorption for transport, countries with the highest energy utilization are very similar. A total of 30% of the share of transport in energy consumption applies to the entire EU. However, the share of energy used in transport varied across states. Figure 3 shows the countries (where 2018 was included for the sake of comparability with Figure 2a), where the absorption of energy from transport to the total energy consumption was relatively the highest.

0.26%

 0.25% 0.21% 0.07%

0.17%

0.06%

**Figure 3.** Transport sector energy use in UE in 2018.

LU SI LT ES GR CY MT PT BG IE AT HR PL GB FR IT DK EE RO LV CZ HU DE SK BE SE NL FI UE With this approach to the problem, among the states from Figure 3, there is no leader in the ranking with the highest energy consumption in the EU. For example, Germany had a lower share of energy utilization in transport than the total EU average. Only Spain was included in the list of large countries. On the other hand, the energy absorption due to transport was significantly higher in Luxemburg and Slovenia than in the EU. For the sake

of completeness, it can be added (which is not shown in Figure 3) that for 2018 the lowest percentage share in energy utilization from transport was in Sweden, the Netherlands and Finland.

Energy consumption in the EU in 2004–2018 was characterized by a rather downward trend (Figure 4a). Considering the 3-year periods, the lowest level of energy absorption in the EU states occurred in 2013–2015. This significant decrease was due to the improvement in energy efficiency. Only in the years 2016–2018 was there an increase in energy utilization by 2.53% compared to the previous 3 annums. In 2016–2018, the economic situation was exceptionally favorable. However, this consumption was still 6.24% lower than in the years 2004–2006. Transport energy absorption has undergone slightly different changes to total energy consumption. The transport sector had made little use of renewable energy sources. It was also less prone to efficiency gains. The transport sector was also closely related to the economic situation in the country. Therefore, these changes were quickly visible in the trend of demand for transport and, consequently, in demand for energy in transport.

**Figure 4.** Final energy consumption in UE 2004–2018 in gigatons of oil equivalent (Gtoe). (**a**) Energy total. (**b**) Transport sector.

#### *3.2. Structure and Trends of Energy Consumption in Transport in the EU Countries*

The next stage presents the results concerning the structure and trends of energy consumption in transport. There was no relatively regular direction in the energy absorption of transportation in the EU countries. Considering the Eurostat nomenclature, the following sectors are distinguished within the energy consumption in the transport sector: rail transport, road transport, domestic aviation, domestic navigation, pipeline shipping and not elsewhere specified. In the EU states, different modes of transport have varying levels of energy intensity. Sometimes these differences were significant.

The largest share in energy consumption due to transport was recorded in road transport, which accounted for over 90% of the total energy utilization in the whole transport sector (Figure 5). For example, the list of individual components in the transport sector 2018 is presented. The results were similar in previous years. This means that the structure of energy absorption in transport within the EU is stabilized. Obviously, from the point of view of sustainable transport development, a large share of energy consumption in road transport is disadvantageous.

**Figure 5.** Energy consumption in different transport in various branches of transport by EU countries in 2018.

0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 800,000

0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000

52,000 53,000 54,000 55,000 56,000 57,000 58,000 59,000 60,000 61,000 62,000

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2004

Mln €

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000

2004

0 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 12,000,000 14,000,000 16,000,000 18,000,000

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2005

Mln €

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2004

Mln €

Mln €

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

0 50,000 100,000 150,000 200,000 250,000 300,000 350,000

Mln €

2004

Mln €

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

To find the relationship between the dynamics of changes related to energy absorption due to transport and the dynamics of changes in economic indicators, it was first decided to compare trends in the value of gross added value in current prices in one million euro gross for individual modes of transport and gross domestic product current prices, euro per capita in 2004–2018 (Figure 6).

**Figure 6.** Trends in changes in gross value added in transport and the trend of gross domestic product in 2004–2018 (current prices, million euro). (**a**) Transportation and storage. (**b**) Land transport and transport via pipelines. (**c**) Water transport. (**d**) Air transport. (**e**) Postal and courier activities. (**f**) Gross domestic product at market prices.

It is easy to notice that the trends of changes in transportation and storage, land conveyance and pipelines shipping and GDP are practically identical. Due to this and the fact that energy consumption in road transport accounted for more than 90% of the total energy utilization in the transport sector, it can be concluded that the other industries are of minor importance in the total energy consumption in transport.

#### *3.3. The Level and Changes in GDP per Capita in the EU Countries*

The next stage presents the differentiation between EU countries in terms of GDP per capita. Comparing the direction of changes in energy consumption due to transport with the dynamics of alterations in GDP per capita, it is worth noting that GDP per capita shows large differences depending on the EU country, as shown in Figure 7, which compares GDP with the average for the EU in general. All the disproportions in this comparison are too visible. Luxembourg is particularly distinct from the EU average, exceeding it more than twice. Most countries had a lower level of GDP per capita than the EU average. This group included all Central and Eastern European countries that joined the EU in 2004 and in subsequent years. Thus, there were large disparities between the EU countries. It is a background to define the dynamics of his trend.

**Figure 7.** GDP per capita in UE countries in 2018 in comparison to the average in UE.

The dynamics of GDP per capita changes against the background of the entire EU can be presented using the so-called overrepresentation maps. Figure 8a shows the dynamics of GDP per capita in an orderly manner. The countries were divided into three groups. First, declining dynamics were found in highly developed countries, such as Spain, Italy, Great Britain and France. Then, in turn, the third group includes the fastest developing countries. First of all, these are the countries of Central and Eastern Europe that develop rapidly because they want to move closer to the level of development of Western European states.

**Figure 8.** Dynamics of the pace of changes in the GDP per capita of the European Union countries in 2004–2018. (**a**) Dynamics in countries. (**b**) Dynamics in group of countries.

Based on scores, it is possible to rank states depending on the strength of the growth dynamics, in this case—GDP per capita compared to the EU. The countries with the corresponding scores and their group members are presented in Table 1. The economically developed states of Western Europe were in the cluster with the lowest GDP growth per capita. The group with the average dynamics included both economically developed and developing countries. This group included Germany, but also Hungary and Slovenia. The bunch of countries with the highest growth of GDP per capita mainly included Central and Eastern European states that joined the EU in 2004 and later. The only exceptions were Malta and Ireland. Such a division into clusters is not surprising. Economically developing countries need to catch up with the differences that separate them from highly developed countries. Such a situation is beneficial for the entire EU, as it leads to more minor differences in the economic development of individual countries. As a result, the area of the EU may be more cohesive in the coming years.


**Table 1.** Ranking of countries by the strength of GDP per capita growth in accordance with the GCA algorithm.

#### *3.4. The Level and Changes in Energy Consumption by Transport per Capita in the EU Countries*

The same operation on the overrepresentation maps as in the case of GDP per capita was repeated for the data on energy consumption in transport. Additionally, to ensure the comparability of the results with GDP per capita, the data on energy absorption from transport in kToe were converted per capita. In this approach, we remove the number of people in countries on energy consumption.

The map clearly shows the width of the row for Luxemburg, which means the highest energy utilization from transport per capita in this country (Figure 9a), and this result is greater than for the other states. The dynamics of changes for this country are not homogenous, but the trend is decreasing compared to the rest of the European states. On the opposite side is Romania, which belongs to the countries with the lowest energy absorption from transport per capita, and its direction of the trend is increasing. The majority of EU states have a moderate value of energy consumption from this industry per capita.

The first group, broken down by the GCA algorithm, consists of countries with lower energy absorption in transport than the average value of consumption in the EU countries. There are countries with stronger dynamics of changes in the third group than the average rate for the EU (Table 2). Developing states such as Romania, Slovakia and Poland are among the countries where more energy is utilized due to transport than the EU average.

**Figure 9.** Overrepresentation maps of energy consumption by transport per capita in European Union countries in 2004–2018. (**a**) Dynamics in countries. (**b**) Dynamics in group of countries.


**Table 2.** Ranking of countries according to the dynamics of energy absorption in transport per capita according to GCA algorithm.

The split carried out is mainly similar to that made in the case of GDP per capita. Highly developed states introduce technological innovations in transport to a greater extent. Of course, the energy efficiency of means of transport changes relatively slowly. Nevertheless, there is a clear advantage of these states over economically developing countries. On the other hand, in developing economies, higher energy consumption in transport is due to greater economic growth. Growing production and absorption in these societies must be handled by transport. These economies do not introduce technological innovations in transport on a large scale. An additional factor may be the expansion of the road network in economically developing countries. Huge funds from the EU have been allocated for this purpose. Another reason could be the rapid increase in the number of vehicles in Eastern European states. The increased wealth of the society and better

roads resulted in greater availability of private cars. Transport companies from Eastern EU successfully competed with enterprises from Western Europe. An example is Poland, which dominated this market. Polish organizations performed about 30% of international road transport.

#### *3.5. Relationships between Energy Consumption by Transport and GDP per Capita in EU Countries*

The next stage presents the relationship between energy consumption by transport and economic growth. Based on two rankings of dynamics of changes in GDP and energy absorption in transport, the rank correlation coefficient was calculated (r<sup>s</sup> = 0.7219). The high dependence can be easily observed in Figure 10.

**Figure 10.** Countries' positions in the rankings of changes in energy consumption by **Figure 10.** Countries' positions in the rankings of changes in energy consumption by transport per capita and GDP per capita.

On the vertical axis, positions from the ranking of alterations in energy consumption from transport per capita are marked (by the maps in Figure 9a). On the horizontal axis, positions are taken in the ranking of changes in GDP per capita. Most of the points representing the positions for States are located close to the diagonal of the square, which reflects the perfect agreement of both rankings. Nevertheless, there are quite significant and clear exceptions to this rule. This applies to Luxembourg, which, despite the GDP per capita growth in line with the pace of changes in the EU, clearly shows a slower pace of changes in energy absorption in transport per capita than the EU average. The situation is similar in Ireland (IE) and less clear but visible in Latvia (LV). It can be considered that these are cases of positive actions compared to the entire EU. Countries for which the points in Figure 9 are above the diagonal of the square and are significantly distant from it are the opposite. These include Slovenia (SI), France (FR) and Croatia (HR). Here, the increase in energy consumption from transport per capita to the GDP per capita growth rate is disproportionately higher than in the EU states. Immediately after these countries is Poland (PL), which also turned out to be the state with the highest growth rate of energy absorption in this sector in the entire EU.

Despite these cases, which can be considered outliers, attention should be paid to the very high value of Spearman's rank correlation coefficient. If the data for the countries with the greatest discrepancies in terms of places in both rankings, i.e., Luxemburg (LU) and Ireland (IE), were removed, the value of this coefficient would increase to the level of r<sup>s</sup> = 0.8114. The results confirm a high correlation in most countries between GDP per capita growth rate and the rate of energy consumption in transport per capita.

#### **4. Discussion**

According to Ibrahiem [112], energy consumption by road transport determines economic growth both in the short and long term. In contrast, economic growth causes energy absorption of road conveyance in the short term. Thus, there are feedbacks. Ibrahiem conducted his research on the example of Egypt in the years 1980–2011. Nasreen et al. [28] examined the relationship between economic growth, freight shipping and energy consumption for 63 developing countries for 1990–2016. Country panel analysis was used. Countries were divided into three sub-panels, namely middle–low income countries, medium–high-income countries and high-income states. The findings showed a two-way causal relationship between economic growth and freight transport for all selected panels and between economic growth and energy absorption for high-income and medium-high income panels. For the lower–middle-income panel, causation was one way, from energy consumption to economic growth. Additionally, the results indicate that the relationship between freight conveyance and energy consumption was bidirectional for high-income countries and one-way from freight to energy consumption for higher-middle-income and lower-middle-income countries. We obtained similar findings in our research. Economically developing states in the EU tended to proportionally absorb more energy (see Figure 10).

Liddle and Lung [113] conducted panel studies on 107 countries covering the years 1971–2009. They found that transport has been an important energy aggregation as transport energy consumption has increased in highly developed and developing countries. They distinguished between three balanced income-based panels, i.e., 40 high-income countries, 39 middle-income states and 28 low-income countries. Energy absorption in transport per capita was the dependent variable, and GDP per capita was an independent variable. The share of countries with significant positive correlations ranged from three-quarters (for high- and low-income panels) to two-thirds (for middle-income panels). However, there was no unanimity. Our research also showed a high correlation between GDP per capita growth rate and the trend of energy consumption in transport per capita. After removing a few outliers, the r<sup>s</sup> correlation coefficient was 0.8114.

Achour and Belloumi [114] explored the causal relationships between transport infrastructure (rail and road), transport value added, gross accumulation and energy intensity of transport in Tunisia in 1971–2012. A one-way relationship between energy consumption in transport and economic growth was found. Infrastructure and population density had a significant impact on the energy consumption of transport. Achour and Belloumi [115] conducted another study on Tunisia's example in 1985–2014. They found that energy intensity played the dominant role in decreasing energy absorption during the study period. Improving the transport intensity exerts a significant effect on saving energy. These studies are interesting and justify why energy utilization grows proportionally slower in the most economically developed countries than in developing countries. We found such patterns in our research.

Rehermann and Pablo-Romero [116] analyzed how the GDP per capita affects transport energy consumption, testing possible nonlinear relationships between variables. The research concerned 22 Latin American and Caribbean countries in 1990–2014. It was found that the elasticity values of transport energy absorption, with respect to GDP per capita, do not show a tendency to decrease in the long term. Saidi et al. [117] explored the impact of transport energy consumption and transport infrastructure on economic growth by utilizing panel data on MENA countries (the Middle East and North Africa region) for 2000–2016. The research confirmed that the causal relationship between energy absorption in transport and economic growth was heterogeneous. There was different flexibility depending on the level of development of the country. Our analysis also showed that the level of economic growth affects the rate of energy consumption in transport. We have

demonstrated it in the example of the EU. As demonstrated by the literature review in other countries and regions, these regularities are similar to our research.

Belke et al. [118] analyzed the long-term relationship between energy consumption and real GDP, including energy prices, in 25 OECD countries in 1981–2007. Energy absorption and economic growth are cross-sectionally correlated. The reason is regional and macroeconomic links, which are manifested through common global economic crises, mutual commercial and financial institutions and local externalities between countries or regions. There is also a division into blocs of states in the EU. One is formed by the economically developed countries of Western Europe, and the other by the developing state of the Eastern EU. Different groups of countries react differently to crises and changes, including in terms of energy consumption. In our research, such divisions were visible.

Gherghina et al. [119] examined the nexus between the main forms of transport, related investments, specific air pollutants and sustainable economic growth. The research concerned the EU countries in 1990–2016. They found that it is important to invest in modern transport infrastructure that facilitates the use of more energy-efficient methods and alternative solutions that positively impact the economy while minimizing negative externalities. This study covered the EU area. Based on our research, it can also be concluded that the key is the use of more energy-efficient methods and alternative technologies in transport. Then, energy consumption in transport will increase less than proportionally to GDP growth.

#### **5. Conclusions and Recommendations**

#### *5.1. Conclusions*

The conducted research allows for a few generalizations.


#### *5.2. Recommendations*

The relationship between energy consumption in transport and the economic situation has not been the subject of systematic research. There are no studies on the association between energy absorption in transport per capita and the level of economic development measured in GDP per capita. The authors found only one project of this type. Furthermore, there were no such studies related to the EU.

The limitations in conducting such academic studies are the lack of available current and detailed data on energy consumption in individual modes of transport. A possible direction of further research is linking energy absorption in transport with environmental pollution and economic development. In this case, it should be based on data concerning per capita. Additionally, the investigation of the interconnections between the various modes of transport would be interesting since EU states differ significantly in this respect. Another direction of academic analysis is the examination of dependencies occurring in regions.

**Author Contributions:** Conceptualization, T.R., G.K. and L.O.; methodology, G.K. and L.O.; software G.K. and L.O.; validation T.R.; formal analysis G.K., L.O. and T.R.; investigation, T.R. and G.K.; resources, T.R., A.B.-B., G.K. and L.O.; data curation, G.K., T.R. and L.O.; writing—original draft preparation, T.R., G.K., L.O., K.W., M.R., H.S., K.M., P.B. and A.B.-B.; writing—review and editing, T.R., G.K., L.O., K.W., M.R., H.S., K.M., P.B. and A.B.-B.; visualization, G.K.; supervision, L.O. and T.R.; project administration, L.O.; funding acquisition, L.O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

