Implication of the EU Countries’ Energy Policy Concerning Scenarios Affecting the Air Quality Improvement

Abstract

Energy policy has a significant impact on the state of the environment and, therefore, on residents’ health and life expectancy, especially in highly urbanized areas. Reducing emissions is currently one of the necessary actions that must be taken at the scale of individual countries to ensure sustainable development. The article aims to identify the best ways to shape energy policy by evaluating development scenarios for air protection and their environmental impact. The realization of the goal is based on the data included in three groups: (1) Economic factors, Health factors, and Demographic factors; (2) Clima-e related economic losses, Renewable Energy sources in electricity, heating, and cooling, Premature deaths due to exposure to fine particulate matter (PM2.5), Health impacts of air pollution, Population change; (3) Demographic balance and crude rates at the national level, GDP per capita in purchasing power PPS, GDP, and principal components; covering 36 EU countries in 2019 and 2021. The study proposes an advanced methodology for assessing development strategies by integrating the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and Bayesian networks (BN) and incorporating them into a multicriteria decision-making (MCDM) support system. The TOPSIS model based on BN allowed for the illustration of the features of many criteria and the identification of relationships between scenarios, allowing for selecting the best way to develop energy policy. The results showed a 60.39% chance of achieving success in extending the life of residents by five years. At the same time, the most favorable development path was the scenario promoting activities aimed at reducing air pollution by introducing renewable energy sources to produce energy used for lighting and preparing domestic hot water urban areas. By presenting possible scenarios and the probability of success, it is possible to achieve the goal of practical energy policy at the level of the country and individual European cities and also by extending the life of city inhabitants, as presented by the authors in this study.

1. Introduction

Observed climate changes are the result of uncontrolled and unsustainable development and urbanization. Rapid urbanization has caused various problems, such as high greenhouse gas emissions, excessive consumption of fossil fuels, and environmental pollution, further deteriorating residents’ health and quality of life [1,2]. The ongoing climate change affects many aspects of the lives of Europeans in a complex and interconnected way [3]. The impact of climate change on health may be direct (e.g., exposure to extreme and unusual temperatures, drought, and floods) or indirect (e.g., through changes in the epidemiology of infectious diseases, ecology, and differences in the availability of medical services as well as in the quality of food, water, and air) [4]. The effects of climate change additionally intensify the negative impact of biological, ecological, and socio-political factors on residents. In the case of European Union countries (EU), susceptibility to climate change is increased by factors such as low demographic growth and aging of society, high urbanization (further intensifying the heat island phenomenon), and mass inflow of people (immigration) as well as slow economic growth.

Finding a way to develop an energy policy may be a way to balance the positive and negative impact of energy on society. Universal access to clean energy can eliminate its previously adverse effects on society, i.e., excessive extraction and combustion of fossil fuels, the risk of climate change, and environmental degradation [5]. The development strategy of countries prioritizing economic growth, accompanied by an increase in the consumption of energy and fossil fuels over environmental issues, leads to significant environmental pollution [6]. There is no consensus that air pollution is an important environmental risk to human health. However, improvements in air quality worldwide have been associated with alleviating numerous health problems, including respiratory infections, cardiovascular disease, and lung cancer [7]. Deteriorating air quality seriously threatens public health in developed and developing countries [8] and degrades the population’s quality of life and well-being [9]. Air quality in cities is crucial for ensuring overall well-being. It requires accurate forecasting of pollutants and, due to the impact on the health of residents, the selection of emission reduction strategies [10]. However, increased air pollution remains beyond individual control and falls on the public sector [11]. Therefore, in this article, an attempt was made to link the air quality, the economic situation, and the health of residents of EU countries in terms of life extension and identification of the most favorable ways of developing cities’ energy policies. The study is essential for the energy transformation policy and, correlated with it, health policy in European countries, with particular emphasis on urbanized areas. The wide-ranging study covers 36 European countries and extends its activities to city dwellers, over 400 million in Europe. Therefore, the authors believe that the approach presented is essential for the economies of these countries, the energy sector, and health care.

The article poses research questions:

Is there a chance of achieving success in extending the life of city residents by five years (reducing/postponing premature death caused by air pollution)?

Which of the three scenarios studied (Figure 1) considers the sustainable development of urban areas to be the most favorable development path? The scenarios are as follows:

–

ScI: Reducing air pollution by introducing renewable energy sources to produce energy used for lighting and preparing hot water.

–

ScII: The gross domestic product will increase (the purchasing power of money will increase), enabling wider access to paid medical services.

–

ScIII: Reducing air pollution by introducing renewable energy sources to produce energy for cooling buildings in connection with upcoming climate changes (warming/overheating).

The study presented in this article proposes a methodology for assessing the ways of developing the energy policy of European countries in terms of extending the life of city residents, based on a combination of Bayesian networks (BN) and TOPSIS and enabling the indication of the degree of probability of success of the adopted scenario. The framework and parameters of the proposed method are based on actual data and expert assessment, and 36 European countries were analyzed as a case study. The rest of the article has the following structure: Section 2 presents a literature review focused on the construction of sustainable energy policies and the associated improvement in the quality of life and decision-making, Section 3 presents the research method and data sources, Section 4 presents the obtained results, and Section 5 and Section 6—discussion and research conclusions.

2. Literature Review

The review synthesizes the current knowledge on the impact of climate change on the health of people living in Europe (EU), with particular emphasis on exposure to extreme heat, water scarcity (drought), air pollution, infectious diseases, and the health of people immigrants. In addition, the authors identify gaps in research knowledge that, if addressed, will help us better understand and monitor the impacts of climate change on human health. The literature does not contain any targeted attempts to calculate the effect of energy policy on air quality in cities, contributing to the extension of the residents’ lives.

In 2015, the United Nations approved the 2030 Agenda and adopted 17 Sustainable Development Goals (SDGs), which: (SDG 7)—achieving clean and affordable energy, (SDG 11)—promoting sustainable cities, and (SDG 13) atmospheric. Energy production based on fossil fuels harms the environment due to the combustion process and the content of large amounts of pollutants that cause smog and acid rain, greenhouse gases, and the destruction of land and water ecosystems [12]. Clean energy has many benefits for the environment, and electricity based on renewable energy sources such as wind or solar energy does not involve carbon dioxide emissions. Simionescu and Cifuentes-Faura [13] assessed the impact of household electricity prices and innovations on CO₂ emissions in the energy supply and waste sectors in the V4 countries (Hungary, Czech Republic, Poland, and Slovakia) in 2010–2021. They observed that higher electricity prices reduce pollution and that economic development increases air pollution in this group of countries.

The growing energy consumption associated with population growth forces the definition of energy development policies or the fight against the effects of pollution, sometimes in the form of spatial strategies aimed at combating low air quality [14,15]. The priority given to spatial policy aimed at improving air quality should be the design of redeveloped and newly developed areas, which will reduce energy demand [16]. Spatial development may be multidimensional and refers to the smaller amount of energy needed for the efficient functioning of households [17] and lower transport needs resulting from the advantages of the 15-minute city [18]. In the case of those parts of urbanized areas where the existing spatial structure will be used, appropriate energy policy ensures sustainable energy use coincides with acceptable social quality of life [19]. Even though a significant part of the municipal and industrial economy is still based on energy from non-renewable sources, due to the thermal condition of buildings throughout the country and the lack of a highly effective alternative to these energy sources, a surplus occurs in the spring and summer periods. Clean heat energy should be used regularly and stored for later use in winter [20]. However, the choice of the development path should not be accidental—access to clean energy should be ensured, which improves living conditions in developed and developing countries and provides their sustainable development [21,22,23].

In recent years, the relationship between renewable energy and economic performance has attracted the interest of many researchers [24,25,26,27], but most analyses focused on economic growth and environmental quality. In the context of sustainable development of urban areas, identifying the interdependencies between urbanization and ecological environmental performance is becoming increasingly important. It has to do with the challenges related to intensive development, consumption of natural resources, climate change, and the general ecological well-being of cities [28,29]. Analyzing and understanding a just energy transition, urbanization, and climate protection is crucial for developing, assessing, and implementing policies that effectively lead to the sustainable development of areas [30]. Chen and Dagestani [1] noted that innovative technologies drive clean energy development by examining the relationship between smart city pilot projects that aim to utilize scarce resources, foster social inclusion, and harmonize economic and environmental development. Keles [3] wrote that planners, architects, and scientists must cooperate closely to deal with phenomena such as rapid urbanization, poverty eradication, rational settlement patterns, sanitary and nutritional problems, exclusions, growth, and development of cities friendly to the environment, not only in developing countries but also in developed countries.

The relationship between renewable energy consumption and economic growth is a relatively new research topic that has gained interest from scientists and policymakers in recent years. The results of many empirical studies support bidirectional causality, both in short- and long-term analyses. Al-mulali et al. [31], using modified least squares (FMOLS), obtained results for 108 countries classified as low-income, lower-middle-income, upper-middle-income, and high-income for the period 1980–2009. Empirical results for most countries (79%) showed a positive bidirectional long-term relationship between renewable energy consumption and GDP growth (confirming the feedback loop).

Wang et al. [32] showed a positive relationship between renewable energy consumption and life expectancy. They determined that life expectancy is also the highest in the high-income group, which means that renewable energy consumption is more helpful in increasing life expectancy in high-income countries [32]. A causal relationship also exists between the high dependence on imported fuels, the need to reduce greenhouse gas (GHG) emissions, and the need to develop a low-carbon economy. The authors [33] draw attention, with Polish conditions, to the economic, ecological, and environmental aspects of the production and use of bioethanol in the context of seeking a balance between the quality and harmony of the natural environment, quality of life, and economic development.

Another study on the presence of greenhouse gases (GHGs) and the potential for reducing GHGs measured by CO₂ emissions through renewable energy development in Malaysia between 1970 and 2011 confirmed that renewable energy production significantly negatively impacts CO₂ emissions [34]. Soukiazis et al. [35] analyzed the connections between the sustainable level of economic development, renewable energy consumption, and atmospheric pollution for 28 OECD countries. The interrelationships between these fundamental variables have demonstrated that deploying renewable energy requires higher human capital skills and reallocating resources to finance projects, developing new ways of achieving cleaner energy. Renewable energy significantly improves economic conditions and sustainable development [35].

Systems supporting decision-making based on many criteria allow decision-makers to take a comprehensive approach to solving problems with many interconnected criteria. The criteria based on which the decision-maker evaluates alternative decisions that can be made may be numerical, linguistic, and characterized by various units of measurement and introduced weights [36,37]. Issues related to multicriteria decision support are occurring in several economic and scientific problems in multiple fields of knowledge [38]. MCDM (multicriteria decision-making) has found applications in solving issues related to economics [39] and process optimization [40,41], choosing the optimal location [1], cybersecurity [42], choosing the appropriate technology [43,44] or in risk assessment [11,45].

MCDM (multicriteria decision-making) has also been used to make decisions and evaluate alternative strategies in energy planning and sustainable use of natural resources and energy [46,47]. Solangi et al. [48], pointing out the need to introduce RES into a sustainable energy management system, identified barriers to the introduction of RES and used the Analytical Hierarchy Process (AHP) and Fuzzy Technique for Order of Preference by Similarity to Ideal Solution, assessed the sustainable development strategy [49]. Using the TOPSIS method, the hierarchization of areas in autonomous communities in Spain was carried out regarding their capabilities and the state of use of RES. The authors emphasize the need to achieve energy self-sufficiency in societies, which should allow for the improvement of social well-being and the development of sustainable energy systems [50]. In turn, the IF-TOPSIS method analyzed the best energy options for renewable energy as part of Turkey’s sustainable development, pointing to solar energy as the best renewable energy source in this country [51]. The TOPSIS method, in conjunction with artificial neural networks, was used to optimize jet ventilation for space heating, which created a framework for designing and controlling ventilation parameters and using the approach in various ventilation scenarios [52]. Ahmad et al. demonstrated the predictive ability of multicriteria models to evaluate solar systems. They used, among others, the TOPSIS method to determine the efficiency and hierarchization of the analyzed systems [53]. However, the possibility of using multicriteria decision support in assessing strategies and identifying solutions to ensure the reliability of energy supplies is presented in [54]. The study by [55] illustrates the use of TOPSIS to classify pellets for energy purposes in Sweden produced from unused forest and agricultural biomass. The authors analyzed various production scenarios and showed that spruce and pine sawdust pellets are the best solution in light of the analyzed criteria related to the state’s energy and climate policy [55]. In turn, Chisale and Lee pointed out the possibility of using the fuzzy TOPSIS method to identify barriers affecting the development of renewable energy sources in Malawi [56]. A similar approach to identifying obstacles to introducing renewable energy was presented in Ghana by Asante et al., who showed that direct actions such as education and training are the most appropriate strategies to remove barriers to renewable energy [57]. The TOPSIS method was also used to evaluate energy strategies used in industry to improve energy efficiency, minimize energy losses, and reduce environmental problems [58,59].

Bayesian networks are widely used to solve problems involving risk, robustness, and probabilistic relationships between different categories of variables [60,61,62]. Combining the advantages of the Bayesian network and the TOPSIS method, a methodology was developed to assess the efficiency of services and compare the degree of risk of accidents [63]. Modeling the impact of air pollution on urban economic development was studied by Wang et al., who used Bayesian regulation, neon networks, and the TOPSIS method to assess many factors affecting air pollution and economic growth [64]. The use of the Bayesian (probabilistic) approach, including the use of Bayesian networks, led to the development of a new approach called the B-TOPISIS method [43,54]. The method is based on a hierarchical model in which the values of the obtained result variables can be divided into conditional dependence or independence. In this way, the model allows for weight estimation based on the relationships between variables. Variables with greater dependency are described as having higher weights. This approach can make decisions more precisely based on the received probabilistic data (uncertain data), thereby minimizing the risks associated with the adopted method of supporting the decision-making process. This modeling can take various forms by “enriching” the models without naming them. One example is a model using the TOPSIS approach to evaluate the aquatic environment [54]. In this way, the model was practically used to determine the threat level. The ability to modify TOPISIS methods is vast and depends on what is essential from the research point of view. An example of the flexibility of the TOPSIS method is the BMW-TOPSIS model, which is based on three-stage decision-making based on previously defined parameters.

The empirical literature is extensive, but most studies have focused on the relationship between conventional economic indicators and energy consumption, and few studies have considered the quality of life improvements or societal benefits of renewable energy consumption. On the other hand, most of the approaches used are quantitative, and no attempt has been made to explain the connection mechanism between economic variables, renewable energy consumption, and residents’ quality of life. There are also many studies on the impact of renewable energy on changes in emissions and air pollution [35]. Still, studies on the effect of renewable energy on the quality of life of city residents are relatively rare. To fill the research gap, the authors included renewable energy consumption, economic growth, and premature death rates from air pollution to examine their relationship. To provide a basis for future energy policy based on an extensive review of available knowledge, in terms of selecting appropriate adaptation and mitigation measures, the authors consider choosing a development path based on selected strategies. The article proposes a framework that identifies a development path in the context of life extension, which may depend on energy policy. These include choosing more empirical development strategies to avoid population exposure to climate change and specific health impacts. The development of appropriate methodologies for assessing how climate change affects the ecological determinants of human life and health makes it possible to find relationships between events based on probability theory. One can look for causal relationships by deepening understanding of the effects of long-term exposure to heat stress and air pollution and assessing the interactions between adaptation and mitigation strategies.

The previously cited studies provide evidence for using the BN-based TOPSIS method in the MCDM (multicriteria decision-making) system. However, there is little research on supporting the decision-making process and assessing energy policy based on the combination of Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and Bayesian networks (BN). The novelty of the proposed approach is the use of nine criteria affecting the quality of life and human health to assess the energy policy of European countries comprehensively. The methodology can be used in the decision-making process to evaluate the effectiveness of different strategies and optimize their implementation at both regional and international levels. Another novelty in the presented study is the MCDM (multicriteria decision-making) problem’s modeling, which considers the interconnections between scenarios and provides information for decision-makers on priority actions in individual scenarios. The framework and parameters of the proposed method are based on actual data and expert judgment. The advantage of the proposed approach is that it relies on experts only at the initial research stage. Expert knowledge was used to identify criteria influencing the success of the implementation of defined scenarios and to create proposals for scenarios whose potential was determined. However, experts from fields such as energy (including renewable energy), construction, spatial planning, architecture, health sciences, environmental protection, and shaping were used to ensure the high quality of the research. The experts represented research units, universities, city decision-makers, non-governmental organizations, and healthcare units. At the same time, based on the collected data, 36 European countries were analyzed as a case study. The study involved separating values for the following criteria: Climate-related economic losses, Renewable Energy sources in electricity, heating, and cooling, Premature deaths due to exposure to fine particulate matter (PM2.5), Health impacts of air pollution, Population change—Demographic balance and crude rates at the national level, GDP per capita in purchasing power PPS, GDP, and principal components. Values for the above criteria were obtained directly from publicly available national and European databases, which were then transformed into values relating to the population living in European cities.

3. Materials and Methods

The article uses an approach that integrates several elements to identify the best scenario for the development of energy policy paths in terms of factors affecting the extension of the life of residents. In the first stage, a database was built and fed with information available in national databases and EUROSTAT [65]. In the next step, analyses were carried out using a Bayesian network. Then, multicriteria decision-making in the TOPSIS model was used to identify the importance of individual criteria and the relationships between scenarios. Ultimately, it allowed us to choose the best way to develop an energy policy. The research scheme is presented in Figure 2.

3.1. Casy Study

According to the first global report on renewable energy in cities [66], cities play a crucial role in transforming energy. It is in cities that diverse transport systems intersect, and the demand for energy for heating and cooling is one of the most energy-intensive. According to the report, more than half of the world’s population lives in cities, responsible for two-thirds of global energy demand and about 75% of global CO₂ emissions.

Taking into account the above, the authors, in their research, also aimed to indicate the dependence of the extension of the life of city inhabitants on the energy policy pursued in Europe. For this purpose, based on available reports ([67] Net Zero Roadmap: A Global Pathway to Keep the 1.5 °C Goal in Reach 2023) and scientific publications [68,69,70], criteria that affect human life and health were indicated with three areas of urban development: economic, health and demographic, which, in the authors’ opinion, create an image of global actions in the field of climate policy. The complete set of criteria is shown in Figure 3.

The research was based on analyzing criteria concerning cities as units with the greatest potential in climate policy modeling. The criteria values for each of the 37 European countries were retrieved from publicly available national and European databases [65]. In the first step, the values for each criterion were indicated at the level of European countries and then converted into values related to the population living in European cities. The data set developed this way was the analysis’s starting point. At the same time, it eliminated possible differences between the compared towns in European countries.

Each identified criteria defines the effects of activities in particular areas of European energy policy. Detailed action scenarios and mechanisms were developed, the implementation of which depended on selected criteria to assess the possibility of extending the life of city residents. In the next step, the weight values of these criteria shaping individual scenarios were determined based on expert knowledge.

3.2. Bayesian Networks

Bayesian networks are a highly developed data analysis tool that offers several advantages:

–

Intuitive representation (Bayesian networks present relationships between variables in a graphical form, which facilitates understanding and interpretation of the model);

–

Flexibility (they can combine expert knowledge with empirical data, which allows the creation of hybrid models.

–

Uncertainty handling (use probability theory to model uncertainty and incomplete data).

–

Bidirectional reasoning (they enable predictive reasoning—from causes to effects and diagnostic reasoning—from effects to causes).

The main aim of the work was to determine the probability of success, which was defined as extending the life of city residents by five years to the original values determined based on previous research. The authors conducted extensive analyses, leading to the selection of three key scenarios related to the energy policies of developed and developing countries. These scenarios resulted from meeting specific criteria or combinations that had to exist to materialize one of the defined event scenarios. Finally, nine criteria were used for further analysis and model construction, divided into three categories (described in detail in Section 3.1 of the work and shown in Figure 3).

The authors proposed three different event scenarios that could lead to success (Figure 1). However, it was assumed that each occurrence is independent at the current research stage. Finally, it was also concluded that the event of any defined scenarios of extending the life of city residents by five years can be understood as a success. Therefore, all combinations that led to a defined success were necessary, regardless of the possibility of implementing one, two, or even three event scenarios in total.

An empirical approach was used to build the Bayesian network, which is commonly used to analyze new research problems [71]. The network topology was based on a simple relationship related to the presence of defined criteria values, the materialization of scenarios, and the implementation of success. It is worth emphasizing that the probability of an event indicates the potential of a given criterion to influence subsequent scenarios. The description of the requirements in linguistic variables suggests the degree of use (implementation) of a given measure in European cities. The relationship is presented in Figure 4. The use of this technique as part of the developed methodology was related to the probabilistic approach to the discussed topic. Following the adopted assumption, the team, based on available data, determined factors relating to the energy policy of cities. Subsequently, a system of various variants of factors was proposed, defining scenarios of events that could occur and affect the quality of human life in the context of activities that could be implemented within the framework of cities’ energy policy. Based on the scenarios, the goal to be pursued was defined, i.e., “a chance to extend the life of city residents by five years”.

3.3. MCDM and TOPSIS

One of the most famous MCDM (multicriteria decision-making) methods is the TOPSIS method, i.e., a method that determines the similarity to the ideal solution. TOPSIS allows you to choose synthetic scores, identified by the distance of each solution (scenario) from the optimal and anti-optimal solution [72,73]. In the classic TOPSIS method, when evaluating and comparing individual variants, it was determined the distance of the m-dimensional Euclidean space (where m is identified as the number of variants) between the value vectors describing a given variant and the vectors corresponding to the optimal and anti-optimal variant. The best variant is the one for which the value vector has the smallest distance from the vector of the optimal variant and the largest from the vector of the anti-optimal variant simultaneously. The TOPSIS method procedure is performed using the steps presented below.

Step 1. A decision matrix creation. Based on the collected data and as a result of the Bayesian network, a decision matrix X is created consisting of m variants (scenarios) and n criteria:

$$X={\left[x_{ij}\right]}_{m\times n}$$

(1)

Step 2. Creation of a standardized decision matrix. Having built the decision matrix X, it can normalize the values of individual variants in terms of separate criteria. It allows the development of a standardized decision matrix R defined as:

$$R={\left[r_{ij}\right]}_{m\times n}$$

(2)

where normalized values r_ij is calculated as:

$$r_{ij}=\frac{x_{ij}}{\sqrt{{\sum }_{i=1}^m{x}_{ij}^2}}$$

(3)

Step 3. Creation of a normalized weighted decision matrix. It can calculate the matrix of normalized weighted values of individual warrants V as:

$$V={\left[v_{ij}\right]}_{m\times n}$$

(4)

where the normalized weighted values v_ij are:

$$v_{ij}=w_j{r}_{ij}, \mathrm{where} w_j=\frac{w_j}{{\sum }_{j=1}^n{w}_j}$$

(5)

It should be noted that w_j is the initial weight of individual criteria determined based on the results of the Bayesian network. Therefore, each weight depends on the probability of success in extending the quality of life if a given scenario is implemented.

Step 4. Determining the optimal and anti-optimal solution. Then, using the adjusted criteria values, it was determined the optimal solution (variant) S⁺:

$$S^{+}=\left\{\begin{array}{cccc}{s}_1^{+}& s_2^{+}& \dots & s_n^{+}\end{array}\right\}$$

(6)

where $s_j^{+}=\left(\left(\underset{i}{\max }{v}_{ij}|j\in C_{+}\right),\left(\underset{i}{\min }{v}_{ij}|j\in C_{-}\right)\right)$ and anti-optimal variant S⁻:

$$S^{-}=\left\{\begin{array}{cccc}{s}_1^{-}& s_2^{-}& \dots & s_n^{-}\end{array}\right\}$$

(7)

where $s_j^{-}=\left(\left(\underset{i}{\min }{v}_{ij}|j\in C_{+}\right),\left(\underset{i}{\max }{v}_{ij}|j\in C_{-}\right)\right)$. In the above equations $s_j^{+}$ and $s_j^{-}$ are values describing the optimal and anti-optimal solution in the context of the j-th criterion. However, C₊ and C₋ are subsets of criteria that have a positive and negative impact, respectively.

Step 5. Calculation of the distance of each variant from the optimal and anti-optimal solution. Once the optimal and anti-optimal solutions are determined, the distances can be calculated D⁺ and D⁻- between them and the individual analyzed variants. These distances were determined as follows:

$$D^{+}=\sqrt{{\sum }_{i=1}^n{\left(v_{ij}-s_j^{+}\right)}^2}$$

(8)

$$D^{-}=\sqrt{{\sum }_{i=1}^n{\left(v_{ij}-s_j^{-}\right)}^2}$$

(9)

Step 6. Calculation of similarity indices of individual variants and creation of a variant ranking. Based on the distances calculated in step 5, the ranking coefficient for individual variants (scenarios) is calculated, which can be defined as:

$$T_i=\frac{D_i^{-}}{D_i^{-}+D_i^{+}}i=1,\dots ,m$$

(10)

The TOPSIS procedure ends with the creation of a ranking of variants in descending order of T_i values.

4. Results

When determining the probability of success of individual energy scenarios analyzed in the article, data was used and obtained as described in Section 3.1 and presented in an aggregated form with individual analysis criteria in

Table 1. Data accepted for analysis (Eurostat—years 2019, 2021).

Country	Criteria
	Economic Factors			Health Factors				Demographic Factors
	E1	E2	E3	H1	H2	H3	H4	D1	D2
Finland	4.7672	16.20771	21.56149	1024	92	2	3,034,855	87	191,886
Sweden	2.6314	31.03864	28.14322	2890	292	3	5,626,602	95	381,496
Estonia	0.6549	12.0294	25.14243	516	45	3	728,651	66	22,212
Ireland	8.1302	14.92523	2.12134	3941	328	7	2,697,332	151	285,086
Iceland	11.1274	40.84707	39.87947	288	27	8	196,345	101	17,622
Luxembourg	155.0402	5.82897	5.29638	596	52	8	337,642	201	49,899
Norway	11.2926	46.6047	13.49884	6789	673	13	2,930,517	118	292,104
Portugal	0.1475	23.95753	17.49921	14,548	1383	14	5,652,139	63	171,500
France	49.6662	10.25656	9.92692	119,585	10,326	16	37,009,759	85	1,950,108
Denmark	13.9948	25.68527	17.0273	10,307	944	17	3,193,345	101	247,621
Spain	33.6536	18.84442	7.13318	91,029	8106	18	25,815,383	73	996,410
Netherlands	21.0984	12.46154	3.1652	36,235	3332	19	9,505,190	102	650,444
Switzerland	22.0778	25.88412	11.931	18,928	1778	21	4,699,490	122	515,555
Malta	0.2124	3.95855	12.85596	1263	104	21	271,457	83	11,349
Belgium	507.1227	10.66533	3.78881	26,722	2496	22	6,300,535	94	382,941
Germany	246.1362	17.90921	6.32958	200,716	18,898	23	45,660,567	97	2,779,288
Austria	45.0229	31.23585	14.54721	20,923	2023	23	4,872,326	101	317,736
Cyprus	3.3217	6.0844	16.95063	3389	320	26	481,744	74	18,541
Türkiye	3.7347	12.136	12.3	306,846	23,201	28	45,102,135	47	542,544
Latvia	1.1505	21.07277	23.52498	6278	552	29	1,055,982	55	24,543
Slovenia	0.0295	14.34016	14.43815	7129	629	30	1,144,499	71	38,826
Slovakia	3.6639	9.17129	8.00361	24,246	1905	35	2,997,732	57	75,543
Czechia	30.8098	5.96304	9.91503	42,730	3819	36	5,857,390	74	180,491
Lithuania	0.472	8.72398	19.93707	11,949	1035	37	1,536,801	67	39,133
Italy	6.2717	14.75836	8.08233	234,093	23,467	39	32,899,170	78	1,437,319
Croatia	1.1918	21.92311	15.59148	20,027	1909	47	2,241,935	54	44,516
Greece	31.5591	14.73294	12.76986	52,120	5215	48	5,898,529	53	146,681
Hungary	1.3806	5.6006	7.35089	54,350	4766	49	5,375,016	58	117,244
Poland	0.3835	7.03806	8.62271	248,853	19,610	52	20,885,047	58	426,004
Romania	0.7906	17.42049	10.03803	119,639	10,160	52	10,677,952	56	179,343
Albania	2.1122	38.71261	8.692	23,398	2186	76	1,574,335	24	11,003
Bulgaria	0.1416	7.7039	10.50625	58,669	5450	78	3,850,021	42	49,247
Montenegro	1.3511	24.78942	26.03418	5360	501	80	342,200	40	3961
North Macedonia	0.4838	8.82812	13.24587	18,648	1851	89	1,142,423	30	9010
Serbia	0.6667	12.25777	14.54393	64,702	6261	90	3,830,070	33	36,804
Bosnia and Herzegovina	1.0325	18.99038	21.85628	33,452	3236	92	1,920,610	26	14,637

. The analyzed set included nine criteria (success factors for oceans of scenarios), divided into three sets: economic, environmental, and demographic criteria, which were converted to values relating to the population living in European cities.

The data presented in Table 1 by criteria are for 2019 and 2021: Economic factors 2021, Health factors, and Demographic factors 2019.

Bayesian networks were used to determine the probability of success by extending the life of residents by five years and then selecting the relationship between the criteria adopted for the analysis. Within the defined topology, nodes were defined, representing the criteria, scenarios, and success situations. A total of 13 nodes were defined in the network: nine corresponding to the criteria, three ScI–ScIII Scenarios, and 1 Success. The previously adopted relationship connected all nodes (Figure 4). The states that can occur are also defined in the nodes. Regarding criteria nodes, these states are concerned with the ranges of values that could happen. Their values were estimated considering the possibilities of implementing individual urban energy improvement scenarios. In the case of scenario nodes and success, they determined their occurrence or not (so-called deterministic nodes) [74]. The state values in the criteria nodes were selected based on previous analyses, and their ranges resulted from the possibility of the occurrence of defined event scenarios. Then, the values of the probability of occurrence of events were entered in the nodes, and the tables of conditional probabilities were filled in. Conditional probabilities resulted from the combination of states of individual criteria. They allowed the achievement of defined energy efficiency improvement scenarios, which successfully extended the life of city residents by five years. The consequences causing the scenarios to occur are presented in

Table 2. Presentation of combinations and variants in scenarios.

Scenario	The Variety of Variants That Cause a Scenario to Occur
ScI	Renewable Energy Share (Low or Medium); GDP per capita in purchasing power (High).
ScII	Premature death from particle pollution <2.5 μm (Medium or High); GDP per capita in purchasing power (Low).
ScIII	Renewable Energy Share (heating and cooling) (Low or Medium); GDP per capita in purchasing power (High).

.

Calculations were performed after entering the probabilities into the network. Based on the collected data, the relationships between the scenarios were analyzed using a Bayesian network, and the relationship between the individual criteria was determined. It was assumed that the values obtained in the nodes of individual criteria would be used in the next step as weights for the analyzed criteria in the TOPSIS method. The results of the Bayesian network are shown in Figure 4.

The constructed network made it possible to observe the probabilities of materialization of individual event scenarios based on defined input parameters based on specific events in nodes. The probability of occurrence of ScI is 4.97%, the occurrence of ScII is 25.20%, and the occurrence of ScIII is 35.79%. Next, the probability of achieving success, i.e., extending the life of city residents by five years, is 61.26%. It can, therefore, be concluded that, based on the analyses, there is a greater probability of extending the life of city residents than not. The network model can also be used to analyze the current situation in a given European country by modeling the existing situation and introducing appropriate variants of nodes. In this way, it is possible to estimate the probability of the occurrence of individual scenarios more precisely, including their success. Individual criteria influencing the life extension of city inhabitants were then analyzed per the TOPSIS method procedure. The weights of each criterion were determined based on the probability of success in the form of life extension obtained using the Bayesian network. The matrix that was generated in the first step of the TOPSIS method, containing assessments consisting of three scenarios and nine criteria, is presented in

Table 3. Decision matrix for three scenarios (ScI–ScIII) and nine criteria.

Analyzed Criteria	Scenario/Weight
	ScI/0.0512	ScII/0.2520	ScIII/0.3477
E1	0.5944	0.0811	0.5946
E2	0.3783	0.1892	0.3783
E3	0.2163	0.2163	0.2163
H1	0.1621	0.1621	0.1621
H2	0.0810	0.3514	0.0810
H3	0.0810	0.3514	0.0810
H4	0.3784	0.3784	0.3784
D1	0.2973	0.5675	0.5675
D2	0.3243	0.3514	0.3514

. The proposed solution, which combines BN and TOPSIS, considers mutual realizations between the scenarios analyzed in BN. Additionally, the weight of each criterion depends on the initial probability of the impact of individual criteria for achieving success. This approach means that the criterion with a higher probability of influencing success also receives greater weight in the decision-making process in MCDM (multicriteria decision-making).

In the next phases, a normalized decision matrix and a normalized weighted decision matrix were calculated (Table 3), the optimal and anti-optimal solution was determined, and the impact of individual criteria on the implementation of the scenarios was determined based on Euclidean distances, the final result is presented in

Table 4. Normalized weighted decision matrix.

Analyzed Criteria	Scenario
	ScI	ScII	ScIII
E1	0.0360	0.0242	0.1744
E2	0.0229	0.0564	0.1109
E3	0.0131	0.0645	0.0634
H1	0.0098	0.0484	0.0475
H2	0.0049	0.1048	0.0238
H3	0.0049	0.1048	0.0238
H4	0.0229	0.1129	0.1110
D1	0.0180	0.1693	0.1664
D2	0.0197	0.1048	0.1031

.

Considering the results obtained, the chance of extending the life of city residents by five years by implementing an appropriate energy policy, identified by actions in the three scenarios presented, is 60.39%. At the same time, research has shown that the most significant impact on achieving success comes from implementing the ScI scenario (

Table 5. Distance values D⁺ and D⁻, indicator T_i and the ranking place.

Scenario	Values
	D+	D−	Ti	Rank
ScI	0.0118	0.2987	0.9619	1
ScII	0.2519	0.1598	0.3880	2
ScIII	0.2677	0.1147	0.3000	3

), which promotes activities aimed at reducing air pollution by introducing renewable energy sources to produce energy for lighting and preparing hot water.

5. Discussion

The study analyzed possible health implications among the environmental and infrastructure interventions of potential urban energy policy paths expressed in three development scenarios. Given the diversity of implementation techniques and decision models in the articles of other researchers whose research the authors refer to, they provide background reference. Against the background of research analysis by different authors, the most important in Wang et al. [32] was the integration of three factors: renewable energy consumption, economic growth, and life expectancy in the same research system to examine their linear and non-linear impact. Variables also included healthcare spending, urbanization, industrialization, and carbon dioxide emissions [32]. Renewable energy consumption in the high-income group significantly impacted life expectancy. At the same time, the relationship between the two was not significant in the higher-middle-income and lower-middle-income groups. Hence, the suggestion is that renewable energy increases life expectancy in more developed countries. Urbanization affects life expectancy regardless of the coal income group [75].

Compared to Soukiazis’s team, the determinants of CO₂ emissions, the authors first confirm the non-linear relationship between the level of pollution and economic development expressed by an inverse U-shaped quadratic form [35]. It can be found that most of the 28 OECD countries in the sample reached the development threshold (86.4 on a scale of 0 to 100), at which action to reduce air pollution is taken. Standard human capital skills and the use of renewable energy have the expected negative impact on CO₂ emissions, supporting the view that educated people better understand the importance of reducing atmospheric pollution and that the way to solve this problem is to use new sources of cleaner energy problem [76]. As expected, total energy consumption increases CO₂ emissions because most of this energy is fossil-based. The result is expected because using renewable energy sources demands higher competencies in human capital and long-term innovative projects.

When creating the network, it was consciously assumed that the occurrence of any of the scenarios would lead to success. However, the authors believe that the joint occurrence of two or three scenarios may increase the consequences of success, i.e., a further potential positive impact on the health of city residents. Quantifying these consequences and determining their probabilities is one of the elements of additional research. Several considerations should be considered when interpreting this study’s results. First, the results from a future modeling study such as this one depend on the quality of the initial health, energy, economic, and environmental data, which, in some cases, is uncertain. Secondly, the differences between the developing economies of the 27 EU countries, which have different assumptions about the projected key sectors, which are also quite uncertain, could lead to significant differences in results.

Finally, the conditional probabilities in the network were influenced by only four nodes (GDP per capita in purchasing power, share of renewable energy (electricity), share of renewable energy (heating and cooling), and premature compliance due to exposure to particulate matter <2.5 μm). Therefore, the probability tables were extensive, but determining their values involved deciding several mutual dependencies between the selected variants in the nodes. These dependencies were adopted following the analysis of energy policy elements that may affect the lives of residents and lead to the occurrence of the Scenarios. As part of further work, it is planned to determine the mutual impact of factors on the defined Scenarios, including their potential independence from each other.

It should also be emphasized that the proposed approach combining BN and TOPSIS is characterized by high universality and the possibility of being used in other research work requiring identifying the most optimal solution. The proposed approach eliminates the limitations related to the large number of criteria and the inability to determine correlations between alternative scenarios. Additionally, the weights of the criteria are defined in a way that does not require the participation of experts, which is also an essential element affecting the universality of the method.

6. Conclusions

The article presents support for decision-making processes based on publicly available data, the Bayesian network (BN), and the TOPSIS method for calculating the probability of success of scenarios. That approach fills the existing gap in the availability of quantitative methods that can consider urban development goals as a tool to improve air protection. The study is relevant to ongoing energy transition policies and, correlated with them, health policies in European countries, with a particular focus on urban areas. The framework and parameters of the proposed method are based on:

–

Actual data (for 36 European countries),

–

Expert evaluation of scenario proposals and identification of criteria affecting the success of their implementation.

Conducting energy policy and defining its strategic goals requires considering many variables. Decision-makers often lack the justification to make a decision, or it is impossible to select (among too many variants) the optimal solution. In this article, the authors defined a clear goal of energy policy (by indicating the best development paths for three different scenarios considering the sustainable development of urban areas), achievable, and meeting the most critical goals of the EU energy policy. There are, i.e., the development of renewable energy reducing emissions of harmful pollutants, mainly CO₂, which contribute to reducing air pollution, translating into health and life expectancy. By presenting possible scenarios and the probability of success, it is possible to achieve the goal of practical energy policy at the level of the country and individual European cities. Such considerations also allow for an analysis of the extension of the city inhabitants’ lives, as presented by the authors in this study. The results show a 60.39% probability of success in extending the life of urban residents by five years, provided that premature death due to air pollution is reduced/postponed. The most favorable development scenario was the ScI, which assumed the reduction of air pollution by introducing renewable energy sources to produce energy used for lighting and preparing hot water.

Therefore, that approach indicates the possibility of estimating and predicting the achievement of success when choosing variant solutions, for example, in the form of the adopted energy development strategy. The results of this study also show the difference in consequences for the decisions of the city’s energy policy. Conclusions of the study may include recommendations for decision-makers shaping development policy in the town and related to the assessment of opportunities for improving the health of residents or, for example, mitigating the effects of climate change. From the perspective of decision-makers creating energy policies for cities to enhance the health of their inhabitants, they gain knowledge regarding recommendations for developing urban development strategies. Additionally, the above research contributed to expanding knowledge related to identifying criteria influencing the development of effective policies consistent with climate protection and sustainable development of urban areas.