**Panel A: Analysis Results in Two-Way Tables**


An observation regarding external reporting is also noteworthy. Compared to the Others segment, E&H companies more often indicated that such actions were taken by the controlling services. Differences are confirmed statistically. Their reasons should probably be sought in the obligations imposed on E&H companies related to reporting to market regulators.

In E&H companies, activities related to the direct involvement of controllers in decision-making processes or preparation of decisions were less frequently recorded. Although there were no differences in the frequency of activities related to process efficiency management, which are rarely undertaken both in the E&H and Others industries, 20% and 25% of respondents, respectively, indicated that controllers undertook this type of activity; the differences in supporting the building of operational efficiency were more visible. Controllers of E&H companies are less likely to engage in activities related to supporting the building of operational efficiency compared to the observations recorded for Others companies.

There were observed surpluses of the observed frequencies over the expected ones for the following categories: (1) analyses are not conducted and (2) up to 25% of working time is devoted to analyses, with the significance level of *p* = <10%. Similarly, it has been noted that E&H companies are less likely to involve controllers in decision-making processes.

Tools supporting the work of controllers were also analysed. The E&H industry is dominated by reporting based on obtaining data from ERP systems and using various tools (spreadsheets, automatically fed spreadsheets, and data warehouses) and their further processing into controlling reports. It has been observed that systems dedicated to management reporting are used less frequently in E&H companies compared to the Others segment.

#### *4.6. Budgeting*

Over 90% of the analysed E&H companies declared that they implement procedures related to budgeting. In this respect, the E&H companies performed slightly better compared to the Others group, where this ratio was 85%. However, one cannot talk about the statistical significance of these differences. Additionally, the results, related to the assessment of the business usefulness of the implemented budget processes, were somewhat surprising, indicated in the synthetic assessment made by the respondents as part of the question whether budget processes fulfil their tasks. For companies from the E&H sector, respondents more often indicated "no" and "rather no" answers compared to the choice of these answers indicated in the Others group. The results obtained in this regard are presented in Table 10.


**Table 10.** Implementation of budgeting processes and assessment of their usefulness.

Source: own study.

In order to assess the reasons for this state of affairs, the occurrence of errors and deficiencies of budgetary processes most often indicated in the literature was analysed. In the questions presented to the respondents, it was verified whether they indicated the following issues:





**Panel A: Analysis Results in Two-Way Tables**


The budget processes implemented in E&H companies do not seem to generate problems related to the mismatch with the company's strategy. However, the situation related to the adequacy of budgets to the changing market situation looks different. E&H respondents in 25% of cases indicated that budgets are mismatched to the current market situation. In the Other group, 10% of respondents indicated a similar mismatch. The conducted analyses confirmed that the mismatch between budgets and the market situation is higher than in companies from the Others sector. Similar results were obtained in the case of the point concerning the lack of flexibility of budgets, which means that they do not adjust to the current market situation. E&H respondents more often pointed to this problem in relation to the Others sector, which confirms the statistical significance of these differences.

The obtained results indicate that the applied budgetary procedures are assessed by the respondents as too detailed and excessively laborious, with an excessive regime of compliance with budget assumptions that does not take into account the current situation. Respondents indicated the occurrence of these problems more often in E&H companies, compared to the responses recorded for the Others group. The differences were confirmed as statistically significant.

Due to the identified challenges of budgeting processes, the scope of created budgets was analysed. Respondents were asked what elements the budget contains, and to indicate those that are present in the company. The question distinguishes: pro-forma balance sheet, depreciation budget, investment budget, direct costs budget, overhead budget, sales budget, salary budget, pro-forma cash flow statement, consolidated budget of costs and revenues in the management layout, consolidated budget in the accounting layout profit and loss account (the so-called pro-forma profit and loss account), budgets of individual organizational units, budgets of individual accounting accounts, quantity budgets (kg, items, m), budgets for individual products, and budgets for individual contractors. Then, how many of the indicated budget elements were used in a given entity was counted. The results obtained for the synthetic variable defined in this way (budget complexity) are presented in Table 12. E&H companies showed a higher content of budgets. The presence of 8 items was most often indicated (average 7.1), while in the Others group, a median of 6 and an average of 6.0 were noted.


**Table 12.** Budget complexity.

Source: own study.

Selected other characteristics of budgeting processes are presented in Table 13 E&H companies used fixed, non-updated annual budgets more often than the Others group, and they used rolling budgets less often, but the significance of these differences was not confirmed with a sufficient level of confidence. There were no differences in the central mode of budgeting or the influence of the manager on budgets when comparing the results for the E&H and Others groups. In turn, clear differences were observed in the processes of budget implementation. No differences were identified in the frequency of reporting deviations, while E&H companies indicate that they are less likely to involve managers in explaining them. E&H companies report the use of IT tools in budgeting more often than Others group.

*Energies* **2023**, *16*, 773


**Table 13.** Characteristics of budgeting processes in the surveyed enterprises.

Source: own study.

#### **5. Discussion and Conclusions**

The Polish energy sector is largely detached from the proven laws of economics and does not follow global development trends. This is important for the economy as it generates very high costs and strives to increase the scale of domestic funding, without introducing changes aimed at adapting to the real economy and global trends. Poland occupies the penultimate place in Europe (per capita) when it comes to obtaining funds for research and development in the field of energy. In addition, the monopolization of the energy sector supports those business models that are based on protecting the state and limiting innovation.

Gradual changes are visible in the economic area, especially in the area of implementing controlling and adjusting it to the specific needs of energy companies. A properly designed controlling system provides the management of an energy company with access to information indicating which product or service is profitable, which creates added value, and which brings losses. The database of information on the costs incurred in connection with the conducted activity enables the optimization of the use of possessed resources and the achievement of the planned goals of this sector of the economy.

Proper management of the effects of operating activities becomes a necessity, and the measurement of the effectiveness of decisions taken is an element of management focused on creating conditions for a stable improvement in economic results. Controlling contributes to supervising the stability of economic results by ensuring that their improvement is linked to the rationalization of the financing costs related to the functioning of the enterprise.

The main purpose of the conducted research was to obtain knowledge on to what extent companies in the energy and heating sector use controlling tools.

The conclusions from original empirical research are:


are used outside of this industry. There were solutions in which more than 10 account segments were used. Most often, this record, as well as cost settlements, are made directly in the accounting system. E&H companies use multiple costing, typically time-based. All this causes the information on costs to become highly complex and, as the conclusions of the conducted research show, it is not easy to report.

• The conducted research indicates that the controlling solutions are of the nature of reporting controlling, not management controlling. Research indicates that controlling tasks rarely focus on support in formulating and implementing decisions. Controllers of E&H companies are less likely to engage in activities related to supporting the building of operational efficiency compared to observations recorded for companies outside this industry. Consuming significantly more time than in other industries, the budgeting process appears to be ineffective in the light of the conducted research. Over 43% of respondents state that budgeting in E&H does not fulfil its tasks and this ratio is definitely higher in the E&H industry than in other groups of enterprises. Detailed budgeting and budgeting procedures are more extensive compared to the other groups. Preparation of one, rigid annual budget, which is not modified during the year, is very common. In most cases, the budget is prepared mainly to fulfil the indicated reporting obligations towards regulators and does not fulfil its management controlling function. This is evidenced by the rare involvement of managers in explaining deviations from budgets, which was revealed in the research.

To improve the efficiency and effectiveness of enterprises operating in the E&H sector, the authors recommend primarily:


of the managerial staff of E&H sector enterprises and it limits benefits resulting from the analysis of controlling data.

• Increasing the emphasis on the use of modern IT tools, business intelligence, and performance management classes supporting the use of management controlling, as well as increasing interest in artificial intelligence algorithms.

**Author Contributions:** Conceptualization, J.N.; Methodology, J.N., M.K.-H. and M.J.K.; Software, M.J.K.; Validation, M.J.K.; Formal analysis, M.J.K.; Investigation, M.J.K.; Writing—original draft, M.K.-H.; Writing—review & editing, J.N. and M.K.-H.; Visualization, J.N.; Supervision, J.N.; Project administration, J.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by a subvention granted to the AGH University of Science and Technology in Krakow, Cracow University of Economics and Wroclaw University of Science and Technology.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available.

**Conflicts of Interest:** The authors declare no conflict of interest. The funder had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

#### *Article* **The Impact of the COVID-19 Pandemic on the Volume of Fuel Supplies to EU Countries**

**Izabela Dembi ´nska <sup>1</sup> , Agnieszka Barczak <sup>2</sup> , Katarzyna Szopik-Depczy ´nska 3,\*, Irena Dul <sup>4</sup> , Adam Koli ´nski <sup>5</sup> and Giuseppe Ioppolo <sup>6</sup>**

	- Institute of Management, University of Szczecin, 71-004 Szczecin, Poland

3


**Abstract:** The COVID-19 pandemic is undoubtedly a destructive factor, strongly affecting the economic fields. From the perspective of the countries affected by the pandemic, almost all sectors of the economy saw declines in economic indicators. First, the lockdown and its social consequences contributed to this. The increasing time perspective since the outbreak of the COVID-19 pandemic implies increasingly more studies analyzing its impact on various economic spheres. The aim of the research is to determine the difference in the level of fuel supplies between a pandemic situation and a situation where a pandemic would not occur. We assumed that the pandemic is a determinant of the decline in fuel supplies. The subjects of the analysis were the following fuels: kerosene-type jet fuel, gas oil and diesel oil, motor gasoline, and oil products. The countries of the European Union were analyzed. Monthly data from 2015–2021 provided by Eurostat were used for the analyses. The forecasts for 2020–2021 were determined using the exponential smoothing method. The assumption was shown to be accurate in the case of kerosene-type jet fuel, gas oil, and diesel oil. In this case, there was a clear drop in the level of supplies. The analysis of forecasts shows that if it were not for the COVID-19 pandemic, in the years 2020–2021, in accordance with the forecasts obtained, approximately 31,495 thousand tons of kerosene-type jet fuel and 11,396 thousand tons of gas oil and diesel oil would have been additionally supplied to the EU countries. For oil products, supply volumes also decreased, but unlike previously mentioned fuels, supply levels had not recovered to pre-pandemic levels by the end of 2021. On the other hand, the forecast of deliveries indicates the volume of 95,683 thousand tons of oil products.

**Keywords:** fuel; fuel supplies; COVID-19; coronavirus; transport; UE; energy economics and management

#### **1. Introduction**

The coronavirus (2019-nCoV-COVID-19) pandemic has affected virtually every economic sector to a greater or lesser extent, and its impact and effects are analyzed in numerous publications, taking into account different perspectives and research objects. Despite this variety of research threads, the authors, however, apart from the statistical reports about the fuel demand, did not find any scientific publications in the field of fuel supply during the pandemic times. A search for publications regarding the fuel supply was carried out in the Scopus database. The authors used various combinations of keywords such as: fuel, fuel supply, COVID-19, coronavirus, transport, and EU, as well as aviation kerosene fuel, diesel, motor gasoline, and petroleum products. The search has been narrowed to journals and articles in the field of research "economics and management", and published in 2020, 2021, or 2022. Using the Scopus database, the authors found only one article in

**Citation:** Dembi ´nska, I.; Barczak, A.; Szopik-Depczy ´nska, K.; Dul, I.; Koli ´nski, A.; Ioppolo, G. The Impact of the COVID-19 Pandemic on the Volume of Fuel Supplies to EU Countries. *Energies* **2022**, *15*, 8439. https://doi.org/10.3390/en15228439

Academic Editors: Bernard Zi˛ebicki and Edyta Bieli ´nska-Dusza

Received: 20 August 2022 Accepted: 9 November 2022 Published: 11 November 2022

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

**Copyright:** © 2022 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/).

the field of social sciences regarding this subject matter. This was the article of Nocera Alves et al. [1]. This article has been cited 2 times so far and is used in our manuscript as a reference. The paper analyses how COVID-19 affected green-fuel supply chain. In comparison with our research problem, the considerations in the cited article concerned Brazil and green fuel. Continuing the analysis of the literature on the fuel market during the pandemic beyond the Scopus base, it was noted that the manuscripts published during those several years were not analyses of the supply factor but only took into account problems related to it in some various areas. The authors grouped them into the following:


In the first area, the authors of analyzed publications showed that the economic slowdown caused by the pandemic resulted in a decrease in fuel consumption and energy demand and made attempts to forecast the trend for the coming years. As a result of the research on the COVID-19 impact on the oil and gas industry conducted by Norouzi [2] for the US market, it was shown that the short-term effect was a decrease in oil consumption by almost 25%, while the long-term effect was a decrease in capital expenditure and R&D investments in the oil and gas market by 30–40%, which resulted in a reduction in the demand for oil extraction from over 800 in 2019 to 265 in 2021. The expected decrease in fuel consumption reasonably implied research that also analyzed the level of CO<sup>2</sup> emissions. The expected decrease in fuel consumption reasonably implied research that also analyzed the level of CO<sup>2</sup> emissions. The natural consequence of introducing the lock down was the reduction of fuel consumption in individual passenger transport, which resulted in a lower level of CO<sup>2</sup> emissions [3]. In the study [4], assessing the impact of the COVID-19 pandemic on global consumption of fossil fuels and CO<sup>2</sup> emissions in the two-year horizon 2020Q1–2021Q4, projections of coal, natural gas, and oil consumption were prepared, depending on GDP growth scenarios on the basis of alternative IMF World Economic Outlook forecasts that were drawn up before and after the outbreak. Forecasts show that fossil fuel consumption and CO<sup>2</sup> emissions will return to, or even exceed, pre-pandemic levels within a two-year horizon, despite significant reductions in the first quarter after the outbreak. Interestingly, there will be stronger growth for emerging economies than for developed economies. It is also worth referring to the Fuels Europe report, prepared on the basis of the International Energy Agency's data for 2021, taking into account the impact of COVID-19 on the demand for individual fuels [5]. It shows that the change in demand did not affect all fuels and all regions equally. In fact, while in the European Union and the United States there was a decrease in the consumption of fossil fuels and an increase in the consumption of renewable energy, in Asian countries the demand for all types of fuels increased in 2021 compared with 2019.

Another group of authors and publications focused on the impact of COVID-19 on the economy, financial markets, the energy market, and the environment and presented the effects of measures introduced by the European Union or selected countries aimed at preventing the spread of the virus while also supporting the economy. Dziembała and Kłos verified the hypothesis that the planned financial instruments, implemented at the EU level, should, to some extent, limit the negative consequences of the pandemic; however, it is necessary to conduct close cooperation between Member States and European institutions in the coordination of actions taken and instruments implemented, conditioning their greater effectiveness. The conclusions reached by the authors indicated that despite the measures introduced by the European Commission to prevent the economic crisis, the burden of counteracting the effects of the epidemic rests mainly with the countries whose governments have introduced anti-crisis packages [6]. Other researchers provided a quantified assessment of the economic and environmental impacts of the blocking measures applied in France over the period of 55 years. They showed that the lockdown led to a

significant decline in economic output by 5% of GDP, but a positive environmental impact, with a reduction of CO<sup>2</sup> emissions by 6.6% in 2020, while concluding that both declines are temporary, and in the coming years they will revert to pre-pandemic levels [7]. Investors' approach and companies' profits from fossil fuels and clean energy were also topics raised by researchers in the context of fuels and energy. The results are not conclusive. One study found that clean energy companies had superiority and greater resilience to the negative effects of the pandemic over fossil fuel companies [8], while another study found that despite viewing clean energy action as more sustainable and less prone to external shocks, fear and coronavirus-induced investor pessimism had also spread to the renewable energy sector, and no more favorable returns on clean energy shares have been observed [9]. The crisis caused by the COVID-19 virus was not indifferent to the natural environment. This relationship was investigated using the Kuznets curve [10], which allowed scientists to formulate a recommendation for economic policy. According to this, an economy in a crisis may take place through investment support and through environmentally friendly legal and organizational solutions [11].

Many authors of the analyzed publications raised the issues related to the supply chains of products and services, as well as related transport or infrastructure in the times of COVID-19. However, no publications dealing with the topic of fuel supplies to the European Union countries have been found, which indicates a literature and research gap in this area. Fuel supply issues were raised, but only in the context of the aviation sector, in an article presenting the relationship between the number of air operations and the volume of aviation fuel delivered to an airport. The analysis showed an increase in the strength of the relationship between the number of air operations performed and the volume of fuel supplies during the pandemic compared with the pre-pandemic period, which brought practical conclusions leading to the optimization of fuel supply chain management in the aviation sector [12]. In general, disruptions in supply chains were investigated by Butt, who showed that manufacturers were fine-tuning production schedules to meet production challenges. Distributors worked with secondary suppliers to meet stock shortages. Finally, procurement firms evaluated the impact of demand by focusing on a short-term demand and supply strategy, preparing for channel changes, opening up additional channels of communication with key customers, and becoming more flexible [13]. Another study found that global supply chains were at that time facing disruptions due to several sources of inherent uncertainty, including natural disasters, war and terrorism, external legal issues, economic and political instability, social and cultural harm, and disease. Flaws in the thencurrent global supply chain have been exposed, resulting in delays, non-delivery, labor shortages, and fluctuations in demand. These supply chain threats have a strong impact on supply chain performance indicators and the magnitude of their impact is amplified in the context of globalization and the COVID-19 pandemic [14].

Many of the supply studies have focused on one of the most problematic sectors, the food sector. The pandemic proved to be an unprecedented challenge for this sector. COVID-19 reduced the security of the food supply chain, increased logistics costs, and radically changed consumer preferences. The industry associated with perishable products, e.g., dairy products, was particularly severely affected by the negative effects of the pandemic. A positive effect of the pandemic is the increased awareness of food waste and the importance of the independence of food production [15–17]. The resilience of supply chains to COVID-19 in the agricultural sector was also examined, showing that agricultural transport systems proved to be extremely robust and able to innovate in real time [18]. Disruptions in supply for medical care [19] were also analyzed, and the impact of the COVID-19 pandemic on the scope and quality of services of courier companies operating on the Polish market was also assessed [20]. Supply chain issues, which may also apply to fuels, were also raised, such as:

• disruptions in freight transport—the changes resulting from the adaptation of the law to the changing situation in the Polish railway sector were analyzed, which led to the conclusion that the introduced legal solutions, although assessed positively, were insufficient [21],

• the problem of preventing COVID-19 infection by employees of hard coal mines through the implementation of safe work systems in mining during a pandemic [22].

Taking into consideration the above, to the best of authors' knowledge, no study has assessed the impacts of COVID-19 on fuel supplies to European Union countries. Therefore, the study presented in this paper is intended to fill the literature and research gap. The aim of the research is, therefore, to determine the difference in the level of fuel supplies between a pandemic situation and a situation where a pandemic would not occur. We assume that the implication of a pandemic would be a decrease in fuel supplies. The impact of the pandemic was quite easy to predict. It had been known from the beginning that the pandemic would have a negative impact on the economy and society. Regarding this context, in our study, we do not ask whether the pandemic will have negative consequences or whether it will cause a decrease in fuel supplies. We are interested mainly in the distribution of changes in fuel supply broken down by fuel groups and in the course of the phenomenon over time. We consider such research justified from the perspective of knowledge about the course of the pandemic, treating the pandemic as a crisis situation. The topic is, therefore, topical and important, although it should be emphasized that our research is cognitive rather than epistemological.

This paper is categorized into different sections as follows: Section 2 provides an explanation of the data and the methodology of this study. The estimated results and discussion are presented in Section 3. The conclusion and limitations of our research are presented in Section 4.

#### **2. Materials and Methodology**

The study covered the volume of fuel supplies to European Union countries, considering kerosene-type jet fuel, gas oil and diesel oil, motor gasoline, and oil products.

Kerosene-type jet fuel is "distillate used for aviation turbine power units. It has the same distillation characteristics at between 150 ◦C and 300 ◦C (generally not above 250 ◦C) and flash point as kerosene. In addition, it has specifications (such as freezing point) which are established by the International Air Transport Association. Includes kerosene blending components. Kerosene type jet fuel is a product aggregate equal to the sum of blended bio jet kerosene (bio jet kerosene in kerosene type jet fuel) and non-bio jet kerosene" [23].

Gas/diesel oil is "primarily a medium distillate distilling at between 180 ◦C and 380 ◦C. Includes blending components. Several grades are available depending on uses. Gas/diesel oil includes on-road diesel oil for diesel compression ignition engines of cars and trucks. Gas/diesel oil includes light heating oil for industrial and commercial uses, marine diesel and diesel used in rail traffic, other gas oil including heavy gas oils which distil at between 380 ◦C and 540 ◦C and which are used as petrochemical feedstocks. Gas/diesel oil is a product aggregate equal to the sum of blended biodiesels (biodiesels in gas/diesel oil) and non-biodiesels" [24].

Motor gasoline "consists of a mixture of light hydrocarbons distilling at between 35 ◦C and 215 ◦C. It is used as a fuel for land-based spark ignition engines. Motor gasoline may include additives, oxygenates and octane enhancers, including lead compounds. Includes motor gasoline blending components (excluding additives/oxygenates), e.g., alkylates, isomerate, reformate, cracked gasoline destined for use as finished motor gasoline. Motor gasoline is a product aggregate equal to the sum of blended biogasoline (biogasoline in motor gasoline) and non-biogasoline" [25].

Oil products is "petroleum products are a product aggregate equal to the sum of refinery gas, ethane, liquefied petroleum gases, naphtha, motor gasoline, aviation gasoline, gasoline type jet fuel, kerosene type jet fuel, other kerosene, gas/diesel oil, fuel oil, white spirit ad SPB, lubricants, bitumen, paraffin waxes, petroleum coke and other products" [26].

Monthly data for 2015–2021 provided by Eurostat were used for the analyses. It should be noted that because Great Britain left the EU structure on 1 February 2020, it was not included in the research. This is due to the fact that the participation of GB in the analyses and its subsequent withdrawal from them may distort forecasts.

The forecasts for 2020–2021 were determined using the exponential smoothing method. Exponential smoothing is a method where the time series of the predicted variable is smoothed using a weighted moving average. This involves replacing each element of the series with a weighted mean of *n* adjacent values, where n is the so-called width of the smoothing window. It should be noted that the weights are determined according to the exponential law. The alpha and delta parameters used in the modeling (the smoothing parameter of the seasonal syntax) were selected using an automatic search for the best value of the parameter using the quasi-Newtonian function minimization procedure (Statistica 13.3). These values are selected on the basis of minimizing the mean square error of the ex-post forecast, i.e., minimizing the sum of squared differences between the empirical values and those forecasted for one period ahead. When carrying out the exponential smoothing process, one can rely on various models, adjusted to the type of components of the predicted time series. Due to the fact that the exponential smoothing method has been extensively described in publicly available publications [27–33], this element was omitted in the study, focusing on the analysis of the obtained results.

The research methodology used might not seem innovative, but the authors believe that due to the fact that the topic itself is an element of novelty in the literature on the subject, there is no need for an innovative statistical method to be used in this respect. We believe that since the literature on the subject lacks publications on fuel supplies to EU countries, the application of the proposed methodology of statistical analyses is sufficient at this point. The authors are already filling the research gap with the topic of the lack of publications in this area toward the number that we would expect.

Due to the fact that the literature contains a multitude of formulas for determining forecast errors, the following formulas were selected for the purposes of the analysis:

Mean error (ME) is derived from the formula:

$$\text{ME} = \frac{1}{\text{m}} \sum\_{\mathbf{t}=1}^{\text{m}} \left( \mathbf{y}\_{\mathbf{t}} - \mathbf{y}\_{\mathbf{t}}^{\text{P}} \right) \tag{1}$$

where:

m—number of observations,

yt—value observed over time,

y P <sup>t</sup> —forecast value at time t.

Mean error is known as the measurement uncertainty or the difference between the measured value and true value.

Mean absolute error (MAE) is derived from the formula:

$$\text{MAE} = \frac{1}{\text{m}} \sum\_{t=1}^{\text{m}} \left| \mathbf{y}\_{t} - \mathbf{y}\_{t}^{\text{P}} \right| \tag{2}$$

Mean absolute error determines by how much, on average, the actual realizations of the forecast variable will deviate—in absolute value—from the forecasts.

Mean percentage error (MPE) is derived from the formula:

$$\text{MPE} = \frac{\sum\_{t=1}^{\text{m}} \text{PE}\_{t}}{\text{m}} \tag{3}$$

where

$$\text{PE}\_{\text{t}} = \frac{\left(\mathbf{y}\_{\text{t}} - \mathbf{y}\_{\text{t}}^{\text{P}}\right)}{\mathbf{y}\_{\text{t}}} \cdot 100 \tag{4}$$

This error determines what percentage of actual implementations of the forecasted variable are forecast errors in the prediction period.

Mean percentage absolute error (MAPE) is derived from the formula: Mean percentage absolute error (MAPE) is derived from the formula:

PE<sup>t</sup> =

*Energies* **2022**, *15*, x FOR PEER REVIEW 6 of 19

variable are forecast errors in the prediction period.

$$\text{MAPE} = \frac{1}{\text{m}} \sum\_{t=1}^{\text{m}} \left| \frac{\mathbf{y}\_{t} - \mathbf{y}\_{t}^{\text{P}}}{\mathbf{y}\_{t}} \right| \cdot 100 \tag{5}$$

∙ 100 (4)

This error determines the average size of forecast errors, expressed as a percentage of the actual values of the forecast variable. This error determines the average size of forecast errors, expressed as a percentage of the actual values of the forecast variable.

(y<sup>t</sup> − y<sup>t</sup> P)

yt

This error determines what percentage of actual implementations of the forecasted

#### **3. Results 3. Results**

where

#### *3.1. Kerosene-Type Jet Fuel Supplies 3.1. Kerosene-Type Jet Fuel Supplies*

In 2015–2019, the volume of aviation fuel deliveries to EU countries was characterized by seasonality, recording increases and decreases in the same months. Stopping or significantly reducing air traffic during the COVID-19 pandemic resulted in a large drop in deliveries in 2020–2021 (Figure 1). Despite the resumption of air traffic after the pandemic stopped, the level of aviation fuel supplies at the end of 2021 was lower by 336.178 thousand tonnes than in 2015, and significantly fewer passengers were handled (Figure 2). The analysis of air traffic in selected EU countries can be found, among others, in work [34]. In 2015–2019, the volume of aviation fuel deliveries to EU countries was characterized by seasonality, recording increases and decreases in the same months. Stopping or significantly reducing air traffic during the COVID-19 pandemic resulted in a large drop in deliveries in 2020–2021 (Figure 1). Despite the resumption of air traffic after the pandemic stopped, the level of aviation fuel supplies at the end of 2021 was lower by 336.178 thousand tonnes than in 2015, and significantly fewer passengers were handled (Figure 2). The analysis of air traffic in selected EU countries can be found, among others, in work [34].

**Figure 1.** The evolution of the volume of kerosene-type jet fuel supplies to EU countries (in thousand tonnes). Source: own elaboration. **Figure 1.** The evolution of the volume of kerosene-type jet fuel supplies to EU countries (in thousand tonnes). Source: own elaboration. *Energies* **2022**, *15*, x FOR PEER REVIEW 7 of 19

Due to the features of the series, we decided to use the exponential smoothing method with additive seasonality. The results obtained with the use of Statistica 13.3 soft-

**Figure 3.** Exponential smoothing of kerosene-type jet fuel supplies. Source: own elaboration.

The values of the forecasts obtained on the basis of this are presented in Table 1, and

**Figure 2.** Number of passengers served (in persons). Source: own elaboration. **Figure 2.** Number of passengers served (in persons). Source: own elaboration.

ware are shown in Figure 3.

the forecast errors are in Table 2.

Due to the features of the series, we decided to use the exponential smoothing method with additive seasonality. The results obtained with the use of Statistica 13.3 software are shown in Figure 3. Due to the features of the series, we decided to use the exponential smoothing method with additive seasonality. The results obtained with the use of Statistica 13.3 software are shown in Figure 3.

**Figure 2.** Number of passengers served (in persons). Source: own elaboration.

2017-01

2017-04

2017-07

2017-10

2018-01

2018-04

2018-07

2018-10

2019-01

2019-04

2019-07

2019-10

2020-01

2020-04

2020-07

2020-10

2021-01

2021-04

2021-07

2021-10

0

2015-01

2015-04

2015-07

2015-10

2016-01

2016-04

2016-07

2016-10

20,000,000 40,000,000 60,000,000 80,000,000 100,000,000 120,000,000 140,000,000

*Energies* **2022**, *15*, x FOR PEER REVIEW 7 of 19

**Figure 3.** Exponential smoothing of kerosene-type jet fuel supplies. Source: own elaboration. **Figure 3.** Exponential smoothing of kerosene-type jet fuel supplies. Source: own elaboration.

The values of the forecasts obtained on the basis of this are presented in Table 1, and the forecast errors are in Table 2. The values of the forecasts obtained on the basis of this are presented in Table 1, and the forecast errors are in Table 2.

**Table 1.** Monthly kerosene-type jet fuel supply forecasts for 2020–2021 (in thousands of tones). Source: own elaboration.


**Table 2.** Forecast errors. Source: own elaboration.


Mean error indicates a slight positive deviation from zero. This means that the obtained forecasts are slightly underestimated. Mean absolute error indicates that the actual deliveries of kerosene-type jet fuel, in absolute terms, will diverge from forecasts by approximately 378.6853 thousand tonnes. It also shows that 15.5358% of actual deliveries of kerosene-type jet fuel are forecast errors in the prediction period (MPE), and the average value of forecast errors is 23.5667% of the volume of kerosene-type jet fuel (MAPE) deliveries. Mean error indicates a slight positive deviation from zero. This means that the obtained forecasts are slightly underestimated. Mean absolute error indicates that the actual deliveries of kerosene-type jet fuel, in absolute terms, will diverge from forecasts by approximately 378.6853 thousand tonnes. It also shows that 15.5358% of actual deliveries of kerosene-type jet fuel are forecast errors in the prediction period (MPE), and the average value of forecast errors is 23.5667% of the volume of kerosene-type jet fuel (MAPE) deliveries.

−100.0174 378.6853 15.5358 23.5667

**Mean Percentage Error (MPE)**

**Mean Percentage Absolute Error (MAPE)**

*Energies* **2022**, *15*, x FOR PEER REVIEW 8 of 19

**Table 2.** Forecast errors. Source: own elaboration.

**Mean Error (ME) Mean Absolute Error** 

**(MAE)**

Source: own elaboration.

**Table 1.** Monthly kerosene-type jet fuel supply forecasts for 2020–2021 (in thousands of tones).

**Period Forecast Period Forecast** 2020-01 1894.612 2021-01 1894.612 2020-02 1739.484 2021-02 1739.484 2020-03 1872.243 2021-03 1872.243 2020-04 1834.793 2021-04 1834.793 2020-05 2119.391 2021-05 2119.391 2020-06 2346.762 2021-06 2346.762 2020-07 2850.966 2021-07 2850.966 2020-08 2929.051 2021-08 2929.051 2020-09 2702.377 2021-09 2702.377 2020-10 2535.197 2021-10 2535.197 2020-11 1988.844 2021-11 1988.844 2020-12 2060.526 2021-12 2060.526

The analysis of the obtained forecasts shows that, had it not been for the COVID-19 pandemic, in the years 2020–2021, in accordance with the forecasts obtained, approximately 31,494.910 thousand tonnes of kerosene-type jet fuel would have been delivered to the EU countries in the years 2020–2021 (Figure 4). The analysis of the obtained forecasts shows that, had it not been for the COVID-19 pandemic, in the years 2020–2021, in accordance with the forecasts obtained, approximately 31,494.910 thousand tonnes of kerosene-type jet fuel would have been delivered to the EU countries in the years 2020–2021 (Figure 4).

**Figure 4.** Comparison of the actual and forecast volumes of delivered kerosene-type jet fuel (in thousand tonnes). Source: own elaboration. **Figure 4.** Comparison of the actual and forecast volumes of delivered kerosene-type jet fuel (in thousand tonnes). Source: own elaboration. *Energies* **2022**, *15*, x FOR PEER REVIEW 9 of 19

#### *3.2. Gas Oil and Diesel Oil Supplies to EU Countries 3.2. Gas Oil and Diesel Oil Supplies to EU Countries*

As in the case of kerosene-type jet fuel deliveries, the volumes of gas oil and diesel oil deliveries decreased significantly as a result of the COVID-19 pandemic. By the end of 2021, the plurality of deliveries of this type of fuel was beginning to return to the pre-pandemic state (Figure 5). As in the case of kerosene-type jet fuel deliveries, the volumes of gas oil and diesel oil deliveries decreased significantly as a result of the COVID-19 pandemic. By the end of 2021, the plurality of deliveries of this type of fuel was beginning to return to the prepandemic state (Figure 5).

**Figure 5.** Changes in the volume of gas oil and diesel oil supplies to EU countries (in thousand tonnes). Source**:** own elaboration. **Figure 5.** Changes in the volume of gas oil and diesel oil supplies to EU countries (in thousand tonnes). Source: own elaboration.

In the case of gas oil and diesel oil supplies, due to the characteristics of the series, we decided to use the exponential smoothing method with additive seasonality. The re-

sults obtained with the use of Statistica 13.3 software are shown in Figure 6.

In the case of gas oil and diesel oil supplies, due to the characteristics of the series, we decided to use the exponential smoothing method with additive seasonality. The results obtained with the use of Statistica 13.3 software are shown in Figure 6. In the case of gas oil and diesel oil supplies, due to the characteristics of the series, we decided to use the exponential smoothing method with additive seasonality. The results obtained with the use of Statistica 13.3 software are shown in Figure 6.

**Figure 5.** Changes in the volume of gas oil and diesel oil supplies to EU countries (in thousand

2018-05

2018-09

2019-01

2019-05

2019-09

2020-01

2020-05

2020-09

2021-01

2021-05

2021-09

2018-01

As in the case of kerosene-type jet fuel deliveries, the volumes of gas oil and diesel oil deliveries decreased significantly as a result of the COVID-19 pandemic. By the end of 2021, the plurality of deliveries of this type of fuel was beginning to return to the pre-

*Energies* **2022**, *15*, x FOR PEER REVIEW 9 of 19

*3.2. Gas Oil and Diesel Oil Supplies to EU Countries*

pandemic state (Figure 5).

tonnes). Source**:** own elaboration.

16,000 17,000 18,000 19,000 20,000 21,000 22,000 23,000 24,000

2015-01

2015-05

2015-09

2016-01

2016-05

2016-09

2017-01

2017-05

2017-09

**Figure 6.** Exponential smoothing of gas oil and diesel oil supplies. Source: own elaboration.

The values of the forecasts obtained on the basis of this are presented in Table 3, and the forecast errors are in Table 4.

**Table 3.** Monthly gas oil and diesel oil supply forecasts for 2020–2021 (in thousand tonnes). Source: own elaboration.


**Table 4.** Forecast errors. Source: own elaboration.


Mean error indicates a slight negative deviation from zero. This means that the obtained forecasts are slightly overestimated. Mean absolute error indicates that the actual deliveries of gas oil and diesel oil, in absolute terms, will diverge from forecasts by approximately 666.4146 thousand tonnes. It also shows that 0.3420% of the actual volume of gas oil and diesel oil supplies are due to forecast errors in the prediction period (MPE), while the average size of forecast errors is 3.3248% of the volume of gas oil and diesel oil (MAPE) supplies.

The analysis of the obtained forecasts shows that if it were not for the COVID-19 pandemic, the EU countries in 2020–2021, according to the forecasts obtained, would have received an additional amount of approximately 11,395.539 thousand tonnes of gas oil and diesel oil (Figure 7). *Energies* **2022**, *15*, x FOR PEER REVIEW 11 of 19 *Energies* **2022**, *15*, x FOR PEER REVIEW 11 of 19

**Figure 7.** Comparison of the actual and forecast volumes of supplied gas oil and diesel oil (in thousand tonnes). Source: own elaboration. **Figure 7.** Comparison of the actual and forecast volumes of supplied gas oil and diesel oil (in thousand tonnes). Source: own elaboration. **Figure 7.** Comparison of the actual and forecast volumes of supplied gas oil and diesel oil (in thousand tonnes). Source: own elaboration.

#### *3.3. Motor Gasoline Supplies to EU Countries 3.3. Motor Gasoline Supplies to EU Countries 3.3. Motor Gasoline Supplies to EU Countries*

Source: own elaboration.

In the case of motor gasoline, the supply volumes did not decrease as a result of the COVID-19 pandemic. Throughout the duration of the COVID-19 pandemic, as well as after its completion, the volumes of deliveries remained at a similar level, taking into account seasonal fluctuations (Figure 8). In the case of motor gasoline, the supply volumes did not decrease as a result of the COVID-19 pandemic. Throughout the duration of the COVID-19 pandemic, as well as after its completion, the volumes of deliveries remained at a similar level, taking into account seasonal fluctuations (Figure 8). In the case of motor gasoline, the supply volumes did not decrease as a result of the COVID-19 pandemic. Throughout the duration of the COVID-19 pandemic, as well as after its completion, the volumes of deliveries remained at a similar level, taking into account seasonal fluctuations (Figure 8).

In this case, the method of exponential smoothing with additive seasonality was also **Figure 8.** Changes in the volume of motor gasoline supplies to EU countries (in thousand tonnes). Source: own elaboration. **Figure 8.** Changes in the volume of motor gasoline supplies to EU countries (in thousand tonnes). Source: own elaboration.

In this case, the method of exponential smoothing with additive seasonality was also used. The results obtained with the use of Statistica 13.3 software are shown in Figure 9.

used. The results obtained with the use of Statistica 13.3 software are shown in Figure 9.

In this case, the method of exponential smoothing with additive seasonality was also used. The results obtained with the use of Statistica 13.3 software are shown in Figure 9. *Energies* **2022**, *15*, x FOR PEER REVIEW 12 of 19

**Figure 9.** Exponential smoothing of motor gasoline supplies. Source: own elaboration. **Figure 9.** Exponential smoothing of motor gasoline supplies. Source: own elaboration.

The values of the forecasts obtained on the basis of this are presented in Table 5, and the forecast errors are in Table 6. The values of the forecasts obtained on the basis of this are presented in Table 5, and the forecast errors are in Table 6.

**Table 5.** Motor gasoline supply forecasts on a monthly basis for 2020–2021 (in thousand tonnes). Source: own elaboration. **Table 5.** Motor gasoline supply forecasts on a monthly basis for 2020–2021 (in thousand tonnes). Source: own elaboration.


**Table 6.** Forecast errors. Source: own elaboration.


39.3798 127.1338 0.5838 2.1901

Mean error indicates a slight positive deviation from zero. This means that the ob-

2020-12 6031.271 2021-12 6031.271


2020-01

2020-02

Mean error indicates a slight positive deviation from zero. This means that the obtained forecasts are slightly underestimated. Mean absolute error indicates that the actual deliveries of motor gasoline, in absolute terms, will deviate from the forecasts by an average of 127.1338 thousand tonnes. This also shows that 0.5838% of the actual deliveries of motor gasoline are forecast errors in the prediction period (MPE), and the average value of forecast errors is 2.1901% of the volume of motor gasoline (MAPE) deliveries. deliveries of motor gasoline, in absolute terms, will deviate from the forecasts by an average of 127.1338 thousand tonnes. This also shows that 0.5838% of the actual deliveries of motor gasoline are forecast errors in the prediction period (MPE), and the average value of forecast errors is 2.1901% of the volume of motor gasoline (MAPE) deliveries. deliveries of motor gasoline, in absolute terms, will deviate from the forecasts by an average of 127.1338 thousand tonnes. This also shows that 0.5838% of the actual deliveries of motor gasoline are forecast errors in the prediction period (MPE), and the average value of forecast errors is 2.1901% of the volume of motor gasoline (MAPE) deliveries.

*Energies* **2022**, *15*, x FOR PEER REVIEW 13 of 19

*Energies* **2022**, *15*, x FOR PEER REVIEW 13 of 19

The analysis of the obtained forecasts shows that despite the apparent maintenance of the trend throughout the entire duration of the COVID-19 pandemic in 2020–2021, according to the forecasts obtained, approximately 11,042.207 thousand tonnes of motor gasoline would have been delivered to the EU countries (Figure 10). The analysis of the obtained forecasts shows that despite the apparent maintenance of the trend throughout the entire duration of the COVID-19 pandemic in 2020–2021, according to the forecasts obtained, approximately 11,042.207 thousand tonnes of motor gasoline would have been delivered to the EU countries (Figure 10). The analysis of the obtained forecasts shows that despite the apparent maintenance of the trend throughout the entire duration of the COVID-19 pandemic in 2020–2021, according to the forecasts obtained, approximately 11,042.207 thousand tonnes of motor gasoline would have been delivered to the EU countries (Figure 10).

**Figure 10.** Changes in the volume of motor gasoline supplies to EU countries (in thousand tonnes). Source: own elaboration. **Figure 10.** Changes in the volume of motor gasoline supplies to EU countries (in thousand tonnes). Source: own elaboration. **Figure 10.** Changes in the volume of motor gasoline supplies to EU countries (in thousand tonnes). Source: own elaboration.

#### *3.4. Deliveries of Oli Products to EU Countries 3.4. Deliveries of Oli Products to EU Countries 3.4. Deliveries of Oli Products to EU Countries*

For oil products, supply volumes have decreased as a result of the COVID-19 pandemic. By the end of 2021, these figures did not return to the pre-pandemic state (Figure 11). For oil products, supply volumes have decreased as a result of the COVID-19 pandemic. By the end of 2021, these figures did not return to the pre-pandemic state (Figure 11). For oil products, supply volumes have decreased as a result of the COVID-19 pandemic. By the end of 2021, these figures did not return to the pre-pandemic state (Figure 11).

**Figure 11.** Shaping the volume of deliveries of oil products to EU countries (in thousand tonnes). Source: own elaboration. **Figure 11.** Shaping the volume of deliveries of oil products to EU countries (in thousand tonnes). Source: own elaboration. **Figure 11.** Shaping the volume of deliveries of oil products to EU countries (in thousand tonnes). Source: own elaboration.

2021-01

2021-05

2021-09

In this case, the method of exponential smoothing with additive seasonality was also used. The results obtained with the use of Statistica 13.3 software are shown in Figure 12. In this case, the method of exponential smoothing with additive seasonality was also used. The results obtained with the use of Statistica 13.3 software are shown in Figure 12.

**Figure 12.** Exponential smoothing of supply volumes of oil products. Source: own elaboration. **Figure 12.** Exponential smoothing of supply volumes of oil products. Source: own elaboration.

The values of the forecasts obtained on the basis of this are presented in Table 7, and the forecast errors are in Table 8. The values of the forecasts obtained on the basis of this are presented in Table 7, and the forecast errors are in Table 8.

**Table 7.** Forecasts of the volume of oil products supplies on a monthly basis for 2020–2021 (in thousand tonnes). Source: own elaboration. **Table 7.** Forecasts of the volume of oil products supplies on a monthly basis for 2020–2021 (in thousand tonnes). Source: own elaboration.


2020-12 40,973.40 2021-12 40,973.40 **Table 8.** Forecast errors. Source: own elaboration.


2020-11 40,545.24 2021-11 40,545.24

**(MAPE)** 36.8596 764.8841 0.0406 1.8740 Mean error indicates a slight positive deviation from zero. This means that the obtained forecasts are slightly underestimated. Mean absolute error indicates that actual oil product

**ror (MPE)**

**(MAE)**

deliveries, in absolute terms, will deviate from forecasts by an average of 764.8841 thousand tonnes. It also shows that 0.0406% of the actual oil product deliveries are due to forecast errors in the prediction period (MPE), while the average forecast error is 1.8740% of the oil product deliveries (MAPE). tained forecasts are slightly underestimated. Mean absolute error indicates that actual oil product deliveries, in absolute terms, will deviate from forecasts by an average of 764.8841 thousand tonnes. It also shows that 0.0406% of the actual oil product deliveries are due to forecast errors in the prediction period (MPE), while the average forecast error is 1.8740% of the oil product deliveries (MAPE).

Mean error indicates a slight positive deviation from zero. This means that the ob-

*Energies* **2022**, *15*, x FOR PEER REVIEW 15 of 19

The analysis of the obtained forecasts shows that if it were not for the COVID-19 pandemic, in the years 2020–2021, in accordance with the forecasts obtained, approximately 95 683.166 thousand tonnes of oil products would have been delivered to the EU countries in the years 2020–2021 (Figure 13). The analysis of the obtained forecasts shows that if it were not for the COVID-19 pandemic, in the years 2020–2021, in accordance with the forecasts obtained, approximately 95 683.166 thousand tonnes of oil products would have been delivered to the EU countries in the years 2020–2021 (Figure 13).

**Figure 13.** Comparison of the actual and forecast volumes of supplied oil products (in thousand tonnes). Source: own elaboration. **Figure 13.** Comparison of the actual and forecast volumes of supplied oil products (in thousand tonnes). Source: own elaboration.

#### *3.5. Discussion 3.5. Discussion*

On 30 January 2020, the World Health Organization (WHO) announced a public health emergency of international concern in connection with the COVID-19 epidemic. At the end of February 2020, WHO issued a recommendation [35] underlining the importance of not imposing any travel or trade restrictions due to the outbreak. Toward the end of the first quarter of 2020, governments around the world began imposing restrictions and then prohibitions on international travel. While these restrictions initially applied primarily to China, where the pandemic began, they quickly spread to other areas as well. On March 11, 2021, the WHO classified the epidemic COVID-19 as a pandemic [36]. The United States announced a suspension of travel from Europe, and a week later, the EU closed its external borders to air passengers [37,38]. As the epidemiological situation worsened, individual countries EU Member States began closing their borders to commercial air traffic from other countries within the EU. By April 2020, 14 Schengen Member States introduced internal border controls. The tourist traffic was closed, as was business travel, which was replaced by on-line business sessions. On 30 January 2020, the World Health Organization (WHO) announced a public health emergency of international concern in connection with the COVID-19 epidemic. At the end of February 2020, WHO issued a recommendation [35] underlining the importance of not imposing any travel or trade restrictions due to the outbreak. Toward the end of the first quarter of 2020, governments around the world began imposing restrictions and then prohibitions on international travel. While these restrictions initially applied primarily to China, where the pandemic began, they quickly spread to other areas as well. On 11 March 2021, the WHO classified the epidemic COVID-19 as a pandemic [36]. The United States announced a suspension of travel from Europe, and a week later, the EU closed its external borders to air passengers [37,38]. As the epidemiological situation worsened, individual countries EU Member States began closing their borders to commercial air traffic from other countries within the EU. By April 2020, 14 Schengen Member States introduced internal border controls. The tourist traffic was closed, as was business travel, which was replaced by on-line business sessions.

While previous crises have slowed down the growth of the aviation industry, none of them led to a complete halt to operations. For example, in 2002, after the terrorist attacks in the United States, air traffic in Europe fell by 2%. The 2009 financial crisis saw a decline of 6.6%, and the eruption of Iceland's Eyjafjallajökull volcano in April 2010 resulted in the cancellation of 111,000 flights [39]. However, it is the COVID-19 pandemic that is unprecedented in history in terms of the magnitude of the effects and the timing of the impact. The International Air Transport Association (IATA) expects a recovery to the pre-pandemic situation no sooner than post-2024. Overall, the number of travelers is projected to While previous crises have slowed down the growth of the aviation industry, none of them led to a complete halt to operations. For example, in 2002, after the terrorist attacks in the United States, air traffic in Europe fell by 2%. The 2009 financial crisis saw a decline of 6.6%, and the eruption of Iceland's Eyjafjallajökull volcano in April 2010 resulted in the cancellation of 111,000 flights [39]. However, it is the COVID-19 pandemic that is unprecedented in history in terms of the magnitude of the effects and the timing of the impact. The International Air Transport Association (IATA) expects a recovery to the prepandemic situation no sooner than post-2024. Overall, the number of travelers is projected to reach 4.0 billion in 2024 (counting multi-sector connecting journeys as one passenger), exceeding pre-COVID-19 levels (103% of the 2019 total) [40].

The lockdown introduced as a result of COVID 19 also influenced the level of mobility in road transport. In this case as well, there was practically no tourist traffic. Restrictions on social distancing and the need to close hotels prevented people from going anywhere. The transition to distance learning and remote work led people to reduce the use of their cars. Interestingly, many companies discovered the advantages of remote work to such an extent that they want to either completely or partially stay with this form of work.

Fuel supplies are a response to the demand for them. It is of a secondary nature, as it results from the demand for transport [41]. This relationship has been highlighted by the pandemic.

An interesting situation occurred with regard to cargo transportation. Despite the freezing of economies in the EU countries, transport companies did not suffer due to the bad economic situation. The main factor behind this situation was the dynamic growth of e-commerce as a result of restrictions in traditional trade. In the EU-27.2, retail sales via mail order houses or the internet increased by 30% in April 2020 compared with April 2019, while total retail sales decreased by 17.9% [42]. The pandemic, together with its limitations, became a strong stimulus for the growth of online purchases, which, in turn, influenced the situation of transport and logistics companies, and consequently also the demand for fuel.

A pandemic is a crisis. It creates uncertainty and unpredictability. Contemporary supply chains, including those in which fuel flows, must be resilient. Resilience is a synergy of flexibility and adaptability [43,44]. The desired state of resilience is security, stability, and functionality. The resilient supply chain is characterized by the ability to reactive response, adaptation, and resilience in changing, unpredictable, and uncertain operating conditions. Unfortunately, analyzing the COVID-19 pandemic in terms of the resilience of fuel supply chains is still an unspoken but necessary discussion.

#### **4. Conclusions**

The COVID-19 pandemic had a significant impact on many sectors of the economy. The fuel sector in the EU was one of the sectors of the economy that felt its direct effects, as it led to massive disruptions to mobility, air traffic, and traditional trade. The aim of our research was to estimate the difference in fuel supplies between the situation where the COVID-19 pandemic occurred and the situation where there was no pandemic. The analysis covered the countries of the European Union. The research was based on the assumption that during the pandemic, fuel supplies had decreased. Our assumption proved to be correct. Only in the case of motor gasoline, the supply volumes did not decrease as a result of the COVID-19 pandemic. Throughout the pandemic, and after its completion, the volumes of deliveries remained at a similar level, taking into account seasonal fluctuations. The decline in fuel supplies was caused mainly by the introduction of lockdown and related social restrictions, which translated into the functioning of almost all sectors of the economy.

This study showed that all lower fuel supplies were, in fact, determined by topdown blockades (hence regional discrepancies) and, thus, "artificially" caused by lower demand. However, the differences between the period—without a pandemic and during a pandemic—are not as great as it seems since the whole world practically stopped. This, in turn, leads to the conclusion that as long as we do not replace fossil fuels with low or zero emission sources, we will not be able to significantly move away from burning fossil fuels (e.g., through energy saving, the pandemic showed that even "forced" did not provide much change) and greenhouse gas emissions.

When reviewing the literature in the research area, it was revealed that there is a huge research gap in this area. There are practically no studies on the impact of COVID-19 on fuel supplies in EU countries. The statistical data are available, however, and this enables various analyses to be performed. The lack of literature is a limitation for our research as no direct comparative analysis is possible. Perhaps a longer time perspective is needed. In addition, as part of the limitations of the study, the research methodology used might not seem innovative, but the authors believe that due to the fact that the topic itself is a novelty, there is no need to use a very innovative method of analysis in this respect. We believe that since the literature on the subject lacks publications on fuel supplies to EU countries, the application of the proposed methodology of statistical analysis is sufficient at this point.

Since the authors find the topic very significant, in terms of further research directions, the plan is to continue the analysis with the use of other statistical methods. The study of the impact of COVID-19 on the volume of fuel supplies has great research potential. In the current study, we wanted to analyze the supply–demand relationship. Having the results of these studies in mind, we will analyze supply and demand in more detail in subsequent studies, and we will certainly pay attention to the segmentation of supply and demand curves. Our research covered the EU countries, but it may be interesting to compare the analyses conducted for other groups of countries, e.g., Asian countries or highly developed countries compared with economically less-developed countries. Forecasts projected for a longer time horizon may also be interesting. The topic of the influence of COVID-19 on various spheres of economic and social life will certainly be both current and attractive for researchers in the coming years.

**Author Contributions:** Conceptualization, I.D. (Izabela Dembi´nska), A.B., K.S.-D., I.D. (Irena Dul), A.K., and G.I.; methodology, A.B.; software, A.B.; validation, Izabela Dembi´nska, A.B., and K.S.-D.; formal analysis, I.D. (Izabela Dembi ´nska), A.B., K.S.-D., I.D. (Irena Dul), A.K., and G.I.; investigation, I.D. (Izabela Dembi´nska), A.B., K.S.-D., I.D. (Irena Dul), A.K., and G.I.; resources, I.D. (Izabela Dembi´nska), A.B., K.S.-D., I.D. (Irena Dul), A.K., and G.I.; data curation, A.B.; writing—original draft preparation, I.D. (Izabela Dembi´nska), A.B., K.S.-D., I.D. (Irena Dul), A.K., and G.I.; writing review and editing, I.D. (Izabela Dembi ´nska), A.B., K.S.-D., A.K., and G.I.; visualization, I.D. (Izabela Dembi ´nska), A.B., and K.S.-D.; supervision, I.D. (Izabela Dembi ´nska) and G.I.; project administration, I.D. (Izabela Dembi ´nska) and K.S.-D.; funding acquisition, A.K. All authors have read and agreed to the published version of the manuscript.

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

**Data Availability Statement:** Not applicable.

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

#### **References**


#### *Review* **Digitalisation of Enterprises in the Energy Sector: Drivers—Business Models—Prospective Directions of Changes**

**Barbara Siuta-Tokarska \*, Sylwia Kruk, Paweł Krzemi ´nski, Agnieszka Thier and Katarzyna Zmija ˙**

Department of Economics and Organization of Enterprises, Institute of Computer Science, Accounting and Controlling, Cracow University of Economics, 31-510 Krakow, Poland **\*** Correspondence: siutab@uek.krakow.pl

**Abstract:** The energy sector has been a pioneer in the use of information and communication technologies for many years, and has undergone enormous changes in recent years as a result of the transition resulting from the fourth industrial revolution. In the paper, we examine and analyse relevant studies and their findings in order to show the current status of research on three selected aspects of digitalisation in energy sector enterprises. The paper sheds light on the diverse drivers influencing the digitalisation processes of energy sector companies. It also provides an overview of business models that are or will soon be implemented in the energy sector thanks to opportunities offered by digitalisation in response to observed trends in the energy market. Finally, it illustrates open research challenges and future dilemmas related to various aspects of energy sector digitalisation. The paper was prepared using the critical literature review method. It covers a large volume of the most recent and relevant literature referring to the three major research areas mentioned above. The literature review allowed us to identify the drivers influencing the digitalisation of energy companies and distinguish between those specific to this sector and those relating to all businesses as a part of the more general phenomena of Industry 4.0 and Industry 5.0. We also show how the digitalisation-based business model innovation presented in the literature empowers new energy producers and consumers through business models. We have also identified the most frequently indicated challenges and dilemmas in the digitalisation of energy companies related to the risk of the destabilisation of the energy market due to decentralisation, new requirements placed on the competences of energy sector workers, the new culture of interaction between energy suppliers and consumers, and the digital security of data used in the energy system.

**Keywords:** digitalisation; drivers; business models; energy sector; changes

#### **1. Introduction**

The problem of the digitalisation of firms is one of the multifaceted and multi-threaded issues which, in the context of the contemporary economy, based on knowledge and modern technologies, are an important and worthwhile discussion. Energy sector companies undergoing digitalisation are strategic entities in the economy, as the energy efficiency improvements occurring with their participation are part of the sustainable development policies of Europe and the world. Therefore, the inclusion of the two above-mentioned areas of analysis is an expression of the scientific and research inquisitiveness of the paper's authors, being part of strategic management.

In this research paper, reference is first made to basic issues such as an explanation of the terms 'digitalisation of enterprises' and 'enterprise 2.0 and 3.0', followed by a review of the literature on the drivers and business models of energy sector companies undergoing digitalisation processes. And thus the bases were created towards the linear, time-based approach towards the directions of changes in light of the emerging challenges and dilemmas in their activities. In the last part of the paper, based on logic and lateral

**Citation:** Siuta-Tokarska, B.; Kruk, S.; Krzemi ´nski, P.; Thier, A.; Zmija, K. ˙ Digitalisation of Enterprises in the Energy Sector: Drivers—Business Models—Prospective Directions of Changes. *Energies* **2022**, *15*, 8962. https://doi.org/10.3390/ en15238962

Academic Editor: Jamie W.G. Turner

Received: 8 October 2022 Accepted: 16 November 2022 Published: 27 November 2022

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

**Copyright:** © 2022 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/).

thinking, the above-mentioned components are combined, presenting an attempt to answer the research questions posed in this Introduction. thinking, the above-mentioned components are combined, presenting an attempt to answer the research questions posed in this Introduction. It should be noted that in the literature on the subject, the term 'digitalisation' does

It should be noted that in the literature on the subject, the term 'digitalisation' does not have one commonly used definition. In the narrow sense, digitalisation is the processing of analogue data into digital form, but in the broader sense, it is a multi-stage process: the identification and selection of documents (information, knowledge), their preparation and ordering, collecting basic metadata, digital conversion, quality control of copies and metadata, providing the user with access to digital documents, the maintenance of digital copies and metadata, backup copies and planning for the future. However, as it applies to firms, this concept is understood as the use of digital technologies related to the configuration of their business models to ensure new opportunities to generate value within the organisation. not have one commonly used definition. In the narrow sense, digitalisation is the processing of analogue data into digital form, but in the broader sense, it is a multi-stage process: the identification and selection of documents (information, knowledge), their preparation and ordering, collecting basic metadata, digital conversion, quality control of copies and metadata, providing the user with access to digital documents, the maintenance of digital copies and metadata, backup copies and planning for the future. However, as it applies to firms, this concept is understood as the use of digital technologies related to the configuration of their business models to ensure new opportunities to generate value within the organisation.

As A. Manterys [1] emphasises, the multitude of realities in which human activities take place, is the constellation of different arrangements and orderings. For the traditional market, a specific dualism can be seen, which, however, is not reflected in the virtual market, which requires analysis in different categories. It should be mentioned here that enterprises, which are market entities, have undergone a kind of evolution (Figure 1). As A. Manterys [1] emphasises, the multitude of realities in which human activities take place, is the constellation of different arrangements and orderings. For the traditional market, a specific dualism can be seen, which, however, is not reflected in the virtual market, which requires analysis in different categories. It should be mentioned here that enterprises, which are market entities, have undergone a kind of evolution (Figure 1).

*Energies* **2022**, *15*, x FOR PEER REVIEW 2 of 22

**Figure 1.** Evolution of market entities. Own-study based on [2] (pp. 21–22), [3,4], [5] (p. 33), [6,7]. **Figure 1.** Evolution of market entities. Own-study based on [2] (pp. 21–22), [3,4], [5] (p. 33), [6,7].

This evolution included successive stages, namely [2] (p. 21): This evolution included successive stages, namely [2] (p. 21):


least slowing down of transfers) of information and knowledge flow [6,7]. We agree with D.J. Hołodnik and K. Perechuda that networking breaks the traditional dual arrangements, such as "profit-loss, effective-ineffective, good-evil" [2] (p. 22). Of course, this does not mean that there are none, but the boundaries between them are slowly blurring, which is influenced by asymmetry and the excess of information (particularly tacit information) between corporations and skilfully dosed "information junk" for the masses (customers and consumers). The confusion and disorientation of an average customer/consumer in the context of the excess of information, in the absence of appropriate tools to process them, forces them to search for specialists who are able to offer We agree with D.J. Hołodnik and K. Perechuda that networking breaks the traditional dual arrangements, such as "profit-loss, effective-ineffective, good-evil" [2] (p. 22). Of course, this does not mean that there are none, but the boundaries between them are slowly blurring, which is influenced by asymmetry and the excess of information (particularly tacit information) between corporations and skilfully dosed "information junk" for the masses (customers and consumers). The confusion and disorientation of an average customer/consumer in the context of the excess of information, in the absence of appropriate tools to process them, forces them to search for specialists who are able to offer them adequate digital services, both information and process-service ones, being part of the characteristics of 2.0 enterprises [8].

them adequate digital services, both information and process-service ones, being part of the characteristics of 2.0 enterprises [8]. In contrast to 2.0 enterprises, the next generation is de-networked, de-empowered enterprises, based on the humanistic paradigm, which is the context of the density of information, knowledge, technological, capital and other densities of contemporary corporations find their place to operate where further network cooperation is not possible—corporations reach the so-called Break-Event-Point of their networking process. However, the management of 3.0 enterprises requires totally new managerial skills, as well as the adoption of a new development perspective [9] (p. 172), for which, however, the majority of modern businesses are not ready yet. Consequently, we can still observe the creation and development of strategies characteristic of 2.0 enterprises [10].

The functioning conditions of 2.0 enterprises focus attention on strategic risk, competition strategy, as well as social sensitivity [11]. The analysis of the critical factors of their development is, in this aspect, an element of the strategy the formulation and adoption of which depends on the position of the enterprise in the network, its network capacity (value creation in the form of a network advantage, relational abilities, as the network participant which favours capturing values [12].

As D.A. Johnston, M. Wade and R. McClean [13] rightly observe, adaptation to the conditions of e-business functioning may be expressed in tangible financial results. However, profits gained may be different depending on the region, type of industry/business sector, as well as the company size class. It can therefore be assumed a priori that there are specific differences in digitalisation:


As P. Depaoli, S. Za and E. Scornavacca [14] indicate, "digital business solutions are commonly adopted with the goal of improving firms' performance". It results from the fact that investment in technologies is one of the key decisions in the context of enterprise strategic operations, transforming its organisational activities, as well as deeply penetrating its broadly understood competitiveness and development. It was also a premise to undertake theoretical research based on the literature of the subject.

#### **2. The Concept of Research and Its Methodology**

The aim of this paper is to review the existing research on the drivers and business models, as well as prospective directions of changes in the energy sector enterprises in the context of their digitalisation. Due to the revealed cognitive gaps, the authors of the paper formulated the following research questions:


We selected keywords from three broad areas, such as drivers, business models and prospective directions of changes in the enterprises of the energy sector to identify research papers that would help to answer the above-mentioned research questions. We searched using the title, keywords and abstract fields according to the searching database. We used Scopus, Google Scholar, Science Direct and Emerald databases as these provided papers that were the most relevant to our research. The time range of the cited scientific sources included, in the vast majority, publications from the last 5 years concerning enterprises in the energy sector.

The methodology of the theoretical research undertaken, in terms of the literature review, was selected in such a way as to enable an objective and structured approach to the study of factors, or more precisely, determinants (drivers), business models and prospective directions of changes in the field of the studied phenomena related to the digitalisation of energy sector enterprises. Therefore, the following research methods were adopted, broken down into research stages:


Taking into account the type of research, it should be emphasized:


As a result, the adopted research methodology allowed for the total compilation of 93 bibliographic items, while the research limitations apply to the adopted four databases indicated above. However, it should be emphasised that the selection of these databases was dictated by the citation rate of their publications and recognition in the scientific community.

The purpose of adopting the research methodology defined in this way was to systematize and also to synthesize the current state of knowledge in this field, and after proper research, it should allow [confirmation or contradiction] and/or extension of the answers to the research questions posed. Thus, the authors' contribution, resulting from the conducted literature research, concerns the systematization and synthetic presentation of the current state of knowledge in the field of the digitalisation of energy sector enterprises, taking into account the detailed issues of factors, including determinants (drivers), business models and prospective directions of changes.

The added value of this paper is the attempt to investigate possibly the most comprehensive appraisal, based on domestic and foreign literature, in the area of social sciences at the scope of relationships between the drivers of digitalisation, new innovative business models of the energy sector enterprises emerging as a result of adapting to changing operating conditions, made possible to introduce digital technologies and the prospective directions of their changes in light of contemporary challenges and dilemmas in their operations.

#### **3. The Drivers of the Digitalisation in the Energy Sector Enterprises**

Beginning in the 1960s, common regulations concerning the activities of energy sector enterprises operating in the EU market began to be introduced. They were included in the "founding treaties": the establishment of EURATOM and the European Coal and Steel Community, followed by the establishment of the International Energy Agency in 1974. In turn, the Single European Act (1986) contributed to research on the European energy sector. The first directives on the liberalization of the energy market were adopted in 1996 (First Energy Package). The Energy Charter Treaty (1998) established a framework for energy investment, trade in energy commodities and energy distribution. The second energy package was adopted in 2003 and the third in 2009, which was related to the further

liberalization of the electricity and gas markets, and in 2009 the fourth energy package was introduced, including the European Union Agency for the Cooperation of Energy Regulators (ACER). The fifth energy package (Delivering the European Green Deal) was published in 2021.

These requirements have become a challenge not only for energy sector companies because the implementation of the Common Energy Market has a direct impact on all areas of life. As A. Czech notes, the process of consolidating the European Member States' markets into one competitive energy market is complex and lengthy. It requires a number of strategic political decisions, including energy liberalization and the adoption of sectoral legislation [15]. The activity of enterprises in the energy sector is determined by the heterogeneous energy infrastructure among EU Member States, as well as the frequent differences in legal environments and institutional and administrative systems.

Some EU countries are characterized by modern and developed electricity and gas infrastructure, while others are clearly lagging behind in this respect. Europe's lack of energy self-sufficiency and import dependence is becoming more and more apparent. According to The Dependency Rate on Energy, the import of energy consumed by EU member states has increased from 47% in 2019 to 53% in 2022 [16]. Moreover, there are large variations between Member States, with energy dependency rate ranges from over 95% for Malta, Luxembourg, and Cyprus to below 15% for Estonia and Denmark [16]. This proves the existing heterogeneity and limited intra-market possibilities of meeting the demand for energy.

As noted by A. Carfora, R.V. Pansini and G. Scandurra, the current direction of EU activities in the field of energy is to reduce energy dependence, which is to be achieved by improving the efficiency and maximum use of domestic energy sources as well as the promotion of renewable energy sources (RES) [16]. Actions taken should be implemented in accordance with the disclosed need to ensure more stable and safe supplies, including increasing the use of renewable energy sources, while reducing greenhouse gas emissions. For these reasons, efforts have recently been focused on ensuring energy security taking into account the principles of sustainable development, thus becoming one of the priorities in defining the new energy policy framework in the EU [17] (pp. 155–163), [18] (pp. 127–136). The process of changes related to the consolidation, research, liberalization and common energy policy in the energy market of the European Union countries is presented in Figure 2. *Energies* **2022**, *15*, x FOR PEER REVIEW 6 of 22

**Figure 2.** Changes related to the research, liberalization, consolidation and common energy policy in the energy market of European Union countries. Own-study based on [15] (pp. 251–262), [17] (pp. 155–163), [18] (pp. 127–136). **Figure 2.** Changes related to the research, liberalization, consolidation and common energy policy in the energy market of European Union countries. Own-study based on [15] (pp. 251–262), [17] (pp. 155–163), [18] (pp. 127–136).

Thus, the activity of energy sector enterprises is very strongly related to socioeconomic conditions. There is an increasing demand for electricity against the backdrop of a decline in non-renewable energy resources. Therefore, this macroeconomic factor should be considered significant and determining the development of enterprises in this sector.

sources in the operation of the energy sector—in the context of global socio-economic conditions—seems to be correct [19]. Lutz, Fischer, Newig and Lang list 19 driving factors of renewable energy implementation processes [20]. They have all been grouped into the

As shown in the above studies, formal networks emerge as a strong driving factor in the studied regions. Regions where formal networks are important see the positive impact of renewable energy use on the regional economy [20]. Therefore, it can be assumed that networking related to the digitalisation processes of energy sector enterprises is an important factor in their development in the context of plans for the use of renewable energy sources. This corresponds to the idea of sustainable development in relation to natural

Due to the complexity of the research topic undertaken, it is worth paying attention to the closely related terms "condition" and "factor". The ability to distinguish between static and dynamic variables in the context of the development of enterprises affects the correct determination of the components of this development and the ability to properly

"Conditions" are a feature or set of features on which the existence of something depends, and a situation prevailing in some area (location, circumstances) in which something happens [21]. In turn, the term "factor" should be understood as the cause of a specific phenomenon, and something that has an impact on the observed phenomenon, actions taken or the existing situation [22]. The word "factor" comes from the concept of

prioritize them in the context of setting development goals.

following four clusters: 1. Planning and process; 2. Exchange and participation; 3. Actors and networks; 4. Economic circumstances.

capital.

Thus, the activity of energy sector enterprises is very strongly related to socioeconomic conditions. There is an increasing demand for electricity against the backdrop of a decline in non-renewable energy resources. Therefore, this macroeconomic factor should be considered significant and determining the development of enterprises in this sector.

The adopted direction of development, related to the use of renewable energy sources in the operation of the energy sector—in the context of global socio-economic conditions—seems to be correct [19]. Lutz, Fischer, Newig and Lang list 19 driving factors of renewable energy implementation processes [20]. They have all been grouped into the following four clusters:


As shown in the above studies, formal networks emerge as a strong driving factor in the studied regions. Regions where formal networks are important see the positive impact of renewable energy use on the regional economy [20]. Therefore, it can be assumed that networking related to the digitalisation processes of energy sector enterprises is an important factor in their development in the context of plans for the use of renewable energy sources. This corresponds to the idea of sustainable development in relation to natural capital.

Due to the complexity of the research topic undertaken, it is worth paying attention to the closely related terms "condition" and "factor". The ability to distinguish between static and dynamic variables in the context of the development of enterprises affects the correct determination of the components of this development and the ability to properly prioritize them in the context of setting development goals.

"Conditions" are a feature or set of features on which the existence of something depends, and a situation prevailing in some area (location, circumstances) in which something happens [21]. In turn, the term "factor" should be understood as the cause of a specific phenomenon, and something that has an impact on the observed phenomenon, actions taken or the existing situation [22]. The word "factor" comes from the concept of action (lat. *facere* = to do) and is one of its elements—the cause [23]. An activity is a form of the active human relation to reality or a process aimed at achieving a goal, the structure of which is shaped depending on the existing conditions [24]. As A. Kaleta notes, the two concepts analyzed above are closely related, particularly regarding the development of an enterprise, the relations between the "condition" and the "factor" are indicated [25]. The literature on the subject emphasizes that "development conditions" have a less direct impact than "development factors", as they are of a potential nature. They may be present in a given place, time and area (legal, economic, social, etc.), they may be favourable and unfavourable, or they may be completely absent. Thus, the existence of appropriate conditions creates the possibility of emergence and development [26]. The disclosure of the development condition contributes to the implementation of certain activities. In turn, the factor triggers some process of action, which is then confronted with the identified conditions. Thus, the factor is an active attitude to reality. The condition, on the other hand, may be of a natural, technical or social nature [23]. Hence, it is assumed that the conditions of development are generally static and are recognized as resources. On the other hand, the factors are dynamic and are treated as streams [27]. In terms of economic capital, the term "conditions", understood as a set of more or less favourable circumstances (conditions) enabling or limiting development, seems to be a more appropriate term than a "condition".

According to S. Chom ˛atowski, conditions (for economic capital these are economic conditions) and factors, including drivers of enterprise development can be understood as what is related to the development of the enterprise and influences this development directly or indirectly. In other words, the conditions for the development of an enterprise are permanent or shorter-lasting possibilities, existing in a given place and time. They include all the opportunities and threats that may favor or limit the development of the enterprise [27]. The conditions of development have a real and material dimension. Their discovery and use contribute to the development of the enterprise, thus becoming its drivers [28].

Returning to today's challenges related to the implementation of sustainable development, three basic conditions useful in the analysis and evaluation of the development of energy sector entities should be indicated. They are part of the "3 Ds of Energy" based on digitalisation, decentralization and decarbonization. These are conditions and factors related to economic, social and natural capital.

Experts from the International Energy Agency [29] (pp. 18–20) indicate that the global demand for electricity will increase by about 2–3% annually [30]. Throughout 2021, global energy demand and emissions increased by 5% compared to 2020, almost reaching pre-COVID-19 levels (~33 Gt energy-related CO<sup>2</sup> equivalent) [30]. As energy consumption increases, so too will greenhouse gas emissions. This could lead to an increase in global temperature of 4–6 ◦C by 2030 [31]. According to industry analyses, even if all countries with net-zero commitments deliver on their aspirations, global warming is projected to reach 1.7 ◦C by 2100 [30]. Such a forecast determines further planning of activities in this regard.

The mentioned conditions and factors are closely related. The activity of enterprises in the energy sector is conditioned by the existence of demand for energy by social capital. This is also an important determinant of how much, how and when energy will be used. The process of obtaining electricity and heat, in turn, involves the use of renewable and non-renewable resources that belong to natural capital. Finally, in order for energy to be delivered, economic capital is needed for energy production and transmission infrastructure.

The development of technology determines the technical capabilities of energy production and transmission infrastructure, as well as the degree of its modernity, reliability and efficiency. It is on the basis of the above-mentioned characteristics that the assessment of this infrastructure is made. The digitalisation process of energy sector enterprises has a direct and indirect impact on these characteristics.

The researchers claim that the energy industry can increase energy efficiency through various changes leading to the integration of individual energy companies and information communication technologies (ICT) supporting processes related to energy generation, transmission and delivery [32]. Therefore, digitalisation plays a key role in the implementation of the new model of the 3 Ds of Energy development. Effective digitalisation will allow the development of more and more advanced technologies enabling the diversified production of energy from decentralized renewable sources. This will make possible the switch from non-renewable sources in the form of coal or crude oil to renewable sources.

Park, Cho & Heo distinguish three main entities which will have a strong influence on digitalisation: (1) energy policymakers, (2) researchers and (3) market participants [33]. Based on empirical research by You & Yi, it has been shown that "self-transcendence knowledge" is very important and critical in the knowledge management and promoting energy companies to react quickly to changes and challenges related to digitalisation [32]. The policy of transforming the energy sector adopted in the EU assumes transferring energy production to renewable sources, reducing the emission intensity of the economy and increasing energy efficiency. The implementation of these plans is to ensure energy and environmental security, enable constant civil development, and the creation of low-emission, highly effective energy based on renewable energy sources. In this respect, digitalisation is treated as a tool for the implementation and development of a new energy policy. As A. Kucharska notes, "digitalisation is an essential element of energy infrastructure management, smart grids and smart meters. It allows better equality of energy distribution, which is a factor in combating energy poverty" [34]. One should be aware that the digitalisation of the energy sector is not a "cure for all diseases" in this sector and is associated with specific threats, which will be presented below. However, digitalisation should be seen as an effective tool for more effective management of the production, distribution and storage of electricity and heat. Digitalisation is a significant factor in the development of enterprises in the energy sector, which is also part of the concept of sustainable socio-economic development.

The above-mentioned drivers in the digitalisation of energy sector enterprises are specific to this industry and fundamental to its development. They are related to the process of implementing the new energy policy based on the assumptions of the 3 Ds of Energy. There are also other drivers in the digitalisation of the activities of energy sector enterprises, which are also noticeable in other industries and result from the so-called digital transformation of business [35].

Both specific and general factors driving the development of the digital economy (digitalisation of activities) of energy sector enterprises as well as enterprises from other industries are presented in Table 1 ([29] (pp. 21–23), [35–37]).

**Table 1.** Specific and general factors in the digitalisation of energy enterprises.


Source: own-study based on [29] (pp. 21–23), [35–37].

The challenges and opportunities related to the above-mentioned factors, which refer straight to the energy sector include [38–40]:


real energy networks to develop new investments and reducing the cost of designing new installations;


Digitalisation in supply chain management, also in the energy sector, allows for the implementation of the most important efficiency criteria in the form of costs, time and service quality. It contributes to maximizing the effectiveness of customer service by delivering the right product (service) as quickly as possible and reducing the cost of service while ensuring high availability and reliability of the service (product) [41]. *Energies* **2022**, *15*, x FOR PEER REVIEW 10 of 22

#### **4. Business Models in Energy Sector Enterprises in the Context of Digitalisation**

V. Akberdina and A. Osmonova note that the digital transformation in energy sector enterprises has been taking place in the world for many years [42,43]. As noted by J. Swiatowiec-Szczepa´nska and B. St˛epie´n, this sector pioneered the implementation of ´ information and communication technologies and over the past few years it has undergone enormous changes as a result of changes related to the fourth industrial revolution [44]. Digitalisation in the energy sector is linked to the creation and use of computerized information and the processing of the huge amount of data that is generated along the energy supply chain. New digital technologies offer great opportunities to improve the efficiency of managing an advanced energy system: from infrastructure design, operation and maintenance, through energy production and transmission, to its consumption. Due to digitalisation in the energy sector, the necessary infrastructure and interfaces are created that enable the efficient functioning of operators and the intelligent and effective implementation of the processes they support because it enables cheaper, faster and better monitoring using "smarter" networks [45]. P.F. Borowski points out that, as research shows, digitalisation enables energy companies to reduce operating costs and improve efficiency, and additionally helps to extend the life of power plants by about 30% [46]. **4. Business Models in Energy Sector Enterprises in the Context of Digitalisation**  V. Akberdina and A. Osmonova note that the digital transformation in energy sector enterprises has been taking place in the world for many years [42,43]. As noted by J. Światowiec-Szczepańska and B. Stępień, this sector pioneered the implementation of information and communication technologies and over the past few years it has undergone enormous changes as a result of changes related to the fourth industrial revolution [44]. Digitalisation in the energy sector is linked to the creation and use of computerized information and the processing of the huge amount of data that is generated along the energy supply chain. New digital technologies offer great opportunities to improve the efficiency of managing an advanced energy system: from infrastructure design, operation and maintenance, through energy production and transmission, to its consumption. Due to digitalisation in the energy sector, the necessary infrastructure and interfaces are created that enable the efficient functioning of operators and the intelligent and effective implementation of the processes they support because it enables cheaper, faster and better monitoring using "smarter" networks [45]. P.F. Borowski points out that, as research shows, digitalisation enables energy companies to reduce operating costs and improve efficiency, and additionally helps to extend the life of power plants by about 30% [46].

The most important global digital trends affecting the energy sector and selected concepts of innovative business models corresponding to them, are presented in Figure 3. The most important global digital trends affecting the energy sector and selected concepts of innovative business models corresponding to them, are presented in Figure 3.

Global digital trends makes energy sector companies look for effective methods of competing, shaping and implementing new strategies and business models, using them to an increasing extent in various types of innovation and cooperation networks. As

mented in specific countries or a group of countries in a specific region [66]. Currently, the key area in the field of strategic management is to ensure that businesses have the capacity to implement quick and abrupt changes that affect the enterprise itself, including its strategy, structures, resources and processes, and its environment. For this reason, the ability to act quickly and flexibly in a turbulent, rapidly changing environment becomes very important. Research conducted by A. Kaleta, J. Radomska and L. Sołoducho-Pelc

Global digital trends makes energy sector companies look for effective methods of competing, shaping and implementing new strategies and business models, using them to an increasing extent in various types of innovation and cooperation networks. As rightly noted by J. Brzóska and M. Krannich, for energy companies, a significant factor influencing their strategic reorientation is also the energy and climate policies implemented in specific countries or a group of countries in a specific region [66]. Currently, the key area in the field of strategic management is to ensure that businesses have the capacity to implement quick and abrupt changes that affect the enterprise itself, including its strategy, structures, resources and processes, and its environment. For this reason, the ability to act quickly and flexibly in a turbulent, rapidly changing environment becomes very important. Research conducted by A. Kaleta, J. Radomska and L. Sołoducho-Pelc shows that in large enterprises, such as energy sector companies, innovation is related to strategic management, and this relationship is related to a strategic response to challenges [67]. According to A. Ates, P. Garengo, P. Cocca, and U. Bititci, large enterprises more often than SMEs design and implement formalized processes to manage operational and managerial activities. In the case of SMEs, the systems and processes related to decision-making and the management of the entire business are less structured [68].

The development of industrial strategies is often associated with the occurrence of conflict situations. Some of them, concerning opposing objectives such as, for example, increased production versus environmental impact, or increasing automation versus reducing employment, are very difficult to resolve. Improving productivity is important for wage growth, and at the same time, it is very often associated with the implementation of advanced technologies that reduce employee involvement. Developing countries often accept environmental issues because they want to counteract labour-saving technology [69].

Tensions related to the necessity to meet the needs of customers, the activities of competition and the level of costs result in the emergence of new methods, tools and theories in managing the development of the energy sector. At the same time, energy sector companies design and reformulate their business models, which is difficult in an unstable and complex environment [70]. S. Küfeo˘glu, G. Liu, K. Anaya and M.G. Pollitt note that the newly adopted business models in the energy industry vary in terms of four business model dimensions [45]:


B. Fattouh, R. Poudineh and R. West note that their effective functioning requires flexible and systematically adapted development strategies [71].

A key driving factor in energy sector digitalisation is the increasingly distributed character of the electricity generation system, which opens up new business opportunities. The decentralization of power generation is part of the transformation based on the so-called 5 Ds (decentralization, digitalisation, decarbonization, democratization, and decreasing consumption), which are the driving forces of the current energy transition [72]. As noted by A. Hirsch, Y. Parag and J. Guerrero, the increasing use of renewable energy sources is playing a huge role in accelerating this trend [73]. S. Küfeo˘glu, G. Liu, K. Anaya and M.G. Pollitt add that the development of renewable energy means that today energy is generated within the electricity distribution grid that includes many more nodes than ever before. The problem, however, is the intermittency of this production and the poor match of its supply to demand [45].

According to A. Idries, J. Krogstie and J. Rajasekharan, the decentralized energy means energy that is generated closer to the place where it is used. It is therefore not drawn from a large centralised power station that is part of the regional or national grid [74].

As rightly noted by J. Silvente, G.M. Kopanos, E.N. Pistikopoulos and A. Espuña, the local generation of energy reduces transmission losses that result from the long distance between production and consumption sites and decreases environmental degeneration as it lowers carbon emissions [75]. It also increases the security of supply for all consumers, as they do not use a single power source or depend on relatively few large and remotely located power plants [76]. A.M. Adil and Y. Ko point out, that energy decentralization has three configurations: distributed generation, micro-grids, and smart micro-grids [77]. The backbone of energy decentralization is distributed generation, which can be defined as "*generating plant connected directly to the grid at distribution level voltage or on the customer side of the meter*" [78,79].

As indicated by H. Khajeh, H. Laaksonen, A.S. Gazafroudi and M. Shafie-khah, decentralized energy is accompanied by the transformation towards a platform-based model, which is a good starting point for service innovation in the energy sector. As a result of this transition, various services will be provided by these platforms, thereby giving many stakeholders and actors the opportunity to benefit from them [80]. Digitalisation helps to better manage the network and congestion, helping to solve the problems of discontinuities related to the production of energy from renewable sources. It makes it possible to use digital platforms to respond to demand, Peer-to-Peer (P2P) energy and carbon credit trading [45]. K. Osmundsen notes that a crucial trend for the digital development of the energy sector is also transforming grid networks into smart grids, which are electricity networks able to process, control, and manage huge data flows [43,81].

As shown in Figure 3, changes important to the existing energy system include business model innovations based among others on the following concepts:


Irregular small-scale energy production and energy exchange between autonomous micro-grid prosumers [48] can be integrated into the energy system with low indirect costs thanks to P2P energy trading. Peer-to-peer (P2P) electricity trading is a new datadriven business model that is currently being tested in the energy sector [49]. Peer-to-peer (P2P) energy trading is essentially based on direct transactions between prosumers and consumers who trade energy with each other without the electricity supplier or retailer as an intermediary. The place of the intermediary is taken by a third-party digital platform, e.g., blockchain-based, which allows prosumers and consumers to interact directly and negotiate better prices for electricity, relative to the offer of a licensed supplier [50,51]. This type of energy trading can be facilitated by the digitalisation of the economy and by the use of advanced digital technologies such as Artificial Intelligence (AI) or Blockchain [45].

Artificial neural networks allow the analysis of information in an innovative manner. These are mathematical and computer models that imitate the work of neurons in the human brain, so they can analyse information and learn to generate proposed solutions. J. Swiatowiec-Szczepa ´nska and B. St˛epie ´n point out that their application is possible in many ´ areas, and they are used at each stage of the energy value chain [44]:


AI can be used to optimize energy storage and supply management, it has the potential of a predictive system and can be the basis for creating systems that respond to

energy changes [52]. The consequence of AI managing energy transmission is Demand Response Models (intelligent networks supported by models of responding to changes in demand) enabling real-time monitoring of changes in energy demand and proper energy distribution [52].

Other digital technologies, such as blockchain, can improve the information flow process, document flow, property management, logistics and security of supply. Blockchain enables, through the controlled use of data, the management of the increasing complexity of the structure and networks of the energy sector and direct interaction between entities through comprehensive monitoring of energy flows at low costs [44]. The greatest benefits associated with the introduction of blockchain technology include the possibility of concluding direct transactions between energy suppliers and their customers. Blockchain technology solutions can also enable the sale of micro volumes of energy between prosumers [52]. Traditional energy trading faces problems such as single points of failure and privacy leakage. Therefore, efficient and transparent models are needed in this area. The advantages of blockchain technology include trust, security and decentralization. This technology maintains a distributed database, eliminating the need to involve third parties, improves economic efficiency, and reduces management costs [53] and ensures supply security by providing information on the origin of energy [44]. In addition, E. Mengelkamp et al. add that blockchain technology provides user-friendly applications that enable them to participate in making decisions about which technology they use to generate energy and who produces it [54].

One of the most reliable and efficient ways to ensure the efficiency, stability and competitiveness of distributed generation is the use of a business model based on the concept of a Virtual Power Plant (VPP). A VPP is a system integrating a few types of energy sources, which provide a reliable overall power supply. Virtual power plants can be described as decentralized energy management systems, which pool the capacity of different units: renewable and non-renewable ones, different storage devices and distributable loads, participating in the energy market and trading energy (and services) with the upstream network [55]. P. Asmus defines VPPs as "software systems, which remotely and automatically dispatch and optimize generation or demand side or storage resources in a single, secure Web-connected system" [56]. Through software innovation, VPPs use existing energy networks to tailor electricity supply and demand services to meet customer needs. This maximises value, both for the end customer and for the distribution company [56]. The VPP buys spare capacity from energy-producing entities and manages the power plant production according to the real-time sales demand. The produced energy is sold simultaneously on multiple markets—on the Power Exchange (PEx), on the balancing energy market and directly to end users. This business model allows small power plants to earn higher margins without incurring additional energy sales and balancing costs [57]. As noted by S. Baidya, V. Potdar, P.P. Ray and Ch. Nandi, the functional capabilities of virtual power plants are, therefore: (1) optimisation of network modes, (2) management of electricity consumption and (3) market management of the reserve capacity [58,59].

The dynamic development of renewable generation and distributed energy resources presents distribution systems with new operational challenges related to intermittency and uncertainty of supply. Addressing these challenges requires distribution system operators to introduce business models that enable more active and flexible management and system control. This results in a shift to a local approach in energy system management. P. Olivella-Rosell et al. point out that local Energy Markets (LEM), also known as micro-markets or local electricity markets [60], use online platforms and smart grid technologies as a means to integrate and coordinate Distributed Energy Resources (DER) at the distribution grid level [61]. According to S. Beattie, W.K. Chan, A. Wei and Z. Zhu, this is important, especially in relation to the local supply of spatially distributed renewable resources that cannot be directly controlled [62]. As noted by P. Olivella-Rosell, G. Viñals-Canal et al. studies show that with too small distribution networks, LEM can effectively improve social welfare compared to other, simpler solutions [60].

The possibility of implementing new business models is also seen in the development of V2G technology. Electric vehicles (EVs) can be classified in terms of their charging methods, energy storage mechanisms or energy sources. The increase in sales of electric vehicles is related to the need to increase the amount of available energy. In addition, the needs of charging electric vehicles result from the profiles of drivers and their mobility. All this affects the allocation of electric vehicle charging demand [63]. V2G technology creates added value for EV users and Distribution System Operators, as some specific services can be provided to the power grid. By connecting am EV to the grid, the bi-directional energy exchange is possible, from the power grid to the electric vehicle and the other way round, from the electric vehicle to the grid [64]. With the growing number of electric vehicles, these services can support the energy industry to better manage energy generation, its management and, ultimately, its consumption. Electricity management by consumers in the technology of two-way energy exchange is a new approach in terms of building a new power grid and the value of available power consumption [65]. As rightly noted by B. Mroczek and A. Kołody ´nska, nowadays more and more attention is paid to the problems of network management. The popularity and effectiveness of V2G services among EV users will depend on the correct definition of electricity demand [65].

In conclusion, it should be stated that the business models introduced by energy sector enterprises are associated with the pressure resulting from the transformation towards the digital economy. In this context, the strategies that relate to gaining a privileged position in the industry are also changing. Changes in the energy sector are influenced by the acquisition of entities from the periphery of the energy sector or those that are still reserved for other products. This mainly refers to big data entities and enterprises using energybased products [70]. R. Trzaska et al. emphasise the need to implement digital technologies in the energy sector due to their potential to increase efficiency, productivity, speed and security, as well as reduce costs. This becomes possible mainly thanks to the use of artificial intelligence, the development of advanced sensors for process monitoring and remote access in various organisational areas [70]. It is important to develop business models that are based on the real added value that distributed energy can bring to the entire energy system. This added value may be a solution that allow for real local real-time power balancing, which requires an advanced data ecosystem and algorithms for managing a distributed system. However, such business models require a number of innovations in many areas.

These factors emphasise the need to introduce digital solutions in the energy sector due to their potential to increase efficiency, productivity, speed, safety and cost reduction.

#### **5. Challenges and Dilemmas of Energy Sector Enterprises Regarding Industry 4.0 and 5.0**

The digital transformation of the energy sector is part of the more general current phenomena of Industry 4.0 and Industry 5.0. Moreover, this process is a component of shaping the information society and the implementation of the global plan of sustainable development [82] (pp. 283–294). It is important to distinguish between two ways of understanding the essence of the digitalisation of the energy industry. Firstly, it is the automation of existing engineering and business practices using digital technologies. Secondly, radical technological and civilizational change is associated with the transformation of the very structure of the industry and the emergence of new practices (*Digital transformation*). We will continue to talk about digitalisation in the latter sense [83]. The effects of digital transformation are shown in three projections: in the area of technology, business and culture [84]. Moreover, digitalisation affects the nature of the relationship between the above-mentioned areas (human–technology, human–business, and business–technology). The following opportunities offered by digital transformation can be listed:


expertise); creating new forms of staff training (gamification, virtual and augmented reality technologies) [85,86];


At the same time, the digitalisation of the energy sector is associated with new challenges and dilemmas concerning, inter alia, the risk of destabilisation of the energy market resulting from its decentralisation, new requirements for the qualifications of people employed in the energy sector, a new culture of the cooperation between energy suppliers and recipients and ensuring the digital security of data used in shaping a common energy system [82] (pp. 283–294). To a large extent, the above-mentioned dangers related to the digitalisation of the energy sphere are also specific to other areas of life where digital transformation takes place. Generally speaking, meeting new challenges and dilemmas is associated with finding a balanced dynamic relationship between the area of technology, business and culture, ensuring the free flow of information between these areas. Below, we will describe some of the difficulties accompanying the digital transformation of the energy sphere:


its user. This applies, inter alia, to ways of shaping flexible tariffs, and the need for clients to know digital technologies. Increasingly, the user is served without an assistant, which is replaced by chatbots. The latter solution, so far, works only in the performance of standard, non-complex tasks. In many cases, solving a problem or doubt requires talking to a person [90] (pp. 189–194). In addition, very often the client's need to get direct feedback from the assistant is still indispensable.

Proposed answers to the above dilemmas and challenges:


of the AI controlling all processes in the plant. This raises more general questions about the place of humans in the new reality shaped by the latest digital technologies;

7. When encountering new digital technologies in the area of customer service, the end user must be convinced that the changes made have contributed to the increase in the convenience and security of contact with the corporation. The user interface has to become more and more intuitive. When proposing new solutions, the corporation should maintain a proper caution, offering clients only proven solutions [91].

The reigning or consciously created social consciousness largely determines our future. We choose how we want to live, and what values and goals are important to us [92]. There may not be only one right strategy for the future. Therefore, when implementing new digital solutions, one should not be radical. Different approaches in the field of energy, more traditional and modern ones, can coexist [91]. One should choose solutions appropriate for a given place and time, solutions effective in solving specific tasks. It is worth following the path of evolution rather than revolution [84]. A rapid transition from one system to another is associated with instability and an increase in risk.

Digital transformation should solve specific problems and bring about concrete improvements in terms of stability, saving, comfort and security. People and their welfare must remain the overriding concern underlying all these changes [91]. It should also be remembered that digital transformation consists not only in the emergence of new business and engineering practices but above all in the transformation of the very culture of production, distribution and use of energy. Digitalisation changes the very paradigm of thinking about energy [93]. In the course of digital transformation, human resources are shifted from existing processes and practices to others, mainly related to the maintenance of information systems. People still remain the main point of digital transformation.

#### **6. Conclusions**

The energy sector is an important sector of the economy, determining the socioeconomic development of each country and, globally, having a strong impact on the environment and climate change. This makes energy an area of the global economy where major technological and organisational transformations are taking place, based on a wide range of digital innovations being implemented. In this paper, we have reviewed existing research on drivers, business models and prospective directions of change of energy sector companies in the context of the digitalisation processes taking place in them.

With reference to the first research question, the conducted literature review allowed us to identify drivers influencing the digitalisation of energy industry enterprises, among which we distinguished those characteristics for this sector and those relating to all business entities, being a part of the more general phenomena of Industry 4.0 and Industry 5.0. We found that a number of sector-specific drivers influence the pace, course and effects of these processes. They include different socio-economic conditions, related on the one hand to the increasing consumption of electricity by economies and societies, and—on the other hand—to shrinking conventional energy resources. They also result from the current energy policy in the European Union, which aims at decarbonising the economy, rising energy efficiency and increasing the share of production from renewable energy sources. The digitalisation of energy companies can greatly facilitate adaptation to these requirements, representing a key factor for their rapid and efficient development, fostering the simultaneous achievement of the three priorities: the sustainability, energy security and competitiveness of the energy sector. As another sector-specific driver, we have identified the increasingly distributed nature of electricity production in the energy sector, in which the voice of prosumers, who want to gain greater control over energy management, use and production, is also increasingly heard. This is shifting more and more attention towards small, distributed, local and dynamic activities and businesses that are changing the fundamentals of the energy market. In parallel, energy sector companies are being impacted by other digitalisation drivers, which are also noticeable in other industries, and are therefore of a universal nature. They stem from the ongoing digital transformation of

the business sector, which is a part of the more general phenomena of Industry 4.0 and Industry 5.0. Our review shows that digitalisation makes it possible to respond to all these challenges in both the supply and demand sides of the market. For energy suppliers, it offers higher system efficiency, the opportunity to make full use of data and implement new strategies and business models, while for consumers and prosumers, it offers comfort and a reduction in energy consumption costs.

The research review showed that a method to improve the efficiency of energy companies is to implement technological innovations, including those enabling their digital transformation, which changes the essence of their operations and affects all aspects of their activities. At the same time, this means introducing new strategies and business models. With reference to the second research question, we have identified several innovative business models in the energy industry, which are or will be actively implemented in the near future thanks to digitalisation as a response to current trends in the energy market. They include new concepts of doing business in the energy sector based on the idea of peer-to-peer (P2P) electricity trading, virtual power plants (VPPs), flexibility management, local energy markets and the vehicle-to-grid concept. Analysing the described emerging innovations in business models in the energy sector as a result of digitalisation, it can be concluded that the boundary between supply and demand may be blurred in a digitised future. The deployment of smart grids, interconnected and interoperable energy trading and management systems, as well as the exploitation of the potential of artificial intelligence, blockchain and other digital technologies, can completely change the position and roles played by suppliers and consumers. New roles and new actors, such as prosumers or aggregators, are emerging in the transforming energy market, and the range of possibilities for their creation seems to be essentially limitless. As the review of research showed, the new business models being implemented in the energy sector encourage, on the one hand, large generators to transform their operations, including through the possibility to use RES, thus enabling them to adapt to the trends observed in the market, while, on the other hand, they support small prosumers in being active in the energy market, which encourages the active building of a demand-supply balance.

Answering the third research question, what prospective directions of changes will determine the paths of development for the energy sector enterprises in the near future, based on a literature review we identified the main challenges and dilemmas in the energy sector in connection with its digitalisation. We have included here: the risk of destabilisation of the energy market resulting from its decentralisation, the new requirements placed on the competences of those working in the energy sector, the new culture of interaction between energy suppliers and consumers and the digital security of data used in the energy system. Digital transformation should correspond closely to these concerns and offer concrete improvements in terms of stability, savings, convenience and security in the energy sector. In this context, this process is a crucial component of shaping the information society and the implementation of the global plan of sustainable development.

Our research contributes to revealing the potential opportunities, threats and challenges of the digital transformation of the energy sector, and all the results obtained and the conclusions expressed should be treated as a stimulus for further in-depth research, especially on the issue of innovation in the field of new business models resulting from digitalisation, which will provide greater opportunities for implementing sustainable energy transformation and on solutions that will be a response to the above-mentioned risks and dilemmas related to the digitalisation processes in this sector. Regarding the directions for future research, it should be pointed out that there is a constant need to monitor the factors forcing the necessary changes or fostering the transformation of the energy sector, especially in the context of the possibility of using digital technologies to improve existing or implement new, innovative business models that meet the complex needs of various stakeholders of the energy system.

**Author Contributions:** Conceptualization, B.S-T., S.K., P.K., A.T. and K.Z.; Formal analysis, B.S.-T., ˙ S.K., P.K., A.T. and K.Z.; Investigation, B.S.-T., S.K., P.K., A.T. and K. ˙ Z.; Methodology, B.S.-T., S.K., ˙ P.K., A.T. and K.Z.; Resources, B.S.-T., S.K., P.K., A.T. and K. ˙ Z.; Writing—original draft, B.S.-T., S.K., ˙ P.K., A.T. and K.Z.; Writing, review & editing, B.S.-T., S.K., P.K., A.T. and K. ˙ Z. All authors have read ˙ and agreed to the published version of the manuscript.

**Funding:** This publication is a component of scientific research carried out at the University of Economics in Cracow as a part of the research of the Potential Program in 2022 no 056/ZZE/2022/POT.

**Data Availability Statement:** Public databases.

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

#### **References**


#### *Article* **The Role of Pro-Innovative HR Practices and Psychological Contract in Shaping Employee Commitment and Satisfaction: A Case from the Energy Industry**

**Katarzyna Gadomska-Lila 1,\* and Anna Rogozi ´nska-Pawełczyk <sup>2</sup>**


**Abstract:** Innovation is a very important attribute of energy companies. Its level largely depends on employees' attitudes and behaviours, which are determined on the one hand by individual factors (e.g., psychological contract) and on the other hand by organisational factors (e.g., human resources (HR) practices). The aim of this article is to identify the relationship between pro-innovative HR practices, psychological contract and employee attitudes—commitment and job satisfaction. The research was conducted in a company which is one of the leading Polish electricity suppliers. Data were collected using a proprietary survey questionnaire. The research involved 402 HR professionals responsible for shaping and implementing pro-innovative HR practices in the company. Hypotheses were tested using the partial least squares structural equation modelling technique (PLS-SEM). The results indicate the existence of a positive relationship between pro-innovative HR practices, psychological contract and employee commitment and satisfaction. Furthermore, they indicate that organisations wishing to increase employee commitment and job satisfaction should strengthen and implement existing psychological contracts. One way to do this is to invest in pro-innovative HR practices, such as competence development, knowledge sharing or creativity-based candidate selection, as they have been shown to be good moderators of these relationships. The obtained results may be of particular importance for HR management specialists and managers responsible for shaping desired attitudes and behaviours of employees.

**Keywords:** energy sector; pro-innovative HR practices; psychological contract; job commitment; job satisfaction

#### **1. Introduction**

Innovation is one of the key determinants of the competitive advantage of companies. The level of organisational innovativeness fundamentally depends on employees' attitudes and behaviours. These attitudes and behaviours are, among others, the result of psychological contracts and the result of implemented management practices, including those identified as pro-innovative HR practices.

The energy sector is one of the most innovative in terms of applied technological solutions, conducted research projects and organisational solutions. The company surveyed is one of the leaders in implementing innovative products and services in many areas of its very broad and diverse activities. Its brand belongs to one of the largest companies in Europe, which makes it universally recognisable. Its well-established position on the Polish market, supported by the trust of institutional and individual clients, as well as its prestige make it attractive for research.

In terms of its wide range of activities, since 2014, it has also held an active concession for trading electricity in Poland. The company has been recognised as a leader in the implementation of new solutions and the use of tools not only in energy production

**Citation:** Gadomska-Lila, K.; Rogozi ´nska-Pawełczyk, A. The Role of Pro-Innovative HR Practices and Psychological Contract in Shaping Employee Commitment and Satisfaction: A Case from the Energy Industry. *Energies* **2022**, *15*, 255. https://doi.org/10.3390/en15010255

Academic Editor: Bernard Zi˛ebicki

Received: 26 November 2021 Accepted: 29 December 2021 Published: 31 December 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/).

and distribution or digital solutions, but also in management practices, especially proinnovative HR practices. The scale of operations and related needs justified the creation of a separate HR structure in the company under study, with several hundred employees specialising in this area. The aim of the article is to identify the relationship between proinnovative HR management practices, psychological contract, and employee commitment and satisfaction. The study undertaken analyses the impact of the application of proinnovative HR practices related to the deliberate implementation of modern technologies within an energy distribution company in Poland on employees' psychological contracts, sense of job satisfaction and employee commitment. The surveyed company is the largest alternative energy retailer in Poland. It is a leader in creating and implementing pioneering solutions in providing access to stable, sustainable and modern energy. In order to ensure the most substantive assessment, the study selected HR department employees, as the most competent to assess the usefulness of specific tools and solutions.

The subject of the research concerns management problems related to shaping certain attitudes and behaviours of employees, which are relevant to the energy sector. We considered this topic to be important because the innovativeness of organisations in this sector largely depends on the attitudes and behaviours of their employees, and these can be successfully shaped and developed using appropriate HR practices. The research subject, which was a leading energy company, is an excellent example of this.

The current study contributes to the extant literature and practice in several ways. First, this article can contribute to a better understanding of the relationship between pro-innovative human resource management practices, the psychological contract, and employee commitment and satisfaction. This will enable the formation of desirable employee attitudes and the selection of effective tools, especially from the group of pro-innovative HR practices. This topic is important, given the impact of attitudes on employee behaviour, including innovative behaviour, and at the same time the unpromising picture of the labour market and employment problems faced by companies in the energy sector. Second, the study has demonstrated the positive impact of the psychological contract on commitment, and confirmed the relationship between commitment and job satisfaction. Pro-innovative HR practices proved to be a good moderator of this relationship—their introduction strengthens the positive relation between the psychological contract and job satisfaction and commitment. Finally, in the utilitarian dimension, the results of the research may be useful to direct the activities of companies in the energy sector consisting in the search for effective ways of shaping employees' attitudes—commitment and job satisfaction, consequently favouring the development of an effective human resources management policy.

The article consists of five parts. After the introduction, the theoretical framework is presented along with the characteristics of the main variables and relations between them, as a basis for deriving research hypotheses. The following sections describe the research methodology and the research results. Finally, in the last section, a discussion and conclusions are formulated along with practical and theoretical implications as well as limitations of the study and further possible research directions.

#### **2. Theoretical Framework**

#### *2.1. Challenges for the Energy Sector in Poland—The Context of Research*

The energy sector is one of the sectors of key importance for the economy of any country. No economy can develop without predictable and reliable energy production, distribution and supply. However, the strategic importance of energy companies, resulting from the need to ensure uninterrupted energy supply, is confronted with the impact of their activities on a wide range of stakeholders and on the environment. Therefore, the issue of sustainable development is becoming a priority, and the directives announced by the UN under the name of "Sustainable Development Goals" are becoming a kind of guide, also for companies in this sector [1]. It is difficult to overestimate how impactful the energy sector is and its significance for sustainable development in social, environmental,

market and consumer issues. In Poland, issues related to the energy sector, such as ensuring energy security, innovativeness of this sector, intelligent energy management, transformation of the mining sector or diversification of supplies, remain a particular area of interest both for state authorities and managers of companies. Data from the Responsible Business Forum [2] show that there are many energy companies that take a responsible and strategic approach to the challenges of sustainable development—they undertake socially responsible initiatives, care for dialogue and transparency of activities, report non-financial data adhering to the best standards, etc. However, there are still many dimensions in which more activity would be desirable, including environmental, market and consumer ones, such as sustainable production and consumption, energy efficiency or renewable energy sources. It is necessary to invest in energy infrastructure and the creation of clean energy technologies, including renewable energy and cleaner fossil fuel technologies, as well as the improvement of technologies for the provision of modern and sustainable energy services. This emphasises the need to implement innovative solutions, drawing special attention to the human resources of companies in this sector. The challenges they are facing confirm that they have to recruit the best specialists and make sure that such specialists are willing to stay with the organisation as long as possible and that they are committed to achieving its goals. At the same time, energy companies are reporting an increasing demand for staff. Construction of new and expansion of existing power units, modernisation of the network, growing sales of energy, etc. generate the need for new specialised employees. In addition, the current situation on the Polish labour market, referred to as the employee market, means that many companies also have problems with retention and must strive to ensure that employees, especially key ones, remain in the organisation. This justifies directing attention to employee attitudes, especially commitment and job satisfaction. An important role is played here by psychological contracts concluded with employees, as well as the activity of HR department employees, who are responsible for shaping and implementing management practices that favour the development of specific attitudes and behaviours of employees. The specificity of the sector, especially the need to create and develop newer and smarter solutions and to implement the latest technologies, induces the search for HR practices which will positively influence employee involvement and satisfaction, as well as being conducive to innovation. This makes it possible, in a way, to address two key objectives from the Development Agenda—Goal 7 on ensuring access to affordable, reliable, sustainable and modern energy, and Goal 8 on promoting stable, sustainable and inclusive economic growth, full and productive employment and decent work for all people.

#### *2.2. Engagement and Job Satisfaction Versus Pro-Innovative HR Practices and Psychological Contract—Hypothesis Development*

Employees' attitudes are a frequent subject of research due to the repeatedly empirically confirmed strong relationship with behaviour and, consequently, with the quality and efficiency of job performance [3,4]. Findings of Harrison, Newman, and Roth [5] suggest that if employees display positive attitudes, they perform better professionally. The consequences of positive attitudes relating to the work environment are specific behaviours (e.g., innovative, proactive, citizenship) as well as job performance. Thanks to the analysis of attitudes, it is also possible to assess the effectiveness of investments in employees. Among the group of attitudes that are particularly effective are job satisfaction and commitment to work. Both attitudes refer to the appraisal and emotional response to the situation in the workplace [6,7], where satisfaction is a more emotional state of well-being resulting from the fulfilment of an employee's needs, and commitment refers to the investment of cognitive, emotional and physical energy in the performance of work activities [8].

Job satisfaction is the most commonly analysed employee attitude in research [9]. It is defined as a general attitude towards work [10] or as an employee's feelings about their job [11]. Robbins and Judge [12] add that this positive feeling about a job stems from an evaluation of its characteristics. As a multidimensional concept, satisfaction

includes several aspects of the job. Among these aspects, most researchers point to: job characteristics, relationship with co-workers, relationship with supervisors, recognition and benefits, promotion opportunities and fairness, working conditions, satisfaction with the company and management, communication [13,14]. Satisfaction is the extent to which an employee perceives job needs as fulfilled [15]. Research by Beal and co-authors [16] has shown that job satisfaction can enhance employees' concentration and attention, and ultimately lead to better job performance. On the other hand, Zahoor, Zia, and Rizwan [17] proved the effect of satisfaction on loyalty. It turns out that employees who feel job satisfaction manifest a significantly lower tendency to leave their jobs [18], while a decrease in the level of satisfaction results in an increase in the desire to leave the organization [19].

Work commitment can be defined as an attitude towards an organisation of identification with its organisational goals, willingness to exert effort on its behalf and a strong desire to maintain membership in the organisation [20]. It provides a measure of the strength of the employee's psychological identification with the job. The stronger the commitment, the more important work is to the person and the more difficult it is to separate self-esteem from the work itself [21]. It is generally accepted that people who are strongly committed to work put more effort into it and perform better. Lv and Xu's [22] research demonstrates that work engagement is positively correlated with job performance at both the firm and individual employee levels. Research confirms strong positive relationships between engagement and a variety of positive outcomes in organisations [23,24]. As work engagement increases, the level of work task completion and job support increases [25]. An engaged employee is less likely to be absent and at the same time less likely to leave the organisation [26]. Engagement influences performance, which in turn influences the taking of new, creative initiatives and the willingness to make changes [27]. In addition, employees who have high levels of engagement are more likely to exhibit innovative behaviour at work [28]. The positive emotions that accompany engagement not only motivate employees to take initiative and try different things [6], but also encourage them to collect feedback on their performance [23].

As a result of low work engagement and individuals' low sense of job satisfaction, low levels of work task completion and work support, as well as citizenship and adaptive behaviour can be predicted [8]. Due to the significant impact of commitment and satisfaction on employee behaviour, it is necessary to search for factors that will positively influence these attitudes. Both individual and organisational variables seem to be important. Literature studies confirm that an important individual factor is the psychological contract, which influences employee attitudes and behaviours as well as organisational performance [29]. The psychological contract is an implicit agreement between the individual and the organization that defines what each person expects to offer and receive. It is a set of people's beliefs about the contract that defines the terms of exchange and the resulting benefits for both employer and employee [30]. It develops from the employee's belief in the mutual obligations between him or her and the employer, becoming the basis of the employment relationship [30,31], directing what the organisation expects of him or her and how it will reciprocate for the effort and actions the employee has made on its behalf [32]. This relates to various aspects of the employee relationship within the organisation [33] and the consequences of accepting and fulfilling mutual obligations to achieve the organisation's goals [34]. The psychological contract complements the formal arrangements, so to speak, and indicates expectations that may relate to economic, organisational or psychological issues. It can help employers understand and predict employee behaviour, induce employees to become more committed, and cause employees to pursue company goals to a greater extent [35]. The implementation of psychological contracts is associated with attitudes such as job satisfaction [36], organisational commitment [37] or innovative behaviour [38]. At the same time, an organisation's failure to fulfil its contractual provisions results in negative attitudes and behaviours, such as intentions to leave [39] or failure to exhibit citizenship behaviours [40]. At the organisational level, the variable that significantly influences the formation of employees' attitudes, as well as skills and behaviours, are HR

practices [41,42]. They have a synergistic effect that influences employees' attitudes towards work [43]. In particular, they affect job satisfaction and commitment [7,44]. Research by Okabe [45] confirms that HR practices can increase job satisfaction, facilitate knowledge sharing among employees, and contribute to the perception of a more welcoming workplace. This in turn can contribute to more innovative behaviour [46]. Effective HR practices make competent and innovative employees contribute to the achievement of organisational goals [47]. Research also indicates that HR practices influence perceptions of psychological contract [35,48,49]. It is one of the strongest influences on the psychological contract, factors [47]. HR practices and the employment relationship play a very important role in forming and shaping contracts [50]. The organisation and its agents [51] communicate promises and expectations to employees in both overt and covert ways. Overt ways usually include formal contracts and all kinds of written communications and direct interactions, while covert forms of communication are more complex and subtle. An employee can learn about mutual commitments and promises by observing the behaviour and reactions of others in the organisation. Recent research by [52] indicates that the formation of a psychological contract in an organisation is influenced by such perceived commitments as job retention, rapid promotion, satisfactory pay, training and career development. These commitments provide the organization with the opportunity for employees to become more involved in organizational activities and goal achievement.

At the same time, a certain group of HR practices is seen as particularly supportive of organisational innovation [53]. In the managerial dimension, innovation is the ability to recognise and respond not only to current, but also future needs and challenges [54,55]. Research by Oladapo and Onyeaso [56], for example, confirms the link between high performance work practices and organisational innovation. This group of practices is referred to as pro-innovation practices. These include, for example, an emphasis on recruiting staff who are creative and open to change, an emphasis on training and development in the context of innovation, or an emphasis on motivating employees to undertake innovative ventures [57]. Pouwels and Koster [58] also add that higher performance in product and service innovation, market and process innovation is achieved by organisations using training and performance-related pay. Furthermore, research by Cheng and Huand [59] indicates that organisations that invest in strategic HRM are better able to manage their internal knowledge. This in turn has a positive impact on the extent to which these organisations are able to innovate. Developing employees' skills and offering ways to optimise their work (through learning, employability, etc.) are mechanisms through which HRM can contribute to organisational innovation [60].

Various factors, both at the level of the individual (psychological contract) and at the level of the organisation (pro-innovative HR practices) influence employee attitudes, especially satisfaction and commitment, and translate into constructive behaviour in the work environment.

Given the considerations presented, we conclude that knowledge of the relationship between pro-innovative HR practices, psychological contract, job satisfaction and commitment is fundamental in the effort to shape desirable employee attitudes and behaviours. Based on the findings, we pose the following hypotheses:

**Hypothesis 1 (H1).** *Psychological contract has positive impact on employees' job satisfaction.*

**Hypothesis 2 (H2).** *Psychological contract has positive impact on employee commitment.*

**Hypothesis 3 (H3).** *Employee commitment has positive impact on employees' job satisfaction.*

**Hypothesis 4 (H4).** *Implementing pro-innovative HR practices strengthens the positive relation between psychological contract and employees' job satisfaction.*

**Hypothesis 5 (H5).** *Implementing pro-innovative HR practices strengthens the positive relation between psychological contract and employee commitment.*

**Hypothesis 6 (H6).** *Implementing pro-innovative HR practices strengthens the positive relation between employee commitment and employees' job satisfaction. between employee commitment and employees' job satisfaction.*  A research model was created to show the relationship between the tested features. This

**Hypothesis 3 (H3).** *Employee commitment has positive impact on employees' job satisfaction.* 

**Hypothesis 4 (H4).** *Implementing pro-innovative HR practices strengthens the positive relation* 

**Hypothesis 5 (H5).** *Implementing pro-innovative HR practices strengthens the positive relation* 

**Hypothesis 6 (H6).** *Implementing pro-innovative HR practices strengthens the positive relation* 

A research model was created to show the relationship between the tested features. This is illustrated in Figure 1. is illustrated in Figure 1.

*Energies* **2022**, *15*, x FOR PEER REVIEW 6 of 21

*between psychological contract and employees' job satisfaction*.

*between psychological contract and employee commitment.*

**Figure 1.** Hypothetical research model. Source: created by the authors. **Figure 1.** Hypothetical research model. Source: created by the authors.

#### **3. Materials and Methods**

#### **3. Materials and Methods**  *3.1. Research Entity*

*3.1. Research Entity*  The surveyed entity belongs to a group that is a leader on the Polish fixed telephony, Internet and data transmission market, which, as an innovative provider of ICT and telecommunications services, has the largest, most technologically advanced infrastructure in the country. For seven years, it has also been selling energy under its own brand. Currently, it is the largest alternative energy seller in Poland, with a customer base of over 120,000. Over 97% of households are within the reach of its offer. The company creates and implements pioneering solutions in the scope of providing access to stable, balanced and modern energy. It vigilantly observes the environment and catches all market signals and business opportunities. Based on its strengths, which are the complexity of solutions and knowledge of customer needs, it constantly creates new solutions and enriches its offer. It focuses on innovation. It carries out investments and pro-environmental projects. Currently, it is investing in obtaining electricity from environmentally neutral sources (two wind farms launched for the company), which is part of a broader, ambitious plan— The surveyed entity belongs to a group that is a leader on the Polish fixed telephony, Internet and data transmission market, which, as an innovative provider of ICT and telecommunications services, has the largest, most technologically advanced infrastructure in the country. For seven years, it has also been selling energy under its own brand. Currently, it is the largest alternative energy seller in Poland, with a customer base of over 120,000. Over 97% of households are within the reach of its offer. The company creates and implements pioneering solutions in the scope of providing access to stable, balanced and modern energy. It vigilantly observes the environment and catches all market signals and business opportunities. Based on its strengths, which are the complexity of solutions and knowledge of customer needs, it constantly creates new solutions and enriches its offer. It focuses on innovation. It carries out investments and pro-environmental projects. Currently, it is investing in obtaining electricity from environmentally neutral sources (two wind farms launched for the company), which is part of a broader, ambitious plan—to achieve climate neutrality by 2040. The company's offer also includes Smart Energy solutions, comprising efficient, modern photovoltaic installations producing green energy from solar radiation and a comprehensive monitoring and management system, unique in the market, and as indicated among others by the research of Gr˛ebosz-Krawczyk and colleagues [61] on consumers' purchase intentions towards the installation of photovoltaic panels, this is

to achieve climate neutrality by 2040. The company's offer also includes Smart Energy

a future-oriented development. The researched company is a leader in introducing innovative solutions in various areas of activity, as well as in the area of human resources management. It encourages its employees to search for and create technological, product and organisational solutions, favourable to energy efficiency, reduction of negative impact on the environment or reduction of resources consumption. It appreciates the importance of human resources. It pursues business objectives with respect for the subjectivity of employees, uses digital solutions, and creates a culture of cooperation in which all employees feel respected and are able to freely pursue their professional objectives.

With the company being recognised as implementing innovative solutions in various areas of activity, including management, the legitimacy of implementing pro-innovative HR practices was analysed in the context of influencing employees' psychological contracts and their attitudes—commitment and satisfaction. In order to ensure the most substantive assessment, HR employees comprising over 400 people were selected for the study.

The choice of HR employees was dictated by the fact that they are best qualified to assess, through their experience, the usefulness of new tools as well as personnel or management solutions. Therefore, the usefulness of these solutions was assessed by all HR department employees, who have the broadest competence in this area and therefore their answers, based on their professional experience, are the most reliable. The research has shown that there is a very large specialised group within this innovative structure, which through its involvement in the research has also in a way confirmed that it was selected accurately.

#### *3.2. Testing Procedure*

To achieve the objective of this research and to verify the hypotheses, the survey was conducted in December 2019. The sampling frame consisted of 402 respondents who represented 100% of the HR department employees of the surveyed company. The personnel of the selected organization was surveyed using the "Employee satisfaction survey" method, which was developed by the Partner in the "AnStat" Business Strategies as per the organization's Strategy of Development for the years 2015–2020. We used a self-administered questionnaire consistent with the needs of the contracting authority and purposive sampling [62].

In the first stage, a pilot test was conducted with ten selected HR department staff. During this research process, respondents checked the content and relevance of each item to ensure every question was adequate and understood. Based on the feedback from the pre-test, minor modifications were made to the questionnaire.

In the second stage, a sample of 402 employees was purposively selected for the study and represented 100% of the HR department employees of the surveyed organisation. The quantitative survey followed a standardised procedure [63], through statements aggregated into a diagnostic tool which used the CATI computer-assisted telephone interview technique. The data collected from each interview were analysed for the questionnaire path that was scripted and also for the consistency of the tool. The average duration of the survey (median) was 19 min with an employee.

Table 1 details the structure of the study sample, which presents: gender, age, level of education, total work experience and length of work experience in the current com-pany, as well as the type of HR position in the professional hierarchy.

The results show that the group working in the HR department is relatively feminised, as almost 65% are women. The large portion of participants employed in the HR department are between 30 and 50 years old. The smallest groups are surveyed employees under 30 years old (10.2%) and 55 and over (5%). The majority of participants (82.3%) are highly educated, having a university degree or above. The company has a preference for recruiting employees to the HR department already with significant work experience, as more than half of the participants have more than 10 years of work experience and a quarter of the surveyed group has worked for up to 10 years in total. In addition, employees working in HR department are characterised by long seniority—66.6% have been working in the

surveyed company for 6 years and more. The distribution of positions in the professional hierarchy is standard, with the majority of survey participants employed as HR assistants and senior HR assistants (74.8%) and a minority of managers and executives (5.2%).


**Table 1.** Demographic description of respondents.

Source: created by the authors.

#### *3.3. Measures, Validation and Reliability Analysis*

Fundamental in the research process was the design of measurement instruments to collect data on the main constructs within the proposed hypotheses. The quantitative study, which used the CATI computer-assisted telephone interview technique to collect empirical material, involved a battery of tests aggregated to a single questionnaire.

The measurement of the study variables was carried out using a set of items derived from adaptations of existing research tools, which were selected on the basis of a literature review. Some tools required cultural adaptation prior to the research. For this purpose, one of the procedures of tool adaptation [64] was used; translation and post-translation from the Polish version of the text to the original (English language) version, in order to faithfully translate the methods used. The translated questionnaires were evaluated by competent judges, who were three independent experts in the fields of English philology, HRM and statistical methods. Attention was also taken to the criterion of equivalence of the questionnaires in the form of: facial, psychometric, functional equivalence and reconstruction validity.

The questionnaire consisted of 35 items taken from the literature measuring basic constructs, such as implementing pro-innovative HR practices, psychological contract, employees' job satisfaction, employee commitment, and the demographic characteristics of respondents (indicated in Table 1), which played the role of control variables. A list of the key constructs used with associated items is presented in Appendix A.

#### *3.4. Implementing Pro-Innovative HR Practices (IPHRP)*

Since the impact of HR practices is not immediate and not always in line with employers' expectations, its effect can manifest in the meanings that employees attribute to these practices. The IPHRP construct intended for opinion by employees measured the implementation of pro-innovative HRM practices implemented in the company. It was divided into 4 subscales: (IPHRP I) Personnel pro-innovative activities in the area of employee recruitment (2 items); (IPHRP II) Personnel pro-innovative activities in the area of professional development (3 items); (IPHRP III) Personnel pro-innovative activities in the area of motivation and remuneration (3 items), and (IPHRP IV) Personnel pro-innovative activities in the area of maintaining positive employee relations (4 items). In developing the items within each subscale, we followed the indications by Kinnie and collaborators [65] and Wang and collaborators [66]. As part of the measurement of IPHRP, employees were asked for their opinion on how satisfied they were with the implementation of pro-innovative company's HR practices. All research measures were carried out by a seven-point Likert scale (where 1 = "not at all" and 7 = "to a very high degree") for each item.

#### *3.5. Psychological Contract (PC)*

The next section of the questionnaire examining the PC was measured by a single variable: the perceived degree to which the psychological contract was fulfilled. The study measured the extent to which obligations under the PC were fulfilled by the employee and the employer. Items to measure the PC were adapted from the study by Rousseau and Tijoriwala [67] and Guest and Conway [68]. To measure the fulfilment of the employer's and employees' obligations two subscales were used, each containing 9 and 8 items, respectively, such as "To what extent do you fulfil the promises and obligations you made to your employer?" and "How would you rate the fulfilment of promises and obligations by your employer?" All items were assessed by respondents on a seven-point Likert scale, where 1 was "not filled at all" and 7 "completely filled".

#### *3.6. Employees' Job Satisfaction (EJS)*

EJS was measured using three items diagnosing general job satisfaction, attitude towards the job and interest in the job. This construct was diagnosed using modified statements taken from the Job Satisfaction Scale by Bettencourt, Gwinner and Meuter [69].

#### *3.7. Employee Commitment (EC)*

EC is conceptualised here as the extent to which the employee personally identifies with his or her work and is intra-orientated and motivated towards work. Employee engagement was assessed using three-item scales which were adapted from the studies by MOW [70] and Saks [71].

For the questions in the tool, regarding EJS and EC, respondents were asked to rate the extent to which they agreed or disagreed with the statements on a seven-point Likert scale from 1—"strongly disagree" to 7—"strongly agree".

#### *3.8. Data Analysis*

The purpose of this paper is to examine how the introduction of pro-innovation practices into an organisation strengthens the role of psychological contracts between the organisation and its employees and influences employees' attitudes at work—commitment and job satisfaction. This objective was achieved through the following research steps:

Stage 1 consisted of desk research on the current situation of the energy sector in Poland and an introduction to the concepts of pro-innovative HR practices, psychological contract, commitment and job satisfaction (chapter: Theoretical Framework).

Stage two focused on the development of 6 research hypotheses (subsection: Engagement and job satisfaction versus pro-innovative HR practices and psychological contract hypothesis development).

In stage 3, the research methodology was developed, the survey questionnaire was constructed and validated, the research sample for the study was planned and the pilot study and proper research were conducted (chapter: Materials and Methods).

Step 4 involved the processing of the statistical data obtained from the quantitative CAWI survey (subsection: Data analysis; chapter: Results)

In step 5 we analysed the results of the survey, synthesised and discussed the results of the survey at the organisational level while presenting the research limitations. (Chapter: Discussion).

Quantitative social survey methodology was used in the data compilation. Descriptive statistics such as frequencies, means and standard deviations were used to describe the distributions of respondents in terms of sociodemographic characteristics. The main statistical techniques used for data analysis were EFA, CFA and SEM.

The research process was conducted with exploratory factor analysis (EFA) to possible reduce items and define patterns of constructs in the data set [72,73]. EFA was used to eliminate measurement elements that had less than 0.30 factor loading or had high cross loading over two or more factors [72].Then, Cronbach's Coefficient Alpha was used to control the reliability of internal consistency items [73]. Exploratory factor analysis (EFA) was conducted on the results of the pilot study.

The matching factors were then used in the next step. We used confirmatory factor analysis (CFA) to develop a measurement model and to verify that each of the variables proposed in the research model is valid for the baseline constructs. This research employed confirmatory factor analysis (CFA) with maximum likelihood estimation [74] to test the significance of items' factor loadings to ensure that the items were not loading onto other constructs and to assess the reliability and validity of the multi-item scales. Discriminant validity was assessed in CFA because factor loadings in each construct can be represented by specific scores. CR and AVE also used for assessment in CFA to represent the quality of the constructs. Tables 2 and 3 set out our empirical results in detail.


**Table 2.** Constructs, measures and factor loadings.


**Table 2.** *Cont.*

Source: created by the authors.

**Table 3.** Convergent validity of the measurement model.


Source: created by the authors.

All factor loadings were higher than 0.7 and all t-tests of the variables were significant at the 0.001 level with t-values greater than 2, suggesting adequate level of reliability. The reliability of the scales was conducted by calculating composite reliability (CR) and Cronbach's α. Cronbach's alpha values for all scales exceeded 0.7 and exceeded the recommended threshold level [73]. Composite reliability (CR) of all constructs exceeded the threshold value of 0.7, suggesting internal consistency and reliability [75]. The average variance extracted (AVE) values were higher than 0.50. In addition, all the square roots of AVE are also greater than inter-construct correlations. Thus, the convergent validity of the scales was confirmed [75].

Table 3 suggests good validity of the measurement model. This is also confirmed by the results of the chi-square test, where *p* < 0.001 was obtained, Further measures confirm good model fit—χ2/df < 5 [76], RMSEA is less than 0.8 [77], and SRMR is less than 0.08 [76], CFI > 0.9 [78], GFI, AGFI, and gamma exceed 0.95 [76], and Bollen's delta is 0.9, and the parsimonious James-Mulaika-Brett index (PGFI) is greater than 0.7. Importantly, for each of the variables, the parameters (factor loadings and measurement error variance) are statistically significant (by item *t*-test, *p* < 0.0001). The correlation between the variables is also significant (*p* < 0.0001).

Summing up, the CFA supported our measurement model and showed that implementing pro-innovative HR practices, psychological contract, employees' job satisfaction and employee commitment are four distinct constructs.

In a final step, partial least squares structural equation modelling (PLS-SEM) was used to determine the assumed relationships between all constructs. In SEM, the direct and indirect effects of endogenous variables provide the overall relationship between all constructs [74]. In addition, discriminant and convergent significance were represented in SEM, indicating the overall validity of the overall constructs. The SEM procedure including data analysis is further discussed in the next section.

The database and statistical tests were conducted using IBM SPSS PS Clementinei IBM SPSS PS Imago computer software. The Statistica statistical package was used for the analyses, which enabled the data to be entered in numerical form and then carried out statistical analyses, facilitating and accelerating the process of compiling the research results.

#### **4. Results**

The theoretical framework of our study became the impact of introducing pro-innovative HR techniques on the relationship between psychological contract, employee commitment and employee's job satisfaction. We assumed that implementing pro-innovative HR practices would have a positive impact on these relationships. In order to test the proposed research hypotheses and to empirically verify the hypothesized research model, we performed partial least squares structural equation modelling. PLS-SEM method is frequently selected to test and develop a theory [79], as well as used in business management or management information systems. Ringle, Sarstedt and Straub [80] argue that the most commonly mentioned reasons for adopting PLS-SEM are associated with non-normal data, small sample sizes and the use of formally measured latent variables. In our study, measures and indicators were developed for the following research constructs: psychological contract, job commitment and employees' job satisfaction.

The size of the research sample (402) that we have selected for the study is also comparatively limited, as the entire Company employs 14,587 people. The study sample thus represents only 2.75% of the total workforce. Therefore, we conclude that PLS-SEM is a better and an appropriate method than other tests to examine the research hypotheses of this study. Table 4 shows that the four hypotheses we developed are supported based on the evaluation of the PLS-SEM model.

Hypothesis H1 must be rejected because psychological contracts have no significant impact on employee's job satisfaction (H1: β = 0.215; t = 1.461, ns). The results obtained are contrary to the research of Khorev and Wechtler [81], who support the idea that positive fulfilment of employers' promises is related to job satisfaction behaviours. Our result suggests a full mediation of employee commitment on the relationship between psychological contract and employee's job satisfaction.

As expected, psychological contract has significant positive impact on employee commitment (H2: β = 0.5040; t = 3.049, *p* < 0.01), which suggests that employees with a sense of fulfilled psychological contract are more committed in their work. Our results confirm that employee commitment has a positive significant impact on employee job satisfaction (H3: β = 0.646; t = 3.647, *p* < 0.05).


**Table 4.** Hypotheses tests results.

Note. \* *p* < 0.1; \*\* *p* < 0.05; \*\*\* *p* < 0.01; ns (not significant). Source: created by the authors.

When testing the moderating effect of implementing pro-innovative HR practices on the relationship among psychological contract, employee commitment and employees' job satisfaction, our study indicates the confirmation of H 4–5. It is found that implementing pro-innovative HR practices strengthens the positive relationship between psychological contract and employee job satisfaction (H4: β = 0.758; t = 5.040, *p* < 0.05). Similarly, implementing pro-innovative HR practices enhances the positive relationship between psychological contract and employee commitment (H5: β = 0.471; t = 3.655, *p* < 0.05).

Our study finds that the implementation of pro-innovative HR practices has no significant impact on the relationship between employee commitment and employees' job satisfaction (H6: β = 0.189; t = 0.467, ns).

#### **5. Discussion**

The aim of this study was to investigate the relationship between employees' perceptions of pro-innovative HR practices, psychological contract and their level of job commitment and job satisfaction. The results generally supported the hypothesis that such relationships exist.

Evidence from existing literature shows that employee commitment is crucial for organizational welfare [15,82]. The results of our study show that employee commitment is positively influenced by psychological contracts and employee job satisfaction.

In particular, we found that the psychological contract has a significant positive impact on employee commitment. This finding reinforces well-established knowledge about psychological contracts in the literature. According to Chambel and Oliveira-Cruz, psychological contract fulfilment has a positive impact on employee commitment [83]. The results of Lv and Xu [22] research analysis confirmed that psychological contract breach has a strong negative impact on employee commitment.

Employees who implement psychological contracts know what conditions and level of support they can expect from their employers. In the short–term relationships, they know that their efforts are rewarded in the form of financial compensation without any further obligations or demands [84]. As long as the amount of financial reward received matches the employees' perceptions of justice and fairness, they will seek to maintain a positive relationship with their employer [85]. Employees who have long–term relationships influence the environment with their attitudes and behaviours. In a similar way, their attitudes and behaviours are influenced by their employer [86]. Thus, employees participate in creating the culture and can influence its change. They then have a sense of ownership because they see that their employer is interested in their professional growth, motivation, adequate remuneration, and that they themselves are working for the benefit of the organisation in which they are employed. Furthermore, the results of our survey show that employee commitment has a direct and positive impact on employee satisfaction. This turns out

to be an important, albeit different, finding, as previous research suggests that higher levels of satisfaction are associated with a number of benefits, such as improved working conditions, interpersonal relationships, or employer branding [87]. This is consistent with Coyle-Shapiro and Conway and Xiong and collaborators [37,88] who argue that employees feel more job satisfaction when they are more committed in their work.

Our results indicate that the relationship between psychological contract and employee job satisfaction is fully mediated by employees' commitment. Our study indicates that without sufficient commitment, employees are less likely to increase their sense of job satisfaction based on their psychological contract.

In the work context, pro-innovative HR practices are strong predictors of psychological contracting because these shape employee behaviour and are the main mechanisms by which employees understand the terms and conditions of employment [89].

Our results show that implementing pro-innovative HR practices increases employee commitment levels. Although this effect has already been verified theoretically [90], to the best of our knowledge, this study is the first to show this effect evidenced through conducted research. The results achieved appear to be relevant, as the increase in employee commitment is linked to the fulfilment of the psychological contract. Thus, we also found that as pro-innovative HR practices were introduced into the company, the positive impact of psychological contracts improved. This is the first study to examine changes in the strength of psychological contracts due to the implementation of pro-innovative HR practices.

The results of the study presented remain consistent with previous studies which have shown that pro-innovative HR practices appropriately selected in an organization, i.e., training and development [91], reward [22], organization of individual and team work [92], communication [93] and stable employment [94], enhance attitudes of job satisfaction and a sense of fulfilment of the psychological contract. In contrast, the results obtained are contrary to suggestions of other researchers [50,95,96] who pointed that implementing pro-innovative HR practices has negative effects of uncertainty at work and weakens employee commitment.

The pressure to use pro-innovative HR practices contributing to fulfilling SDG 7 and 8 will continue to intensify, whether due to high levels of competitiveness, changing working conditions, or consumer preferences [97,98]. Implementations of pro-innovative HR practices are gaining significant support from top managers and HR professionals, particularly in energy sector organisations. Employees need to feel a sense of fulfilling the psychological contract and be positively committed in the implementation of innovative HR practices in order to fully gain the advantages. Our study shows how employee commitment increases as a result of adopting and implementing pro-innovative HR practices. This study has demonstrated the importance of psychological contracts when introducing pro-innovative HR practices. This means that psychological contracts will play a greater role in shaping the relationship between employers and employees as pro-innovative HR practices spread more widely. Finally, the implementation of pro-innovative HR practices leads to employees fulfilling psychological contracts while being perceived as committed and satisfied with their work, which will move organisations with a particular focus on the energy sector towards achieving SDGs 7 and 8.

#### *5.1. Theoretical Implications*

The theoretical contribution of this study is manifested in several aspects. First, we identified relationships between pro-innovative human resource management practices, psychological contract, commitment and job satisfaction. We confirmed that the psychological contract positively influences commitment and that commitment is associated with job satisfaction.

Second, we found that pro-innovative HR practices are a good moderator of this relationship; the introduction of pro-innovative HR practices strengthens the positive relationship between psychological contract and job satisfaction and commitment.

It is also important that this study integrates individual variables (psychological contract) and organizational variables (pro-innovative HR practices) and employee attitudes, which enriches the existing literature on the mechanism of interaction and extends the researchers' theoretical view.

Finally, this study integrates HR practices into a group of pro-innovation practices and assesses employee perceptions of these practices, which demonstrates that HR practices should be analysed holistically rather than as independent practices, because, as Snape and Redman [99] found, they can substitute, complement or interfere with each other in predicting employee behaviour.

#### *5.2. Practical Implications*

The completed research also has practical implications. First of all, the results of the research may direct the activities of companies from the energy sector, regarding the search for effective ways of shaping employees' attitudes—commitment and job satisfaction, which may contribute to shaping desired behaviours, including innovative ones. They provide, especially to employees responsible for human resources management, knowledge in the selection of tools and solutions to stimulate commitment and job satisfaction. This may be helpful in developing a more effective human resources management policy. The project also draws attention to the importance of psychological contracts and thus encourages respect for them, as well as the design of HR practices in such a way that employees feel supported by the organisation and reciprocate this support through higher levels of commitment to the organisation's goals.

#### *5.3. Limitations and Further Research*

Like most studies, this study also has several limitations. One of the limitations of the study is related to the research methodology used, which was limited to using data from a single source (i.e., reports from employees) through survey methods. Therefore, it would be advisable to conduct further research using, for example, longitudinal studies, which could show a more precise picture of the relationships found and would allow the directions of causality to be explored. Furthermore, the method of data collection may result in a common method bias, which may affect the magnitude of the relationship between the variables under study. To overcome this limitation, all variables were measured using fixed scales, which may reduce measurement error and thus reduce common method bias.

Another limitation stems from the fact that the subject of the study included employees of a single company, albeit thriving in the energy industry. It would therefore be useful to widen the circle of research, which would allow us to establish how these practices are perceived by employees of other energy companies. Moreover, the data were collected in a company operating in specific—Polish—cultural conditions. On the one hand, this may be a limitation of the research, as it narrows the possibility of generalising the results to other countries. On the other hand, however, it provides knowledge on attitudes and behaviours related to the application of pro-innovative HR practices and psychological contracts, characteristic of this cultural circle.

In the course of the research investigation, further interesting research directions also emerged. Firstly, as we focused on employee engagement and satisfaction, it would be worthwhile to widen the circle of research and verify how pro-innovative HR practices and the psychological contract affect other attitudes and behaviours of energy sector employees, e.g., attachment to the organisation. Another interesting direction of research appears to be the analysis of different types of the psychological contract—relational and transactional.

The study was limited to HR professionals of one energy organisation and therefore the results cannot be generalised to other professions. The participants in this study were senior and mid-level HR professionals. Future research recommends including more highlevel HR professionals as well as Industry 4.0 experts in the sample to understand the role of HR in Industry 4.0. Future research may seek to use quantitative research methods to validate the findings and conclusions of this study.

#### **6. Conclusions**

The issues addressed, supported by the research discussed in the text, made it possible to grasp the relations between the key individual (psychological contract) and organisational (pro-innovative HR practices) variables in a broader perspective, both practical and theoretical. They made it possible to give them a fuller context by capturing the relationships and measuring their strength in a specific way. This justifies the derivation of a positive prognosis as to the topicality of the undertaken issues and the adoption of this research as a starting point for further scientific exploration. Research confirmation of the relationships between these variables may prove useful for organisational practice in enterprises basing their operational and organisational activity on innovative solutions (similarly to the research subject). For the above reasons, the authors remain convinced of the relevance of the undertaken subject matter, its scientific significance and practical usefulness for the functioning of entities operating in the field of modern technological solutions, manifesting their readiness to use innovative organisational solutions, which undoubtedly include energy companies using modern technologies.

**Author Contributions:** General concept, K.G.-L. and A.R.-P.; theory, K.G.-L. and A.R.-P.; methodology, K.G.-L. and A.R.-P.; validation and formal analysis, A.R.-P.; investigation and data curation, A.R.-P.; preparation of the draft version, K.G.-L. and A.R.-P.; preparation of the final version, K.G.-L. and A.R.-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 A.R.-P.

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

#### **Appendix A. Key Constructs, Subscales and Items**

Implementing Pro-innovative HR Practices (IPHRP) [4,5]

(IPHRP I) Personnel pro-innovative activities in the area of employee acquisition related to:

(IPHRP I\_1) cooperation with newly recruited employees

(IPHRP I\_2) the specificity and nature of one's work

(IPHRP II) Personal pro-innovative activities in the area of professional development related to:

(IPHRP II\_1) development opportunities offered by the company

(IPHRP II\_2) the implementation of programmes designed to develop high-potential employees (talents)

(IPHRP II\_3) flexible and task-based working conditions offered by the company

(IPHRP III) Personal pro-innovative activities in the area of employees' motivation remuneration of related to:

(IPHRP III\_1) the remuneration

(IPHRP III\_2) the personalised additional package offered (financial and non-financial total rewards)

(IPHRP III\_3) the functioning of the performance evaluation system

(IPHRP IV) Personal pro-innovative activities in the area of maintaining positive employee relations related to:

(IPHRP IV\_1) maintaining relations and cooperation with employers in the performance of their professional duties

(IPHRP IV\_2) maintaining relations and cooperation with co-workers as part of their work on innovative projects

(IPHRP IV\_3) opportunities for involvement in decision-making concerning important issues of the company

(IPHRP IV\_4) opportunities to create new and innovative products and services for the company's clients

Psychological Contract (PC) [6,7]

(PC I) Fulfilment of the psychological contract by employees towards the employer in the opinion of employees

(PC I\_1) Promote and maintain a positive employer branding (company reputation)

(PC I\_2) Demonstrate loyalty to the organisation

(PC I\_3) Develop new knowledge, skills and experience

(PC I\_4) Flexible thinking and proactive approach to solving arising problems

(PC I\_5) Be courteous to customers, clients and other colleagues

(PC I\_6) Take initiative and put forward new, innovative ideas

(PC I\_7) Achieving expected results of strictly defined tasks (hard work for the organisation)

(PC I\_8) Maintain high level of attendance and punctuality

(PC I\_9) Willingness to work additional hours if the needs of the employer change

(PC II) Fulfilling the psychological contract by the employer towards employees in the employees' opinion

(PC II\_1) Opportunities to participate in attractive and innovative trainings, mentoring and opportunities for professional development within the organisation

(PC II\_2) Opportunities for promotion within the organisation

(PC II\_3) Providing acknowledgement to the employee for initiating and implementing new ideas

(PC II\_4) Offering the employee interesting tasks to perform

(PC II\_5) Guaranteeing a pleasant and safe working environment

(PC II\_6) Ensuring job security and stability

(PC II\_7) Providing total rewards commensurate with outputs achieved

(PC II\_8) Offering a non-wage benefits package

Employees' Job Satisfaction (EJS)

(EJS 1) I like my job more than the typical employee

(EJS 2) I am enthusiastic about my work most days

(EJS 3) I feel that my work is more interesting than others

Employee Commitment (EC)

(EC 1) I work according to the values of the company

(EC 2) I am happy to be part of the organisation

(EC 3) The organisation where I work encourages me to improve my work performance

#### **References**


**Joanna Toborek-Mazur and Magdalena Wójcik-Jurkiewicz \***

Department of Accounting, Institute of Management, College of Management Sciences and Quality, Cracow University of Economics, 31-510 Kraków, Poland; toborekj@uek.krakow.pl **\*** Correspondence: magdalena.wojcik-jurkiewicz@uek.krakow.pl

**Abstract:** The motive for the functioning of enterprises in the market economy is the maximization of benefits by achieving positive financial results with the minimum possible involvement of capital. The positive results achieved are perceived very positively by the stakeholders. To achieve this, it is necessary to carry out reorganization processes (mergers and acquisitions) within capital groups, which are perceived as necessary (positive) measures. These phenomena have been observed in the energy sector for many years, and the years 2010–2021 were adopted as the research period. The aim of this article is to attempt to answer the question of whether the multi-energy concern is the result of activities carried out in light of the 2030 Agenda. The effect of merging the entities is capital concentration. This fact has been verified by reviewing changes in the structures of capital groups in the energy sector against the background of the largest mergers and acquisitions in Poland in terms of value. The theoretical part presents the differences in the terminology of mergers and acquisitions and the motives for their creation and operation, with an emphasis on energy groups. Therefore, the authors undertook research aimed at identifying and assessing the main premises and effects of capital concentration through mergers and acquisitions in the energy sector in Poland.

**Keywords:** multi-energy concern; Agenda 2030; the energy sector in Poland; capital concentration; mergers and acquisitions

#### **1. Introduction**

Poland, along with nearly 190 other countries around the world, joined the Paris Agreement on climate protection in 2015. The purpose of the agreement is to reduce the negative impact of greenhouse gas emissions on the natural environment by implementing specific solutions on a national (national action plans) and international scale (e.g., providing support to countries in reducing the effects of climate change). During the COP26 (Conference of the Parties) climate summit held in November 2021 in Glasgow, Poland committed to a gradual shift away from coal by the 40s of the 21st century, using, inter alia, the resignation from investments in coal, a just transformation of coal sector employees (on 28 May 2021, a social agreement was signed with miners assuming a schedule for the gradual closure of mines), or an increase in spending on the development of clean technologies [1–3].

An indispensable element that determines the possibility of Poland gaining a competitive advantage in Europe is the concern to ensure an appropriate level of energy security, understood as a guarantee of resistance to any unfavorable and unpredictable events that may threaten the physical integrity of electricity flows. One such unfavorable phenomena may be the limited availability of energy and the possibility of purchasing it at reasonable and acceptable prices. It is worth emphasizing that for many years, the level of energy demand has been increasing, which should encourage the representatives of the authorities to provide an appropriate system for diversifying electricity sources. Preferably, it should be supplied from as many sources as possible and preferably be renewable, thus ensuring adequate flexibility to meet the collective needs of the state [4–6].

**Citation:** Toborek-Mazur, J.; Wójcik-Jurkiewicz, M. Multi-Energy Concern as an Example of the Implementation of Agenda 2030: Poland as a Case Study. *Energies* **2022**, *15*, 1669. https://doi.org/10.3390/ en15051669

Academic Editor: Edmundas Kazimieras Zavadskas

Received: 13 December 2021 Accepted: 14 February 2022 Published: 23 February 2022

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

**Copyright:** © 2022 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/).

The aim of the article is to try to answer the question of whether the multi-energy concern created in Poland is the result of activities carried out in the light of the 2030 Agenda [7–12]. The effect of merging the entities is capital concentration. This fact was verified by reviewing changes in the structures of capital groups in the energy sector against the background of the largest mergers and acquisitions in Poland in terms of value. The analysis of investment motives in the merger processes allowed the authors to refer to the goals and effects of capital concentration in the energy sector.

Mergers and acquisitions are phenomena that are also carried out in the energy system. This system, due to its complexity, as well as economic and geopolitical importance, is an essential link for the development of the economies of many countries [13–15]. This is especially relevant for the post-industrial countries (including Poland), where electricity ensures the functioning of production processes, communication, and trade in goods [16–22].

Creating a concern in the energy system constitutes a big challenge. It can be concluded that this is an energy revolution. The direction of the change has been set, but without defining all the goals.

In the article, the authors analyze the process of establishing energy groups and the creation of a multi-energy concern based on oil, gas, petrochemicals, and energy, which is a response to signed agreements determined by political agreements. Building a new concern will allow for the better management and control of the processes which Poland has committed itself to implement. Its development is ensured by further consolidation within the energy sector. These actions are necessary because market forecasts also indicate that, after 2030, oil consumption will drop significantly in favor of alternative fuels. It is an opportunity to reduce the climate crisis.

Synergy in the economy is a phenomenon consisting of the integration of two or more entities and their activities in such a way that as a result, they would bring greater benefit than if these entities functioned separately. This phenomenon is quite well illustrated by the relation in which 2 + 2 > 4. Due to the use of synergy, it is possible to achieve effects greater than the sum of the results of each of the entities before the merger process. The most important positive results of synergy effects include savings, which are achieved by reducing the duplication of processes, organizational optimization, i.e., more effective use of available resources, as well as financial costs, such as the reduction of costs and tax burdens. An important synergy effect is also complexity, which enables more effective coordination of the activities of internal departments in the company. The synergy effect may be achieved in various areas of the business entity's operation as pure synergy and cost synergy. An example of the latter is the development of a common accounting system (especially management). It is a very important factor in persuading strategic investors to make mergers and acquisitions in 2018–2020 [23].

#### **2. Literature Review**

Mergers and acquisitions have been the subject of research by many authors [24–38]. Observations and research have unanimously proven that mergers generally bring positive effects.

Investors very carefully evaluate the market value of functioning economic entities; therefore, mergers and acquisitions may not only lead to capital concentration but also make it easier to obtain it [17]. It is worth noting that this phenomenon is currently the main cause of capital changes in the world [24] pp. 395–399.

In the context of the 2030 Agenda for Sustainable Development, adopting the EU

Renewable Energy Directive and the European Green Deal, the European Union aims for the extremely ambitious goal to become climate neutral by 2050 [7–12].

There are two basic methods of merging companies in the Commercial Companies Code. One of them is an acquisition, and the other is a merger as a method of creating a new company from two existing ones [19]. It should be remembered, however, that both of these methods are fundamentally different from each other, and their use results from specific economic conditions, as well as stakeholders' expectations as to the concept of the development of combined enterprises [31].

Based on accounting reports, many authors prove the positive impact of mergers on the condition of entities participating in these processes. Only the determination of the benefits allows for you to start the connection [39]. The value resulting from this procedure is defined as a significant increase in profitability and as an increase in the value of the combined enterprise [40].

The common denominator of mergers is undoubtedly the pursuit of capital concentration, due to the benefits resulting from increased efficiency in managing the available resources or the synergy effect. Apart from concentration, external development also involves integration (with suppliers or recipients), but also diversification (related and unrelated) [41] p. 19.

The main purpose of the article is to review the changes in the structures of capital groups in the Polish energy sector. The main research hypothesis was the following: the formation of a multi-energy concern is conducive to mergers and acquisitions.

The authors draw attention to the terminological disorder in theory and practice in the nomenclature of various, often different, processes of economic unit mergers. The article uses research methods such as analysis and criticism of the literature and the analysis of financial documents of the studied entities.

In the conducted empirical research, the following research questions will be verified, which at the same time constitute research problems:

Q1: Were there any mergers and acquisitions in the energy sector in the analyzed period, and what were their causes?

Q2: What is the main cause of capital concentration processes in the energy sector, and does it align with Agenda 2030?

Q3: How have the capital structures in the energy sector changed under the influence of these processes?

In addition, the following research hypotheses were put forward:

**Hypothesis 1 (H1).** *Mergers and acquisitions strengthen a multi-energy concern*.

**Hypothesis 2 (H2).** *Mergers and acquisitions in the energy sector are processes that increase capital concentration and share in the production and sale of electricity*.

**Hypothesis 3 (H3).** *Mergers and acquisitions transactions changed the shareholding structure in energy groups and thus increased the influence of the State Treasury during the period considered*.

The global coronavirus pandemic that prevails to this day has caused a disorder of the economic order, for example, by slowing down the economy, causing stagnation in money markets, or, finally, limiting the wealth of investment portfolios of many international companies, which results in a limited capital (resulting from greater investment risk and the need to accumulate capital in a period of uncertainty) economic) with the need to carry out mergers and acquisitions. This is especially visible in the catering, hotel, tourism, and automotive industries [1–3,23]. On the other hand, there is no similar tendency in the energy sector, which prompts the assessment of this phenomenon in this study.

The authors have attempted to evaluate mergers and acquisitions in recent years, paying attention to the acquisition of Energa by Orlen.

#### **3. Theoretical Approaches to Mergers and Acquisitions as a Manifestation of Capital Concentration**

The external development of the enterprise as an objective is very important in a dynamic and competitive economy. This development is manifested by business combinations. It is thanks to mergers that companies obtain synergy effects that allow them to implement many new ventures.

A merger is a basic form of merging two separate enterprises into one entity. The distinguishing feature of a merger is that, as a rule, it takes place as a result of the concerted and free action of partners [42].

A merger transaction takes place when the number of companies is reduced as a result of the merger. Such an undertaking can be carried out in two ways. Consolidation (also known as an acquisition sensu stricto) takes place when two entities merge and, in the end, a new economic entity is created. Such a procedure can be presented in the form of the equation [(Subject) A + B = Subject C] [13,14,16]. In general, one could say that "*a merger is defined as a voluntary merger of two business units into a new legal entity under similar conditions for both of the merging entities*" [43].

The second way is attachment (also known as a takeover sensu largo). This situation occurs when one enterprise is liquidated, losing its legal personality, and the assets are bought back. This can be written as [(Subject) A + B = Subject B]. According to the trade law, balance law, and numerous directives of the European Union, it can be called incarnation or incorporation [13,14,16].

The main difference comes down to the statement that two relatively economically equivalent entities are linked by consolidation, while by incorporation, a larger entity (economically, financially, organizationally) incorporates a smaller entity, although it is not an exclusive rule (there are cases reverse) [44–46].

Considering the definitions of an acquisition in a general sense, one can assume, following the dictionary of the Polish language, that it is "*taking what previously belonged to someone else*". It may mean that there is still the same number of things in the physical sense, only some invisible bond between them changes. Similar importance can be noted in economics and management. In management sciences, a merger is defined as a combination of two or more entities, as a result of which all but one of the merging entities lose their legal personality, under the name of which the entity resulting from the merger will function. The merger must be approved by the shareholders or stakeholders, and the merger process itself takes place by way of an exchange of stocks or shares [47].

A comparison of both unit combinations based on the Code of Commercial Companies in Poland is shown in Figure 1.

It can be concluded that the process of economic mergers always results from the merger of two or more companies. In addition, the transaction deprives the other entities of legal personality, as only one of the merger participants retains it. Most of the definitions suggest that the effect of the merger is the creation of a completely new economic entity. Taking into account the fact that the merger increases the resources of the new legal entity, as well as the modification of the organizational structure of the resulting entity, it should be assumed that it is a "new" entity that is the successor of the previously existing entity. Thus, the continued operation of an economic unit under the earlier name does not contradict the condition specified in the other definitions. Although, the functioning of a new entity under the name of an already existing entity is not appropriate, according to the authors. By acquisition, the authors understand the transfer of control from the acquired enterprise to the acquiring enterprise. It often occurs as a result of shareholders accepting an offer to buy their shares. Table 1 presents the most frequently mentioned buyer motives in connection processes.

During the analysis of the above classification, attention should be paid to the financial motives. An interesting example of the implementation of these premises is the separation of a tax group for tax purposes. Such a process is also a merger of several companies. An example is the tax capital group operating in Tauron Polska Energia. In 2011, Tauron completed the registration process of the Tax Capital Group (PGK), which included 11 companies from the capital group. On 14 December 2020, another tax agreement of the capital group for the years 2021–2023 was registered [48].

**Figure 1.** Sample characteristics by Source: author's study based on art. 693 Polish Code of Commercial Companies. **Figure 1.** Sample characteristics by Source: author's study based on art. 693 Polish Code of Commercial Companies.

**Table 1.** Motives for M&A **No. Group of Motives Benefits/Effects Table 1.** Motives for M&A.


Source: author's study based on: [7,10].

As already noted, the basic difference between a merger and an acquisition is that the former occurs as a result of mutual consent and cooperation between companies.

On the other hand, according to the dictionary of the Polish language, seizure means taking what previously belonged to someone else or detaining someone or something that is on the way illegally. Therefore, takeovers are often considered to be hostile, i.e., those in which the basis is the desire of one of the parties to buy out a controlling stake, without the prior determination of this fact, with the acquired party [13,14,16].

However, due to the closeness of the two terms, merger and acquisition are often used interchangeably, hence distinguishing them in the language of economists is often conventional. Examples of authors who use only one term—"merger"—are Eugene F. Brigham and Louis C. Gapenski in "Financial Management". They explain their choice to use only the term "merger" by the fact that they are solely interested in the process and focus on the fundamental economic and financial aspects of mergers, not the legal ones [49].

The initial years of the regime change in Poland were dominated by the privatization of state-owned enterprises when these entities were massively transformed into limited liability companies [50]. Currently, after thirty years of transformations, interesting changes in the concentration of capital in the mergers and acquisitions market in the energy sector can be observed.

#### **4. The Energy Sector in the Context of Mergers and Acquisitions in Poland**

To conduct observations and research, the financial statements of entities participating in mergers and acquisitions from 2010–2020, as well as information from databases, industry reports FORDATA and Navigator Capital and KPMG, the Warsaw Institute of Entrepreneurship, and available studies by the Ministry of State Assets, Internet articles, and press releases were obtained. The first part of this section focuses on capital groups in the energy sector.

The power sector is defined as a part of the fuel and energy sector related to the production, transmission, and distribution of, as well as trading in, electricity and heat. Electricity is the so-called "Base sector", and thus has a significant impact on the development and competitiveness of other areas of the economy in a particular industry [51] p. 53. This has its practical justification because the strong energy demand has been a phenomenon that has been noticeable for many years, not only in Europe but also all over the world. Data transmission via power lines is a pillar of the development of enterprises whose production systems, production automation, or teleinformation services are based on the existence of stable power networks. The power industry currently needs new sources of energy supply. This phenomenon gains additional significance at a time of drastically diminishing natural resources (conventional sources) and the need to search for new, but alternative (unconventional), energy sources.

The assessment was also influenced by the epidemiological situation related to the spread of the SARS-CoV-2 virus on a global scale [52–54]. This event influenced the development of many branches of economies, which, on the one hand, contributed to the improvement of the effectiveness of new distribution and sales channels, and, on the other hand, contributed to the paralysis of those industries that were not able to cope with the new reality. With all these issues in mind, for a more complete understanding of the problem, the authors assessed the situation in the mergers and acquisitions market in 2021 [55–57].

Changes in the energy system are closely related to ensuring energy security, understood as a guarantee of energy supplies and defined as the system's resistance to exceptional and unpredictable events that may threaten the physical integrity of the energy flow or lead to an unstoppable increase in its price, regardless of economic grounds. Therefore, it is part of the national security system because reliable and constant access to energy sources, at costs that can be borne by society, is an essential element of any modern economy. More generally, energy security is such a state of the economy that ensures that the current and future customer demand for fuels and energy is covered in a technically and economically

justified manner, with a minimal negative impact of the energy sector on the environment and living conditions of the society.

For many years, a strong multi-energy concern has been created in Poland to meet the challenges related to the energy transformation of Poland. Effective integration within the group through mergers and acquisitions is key to this process. Of course, apart from the basic factors stimulating the achievement of an appropriate level of energy security in Poland, the availability of infrastructure, transport, and, finally, the concentration of suppliers is also important.

The answers to the questions and the verification of the research hypotheses were made based on numerous tables and charts.

The list of mergers and acquisitions in Poland from 2010 to 2019 is presented in Table 2. As can be seen in Table 2, over the years 2010–2019, the number of mergers and acquisitions has decreased significantly. An important justification for the decreasing number of connections is investment caution resulting from the uncertainty of economic events, and thus difficulties in assessing the long-term prospects of the functioning of enterprises. This problem is more visible in the era of the coronavirus pandemic, which prompts companies to concentrate their accumulated capital rather than make costly business decisions, which undoubtedly include mergers and acquisitions.

**Table 2.** Merger and acquisitions in Poland from 2010–2019.


Source: Raporty M&A Index Polska za Lata 2009–2020, Raporty Branzowe. ˙ https://fordata.pl/category/ blog/raporty-branzowe, (accessed on 2 July 2021) [58]; M&A Index Poland. Fuzje i przej˛ecia w 2018 Roku. Raport Przygotowany Przez Firmy Navigator Capital oraz FORDATA Grudzie´n 2018 Roku. http://www. navigatorcapital.pl/wp-content/uploads/2018/12/Raport\_MnAIndexPoland\_rok2018\_PL.pdf (accessed on 22 November 2021) [58,59].

Continuing the analysis, it can be noted that the number of transactions does not always translate into a correspondingly higher or lower total value of mergers and acquisitions. This regularity can be seen taking into account the years 2010 and 2019. In 2010, with the number of transactions at 581, a total value of mergers and acquisitions was recorded at the level of EUR 21 billion, while in 2019, with only 258 transactions, the total value of transactions was recorded at the level of EUR 10.9 million. These figures illustrate a situation where more costly transactions were made despite a reduced number of transactions in general. In other words, these are, financially, more spectacular operations than single transactions carried out in other years.

From the observations carried out, it can be concluded that the transactions in 2010–2019 were characterized by a much higher value than those from the end of the second decade of the 21st century.

The most valuable connections in Poland from 2010 to 2019 are presented in Table 3.

The largest transaction (from the financial point of view) in the analyzed period was the merger process in 2011, as a result of which Polkomtel acquired Spartan Capital Groups. Its value was estimated at PLN 18.6 billion. In turn, the "cheapest" transaction was carried out in 2015. As a result, the American concern Scripps Network Interactive acquired shares in TVN.

The median of the set of all mergers and acquisitions in 2010–2019 is PLN 4.7 billion. On the other hand, the arithmetic mean is PLN 6.68 billion, and the fashion can be estimated (only with the assumption of rounding to the full value in PLN billion) at PLN 4 billion, with an average deviation of PLN 3.59 billion, and a standard deviation from the population of PLN 4.80 billion. The standard deviation from the sample is PLN 5.06 billion, the range of all values is PLN 16.2 billion, and the statistical error is PLN 1.60 billion. These data also

indicate that, in the analyzed years, the value of transactions exceeded PLN 10 billion only in two periods, i.e., in 2011 and 2016. The other transactions individually did not exceed the arithmetic mean of all transactions, amounting to PLN 6.68 billion. The above results are presented in Table 4.


**Table 3.** List of the largest mergers and acquisitions in Poland from 2010–2019, in Bln zł.

Source: Raporty M&A Index Polska za Lata 2009–2020, Raporty Branzowe. ˙ https://fordata.pl/category/blog/ raporty-branzowe/, (accessed on 2 July 2021) [58]; M&A Index Poland. Fuzje i przej˛ecia w 2018 Roku. Raport Przygotowany Przez Firmy Navigator Capital oraz FORDATA Grudzie´n 2018 Roku. http://www.navigatorcapital.pl/ wp-content/uploads/2018/12/Raport\_MnAIndexPoland\_rok2018\_PL.pdf (accessed on 22 November 2021) [58,59].

**Table 4.** Statistical indicators of measures of the largest mergers and acquisitions from 2010–2019.


Source: author's study based on Table 2.

It is certainly important that in the period under review the companies were made intra-group acquisitions. Among others, Polkomtel became part of Cyfrowy Polsat. The former was taken over by Spartan Holdings Capital, and, therefore, the entire venture is owned by Cyfrowy Polsat, which is at the forefront of the telecommunications services market in Poland.

In 2020, a total of 229 mergers and acquisitions were made, so there is still a continuing downward trend in terms of their number. The largest transaction was worth PLN 9.6 billion and concerned the takeover of Play Communications by Iliad S.A. (Fordata, Pozna ´n, Poland, 2020) [59].

A similar phenomenon can be observed in the energy sector. It is one of the most important areas for the development of the economy of many countries. The concern of ensuring an appropriate level of energy security has become the reason for many mergers and acquisitions. The Polish energy economy is based primarily on fossil fuels. According to 2011 data from the Ministry of Economy, the share of coal in the primary energy demand structure is the highest and amounts to 56%. It is followed by crude oil (25%), natural gas (13%), and biomass and waste (6%), respectively.

Figure 2 shows the structure of primary energy demand by source.

market in Poland.

(Fordata, Poznań, Poland, 2020)[59].

Figure 2 shows the structure of primary energy demand by source.

natural gas (13%), and biomass and waste (6%), respectively.

**Figure 2.** The structure of primary energy demand is classified. Source: author's study based on report "Raport sektorowy—energetyka", by Jan Sekuła posted 31 October 2021, https://profitjournal.pl/raport-sektorowy-energetyka/ access date: 2 February 2022 [60]; bankier.pl [37] (accessed on 23 November 2021). **Figure 2.** The structure of primary energy demand is classified. Source: author's study based on report "Raport sektorowy—energetyka", by Jan Sekuła posted 31 October 2021, https://profit-journal. pl/raport-sektorowy-energetyka/ access date: 2 February 2022 [60]; bankier.pl [37] (accessed on 23 November 2021).

It is certainly important that in the period under review the companies were made intra-group acquisitions. Among others, Polkomtel became part of Cyfrowy Polsat. The former was taken over by Spartan Holdings Capital, and, therefore, the entire venture is owned by Cyfrowy Polsat, which is at the forefront of the telecommunications services

In 2020, a total of 229 mergers and acquisitions were made, so there is still a continuing downward trend in terms of their number. The largest transaction was worth PLN 9.6 billion and concerned the takeover of Play Communications by Iliad S.A.

A similar phenomenon can be observed in the energy sector. It is one of the most important areas for the development of the economy of many countries. The concern of ensuring an appropriate level of energy security has become the reason for many mergers and acquisitions. The Polish energy economy is based primarily on fossil fuels. According to 2011 data from the Ministry of Economy, the share of coal in the primary energy demand structure is the highest and amounts to 56%. It is followed by crude oil (25%),

This state of affairs is largely due to the geographical location of Poland. It has large stores of fossil fuels, mainly hard coal and lignite. According to data from Agencja Rynku Energii S.A. in August 2020 [61], the total share of hard coal and lignite in the structure of consumption of basic fuels in the commercial power industry in the period of January– August 2020 (a decrease by about 2 percentage points compared to the corresponding period in 2019) amounted to nearly 87.8%, while the share of gaseous fuels was only 6.87% (an increase by 1 percentage point compared to the corresponding period in 2019). Admittedly, the share of biogas/biomass also increased from 3.92% in 2019 to 5.10% in This state of affairs is largely due to the geographical location of Poland. It has large stores of fossil fuels, mainly hard coal and lignite. According to data from Agencja Rynku Energii S.A. in August 2020 [61], the total share of hard coal and lignite in the structure of consumption of basic fuels in the commercial power industry in the period of January– August 2020 (a decrease by about 2 percentage points compared to the corresponding period in 2019) amounted to nearly 87.8%, while the share of gaseous fuels was only 6.87% (an increase by 1 percentage point compared to the corresponding period in 2019). Admittedly, the share of biogas/biomass also increased from 3.92% in 2019 to 5.10% in 2020.

2020. The analysis of the above values leads to a thesis that, currently, Poland is not a sufficiently independent recipient of energy sources because the production of electricity The analysis of the above values leads to a thesis that, currently, Poland is not a sufficiently independent recipient of energy sources because the production of electricity is still largely based on coal (hard and lignite).

is still largely based on coal (hard and lignite). According to the analysis of the National Energy and Climate Plans (NECP), Poland is among the 7 EU countries that are the most delayed in the implementation of the coal transition program [62,63]—and by 2030 it is to be the most coal-dependent country in the community. From a long-term perspective, this approach requires a fundamental change [64,65]. This is dictated not only by the growing awareness of the need to protect the According to the analysis of the National Energy and Climate Plans (NECP), Poland is among the 7 EU countries that are the most delayed in the implementation of the coal transition program [62,63]—and by 2030 it is to be the most coal-dependent country in the community. From a long-term perspective, this approach requires a fundamental change [64,65]. This is dictated not only by the growing awareness of the need to protect the natural environment (in the context of limited non-renewable energy resources), but also by the current legislative trends in the European Union. Poland's independence from fossil fuels will not only positively contribute to the improvement of the quality of the natural environment (including polluted air), but should also allow for the building of an appropriate level of energy security.

Renewable energy sources are at the forefront of solutions aimed at reducing the unfavorable share of fossil fuels in the structure of electricity demand in Poland. The gradual growing share of renewable sources in this structure should be considered a positive effect of the current legal and social trends, the final result of which is the need to become independent from non-renewable sources. For comparison, at the end of August 2021, the capacity of installed renewable energy sources increased, compared to the same period from 2020, by 33.9%, amounting to 14.9 GW. Therefore, taking into account the total capacity of all installed electricity sources in Poland (approximately 53.4 GW), the share of renewable sources is nearly 28%. The largest share of the available renewable sources in Poland is wind farms at around 6790 MW and photovoltaics at 5970.8 MW. Table 5 presents a detailed summary of the given types of electricity sources, with an installed capacity in MW, in August 2021 [66].


**Table 5.** Types of energy sources in Poland as of 30 August 2021.

Source: The types of energy sources in Poland as of 30 August 2021. Source: author's study based on report "Raport sektorowy—energetyka", by Jan Sekuła posted 31 October 2021, https://profit-journal.pl/raport-sektorowyenergetyka/, access date: 2 February 2022 [60]; bankier.pl (accessed on 23 November 2021).

Aside from the above considerations, it is worth noting that another solution proposed by the experts is the transition to nuclear energy, which will accelerate the process of decarbonizing Poland's energy structure and thus contribute to the improvement of the country's energy security [67–69].

Although many different entities in Poland have a license to trade in electricity, the entire energy market is divided into 4 entities, which end suppliers choose depending on their geographic location. When characterizing this sector, it should be mentioned that under the government program, four capital groups (the so-called big energy four) were created: PGE, Tauron, Enea, and Energa, which generate about 70% of the country's energy. In addition, Innogy Polska, which manages the capital's electricity network, runs its business in Warsaw.

The goal of the government program, which is to consolidate the Polish energy market, is implemented by linking economically strong enterprises. As a result of this consolidation, these entities will be able to incur investment costs, compete, and balance the energy market.

In the first stage, the Transmission System Operator was separated from the structures of PSE (PSE Operator) and Distribution System Operators (including energy companies, e.g., ZKE Dystrybucja, RWE Stoen Operator, etc.) as independent legal entities [70]. The Transmission System Operator, along with the transmission network, has been handed over to the State Treasury. In turn, PSE was merged with the Polish Energy Group, which is one of the four energy groups operating on the market.

Polska Grupa Energetyczna (PGE) was established on the basis of the BOT Górnictwo i Energetyka SA holding, Zespół Elektrowni Dolna Odra SA, with the assets remaining after the separation of the Transmission System Operator from PSE SA, along with its assets and 8 distribution companies from central and eastern Poland without RWE Stoen.

An important entity forming the energy sector is Tauron Polska Energia SA. It is the second-largest enterprise in Poland after Polska Grupa Energetyczna. It is a jointstock company listed on the Warsaw Stock Exchange since 2010. The Tauron Group is one of the largest economic entities in Poland and is one of the largest energy holdings in Central and Eastern Europe. It operates in all areas of the energy market, from coal mining, to the generation, distribution, and sale of electricity and heat, as well as customer service. On a smaller scale, the holding also conducts the wholesale of fuels and derivative products (trading in coal and biomass). In 2014, Tauron entered the gas fuel trading market. Currently, it is active in the field of renewable energy sources.

Another energy group is Enea. According to the government program, this group included companies under privatization or consolidation. Enea S.A. is a parent company that operates in the full value chain of the energy market: hard coal mining, electricity generation, electricity distribution, and energy trading. The Group also operates in the area of heat generation, distribution, and sale, as well as in the field of lighting services. It is the vice-leader in electricity production in Poland. In 2018, it generated 26.5 TWh. It was created from the merger of five power plants (Pozna´n, Bydgoszcz, Szczecin, Zielona Góra, and Gorzów Wielkopolski), as well as from subsequent acquisitions (including the Kozienice Power Plant and Połaniec Power Plant).

Energa SA is a group that produces and sells thermal and electric energy as well as gas. Currently, after the acquisition by PKN ORLEN S.A., which holds 80.01% of its share capital, the main activities of the group include the generation, distribution, and trade in electricity and heat. The Gda´nsk entity has the largest share of RES in energy production among all Polish energy groups (over 40%). The production capacities include a system power plant in Ostroł˛eka, two combined heat and power plants, forty-seven hydropower plants, a pumped-storage power plant, five wind farms, and two photovoltaic farms. The Group's distribution network covers an area of approximately 1/4 of the country [71].

The creation of a multi-energy concern favors mergers and acquisitions. There are also other companies in Poland operating within the energy sector, the market values of which are shown in Table 6.

It is worth mentioning that the ownership structure of other holdings is different from the above-mentioned four Polish energy groups. The owner of the CEZ group, whose ˇ capitalization is the highest and exceeds PLN 40 billion, is the Czech Republic, which holds 69.78% of the property rights [72]. The volume share of the produced energy for each company is presented in Figure 3. *Energies* **2022**, *15*, 1669 12 of 23

**Figure 3.** Share of capital groups in the volume of electricity fed into the grid in 2021. Source: author's study based on: https://profit-journal.pl/raport-sektorowy-energetyka/ (access date: 15 August 2021) [60]; bankier.pl (accessed on 23 November 2021). **Figure 3.** Share of capital groups in the volume of electricity fed into the grid in 2021. Source: author's study based on: https://profit-journal.pl/raport-sektorowy-energetyka/ (access date: 15 August 2021) [60]; bankier.pl (accessed on 23 November 2021).


27% of the State Treasury's share in the shareholding structure.


The exception among large Polish companies is Energa in Gdańsk, for which the majority shareholder is PKN Orlen. It is worth noting that PKN Orlen itself holds over Source: author's study based on: https://profit-journal.pl/raport-sektorowy-energetyka/ (access date: 15 August 2021) [60]; bankier.pl (accessed on 23 November 2021).

Gdansk-based company. The purchase process was completed in just 4 months. PKN ORLEN announced a tender offer for 100% of Energa Group shares on 5 December 2019. Due to, among other things, the situation caused by the COVID-19 outbreak, the time for accepting subscriptions for shares in the Pomeranian group has been extended until 22 April 2020. The concern received approval from the European Commission for the transaction on 31 March 2020 [74]. The transaction was finalized on 30 April 2020, with the acquisition of Energa shares representing approximately 80% of the company's share capital and approximately 85% of the total number of votes at the general meeting. The price of all shares acquired amounted to approximately PLN 2.77 billion and was covered by the company with cash from its resources and an available syndicated loan [75]. The financial results of Energa before the merger with PKN ORLEN are presented in Figure 4.

**5. Multi-Energy Concern—A Case Study from Poland (PKN Orlen and Energa)**

Polska Grupa Energetyczna (PGE S.A.), as confirmed by the chart above, supplies over 40% of the energy for the entire country. This means that the share of this group in the structure of diversification of electricity supplies in Poland is significant, and may, in the long term, contribute to a significant concentration of energy capital in the hands of the State Treasury. Due to their strategic nature for the development of the economy, the largest Polish energy companies have a majority share of the State Treasury in their shareholding structure. The state's share in PGE is 57.39%, is 54.49% in Tauron Polska Energia, and is 51.5% in Enea [73].

The exception among large Polish companies is Energa in Gda´nsk, for which the majority shareholder is PKN Orlen. It is worth noting that PKN Orlen itself holds over 27% of the State Treasury's share in the shareholding structure.

#### **5. Multi-Energy Concern—A Case Study from Poland (PKN Orlen and Energa)**

The largest acquisition of the Polish fuel and energy market is the purchase of 80% of shares in Energa Group by PKN ORLEN. This means that the concern owns the Gdanskbased company. The purchase process was completed in just 4 months. PKN ORLEN announced a tender offer for 100% of Energa Group shares on 5 December 2019. Due to, among other things, the situation caused by the COVID-19 outbreak, the time for accepting subscriptions for shares in the Pomeranian group has been extended until 22 April 2020. The concern received approval from the European Commission for the transaction on 31 March 2020 [74]. The transaction was finalized on 30 April 2020, with the acquisition of Energa shares representing approximately 80% of the company's share capital and approximately 85% of the total number of votes at the general meeting. The price of all shares acquired amounted to approximately PLN 2.77 billion and was covered by the company with cash from its resources and an available syndicated loan [75]. The financial results of Energa before the merger with PKN ORLEN are presented in Figure 4. *Energies* **2022**, *15*, 1669 13 of 23

**Figure 4.** Financial results of the Energa group from 2010–2018. Source: author's study based on https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August 2021). **Figure 4.** Financial results of the Energa group from 2010–2018. Source: author's study based on https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August 2021).

As can be seen from the chart, even though the EBITDA [76] value of the Gdanskbased Group did not change in recent years before the merger transaction, the market value of the Group dropped significantly. In 2015, the market value of Energa is almost PLN 11.5 billion. The following years brought a significant reduction in the market capitalization of the acquired company. First and foremost, misguided investment activities contributed to this valuation, which resulted in a liquidity crunch for the As can be seen from the chart, even though the EBITDA [76] value of the Gdansk-based Group did not change in recent years before the merger transaction, the market value of the Group dropped significantly. In 2015, the market value of Energa is almost PLN 11.5 billion. The following years brought a significant reduction in the market capitalization of the acquired company. First and foremost, misguided investment activities contributed to this valuation, which resulted in a liquidity crunch for the acquired company.

acquired company. When analyzing Energa Group's financial results for 2019, it is worth noting the significant reduction in operating profit, which amounted to PLN 459 million compared to PLN 1.18 billion in the previous year. EBITDA was PLN 2.04 billion compared to PLN When analyzing Energa Group's financial results for 2019, it is worth noting the significant reduction in operating profit, which amounted to PLN 459 million compared to PLN 1.18 billion in the previous year. EBITDA was PLN 2.04 billion compared to PLN 1.88 billion a year earlier. The Energa Group suffered a net loss of PLN 1 billion in 2019,

1.88 billion a year earlier. The Energa Group suffered a net loss of PLN 1 billion in 2019, compared to PLN 744 million profit for 2018, and the reason for the company's declining market capitalization is visible. The group's parent company's net loss for 2019 is PLN 952

million [77]. A decrease in the market value of Energa before the merger with PKN

Negative financial results of the Energa group were significantly influenced by the impairment test conducted by Ostrołęka power plant and recognition of impairment loss for PLN 1.03 billion, as well as an impairment loss on PGG shares and recognition of a provision for future tax liabilities. It was the investment related to the construction of a new coal unit in the Ostrołęka Power Plant that so significantly reduced the value of the acquired group [78]. That is why PKN Orlen offered a maximum of PLN 2.9 billion for the

The next charts show the energy market share of Polish energy companies after the merger of PKN ORLEN and Energa Group and their impact on sales to end-users. The analysis of the graphs clearly shows that the transaction has strengthened the energy

From this analysis, we can see the increasing share of merged entities in the energy

buyout. Orlen's current capitalization is PLN 37.5 billion.

concern and changed its structure, as presented in Figure 6.

market. This supports research hypothesis no. 2.

ORLEN is shown in Figure 5.

0

18

35

**PRODUCTION (TWH/YEAR)** 

0

3

6

**COMPANY VALUE AND SHARE PRICE**

53

70

7

compared to PLN 744 million profit for 2018, and the reason for the company's declining market capitalization is visible. The group's parent company's net loss for 2019 is PLN 952 million. It is worth mentioning that the company ended 2018 with a profit of PLN 739 million [77]. A decrease in the market value of Energa before the merger with PKN ORLEN is shown in Figure 5. *Energies* **2022**, *15*, 1669 14 of 23

**Figure 5.** Market value of the Energa concern before the merger by PKN ORLEN. Source: author's **Figure 5.** Market value of the Energa concern before the merger by PKN ORLEN. Source: author's study based on https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August 2021). 11.5 6.9 6.5 7 8

study based on https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August

2021). 66 Negative financial results of the Energa group were significantly influenced by the impairment test conducted by Ostroł˛eka power plant and recognition of impairment loss for PLN 1.03 billion, as well as an impairment loss on PGG shares and recognition of a provision for future tax liabilities. It was the investment related to the construction of a new coal unit in the Ostroł˛eka Power Plant that so significantly reduced the value of the acquired group [78]. That is why PKN Orlen offered a maximum of PLN 2.9 billion for the buyout. Orlen's current capitalization is PLN 37.5 billion. 2.9 2.7 2.9 0 Pre-debut valuation December 2013) The highest valuation (April 2015) Valuation in the last 6 months Valuation in the last 3 months Max valuation by Orlen (December 2019)

> 27 16 4 8 PGE Enea Tauron Energa + Orlen The next charts show the energy market share of Polish energy companies after the merger of PKN ORLEN and Energa Group and their impact on sales to end-users. The analysis of the graphs clearly shows that the transaction has strengthened the energy concern and changed its structure, as presented in Figure 6. **Figure 5.** Market value of the Energa concern before the merger by PKN ORLEN. Source: author's study based on https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August 2021).

energy sources, including, among others, wind farms and hydroelectric power plants. On the other hand, by using PKN Orlen's production surpluses, it is possible to reduce operating costs, and the combined customer base will generate additional sales potential. Figure 7 clearly shows the increased energy market share (sales to end users) of the two **Figure 6.** Electricity market after Energa acquisition by Orlen–production [TWh/Year]. Source: author's study based on: https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August 2021). **Figure 6.** Electricity market after Energa acquisition by Orlen–production [TWh/Year]. Source: author's study based on: https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August 2021).

The market objective of the concentration in the multi-energy sector is to gain an

create an appropriate level of resistance of the company to market fluctuations. Finally, it should be noted that Energa Group owns a wide range of assets for acquiring renewable energy sources, including, among others, wind farms and hydroelectric power plants. On the other hand, by using PKN Orlen's production surpluses, it is possible to reduce operating costs, and the combined customer base will generate additional sales potential. Figure 7 clearly shows the increased energy market share (sales to end users) of the two

combined entities. This supports research hypothesis no. 2.

combined entities. This supports research hypothesis no. 2.

From this analysis, we can see the increasing share of merged entities in the energy market. This supports research hypothesis no. 2.

The market objective of the concentration in the multi-energy sector is to gain an appropriate competitive advantage through the diversification of revenue sources, which not only allows for the building of additional value for customers, but primarily serves to create an appropriate level of resistance of the company to market fluctuations. Finally, it should be noted that Energa Group owns a wide range of assets for acquiring renewable energy sources, including, among others, wind farms and hydroelectric power plants. On the other hand, by using PKN Orlen's production surpluses, it is possible to reduce operating costs, and the combined customer base will generate additional sales potential. Figure 7 clearly shows the increased energy market share (sales to end users) of the two combined entities. This supports research hypothesis no. 2. *Energies* **2022**, *15*, 1669 15 of 23

**Figure 7.** Electricity market after Energa acquisition by Orlen-end-user sales [TWh/Year]. Source: author's study based on: https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August 2021). **Figure 7.** Electricity market after Energa acquisition by Orlen-end-user sales [TWh/Year]. Source: author's study based on: https://wysokienapiecie.pl/24792-orlen-chce-przejac-energe/ (access date: 15 August 2021).

The multi-energy concern that is being created will likely more easily meet the challenges of Poland's energy transformation. The effective concentration of capital within the Group is critical to this process. Therefore, PKN ORLEN initially wanted to purchase 100% of Energa group shares. Such capital control always means tangible benefits both organizationally and operationally. Although control has not been fully acquired, the participation of the remaining shareholders does not jeopardize the strategic The multi-energy concern that is being created will likely more easily meet the challenges of Poland's energy transformation. The effective concentration of capital within the Group is critical to this process. Therefore, PKN ORLEN initially wanted to purchase 100% of Energa group shares. Such capital control always means tangible benefits both organizationally and operationally. Although control has not been fully acquired, the participation of the remaining shareholders does not jeopardize the strategic objectives set by the acquirer [79].

objectives set by the acquirer [79]. According to the authors, this merger is another important stage in the creation of a strong, multi-energy concern that will strengthen the competitive and financial position of the participating entities, as well as the country's energy security and, consequently, the Polish economy (Hypothesis 1). These conclusions are also confirmed by other According to the authors, this merger is another important stage in the creation of a strong, multi-energy concern that will strengthen the competitive and financial position of the participating entities, as well as the country's energy security and, consequently, the Polish economy (Hypothesis 1). These conclusions are also confirmed by other analysts (Polish Multi-energy concern, March 2021).

analysts (Polish Multi-energy concern, March 2021). PKN ORLEN has announced a tender offer for a minority stake in Energa S.A. This will further facilitate the effective integration of the assets of ORLEN and Energa Groups. It will also bring tangible benefits to ORLEN Group shareholders and investors, who will be able to fully benefit from the synergies of the merger. It is worth remembering that the purchase alone does not complete the effective integration of the two entities, which sometimes takes several years, especially since the transaction itself was completed in a PKN ORLEN has announced a tender offer for a minority stake in Energa S.A. This will further facilitate the effective integration of the assets of ORLEN and Energa Groups. It will also bring tangible benefits to ORLEN Group shareholders and investors, who will be able to fully benefit from the synergies of the merger. It is worth remembering that the purchase alone does not complete the effective integration of the two entities, which sometimes takes several years, especially since the transaction itself was completed in a very short time.

very short time. The company plans to develop the areas in which Orlen and Energa are already active, such as renewable energy and electromobility, but also to enter new projects, such as offshore wind farms. The completion of this transaction should result in a more efficient use of the companies' combined potential. The group owns more than 50 assets that produce energy from renewable sources, primarily including hydroelectric power plants, onshore wind farms, and photovoltaic farms. More than 30% of the electricity produced by the Energa Group comes from renewable sources and this is the highest share among The company plans to develop the areas in which Orlen and Energa are already active, such as renewable energy and electromobility, but also to enter new projects, such as offshore wind farms. The completion of this transaction should result in a more efficient use of the companies' combined potential. The group owns more than 50 assets that produce energy from renewable sources, primarily including hydroelectric power plants, onshore wind farms, and photovoltaic farms. More than 30% of the electricity produced by the Energa Group comes from renewable sources and this is the highest share among its main competitors. For PKN ORLEN, it is an interesting RES portfolio that balances the company's

its main competitors. For PKN ORLEN, it is an interesting RES portfolio that balances the company's conventional assets, such as the steam-gas units in Płock and Włocławek. Additionally, the diversification of revenue sources increases the company's resilience to

The authors believe that the merger of these entities will not only benefit both companies, but the entire Polish economy. The beneficiaries of the capital takeover of the Gdansk group are also Pomerania and its inhabitants. Thanks to the planned investments, the number and scale of orders will increase. From a fiscal point of view, Energa will remain fully separate, which means more income for the local government budget and influence over the development of the Pomeranian area. Shares of companies in the WIG-

ENERGIA index after the merger are illustrated in Figure 8.

conventional assets, such as the steam-gas units in Płock and Włocławek. Additionally, the diversification of revenue sources increases the company's resilience to market fluctuations and changes in the macroeconomic environment. In this way, additional value is built for customers and shareholders.

The authors believe that the merger of these entities will not only benefit both companies, but the entire Polish economy. The beneficiaries of the capital takeover of the Gdansk group are also Pomerania and its inhabitants. Thanks to the planned investments, the number and scale of orders will increase. From a fiscal point of view, Energa will remain fully separate, which means more income for the local government budget and influence over the development of the Pomeranian area. Shares of companies in the WIG-ENERGIA index after the merger are illustrated in Figure 8. *Energies* **2022**, *15*, 1669 16 of 23

**Figure 8.** Shares of companies in the WIG-ENERGIA index as of 30 October 2020. Source: author's study based on: https://profit-journal.pl/raport-sektorowy-energetyka/ access date: 25 August 2021 [60]; bankier.pl (accessed on 23 November 2021). **Figure 8.** Shares of companies in the WIG-ENERGIA index as of 30 October 2020. Source: author'sstudy based on: https://profit-journal.pl/raport-sektorowy-energetyka/ access date: 25 August 2021 [60]; bankier.pl (accessed on 23 November 2021).

According to Figure 8, the largest share in the companies' structure in the WIG-Energia index on 30 October 2020 was in PGE S.A. and amounted to 40.93%. Tauron is in second place with a result of approximately 20.88%. The next two largest shares include Enea at 11.29% and Energa at 7.74%. These data only confirm the thesis that among many entities in the energy market, only the 4 largest entities forming the concern play a key role in ensuring an appropriate level of energy security. This supports research hypothesis According to Figure 8, the largest share in the companies' structure in the WIG-Energia index on 30 October 2020 was in PGE S.A. and amounted to 40.93%. Tauron is in second place with a result of approximately 20.88%. The next two largest shares include Enea at11.29% and Energa at 7.74%. These data only confirm the thesis that among many entities in the energy market, only the 4 largest entities forming the concern play a key role in ensuring an appropriate level of energy security. This supports research hypothesis no. 3.

#### no. 3. **6. Top Motives for Acquisitions and Mergers by Investors in the Energy Sector**

**6. Top Motives for Acquisitions and Mergers by Investors in the Energy Sector**  The motives behind mergers and acquisitions are varied. Many authors point out that investment motives vary depending on the type of entity that conducts the transactions. Strategic investors who acquire companies want to integrate the entity with the rest of the group and have a real impact on its management. Financial investors, in turn, involve The motives behind mergers and acquisitions are varied. Many authors point out that investment motives vary depending on the type of entity that conducts the transactions. Strategic investors who acquire companies want to integrate the entity with the rest of the group and have a real impact on its management. Financial investors, in turn, involve capital in an entity, expecting economic benefits from its future sale.

capital in an entity, expecting economic benefits from its future sale. In a survey conducted by KPMG, both strategic and financial investors took part. In response to investment motives, strategic investors most often indicated the willingness to expand geographically and take over the portfolio of products and services (79% of In a survey conducted by KPMG, both strategic and financial investors took part. In response to investment motives, strategic investors most often indicated the willingness to expand geographically and take over the portfolio of products and services (79% of responses) as the main or the most frequently considered reason for the investment.

responses) as the main or the most frequently considered reason for the investment.

Figure 9 illustrates the main M and A motives listed by investors.

**Factor—Category** 

available)

Acquisition of significant intangible assets (licenses, brand, R&D, etc.)

Figure 9 illustrates the main M and A motives listed by investors.

*Energies* **2022**, *15*, 1669 17 of 23

**Figure 9.** Main investment motives in mergers and acquisitions. Source: author's study based on report "Charakterystyka procesów transakcyjnych w Polsce. Oczekiwania a rzeczywistość", KPMG, Warszawa 2021, p. 13. **Figure 9.** Main investment motives in mergers and acquisitions. Source: author's study based on report "Charakterystyka procesów transakcyjnych w Polsce. Oczekiwania a rzeczywisto´s´c", KPMG, Warszawa 2021, p. 13.

Cost synergies of industry consolidation and acquisition of customers and distribution channels are also important to them. Almost 80% of strategic investors (Figure 9), when making an equity investment, do not consider the subsequent resale of the object of the transaction to make a profit on such a transaction. However, the vast majority of financial investors (82%) say that the main or usually considered factor when making a transaction is the desire to profit from the future resale of the entity. The vast majority of financial investors (82%) say that the primary or usually considered factor in making a transaction is the desire to profit from the future resale of the entity. In addition, more than two-thirds are making acquisitions to consolidate the industry. Table 7 presents the distance between investment motives in M and A. Cost synergies of industry consolidation and acquisition of customers and distribution channels are also important to them. Almost 80% of strategic investors (Figure 9), when making an equity investment, do not consider the subsequent resale of the object of the transaction to make a profit on such a transaction. However, the vast majority of financial investors (82%) say that the main or usually considered factor when making a transaction is the desire to profit from the future resale of the entity. The vast majority of financial investors (82%) say that the primary or usually considered factor in making a transaction is the desire to profit from the future resale of the entity. In addition, more than two-thirds are making acquisitions to consolidate the industry. Table 7 presents the distance between investment motives in M and A.

**Table 7.** Distance between investment motives in mergers and acquisitions undertaken by financial and strategic investors**. The Importance of the Category for the Financial Investor The Importance of the Category for the Strategic Investor The Importance of Category for both Groups of Investors (Arithmetic Mean)**  Access to resources 0.3 0.3 0.3 Assumed profit on future resale 0.8 0.3 0.55 Opportunistic approach (applies when a potential acquisition target suddenly becomes 0.8 0.3 0.55 As can be seen by analyzing Table 7, both strategic and financial investors are driven by similar motives in M and A transactions. However, because of their expectations, their importance is different. A strategic investor looking to integrate an entity expects to realize the benefits of capital and personnel ties. When a financial investor involves capital in an entity, they expect quick economic benefits from its future sale. In many aspects, a certain degree of agreement between the two investor groups can be observed, e.g., on the issue of access to raw materials, acquisition of significant intangible assets, business diversification, acquisition of customer portfolio or distribution channels, and industry consolidation. However, the biggest differences between investors are in the issues of expected return on investment or the success of the acquisition target. A very important motive for both groups of investors (having an arithmetic mean of 0.7) is the consolidation of the industry, which is confirmed by the process of strengthening the multi-energy concern (Hypothesis 1).

0.3 0.3 0.3

Business diversification 0.6 0.6 0.6


**Table 7.** Distance between investment motives in mergers and acquisitions undertaken by financial and strategic investors.

Source: author's study based on Figure 9.

#### **7. Discussion**

Analyzing the program of creating a multi-energy concern in Poland, one can confirm that its transformational goals are ambitious. PGE has announced that it will achieve climate neutrality by 2050 by acquiring 20 gigawatts (GW) of power from renewable energy sources. Currently, the company's total installed capacity is 17 GW and it is received mainly from coal assets. The three directions of development for the next decade—offshore, onshore, and photovoltaic—are expected to contribute to achieving this goal. Particularly interesting news for the energy industry was coming out in the second quarter of 2021. Namely, the Ministry of State Assets plans to consolidate the incumbent on energy companies mining assets and create a new entity. We are talking about the planned spin-off of coal assets from the Treasury's energy companies.

The separation of selected assets also has a practical justification. Other companies that have suffered as a result of the financial crisis caused by the COVID-19 pandemic are also using such strategies. They take defensive and offensive tools that optimize the internal investment portfolio.

Among the most important defensive activities of M and A operations are the possibility of selling some part of ineffective assets. Another solution is the ability to look for strategic partnerships and alliances with other market players as part of mergers and acquisitions [80].

This strategy is in line with trends in the global energy industry, where asset stripping can be seen due to shrinking synergies between areas of the value chain. Sellers conclude that disposing of less desirable assets is more profitable than the gains from owning them.

An example of such an activity is the sale of the REN power grid operator as part of Portugal's Stabilization and Growth Program. There is a closing of investment cycles by financial investors. Thus, Private Equity investors who made investments in the electricity market in previous years are choosing to realize profits and sell selected assets from their portfolios. The sale of AES Red Oak LLC, an electricity generator, by U.S. fund PE Energy Capital Partners LLC is one of many such deals [81].

The proposed actions of the Ministry of State Assets in Poland will improve the profitability of these companies and allow for new sources of financing for RES investments. They will be able to obtain bank loans more easily.

The news about removing mines from power companies will probably be well valued by the market, which can be seen from the charts of all the power companies, especially Tauron, whose subsidiary Tauron Wydobycie generated almost PLN 84 million in losses on PLN 236 million revenue in Q1 2021 alone. According to PGE's CEO, the optimal date for spinning off coal assets from current energy groups is the end of 2021. According to the authors, the fact of this announcement should positively influence the improvement of the financial results of these companies [82].

It is worth emphasizing that the accelerated monopolization of the energy sector by state-owned corporations will probably result in a significant increase in energy prices in the future, with simultaneous attempts to interfere in the system of rates and charges, e.g., by offering compensations or the possibility of postponing the increase in gas prices for households. This result is compounded simultaneously by several important factors. First, in 2021, a power fee was introduced, charged to final energy consumers for financing new or the modernization of existing units by power companies. Second, the effects resulting from the prevailing global coronavirus pandemic have been severe on the economy as a whole—demand for energy and gas has increased after the lockdown period. Additionally, there is a shift developing in Europe from traditional fuels to renewable sources. Finally, the effect of Russia's unpredictable gas policy, which seeks to maintain its influence in EU states, if only by exerting pressure on the start-up of the Nord Stream 2 pipeline [83].

#### **8. Conclusions**

Continuing the discussion, it is worth mentioning the next steps in strengthening the multi-energy concern. Currently, there are plans to merge three state-owned holding companies [84]. On 12 May 2021, a cooperation agreement was signed between PKN Orlen and Lotos Group, PGNiG, and the State Treasury concerning the acquisition of control by the Płock concern over its competitor from Gda´nsk and PGNiG. As a result of this transaction, all assets of the acquired companies will be transferred to PKN Orlen. The State Treasury and PKN Orlen will work together to develop analyses aimed at the final confirmation of this scenario as optimal from the point of view of the State Treasury [85].

The agreement executed between the parties of the transaction postulates that, as a result of the merger, shareholders of the Lotos Group and PGNiG will receive new shares issued by Orlen in exchange for their shares. This will occur when it assumes the rights and obligations of the Lotos Group and PGNiG, respectively, under universal succession. The State Treasury will increase its shareholding in the entity created as a result of the acquisition of Lotos and PGNiG by PKN Orlen to 50%, which will protect it against hostile takeovers. The acquisition by Orlen will be non-cash and the State Treasury's stake in PKN Orlen will be increased. They are currently below 30% [79]. These are further facts that will support the accepted theses in this paper.

In July 2021, the Ministry of State Assets announced the planned merger of three energy companies—PGE, Tauron, and Enea—in the coming years. The strategy related to the divestment of coal mines from the assets of energy companies is very important due to the inevitable transition towards green energy and the resulting preferential conditions for raising capital from banks and EU programs [86]. This type of financing is essential for the implementation of the long-term energy transition process, which aims for a 55% reduction in Poland's emissions at an estimated cost of €136 billion [68]. This fact fully confirms that the strengthened multi-energy concern can contribute to the implementation of activities carried out in light of the 2030 Agenda. Thus, it can be considered that the scale of M and A transactions in the sector, which take place in the energy sector are due to signed agreements and are the result of actions carried out in light of the 2030 Agenda [7–12].

Summarizing the considerations, it can be concluded that the transferred transactions of capital concentration significantly increase the participation of the Treasury in energy companies and contribute to ensuring national energy security. These processes strengthen the multi-energy concern and reduce the number of capital groups that are operators of energy services in the country. A strengthened concern arguably has more investment potential. In turn, entering an integrated holding company is a major development opportunity for any entity. The creation of multi-energy concerns is in line with megatrends and actions implemented by other, international oil companies. An example of such a

state-controlled multi-sector conglomerate is Italy's ENI. Today it is the largest company in Italy and one of the world's leading integrated oil and gas producers. Its Polish counterpart would be the planned concern combining ORLEN, LOTOS, and ENERGA [87].

The merger of PKN Orlen with PGNiG and Grupa Lotos is a complex project, but, because of it, an entity will be created that will match the potential of the euro with its competitors. As assessed by the participants of the 6th edition of Congress 590, after consolidation, it will have a much greater investment potential and will be able to increase its involvement in low- and zero-emission energy investments [88].

PKN Orlen contributes to the strengthening of the development of capital concentration, which will not burden the natural environment. However, supporting the development of renewable energy sources in the energy sector or alternative fuels in transport requires large investment outlays. This is possible only thanks to building a strong multienergy concern. Thus, a multi-energy concern has a chance to make a green transformation.

The new entity is a response to the revolutionary challenges faced by the energy market today [89].

**Author Contributions:** Conceptualization, J.T.-M. and M.W.-J.; Data curation, J.T.-M.; Formal analysis, M.W.-J. and J.T.-M.; Funding acquisition, J.T.-M. and M.W.-J.; Methodology, M.W.-J. and J.T.-M.; Project administration, J.T.-M. and M.W.-J.; Resources, J.T.-M. and M.W.-J.; Software, M.W.-J.; Supervision, J.T.-M. and M.W.-J.; Validation, J.T.-M. and M.W.-J.; Visualization, M.W.-J.; Writing—original draft, J.T.-M. and M.W.-J.; Writing—review and editing, J.T.-M. and M.W.-J. All authors have read and agreed to the published version of the manuscript.

**Funding:** The publication was financed from the subsidy granted to Cracow University of Economics.

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data is contained within the article.

**Acknowledgments:** The authors thank the anonymous Reviewers and the Editors for their valuable contributions that significantly improved this manuscript.

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

#### **References**


#### *Article* **The Business Model in Energy Sector Reporting—A Case Study from Poland: A Pilot Study**

**Beata Sadowska <sup>1</sup> , Magdalena Wójcik-Jurkiewicz 2,\*, Grzegorz Zimon 3,\*, Adam Lulek <sup>1</sup> , Nina St˛epnicka <sup>4</sup> and Robert Walasek <sup>5</sup>**


**Abstract:** The business model is understood as a way of generating revenue and creating value. This article aimed to define the framework and detailed characteristics of the business model in corporate reporting in the energy sector. The study also addressed the issue of value creation and examined the correlation between reporting information about the business model and the value of the enterprise, calculated using selected accounting measures. The research was conducted in 2021 on all listed companies in the energy sector included in the WIG20 index in 2019–2020. The research methodologies included critical analysis of the literature, analysis of the content of corporate publications, comparative analysis, analysis of phenomena and synthesis of results, Spearman's rank correlations, and graphical data presentation. The originality of our research concerns directing attention to the disclosure and reporting of information about the business model in corporate reports in the energy sector in Poland. This is the first such study in Poland conducted on companies in the energy sector; therefore, it is treated by the authors as a pilot study. The results show that companies in the energy sector included in the WIG20 index do not disclose information about their business models in reports and the structure of published reports is very diverse, which makes it even more difficult to compare financial and non-financial data. Enterprises in the energy sector disclose financial and non-financial information in financial statements and the reports of the management board on the activities of the company and the capital group. The financial information disclosed by the companies most often included financial and productive capital. Energy companies are reluctant to disclose information about by-products and waste.

**Keywords:** business model; integrated reporting; non-financial reporting; energy sector; Warsaw Stock Exchange (WSE)

#### **1. Introduction**

In the past decade, there has been a huge amount of research related to the business model in terms of both management and accounting [1,2]. Especially in the last twenty years, the business model has become a popular topic in research and practice [1–3]. Many articles discuss the theoretical and practical aspects of creating a company's business product in the context of increasing competition around the world [4–6]. Various aspects have been analyzed, for example, the impact of the company's openness to stakeholders on the effectiveness of the company's business model [7], the business model in agricultural enterprises [8], the relationship between knowledge orientation and business analytics

**Citation:** Sadowska, B.;

Wójcik-Jurkiewicz, M.; Zimon, G.; Lulek, A.; St˛epnicka, N.; Walasek, R. The Business Model in Energy Sector Reporting—A Case Study from Poland: A Pilot Study. *Energies* **2023**, *16*, 1955. https://doi.org/ 10.3390/en16041955

Academic Editor: Krushna Mahapatra

Received: 8 January 2023 Revised: 9 February 2023 Accepted: 13 February 2023 Published: 16 February 2023

**Copyright:** © 2023 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/).

capabilities in driving business [9], integrated management of multiple business models, and the strategic management of business model innovation [10].

This fact means that research on the business model has not been established based on economic research or management. The ambiguity of placing the business model in scientific research leads to a differentiated assessment of its usefulness and value creation architecture [11]. The business model is one of the most commonly used terms in strategic management and, more recently, in accounting [12–14]. Although the term was initially used in the context of e-commerce, its application is currently comprehensive and includes research in many sectors, such as energy, manufacturing, healthcare, and biotechnology [15–19]. The use of the business model concept has increased as researchers studying strategic management and the development of modern business forms, such as business groups, have begun to take an interest [20–23]. Integrated reporting is often combined with the concept of integrated thinking [24] in the context of creating a sustainable business model [25–29]. This is especially important for companies that have a significant impact on the environment, such as transport companies or energy operators. Each individual company, depending on its proximity and distance to impact on the environment, formulates its strategy, vision and mission, accounting policy, and business model. The energy sector is not isolated in the implementation of these activities. Entrepreneurs conducting business activity in the field of logistics services represent the group with the fastest reaction to changes taking place in the market, thus defining their own business model. In the case of enterprises in the energy sector, the business models are a combination of strategic orientation and technology, which is an essential element of achieving business goals and generating positive results from operations [30,31]. However, there is a noticeable gap in the research concerning integrated reporting on applied business models in this sector. A few studies have explored the value creation process of the low-cost carrier business model in the aerospace industry [32–34]. Another case study refers to research on the development of the sustainability habitus in the Arab Middle East (ME) using the example of a leader in the energy sector. In these studies, a comparative analysis of integrated reporting and environmental reporting was performed [35]. Ciubotariu et al. [36] suggested that based on sensitivity function value, sectors of activity such as transport, infrastructure, services, and trade have a higher sensitivity preference for IR (integrated reporting) due to quantification of the stakeholders' interest in terms of performance-positive trend indexes.

In our research, we adopted the theoretical framework of the business model in terms of corporate accounting and financial reporting and related it to enterprises in the energy sector in Poland [37–40]. On 9 December 2013, the International Integrated Reporting Council (IIRC) released the first internationally recognized IR framework. IR is understood as "a process founded on integrated thinking that results in a periodic integrated report by an organization about value creation over time and related communications regarding aspects of value creation" [41]. In the literature, several theoretical papers explain why disclosing the business model could be useful in allowing investors to predict future earnings and cash flow [42–46]. Moreover, the business model is one of the most important elements in the strategy of communication between enterprises and current and potential investors. There are many approaches to the category of "business model," among which is a link between economic theory and accounting valuation and thus with financial and non-financial reporting [18]. From a business model perspective, it is acknowledged that any asset can be utilized in different ways and that value creation, another concept key to IR, will ultimately vary between firms. The challenge at hand is the alignment of the financial perspective with the value proposition of the enterprise [14,32].

This paper contributes to the accounting literature as it provides evidence that the voluntary disclosure of macro components encounters many limitations related to the communication of the activities of the enterprise, especially in the case of environmental impact [47]. This was confirmed by research conducted in Russia on enterprises in the transport sector, which proved that there was a lack of transparency in communicating the activities and business models of these enterprises [48–51]. Our results are of interest not only to researchers but also to practitioners and standard setters. In particular, they highlight the importance of disclosing information concerning business models related to this sector in Poland in order to improve the usefulness of accounting [52–56].

Our research aimed to define the framework and detailed characteristics of the business model in corporate reporting in the energy sector. Our research was conducted on a sample of listed companies in the energy sector included in the WIG20 index for the years 2019–2020. We used disclosure maps proposed by CIMA and IIRC [32], which allowed us to conduct a detailed analysis of disclosures regarding the business models used.

The business model describes how ways of creating and delivering value can evolve as customer needs and preferences change [57]. It is a structure that provides information about the enterprise's prospects. It can be assumed that the business model describes what the company offers to its customers, how it reaches them with the offer, how it maintains relations with them, what resources it engages, what actions it takes for this purpose, with whom it enters into a partnership and, consequently, how it achieves revenues [58–60].

Progressive changes, including the deregulation of markets, development of technology, provision of services, flow of information and communication, and coordination of markets and economies make the business model increasingly necessary for enterprises to achieve market success and a sustainable competitive advantage. It is a determinant of creating the value of an individual enterprise [61]. The value category can be considered from a broad perspective, both financial and non-financial. The key in the process of creating a business model by enterprises in the energy sector appears to be understanding the essence of this concept and the correct selection of components. It should be noted that it is difficult and rather impossible to clearly define the business model [62,63] as the definitions are numerous and include diverse elements.

This paper proceeds as follows. The "Related Literature" section reviews the literature on the topic investigated and describes our research hypotheses. The "Materials and Methods" section provides details about our sample selection strategy and research design. The "Results" section presents our research results, while the "Discussion" and "Conclusions" sections contain a discussion of the implications, limitations, and possible future developments of the research.

#### **2. Related Literature**

The topic of the business model has attracted the attention of researchers since the 1990s, which is primarily related to the development of new information and communication technologies and the phenomenon of globalization [61–66]. The essence of the business model is to explain how organizations create, deliver, and capture value [67–71]. The business model is perceived as a key factor for creating and maintaining a competitive advantage as well as achieving corporate financial results. Business models evolve as managers both introduce new ideas and adjust in response to external factors [1–3]. As research shows, it can also communicate the business logic of the enterprise to a wide group, such as shareholders [45]. This unique ability of the business model to communicate value has resulted in the fact that various accounting organizations [20,51–54] currently perceive it in terms of increasing the transparency and clarity of financial disclosures. Despite its undeniable advantages, the concept of the business model also has several important gaps, both cognitive and technical [55]. The cognitive gap refers to the lack of agreement on the unambiguous understanding of the term business model, which is the reason for the frequent development of a specific understanding of this term dependent on the academic center formulating it. In turn, the technical gap is generally expressed by the lack of guidance in the presentation of information within the business model. At the same time, the business model considered from the point of view of the economic theory of the company provides practical difficulties in the area of financial reporting measurement [72,73]. Sustainable business models strive to balance economic prosperity, social well-being, and environmental preservation in their operations and strategies [73].

In our article, based on the concept of a business model in corporate reporting, we have formulated the following research hypotheses:

**Hypothesis 1 (H1).** *Do companies in the energy sector use the business model concept in their reports?*

**Hypothesis 2 (H2).** *How are business model disclosures identified in corporate reports in the energy sector?*

**Hypothesis 3 (H3).** *Is the structure of the reports published by companies in the energy sector homogeneous or diversified, and do these reports allow for easy comparison of financial and nonfinancial data?*

**Hypothesis 4 (H4).** *Do companies in the energy sector have specificities compared to other sectors?*

**Hypothesis 5 (H5).** *Are there strong correlations between the disclosure of a business model's information and the entity's value?*

In the IIRC framework structure, the term "business model" is defined as "a selected system composed of inputs, business activities, outputs, and outcomes, designed to create a specific value in the short, medium and long term" [50,68]. Reporting and disclosure in corporate reporting should answer the following question: What is the business model of a given organization and to what extent is it sustainable? The components, categories, and detailed characteristics of the business models are presented in Table 1.

**Table 1.** Components, categories, and detailed characteristics of the business models.


Source: Ref. [18], pp. 189–202.

The benefits of integrated reporting apply to many groups interested in creating corporate value, not only financial capital providers. It is proven that integrated reporting combines financial and non-financial reporting processes, paving the way for the transformation of the internal reporting system [74–78]. The process of creating corporate value depends to a large extent on the adopted business model understood as a "system of transforming inputs, through its business activities, into outputs and outcomes" [62]. Enterprises in the energy sector conduct business activities based on diverse business models [79]. The business model is a method of doing business that ensures further operation for the entity. It enables communication with stakeholders, which contributes to reflecting the true value of the company. The lack of complete information (e.g., non-financial) in an integrated report on the business model may lead to a distortion (underestimation) of the value of the economic entity. Enterprise value is also created through communication about the business model.

The business model is a method adopted by the company to increase and use resources, which is presented to customers with an offer of products and services; it is a conceptual tool containing elements and relationships between them. The business model describes how ways of creating and delivering value can evolve as customer needs and preferences change. An unambiguous definition of the business model is difficult and rather impossible because there are many concepts defining the business model, thus it is understood differently by economic units. The business model is often reported in an integrated report.

The business model should inform about the use of the resources by the enterprise in business activities, whereas the key elements of the business model include segments of stakeholders, business partners, customers, value propositions, customer communication channels, stakeholder relations, revenue streams, key resources, key success factors, key partnerships, and cost structure. The business model is built based on these assets shaping the so-called value dynamics. The business model is multidimensional; it is a unique combination of tangible and intangible assets aimed at maximizing the value of the organization.

A key issue, especially for enterprises in the energy sector, is non-financial reporting related to areas such as corporate social responsibility and environmental impact. This issue concerns the transparency of non-financial reporting [80–84]. As evidenced by previous research in Poland, the lack of compulsory environmental disclosure has discouraged managers from voluntarily disclosing such information [82–86]. However, the current discourse on business models highlights an increase in levels of disclosure of both financial and non-financial information [87–90]. In particular, relating non-financial reporting to value in the long term is problematic, taking into account intangible values that are not measured or communicated [90,91].

Communicating information about the business model and non-financial capital should be the next stage in the evolution of the company's reporting and this type of disclosure should be carried out in a specific reporting format, called the integrated report [91,92]. However, there is currently a turbulent scientific discourse about the essence of understanding the business model, in terms of referring it to management paradigms, embedding it in the criterion of environmental impact, as well as cause-and-effect relationships and correlations between other ontological entities. The ambiguity of the definitions and the multiplicity of elements forming the business model structure stem, on the one hand, from the variability of the realities of running a business and their perception by decision-makers, and, on the other hand, from the evolution of concepts and theories explaining these processes and phenomena.

#### **3. Materials and Methods**

In our research, we used financial statements and reports of companies from the energy sector listed on the Warsaw Stock Exchange (WSE). There are seven listed companies on the WSE. Table 2 presents the surveyed companies and additionally indicates the industries of each of the companies in the energy sector and the detailed scope of their activities. The leading activity of the studied companies was transport. The companies conducted diversified business activities related to the domestic market. Table 1 also indicates the business model disclosures analyzed in the reports.

**Table 2.** List of companies and reports in which financial and non-financial information was disclosed.


Source: Study based on our own research.

The survey was conducted in 2021 based on financial and non-financial data available for 2019–2020. The layout and content of the study were subordinated to the implementation of the assumed purpose of the study. The authors used a pilot testing method and combined it with other methods, such as experiments and observational studies, to provide a more comprehensive understanding of the research questions. As a research method, we empirically analyzed the material, which allowed us to verify the research hypotheses posed. The methodology process involved several steps. The first step included a selection and review of the literature. This included identifying texts that were relevant to the research question or topic being studied. The authors also organized articles and identify patterns and connections between them. The next step was to combine the information from the texts to create a comprehensive understanding of the topic or research question. Finally, conclusions and interpretations were drawn.

This research approach is widely accepted in studies exploring capital information disclosures in business models [26]. It should be emphasized that the small research sample was conditioned by the small number of enterprises that currently publish comprehensive information about their business model. The business model disclosures were drawn from various reports as there is no single unified concept of business model reporting. We did not have comparable and extensive research on business model disclosures by the industry. Sometimes, the disclosures were fragmented, incomplete, and impossible to compare.

The study also addressed the issue of value creation and examined the correlation between reporting information about the business model and the value of the enterprise, which was calculated using selected accounting measures. As part of the current study, the interdependencies of the detailed components of the business model were examined i.e., the correlation relationships that occurred between the sum of disclosures about the business model in selected components for the surveyed enterprises and the value of the enterprise, measured using selected accounting methods, achieved by these entities. These

included total assets, financial results, and equity. Statistical analysis of the correlations was carried out based on the Spearman correlation method, i.e., coefficients were calculated for the ranks of variables. The argument for using this statistical method was the small size of the group; as this was a preliminary, pilot study that will be continued on a large group of entities this year. As an additional argument for choosing this statistical method, we also considered the fact that the correlation coefficients calculated for the ranks of variables showed any monotonic relationship, and the distribution of the research sample was not linear.

Among the surveyed companies, there were no significant differences in the structures of the reports in which the information about the business models was disclosed. Two of the seven companies in the energy sector listed on the Warsaw Stock Exchange had not published current financial statements or non-financial reports for the years 2019–2020. One company, COM4, drew up an integrated report, while another one, COM3, reported the non-financial information of the capital group, and three companies disclosed information about the results of their activities in the consolidated financial statements and in the management board's report on the activities of the company and the capital group (Table 3).

**Table 3.** The number of disclosures on components of the business model of companies in the energy sector.


Source: Study based on our own research.

The study of the number of disclosures indicated that logistics companies most often report information about the input elements (capital), including mainly financial elements and business outcomes, both financial and non-financial. Slightly less often, enterprises disclose information about business activities (most commonly about planning, design, and risk). Some components of the business model were not disclosed in single cases—no data were reported in the period under review. A characteristic situation for the entire research sample was that the information highlighted in the components of the business model was at a similar level in 2019 and 2020. This was reflected in the score of the individual components (Table 4).

In the next stage of the study, detailed elements of the components of the business model and information regarding financial (numerical) disclosures were subjected to indepth content analysis for all surveyed logistics companies (Table 4). It should be noted that the empirical data confirmed the dominant verbal nature of disclosures about the business model. There was a much greater number of financial (numerical) information disclosures in the surveyed logistics companies than those of a descriptive nature. The

largest percentage of financial disclosures among the total information was recorded for the business activities component in 2019 (i.e., 74%), whereas the smallest percentage of financial disclosures was recorded for the business activities component in 2020 (i.e., 27%).

**Table 4.** The number of disclosures on components of the business model of companies in the energy sector—only financial information.


Source: Study based on our own research.

#### **4. Results**

In our research, we first collected the disclosures reported by logistics companies under four main components of the business model. For this purpose, the disclosure map proposed by CIMA and IIRC was used, according to which the business model consists of 4 components:

	- 4. Outcomes, including customer satisfaction, profit/loss, return to shareholders, asset consumption, employment creation, employee development and engagement,

improvement of living standards, environmental impact, licenses/certificates, and contribution to the local economy.

Within the indicated components of the business model, 22 types of disclosures were listed, which formed the basis for further analysis. The subject of the study included their descriptive form and disclosures expressed in a monetary measure. The analyses were carried out based on a study of the content of financial statements and reports available on the website of the Warsaw Stock Exchange.

The content analysis was primarily qualitative. Content analysis is defined as a research technique that allows obtaining reproducible and correct conclusions from texts in the context of their use. Its capabilities also concern the use of the quality index as a basis for searching for evidence that the item is or is not disclosed. Detailed results of the analyses are presented in Tables 3–6. For each of the selected parameters describing the business model, the following score was awarded: 1—when such information was disclosed by companies, 0—when there were no disclosures. As a result, the maximum number of points that a given company could be awarded was 22 for one year. In total, it was therefore possible for a company to obtain a maximum of 44 points for both analyzed years. Table 5 presents the results of analyses concerning the input elements, i.e., the capital of the business models of enterprises in the energy sector for the years 2019–2020.

**Table 5.** Input elements—the capital of the business models of enterprises in the energy sector for the years 2019–2020.


Source: Study based on our own research.

**Table 6.** Business activities.


Source: Study based on our own research.

When analyzing capital reporting among the seven companies surveyed, two did not include such information at all, as they did not publish any reports. One of the companies only disclosed information about their financial capital, whereas two companies disclosed information about all of their capital, but only for 2019. They had not yet published reports for 2020. One of the companies disclosed information about financial and productive capital and one enterprise disclosed information about all capital except data on natural and social capital.

In the area of reporting on business activities, companies made various disclosures, as shown in Table 6.

Two companies reported data for both years studied, while the other two had not yet published reports for 2020. Planning and production activities in a descriptive dimension were disclosed by two companies in both years and by one company in 2019. Risk was described by three companies over a period of two years and by two companies in 2019. Information about training, corporate research and development, and relationship management were included in the reports of only two companies in the energy sector and only in 2019. Diversification of services was reported by three companies, two of which disclosed information for 2019.

The output elements of the energy sector enterprises in the business models were also subject to significant variation, as shown in Table 7.


**Table 7.** Output elements of enterprises in the energy sector for 2019–2020.

Source: Study based on our own research.

It is clear that companies in the energy sector are reluctant to disclose information about by-products, both in financial and descriptive terms. Disclosure of information about waste was at a similar level. Only one company disclosed such information. It can be presumed that in the case of typical service activities, such as logistics activities, these two categories were of negligible importance in the activities of these enterprises. However, the detailed scope of the activities carried out by some companies indicated that production activities or other service activities may have generated waste and by-products in some cases. This requires additional extended research.

Table 8 presents the output outcomes—the capital of the business models of enterprises in the energy sector for the years 2019–2020.

**Table 8.** Outcomes—the capital of the business models of enterprises in the energy sector for the years 2019–2020.


Source: Study based on our own research.

Of the five companies that published their reports, all of them disclosed information about the profit they achieved in 2019. There were no data reported for the two companies in 2020. Asset consumption was reported by three companies during the period under review, and the result of corporate research and development was only reported for one company. Information about employee development and employment creation was published by one energy sector enterprise in both years surveyed, while two companies disclosed these data for 2019. With regard to the information about the impact of the company's activities on the natural environment and the level of customer satisfaction, such information was reported by only two companies in 2019.

It is recommended that the current status of the surveyed WIG20 index companies be reviewed, and integrated reporting should be implemented to disclose, in addition to financial information, non-financial information about corporate social responsibility. The results of the current study confirmed that the surveyed companies included in the WIG 20 index do not disclose information about their business model in annual statements/reports, and disclosures about the business model have a significant role in the value of the enterprise. This was confirmed in research conducted by Kannenberg and Schreck [46] and Jensen and Berg [47].

It is recommended that further research be conducted in this area and new hypotheses are put forward. For example, it should be explored whether larger enterprises, more developed entities with a stronger market position in which the size is scaled with the sum of their assets, use new tools for financial and non-financial reporting on the business model. It is worth continuing research into how small and medium-sized enterprises react and whether they see the need for change. Further research is necessary due to the development of integration of enterprise management systems with accounting systems. This provides new challenges for accountants and controllers and will determine the development of the accounting profession as it evolves into a creative profession.

#### **5. Discussion**

In conclusion, it should be noted that in the case of companies in the energy sector, there is no positive approach to reporting on the business model, including non-financial reporting. There is also a significant reduction in the narrative, which could have a negative connotation. Companies disclose mainly financial information in consolidated financial statements and corporate information in the management board's report on activities. In these documents, non-financial information about the business model is limited, practically non-existent, a fact confirmed by our research (Hypothesis 1). Companies should use business models to explain how they generate revenue, create value for their customers, and plan to sustain their competitive advantage. The business model concept is also useful for energy companies to communicate their strategies to stakeholders, such as investors and regulators [47].

As part of the current research, interrelationships between the detailed components of the business model were also examined for the years 2019–2020—i.e., correlations between the sum of disclosures about selected components of the enterprise's business model and the value of the enterprise, measured using selected accounting methods. These include total assets, financial results, and equity. Statistical analysis of the correlations was carried out based on the Spearman correlation method, i.e., coefficients were calculated for the ranks of the variables. The argument for using this statistical method was the small size of the group. The results of the dependence are presented in Tables 9 and 10.

Table 9 shows Spearman's rank correlations for the sum of disclosures of the specified components of the business models of the surveyed enterprises with the selected accounting methods of measuring value.


**Table 9.** Spearman's rank correlations for all disclosures.

Source: Own study.

**Table 10.** Spearman's rank correlations calculated for financial information.


(In brackets the *p*-value for the Spearman's coefficient significance test was given). Source: Own study.

The overall impact of the considered components on the enterprise's value against the background of the parameters published by the analyzed companies describing their financial situation indicated a statistically significant (with *p*-value < 0.05) relationship between the sum of assets and results and between the financial result and inputs and overall values of the components. At the same time, it should be noted that these relationships were only significant in 2019. In 2020, in each case, a decrease was observed in the strength of the relationship between the total value of the components and the financial situation of the company described by the parameters, such as total assets, financial result, or equity capital. In the case of total assets, there was also a change in the direction of the relationship from positive in 2019 to negative in 2020, while there was a substantially reverse change in the relationship from negative to positive in the case of the financial result (Hypothesis 2). Table 10 presents the correlations between pure financial disclosures. It should be noted that companies should include business model disclosures in the annual report, strategic plan, or sustainability report [93,94] There should be a separate section about the business model in reports or investor presentations. In this section, companies should explain how they generate revenue, create value for their customers, and plan to sustain their competitive advantage.

When examining the relationship between individual financial components and the financial parameters discussed, statistically significant relationships were only observed in 2019. These dependencies concerned the relationship between all the results and the sum of assets as well as the values of input elements and the financial result achieved by the examined companies. As in the case of the total components, there was a change in the direction of dependence in 2020 compared to 2019 for the relationship between individual components and the sum of assets and the financial result.

To sum up this stage of the research, our findings indicate that disclosures about the business model do not represent a significant value for the energy sector (Hypothesis 5). In disclosures about the business model, this sector does not perceive categories that create value for the enterprise. Studies have shown that companies that provide more detailed and transparent information about their business model in their financial reports tend to be viewed as more valuable by investors and other stakeholders (Hypothesis 5) [46,47,92]. Further research in this area should be carried out, taking into account the outbreak of the COVID-19 pandemic and looking for determinants that contribute to this state of affairs. Generally, the structures of the annual reports and sustainability reports of companies in the energy sector can be quite diverse, making it difficult to compare financial and non-financial data across different companies. Some companies may provide detailed information about their environmental and social performance in their sustainability reports, while others may not (Hypothesis 3).

#### **6. Conclusions**

The business model has been identified by the International Integrated Reporting Council (IIRC) as a basic element of integrated reporting, and integrated reporting should focus on the business model of a given enterprise, including its resources, relationships, and financial and non-financial aspects. It was also found that investors promote the position that high-quality reporting describing an enterprise's business model is essential in better understanding its performance, the impact of the external environment on the entity, and how entities create and "maintain" value. Our research allowed us to determine the framework, place, and detailed characteristics of the business model in corporate reporting in the energy sector.

Enterprises in the energy sector disclose financial and non-financial information in financial statements and reports of the management board on the activities of the company and the capital group. Only one company published an integrated report that included information about the business model (partial and only for 2019). Because these entities did not draw up integrated reports, they did not directly use the business model concept or indicate the characteristic components of the business model in their reports. Thus, the main objective of the study was achieved, i.e., we defined the framework, place, and detailed characteristics of the business model in corporate reporting in the energy sector. Our research recommends that the current state of companies in the energy sector in the WIG20 index be verified in order to implement integrated reports. Therefore, in addition to financial information, non-financial information regarding corporate social responsibility should be disclosed by these companies. This finding coincides with research conducted by ref. [95] recommending that sustainable business models should adopt the circular economy concept and report on biodiversity and extinction accounting in a more structured and mandatory way via producing integrated reports.

Companies in the energy sector most often disclose information about financial and productive capital. They rarely disclose information regarding the category of business activity, indicating mainly the risk related to the business they conduct. Companies in the energy sector are reluctant to disclose information about by-products and waste, both in financial and descriptive terms. In terms of results, the surveyed companies in the energy sector practically do not disclose information about the effects of research and development, employee development and employment creation, environmental impact, and customer satisfaction. Financial statements and the management board's report on the company's activities are obligatory documents legitimized by the Accounting Act. Therefore, it should be noted that the structure of the reports published by companies in the energy sector should be homogeneous to allow for easy comparison of financial data (Hypothesis 3). Companies in the energy sector have a diverse set of business models, for the generation and distribution of electricity, natural gas, oil exploration and production, and renewable energy projects, among others (Hypothesis 4) [93,94].

Our research shows that companies in the energy sector included in the WIG20 index do not disclose information about their business model in their annual reports. There is little descriptive content regarding the business model in the analyzed documents. This information is very important for making investment decisions by stock investors. The listed companies, in particular, should publish qualitative descriptive information in their reports. Emphasis on publishing relevant information about the business model would help to strengthen the knowledge and decision-making process of report recipients.

It should be emphasized that this was a pilot study, and the originality of this research is in focusing attention on the disclosure and reporting of information about the business model in corporate reports of the energy sector in Poland. This is the first such study in Poland conducted on companies in the energy sector; therefore it is treated by the authors as a pilot study. The research showed that companies in the energy sector included in the WIG20 index do not disclose information about the business model in their reports and the structure of the published reports is very diverse, making it even more difficult to compare financial and non-financial data. Therefore, it should be clearly emphasized that the results of the study, unfortunately, cannot be generalized. The main limitations of the study were the size of the research sample, despite focusing on the largest "list players" belonging to the WIG20 index in Poland, and the availability and quality of the financial and non-financial information about the business models presented in the reports of the energy companies and their indicated dispersion. The theoretical implication of the study includes the influence of the business model on the strategic and financial performance of companies. For example, a company that utilizes a decentralized business model may have a greater ability to respond quickly to changes in the market and adapt to new technologies, leading to improved performance. In practice, such companies may, for example:


Additionally, the choice of business model can also impact the regulatory environment in which energy companies operate, as different models may be subject to different rules and regulations. In terms of reporting, the business model can affect the level of transparency and disclosure of information to stakeholders, such as investors and regulators.

**Author Contributions:** Conceptualization, B.S., M.W.-J., G.Z., A.L., R.W. and N.S.; Data curation, B.S., M.W.-J., G.Z., N.S. and A.L.; Formal analysis, B.S., M.W.-J., G.Z., A.L., R.W. and N.S.; Funding acquisition, B.S., M.W.-J., R.W., G.Z., A.L. and N.S.; Methodology, B.S., M.W.-J., G.Z., N.S., R.W. and A.L.; Project administration, B.S., M.W.-J., G.Z., A.L., R.W. and N.S.; Resources, B.S., M.W.-J., G.Z., A.L., R.W. and N.S.; Software, B.S., M.W.-J., G.Z. and A.L.; Supervision, B.S., M.W.-J., G.Z. and A.L.; Validation, B.S., G.Z., M.W.-J., A.L., R.W. and N.S.; Visualization, B.S., M.W.-J. and G.Z.; Writing—original draft, B.S., M.W.-J., G.Z., A.L., N.S. and R.W.; Writing—review and editing, B.S., M.W.-J., G.Z., A.L., R.W. and N.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This project was co-financed within the framework of the program of the Ministry of Science and Higher Education under the "Regional Initiative of Excellence" in 2019–2022 (project number 018/RID/2018/19, amount of funding PLN 10 788 423,16) and the Research Excellence Program (contract no. 013/ZZR/2022/DOS).

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data are contained within the article.

**Acknowledgments:** The work was carried out as part of the statutory activity of the University of Szczecin, Cracow University of Economics, University of Lodz, Rzeszow University of Technology, and Jan Kochanowski University in Kielce.

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

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

#### *Article* **Regional Specialization, Competitive Pressure, and Cooperation: The Cocktail for Innovation**

**Bartolomé Marco-Lajara, Eduardo Sánchez-García \*, Javier Martínez-Falcó and Esther Poveda-Pareja**

Department of Business Organization, University of Alicante, 03690 San Vicente del Raspeig, Spain; bartolome.marco@ua.es (B.M.-L.); javier.falco@ua.es (J.M.-F.); esther.poveda@ua.es (E.P.-P.) **\*** Correspondence: eduardo.sanchez@ua.es; Tel.: +34-96-590-3400 (ext. 3153)

**Abstract:** The main aim of this paper is to analyze the effect of industrial agglomeration on the degree of interorganizational cooperation and the innovative performance of firms of the electricity supply sector in Spain. For this purpose, the agglomeration coefficient in each of the 50 provinces of Spain is calculated, based on secondary data from SABI database. Subsequently, primary data are obtained from a sample of 197 companies through a structured questionnaire. In this case, the PLS-SEM technique is used. The results show that there is a positive and significant relationship between the variables analyzed. It is concluded that industrial agglomeration and cooperation are relevant external factors that boost the innovative performance of firms and that business associations foster interorganizational cooperation.

**Keywords:** agglomeration; cluster; specialization; cooperation; innovation; PLS-SEM; energy

**JEL Classification:** R12; O36; R30; L94

**Citation:** Marco-Lajara, B.; Sánchez-García, E.; Martínez-Falcó, J.; Poveda-Pareja, E. Regional Specialization, Competitive Pressure, and Cooperation: The Cocktail for Innovation. *Energies* **2022**, *15*, 5346.

Academic Editors: Bernard Zi˛ebicki and Edyta Bieli ´nska-Dusza

https://doi.org/10.3390/en15155346

Received: 22 June 2022 Accepted: 20 July 2022 Published: 23 July 2022

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

**Copyright:** © 2022 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/).

#### **1. Introduction**

Innovation is a fundamental factor for economic development and favors the generation and exploitation of opportunities by companies and boosts their performance [1–3]. Its importance has increased in recent years in academia, as a fundamental factor for the survival, and competitiveness of firms [4–6]. The development of innovations requires the interaction and cooperation of multiple actors in terms of information, knowledge, and other resources in favor of particular social or economic goals [7,8]. Cooperation allows organizations to analyze the market situation, generate new ideas, and expand their knowledge base on an ongoing basis [9]. In this way, the establishment of reciprocal cooperative relationships favors the absorption of knowledge by the participating companies and its application to the development of innovations [10,11]. Thus, the exchange of knowledge with agents in the environment is essential to boost the innovative performance of firms [12,13].

Industrial agglomeration favors communication and the establishment of cooperative– competitive relationships between companies in the sector (whether direct competitors or not) that are located geographically close to each other, which reduces the costs and risks of the process and has positive effects on communication, the generation of trust, access to and creation of new knowledge, and innovative performance [14,15]. In this sense, Porter [16] highlights a driver of innovation—in addition to the intensity of local competition— as being the existence of greater opportunities for cooperation. In this regard, there are certain agents, such as business associations, that facilitate the implementation of joint innovation processes acting as intermediary entities [17,18]. Belonging to an association of entrepreneurs promotes interaction and cooperation among its members, and favors the establishment of relationships of trust, acting as a tool of cohesion and intermediation [17], which favors the acquisition and assimilation of new knowledge [19].

In this research paper we study the influence of the agglomeration of the sector on the innovative performance of firms. In addition, we analyze the mediating effect of the degree of cooperation in this relationship and the membership of an association of entrepreneurs as an antecedent variable of the degree of the cooperation of firms with entities in their environment. Specifically, the aim of this paper is to determine the effect of industrial agglomeration on the innovative performance of firms in the sector under study, and the influence of cooperation as a mediating variable in this relationship.

The analysis is carried out in two phases. In the first one, data of the electricity supply sector were extracted from the SABI database, from which the number of companies and employment in each of the 50 provinces that make up the Spanish territory were calculated. Based on this data, the degree of agglomeration of the sector in each province was calculated in relation to the national average. Subsequently, in the second phase, a theoretical model was developed based on the hypotheses put forward. A structured questionnaire was designed, composed of validated scales for the calculation of the variables included in the proposed model, and was distributed among the companies of the sector being studied, obtaining a total of 197 valid responses, and the use of the PLS-SEM method was chosen to estimate the proposed relationships. Specifically, the responses were coded and analyzed using the SmartPLS software, in its version 3.3.3, SmartPLS GmbH, Oststeinbek, Germany.

The structure of the research is the following. First, there is a review of the literature related to the studied variables and relationships, from which the research hypotheses and the nomogram of the model are proposed. Then, the methodology used is described and, later, the results of the study are set out. Finally, the results are discussed, and the conclusion drawn, emphasizing the necessity to further deepen the analysis of the effects of closeness and cooperation dynamics on innovation.

#### **2. Literature Review**

#### *2.1. Industrial Clustering and Innovation*

Although by applying logic it could be determined that the accelerated process of the globalization of the economy, the reduction in the costs of transporting goods, and the development of information and communication technologies reduce the importance of location as a driver of business performance, the reality seems to indicate that the importance of the local environment has progressively increased [20] (p.256). Thus, nowadays the choice of the environment in which to locate business activities is a key strategic decision, which determines the characteristics of external agents that can favor the generation of certain localization economies and, consequently, a comparative advantage with respect to firms in geographically dispersed locations.

According to Baldwin and Von Hippel [21], it is essential to shift the focus of innovation generation and development from within firms to an open and collaborative path, in which firms and various agents and stakeholders, especially customers, work together in the design and development of innovations. This is because the capture of ideas and knowledge by companies through interaction with different specialized agents in the environment is a potential source of valuable resources that can lead to interesting opportunities for innovation [22]. In this sense, industrial agglomeration favors the generation of cooperative– competitive environments, in which the development of the relational networks of firms is promoted, alongside the pooling of a series of complementary resources and capabilities in favor of the achievement of shared objectives [23–25].

Moreover, productive specialization is one of the essential externalities derived from industrial agglomeration [26], which favors the technical specialization of processes and knowledge, not only of those companies dedicated to the industry in question, but also of those others that perform complementary activities [27]. In this sense, collaboration between various specialized economic agents, belonging to a given geographical area, increases the likelihood of the success of the innovation development processes undertaken by firms [28–30].

In addition, the exchange of knowledge between these specialized agents located geographically close to each other avoids duplication of efforts for the development of new knowledge and innovations [15]. Based on the above, the following hypothesis is proposed:

**Hypothesis 1 (+):** *There is a positive and significant relationship between the degree of agglomeration of the sector and the innovative performance of firms.*

#### *2.2. Business Associations and Interorganizational Cooperation*

According to Mejía-Villa, Tanco, and San Martín [19], the role of business associations as stimulators of innovation has been scarcely studied, despite the particularities conferred by their associative nature, which do not pursue lucrative purposes but simply defend the interests of their members and promote their positioning in the market, as well as enhance the dissemination of knowledge and the innovation capacity of their associates.

Business associations act as the promoters of collaborative innovation [19]. They are non-profit organizations [17] that usually bring together a large part of the entrepreneurs of a specific sector and region, in the first instance, exercising a certain degree of cohesion among them, in that their membership of the association is due to very similar motives and/or needs.

According to Nonaka and Konno [31], through continuous interaction at different levels between individuals and the sharing of their individual experiences, knowledge, and visions in a specific environment and time, the right conditions are established for the generation of trust, ideas, and new knowledge, as well as their development and integration. This can lead to the adoption of efficient, effective, and innovative solutions to their common problems. The effective development of open innovation processes requires that the communication channels work smoothly, as well as the existence of trust between the different individuals or agents involved in the process [32].

In this regard, socialization is an essential element for the generation of trust and the exchange and combination of ideas, information, and knowledge, especially tacit knowledge [33,34]. The latter is rooted in the experiences, aspirations, and values of individuals, which makes its dissemination difficult, although through continuous interaction between individuals in a specific place and time, its transfer can be promoted [31].

Zheng [35] concluded that various characteristics of the relational networks established by firms with agents in their environment, such as the strength of ties, trust, and shared norms and vision, can influence their ability to develop innovations. However, it is necessary to combine the agglomeration approach with other tools to accelerate results in terms of cooperation and innovation [36]. In this respect, business associations act as intermediary agents, favoring continuous interaction between entities with complementary resources and capabilities.

According to Sun, Zhang, and Zhang [37], the density of relational networks and the intensity of cooperation have a positive effect on the innovative performance of participants. Thus, they can influence the quality and quantity of cooperation agreements established by the company with agents in the environment and the transmission of tacit knowledge, especially among entities belonging to the association itself. Based on the above, the following hypothesis is proposed:

**Hypothesis 2 (+):** *There is a positive and significant relationship between active membership in a business association and the degree of cooperation of firms with surrounding agents.*

#### *2.3. Industrial Clustering, Cooperation and Innovation*

In today's knowledge society, access to external sources of valuable information and knowledge is essential for the survival of firms [38,39]. According to Audretsch, Hülsbeck, and Lehmann [40], collaborative networks contribute to the increased competitiveness of countries and regions through the pooling of resources and capabilities for the joint development of innovations.

Intangible resources based on knowledge drive the creation of value by companies [41] and favor their constant adaptation to current dynamic environments, in which it is difficult to make reliable forecasts [42]. This flexibility, in turn, facilitates access to new technologies and knowledge, as well as their assimilation and exploitation [43].

Geographical proximity does not automatically imply the establishment of cooperation agreements between entities located in a given region [44]. Interorganizational learning and the joint development of innovations seems to require, in addition to physical proximity, the existence of social and cognitive proximity between specialized economic agents located geographically close, which allows them to communicate effectively [45–47]. In this respect, industrial agglomeration implies the existence of a large number of entities that, in addition to being located geographically close, are specialized in a main industry, which can have a positive impact on social and cognitive proximity.

This can favor the membership of certain specialized social networks that, according to Alguezaui and Filieri [48], allows firms to access a wide set of valuable resources and capabilities shared by their members, especially new knowledge, which is an essential factor in the development of innovations.

Industrial agglomeration fosters the generation of a favorable environment for interfirm cooperation [49]. The geographical proximity derived from industrial agglomeration fosters the generation of links between economic agents belonging to a main industry, and boosts the strength of relationships, trust, the existence of shared values, and the efficiency and effectiveness with which knowledge is transferred [50–52].

Thus, the interaction between individuals derived from collaboration with external agents allows firms to increase their knowledge base and establish mutually beneficial relationships [53–55]. By making constant efforts in this area, collective learning dynamics are generated that drive the development of innovations [56–58]. In particular, the effective exploitation of external knowledge sources has a positive effect on the innovative performance of firms [59].

Although studies linked to the analysis of the so-called "New Economic Geography" provide certain theoretical explanations in this regard [60], at present there is still some confusion when trying to determine which of the positive externalities generated by industrial agglomeration favor the development of firms and their innovative performance [61]. Thus, this paper proposes that cooperation is one of the main factors derived from industrial agglomeration that favors the innovative performance of firms. Based on this, the following hypothesis is proposed.

**Hypothesis 3 (+):** *The degree of cooperation of firms with surrounding agents mediates the relationship between the degree of agglomeration of the sector and the innovative performance of firms.*

#### **3. Methodology**

#### *3.1. Population and Sample*

The population being studied is made up of the companies belonging to the electricity supply sector in Spain. According to the "Sistema de Análisis de Balances Ibéricos (SABI)" database, the number of active companies in Spain for the year 2019 is 13,339.

The sample is comprised of 197 companies in the sector, located in different regions of Spain. Although this sector only employs 2% of the total number of workers in Spain, its activity generates 13.3% of the gross added value, which positions it as the second most important sector, being only surpassed by the "Food, beverages and tobacco" sector, as well as presenting the highest average productivity per employee (449,800 euros) (According to the data issued by the National Statistics Institute for the financial year 2017: https://www.ine.es/, accessed on 3 February 2022).

#### *3.2. Data Collection and Measurement of Variables*

The data used to test the hypotheses come from both primary and secondary sources. In relation to the former, the data were obtained through the design and distribution of a

questionnaire to the companies in the population under study (The questionnaire is described in detail in Appendix A). A total of 11,757 emails were sent out, addressed to the companies' management staff, as they were considered to have a broad knowledge of the general functioning of the organization, as well as of the main decision-making bodies. The tools used for the design and distribution of the questionnaire were "Qualtrics" and "Microsoft Outlook" software, version January 2021, Qualtrics, Washington, DC, USA, respectively.

The questionnaire distribution process covered a period of 4 months, specifically from September to December 2020, during which, in addition to the initial mailing, several reminders were sent out, as well as telephone calls to encourage participants to collaborate with the research. After analyzing the completed questionnaires to determine their statistical validity, and after discarding any that were not considered valid (for different reasons, such as the existence of a large number of missing values, the presence of response patterns or a high number of single-value responses), a sample of 197 valid responses was collected. Hair, Hult, Ringle, and Sarstedt [62] establish, through their "minimum R2" method, that for a minimum R<sup>2</sup> value of the model equal to 0.500 and a maximum number of predictors of 2, the minimum sample size is 33 cases.

Regarding the secondary sources, we used the SABI database to determine the degree of agglomeration of companies and employment in the sector, as it allows us to determine the exact number of employees per company, as well as the number of companies per province, which increases the precision of the study. Due to the peculiarities inherent to the territorial organization of Spain, we decided to take the province as the territorial delimitation. In this way, the fifty provinces and the autonomous cities of Ceuta and Melilla were considered.

Innovative performance (dependent variable): A 7-point Likert-type scale with 13 items was used to measure this reflective variable. Specifically, validated scales of 5, 4, 3, and 1 items were used to measure the innovative performance of product, process, marketing, and management, respectively, based on the works of Prajogo and Ahmed [63] and Škerlavaj, Song, and Lee [64].

Degree of agglomeration of the sector (independent variable): This formative variable determines the degree of concentration of the sector under study in each of the regions. The geographical concentration of firms is measured in different ways in the agglomeration literature. Some studies use the density of firms in each specific industry and geographical area [65–67], while others use employment data [68,69]. In this paper, both indicators were used to calculate this construct. This type of coefficient has been widely used in many empirical works related to the study of clusters. Although they are simple indicators, they make it possible to determine in a clear and comprehensible way the regional distribution of companies and employees belonging to a given sector or industry. As territorial units of analysis we used the 50 provinces and 2 autonomous cities of Spain. Once the territorial division was established, the degree of regional concentration of employment and companies in the sector was analyzed in relation to the national average, by using the following coefficient

$$\text{AC} = \text{(a:b):(c:d)} \tag{1}$$

where AC: agglomeration coefficient. a: Sector units at regional level. b: Total units at regional level. c: Sector units at national level. d: Total units at national level [65–69].

This coefficient should be interpreted as follows. Regions that present a value greater than 1 have a degree of concentration of employees and/or companies in the sector that is greater than the national average. The higher the value of the coefficient, the greater the degree of concentration in the region in question.

Cooperation (mediating variable): Firms mainly have access to new knowledge through internal development or from external sources [70]. In this line, several authors highlight the role of cooperation with accessible external sources as a tool that allows the generation of ideas and knowledge sharing and, consequently, favors the increase of the knowledge base of the participating companies [71–73]. Thus, it is proposed that the degree of cooperation between firms and the different nodes of their network of relationships can be favored by belonging to a business association. This paper proposes that cooperation mediates the relationship between the degree of agglomeration of the sector and the innovative performance of firms. This reflective variable refers to the extent to which firms cooperate with the main stakeholders in their environment (competing firms, suppliers, customers, universities, technology centers, and other entities). It was measured by means of a Likert-type scale of 7 points and 6 items, which was elaborated based on the work of Claver-Cortés, Marco-Lajara, and Manresa-Marhuenda [74].

Association (antecedent variable): According to Dalziel [17], associations are nonprofit entities that bring together a high proportion of entrepreneurs in each sector and region. These associations defend the interests of their members and promote cooperation among them. Likewise, membership and participation in the association can favor the generation of relationships of trust between members, which in turn can influence their willingness to cooperate in a wide range of areas, pooling their resources and capabilities to achieve shared objectives. Association was established as an antecedent variable of cooperation. This reflective variable was measured by means of a 7-point Likert-type scale with 3 items, reflecting membership or not of an association and the type of participation, active or passive, of the associates.

#### *3.3. Analysis Technique*

To test the hypotheses, structural equation modelling (SEM) was used; specifically, the second-generation multivariate analysis technique of partial least squares (PLS), which according to Hair, Sarstedt, Pieper, and Ringle [75], has gained great relevance in recent years among researchers in the field of strategic management of the company.

For this purpose, the SmartPLS software, version 3.3.3 was used [76]. According to Hair, Hult, Ringle, Sarstedt, Castillo Apraiz, Cepeda Carrión, and Roldán [77], it is a suitable technique for predictive analytics, especially in the field of social sciences. In addition, it allows us to test models of linear relationships between variables, including those of a latent nature.

The PLS-SEM technique uses the maximization of the explained variance of the observable and unobservable variables to estimate the parameters of the established model [78]. According to these authors, due to the above, this technique is particularly suitable for research in the field of social sciences. It has been evidenced that the PLS-SEM method obtains greater flexibility and robustness than traditional approaches [79].

This method of analysis has been chosen for different reasons. Fundamentally, because the study is predictive in nature, making it suitable for the use of the PLS-SEM technique [80]. In addition, it is an efficient tool for the estimation of complex models, which allows working with relatively small sample sizes, and with data that do not follow a specific distribution [77].

Finally, the proposed model includes second-order latent variables, and the PLS technique allows us to efficiently estimate this type of multidimensional model [81].

#### **4. Data Analysis and Results**

#### *4.1. Data Analysis and Results*

The results of the first phase of analysis, relating to the distribution of the sector in Spain, are presented in Table 1 below. Specifically, the values corresponding to the degree of agglomeration of the sector (companies and employment) at the provincial level, in comparison with the national average, are shown. The distribution of the sample according to the degree of agglomeration of the sector in the province in which they are located, in relation to the national average, is also shown.


**Table 1.** Distribution of the sample in relation to the degree of agglomeration of the sector under study, in absolute and relative values.

Table 2 details the distribution of the population of companies compared to the sample in relative terms, according to their location or not, in a region that presents agglomeration of firms or employment in the sector, in comparison with the national average.

**Table 2.** Comparison of the distribution of the population and the sample in relation to the degree of agglomeration of the sector under study, in relative values.


Table 3 shows brief descriptive statistics of the data obtained in the questionnaire.


**Table 3.** Descriptive statistic of the data obtained in the questionnaire.

Note: D.A.S.: Degree of agglomeration of the sector. I.P.: Innovative performance. S.D.: Standard deviation.

Finally, Figures 1 and 2 show in detail the agglomeration coefficient of the sector under study in each of the provinces, in relation to the number of companies and employment, respectively. To facilitate its reading, a color range has been established, which varies from dark red, for levels of agglomeration much lower than the national average, to dark green, for high levels of agglomeration. The other colors indicate intermediate levels.

*Energies* **2022**, *15*, x FOR PEER REVIEW 8 of 17

**Figure 1.** Degree of agglomeration of companies in the sector at the provincial level in Spain. **Figure 1.** Degree of agglomeration of companies in the sector at the provincial level in Spain. **Figure 1.** Degree of agglomeration of companies in the sector at the provincial level in Spain.

*4.2. Model Evaluation*  **Figure 2.** Degree of agglomeration of employment in the sector at the provincial level in Spain. *4.2. Model Evaluation* **Figure 2.** Degree of agglomeration of employment in the sector at the provincial level in Spain.

#### *4.2. Model Evaluation*

The model under analysis includes multidimensional constructs that, according to [82], are composed of different related dimensions that can be analyzed as a single theoretical concept. According to Van Riel, Henseler, Kemény, and Sasovova [83], a first analysis should be performed to obtain the scores of the first-order latent variables, which will be used in the subsequent analysis to model the second-order constructs. PLS is an effective tool for process [84]. Ringle, Sarstedt, and Straub [85] state that this two-stage process, commonly used in social science research, allows the second-order construct to be established in an endogenous way within the structural model.

The results obtained after evaluating the research model using PLS-SEM, which, according to Hair et al. [77], should be carried out in two stages, the first corresponding to the measurement model and the second to the structural model, which are presented below. The final models, both saturated and estimated, present a good fit, as they have a standardized root mean square residual [SRMSR] value of 0.057 < 0.08 [86].

#### 4.2.1. Evaluation of the Formative and Reflective Measurement Models

When evaluating measurement models, different criteria must be followed depending on the type of construct involved, formative or reflective [77]. For the evaluation of the formative model (degree of sector agglomeration), a single item representing the essence of the latent variable that formative indicators seek to measure is used [87]. In this first step, which is also called "redundancy analysis" [88] and determines the convergent validity of the model, the degree of correlation between the different measures of the same construct is assessed by using different indicators. For this, the formative latent variable is used as an exogenous variable, which acts as a predictor of another endogenous construct that uses other indicators of a reflective nature.

Although, generally, the use of individual indicators is not recommended when using the PLS technique, in the case of redundancy analysis it is appropriate, since its objective is not to capture the total content of the construct but simply its fundamental elements, to have a standard of comparison [77].

Thus, this item, called "PSecGDP", indicates the percentage share of the sector under study of the GDP at the provincial level (For its calculation, secondary data obtained from the SABI database and the National Statistics Institute (INE) are used). According to Hair et al. [77], the value of the path coefficient between the two constructs should be greater than 0.7 and the value of R<sup>2</sup> greater than 0.5. The path coefficient between both formative and reflective indicators takes the value 0.916 > 0.8, and the R<sup>2</sup> amounts to 0.839 > 0.5, which means that the formative measurement model meets the convergent validity criterion.

The degree of collinearity of the formative indicators is significantly lower than the critical values set, as the VIF value amounts to 1.192 < 3 [89]. Finally, the significance and relevance of the formative indicators are assessed.

After running the bootstrapping process, in full mode and 5,000 random subsamples, it was found that both the weights and the external loadings of the formative indicators have significantly different values from zero, both in relative terms (External weights: L1 = 0.583 (Employment agglomeration coefficient); L2 = 0.612 (Agglomeration coefficient of the companies)) and in absolute terms (External loadings: L1 = 0.828; L2 = 0.846), which indicates that their contribution to the construct is high.

To evaluate the reflective model, internal consistency and convergent and discriminant validity must be analyzed [77]. According to these authors, in the first case, three methods are used: Cronbach's alpha (Tends to underestimate internal consistency reliability) (α), composite reliability (Tends to overestimate internal consistency reliability) (ρc) and Dijkstra–Henseler's rho (It is considered a measure of consistent reliability) (ρA).

As can be seen in Table 4, all values are significantly higher than 0.7 [77,88,90]. To confirm convergent validity, the measurement is performed by assessing the reliability of the indicators, that is, the size of the external loadings (λ) and the average variance extracted (AVE), which refers to the total mean value of the loadings of the indicators belonging to the same construct squared [77]. Furthermore, it is observed that the value of the external loadings is greater than 0.707, and the AVE > 0.5, so this criterion is also met [77,91].


**Table 4.** Assessment of internal consistency and convergent validity.

Note: COOP: Cooperation. Tech.: Technology. I.P.: Innovative performance.

Traditionally, two methods are used: cross-loading analysis and the Fornell and Larcker method. Although both criteria are fulfilled, according to [91], these methods have certain shortcomings that affect the detection of discriminant validity problems. These authors determine that the heterotrait–monotrait ratio (HTMT) is a more effective tool for this purpose.

According to Kline [92], the value of the HTMT ratio should be less than 0.85. Table 5 shows that the model largely meets this requirement.

**Table 5.** Evaluation of discriminant validity.


Note: I.P.: Innovative performance.

#### 4.2.2. Evaluation of the Structural Model

The evaluation of the structural model allows us to determine the predictive capacity of the model and the type of relationships existing between the different latent variables that compose it and, consequently, to test the hypotheses put forward in the theoretical framework.

To this end, according to Hair et al. [77] the following elements must be analyzed: the significance level and relevance of the relationships, the value of the coefficients of determination (R<sup>2</sup> ), collinearity, effect size ƒ<sup>2</sup> , and predictive relevance (Q<sup>2</sup> ).

Figure 3 shows the nomogram, where the path coefficients obtained by running the bootstrapping process in full mode and 5000 random subsamples can be observed. Tables 6 and 7 show data corresponding to the direct and indirect effects, respectively. Figure 3 shows the nomogram, where the path coefficients obtained by running the bootstrapping process in full mode and 5000 random subsamples can be observed. Tables 6 and 7 show data corresponding to the direct and indirect effects, respectively.

*Energies* **2022**, *15*, x FOR PEER REVIEW 11 of 17

**Figure 3.** Path coefficients and significance levels of the proposed model. **Figure 3.** Path coefficients and significance levels of the proposed model.



Note: C.I.: Confidence interval. D.A.S.: Degree of agglomeration of the sector. I.P.: Innovative performance. Coef.: Coefficient. S.D.: Standard deviation. \*\* Statistically significant at 1%. Note: C.I.: Confidence interval. D.A.S.: Degree of agglomeration of the sector. I.P.: Innovative performance. Coef.: Coefficient. S.D.: Standard deviation. \*\* Statistically significant at 1%.

When analyzing the data, it is observed that there is no collinearity, as all VIF values **Table 7.** Summary of indirect effects.


tively. Note: C.I.: Confidence interval. D.A.S.: Degree of agglomeration of the sector. I.P.: Innovative performance. Coef: Coefficient. \*\* Statistically significant at 1%.

**Table 7.** Summary of indirect effects. **Total Effect of D.A.S. on I.P. Direct Effect of D.A.S. on I.P. Indirect Effect of D.A.S. on I.P. Results Coef. (β) t0.005, 4999 Coef. (β) t0.005, 4999 Estimate t0.005, 4999 C.I. 99%**  0.612 \*\* 14.241 0.412 \*\* 4.991 0.200 3.028 [0.067–0.371] H3 (+) When analyzing the data, it is observed that there is no collinearity, as all VIF values are less than three [89]. There is a positive and significant direct effect of the degree of industry agglomeration on the innovative performance of firms [0.412, *p* = 0.000]. In addition, there is a positive and significant indirect effect produced by the mediation of the variable "Cooperation" [0.200, *p* = 0.000]. The proposed model explains 53.7% and 50.8% of the variance of the constructs "Cooperation" and "Innovative Performance", respectively.

> Note: C.I.: Confidence interval. D.A.S.: Degree of agglomeration of the sector. I.P.: Innovative performance. Coef: Coefficient. \*\* Statistically significant at 1%. Furthermore, the contribution of the exogenous construct "Degree of industry agglomeration" to the R2 value of the endogenous latent variables "Cooperation" and "Innovative Performance" (ƒ2) is large [0.397] and median [0.167], in this order, based on the values proposed by [93]. In turn, the contribution of the exogenous construct "Cooperation" to the R2 value of the endogenous latent variable "Innovative Performance" is median [0.115]. The data point to the existence of a direct and positive effect of belonging to Furthermore, the contribution of the exogenous construct "Degree of industry agglomeration" to the R<sup>2</sup> value of the endogenous latent variables "Cooperation" and "Innovative Performance" (ƒ 2 ) is large [0.397] and median [0.167], in this order, based on the values proposed by [93]. In turn, the contribution of the exogenous construct "Cooperation" to the R<sup>2</sup> value of the endogenous latent variable "Innovative Performance" is median [0.115]. The data point to the existence of a direct and positive effect of belonging to an association on the degree of cooperation of firms [0.199, *p* =0.006], although the contribution of the exogenous construct "Association" on the endogenous latent variable "Cooperation" is small [93].

Finally, the values of the endogenous variables' "cooperation" and "innovative performance" have Q<sup>2</sup> values 0.335 and 0.341, respectively, which indicates that the model

has a moderate predictive relevance on the mentioned variables, in accordance with the values established by Hair, Risher, Sarstedt, and Ringle [94]. Based on these data, the three hypotheses are confirmed:

 **H1 (+):** *There is a positive and significant relationship between the degree of agglomeration of the sector and the innovative performance of firms.*

 **H2 (+):** *There is a positive and significant relationship between active membership in a business association and the degree of cooperation of firms with surrounding agents.*

 **H3 (+):** *The degree of cooperation of firms with surrounding agents mediates the relationship between the degree of agglomeration of the sector and the innovative performance of firms.*

#### **5. Conclusions**

This paper contributes to the existing literature on the analysis of positive externalities derived from industrial agglomeration, particularly those related to cooperation and the innovative performance of firms. The three hypotheses are confirmed, thus establishing a positive and significant relationship between the variables of the proposed model. Based on these findings, three main conclusions are established. First, the positive externalities derived from industrial agglomeration boost the innovative performance of firms. Second, cooperation between specialized entities is established as an important positive externality derived from industrial agglomeration, which has a positive effect on the innovative performance of firms. Finally, business associations foster cooperation, especially among their associates, and act as representative entities that facilitate the establishment of agreements at the superstructure level.

Collaboration with external actors in knowledge sharing and the development of collaborative innovation processes increases the chances of superior innovative performance [95]. However, firms should carefully select the partners with whom they cooperate, as this is an element that directly influences the outcomes of partnerships [96]. In this regard, belonging to a region with a high degree of agglomeration in each sector implies a high degree of specialization of the linked firms [27].

Thus, industrial agglomeration derives from the existence of specialized entities located geographically close, which translates into the availability of potentially valuable partners with which to establish cooperation agreements. In addition, business associations play an intermediary role that favors cooperation at different levels. In this context, continuous interaction with specialized industry players boosts the innovative performance of firms [97].

The results obtained in this study highlight the importance of location and cooperation as drivers of firms' innovative performance. The environment in which they are located can provide opportunities for access to potential sources of valuable resources and capabilities, especially new knowledge. Therefore, despite globalization and the accelerated development of information and communication technologies in recent decades, and especially in recent years, there are certain elements associated with geographical and cognitive proximity that favor the effective transfer of knowledge and the innovative performance of companies. For this reason, it is necessary to continue to study industrial agglomeration in depth as a tool for regional economic development and competitiveness.

In relation to the limitations of the work, it is worth highlighting the eminently external focus of the study. Thus, in future work it would be interesting to incorporate the influence of the internal factors of the companies in the study of the determinants of their innovative performance. In particular, it would be interesting to incorporate the absorptive capacity of firms, which could influence the degree to which firms benefit from collaboration with external agents [98].

**Author Contributions:** Conceptualization, B.M.-L. and E.S.-G.; Data curation, B.M.-L., E.S.-G. and J.M.-F.; Formal analysis, E.S.-G. and E.P.-P.; Investigation, B.M.-L., E.S.-G. and J.M.-F.; Methodology, J.M.-F. and E.P.-P.; Project administration, B.M.-L.; Resources, B.M.-L., E.S.-G. and J.M.-F.; Software, E.S.-G. and E.P.-P.; Validation, E.S.-G. and E.P.-P.; Writing—original draft, B.M.-L. and E.S.-G.; Writing—review & editing, J.M.-F. and E.P.-P. All authors have read and agreed to the published version of the manuscript.

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

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Not applicable.

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


#### **Appendix A**


#### **References**


#### *Article* **Optimizing Multi Cross-Docking Systems with a Multi-Objective Green Location Routing Problem Considering Carbon Emission and Energy Consumption**

**Ieva Meidute-Kavaliauskiene 1,\*, Nihal Sütütemiz <sup>2</sup> , Figen Yıldırım <sup>3</sup> , Shahryar Ghorbani <sup>4</sup> and Renata Cinˇcikait ˇ e˙ 1**


**Abstract:** Cross-docking is an excellent way to reduce the space required to store goods, inventory management costs, and customer order delivery time. This paper focuses on cost optimization, scheduling incoming and outgoing trucks, and green supply chains with multiple cross-docking. The three objectives are minimizing total operating costs, truck transportation sequences, and carbon emissions within the supply chain. Since the linear programming model is an integer of zero and one and belongs to NP-hard problems, its solution time increases sharply with increasing dimensions. Therefore, the non-dominated sorting genetic algorithm-II (NSGA-II) and the multi-objective particle swarm optimization (MOPSO) were used to find near-optimal solutions to the problem. Then, these algorithms were compared with criteria such as execution time and distance from the ideal point, and the superior algorithm in each criterion was identified.

**Keywords:** non-dominated sorting genetic algorithm-II (NSGA-II); multi-objective particle swarm optimization (MOPSO); cross-docking

#### **1. Introduction**

Logistics strategy is an essential advantage for supply chain management operations that requires centralized planning and adjustment of procedures to reduce costs and increase customer satisfaction [1,2]. Classic supply chain management for transporting goods from suppliers to the distribution centre requires warehousing for storage [3] and then packaging and shipping to customers [4], according to their demand [5]. This form of supply chain management has high inventory maintenance costs and human resources costs [6,7]. In other words, there are four stages of receiving, storage, order selection, and sending in traditional warehouses [8]. Still, in the cross-dock, the storage stage, one of the main obstacles of inventory systems, is removed [9,10]. Therefore, it reduces inventory and increases customer satisfaction [11]. Cross-dock is a logistic method that distributes products directly from a supplier or factory [12]. The cross-dock method ensures timely transportation and better coordination between supply and demand [13]. In recent years, one of the issues facing the supply chain was clean production and environmental protection [5]. One of the critical issues of clean output is using supply chains that pay attention to environmental issues in all supply chain processes [14,15].

In this research, we investigate the operational cross-docking decisions that concern the optimization of short-term decisions such as CO<sup>2</sup> emission and energy consumption that are directly related to the transit of products from inbound trucks to outbound trucks.

**Citation:** Meidute-Kavaliauskiene, I.; Sütütemiz, N.; Yıldırım, F.; Ghorbani, S.; Cinˇcikait ˇ e, R. Optimizing Multi ˙ Cross-Docking Systems with a Multi-Objective Green Location Routing Problem Considering Carbon Emission and Energy Consumption. *Energies* **2022**, *15*, 1530. https://doi.org/10.3390/en15041530

Academic Editors: Bernard Zi˛ebicki and Edyta Bieli ´nska-Dusza

Received: 21 December 2021 Accepted: 16 February 2022 Published: 18 February 2022

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

**Copyright:** © 2022 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/).

We focus on the optimization of the truck to door assignment problem, which is one of the key issues in cross-docking. It attempts to find the optimal assignment for each incoming inbound or outbound truck to the appropriate inbound or outbound dock according to the cross-dock characteristics. When the problem of truck-to-door assigning is mentioned, two different questions should be answered: at which door and in which order the trucks should be docked. The first question was answered by many researchers, but the second question receives less attention and this research tries to investigate the answer to this question. The cross-docking system targets to reduce the cost of inventory holding and minimize the delivery time from suppliers to retailers in the supply chain. However, this paper addresses the cross-dock selection and optimization problem, which was first considered with the vehicle routing problem (VRP) by Maknoon and Laporte (2017), to make optimal decisions on route construction and freight consolidation cross-docks. In other words, selecting cross-docks as freight consolidation points leads to gaining economies of scale. The products are processed and transported to the customers by at least one cross-dock in the CDS problem. It has a vast application in production and retail companies and logistic service providers that handle various shipments on large networks.

Dondo and Cerda (2014) took a different approach to applying the time window so that a variable was used to determine the time window interval [16]. If there is a time window, this variable will be greater than zero for error; otherwise, it will be zero [13]. Mohtashami et al. (2015), in a study to provide and optimize a multi-objective mathematical model in the cross-dock and the entire supply chain, assumed that the supplier can also send the product directly to the customer and send it to the cross-dock [15]. On the other hand, multiple objectives, such as minimizing the total transport time of trucks, total cost, number of shipments, and model solving with NSGA-II and MOPSO with unsuccessful sorting and comparing the results of the above two algorithms, were included in their research. Mohtashami (2015) presented a new dynamic genetic algorithm based on the vehicle scheduling method in the cross-dock system to reduce the operation time. This algorithm assumed the existence of a temporary warehouse for receiving and sending goods, frequent entry and exit of trucks, and two types of chromosomes determined for incoming and outgoing trucks, providing a shorter solution for the operating time [17]. Ponboon et al., 2016, evaluated the cost structure and the impact of parameters on location issues and time windows in the distribution network. Nine scenarios of the Osaka distribution network of freight cargo with different warehouse locations were tested with two criteria, warehouse size and vehicle size. Therefore, using a large warehouse with a large vehicle will reduce costs. Warehouse features, vehicles, and shipping information were also discussed in detail [18]. Gomes et al. (2018) conducted bibliometric research in warehouse management systems from 2006 to 2016 [19].

Birim (2016) investigated the vehicle routing problem with cross-dock, considering heterogeneous vehicles with variable capacity and seeking to minimize the objective function of the total cost of transportation. The problem was solved with a simulated annealing algorithm, and the best answer is compared [20]. Yin and Chuang (2016) emphasized that we must focus on the green supply chain for sustainable development in the supply chain and pay attention to the friendly environment and business values. This paper deals with the minimum cost of green vehicle routing in transporting final products from suppliers to customers through a cross-dock with limited carbon dioxide emissions. Load management for long distances with the lowest cost and carbon dioxide emissions determine high fuel efficiency with the tabu search algorithm [21]. Wisittipanich and Hengmeechai (2017) said that one of the fundamental operational management problems is scheduling trucks in assigning them to cross-dock doors and arranging all domestic and foreign trucks for loading and unloading. This paper presents a mathematical model of integer hybrid programming for allocating and setting trucks in a multi-door cross-docking system. The purpose of this model is to minimize the total operation time. Then, a modified particle swarm algorithm is proposed and optimized with special unique designs, coding, and decoding to solve the problem of truck scheduling in a cross-docking system [16]. Zulaga et al. (2017) showed

that this type of dock has results such as reducing cost time and improving management information in reverse processes. Sensitivity analysis of this model helps companies enhance their competitive position by providing flexibility concerning products, reducing the likelihood of product returns from the secondary market, and combining product returns and cross-dock costs relative to traditional warehousing [14]. Mohtashami and Najafabadi (2014) examined the reduction of operating time in the entire supply chain in their article. In the proposed model, the incoming trucks move directly towards the customers or the cross-dock in the first stage after loading the products from the suppliers. The products are unloaded in the cross-dock and loaded in the output trucks in the second stage. Then, the products are transferred to the customers [22].

Table 1 summarize the literature regarding optimizing multi cross-docking systems (MCDS) and compare the existing problems with the problem proposed in this paper.

**Table 1.** Summary of literature regarding MCDS and its different variants.


Since cross-docks and related transportations have significant roles in increasing the efficiency of large-scale supply chain distribution networks, considering the other real-world essential factors such as reliability and pollution is attractive in designing or redesigning this network. Reliability maximization leads to customer satisfaction and earning a lot of market shares, whereas pollution minimization creates eco-friendly industries.

The research conducted was generally related to cross-dock and focused on optimizing the timing of trucks and the cost of transportation inside the dock. In this research, by considering several cross-docks and concentrating on green management with a norm time window approach for customers, issues related to out-of-cross-dock and the entire supply chain operations are also addressed. This study focuses on optimizing the cost of transporting trucks by locating the cross-dock, the sequence of transporting trucks, and the amount of carbon dioxide emissions in the environment simultaneously throughout the supply chain. In summary, the contributions of this research are described as follows.


#### **2. Problem Statements**

In this paper, three objective functions, including minimizing the total cost of operation, truck transport sequence, and carbon dioxide emissions, are examined. The proposed model also examines the relationship between (cross-dock–suppliers), (cross-dock–customers), (suppliers–customers), (suppliers–suppliers), (customers–customers), and (cross-dock– cross-dock) in terms of fuzzy scheduling with a norm time window approach for truck transport and reducing greenhouse gas emissions. The assumptions of the model are:


To better understand and explain the research problem, the flow of operations in the supply chain is drawn in Figure 1. This figure considers several cross-dock distribution and transportation operation flows and several customers and suppliers. Vehicles with different capacities are also used to transport products in the supply chain.

**Figure 1.** Operation process in the cross-dock.

After loading the products from the suppliers, the incoming trucks move directly to the customers, move to another supplier, move to one of the cross-docks, and unload the products in the cross-dock. They are then loaded onto output trucks and delivered to customers [36]. It should also be noted that a truck can load products from more than one supplier, and a truck can move to more than one customer and unload products between them. Three trucks with different capacities are used to transport the products.

#### *2.1. Mathematical Model*

The main variables and important parameters in the model and algorithms are presented in Table 2:

**Table 2.** Sets, parameters, and decision variables of the model.


#### **Table 2.** *Cont.*


#### **Table 2.** *Cont.*


The first objective function minimizes the total cost of operations within the supply chain (including transportation, fuel, delay, shortage, holding, and opening costs). The second objective function minimizes the truck transport sequence between suppliers, crossdocks, and customers, and the third objective function minimizes CO<sup>2</sup> emissions from trucks throughout the system.

Min Z<sup>1</sup> = " <sup>K</sup> ∑ i∈*Kin*=1 S ∑ i∈s=1 Z ∑ i∈cd=1 G ∑ i∈G=1 (C i scdG + C i sGcd) ∗ X i scdG! + K ∑ i∈*Kin*=1 S ∑ i∈s=1 C ∑ i∈c=1 G ∑ i∈G=1 C i scG∗X i scG! + K ∑ i∈*kin*=1 S ∑ s=1 G ∑ i∈G C i <sup>s</sup>msnG∗ X i smsnG ! + K ∑ i∈*Kin*=1 S ∑ i∈s=1 Z ∑ i∈cd=1 G ∑ i∈G=1 (C i cdsG∗ X i cdsG! + G ∑ i∈G=1 C ∑ i∈c=1 K ∑ i∈*kin* C i <sup>c</sup>mcnG∗ X i cmcnG ! + K ∑ j∈*kout*=1 Z ∑ i∈cd=1 C ∑ i∈c=1 G ∑ i∈G=1 (C j cdcG∗ X j cdcG! K ∑ j∈*Kin*=1 i∈ Z ∑ i∈cd=1 G ∑ i∈G=1 C i cdmcdnG ∗ X i cdmcdnG ! + K ∑ j∈*Kout*=1 i∈ C ∑ c=1 G ∑ i∈G=1 C j cmcnG ∗ X j cmcnG ! + K ∑ j∈*Kout*=1 C ∑ i∈c=1 Z ∑ i∈cd=1 G ∑ i∈G=1 c j ccdG∗X j CCdG! + K ∑ j∈*Kin*=1 C ∑ i∈c=1 S ∑ i∈s=1 G ∑ i∈G=1 c i csG∗X i CsG! +( Z ∑ i∈cd=1 *O*i*y*<sup>i</sup> + S ∑ i∈s=1 I ∑ i∈I=1 G ∑ i∈G=1 *F*<sup>i</sup> ∗ X i scdG) + [ K ∑ i∈*Kin*=1 S ∑ i∈s=1 C ∑ i∈c=1 G ∑ i∈G=1 *PEkin* ∗ (γ<sup>i</sup> + θi) ∗ X i scG# +[ K ∑ *j*∈*Kout*=1 I ∑ i∈I=1 C ∑ i∈c=1 G ∑ i∈G=1 *PEkOUT* ∗ (γ<sup>i</sup> + θi) ∗ X j cdcG] (1)

$$\text{Min\,Z\,m} = \text{F} = \sum\_{\text{i}=\text{s}=1}^{\text{S}} \sum\_{\text{i}=\text{e}=1}^{\text{C}} \sum\_{\text{i}=\text{e}l}^{\text{C}} \sum\_{\text{i}=\text{k}u=1}^{\text{K}} \sum\_{\text{i}=\text{k}u}^{\text{K}} \sum\_{\text{i}=\text{G}+1}^{\text{G}} \lambda\_{\text{scG}}^{\text{i}} + \lambda\_{\text{scG}G}^{\text{i}} + \lambda\_{\text{scG}G}^{\text{i}} + \lambda\_{\text{scG}G}^{\text{i}} + \lambda\_{\text{cdcG}}^{\text{i}} + \lambda\_{\text{cdcG}}^{\text{i}} + \lambda\_{\text{cdcG}}^{\text{i}} \tag{2}$$

Min Z<sup>3</sup> = " [ S ∑ i∈s=1 Z ∑ i∈cd=1 K ∑ i∈*kin*=1 C*kin* ∗g i scd∗dscd∗X i scd! +( S ∑ i∈s=1 S ∑ i∈s=1 G ∑ i∈G=1 K ∑ i*ekin* C / *kin* ∗ *Qkin* ∗fscdG∗WG∗dSCdG∗X i scdG)] + [( K ∑ i∈*kin*=1 S ∑ i∈s=1 C ∑ i*e*c=1 C*kin* ∗g i sc∗dsc∗X i sc) +( S ∑ i∈s=1 C ∑ i∈c=1 G ∑ i∈G=1 K ∑ i*ek*i*<sup>n</sup>* C / *k*i*<sup>n</sup>* ∗ *Qk*i*<sup>n</sup>* ∗fscG∗WG∗dSCG∗X i scG)] + [( K ∑ i∈*k*i*n*=1 S ∑ i∈s=1 C*k*i*<sup>n</sup>* ∗g i sms<sup>n</sup> ∗ dsms<sup>n</sup> ∗X i sms<sup>n</sup> ) +( K ∑ i∈*kin*=1 S ∑ i∈s=1 G ∑ i∈G=1 C / *kin* ∗ *Qkin* ∗fsmsnG∗WG∗dsmsnG∗X i smsnG )] + [( K ∑ i∈*kin*=1 Z ∑ i∈cd=1 C*kin* ∗g i cdmcd<sup>n</sup> ∗dcdmcd<sup>n</sup> ∗X i cdmcd<sup>n</sup> ) +( K ∑ i∈*kin*=1 Z ∑ i∈cd=1 G ∑ i∈G=1 C / *kin* ∗ *Qkin* ∗fcdmcdnG∗WG∗dcdmcdnG)∗X i cdmcdnG ] + [( Z ∑ i∈cd=1 C ∑ i∈c=1 K ∑ i∈*kout*=1 C*kout* ∗g j cdc∗dcdc∗X j cdc) +( Z ∑ i∈cd=1 C ∑ i∈c=1 G ∑ i∈G=1 K ∑ i*ekout* C / *kout* ∗ *Qkout* ∗fcdcG∗WG∗dCdcG∗X j cdcG)] + [( K ∑ i∈*kout*=1 C ∑ i∈c=1 C*kout* ∗g j cmc<sup>n</sup> ∗dcmc<sup>n</sup> ∗X j cmc<sup>n</sup> ) +( K ∑ i∈*kout*=1 C ∑ i∈c=1 G ∑ i∈G=1 C / *kout* ∗ *QkOUT* ∗fcmcnG∗WG∗dcmcnG∗X j cmcnG )]# (3)

*S.t:*

$$\sum\_{\mathbf{i}\in\mathcal{K}\_{\rm in}}^{\mathbf{K}} \sum\_{\mathbf{i}\in\mathbf{c}=1}^{\mathbf{C}} \sum\_{\mathbf{i}\in\mathbf{s}=1}^{\mathbf{S}} \sum\_{\mathbf{i}\in\mathbf{G}=1}^{\mathbf{G}} \mathbf{q}\_{\mathbf{G}\mathbf{s}} \* \lambda\_{\mathbf{s}\mathbf{C}}^{\mathbf{i}} = \mathbf{D}\_{\mathbf{c}\mathbf{G}} \tag{4}$$

$$\sum\_{\mathbf{i}:\mathbf{i}\in\mathbf{d}=1}^{\mathbf{Z}}\sum\_{\mathbf{j}\in k\_{\mathrm{off}}=1}^{\mathbf{K}}\sum\_{\mathbf{i}\in\mathbf{C}=1}^{\mathbf{C}}\sum\_{\mathbf{i}\in\mathbf{C}=1}^{\mathbf{C}}\mathbf{q}\_{\mathrm{Ccd}}\,\mathop{\mathbf{\*}\mathbf{X}}\_{\operatorname{cdd}\mathbf{G}}\,\mathop{\mathbf{\*}\mathbf{X}}\_{\operatorname{cdd}\mathbf{G}} + \sum\_{\mathbf{i}\in\mathbf{K}\_{\mathrm{ll}}=1}^{\mathbf{K}}\sum\_{\mathbf{i}\in\mathbf{c}=1}^{\mathbf{S}}\sum\_{\mathbf{i}\in\mathbf{c}=1}^{\mathbf{C}}\sum\_{\mathbf{i}\in\mathbf{G}=1}^{\mathbf{G}}\mathbf{q}\_{\mathrm{C}\mathbf{s}}\,\mathop{\mathbf{\*}\mathbf{X}}\_{\operatorname{s\mathbf{G}}}^{\mathbf{i}} = \mathbf{P}\_{\mathbf{G}}\tag{5}$$

$$\sum\_{\substack{\mathbf{i}\in K\_{\text{int}}=1}}^{K} \sum\_{\substack{\mathbf{i}\in s=1}}^{S} \sum\_{\substack{\mathbf{i}\in c=1}}^{C} \sum\_{\mathbf{i}\in G=1}^{G} \lambda\_{\text{scG}}^{i} \ge 1 \tag{6}$$

$$\mathbf{X}\_{\text{cls}G}^{\text{i}} = \sum\_{\mathbf{i} \in \mathbf{X}\_{\text{ll}} = 1}^{\text{K}} \sum\_{\mathbf{i} \in \mathbf{s} = 1}^{\text{S}} \sum\_{\mathbf{i} \in \mathbf{G} = 1}^{\text{G}} \mathbf{X}\_{\text{s} \text{n}}^{\text{i}} \mathbf{G} + \sum\_{\mathbf{i} \in \mathbf{K}\_{\text{il}} = 1}^{\text{K}} \sum\_{\mathbf{i} \in \mathbf{s} = 1}^{\text{S}} \sum\_{\mathbf{i} \in \mathbf{c} = 1}^{\text{G}} \sum\_{\mathbf{i} \in \mathbf{c} = 1}^{\text{G}} \mathbf{X}\_{\text{scdG}}^{\text{i}} + \sum\_{\mathbf{i} \in \mathbf{K}\_{\text{il}} = 1}^{\text{K}} \sum\_{\mathbf{i} \in \mathbf{c} = 1}^{\text{S}} \sum\_{\mathbf{i} \in \mathbf{c} = 1}^{\text{G}} \sum\_{\mathbf{i} \in \mathbf{G} = 1}^{\text{G}} \mathbf{X}\_{\text{scdG}}^{\text{i}} \tag{7}$$

$$\mathbf{X}\_{\mathbf{s}\_{\mathrm{in}}\mathbf{s}\_{\mathrm{in}}\mathbf{G}}^{\mathrm{i}} = \sum\_{\mathbf{i}\in\mathcal{K}\_{\mathrm{in}}}^{\mathrm{K}} \sum\_{\mathbf{i}\in\mathbf{s}=\mathbf{1}}^{\mathrm{S}} \sum\_{\mathbf{i}\in\mathbf{G}=\mathbf{1}}^{\mathrm{G}} \mathbf{X}\_{\mathbf{s}\_{\mathrm{in}}\mathbf{G}}^{\mathrm{i}} + \sum\_{\mathbf{i}\in\mathbf{K}\_{\mathrm{in}}=\mathbf{1}}^{\mathrm{K}} \sum\_{\mathbf{i}\in\mathbf{s}=\mathbf{1}}^{\mathrm{S}} \sum\_{\mathbf{i}\in\mathbf{c}=\mathbf{1}}^{\mathrm{Z}} \sum\_{\mathbf{i}\in\mathbf{G}=\mathbf{1}}^{\mathrm{G}} \mathbf{X}\_{\mathbf{s}\_{\mathrm{out}}\mathbf{G}}^{\mathrm{i}} + \sum\_{\mathbf{i}\in\mathbf{K}\_{\mathrm{in}}=\mathbf{1}}^{\mathrm{K}} \sum\_{\mathbf{i}\in\mathbf{c}=\mathbf{1}}^{\mathrm{G}} \sum\_{\mathbf{i}\in\mathbf{c}=\mathbf{1}}^{\mathrm{G}} \sum\_{\mathbf{i}\in\mathbf{G}=\mathbf{1}}^{\mathrm{G}} \mathbf{X}\_{\mathbf{s}\mathbf{G}}^{\mathrm{i}} \tag{8}$$

$$\mathbf{X}\_{\text{cdcG}}^{\text{j}} = \sum\_{\substack{\mathbf{i} \in \mathbb{K}\_{\text{out}} = 1}}^{\text{K}} \sum\_{\mathbf{i} \in \mathbb{c} = 1}^{\text{C}} \sum\_{\mathbf{i} \in \mathbb{G} = 1}^{\text{C}} \mathbf{X}\_{\text{cm} \text{Cr} \mathbf{G}}^{\text{j}} + \sum\_{\substack{\mathbf{i} \in \mathbb{K}\_{\text{out}} \ \mathbf{i} \in \mathbb{c} = 1}}^{\text{K}} \sum\_{\mathbf{i} \in \mathbb{c} = 1}^{\text{C}} \sum\_{\mathbf{i} \in \mathbb{c} \mathbf{d} = 1}^{\text{C}} \sum\_{\mathbf{i} \in \mathbb{G} = 1}^{\text{G}} \mathbf{X}\_{\text{cdG}}^{\text{j}} \tag{9}$$

$$\mathbf{X}\_{\mathbf{c}\_{\mathrm{Im}\,\mathrm{G}}\mathbf{G}}^{\mathrm{i}} = \sum\_{\mathbf{i} \in \mathcal{K}\_{\mathrm{in}}}^{\mathrm{K}} \sum\_{\mathbf{i} \in \mathbf{c} = \mathbf{1}}^{\mathrm{C}} \sum\_{\mathbf{i} \in \mathbf{C} = \mathbf{1}}^{\mathrm{C}} \sum\_{\mathbf{i} \in \mathbf{C} = \mathbf{1}}^{\mathrm{G}} \mathbf{X}\_{\mathrm{c}\_{\mathrm{Im}\,\mathrm{G}}\mathbf{G}}^{\mathrm{i}} + \sum\_{\mathbf{i} \in \mathbf{K}\_{\mathrm{in}} = \mathbf{1}}^{\mathrm{K}} \sum\_{\mathbf{i} \in \mathbf{c} = \mathbf{1}}^{\mathrm{C}} \sum\_{\mathbf{i} \in \mathbf{c} \mathbf{d} = \mathbf{1}}^{\mathrm{Z}} \sum\_{\mathbf{i} \in \mathbf{c} \mathbf{d} = \mathbf{1}}^{\mathrm{G}} \mathbf{X}\_{\mathrm{c} \mathrm{d} \mathbf{G}}^{\mathrm{i}} \tag{10}$$

$$\boldsymbol{\lambda}\_{\mathbf{c}\mathbf{n}\mathbf{c}\mathbf{G}}^{\mathbf{j}} = \sum\_{\substack{\mathbf{j}\in\mathbf{k}\_{\mathrm{out}}\ \mathbf{i}}}^{\mathbf{K}} \sum\_{\mathbf{i}\in\mathbf{c}=\mathbf{I}}^{\mathbf{C}} \sum\_{\mathbf{i}\in\mathbf{G}=\mathbf{I}}^{\mathbf{C}} \lambda\_{\mathbf{c}\mathbf{n}\mathbf{c}\mathbf{G}}^{\mathbf{j}} + \sum\_{\substack{\mathbf{j}\in\mathbf{k}\_{\mathrm{out}}\mathbf{G}=\mathbf{I}}}^{\mathbf{K}} \sum\_{\mathbf{i}\in\mathbf{c}=\mathbf{I}}^{\mathbf{C}} \sum\_{\mathbf{i}\in\mathbf{c}\mathbf{d}=\mathbf{I}}^{\mathbf{C}} \sum\_{\mathbf{i}\in\mathbf{G}=\mathbf{I}}^{\mathbf{G}} \lambda\_{\mathbf{c}\mathbf{d}\mathbf{G}}^{\mathbf{j}} + \sum\_{\substack{\mathbf{j}\in\mathbf{k}\_{\mathrm{out}}\mathbf{G}=\mathbf{I}}}^{\mathbf{K}} \sum\_{\mathbf{i}\in\mathbf{c}=\mathbf{I}}^{\mathbf{C}} \sum\_{\mathbf{i}\in\mathbf{c}\mathbf{d}=\mathbf{I}}^{\mathbf{C}} \lambda\_{\mathbf{c}\mathbf{d}\_{\mathrm{in}}\mathbf{d}\_{\mathrm{in}}\mathbf{G}}^{\mathbf{j}} \tag{11}$$

$$\boldsymbol{\lambda}\_{\text{scG}}^{i} = \sum\_{\text{i} \in \mathcal{K}\_{\text{in}} = 1}^{\text{K}} \sum\_{\text{i} \in \text{c} = 1}^{\text{C}} \sum\_{\text{i} \in \text{G} = 1}^{\text{C}} \lambda\_{\text{c}, \text{c}\_{\text{h}} \text{G}}^{\text{G}} + \sum\_{\text{i} \in \text{K}\_{\text{ll}} = 1}^{\text{K}} \sum\_{\text{i} \in \text{c} = 1}^{\text{C}} \sum\_{\text{i} \in \text{c} \mathbf{d} = 1}^{\text{Z}} \sum\_{\text{i} \in \text{c} \mathbf{d} = 1}^{\text{G}} \lambda\_{\text{c} \text{c} \mathbf{d} \mathbf{G}}^{i} \tag{12}$$

$$\sum\_{\mathbf{i}\in\mathcal{K}\_{\mathrm{ll}}=1}^{\mathrm{K}}\sum\_{\mathbf{i}\in\mathbf{s}=\mathbf{1}}^{\mathrm{S}}\sum\_{\mathbf{i}\in\mathrm{cd}=1}^{\mathrm{Z}}\sum\_{\mathbf{i}\in\mathrm{cd}=1}^{\mathrm{G}}\mathbf{q}\_{\mathrm{G}\mathbf{s}}\ast\lambda^{\mathrm{i}}\_{\mathrm{scd}\mathbf{G}}=\sum\_{\mathbf{i}\in\mathcal{K}\_{\mathrm{ll}}}^{\mathrm{K}}\sum\_{\mathbf{i}\in\mathbf{c}=1}^{\mathrm{C}}\sum\_{\mathbf{i}\in\mathrm{cd}=1}^{\mathrm{Z}}\sum\_{\mathbf{i}\in\mathrm{cd}=1}^{\mathrm{G}}\mathbf{q}\_{\mathrm{G}\mathbf{c}\mathbf{d}}\ast\lambda^{\mathrm{i}}\_{\mathrm{cd}\mathbf{d}\mathbf{G}}\tag{13}$$

$$\sum\_{\mathbf{i}=\mathbf{s}=1}^{\mathbf{S}} \sum\_{\mathbf{i}\in\mathbf{cd}=1}^{\mathbf{Z}} \sum\_{\mathbf{i}\in\mathbf{G}=1}^{\mathbf{G}} \lambda\_{\text{scd}\mathbf{G}}^{\mathbf{i}} \* \mathbf{q}\_{\text{Gs}} \le \mathbf{Q}^{\mathbf{i}} \tag{14}$$

$$\sum\_{\mathbf{i}:\mathbf{j}\in\mathbf{c}=1}^{\mathbf{C}} \sum\_{\mathbf{i}\in\mathbf{c}\mathbf{d}=1}^{\mathbf{Z}} \sum\_{\mathbf{i}\in\mathbf{G}=1}^{\mathbf{G}} \lambda\_{\mathbf{c}\mathbf{d}\mathbf{c}\mathbf{G}}^{\mathbf{j}} \* \mathbf{q}\_{\mathbf{G}\mathbf{c}\mathbf{d}} \le \mathbf{Q}^{\mathbf{j}} \tag{15}$$

$$\sum\_{\mathbf{i}\in\mathcal{G}=1}^{\mathcal{S}} \sum\_{\mathbf{i}\in\mathcal{C}=1}^{\mathcal{C}} \sum\_{\mathbf{i}\in\mathcal{G}=1}^{\mathcal{G}} \left(\mathbf{A}\_{\text{scG}}^{\mathbf{i}} + \mathbf{T}\_{\text{scG}}^{\mathbf{i}} + \mathbf{U} \mathbf{L}\_{\text{scG}}^{\mathbf{i}} + \mathbf{w}\_{\text{i}}\right) \* \mathbf{X}\_{\text{scG}}^{\mathbf{i}} \leq \mathbf{A}\_{\text{cS}G}^{\mathbf{i}} \tag{16}$$

$$\sum\_{\mathbf{i}\in\text{cd}=1}^{Z} \sum\_{\mathbf{i}\in\text{C}=1}^{C} \sum\_{\mathbf{i}\in\text{G}=1}^{G} (\mathbf{A}^{\mathbf{j}}\_{\text{cdcG}} + \mathbf{T}^{\mathbf{j}}\_{\text{cdcG}} + \mathbf{U}\mathbf{L}^{\mathbf{j}}\_{\text{cdcG}} + \mathbf{w}\_{\text{i}}) \* \mathbf{X}^{\mathbf{j}}\_{\text{cdcG}} \leq \mathbf{A}^{\mathbf{i}}\_{\text{cG}} \tag{17}$$

$$\sum\_{\mathbf{j}\in\mathcal{K}\_{\text{out}}=1}^{\mathbf{K}}\sum\_{\mathbf{i}=\mathbf{c}=1}^{\mathbf{C}}\sum\_{\mathbf{i}\in\text{cd}=1}^{\mathbf{Z}}\sum\_{\mathbf{i}\in\text{cd}=1}^{\mathbf{G}}\lambda\_{\text{cdcG}}^{\mathbf{i}} + \lambda\_{\text{CCdG}}^{\mathbf{i}} \le 1 + \mathbf{F}\_{\text{cdc}}\tag{18}$$

$$\sum\_{\mathbf{i}\in\mathcal{K}\_{\text{l}\mathbf{s}=1}}^{\mathbf{K}} \sum\_{\mathbf{i}\in\mathcal{s}=1}^{\mathbf{S}} \sum\_{\mathbf{i}\in\mathcal{C}=1}^{\mathbf{Z}} \sum\_{\mathbf{i}\in\mathcal{G}}^{\mathbf{G}} \lambda\_{\text{scdG}}^{\mathbf{i}} \le \mathbf{y}\_{\mathbf{i}} \tag{19}$$

$$\mathbf{y}\_{\text{i}} \ge \mathbf{E}\_{\text{i}} - \mathbf{A}\_{\text{scG}}^{\text{i}} \tag{20}$$

$$\Theta\_{\rm i} \ge \mathcal{A}\_{\rm scG}^{\rm i} - \mathcal{L}\_{\rm i} \tag{21}$$

$$\left(\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\boldsymbol{\alpha}}}\,\boldsymbol{\chi}\_{\mathrm{s}\mathbf{n}\mathbf{c}\mathbf{G'}}^{\dot{\boldsymbol{\alpha}}}\,\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\mathbf{s}}}\,\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\mathbf{s}}}\,\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\mathbf{s}}}\,\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\mathbf{s}}}\,\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\mathbf{s}}}\,\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\mathbf{s}}}\,\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\mathbf{s}}}\boldsymbol{\chi}\_{\mathrm{cdsG'}}^{\dot{\mathbf{s}}}\,\boldsymbol{\chi}\_{\mathrm{i}}\right)\in\{0,1\},\tag{22}$$

$$\left(\boldsymbol{\chi}\_{\mathrm{cds}\prime}^{i}\,\boldsymbol{\chi}\_{\mathrm{s}\_{\mathrm{n}}\mathrm{s}\_{\mathrm{n}}\prime}^{i}\,\boldsymbol{\chi}\_{\mathrm{cds}\prime}^{i}\,\boldsymbol{\chi}\_{\mathrm{c}\prime}^{i}\,\boldsymbol{\chi}\_{\mathrm{c}\_{\mathrm{n}}\mathrm{c}\_{\mathrm{n}}\prime}^{i}\,\boldsymbol{\chi}\_{\mathrm{c}\prime\prime}^{j}\,\boldsymbol{\chi}\_{\mathrm{c}\_{\mathrm{n}}\mathrm{c}\_{\mathrm{n}}\prime}^{j}\,\boldsymbol{\chi}\_{\mathrm{c}\prime\prime}^{i}\,\boldsymbol{\chi}\_{\mathrm{c}\prime\prime}^{j}\,\boldsymbol{\chi}\_{\mathrm{c}\_{\mathrm{d}}\mathrm{d}\_{\mathrm{n}}\prime}^{j}\,\mathrm{F}\_{\mathrm{c}\mathrm{d}\prime}\right)\in\{0,1\}\tag{23}$$

$$\mathbf{I} = 1, 2, 3, \dots, \mathbf{I}, \mathbf{c} = 1, 2, 3, \dots, \mathbf{C}, \mathbf{G} = 1, 2, 3, \dots, \mathbf{G} \mathbf{j} = 1, 2, 3, \dots, \mathbf{J} \tag{24}$$

$$\text{All variables } \ge 0, \forall \text{ i,s,c and } h \tag{25}$$

#### *2.2. Description of Objective Functions and Constraints*

The first objective function minimizes the cost of transporting vehicles throughout the system. The second objective function minimizes the transport sequence of trucks throughout the system. The objective is utilized to sequence inbound and outbound trucks to minimize the makespan. The third objective function minimizes the amount of carbon dioxide emitted from vehicles throughout the system. Each route between origin and destination has two sections. In the first part, the truck is unloaded, and in the second part, the truck is loaded.

Constraint (1) ensures that the total number of units of type G order product loaded from the supplier (s) in the truck (i) is precisely equal to the number of units of type (h) order product that the customer (c) needs. Constraint (2) ensures that the total number of order types (h) that are loaded from the supplier (s) in the entry truck (i) and move to the customer (c), as well as the total number of order types (h) that are loaded from the crossdocks (cd) in the output truck (j), are loaded and transferred toward the customer (c) and are exactly equal to the sum of the number of h-type order products which are produced by the supplier (s). Constraint (3) ensures that at least one of the incoming trucks (i) loads the supply (s) of type G product and transports all products to the customer (c) at once. Constraint (4) ensures that if the incoming truck (i) moves towards the supplier (s), it selects three modes to exit the supplier. The first case is that the truck moves towards another supplier, the second one moves towards one of the cross-docks (cd), and the last one moves towards the customer (c). Constraint (5) ensures that if the incoming truck (i) moves from the source supplier (Sm) to the destination supplier (Sn), it selects one of three modes to exit the supplier (Sn). The first case is that the truck is moving towards another supplier, the second one is moving towards one of the cross-docks (cd), and the third one is moving towards the customer (c). Constraint (6) ensures that if the outgoing truck (j) moves towards the customer (c), it selects one of two modes to exit the customer (c). The first case is that the truck moves towards another customer and the second moves towards one of the cross-docks (cd). Constraint (7) ensures that if the incoming truck (i) moves from the source customer (cm) to the destination customer (cn), it selects one of two modes to exit the destination customer (cn). The first case is that the truck moves towards another customer and the second moves towards one of the cross-docks (cd). Constraint (8) is similar to constraint (7), and the difference between these two constraints is that in constraint (8), the outgoing truck (j) is considered in the constraint, and the third case, the outgoing truck crosses from a cross-dock to another cross-dock. Constraint (9) ensures that if the incoming truck (i) moves from the supplier (s) to customer (c), it selects one of two modes to exit the customer (c). The first case is that the truck is moving towards another customer, and the second case is that the truck is moving towards the cross-dock (cd). Constraint (10) ensures that the number of units of product that the incoming truck (i) transports from the supplier (s) to one of the cross-docks (cd) and unloads at the cross-dock is precisely equal to the number of units of products in the outgoing truck (j) loaded for the customer (c). Constraints (11) and (12) relate to the capacity of incoming and outgoing trucks and ensure that trucks do not load products beyond their capacity. Constraint (13) ensures that the total time of arrival of the vehicle to the customer, the time elapsed between the manufacturer and the customer, the time of unloading the cargo at the customer's place, and the waiting time of the vehicle should not exceed the specified time of departure for each customer. Constraint (14) guarantees that the total time of arrival of the vehicle to the customer, the time elapsed between the cross-dock and the customer, the time of unloading the cargo at the customer's place, and the waiting time of the vehicle should not exceed the specified time of departure of the vehicle from the customer's location. Constraint (15) ensures that a

customer is assigned to a cross-dock if a truck connects the two. Constraint (16) ensures that inbound truck (i) from the supplier (i) only moves to a cross-dock. Constraints (17) and (18) are soft to calculate the number of early and late trucks to provide customer service in the norm time window.

#### **3. Solution Methodology**

In a small example, the designed mathematical model was first solved in GAMZ software by the є-constraint method [37]. Then, with the development of the example, due to the large volume of model variables over time, it was not possible to solve with GAMZ software. Therefore, it was solved using NSGA-II MOPSO meta-heuristic algorithms.

#### *3.1. Genetic Algorithm*

Charles Darwin's theory, proposed in 1859, occupied a special place in optimization problems [38]. This theory is based on the best evolution, and it can be considered a starting point for evolutionary calculations. Genetic algorithms are random search algorithms whose idea is derived from nature. Genetic algorithms for classical optimization have successfully solved linear, convex, and similar problems, but genetic algorithms are much more efficient for solving discrete and nonlinear problems [39]. In nature, better generations emerge from a combination of better chromosomes. In the meantime, there are sometimes mutations in the chromosomes that may improve the next generation. Genetic algorithms also use this idea to solve problems [40]. Figure 2 show the schematic procedure of the NSGA-II algorithm, and then the details of the proposed genetic algorithm are described.


**Figure 2.** A schematic procedure of the NSGA-II algorithm.

#### 3.1.1. Chromosome Structure

Due to the complexity of the problem and the model, the chromosome consists of three components. These three components include the truck's movement between the origins and destinations and the loading modes of various products. First, how to produce the problem chromosome is described, and then for each example of the problem, in three levels: small, medium, and large, according to the number of trucks, manufacturers, cross-docks, customers, determination and capacity of trucks, time loading, unloading and handling at the cross-dock, distances between origins and destinations, and customer demand, the amount of production of producers are produced as a function of uniform distribution.

Table 3 shows the truck movement modes between the sources. Trucks identify products and loading locations before starting from the cross-dock. In Table 3, for example, four primary sources are considered: two suppliers and two cross-docks. According to Table 3, if the truck moves to the supplier or the cross-dock, it is assigned the number one; otherwise, it is given the number zero. The second column of Table 3, which contains the numbers (0, 0, 0, 1), states that the truck moves only towards supplier 1 and loads the products of supplier 1. Briefly in the table below, S1: supplier 1, S2: supplier 2, CD1: cross-dock 1, and CD2: cross-dock 2.

**Table 3.** Truck motion matrix at the origins.


Table 4 show the trucks' destinations after passing through the sources. For example, in this table, four main destinations are considered, including two customers and two cross-docks, and trucks can transport different modes of products to their destinations. According to Table 4, if the truck moves to the customer or the intersection dock, it is assigned the number one; otherwise, it is given the number zero. The second column of Table 4, which contains the numbers (0, 0, 0, 1), states that the truck moves only to customer one and unloads the customer's products. Briefly in Table 4, C1: customer 1 and: C2: customer 2 and CD1: cross-dock 1 and CD2: cross-dock 2.

**Table 4.** Truck movement matrix at destinations.


For example, four different types of products are used. These four types of products are delivered to customers by production suppliers and trucks. Table 5 show the different modes of loading products from suppliers in trucks. A truck can load other methods of

products in its truck. According to Table 5, if the truck loads one of the product ordering modes, it will be number one; otherwise, it will be zero. The second column of Table 5, which contains the numbers (0, 0, 0, 1), states that the truck only loads product A.


**Table 5.** Matrix of product ordering modes.

#### 3.1.2. The Structure of the Cross Method

In the cross method, the production of new children is from the elite parents of the current generation. New children should inherit the main characteristics of parents and be more privileged than them. How children are produced depends on the structure of the chromosome. Using the tournament method as a selection mechanism, two chromosomes are selected as parents. The single-point cross is then performed separately for each part of the chromosome using row (horizontal) and columnar (vertical), randomly selected and described as follows. Table 6 show two samples of the parent chromosome to produce new offspring.

**Table 6.** Two hypothetical parent chromosomes.


**A—Array cross operator:** In this case, we produce a random number in the range of the number of random devices. The generated random number indicates the array location of the cross in the parent chromosomes. Then, we replace the corresponding genes of the parents from the desired array, and new offspring are produced. For example, array 19 is selected at the cross of the array site 19 in the parent chromosomes, and then the corresponding genes of the parents are shifted from the array site 19 times in Table 7, and new offspring are produced.

**B—Horizontal cross operator**: In this case, we generate a random number in the range of the number of rows in the matrix at random. The generated random number indicates the location of the cross in the parent chromosomes. The selected line then replaces the corresponding genes of the parents, and new offspring are produced. For example, if the second row is selected, the junction of the second row is selected on the parent chromosomes, and then the corresponding parent genes are switched from the second row, as shown in Table 8, and new offspring are produced.


**Table 7.** Children of the cross-array operator.

**Table 8.** Children of linear (horizontal) cross operator.


**C—Vertical cross operator**: In this case, we generate a random number in the range of matrix columns. The generated random number indicates the location of the cross in the parent chromosomes. The selected column then replaces the corresponding genes of the parents, and new offspring are produced. For example, the fourth column is selected, the intersection is performed from the location of the fourth column in the parent chromosomes, and then the corresponding parent genes are switched from the location of the fourth column according to Table 9, and new offspring are produced.

**Table 9.** Children of the linear (vertical) cross operator.


#### 3.1.3. Structure of the Mutation Method

The mutation operator is mainly used to bring about transformation diversity and prevent divergence in the population. Due to the chromosome structure, the mutation operator, similar to the crossover operator, is performed as one of the array methods: row (horizontal) and columnar (vertical), which are randomly selected by the algorithm.

**A—Array mutation operator:** In this case, we produce a random number in the range of random arrays. The generated random number indicates the location of the mutation on the parent chromosome. Then, if the number in the array location is zero, it becomes one, and if it is one, it becomes zero. For example, array 19 is selected in parent one and the mutation is performed from array location 19 in the parent chromosomes. Then the corresponding parent gene from array location 19 is replaced by Table 10, and a new chromosome is produced.

**Table 10.** New chromosome derived from the array mutation operator.


**B—Horizontal mutation operator:** In this case, we generate a random number in the range of the number of rows in the matrix. The generated random number indicates the location of the mutation on the parent chromosome. The corresponding parent genes are then selected from the row location, which has values of zero and one, and we move the values of the two, and a new child is produced. For example, the second row is selected in the first parent and the mutation is performed from the second-row location in the parent chromosome. Then the corresponding parent genes are moved from the second-row location in Table 11, and a new chromosome is produced.

**Table 11.** New chromosome derived from the horizontal mutation operator.


**C—Vertical mutation operator:** In this case, we generate a random number in the range of the number of columns in the matrix. The generated random number indicates the location of the mutation on the parent chromosome. The corresponding parent genes are then selected from the column location, which has values of zero and one, and we move the values of the two, and a new child is produced. For example, the seventh column is selected in the first parent and the mutation is performed from the seventh-column location in the parent chromosome. Then the corresponding parent genes are moved from the seventh-column location in Table 12, and a new chromosome is produced.



#### *3.2. Particle Swarm Optimization Algorithm—PSO*

Kennedy and Eberhart (1995) presented the PSO algorithm in two papers for optimization problems whose continuous nature indicates the answers [41]. The population of answers is called a *swarm*. Each answer is like a *bird* in a group of birds and is called a *particle*. It is similar to the chromosome in the genetic algorithm. All particles have a *fitness value* calculated using the fitness function, and the particle fitness function must be optimized. The velocity vector of that particle determines the direction of motion of a particle. The PSO algorithm starts with random solutions (particles) and then seeks the optimal solution by synchronizing the particles in each iteration. If the decision variables, also their particles, are zero and one, each particle's velocity and position vectors in each iteration of the algorithm are calculated according to Equations (26)–(29).

$$\mathbf{V\_{it}} = \mathbf{w} . \mathbf{V\_{it-1}} + \mathbf{c\_1} . \mathbf{r\_1} . (\mathbf{pBest\_i} - \mathbf{x\_{it}}) + \mathbf{c\_2} . \mathbf{r\_2} . (\mathbf{nBest\_i} - \mathbf{x\_{it}}) \tag{26}$$

$$-\mathbf{V}\_{\text{max}} \le \mathbf{V}\_{\text{it}} \le \mathbf{V}\_{\text{max}} \tag{27}$$

$$\mathbf{s}\_{\mathbf{i}} = \mathbf{1}/\mathbf{1} + \mathbf{e}^{\mathbf{V}\_{\mathbf{i}\mathbf{t}}} \tag{28}$$

$$\mathbf{x}\_{\text{if}} = \begin{cases} \ 1 \ \not\le \mathbf{s}\_{\text{i}} \\ \ 0 \text{ otherwise} \end{cases} \tag{29}$$

According to the relation of 19, new velocity vectors of each particle based on the previous velocity of the particle itself (Vit−1), the best position the particle has ever reached (pBest<sup>i</sup> ) and the position of the best particle in the neighborhood of the particles obtained so far (nBest<sup>i</sup> ), are calculated. If the neighborhood of each particle contains all the particles in the group, then nBest<sup>i</sup> indicates the position of the best particle in the group, denoted by gBest. r<sup>1</sup> and r<sup>2</sup> are two random numbers with a uniform distribution between [0, 1] generated independently of each other. c<sup>1</sup> and c<sup>2</sup> are referred to as learning coefficients, they control the effect of pBest and nBest during the search process. *w* represents the weight coefficient of inertia. The value limits the particle velocity vector (Vmax). Vmax is a constraint that controls the global search capability of a particle group. Using the relation of 20 vectors, the velocity of each particle is converted to the change probability vector. In the above relation, s<sup>i</sup> indicates the probability that xit is equal to 1. Then, using the relation of 22 vectors, the position of each particle is updated. In the above relation, ρ is random with a uniform distribution between zero and one. The pseudo-code of group PSO is shown in Table 13.

**Table 13.** PSO pseudo-code.

```
For each particle
Initialize particle
End For
Do
For each particle
Calculate fitness value of the particle fp
/*updating particle's best fitness value so far)*/
If fp is better than pBest
set current value as the new pBest
End For
/*updating population's best fitness value so far)*/
Set gBest to the best fitness value of all particles
For each particle
Calculate particle velocity according to equation
Update particle position according to equation
End For While maximum iterations OR minimum error criteria is not attained
```
#### *3.3. Validation of the Designed Model*

Although many articles studied the issue of truck scheduling and transportation time minimization at the intersection dock, the issue of cost minimization (transportation, loading, unloading, and relocation) throughout the supply chain, considering the relationships and assumptions, was not previously studied. This article represents a new beginning for future work in this field. Therefore, since it is impossible to compare the results with other existing articles, a numerical example is presented in the next section, and the answers to the problem are compared using genetic algorithms and particle swarm.

Method of Generating Random Problems

Sample production problems are randomly generated in small, medium, and large groups. From each dimension of 7 sample problems, 21 sample problems are described according to Table 14 of the production. Then, the computational results will be presented, and finally, by repeating the algorithm implementations and using the uniform distribution function, the problem parameters are generated, and new numbers are generated in each sample. Additionally, the sensitivity analysis of the results is analyzed using the number of trucks, and in the end, the two algorithms are compared using different performance criteria. The solution of the model was obtained using MATLAB (R2019a) software. The problem was that the trucks would move from suppliers to customers or cross-dock and unload the loaded products after loading the products. To better understand how the two algorithms work, the problem-solving details of Example 6 are shown in a small size, which is almost the most straightforward example problem.


**Table 14.** Fixed parameters of the problems.

Several parameter values are generated using the uniform distribution function in this problem. Parameters that are generated randomly include the capacity of the trucks, the amount of customer demand, the amount of production of the manufacturers, the number of products, the fuel consumption rate, and the distance traveled by the trucks between the origins and the destinations. Sample parameter values are created and given in the relevant tables for each problem. Table 15 list the minimum and maximum values that the parameters can be assigned.

**Table 15.** Random parameters of the problem.


Two meta-heuristic algorithms for solving multi-objective problems, including the NSGA-II faulty sorting genetic algorithm and the MOPSO multi-objective particle swarm algorithm, were used to solve the proposed model. The Taguchi method was used to adjust the parameters of the algorithms. The advantage of the Taguchi method over other experimental design methods is that in addition to cost, obtaining optimal levels of parameters can be achieved in less time. One of the most important steps in this method is to select an orthogonal array that estimates the effects of factors on the mean response and variation. In this paper, the most suitable three-level test design is identified, and according to the Taguchi standard orthogonal arrays, the L9 array is selected as the appropriate test design to parameterize the proposed algorithms. A statistical measure of performance is considered the *S*/*N* (*signal to noise*) ratio to adjust the optimal parameters, which includes the average and changes. This ratio is desirable. The considered response variable is the average of the four standard indices MID (mean ideal distance), MD (maximum spread or diversity), spacing, and NPS (number of pareto solution) for multi-objective meta-heuristic algorithms. Since this response variable is less, its corresponding *S*/*N* ratio is considered as Equation (30). The proposed meta-heuristic algorithms were implemented for each Taguchi experiment, and then the *S*/*N* ratios were calculated using Minitab 18 software. The optimal values of the parameters of each algorithm are shown in Table 16.

$$S/NRatio = -10 \log(\frac{sum(y^2)}{n}) \tag{30}$$

**Table 16.** Optimal values of parameters in algorithms.


#### **4. Results**

Sample problem 6 was solved at the small size level with the fixed parameters listed in Table 4 and the random parameters listed in Table 6 with the four algorithms NSGA-II and MOPSO. Based on this, the sequence of truck transportation operations in sample case 6, for all four algorithms, is presented as the output analysis of the variables according to the following tables. Table 17 show the transport sequences of the first truck in sample problem 6 by solving the NSGA-II algorithm, and Table 18 show the transport sequences of the first truck in sample problem 6 using the MOPSO solution method.

**Table 17.** Truck transport sequence 1 with NSGA-II algorithm for sample problem 1.



**Table 17.** *Cont.*

**Table 18.** Truck transport sequence 1 with MOPSO algorithm for sample problem 1 (small level).


#### *4.1. Comparison of NSGA-II and MOPSO Algorithms*

Tables 19 and 20 show the values related to the criteria of several answers, time to solve problems, distance from the ideal point, distance, and maximum coverage to compare NSGA-II and MOPSO algorithms.

**Execution time**—One of the crucial criteria for measuring the performance quality of an algorithm is its execution time, which in some articles is also referred to as execution speed. This criterion becomes essential when the dimensions and complexity of the problem increase. The execution time of these four methods in the three levels of small, medium, and large sample problems are given in Tables 19 and 20 and Figure 3. In general, the NSGA-II algorithm is more acceptable in terms of average runtime at both medium and large sample problem levels.


**Table 19.** Criteria values of the NSGA-II algorithm.

**Table 20.** Criteria values of the MOPSO algorithm.


**Figure 3.** Comparison between the average execution time of NSGA-II and MOPSO.

#### 4.1.1. Distance from the Ideal Point

The MID index is used to calculate the mean distance of Pareto solutions from the origin's so-called ideal point. Considering the MID, it is clear that the lower the value of this index, the higher the efficiency of this algorithm. The ideal point distance criterion, which measures the optimal close Pareto proximity measure, was calculated for 21 sample problems for the NSGA-II and MOPSO algorithms. In these 21 problems, the average values of the NSGA-II algorithm in small and medium problems were less than for the MOPSO algorithm. The DM index shows the variety of Pareto solutions. The higher the DM value is, the better the algorithm performs.

#### 4.1.2. Spacing

The spacing criterion for each sample problem was calculated for the NSGA-II MOPSO algorithms, and the comparison of the mean of these two values for the sample problems was performed at three levels to determine the most desirable algorithm. The spacing criterion in the MOPSO algorithm was smaller than the NSGA-II algorithm and had better performance.

#### 4.1.3. The Most Expansion

The maximum expansion criterion for the 21 sample problems was calculated for the NSGA-II and MOPSO algorithms. The criterion measures the maximum expansion density of Pareto archive solutions, and a simple method was used to calculate it, i.e., without considering the intermediate solutions, only the boundary solutions were used to plot the extent of Pareto archive solutions. In this criterion, the average of the calculations shows that the NSGA-II algorithm offers more comprehensive answers than the MOPSO algorithm.

#### **5. Conclusions and Suggestions**

The problem studied in this article was used to provide a model for the optimal sequence of trucks and the cost of operations within the supply chain concerning three objectives (minimizing the cost of transportation, minimizing the sequence of transporting trucks, and minimizing carbon dioxide emissions) [42–45]. The mathematical model approach is complicated due to the high number of variables and limitations related to the number of trucks sending and receiving and the number of products required to solve the problem. When increasing these variables, the dimensions of the problem and

the solution time increase exponentially [46,47]. Therefore, two meta-heuristic algorithms NSGA-II and MOPSO, were used to solve the model. The results of the calculations were compared using four criteria. On average, the NSGA-II algorithm is more acceptable in runtime for 21 sample problems at the three small, medium, and large levels. In terms of distance from the ideal point, the NSGA-II algorithm is better. In the spacing criterion, which measures the density of Pareto archive solutions, the MOPSO algorithm scores better than the NSGA-II algorithm. The NSGA-II algorithm offers broader answers than the MOPSO algorithm at the maximum expansion criterion. Because research in this area is in its infancy, there is much opportunity for future research. Several areas in which research can be expanded are listed below.

Generalization of the proposed mathematical model for modelling in the presence of cross-dock systems with many entrances and exit doors.

It was assumed that temporary storage has unlimited capacity, while temporary storage is usually limited in practice. This real-world assumption can be added to the model.

Generalization of the proposed mathematical model by considering the multiperiod mode.

Generalization of the proposed mathematical model of the problem by considering uncertain parameters (for example, considering fuzzy demand for customers) given the complex nature of docking systems.

**Author Contributions:** Conceptualization, I.M.-K. and N.S.; methodology, S.G. and N.S.; software, F.Y.; validation, I.M.-K.; formal analysis, S.G. and F.Y.; investigation, I.M.-K. and R.C.; resources, S.G. ˇ and R.C.; writing—original draft preparation, S.G. and N.S.; writing—review and editing, I.M.-K.; ˇ visualization, S.G. and R.C.; supervision, I.M.-K. All authors have read and agreed to the published ˇ 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 of this study are available from the authors upon request.

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

#### **References**


#### *Article* **Sustainable Rural Electrification Project Management: An Analysis of Three Case Studies**

**Laura Del-Río-Carazo \*, Emiliano Acquila-Natale, Santiago Iglesias-Pradas and Ángel Hernández-García**

Departamento de Ingeniería de Organización, Administración de Empresas y Estadística, ETSI de Telecomunicación, Universidad Politécnica de Madrid, Av. Complutense 30, 28040 Madrid, Spain; emiliano.acquila@upm.es (E.A.-N.); s.iglesias@upm.es (S.I.-P.); angel.hernandez@upm.es (Á.H.-G.) **\*** Correspondence: laura.delrio@upm.es

**Abstract:** Universal access to energy is a global challenge for sustainable development that requires granting last-mile access to energy services to rural and isolated communities. However, achieving access is not sufficient: it must be done affordably, reliably and with an adequate quality. Universal access to energy goes beyond the mere selection of a technical solution or infrastructure; it demands being able to design management models for projects aiming to guarantee that households may access energy services in a sustainable way. This study analyzes the main elements (i.e., governance, technological and business models) of management models in universal access to energy projects and their impact on the different dimensions of sustainability (i.e., social, environmental, and economic). The study then presents three case studies of rural electrification projects having different configurations of the management model, with special focus on the differences in the business model, and it analyzes their outcomes from a sustainability perspective. The analysis of the three case studies suggests that the choice of the business model is key to ensuring sustainability, with fee-for-service models giving the best results. The analysis also highlights the importance of collaboration and involvement of the communities in projects engaging multiple agents with different roles.

**Keywords:** management model; rural electrification; sustainability; governance; technology; business model

#### **1. Introduction**

Universal access to energy is essential to foster and guarantee the progress of lessfavored and isolated rural communities [1]. Energy poverty has become a world problem and a threat to the environment, public health, safety, and the world economy [2]. According to the United Nations, the 17 Sustainable Development Goals (SDGs) cannot be achieved without sustainable energy, which is why high priority was given to affordable and nonpolluting energy in SDG 7 [3].

The world population with access to energy has increased from 83 percent in 2010 to 90.2 percent in 2021. However, and despite the overall improvement in access to energy in the last years, rural areas in developing countries still show access levels that are far from the objectives declared in the 2030 Agenda [4]. In these cases, the main challenge still remains to address the "last mile" problem; that is, to reach to the most vulnerable and isolated communities: 759 million people do not have access to energy yet, of which most of them live in rural areas, representing a total of 85 percent of the global deficit in access to energy services [5].

The SDG 7 states that universal access to energy must be achieved by using renewable energies, improving the efficiency of energy systems, increasing international collaboration, extending the infrastructure and developing technologies [3]. Therefore, guaranteeing access to energy and energy services is not enough: it is imperative to provide reliable and affordable energy of sufficient quality [6]. Additionally, it is necessary to increase the production and delivery of energy services in a way that is compatible with the SDGs, a

**Citation:** Del-Río-Carazo, L.; Acquila-Natale, E.; Iglesias-Pradas, S.; Hernández-García, Á. Sustainable Rural Electrification Project Management: An Analysis of Three Case Studies. *Energies* **2022**, *15*, 1203. https://doi.org/10.3390/en15031203

Academic Editors: Bernard Zi˛ebicki and Edyta Bieli ´nska-Dusza

Received: 30 December 2021 Accepted: 3 February 2022 Published: 7 February 2022

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

**Copyright:** © 2022 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/).

reason why the design of such provision must be made under the principles of universality, diversity and sustainability [3,7].

The origin of the "last mile" challenge is that it makes it very expensive, often prohibitive and not affordable, to provide remote rural communities with energy because the demand for energy from these communities is relatively low. Therefore, energy providers (ESCOs) discard those investments due to the impossibility to make them profitable without increasing their tariffs or having them subsidized by the state; instead, they focus on urban areas, densely populated and with higher average income population [8]. However, for universal access to energy to happen, it is necessary not only to provide access to energy, but also that the energy services are accessible: even when households may access a network or purchase individual energy generation systems, access rates may remain low due to the high initial costs, interconnection fees, wiring, and electric appliances. Hence, universal access to energy requires the development of business models that make access affordable, be it by means of subsidies or payment methods that are compatible with the economy of the households. Long-term sustainability of universal access to energy must also consider the local context and the different agents involved; for this reason, the underlying business models must be tailored to the characteristics of the population and be flexible enough to be consistent with local needs. Therefore, there is no single business model in universal access to energy, but rather different possibilities and characteristics that may be particular to each project. The big challenge is then being able to provide tailored solutions that do not slow down the projects aiming to achieve universal access to energy [5].

Because of the abovementioned barriers, project managers not only need to focus on using high-quality technology that may address local needs but must also ensure that the project's business model is consistent with cultural values, awareness, and the agents involved in the planning and execution of the project [9]. In other words, a correct management of the project will ensure its sustainability through the provision of a reliable, high-quality and affordable access to energy in rural areas [10].

To further clarify what constitutes a correct management in sustainable rural electrification projects, the main objective of this study is the elaboration of a comprehensive management model to analyze how different configurations of the model may affect sustainability. This is one of the main contributions of this study, as existing literature on rural electrification rarely adopts a managerial view of the projects. The analysis involves the examination of three different studies of real projects of rural electrification led by private entities. The remainder of the study is structured as follows. Section 2 lays out the possible configurations of the management model, with three different components (i.e., governance model, technological model, and business model), after observation of previous academic research on rural electrification projects. Furthermore, it links the impact of the different components of the management model with the three pillars of sustainability (social, environmental, and economic). Section 3 analyzes in further detail three study cases of rural electrification carried out by a private entity. Section 4 summarizes and discusses the main findings of the research. Finally, Section 5 presents the conclusions and limitations of the study.

#### **2. Conceptual Framework**

#### *2.1. Management Modeling of Rural Electrification Projects*

The configuration of a management model for rural electrification projects is complex and multifaceted. Real cases of rural electrification projects cover primarily the technological elements, project funding and the roles of the different agents involved [11], even though the terminology varies from one case to another [12]. From a wider perspective, the configuration of the management model requires the consideration of aspects relative to ownership, operation and maintenance, selection of technologies, viability and feasibility, funding, or stakeholders' roles, among others [11–14]. This study synthetizes these elements in three fundamental components: governance model, technological model, and business model (Figure 1). While governance may often be equated to the broader

idea of project management, in rural electrification projects its scope extends to ownership of the initiative or project, the operation of energy provision and the relationships between actors/stakeholders [13,14]. A similar clarification is required in the case of the business model; while some may limit this component to the funding of the project, the business model covers more aspects, such as system acquisition, maintenance, or payment methods [12]. project management, in rural electrification projects its scope extends to ownership of the initiative or project, the operation of energy provision and the relationships between actors/stakeholders [13,14]. A similar clarification is required in the case of the business model; while some may limit this component to the funding of the project, the business model covers more aspects, such as system acquisition, maintenance, or payment methods [12].

the configuration of the management model requires the consideration of aspects relative to ownership, operation and maintenance, selection of technologies, viability and feasibility, funding, or stakeholders' roles, among others [11–14]. This study synthetizes these elements in three fundamental components: governance model, technological model, and business model (Figure 1). While governance may often be equated to the broader idea of

*Energies* **2022**, *15*, x FOR PEER REVIEW 3 of 21

**Figure 1.** Management model for rural electrification projects. **Figure 1.** Management model for rural electrification projects.

#### 2.1.1. Governance Model 2.1.1. Governance Model

Based on classifications of governance models of rural electrification projects found in prior research, we can highlight three main elements: ownership of the project, operation of the service and the relationships between agents [13,15,16]. Regarding ownership, the projects may be *state owned, privately owned, municipal property, community property,* and *cooperative property* [15]. *State ownership* refers to projects planned by the state and generally managed by existing electric companies [17]; in contrast, *privately owned* projects are those in which the systems and/or network that grant access to energy belong to the end users or a private company [18–20]. Ownership falls in *municipality* or local authority, or in the *community*, when the energy systems or network have been donated by an entity (public or private) but the control and management of the electric service has been transferred to the local authorities or to the community, respectively [21]. A final option is that of *cooperative* ownership of the project, in which a group of individuals oversees, imple-Based on classifications of governance models of rural electrification projects found in prior research, we can highlight three main elements: ownership of the project, operation of the service and the relationships between agents [13,15,16]. Regarding ownership, the projects may be *state owned*, *privately owned*, *municipal property*, *community property*, and *cooperative property* [15]. *State ownership* refers to projects planned by the state and generally managed by existing electric companies [17]; in contrast, *privately owned* projects are those in which the systems and/or network that grant access to energy belong to the end users or a private company [18–20]. Ownership falls in *municipality* or local authority, or in the *community*, when the energy systems or network have been donated by an entity (public or private) but the control and management of the electric service has been transferred to the local authorities or to the community, respectively [21]. A final option is that of *cooperative* ownership of the project, in which a group of individuals oversees, implements and collectively owns the energy provision service to satisfy their energy access needs [22].

ments and collectively owns the energy provision service to satisfy their energy access needs [22]. Regardless of ownership, the owner of the project is not necessarily responsible for the operation and maintenance of the service; from the perspective of operation and economic management of the service, there are two different options: either the service is operated *by the owner* or *by a third party* [13,23]. When projects are privately owned and Regardless of ownership, the owner of the project is not necessarily responsible for the operation and maintenance of the service; from the perspective of operation and economic management of the service, there are two different options: either the service is operated *by the owner* or *by a third party* [13,23]. When projects are privately owned and operated by the owner, owners are responsible for the operation of the service, even though they may delegate preventive maintenance to end users and corrective maintenance to expert staff [24]; in *state-owned* projects, operation and preventive and corrective maintenance are generally subcontracted to electric companies; in the case of *municipal* or *community* ownership, citizens generally oversee operation and maintenance; finally, when service operation is overseen by a *third party*, the responsibility for operation and maintenance of the service is transferred to an entity that works autonomously from the owner; the staff of

the third party has (or receives) the training and qualification to perform preventive and corrective maintenance, and the entity receives economic payments from periodic (weekly, monthly, yearly) service fees [25,26].

The third element of the governance model involves the relationships between the agents participating in the project. Because of the need to mobilize resources to broaden the coverage of the energy service, there are several actors involved in the energy service provision in rural electrification projects in remote areas [11]. From the review of prior experiences, three different forms of participation may be found: non-existent (the owner of the project executes it autonomously and independently), *public–private partnerships* (*PPPs*), and *pro-poor public–private partnerships* (*5P*) [27]. A PPP consists of agreements between the public and private sector, where resources from both are combined to provide the service; in a PPP, the public sector lays out the regulatory framework and protects the end users' rights, while the private participants contribute to the project with resources, knowledge and effective operation. The rise of initiatives oriented to improve universal access to basic services in developing countries has led to the emergence of a new type of relationship (5P), where not only states and private companies collaborate, but also incorporate other agents to the project, such as microfinancing institutions, banks or credit unions, universities and nonprofit, not-for-profit, and non-government organizations (NGOs) [17,21,27]. The main objective of 5P is the provision of services to poor communities, often ignored in PPP due to business risks. In 5P, poor people are not only users but also project partners; private companies may develop corporate social responsibility initiatives, public bodies may fulfill their obligation to provide universal access to basic services, and the communities may collaborate in broadening such access.

#### 2.1.2. Technological Model

Rural electrification projects may rely on different technologies, depending on how dispersed are the potential users or their proximity to an existing electric network [28]. When the citizens and the communities live in areas that are close to an existing grid, and the orography is not a problem, *on-grid* systems (i.e., extending the existing network) comes up as an optimal solution [29]. On-grid systems provide enough electric power to connect several lighting devices, such as lamps, and other electric appliances [30]. In these cases, existing electric companies take on the initial infrastructure investment and operate the service, including the maintenance of the different systems [31]. When expanding the grid is not possible due to the fact of long distances between the users and the existing network, *off-grid* solutions are necessary; off-grid systems are independent systems, able to self-generate and store energy [32,33]. Among off-grid systems, there are different options depending on the dispersion of users. If the dispersion is low, *microgrids* generate energy at a local level and supply electricity to a reduced number of users, who are interconnected through a shared distributed system [19,34,35]; when the dispersion is high, *standalone systems* (e.g., *solar home-systems*) are the best solution, especially considering that the cost of grid connection increases with distance from grid [18,36–38].

Energy generation technologies vary depending on the chosen system. Based on the source of energy, they can be categorized as *conventional*, *non-conventional* (*renewable*), and *hybrid*. Conventional technologies use fossil fuel [38,39], whereas non-conventional technologies refer to those that use exclusively renewable energy sources [17,21,38,40,41] and hybrid technologies combine both [20,28,38,42]. As the availability of renewable sources of energy may be variable, non-conventional technologies make it necessary to use energy storage systems, such as batteries; hybrid systems generally try to avoid the use of batteries by incorporating diesel generators and reducing the capacity of storage systems [43].

#### 2.1.3. Business Model

Ensuring that the business model of service provision is successful entails the collaboration of different agents (e.g., the state, donors, NGOs, or private companies) [44], and different alternatives can be offered to beneficiaries [45]. For instance, ownership of the energy generation system may fall on the end user (customer) or in the owner of the project (if state or privatelyowned). When customers own the system, they may have made a payup-front purchase or have their purchase financed in different payment installments using microcredits [46]. They are also responsible for the maintenance of the system, bearing the costs associated with maintenance and repairs; to avoid exceptional high expenses, another option is to make periodic payments (a maintenance fee), regardless of whether the systems needs repairing or not [45].

If the customer does not own the system, an *energy-as-a-service* (EaaS) model may be used. EaaS involves that the customer makes a periodic payment for energy provision that includes maintenance and substitution of the system when it reaches end-of-life or end-ofservice-life [18,47]. EaaS improves the affordability of energy access because the energy provider maintains ownership of the system and has a duty to keep it operational [45]. Additionally, in EaaS there are two possible payment methods: *prepaid* and *postpaid* [48]; the former allows customers to pay only for the energy used, avoiding expenses when they do not have the money to make the payment (they would not have access to energy in the meanwhile), whereas the latter allows consumers to have access to energy even when they cannot make the payment occasionally (settling the debt in later installments) [49].

Finally, financing of all payments related to access to energy in rural electrification projects generally are subsidized either by the state or other public entities, or by private companies that bear a fraction of the cost of the equipment or the periodic fees; the subsidy may be equal for all customers or proportional to their income.

#### *2.2. Rural Electrification Management Models and Sustainability*

Achieving sustainability in rural electrification projects is one of the main challenges for developing countries [50,51]. The use of technologies that are tailored to rural areas widens access to energy [7], but selecting the right technology might not be enough, as it is necessary to consider all aspects related to sustainability throughout all the phases of the project, both at a technical and management level [52,53]. For example, choosing the right business model, achieving effective knowledge and technology transfer and including the beneficiaries in the decision making process facilitate the sustainability of the project [18,54,55].

Sustainable development refers to meeting "the needs and aspirations of the present generation without destroying the resources needed for future generations to meet their needs" [55]. A common view of sustainability is that of the three pillars, which differentiates between social, environmental, and economic sustainability [56]. Social sustainability refers to the ability of social identities, relationships, and values to continue in the future [57], which requires that social cohesion is sustained and that individual needs are met [58]. Environmental sustainability represents the ability to improve human well-being, while protecting the sources of raw materials and guaranteeing the integrity of ecosystems [59]. Economic sustainability refers to the ability of a project to sustain a given level of economic production indefinitely [60,61]. To better guide the decisions of rural electrification projects, the following subsections analyze the effect of the different possible configurations of the management model presented in Section 2.1 in each of the pillars of sustainability, by taking different frameworks defined to assess sustainability, either with a project management approach [62,63] or a rural electrification view [64–67].

#### 2.2.1. Social Sustainability

The social dimension focuses on the improvement in the quality of life of the population, and it is associated with the fulfilment of their basic needs and their active participation in society. It is essential that the beneficiaries are involved in the development of the project, from its design to its implementation, so that the projects are adapted to the particular characteristics of the area and it is easier for the communities to appropriate the technology, thus ensuring their awareness and actual use of the solution [54,68].

From the management model, all three components (i.e., governance, technology, and business models) have a direct effect on different aspects related to social sustainability found in prior literature: funding, energy equity, availability of services, availability of support, profile of operation staff, degree of local participation, project duration, and default rate, or affordability.

Funding helps assess whether the beneficiaries can take care of the periodic expenses associated with access to energy, whether they should own the systems or whether the cost of the service is subsidized to facilitate affordable access to the least favored [68]. These decisions are affected by the governance model, through leadership and relationships between actors, and by the business model, when establishing the modality of acquisition.

Energy equity refers to the requirement that the project must focus on poorest communities and with higher barriers in access to energy, regardless of their social condition or the dispersion of such communities [63], which is highly related to the governance model, depending on the involvement of the community in decision making, and to the technology model, in system selection.

The service also must meet a minimum level of service offer and reliability, in what constitutes the availability of services. Such level varies depending on the characteristics of the community and must be analyzed as part of the project planning, considering the specific needs of each potential user. Availability of services may be assessed based on duration of energy shortages or blackouts as well as users' perceptions [64,69,70], and it is highly associated with the technological model. Additionally, beneficiaries may be provided with the necessary resources to repair the systems or replace them if they are beyond repair; in other words, they need to have available support, which must be both accessible and affordable to users [20,69]. Availability of support is therefore associated with the governance model, by defining who is responsible for service operation, and the business model, by the defined maintenance fees.

Regarding the profile of operating staff, the project must include personnel that has the required skills to operate and maintain the project, so that effective knowledge and technology transfer to the beneficiaries is possible [64]. Additionally, the project should include and empower women in such operation and maintenance tasks [68]. These aspects are affected by the governance model, depending on how community participation is incorporated into the project, which also affects the number of local population involved in the design and operation of the project [62,69].

Choices in ownership (i.e., governance model) and service modeling (i.e., business model) have an effect on the duration of the project. Rural electrification projects require an adoption and adaptation period among end users, as well as additional maturing time that facilitates analysis and fine tuning, which is why short and fast implementation projects with no subsequent monitoring should be avoided [68].

Finally, it must be possible to assess whether access to energy is really affordable for the beneficiaries or, on the contrary, it is necessary to revise the business model to change system acquisition or service provision to prevent abandonment by the users, measured as default rate [20,68,69].

#### 2.2.2. Environmental Sustainability

Environmental sustainability considers the different aspects that impact natural resources and local ecosystems of the areas where the project will be implemented [6]. Decisions in the technology component of the management model are the ones with the most direct effect on environmental sustainability, which may be assessed through CO<sup>2</sup> emissions, impact on the ecosystem, replacement of other polluting sources or compatibility with future grid services.

CO<sup>2</sup> emissions facilitate quantitative comparison and selection of the most appropriate technology to achieve lower levels of pollution. It is necessary that the project management considers which is the right choice among the possible energy sources and systems [61]. Ecosystem impact is based on a qualitative and quantitative analysis of how respectful with the environment is the service infrastructure of the project, both at visual and acoustic levels. At the initiation phase of the project, the decision of the technology must consider and analyze the characteristics of the area and the population, as well as the dispersion of the community [62,64]. When developing rural electrification projects, current use of energy sources must be considered, so that the system to be introduced provides the highest reduction in the use of conventional sources and maximizes the use of non-polluting natural resources in the area [68]. Finally, the solution must guarantee the compatibility of the service with future developments and avoid the generation of technological waste due to obsolescence or abandonment of the system [68]; that is, the technology should be flexible and adaptable.

#### 2.2.3. Economic Sustainability

The top priority of the project in the economic sustainability dimension of a rural electrification project should be to have the ability to autonomously bear the costs associated with the provision of the service in the future. Achieving economic sustainability requires that the costs of the electric service are high enough to maintain the systems operational in the long term but also low enough to guarantee affordability. Decisions of the business model directly affect economic sustainability and, in general, different business models are piloted, helping fine-tune the model until an optimal balance is achieved [5]. As seen in Section 2.1.3, decisions in the business model include system ownership and acquisition, maintenance mode, and payment methods. In developing countries, system acquisition is often prohibitive to the residents in rural areas. Therefore, ensuring economic sustainability involves the adoption of non-conventional business models where the end users only assume a percentage of the total cost.

Broadly, economic sustainability may be assessed by observing the profitability of the project, operation and maintenance costs, and capital and installation costs [68]. That is, the project must (1) be profitable enough to facilitate scaling up, which is why the business model must consider all the associated costs when fees and prices are set-up; (2) observe the economic resources necessary to ensure continuance of the project (higher economic resources entail higher fees and costs for beneficiaries); (3) consider the initial investment to deploy the technical system for energy provision (higher investment in more difficult implementations).

Table 1 shows a summary of the potential impact of decisions in the configuration of the business model on the three pillars of sustainability. The following section analyzes the effect of different decisions in the configuration of the management model in the sustainability of three real projects of rural electrification.


**Table 1.** Impact of decisions in the management model on sustainability in rural electrification projects.

#### **3. Case Studies**

#### *3.1. Overview and Previous Considerations*

This section showcases the analysis of three rural electrification projects in remote areas of Peru and Mexico. The three projects were led and developed by acciona.org, the corporate foundation of ACCIONA. Acciona.org's projects are based on four fundamental principles: (1) the projects are aligned with the main activity of the corporation with impact on human development (energy, water, and sanitation), which helps take advantage of the knowledge, expertise, and developed networks on these areas in developing regions while also empowering local employees; (2) prioritization of projects in areas that are less attended by other institutions, primarily impoverished rural and remote areas, where service provision is not planned; (3) the main objective of the projects must be the design of solutions that are sustainable in the long term. To do so, affiliate non-for-profit organizations are established; the affiliate organizations act as service providers, which facilitate the creation of long-term relationships with the beneficiaries; (4) collaboration with other local, national, and international organizations, including public, private and academic institutions, as well as other civil society organizations and local communities.

The three projects share some characteristics in their configuration of the management model. The governance model is based on private ownership of the projects, initiated by acciona.org and executed by their national affiliates, in this case acciona.org Perú and acciona.org México. The relationship model is 5P, with different collaborating actors at all levels. An advantage of the 5P configuration is that it guarantees the support of larger public and private entities to promote the project while bringing closer the potential beneficiaries by means of existing communication channels that facilitate adaptation of the project (e.g., local population, banks and credit unions, NGOs) as well as innovation through collaboration with academic and research institutions. 5P models also help overcome a big limitation of purely private ownership with no relations, where the design and the decisions in the other elements of the management model are mostly driven by the risk and low profitability of this kind of projects, which hinders the involvement of the local communities, the understanding of their needs and adaptation of the projects [48,71].

Except for slight differences, the three case studies include similar configurations of their management models in the governance (privately owned, 5P) and technology components (non-conventional, off-grid standalone systems; the technology of the three projects is based on solar home systems due to the dispersion of the communities and long distances to the electric network). Thus, even though the main implications of this configuration on the sustainability of the project will be analyzed, the main focus of the analysis will be the impact of the decisions related to the configuration of the business model on the sustainability of the projects. The following subsections introduce the setting of the three projects and analyze their management model according to the components presented in Section 2.

#### *3.2. Case Study 1: Luz en Casa Cajamarca*

Peru had a poverty rate of 30.1 percent in 2020. Access to electricity now extends to 98.35 percent of the country, but this figure drops to 92.45 percent in rural areas [71]. The department of Cajamarca, with 1.5 million inhabitants, has the lowest electrification rate in the country due to the wide dispersion of households in rural communities, which account for 70 percent of the population [72]. The "Luz en Casa Cajamarca" project (Table 2) was developed by the Spanish acciona.org foundation and the Peruvian civil association acciona.org Peru—the affiliate of the acciona.org foundation in Peru. The main objective of the project was to improve access to energy in the region of Cajamarca.

**Table 2.** "Luz en Casa Cajamarca": Project overview.


#### 3.2.1. Governance Model

"Luz en Casa Cajamarca" is a privately owned project led by acciona.org. In this project, members of the community are selected to work for acciona.org as freelance technical staff, and they oversee installation, set-up, and corrective maintenance of the systems, and provide support in shortage and reconnection tasks.

Local management of the project is performed by acciona.org, which assumes the project's operational tasks. Acciona.org is supported by a network of volunteers who provide occasional collaboration and by the Committees of Photovoltaic Electrification (CEFs). CEFs are a key agent in the operation of the project, acting as intermediaries between acciona.org and end users, helping in system inspection tasks and preventive maintenance, and collecting, transporting, and depositing the periodic fees associated with the service.

The relationship model can be labeled as 5P, including the collaboration of the acciona.org Foundation as promoter, funding body, coordinator, and supervisor; acciona.org as promoter and long-term operating agent; Ministerio de Energía y Minas (MEM) and Organismo Supervisor de la Inversión en Energía y Minería (OSINERGMIN) as regulators; Banco Interamericano para el Desarrollo (BID) as co-financer; Universidad Politécnica de Madrid (UPM) as academic and technical advisor; municipalities as partners and communities; CEFs as users and representatives.

#### 3.2.2. Technological Model

Electricity was generated by second-generation solar home systems (SHS), standalone portable equipment that use renewable energies and allow to reach homes that fall out

of the grid extension plans. SHS are solar photovoltaic generators rated between 11 and 100 Wp (Watt-peak) and have a storage battery [73]. Second-generation SHS are equipment specifically designed to optimize solar energy use, with higher capacity and efficiency than first-generation systems but lower required power. A novelty of these systems is that they feature a DC-DC voltage adapter for radio to allow radio connection of different voltages below DC 12 V and a socket to charge mobile phones [74]. Therefore, the service provided by the SHS in the project allows up to four hours of light using spotlights and the charge and use of entertainment and communication equipment, such as mobile devices. different voltages below DC 12 V and a socket to charge mobile phones [74]. Therefore, the service provided by the SHS in the project allows up to four hours of light using spotlights and the charge and use of entertainment and communication equipment, such as mobile devices. 3.2.3. Business Model The business model was based on EaaS, where a monthly fee is charged for service provision. In this model, the users did not buy the systems: the systems were owned by

Electricity was generated by second-generation solar home systems (SHS), standalone portable equipment that use renewable energies and allow to reach homes that fall out of the grid extension plans. SHS are solar photovoltaic generators rated between 11 and 100 Wp (Watt-peak) and have a storage battery [73]. Second-generation SHS are equipment specifically designed to optimize solar energy use, with higher capacity and efficiency than first-generation systems but lower required power. A novelty of these systems is that they feature a DC-DC voltage adapter for radio to allow radio connection of

*Energies* **2022**, *15*, x FOR PEER REVIEW 10 of 21

#### 3.2.3. Business Model acciona.org, which also oversaw maintenance and replacement in case of system failure

3.2.2. Technological Model

The business model was based on EaaS, where a monthly fee is charged for service provision. In this model, the users did not buy the systems: the systems were owned by acciona.org, which also oversaw maintenance and replacement in case of system failure or end-of-service-life. The decision to implement EaaS was made upon a cost analysis that showed that it was the only option that could support sustainable operation and maintenance costs, even assuming that the complete initial investment would not be recovered. To ensure affordable tariffs, it was necessary to negotiate with the Peruvian government to design a policy and regulatory framework for regulated tariffs in rural electrification with off-grid photovoltaic systems, and to incorporate the tariff into the national cross-subsidy system. The system established that 80 percent of the fee was paid by the state, and end users payed the remaining 20 percent of the fee. These amounts were set after consideration of the average household expenditure of 16.4 soles in access to energy (kerosene, candles, batteries) prior to the start of the project and the calculation of operation costs of the project to ensure sustainability. Prior to the coming into effect of the framework, the service fee was 15 soles per month, whereas after the implementation of the project, the fee was reduced to 10 soles per month, which was paid periodically once the service was provided (i.e., a postpaid system). Figure 2 shows a summary of the management model of the "Luz en Casa Cajamarca" project. or end-of-service-life. The decision to implement EaaS was made upon a cost analysis that showed that it was the only option that could support sustainable operation and maintenance costs, even assuming that the complete initial investment would not be recovered. To ensure affordable tariffs, it was necessary to negotiate with the Peruvian government to design a policy and regulatory framework for regulated tariffs in rural electrification with off-grid photovoltaic systems, and to incorporate the tariff into the national crosssubsidy system. The system established that 80 percent of the fee was paid by the state, and end users payed the remaining 20 percent of the fee. These amounts were set after consideration of the average household expenditure of 16.4 soles in access to energy (kerosene, candles, batteries) prior to the start of the project and the calculation of operation costs of the project to ensure sustainability. Prior to the coming into effect of the framework, the service fee was 15 soles per month, whereas after the implementation of the project, the fee was reduced to 10 soles per month, which was paid periodically once the service was provided (i.e., a postpaid system). Figure 2 shows a summary of the management model of the "Luz en Casa Cajamarca" project.

**Figure 2.** Configuration of the management model of the "Luz en Casa Cajamarca" project. **Figure 2.** Configuration of the management model of the "Luz en Casa Cajamarca" project.

#### 3.2.4. Main Results and Impact on Sustainability

"Luz en Casa Cajamarca" provides energy to 3910 households spread through 188 different communities after 12 years. With the project, acciona.org set in motion a social model that is close to the community and the users, who are not only beneficiaries of the service but can also take part in the operation and management of the project. The data from quantitative questionnaire-based satisfaction surveys of all users showed high rates of acceptance of the project among beneficiaries; additionally, the model ensured affordable access to energy, as default rates dropped from 3.03% to 0.19%. Awareness and sensitivity activities in the communities have increased credibility and seem to validate the management model. Users highlight time availability as one of the main advantages of the system, as they now have study and cooking time at night.

The project contributed to prevent the emission of 1544 tons of CO<sup>2</sup> per year and the generation of 1.65 tons of battery waste, with a minimal environmental impact due to the use of SHS that do not require the deployment of any additional infrastructure. Energy savings for households was estimated to be 656,880 soles per year (approximately EUR 163,000 per year).

The project made energy fees affordable for users with a balance between revenue and costs that ensures the economic viability and continuity of the project. The technology also helped reduce initial costs associated with infrastructure deployment: standalone SHS cost 2000 soles, compared to a cost of 70,000 soles for grid extension in these areas.

From a sustainability perspective, we observed two main shortcomings of the project. The first of them was related to corrective maintenance and operation, both at economic and human levels; because of the isolated and remote location of households, it is very difficult to supply repair parts, as well as to effectively repair faulty systems and collect service fees. The second shortcoming had to do with the low availability of qualified technical staff; even though the project put emphasis on the identification of users with communication and social mobilization skills to increase awareness and reach all potential beneficiaries, it is still necessary to identify residents with the minimum required technical skills.

#### *3.3. Case Study 2: Luz en Casa Oaxaca*

In 2010, 3 million people (2.6 percent of the population) did not have access to energy in Mexico yet. In Oaxaca, this percentage rose to 5.3 percent due to the remote and isolated location of the households [75]. The "Luz en Casa Oaxaca" program (Table 3) was ideated by the acciona.org foundation and the local acciona.org México, analogously to "Luz en Casa Cajamarca". The main objective of the project was to improve access to energy in the state of Oaxaca. The project is now included in the broader EncASa Oaxaca program, which seeks to provide integral and universal access to different basic services: electricity, drinking water, sanitation, and clean cooking.


**Table 3.** "Luz en Casa Oaxaca": Project overview.

#### 3.3.1. Governance Model

Same as "Luz en Casa Cajamarca", this project was privately owned by the acciona.org Foundation and acciona.org Mexico. Acciona.org Mexico is formed by a team of local technical staff from the country, which facilitates the communication with the communities (indigenous habitants represent 35 percent of the population in Oaxaca and the region is the most linguistic diverse in the country, with more than seven languages and dialects spoken, such as Mixteco, Zapoteco, and Chinanteco). Acciona.org México is in charge of local management and operation of the project.

The relationship model was 5P, with the collaboration of the acciona.org foundation as promoter, funding body, coordinator and supervisor; acciona.org México as promoter and responsible of the project in the long term; Gobierno del Estado de Oaxaca, Agencia Española de Cooperación Internacional para el Desarrollo (AECID), and Agencia Mexicana de Cooperación International para el Desarrollo (AMEXCID) as co-financers and intermediaries. The international organization KIVA participated as micro-financers, local municipalities acted as partners, and the communities and CEFs participated as end users and representatives.

#### 3.3.2. Technological Model

The project consisted of the deployment of third-generation SHS, or pico photovoltaic systems. An additional benefit of third-generation systems when compared to secondgeneration SHS (such as those used in "Luz en Casa Cajamarca") is that they are more compact, lighter, easier to install and use (plug and play), and that they have integrated LED lights that allow for more efficient use of energy, in turn extending the duration of the batteries [74]. Third-generation SHS also use lithium batteries (instead of lead batteries), further reducing the weight of the system which, in turn, facilitates transportation to the communities [76].

#### 3.3.3. Business Model

In "Luz en Casa Oaxaca", the end users acquired and owned the systems due to the impossibility of implementing EaaS models when the project started due to regulatory constraints. In this user-ownership model, end users paid for half of the cost of the system and the other half was subsidized by the project's funding bodies. Additionally, if a user could not afford the payment up front, different installments could be made by participating in a micro-credit program directed by KIVA (this program was only active in 2014). As end users owned the systems, they were responsible for maintenance, which entails additional costs beyond the warranty period of the systems.

The model requires higher availability of maintenance than EaaS because the end users have to be able to reach acciona.org's technical staff for diagnostic and repairing tasks. This decision also has an impact on the governance model, by delegating operational tasks in each area to local entrepreneurs who, from the so-called Centros Luz en Casa (CLCs), provide support and receive fee payments; they also establish franchises to sell appliances that are compatible with the SHS in order to extend the functionalities of the systems beyond lighting. CLCs are usually already existing local businesses located in reference sites that dedicate part of their space and time to support "Luz en Casa Oaxaca" users. Local entrepreneurs receive training to provide effective corrective maintenance and keep constant communication with the project's management bodies for fee and stock management; in return, they receive a commission fee for each repaired system or sold item. Currently, the project has five CLCs and four Centros EncASa (the latter offer support for sanitation, cooking and drinking water services, in addition to SHS systems and compatible appliances).

Figure 3 summarizes the management model of the "Luz en Casa Oaxaca" project.

**Figure 3.** Configuration of the management model of the "Luz en Casa Oaxaca" project. **Figure 3.** Configuration of the management model of the "Luz en Casa Oaxaca" project.

#### 3.3.4. Main Results and Impact on Sustainability 3.3.4. Main Results and Impact on Sustainability

The project, including Luz en Casa Oaxaca and EncASa Oaxaca, provided energy access to 8525 households in 628 communities after 9 years. Acciona.org subsidized the systems through a partnership involving several public and private entities, ensuring affordability for the most needed and improving universal access. The multi-agent configuration, supported by the CEFs and CLCs, provided higher availability of repairing staff than the "Luz en Casa Cajamarca" project through the identification and training of personnel with technical skills, in addition to awareness and social sensibilization skills. Three out of the nine local entrepreneurs were women who have become referents in the communities they provide support to, advancing towards gender equality. The partnership with public agents helped not only subsidize the systems but also improve and facilitate communication with the communities and a previous identification of the commu-The project, including Luz en Casa Oaxaca and EncASa Oaxaca, provided energy access to 8525 households in 628 communities after 9 years. Acciona.org subsidized the systems through a partnership involving several public and private entities, ensuring affordability for the most needed and improving universal access. The multi-agent configuration, supported by the CEFs and CLCs, provided higher availability of repairing staff than the "Luz en Casa Cajamarca" project through the identification and training of personnel with technical skills, in addition to awareness and social sensibilization skills. Three out of the nine local entrepreneurs were women who have become referents in the communities they provide support to, advancing towards gender equality. The partnership with public agents helped not only subsidize the systems but also improve and facilitate communication with the communities and a previous identification of the communities that would benefit from the project.

nities that would benefit from the project. The SHS helped prevent the emission of 1075 tons of CO2 per year and 12.42 tons of uncontrolled battery waste, with low impact on the environment. Household energy sav-The SHS helped prevent the emission of 1075 tons of CO<sup>2</sup> per year and 12.42 tons of uncontrolled battery waste, with low impact on the environment. Household energy savings were estimated at 18.75 M pesos per year (approximately EUR 758,200 per year).

ings were estimated at 18.75 M pesos per year (approximately EUR 758,200 per year). The subsidization program reduced cost acquisition in half, facilitating affordability, The subsidization program reduced cost acquisition in half, facilitating affordability, and the SHS reduced initial and deployment costs.

and the SHS reduced initial and deployment costs. We observed two elements that could be improved from a sustainability perspective. First, maintenance costs may be high for end users. Because of the business model, costs associated with maintenance and repairing are not included in the price and, therefore, must be bore by the users, increasing costs in the case of system failure. Consequently, repair rates were low and, in some cases, the systems were abandoned when they stopped working because users could not afford the cost associated with repairing or replacing the system. Second, subsidization of system acquisition makes it necessary to increase awareness and sensibilization, as having the system subsidized increases the perception of higher costs if replacements are needed, which in turn may increase system abandonment rates. However, the alternative is to use conventional sources of energy, which are more We observed two elements that could be improved from a sustainability perspective. First, maintenance costs may be high for end users. Because of the business model, costs associated with maintenance and repairing are not included in the price and, therefore, must be bore by the users, increasing costs in the case of system failure. Consequently, repair rates were low and, in some cases, the systems were abandoned when they stopped working because users could not afford the cost associated with repairing or replacing the system. Second, subsidization of system acquisition makes it necessary to increase awareness and sensibilization, as having the system subsidized increases the perception of higher costs if replacements are needed, which in turn may increase system abandonment rates. However, the alternative is to use conventional sources of energy, which are more harmful for both end users and the environment.

#### harmful for both end users and the environment. *3.4. Case Study 3: Luz en Casa Amazonía (Napo)*

*3.4. Case Study 3: Luz en Casa Amazonía (Napo)*  In the Amazon rainforest area, only 82.7 percent of the population has access to electric energy [71], which means that about one in five people do not have access to energy. In the Amazon rainforest area, only 82.7 percent of the population has access to electric energy [71], which means that about one in five people do not have access to energy. The "Luz en Casa Amazonía" Project (Table 4), designed by the acciona.org foundation and

acciona.org Perú, aimed to improve access to electric energy of indigenous communities in the Peruvian Amazon rainforest.



#### 3.4.1. Governance Model

Analogously to the previous two projects, "Luz en Casa Amazonía" was privately owned by the acciona.org foundation and acciona.org Peru. Acciona.org Peru is supported by CEFs and has CLCs for local management and operational tasks, including fee collection and payment, repairs, and sales.

The relationships are 5P, with the collaboration of the acciona.org foundation as promoter, funding body, coordinator, and supervisor; acciona.org Peru as promoter and long-term operator; Ministerio de Energía y Minas (MEM) and Organismo Superior de la Inversión en Energía Minera (OSINERGMIN) as regulators and policymakers; the Peruvian Government, Agencia Española de Cooperación Internacional para el Desarrollo (AECID), and Fondo Nacional de Desarrollo Científico, Tecnológico y de Innovación Tecnológica (FONDECYT) as co-financers; Universidad Politécnica de Madrid (UPM), Fundación de Ingenieros del ICAI para el Desarrollo (FICAID), and the Instituto de Investigación Tecnológica (IIT) as academic and technical advisors and partners; municipalities as partners; communities, CLCs (four), and CEFs as users, representatives, and micro-franchises.

#### 3.4.2. Technological Model

The technology applied for service provision was based on third-generation SHS (same as in "Luz en Casa Oaxaca"), but using the pay-as-you-go (PAYG) functionality of the system to simplify the payment process. According to the World Bank, PAYG is one of the most effective solutions to provide decentralized access to energy in remote communities of developing countries [77]. PAYG systems, which were not used in the previous two projects, allow for a prepaid system of energy provision in which the end user pays a fee for a code that unblocks the SHS for a period of time. Once that period expires, the user must complete another payment and a new unblocking code is generated, or else the SHS will be blocked. Therefore, PAYG systems facilitate remote management of off-grid systems, reducing default rates and the risk associated with the initial investment [78].

#### 3.4.3. Business Model

As in the "Luz en Casa Cajamarca", the energy was provided as EaaS. The business model improves the affordability of the technology: users can also benefit from cross subsidies and the tariff is regulated; therefore, end users only pay 20 percent of the total costs (approximately 9 soles per month). The main difference with the "Luz en Casa Cajamarca" project was that PAYG systems give more freedom to end users, which can accommodate their consumption to their disposable income.

Figure 4 summarizes the management model of the "Luz en Casa Amazonía" project.

**Figure 4.** Configuration of the management model of the "Luz en Casa Amazonía" project. **Figure 4.** Configuration of the management model of the "Luz en Casa Amazonía" project.

Figure 4 summarizes the management model of the "Luz en Casa Amazonía" project.

3.4.4. Main Results and Impact on Sustainability

3.4.4. Main Results and Impact on Sustainability The project provided access to energy to 1518 households in 58 communities after 5 years. From a social sustainability perspective, acciona.org spread out a clean electrification solution for remote rural indigenous communities that minimizes the use of fossil fuel. According to satisfaction surveys collected by the project promoter with replies to a quantitative questionnaire of all participants in the pilot project, all the beneficiaries (100 percent) reported the highest level of satisfaction with the use of the SHS, and the majority of them (98 percent) were also satisfied with their energy savings. The multi-agent model and the incorporation of the CEFs and CLCs make service management and operation viable, as the households are only accessible by the river, and the weather conditions (tem-The project provided access to energy to 1518 households in 58 communities after 5 years. From a social sustainability perspective, acciona.org spread out a clean electrification solution for remote rural indigenous communities that minimizes the use of fossil fuel. According to satisfaction surveys collected by the project promoter with replies to a quantitative questionnaire of all participants in the pilot project, all the beneficiaries (100 percent) reported the highest level of satisfaction with the use of the SHS, and the majority of them (98 percent) were also satisfied with their energy savings. The multi-agent model and the incorporation of the CEFs and CLCs make service management and operation viable, as the households are only accessible by the river, and the weather conditions (temperature and rain) are extreme, which further complicates distribution and support.

perature and rain) are extreme, which further complicates distribution and support. The implementation of PAYG systems also has a positive impact in operation and management because prepaid installments help reduce default rates (which are close to zero) and delays in fee payments (a reduction of 83 percent, from 68 to 11 days). CLCs have made 80 percent of system repairs (including a 13.2 percent rise in the last year), thus The implementation of PAYG systems also has a positive impact in operation and management because prepaid installments help reduce default rates (which are close to zero) and delays in fee payments (a reduction of 83 percent, from 68 to 11 days). CLCs have made 80 percent of system repairs (including a 13.2 percent rise in the last year), thus reducing trip cost and time of local entrepreneurs and facilitating maintenance.

reducing trip cost and time of local entrepreneurs and facilitating maintenance. With a minimal environmental impact, the project has helped reduce 158 tons of CO2 emissions per year and prevented 13 tons of uncontrolled battery waste. The estimated energy savings for all households was 216,133 soles per year (approximately EUR 53,365 per year). The EaaS model also favors environmental sustainability because acciona.org was responsible for the management of faulty systems, which were collected and sent to With a minimal environmental impact, the project has helped reduce 158 tons of CO<sup>2</sup> emissions per year and prevented 13 tons of uncontrolled battery waste. The estimated energy savings for all households was 216,133 soles per year (approximately EUR 53,365 per year). The EaaS model also favors environmental sustainability because acciona.org was responsible for the management of faulty systems, which were collected and sent to the provider of the SHS for adequate recycling of each component.

the provider of the SHS for adequate recycling of each component. Fees have been made affordable, and the balance between revenue and costs ensures Fees have been made affordable, and the balance between revenue and costs ensures the continuity and economic sustainability of the project.

the continuity and economic sustainability of the project. Despite the good results of the project, we observed a missed opportunity, because EaaS with prepaid modes may enable the use of information and communication technologies to improve fee collection and management, as well as the delivery of activation Despite the good results of the project, we observed a missed opportunity, because EaaS with prepaid modes may enable the use of information and communication technologies to improve fee collection and management, as well as the delivery of activation codes to users.

#### codes to users. **4. Discussion**

Upon the analysis of the three case studies of rural electrification project, we can assess the impact of different configurations of the management model in the sustainability of the projects.

#### *4.1. Social Sustainability*

First, we observed a strong influence of the governance model in the social sustainability of the project, with better results in the projects where the number of agents involved in the operation of the project was higher. From the three projects under analysis, it is recommended to incorporate actors that (a) facilitate awareness and social sensibilization, (b) facilitate knowledge and technology transfer, and (c) know the local culture and customs [11,71]. For example, the incorporation of the CLCs in the latter two projects was so successful that CLCs have been added to the "Luz en Casa Cajamarca" project, which now has 12 active CLCs. The inclusion of entrepreneur profiles and CEFs in the projects improves the availability of maintenance, allows for training of local users, increases the participation of women in technical tasks and, in the end, facilitates the continuity of the project by generating a socially sustainable model.

The business model also has an effect on the social sustainability of the project. From the analysis of the three case studies, EaaS seems to offer better results in improving affordability, which is one of the main challenges in universal access to energy [43]. Besides, EaaS guarantees maintenance and replacement of faulty systems which, in turn, contributes to continued use and (again) to universal access because repairs do not entail additional cost for end users [77].

Funding and subsidization are also key for social sustainability. Both are heavily dependent on the agents involved in the project (states, regulators, and policymakers) and the resulting business model [79,80]. For example, only after the negotiations of acciona.org in Peru it was possible to create a regulated EaaS tariff for photovoltaic systems, which in turn made it possible to subsidize the fee and improve affordability of the systems.

#### *4.2. Environmental Sustainability*

In all three cases the results were similar, a result consistent with the use of the same technological model in the three projects (standalone systems using non-conventional or renewable sources). The positive results confirm previous literature, which identifies this kind of systems as the most adequate choice in rural electrification projects because they can reach remote and disperse communities, have high reliability and provide a more stable access to energy that does not depend on conventional grids nor requires complex infrastructure [55,81,82].

However, the use of systems that have a limited service life (mainly because of the batteries) requires the establishment of protocols to facilitate recycling of the different elements of the SHS, which otherwise would generate toxic waste and damage to the ecosystems [43]. Current trends in rural electrification assess whether it is possible to use fewer batteries to reduce the waste generated by SHS [82] by using independent microgrids that rely on hybrid sources (e.g., solar and diesel) and require much lower battery capacity [69]. For instance, the acciona.org foundation is currently piloting a project in Copal Urco (Peru) to validate the sustainability of the technological model with hybrid sources (using diesel only as a backup source of energy generation).

#### *4.3. Economic Sustainability*

Economic sustainability is affected by the three main components of the model, as every aspect had a direct impact on the total costs and profitability of the project. First, the use of standalone systems has lower initial and infrastructure deployment costs, which favors economic sustainability [82]. Second, multi-agent governance models also have a positive effect on the economic sustainability of the projects; for instance, when the project incorporates the CLCs, maintenance costs are reduced due to the fact of less frequent trips and higher repair rates [79]. Third, projects operated as EaaS achieve a balance between revenue and costs that facilitates economic sustainability [81], even though profitability is only possible when a regulated tariff and subsidy are in place. This seems to suggest that fee-for-service models (both for system acquisition and maintenance) offer better results regarding economic sustainability.

#### **5. Conclusions**

The first part of this study showed a great variety of possible configurations of the management model in rural electrification projects. From a review of the literature, we elaborated a model that showed the different elements of the management model with three main components: governance, technological and business models. The governance model included all aspects relative to the actors involved in the project: owners/promoters or operators and their relationship with other agents, such as financers, regulators, academic institutions, and the communities and users that gain access to energy. The technological model considers choices in the technology based on the location of the communities, their dispersion, proximity to an existing grid, and available sources of energy generation. The business model includes how the systems are acquired and owned, maintenance costs, and the type of contract for service provision.

Our review highlights the concerns of rural electrification project promoters about achieving sustainability and affordability of the service to provide effective universal access while considering the diversity of each context that requires flexible and adaptable solutions. It is very complex to make decisions about the management model that improve the three dimensions of sustainability (social, environmental, and economic) because, most often, economic sustainability (namely, the profitability of the project) becomes the most relevant variable. It is not rare to observe decisions that improve the profitability but at the expense of affordability for end users.

Our analysis of the configurations of the management model of rural electrification projects identified 16 elements and their impact on the dimensions of sustainability. From the analysis, we observed that the environmental sustainability was mostly dependent on the technological model, but the business model had the higher impact on the social and economic sustainability of the project. The analysis of the three case studies under the lens of the management model and its impact on the sustainability of the project facilitated the identification of the strengths, weaknesses and areas of improvement of each project.

The main finding of the study is the key role of the business model and the collaboration between the agents involved in the sustainability of the projects. The choice of the business model is critical, and the analysis showed that fee-for-service models seem more adequate to ensure sustainability of the project. Socially, it allows for affordability of the systems to users in extreme poverty, by combining the fees with subsidies. Economically, in combination with an adequate regulatory framework and subsidization, it facilitates profitability and reduces operation and maintenance costs notably.

Collaboration between agents is essential to achieve sustainability. The analysis of the case studies suggests that variety of roles is important and that the communities and end users must have an active role in the project. In this sense, 5P and involvement of the communities in operational tasks seem to highly favor social sustainability.

The study is not exempt from limitations. Most notably, the three case studies had very similar configurations of the governance and technological models. While this choice allowed us to better understand the effect of the different configurations of the business model on the sustainability of rural electrification projects, a wider selection of case studies with different configurations of the governance and technological model are necessary to better understand the impact of each element of the management model on the different dimensions of sustainability.

**Author Contributions:** Conceptualization, L.D.-R.-C., S.I.-P. and Á.H.-G.; Methodology, L.D.-R.-C. and E.A.-N.; Validation, L.D.-R.-C., E.A.-N. and S.I.-P.; Formal analysis, L.D.-R.-C. and Á.H.-G.; Investigation, L.D.-R.-C., E.A.-N. and Á.H.-G.; Resources, L.D.-R.-C. and S.I.-P.; writing—original draft preparation, L.D.-R.-C. and Á.H.-G.; writing—review and editing, E.A.-N. and S.I.-P.; Visualization, L.D.-R.-C.; Supervision, S.I.-P.; Project administration, S.I.-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:** Not applicable.

**Acknowledgments:** The authors would like to thank the acciona.org foundation for their help and support in the elaboration of this study.

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

#### **References**


#### *Article* **Appraisal Modeling for FSRU Greenfield Energy Projects**

**Dimitrios Dimitriou \* and Panagiotis Zeimpekis**

Department of Economics, Democritus University of Thrace, 69100 Komotini, Greece; panazeib@econ.duth.gr **\*** Correspondence: ddimitri@econ.duth.gr

**Abstract:** Floating storage and regasification units (FSRU) provide a flexible and competitive energy distribution option when it comes to the regasification of liquefied natural gas (LNG). FSRU projects have become more and more popular, attracting the interest of investors, energy authorities, and governments; therefore, the project feasibility in terms of risks and profitability are a major concern. This paper deals with the appraisal of a greenfield LNG infrastructure project, where usually, decision complexity deals with the high number and different expectation of stakeholders, the capital-intensive financing nature, and the business risks in the project life cycle. Conventional wisdom is to provide a coherent, compact, and well-structured appraisal modelling framework, adjusted to FSRU technical, structural, and operational features on one hand; and business risks, long-term life cycle, and investment attractiveness on the other. Appraisal modelling structure and outputs are considered to provide key messages to the decisions involved and interested parties toward the project feasibility and the associated investment risks for the implementation of the FSRU project. The proposed modelling framework was applied to the Alexandroupolis FSRU project, where the first discussion was many years ago, but the existing conditions in the energy market are raising the interest for developing energy distribution facilities globally.

**Keywords:** FSRU; LNG; appraisal modeling; modeling energy project appraisal; economic assessment; project incentives; energy project accounting; greenfield project planning; project due-diligence

**1. Introduction**

As the need for a more sustainable and decarbonized future is continuously growing [1], natural gas (NG) has become one of the main primary energy sources in the world, due to its environmentally friendly nature and its multiple uses across a number of sectors. In the European Union (EU), NG is mainly used to produce electricity and for heating purposes, but it can be expanded to new fields such as agriculture [2] and aviation [3], and it is the second most commonly used energy source, combining for 23.1% of total energy consumption for the year 2018 [4].

One of the biggest challenges in energy consumption that places such as Europe have faced is the extremely high dependency in NG imports. Since the only EU members which have facilities for the production of NG are Norway and the Netherlands, the import dependency of NG has reached almost 90% for 2019, with Russia being the main supplier, providing 41.3% of the NG used in European Union countries for 2019 [4], although the NG flow from Russia to the European Union is projected to significantly decrease in the coming years [5,6]. This relationship strongly affects the European gas market [7,8]. The EU, in need of securing its energy stability and primary sources availability, set the diversification of energy sources to supply one of the main aims of its energy policy [9].

For NG, this can be succeeded by Liquefied Natural Gas (LNG), which allows NG to be transported in large quantities all over the world [10], without the location-specific limitations that underground pipelines create. The traditional supply chain of LNG includes gas production, liquefaction, shipping, storage, and regasification [11]. LNG is a liquid state of natural gas which is cooled to below 113.1 K, and its volume is 600 times smaller than that of NG [12]. Upon reaching its destination, LNG is reverted to natural gas at import

**Citation:** Dimitriou, D.; Zeimpekis, P. Appraisal Modeling for FSRU Greenfield Energy Projects. *Energies* **2022**, *15*, 3188. https://doi.org/ 10.3390/en15093188

Academic Editors: Bernard Zi˛ebicki and Edyta Bieli ´nska-Dusza

Received: 6 March 2022 Accepted: 21 April 2022 Published: 27 April 2022

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

**Copyright:** © 2022 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/).

terminals and distributed via pipelines to the final consumers [13,14]. The conversion from the liquid phase of LNG to gas can be implemented using onshore facilities or regasification ships moored at specially designed docks for this purpose, allocating significantly fewer capital expenses than the onshore terminals [15,16]. In the last case, the terminal is based on a Floating Storage and Regasification Unit (FSRU) permanently moored at the jetty and periodically supplied by an LNG carrier. FSRU, for small- and mid-sized markets, typically offer a more cost-effective, faster, and flexible means to bring natural gas to consumers compared to shore-based terminals [17–19]. In 2020, there were 34 FSRU operating, and 12 more are under construction [20]. Most of these stations are located in Eastern Asia and America. Europe are currently operating only five FSRU: in Toscana, Italy; in Klaipeda, Lithuania; in Aliaga and Dortyol, Turkey; whereas the most recently constructed FSRU is the one in Krk, Croatia, which started operation in 2021 [21,22]. Moreover, the only FSRU that is currently constructing in Europe is in Cyprus [21].

Greece is one of the EU members that is most dependent in NG imports, since it imports the 100% of its needs. The energy transportation to renewable sources has turned NG to one of the main energy sources used in Greece, increasing the usage from 3.500 mil NM3 in 2009 to almost 5000 mil NM3 in 2019 [23]. The need for the diversification of sources has also changed the ways that Greece imports its NG quantities, and increased the needs for LNG. The LNG supplied to Greece for 2019 was 47.6% of the whole NG imported, whereas it was only 19.7% in 2018 [23]. According to Greece's yearly National Natural Gas System (NNGS) allocation data, the LNG was supplied mainly from USA (48.26%) and Qatar (22.36%) [24].

While the demand is growing so fast, the necessity for more expanded and welldeveloped natural gas distribution infrastructure is rising as well. Greece currently disposes an onshore terminal facility for the regasification process in Agia Triada, next to Athens. Besides this, a new energy projects initiative with the approval of the government decided on the construction of a new FSRU station at the Alexandroupolis sea area, located in the northern part of Greece.

This paper sets a comprehensive dynamic framework to evaluate the feasibility and economic footprint for the development of a new FSRU Station, considering the energy market characteristics, the project development phases, and the financing volatility. By a systemic approach, the appraisal modelling is presented, and it is applied to the Alexandroupolis city FSRU project in Greece. The output sensitivity, reviewed based on the FSRU station usage rate, linked the project viability with the energy demand risks. Additionally, the analysis provides key messages to planners and decision-makers regarding the incentives towards FSRU project implementation, promoting risk-sharing measures towards investment attractiveness.

This paper's novelty is to provide a coherent FSRU appraisal modelling structure adjusted for a new energy infrastructure project with a long-term payback period and high volatile business environment, which may support planners, economists, and decisionmakers to illustrate conclusions on project feasibility, compare it with other projects, and apply it to similar cases. The analysis output sensitivity, reviewed based on the energy FSRU station usage rate, linked the project viability with the energy demand risks. Additionally, the analysis provides key messages to planners and decision-makers regarding the incentives towards FSRU project implementation, promoting risk-sharing measures towards investment attractiveness. This paper's novelty is to provide a coherent FSRU appraisal modelling structure adjusted for a new energy infrastructure project with a longterm payback period and high volatile business environment, which may support planners, economists, and decision-makers to illustrate conclusions on project feasibility, compare it with other projects, and apply it to similar cases.

#### **2. Alexandroupolis FSRU Station Features**

The Alexandroupolis, Greece, FSRU Station is planned to be constructed in the sea area southwest of the town of Alexandroupolis at a distance of 17.6 km from Alexandroupolis

Port (Figure 1). The floating unit is planned to have four LNG storage tanks with a total capacity of up to 170,000 cubic meters. The dimensions of the floating unit will be approximately 300 m in length (LOA), 32.5 m in breadth, and 26.5 m in height (depth to upper deck) [25]. • Supply the Turkish market via reverse flow functionality of the existing Turkey– Greece Interconnection Pipeline. • Connect with the European South Corridor gas projects such as TAP. The exact location of the station and the interconnections with the neighboring countries are presented in Figure 1.

The Alexandroupolis, Greece, FSRU Station is planned to be constructed in the sea area southwest of the town of Alexandroupolis at a distance of 17.6 km from Alexandroupolis Port (Figure 1). The floating unit is planned to have four LNG storage tanks with a total capacity of up to 170,000 cubic meters. The dimensions of the floating unit will be approximately 300 m in length (LOA), 32.5 m in breadth, and 26.5 m in height

The NG after the regasification process will be transferred to the Greek National NG System (NNGS) via a 24 km subsea pipeline, which will be laid on a sea depth over 15 m, and a 4 km, 30", and 100 bar operating pressure onshore underground pipeline [25]. A new Metering Station is constructed to act as the connection point with the (NNGS).

• Feed directly into the Greek NG Transmission System (NNGS) to supply the Greek

• Provide direct access to the Bulgarian market via the under-construction Greek– Bulgarian Interconnector Pipeline (IGB) and through the Bulgarian pipelines to the

*Energies* **2022**, *15*, x FOR PEER REVIEW 3 of 22

The key activities for the new FSRU Station are:

Balkan region, Hungary and Eastern Europe markets.

**2. Alexandroupolis FSRU Station Features** 

(depth to upper deck) [25].

market.

**Figure 1.** (**a**) Alexandroupolis FSRU interconnections [25]; (**b**) Alexandroupolis FSRU location [25]. **Figure 1.** (**a**) Alexandroupolis FSRU interconnections [25]; (**b**) Alexandroupolis FSRU location [25].

**3. FSRU Appraisal Modelling Framework**  The appraisal modelling framework was developed by a breakdown of the key unit cost in the project lifecycle. The FSRU project characteristics taken into consideration for the definition the project unit cost distinguished the lifecycle in two time periods: the The NG after the regasification process will be transferred to the Greek National NG System (NNGS) via a 24 km subsea pipeline, which will be laid on a sea depth over 15 m, and a 4 km, 30", and 100 bar operating pressure onshore underground pipeline [25]. A new Metering Station is constructed to act as the connection point with the (NNGS).

construction and the operation. In the operation period, the energy market volatility has The key activities for the new FSRU Station are:


The exact location of the station and the interconnections with the neighboring countries are presented in Figure 1.

#### **3. FSRU Appraisal Modelling Framework**

The appraisal modelling framework was developed by a breakdown of the key unit cost in the project lifecycle. The FSRU project characteristics taken into consideration for the definition the project unit cost distinguished the lifecycle in two time periods: the construction and the operation. In the operation period, the energy market volatility has taken into consideration the promotion of a dynamic formulation of the equations. In the following part, the modelling structure and the associated explanations are expressed.

#### *3.1. Initial Capital Cost*

The total capital cost required during the construction period includes the construction cost and the project implementation cost. The construction cost includes the FSRU Vessel purchase, which is usually the larger unit cost; the mooring of the Vessel; the construction of subsea and inland pipelines to transfer the NG to the larger transmission pipeline; and the cost of other equipment needed for the FSRU Station operation. Project implementation cost is considered the cost of the project team, management support services, the specialists, and consultants prior to the final investment decision; and the administrative costs, such as the contract preparation, the procurement, the permits, and the licenses from authorities.

Therefore, the project capital cost calculated by the following equation:

$$\text{LCC} = \text{Con} + \text{Own} + \text{Cont} \tag{1}$$

where,

ICC is the total Initial Capital Cost during the construction period; Con is the Construction Cost;

Own is the project implementation costs undertaken by the FSRU project shareholders; and Cont is the cost of any contingency that may happen.

#### *3.2. Financing*

In such a large investment, usually, the funding for the capital costs is undertaken either by the shareholders equity or by a project financing tool, such as bank loan, bond issue, or by a combination of shareholders and bank or market funds. Therefore, the total capital cost is:

$$\text{ICC} = \text{E} + \text{B} \tag{2}$$

where,

ICC is the total Initial Capital Costs during the construction period (see Equation (1)); E is the Shareholders Equity; and

B is the funds retrieved by a bank loan.

In case of a bank loan, the annual installment is determined by the following equation:

$$\mathbf{L}\_{\mathbf{i}} = \frac{\mathbf{B} \ast \mathbf{i}\_{\mathrm{RL}} \ast (\mathbf{1} + \mathbf{i}\_{\mathrm{RL}})^{\mathrm{n}}}{\left(\mathbf{1} + \mathbf{i}\_{\mathrm{RL}}\right)^{\mathrm{n}} - 1} \tag{3}$$

where, i ∈ N and 1 ≤ i ≤ n;

Li is the annual loan installment in the year i;

B is the total funds retrieved by the bank loan;

iRL is the loan interest rate; and

n is the years that the loan must be paid.

For the project financing lifecycle, the total loan cost is:

$$\text{TL} = \sum\_{i=1}^{n} \text{L}\_{i} \tag{4}$$

where, i ∈ N and 1 ≤ i ≤ n;

TL is the total loan cost during the whole period of the loan payment; Li is the annual loan installment in the year I (see Equation (3)); and n is the years that the loan must be paid.

The annual loan installment includes both an interest and a principal amount. The interest amount is used in order to calculate the net income before taxes, so it must be estimated. The mathematic formula is given below:

$$\mathbf{L}\_{\rm li} = \mathbf{L}\_{\rm i} + \left(\mathbf{1} + \mathbf{i}\_{\rm RL}\right)^{\rm i - 1} \* \left(\mathbf{B} \* \mathbf{i}\_{\rm RL} - \mathbf{L}\_{\rm i}\right) \tag{5}$$

where, i ∈ N and 1 ≤ i ≤ n;

LIi is the interest paid amount of the loan that is given during the year i; Li is the annual loan installment in the year I (see Equation (3));

B is the total funds retrieved by the bank loan (see Equation (2)); and iRL is the loan interest rate.

The principal paid amount in each annual installment is given by the following formula:

$$\mathbf{L}\_{\rm Pi} = \mathbf{L}\_{\rm i} - \mathbf{L}\_{\rm li} \tag{6}$$

where, i ∈ N and 1 ≤ i ≤ n;

LPi is the principal paid amount of the loan that is given during the year i; LIi is the interest paid amount of the loan that is given during the year i (see Equation (5)); Li is the annual loan installment in the year i (see Equation (3)).

#### *3.3. Sales Income*

The FSRU revenue streams deal with the regasification services that it provides. The extent of the sales income depends on the capacity of the FSRU's Vessel, due to the fact that the larger the storages are, the more LNG quantity can be degasified. Moreover, it depends on the quantity of the LNG carriers that use the facilities, which, in this analysis, is called "usage rate" of the Station, and is expressed as a percentage of the Station's yearly usage. Finally, it depends on the tariff that the Station charges its customers for the regasification services, and is usually expressed on price per normalized cubic meter (Nm<sup>3</sup> ). The Station will operate for (T) years, and given the above, the annual Sales income is given by:

$$\text{RV}\_{\text{i}} = \text{C}\_{\text{V}} \ast \text{uf}\_{\text{i}} \ast \text{TR}\_{\text{i}} \tag{7}$$

where, i ∈ N and 1 ≤ i ≤ T;

T is the years of operation of the FSRU Station;

RV<sup>i</sup> is the annual revenue for the operation year i;

C<sup>V</sup> is the total annual capacity of the Vessel in Nm<sup>3</sup> ;

uf<sup>i</sup> is the usage rate of the FSRU Station for the operation year i;

TR<sup>i</sup> is the tariff price charged per normalized cubic meter or LNG for the operation year i.

The total sales income for the whole operation period of the Station is:

$$\mathbf{R} = \sum\_{\mathbf{i}=1}^{\mathbf{T}} \mathbf{R} \mathbf{V}\_{\mathbf{i}} \tag{8}$$

where, i ∈ N and 1 ≤ i ≤ T;

RV<sup>i</sup> is the annual revenue for the operation year i.

#### *3.4. Operating and Capital Costs*

The main operational cost covers the personnel working in the Station, usually around 120 persons, who work on the onshore facilities, in the marine section for the transportation of the personnel, in the operating department, and the maintenance department of the FSRU. Furthermore, the operational costs include the power and steam generation needed for the FSRU's equipment; the maintenance and inspection costs; the underwater inspections; the materials, chemicals, and spare part supply; the insurance; the harbor fees; the service boat maintenance; the dredging inspection; and finally, the head office support to the operations.

The operational cost for the power and steam generation during the regasification process is estimated according to the below formula:

$$\text{PowerCost}\_{\text{i}} = \text{PowerCost}\_{\text{Y}} \ast \text{uf}\_{\text{i}} \tag{9}$$

where, i ∈ N and 1 ≤ i ≤ T;

PowCost<sup>i</sup> is the power and steam generation cost for the operation year i; PowCost<sup>Y</sup> is the power and steam generation cost if the usage rate was 100%; and uf<sup>i</sup> is the usage rate of the FSRU Station for the operation year i.

In addition, the operational costs depending on the usage rate of the Station are essential to take into consideration. To quantify the variable cost that is related to the usage rate of the FSRU, it is estimated by correlating it with the operational expenses when the usage rate is 100%, and the usage rate through a variable named b<sup>i</sup> . Respectively, the fixed costs to the operational costs of 100% usage rate use a variable named a<sup>i</sup> . Therefore, the fixed and variable costs, as well as the operational costs given from the below group of equations:

$$\text{FC}\_{\text{i}} = \text{a}\_{\text{i}} \* \text{OPE} \text{X}\_{100\%} \tag{10}$$

$$\text{VC}\_{\text{i}} = \text{b}\_{\text{i}} \* \text{OPE} \mathbf{X}\_{100\%} \* \text{uf}\_{\text{i}} \tag{11}$$

$$\text{OPEX}\_{\text{i}} = \text{FC}\_{\text{i}} + \text{VC}\_{\text{i}} \tag{12}$$

where, i ∈ N and 1 ≤ i ≤ T;

OPEX100% is the annual operating cost for 100% usage rate of the FSRU Station;

OPEX<sup>i</sup> is the annual operational cost except from the LNG consumption cost for the operation year i;

FC<sup>i</sup> is the fixed cost for the operation year i;

VC<sup>i</sup> is the variable cost, for the operation year i;

a<sup>i</sup> and b<sup>i</sup> are numbers that correlate the fixed and variable costs with the operational costs, ai ,b<sup>i</sup> ∈ R, 0 < a<sup>i</sup> < 1, 0 < b<sup>i</sup> < 1, and a<sup>i</sup> + b<sup>i</sup> = 1.

The annual capital expenditures include all the costs related to the upgrade of the existing facilities and equipment that are necessary for the operation of the FSRU. Furthermore, they include the costs of acquiring new assets, and research and development (R&D) costs. The annual capital costs are usually correlated to the initial investment cost, and can de expressed as:

$$\text{CAPEX}\_{\text{i}} = \text{c}\_{\text{i}} \* \text{ICC} \tag{13}$$

where, i ∈ N and 1 ≤ i ≤ T;

CAPEX<sup>i</sup> is the capital expenditures for the operation year i;

ci is a number that correlates the annual capital expenditures with the capital costs of the construction period, c<sup>i</sup> ∈ R, 0 < c<sup>i</sup> < 1.

#### *3.5. Depreciation and Amortization*

It is important to analyze and distinguish the capital expenditures from the operating cost, because the first is depreciated and reduces the final payable tax. Moreover, the depreciation depicts the value of the project assets over time, providing the market value of the project. To calculate the depreciation, we have to add the depreciation of the initial capital cost to the depreciation of the annual capital expenditures produced every operating year. This means that to estimate the depreciation cost for the 10th year of operation, the following are added: the depreciation of initial capital cost during the 10th year, the depreciation of capital expenditure of 1st year of operation during the 10th year, the depreciation of capital expenditure of 2nd year of operation during the 10th year, etc. In real-life applications, it uses methods to account for depreciation, such as straight line depreciation, declining balance, double declining balance, sum of the year's digits, and units of production. In this analysis, the straight line method approach is selected, because the project production ability is not stabile over time, and it provides more accuracy for a long lifecycle project with a long payback period.

Using the straight line method, the salvage value of the asset after the depreciation period and the depreciation rate must be set according to accounting norms and the used materials and equipment in the project. In this analysis, the depreciation ratio received was stable over time, and it covered the overall project lifecycle. This means that the depreciation amount for each infrastructure and equipment unit cost will be stable for all of the operating period, except from the last year that it can get a lower price, until it matches its salvage value.

Therefore, the depreciation for the construction cost is estimated by:

$$\text{If } \mathbf{i} = \mathbf{1}, \text{ then } \mathbf{D}\_{\text{(i)Con}} = \mathbf{d}\_{\text{Con}} \* (\text{Con} - \text{SAL}\_{\text{Con}}) \tag{14a}$$

$$\begin{array}{c} \text{If } i > 1 \text{ and } \text{REM}\_{(i-1)\text{Con}} - \text{D}\_{(i-1)\text{Con}} > \text{SAL}\_{\text{Con}}, \text{ then} \\ \text{D}\_{(i)\text{Con}} = \text{d}\_{\text{Con}} \* (\text{Con} - \text{SAL}\_{\text{Con}}) \end{array} \tag{14b}$$

$$\begin{array}{l} \text{If } i > 1 \text{ and } \text{REM}\_{(i-1)\text{Con}} - \text{D}\_{(i-1)\text{Con}} \le \text{SAL}\_{\text{Con}}, \text{ then} \\ \text{D}\_{(i)\text{Con}} = \text{REM}\_{(i-1)\text{Con}} - \text{SAL}\_{\text{Con}} \end{array} \tag{14c}$$

where, i ∈ N and 1 ≤ i ≤ T;

D(i)Con is the depreciation amount of the initial construction capital cost for the operating year i;

dCon is the depreciation rate, expressed as percentage (%) of the initial construction capital cost;

Con is the initial construction capital cost (see Equation (1));

SALCon is the salvage value of the initial construction capital cost; and

REMV(i)Con is the remaining value of the initial construction capital cost after depreciation in the operating year i, and is given by the below equation:

$$\begin{array}{l}\text{If } \text{Con}-\text{i} \ast \text{d}\_{\text{Con}} \ast \left(\text{Con}-\text{SAL}\_{\text{Con}}\right) > \text{SAL}\_{\text{Con}}, \text{ then} \\ \text{REM}\_{\text{(i)Con}} = \text{Con}-\text{i} \ast \text{d}\_{\text{Con}} \ast \left(\text{Con}-\text{SAL}\_{\text{Con}}\right) \end{array} \tag{15a}$$

$$\begin{array}{c} \text{If } \mathsf{Con}-\mathsf{i} \ast \mathsf{d}\_{\mathsf{Con}} \ast (\mathsf{Con}-\mathsf{SAL}\_{\mathsf{Con}}) \leq \mathsf{SAL}\_{\mathsf{Con}}, \text{ then} \\ \text{REM}\_{\mathrm{(i)}\mathrm{Con}} = \mathsf{SAL}\_{\mathsf{Con}} \end{array} \tag{15b}$$

The depreciation amount for the capital expenditure costs produced during the (i) year of operation is calculated using the same logic as above, and given according to the following equations:

$$\text{If } \mathbf{i} = \mathbf{j}, \text{ then } \mathbf{D}\_{\text{(i)} \text{CAPEX}\_{\mathbf{j}}} = \mathbf{d}\_{\text{CAPEX}\_{\mathbf{j}}} \ast \left( \mathbf{CAPEX}\_{\mathbf{j}} - \text{SAL}\_{\text{CAPEX}\_{\mathbf{j}}} \right) \tag{16a}$$

$$\begin{aligned} \text{If } \mathbf{i} > \mathbf{j} \text{ and } \mathbf{E} \mathbf{M} \mathbf{V}\_{\text{(i-1)CAPEX}\_{\uparrow}} - \mathbf{D}\_{\text{(i-1)CAPEX}\_{\uparrow}} > \mathbf{S} \mathbf{A} \mathbf{L}\_{\text{CAPEX}\_{\uparrow}} \text{ then} \\ \mathbf{D}\_{\text{(i)CAPEX}\_{\uparrow}} = \mathbf{d}\_{\text{CAPEX}\_{\uparrow}} \* \left( \mathbf{C} \mathbf{A} \mathbf{P} \mathbf{E} \mathbf{X}\_{\uparrow} - \mathbf{S} \mathbf{A} \mathbf{L}\_{\text{CAPEX}\_{\uparrow}} \right) \end{aligned} \tag{16b}$$

$$\begin{array}{l} \text{If } i > j \text{ and } \text{REM}\_{(i-1)\text{CAPEX}\_{\natural}} - \text{D}\_{(i-1)\text{CAPEX}\_{\natural}} \le \text{SAL}\_{\text{CAPEX}\_{\natural}} \text{ then} \\ \text{D}\_{(i)\text{CAPEX}\_{\natural}} = \text{REMV}\_{(i-1)\text{CAPEX}\_{\natural}} - \text{SAL}\_{\text{CAPEX}\_{\natural}} \end{array} \tag{16c}$$

where, i ∈ N, 1 ≤ i ≤ T, j ∈ N, 1 ≤ j ≤ T, and i ≥ j;

D(i)CAPEX<sup>j</sup> is the depreciation amount for operating year i of the capital cost that was produced during operating period j, i ≥ j, i ∈ N, j ∈ N;

dCAPEX<sup>j</sup> is the depreciation rate of the capital cost that was produced during operating period j, j ∈ N;

SALCAPEX<sup>j</sup> is the salvage value of the capital cost that was produced during operating period j, j ∈ N;

REMV(i)CAPEX<sup>j</sup> is the remaining value of the capital cost that was produced during operating year j, after depreciating in the operating year i, and is given by the below equations:

$$\begin{aligned} \text{If } \text{CAPEX}\_{\text{j}} - (\text{i} - \text{j}) \ast \text{d}\_{\text{CAPEX}\_{\text{j}}} \ast \left( \text{CAPEX}\_{\text{j}} - \text{SAL}\_{\text{CAPEX}\_{\text{i}}} \right) &> \text{SAL}\_{\text{CAPEX}\_{\text{j}}} \text{ then} \\ \text{REMV}\_{\text{(i)Con}} &= \text{CAPEX}\_{\text{j}} - (\text{i} - \text{j}) \ast \text{d}\_{\text{CAPEX}\_{\text{j}}} \ast \left( \text{CAPEX}\_{\text{j}} - \text{SAL}\_{\text{CAPEX}\_{\text{i}}} \right) \end{aligned} \tag{17a}$$

$$\begin{aligned} \text{If } \mathsf{CAPEX}\_{\text{j}} - (\mathsf{i} - \mathsf{j}) \ast \mathsf{d}\_{\text{CAPEX}\_{\text{j}}} \ast \left( \mathsf{CAPEX}\_{\text{j}} - \mathsf{SAL}\_{\text{CAPEX}\_{\text{j}}} \right) \leq \mathsf{SAL}\_{\text{CAPEX}\_{\text{j}}} \text{ then} \\ \text{REMV}\_{\text{(i)Con}} = \mathsf{SAL}\_{\text{CAPEX}\_{\text{j}}} \end{aligned} \tag{17b}$$

Given the Equations (14) and (16), the depreciation amount for all capital costs, including initial capital costs and annual capital costs, for each operating year is expressed as:

$$\mathbf{D}\_{\mathbf{i}} = \mathbf{D}\_{\mathrm{(i)Con}} + \sum\_{\mathbf{j}=1}^{\mathbf{i}} \mathbf{D}\_{\mathrm{(i)CAPEX\_{\mathbf{j}}}} \tag{18}$$

where, i ≥ j, i ∈ N, 1 ≤ i ≤ T j ∈ N, and 1 ≤ j ≤ T.

Similarly, according to Equations (15) and (17), the remaining value after depreciation of the whole project's capital costs for each operating year can be expressed as:

$$\text{REMV}\_{\text{i}} = \text{REMV}\_{\text{(i)Con}} + \sum\_{\text{j}=1}^{\text{i}} \text{REMV}\_{\text{(i)CAPEX}\_{\text{j}}} \tag{19}$$

The amortization consists of the loan's amortization and the intangible asset amortization. The loan's amortization is the interest-paid amount of the loan given using Equation (5). The amortization of the intangible assets is calculated in the same way used above for the calculation of depreciation. Finally, the amortization is expressed as follows:

$$\mathbf{A}\_{\mathbf{i}} = \mathbf{L}\_{\text{li}} + \mathbf{A}\_{\text{(i)Con}} + \sum\_{\mathbf{j}=1}^{\text{i}} \mathbf{A}\_{\text{(i)CAPEX}\_{\text{j}}} \tag{20}$$

where, i ≥ j, i ∈ N, 1 ≤ i ≤ T j ∈ N, and 1 ≤ j ≤ T.

#### *3.6. Income Statement*

Taking into consideration the sales revenues (Equation (7)), the operating (Equation (12)) and capital expenditures (Equation (13)), the interests (Equation (5)), the depreciation (Equation (18)), and the amortization (Equation (20)), the earnings before taxes (EBT) for the operating year i are given by:

$$\rm EBT\_i = RV\_i - PowerCost\_i - OPEX\_i - D\_i - A\_i - LI\_i \tag{21}$$

where, i ∈ N, 1 ≤ i ≤ T.

For the EBT of each year, a certain tax applied, being set by the government and local authorities. The tax is applied only when the EBT is higher than 0, and the company has earnings in that specific tax year:

$$\text{TAXES}\_{\text{i}} = \begin{cases} \text{tax}\_{\text{i}} \* \text{EBT}\_{\text{i}\prime} \text{ if } \text{EBT}\_{\text{i}} > 0\\ \quad \text{0, if } \text{EBT}\_{\text{i}} \le 0 \end{cases} \tag{22}$$

where, i ∈ N, 1 ≤ i ≤ T;

TAXES<sup>i</sup> is the tax expense for the operating year i;

tax<sup>i</sup> is the tax percentage set by the government for operating year i.

The net earnings (NE) for each year reflect the earnings of the organization after the imposed taxes, and it is an important element to evaluate the profitability of an investment because it takes into account all revenues and expenses the company has. For the operating year i, the net earnings are calculated as follows:

$$\rm{NE}\_{\rm i} = \rm{EBT}\_{\rm i} - \rm{TAXES}\_{\rm i} \tag{23}$$

where i ∈ N, 1 ≤ i ≤ T.

#### *3.7. Cash Flow*

The total cash inflows for the FSRU Station of the current study come only from the revenues we have calculated from Equation (7). As such, total cash inflows are:

$$\rm CFI\_{\dot{1}} = RV\_{\dot{1}} \tag{24}$$

where CFI<sup>i</sup> is the cash inflows for the operating year i, i ∈ N.

The total cash outflows come from the power generation costs during the regasification process (see Equation (9)), the operating (see Equation (12)) and capital expenses (see Equation (13)), and the interest (see Equation (4)) and the taxes expenses (see Equation (22)), and they are calculated by the following formula:

$$\text{CFO}\_{\text{i}} = \text{PowerCost}\_{\text{i}} + \text{OPEX}\_{\text{i}} + \text{CAPEX}\_{\text{i}} + \text{D}\_{\text{i}} + \text{A}\_{\text{i}} + \text{L}\_{\text{i}} + \text{TAXES}\_{\text{i}} \tag{25}$$

where CFO<sup>i</sup> is the cash outflows for the operating year i, ieN.

From the above Equations (24) and (25), the total cash flows are expressed by:

$$\rm CF\_i = \rm CFI\_i - \rm CFO\_i \tag{26}$$

where CF<sup>i</sup> is the cash flows for the operating year i, ieN.

#### *3.8. NPV and IRR*

To be able to conduct an economic assessment of the project, there are several financial techniques. The most common methods include Break Even Point Analysis, Payback Time Period, Net Present Value (NPV), and Internal Rate of Return (IRR). In the present paper, we have conducted the NPV and IRR analysis in order to assess the results of our study. The main purpose of the financial analysis is to use the project cash flow forecasts to calculate suitable net return indicators [26–29].

To calculate the NPV, we use the cash flows for each operating year, and then discount the stream back to the present time period. The formula for deriving NPV is:

$$\text{NPV} = \left(-\text{ICC}\right) + \sum\_{i=1}^{T} \frac{\text{CF}\_{i}}{\left(1 + \text{DR}\right)^{i}} \tag{27}$$

where i ∈ N 1 ≤ j ≤ T; and

DR is the discount rate.

Financial sustainability is ensured if NPV of cumulated cash flow is positive for all the years considered [26].

For the IRR calculation, in Equation (27), we set NPV equal to zero and DR=IRR. Then, we can solve to find the IRR:

$$\text{NPV} = \left(-\text{ICC}\right) + \sum\_{i=1}^{T} \frac{\text{CF}\_{i}}{\left(1 + \text{IRR}\right)^{i}} = 0 \tag{28}$$

The IRR should match a certain price that is set by the shareholders in order to assess positively the investment, and, at least, should be greater than the discount rate.

#### *3.9. Tax Incentives*

In this section, the key question is if, and in what level of capitals, the state's authorities could promote the investment by offering tax benefits and making the investment more attractive to potential investors. Large infrastructure projects develop, assess, and support activities providing a positive contribution to the regional economy [26–28]. This can boost the socioeconomic development of regional and remote areas of a state, such as Alexandroupolis [29–31]. Therefore, the state would favor to promote and facilitate the construction of such a project.

The methodology used includes estimating the economic benefits that the state will indirectly gain from the operation of the Station. "Indirectly" benefits include all the other sources of income the FSRU will generate, except from the taxes that will be paid to the state at the end of every tax year. Such sources can be the taxation of works and activities related to the Station, the taxation of the employees' salaries, and others. In this analysis, firstly, these activities will be pointed out, and then, the total turnover of each activity will

be estimated, and the taxation incomes will be found out. Finally, these profits that the state will gain from the FSRU's operation can be given to the FSRU as tax reliefs.

Specifically, the state will gain tax incomes during the construction period. For the construction of the Station's heavy equipment, such as the FSRU vessel or the NG pipelines, these will be supplied and imported to the country, and works and services will be performed, such as the construction of the pipeline, the dredging of the vessel, the Project's reporting, and others. All of these supplies and works will be taxed by the government. Furthermore, extra jobs will be created, and the state will be benefited by the taxation on their salaries.

During the operation period, the Station shall employ around 120 people personnel. The state gains taxes from the salaries that the FSRU pays to these employees. These taxes come not only from the direct taxation, but also from the fact that these people will spend their wages, and they will be taxed for these purchases.

Moreover, during the operating period, each year, the Station generates operating and capital expenditures, which concern the supply of goods and services which include taxes to the state. These expenses for each operating year i are the sum of OPEX<sup>i</sup> and CAPEX<sup>i</sup> (Equations (12) and (13)) minus the wages of the personnel mentioned above, and expressed from the following formula:

$$\text{FSRUEXP}\_{\text{i}} = \text{OPEX}\_{\text{i}} + \text{CAPEX}\_{\text{i}} - \text{Sal}\_{\text{i}} \tag{29}$$

where, i ∈ n, 1 ≤ i ≤ T;

FSRUEXP<sup>i</sup> is the total turnover generated by the annual operating and capital costs of the FSRU for the operating year i;

OPEX<sup>i</sup> is the operating expenses during operating year i;

CAPEX<sup>i</sup> is the capital costs during operating year i;

Sal<sup>i</sup> is the total salaries paid to the FSRU personnel during operating year i.

Finally, the Station produces some extra activities. An FSRU needs, for its maintenance and inspection, someone with technical expertise, who will probably need to come from other areas of Greece or even from abroad. These people will generate turnover by their transport, food, and accommodation expenses. Moreover, the regasification process usually lasts almost a day. This means that the LNG carriers' personnel (around 20 people) will stay for one day in the town of Alexandroupolis, spending money on food, coffee, gifts, and others. This tax income depends on the usage rate of the FSRU.

The tax income from all the above-mentioned cases can be expressed by the following equation:

$$\begin{array}{c} \text{STATEINC} = \text{Con} \ast \text{TAX}\_{\text{ICC}} + \sum\_{i=1}^{\text{T}} [\text{Sal}\_{i} \ast \text{Tax}\_{\text{Sali}} + \text{FSRUEXP}\_{i} \ast \text{L} \text{L}^{\ast}] \\ \text{Tax}\_{\text{FSRUEXPi}} + \text{Local}\_{i} \ast \text{Tax}\_{\text{Locali}}] \end{array} \tag{30}$$

where, i ∈ N, 1 ≤ i ≤ T;

STATEINC is the total income from taxes generated from FSRU during the construction and operating period, without taking into account the taxes of FSRU;

TAXICC is the total tax applied to the income generated during the construction period; Sal<sup>i</sup> is the total yearly wages paid to the FSRU personnel during the operating year i; TaxSali is the total tax applied to the salaries of the FSRU personnel for operating year i; TaxFSRUEXPi is the tax applied to the operating and capital expenditures of the FSRU during the operating year i;

Local<sup>i</sup> is the total income generated for local society during operating year i; TaxLocali is the tax applied to the income for local society during operating year i.

After estimating the total income from the taxes, the state may decide whether it is in its interest to proceed with offering tax incentives. Then, the taxes of the FSRU are given by Equation (22), modified according to the below formula:

$$\text{TAXES}\_{\text{i}} = \begin{cases} \text{tax}\_{\text{i}} \star \text{EBT}\_{\text{i}} - \text{TAXrelief}\_{\text{i}} & \text{if } \text{EBT}\_{\text{i}} > 0 \\ & 0, \text{ if } \text{EBT}\_{\text{i}} \le 0 \end{cases} \tag{31}$$

where, TAXrelief<sup>i</sup> is the tax relief that the state offers for operating year i. Finally, the NE, NPV, and IRR are re-adjusted according to Equation (31).

#### **4. Numerical Application Results**

Based on above modelling framework for the evaluation of an FSRU Project, the numerical application deals with the development of the new Alexandroupolis FSRU Project. The primary referenced data that relate to the specific characteristics of a FSRU Station, such as basic operating and construction costs, are retrieved from "The Outlook for Floating Storage and Regasification Units (FSRUs)" and "Floating LNG Update-Liquefaction and Import Terminals", both published by the Oxford University for Energy Studies [32,33]. We have also used data from the operating OLT Offshore LNG Toscana FSRU Station located in Toscana, Italy [34].

Subsequently, a sensitivity analysis subject to the FSRU usage rate has to be conducted supporting decisions regarding the authorities' taxation, and in what range a tax reduction scheme or mitigation policies could be increase the capital leverage, providing a win–win risk mitigation scheme between authorities and private funds.

#### *4.1. Economic Assessment of the Alexandroupolis FSRU Station*

Economic analysis provides decision-makers with a way of assessing the overall value of a project, including the financial viability and the investment's likely productivity and effectiveness indexes [29,30].

The FSRU of Alexandroupolis is the first FSRU Station being constructed in the Greece territory, characterized as a totally greenfield project, and there are not existing baseline scenarios or data to be used for the economic analysis. The available data have already been addressed, mainly related to FSRU key operational characteristics, such as the Vessel's capacity. Therefore, the analysis structure developed by analyzing all the necessary factors and rates is necessary to be taken into consideration to assess the economic output of the project. For the readers and users of the proposed methodology framework, the formulation provides a dynamic modelling formulation structure, where key assessment variables are adjusted over the project lifecycle, the technical features, the financial market conditions, and the project features.

For the calculations presented in this paper, most of the assessment variables have been retrieved from the literature and relevant bibliography [32]. However, key assumptions are taken, providing the analysis hypothesis, based on the nature of the Station and the local characteristics, without reducing the value of the methodology, the modeling structure, and the application outputs.

#### 4.1.1. Case Study Assumptions

The values we used to run this analysis are:


#### 4.1.2. Case Study Results

Applying the above data and values, we get the economic results. The most important of them are presented in Table 1.


**Table 1.** Results of economic assessment.

Continuing the analysis, we proceed with the calculation of the economic indicators of NPV and IRR, using the Equations (27) and (28). We set the discount rate at 4.9%, as was given for Greece from the European Commission for 1 January 2022, and we consider that the construction period will last for 2 years. Given the above, the NPV and IRR are calculated as follows:


decommissioning period also.

in relation to various parameters of the FSRU.

*4.2. Sensitivity Analysis* 

4.2.1. Usage Rate

Figure 2 depicts the NPV for the Alexandroupolis FSRU.

**Figure 2.** NPV for the Alexandroupolis FSRU Project. **Figure 2.** NPV for the Alexandroupolis FSRU Project.

Finally, with the help of Equations (15) and (17), the remaining value of the assets of the Station are calculated at 168,884,352 €. Finally, with the help of Equations (15) and (17), the remaining value of the assets of the Station are calculated at 168,884,352 €.

Discussing the results of the analysis, we can ascertain that the economic output of the Project, given the assumptions we made, can definitely be positive and profitable. The

Evaluating the assessment using the economic indicators of NPV and IRR, the Project is also characterized as profitable and viable. The NPV at the end of operation period is not only positive, but has reached around 214 million €, which is 54.55% of the ICC. This means that, even estimating the discount rate of money for all 20 years of operation, the investment can give a high percentage of profit. As far as it concerns the IRR, it reaches 9.46%, which is 1.93 times higher than the discount rate of Greece (4.9%) at this time.

Last but not least, the remaining value of the assets at the end of the operation period, including both initial capital cost and capital costs during each year of operation, is still high enough, giving the investors the opportunity to gain high profits during the

The above results were calculated using a relatively small usage rate of 42% of the FSRU. This means that there are bigger profit margins if this rate increases. On the other hand, if it drops even more, there is a doubt if the Project can remain economically viable or not. The same questions will be raised if other rather unstable parameters, such as USD/EUR exchange rate or tariff price, change. In this section, a sensitivity analysis is conducted to fully understand and highlight the dependence of the Project's profitability

Implementing infrastructure projects, the demand curves are essential and critical to assess feasibility and economic viability. Therefore, the sensitivity analysis should keep all the values stable, and change the demand of the Station, which is expressed by the usage rate. Applying a series of usage rates, the NPV, IRR, and net earnings using the methodology of Section 3 are calculated. The usage rate that was adopted during the presented (basic) scenario, based on values resulting from a market sounding survey

Discussing the results of the analysis, we can ascertain that the economic output of the Project, given the assumptions we made, can definitely be positive and profitable. The net earnings, even after the taxes and expenses, can attribute to a very high profitability, which makes the project attractive enough to investors. The NE of the 1st year of operation reaches 12.46% of the ICC. This means that in only a few years, the ICC can be recuperated.

Evaluating the assessment using the economic indicators of NPV and IRR, the Project is also characterized as profitable and viable. The NPV at the end of operation period is not only positive, but has reached around 214 million €, which is 54.55% of the ICC. This means that, even estimating the discount rate of money for all 20 years of operation, the investment can give a high percentage of profit. As far as it concerns the IRR, it reaches 9.46%, which is 1.93 times higher than the discount rate of Greece (4.9%) at this time.

Last but not least, the remaining value of the assets at the end of the operation period, including both initial capital cost and capital costs during each year of operation, is still high enough, giving the investors the opportunity to gain high profits during the decommissioning period also.

#### *4.2. Sensitivity Analysis*

The above results were calculated using a relatively small usage rate of 42% of the FSRU. This means that there are bigger profit margins if this rate increases. On the other hand, if it drops even more, there is a doubt if the Project can remain economically viable or not. The same questions will be raised if other rather unstable parameters, such as USD/EUR exchange rate or tariff price, change. In this section, a sensitivity analysis is conducted to fully understand and highlight the dependence of the Project's profitability in relation to various parameters of the FSRU.

#### 4.2.1. Usage Rate

Implementing infrastructure projects, the demand curves are essential and critical to assess feasibility and economic viability. Therefore, the sensitivity analysis should keep all the values stable, and change the demand of the Station, which is expressed by the usage rate. Applying a series of usage rates, the NPV, IRR, and net earnings using the methodology of Section 3 are calculated. The usage rate that was adopted during the presented (basic) scenario, based on values resulting from a market sounding survey conducted before the COVID-19 pandemic by the FSRU investors, constitute the most reliable scenario. Furthermore, higher demand scenarios, even it seems too optimistic for a greenfield energy infrastructure project, are considered in this numerical application, taken into account the new conditions in European energy market where new infrastructure for LNG supply is in major attention. However, the recent conflicts in North-East Europe where Russia is acting may provide a great opportunity for new business, and generate additional demand.

Initially, in the analysis, the hypothesis of a light volatility around the regular scenario's usage rate value was taken into consideration. Assuming that the volatility will be 20%, the values of usage rate to be examined are 33.6% for the high-risk scenario, and 50.4% for the positive market reaction scenario. Furthermore, the calculation assumed the hypothesis of a high increase for LNG demand that increases the demand 1.5 times higher than the regular scenario demand, forming the usage rate at 63%.

The results issued for each scenario are presented in Table 2.

Euro).

LNG High Demand Increase—


**Table 2.** Results of demand-usage rate sensitivity analysis. than the regular scenario demand, forming the usage rate at 63%. The results issued for each scenario are presented in Table 2.

hypothesis of a high increase for LNG demand that increases the demand 1.5 times higher

conducted before the COVID-19 pandemic by the FSRU investors, constitute the most reliable scenario. Furthermore, higher demand scenarios, even it seems too optimistic for a greenfield energy infrastructure project, are considered in this numerical application, taken into account the new conditions in European energy market where new infrastructure for LNG supply is in major attention. However, the recent conflicts in North-East Europe where Russia is acting may provide a great opportunity for new

Initially, in the analysis, the hypothesis of a light volatility around the regular scenario's usage rate value was taken into consideration. Assuming that the volatility will

*Energies* **2022**, *15*, x FOR PEER REVIEW 15 of 22

business, and generate additional demand.

Moreover, Figure 3 presents the NPV during the operation period for each scenario. Moreover, Figure 3 presents the NPV during the operation period for each scenario.

**Figure 3.** NPV during operation period for alternative usage rate scenarios. (NPV in millions **Figure 3.** NPV during operation period for alternative usage rate scenarios. (NPV in millions Euro).

Assessing the above results, it can be highlighted that the project profitability is very sensitive to the demand curves (usage rate). A small increase in the demand annual volume from the level of 42% (regular scenario) to 50.4% can improve essential the IRR to Assessing the above results, it can be highlighted that the project profitability is very sensitive to the demand curves (usage rate). A small increase in the demand annual volume from the level of 42% (regular scenario) to 50.4% can improve essential the IRR to a level above 12%, stimulating the investors' interest. Furthermore, in the LNG high-demand scenario, where the demand increases to a 63% usage rate, the IRR reached the level of 15.6%. On the contrary, in the high-risk scenario, where the demand reduces to 33.6%, the IRR can drop to levels close to the discount rate, leading the investment to be non-profitable enough. It is noteworthy that although the change in demand is the same (20%) in both the high-risk and positive market scenario, the variation of IRR is bigger when the demand decreases (34.9% to regular scenario), compared to when the demand increases (29.8% to regular scenario).

> The same conclusions also apply for the NPV. In the LNG high-demand scenario, NPV turns positive after only 6 years, which in an operation period of 20 years, can lead to extremely high profitability. In the case of the high-risk scenario, where a small decrease occurs, NPV remains negative for a long period, until it turns positive near the end of operation period at around 17 years.

#### 4.2.2. USD/EUR Exchange Rate

For the second part of the sensitivity analysis, the fluctuation of the USD to Euro Exchange Rate (XR) is taken into consideration in the analysis. This rate is affected by exogenous issues to the project e.g. accidents, political instability, etc.) and changes in energy market result higher price volatility on one hand; and higher construction and operational cost on the other.

The exchange rate values for this analysis tracked, by extrapolate, the historical data of XR from the commencement of the Euro circulation in 2000 until 2021, where:


As we can notice, the average rate (0.83) is close enough to the rate that we used during our regular scenario (0.87). The results for each scenario are presented in Table 3, and the NPV during the operation period is given in Figure 4.


**Table 3.** Results of USD/EUR exchange rate sensitivity analysis.

**Figure 4.** NPV during operation period for alternative USD/EUR exchange rate scenarios. (NPV in million Euro).

Appreciating the above data, we can comprehend that the Project's economic rates are volatile enough to the XR's changes. Both IRR and NPV fluctuate in large ranges that could threaten the Project's viability. The IRR ranges between 5.51% and 13.70%, whereas the NPV ranges from 32 mil € to over 362 mil €. Especially in the high XR scenario, the NPV remains negative for almost all of the operation period of the Project. Decrease 30% 3.58% −61.33% −5,105,148,745 **Figure 4.** NPV during operation period for alternative USD/EUR exchange rate scenarios. (NPV inmillion Euro). Appreciating the above data, we can comprehend that the Project's economic rates are volatile enough to the XR's changes. Both IRR and NPV fluctuate in large ranges that could threaten the Project's viability. The IRR ranges between 5.51% and 13.70%, whereas the NPV ranges from 32 mil € to over 362 mil €. Especially in the high XR scenario, the NPV remains negative for almost all of the operation period of the Project.

such as the competition, the demand, etc. In this analysis we try to set the boundaries where the tariff price may range, and where the Project will still be economically viable. This is why we have chosen decreased values compared to the regular scenario. The used

The results for each scenario are given in Table 4, whereas the NPV during the

**Regular Scenario NPV (€)** 

tariff price's values are for a 10% decrease, 20% decrease, and 30% decrease.

Regular scenario 9.46% - 214,463,908 Decrease 10% 7.72% −16.66% 125,598,776 Decrease 20% 5.79% −37.48% 37,453,644

4.2.3. Tariff Price—Sensitivity Analysis

operation period is depicted in Figure 5.

**Scenario of Tariff Price IRR Change of IRR Compared to** 

**Table 4.** Results of tariff price sensitivity analysis.

#### 4.2.3. Tariff Price—Sensitivity Analysis

Finally, a tariff price sensitivity analysis is conducted. The tariff is a parameter that, although it is determined by the FSRU, also takes into consideration exogenous factors, such as the competition, the demand, etc. In this analysis we try to set the boundaries where the tariff price may range, and where the Project will still be economically viable. This is why we have chosen decreased values compared to the regular scenario. The used tariff price's values are for a 10% decrease, 20% decrease, and 30% decrease.

The results for each scenario are given in Table 4, whereas the NPV during the operation period is depicted in Figure 5.


#### **Table 4.** Results of tariff price sensitivity analysis.

**Figure 5.** NPV during operation period for alternative tariff price scenarios. **Figure 5.** NPV during operation period for alternative tariff price scenarios.

As we see from the analysis's results, in the first scenario, although the decrease is in the range of 10%, the Project can still produce satisfied earnings with 7.72% IRR. When the decrease goes to 20%, the NPV remains marginally positive, but with an extremely small amount of earnings. On the contrary, when the tariff's decrease reaches 30%, the As we see from the analysis's results, in the first scenario, although the decrease is in the range of 10%, the Project can still produce satisfied earnings with 7.72% IRR. When the decrease goes to 20%, the NPV remains marginally positive, but with an extremely small amount of earnings. On the contrary, when the tariff's decrease reaches 30%, the Project turns undoubtedly unviable.

Project turns undoubtedly unviable. The analysis proves that the FSRU may decrease its tariff price until around 10% and still maintain satisfactory earnings. If the external circumstances require a bigger decrease, then the Project will either lose much more of its economic efficiency or will be The analysis proves that the FSRU may decrease its tariff price until around 10% and still maintain satisfactory earnings. If the external circumstances require a bigger decrease, then the Project will either lose much more of its economic efficiency or will be economically unviable.

economically unviable. The extremely high sensitivity of the investment to all the checked parameters The extremely high sensitivity of the investment to all the checked parameters indicates that before the final investment decision is taken, the shareholders group must acquire not

• The construction cost has been set at 314,496,000 €, as mentioned above in Section 4.1. • The tax rate during the construction can be set at 22%, which was the tax rate for

• The FSRU employs around 120 people. If we consider that the average wage will be 1500 € per month, which is an average technical wage for Greece, then the total yearly salaries of FSRU shall be 2,520,000 €. In this amount, we have also included two extra

• The tax rate for the wage of 1500 € in Greece is around 25%. Assuming also that the employees will spend 2/3 of their income on purchases and services, they will generate extra turnover which is taxed by 24%. This way, they produce extra taxes in

Based on the formulation of Section 3.9, the extra tax income the state could gain from the FSRU Project is estimated by using Equation (30) under the following data and

external factors, such as the USD/EUR exchange rate, are in favor of the Project.

*4.3. Project Incentives* 

companies in Greece for 2021.

wages for the Christmas and summer periods.

assumptions:

indicates that before the final investment decision is taken, the shareholders group must acquire not only all the necessary data that concern the FSRU, but also be sure that the only all the necessary data that concern the FSRU, but also be sure that the external factors, such as the USD/EUR exchange rate, are in favor of the Project.

#### *4.3. Project Incentives*

Based on the formulation of Section 3.9, the extra tax income the state could gain from the FSRU Project is estimated by using Equation (30) under the following data and assumptions:


Applying these data to Equation (30), the total state income from these activities is 116,349,082 €, as shown in Table 5.


**Table 5.** Taxes generated by FSRU Project over years.

The state should carefully consider how to use this amount. The state may decide to give the entire amount back to the FSRU, or a big or a small part of it. It may decide to give it back all at once during the first years of operation, or give it back gradually with a small amount every year. Furthermore, the state, knowing that they will gain extra income, may decide to finance the Project, or to become a shareholder by acquiring a percentage of the investment.

In the analysis, the project implementation incentives by the state in terms of tax relief over the years, are examined. This way, if the operating period is 20 years, the state may offer a tax relief of 5,817,454 € for each operating year, and the taxes that the FSRU must pay are given by Equation (31). Using these taxes, the NPV and IRR are recalculated using the equations of Section 3.8, and the adjusted results reach the values of 280,909,698 € for NPV and 10.67% for IRR. Compared to the results of the regular scenario, the increase is 30.98% for NPV and 12.79% for IRR, respectively.

#### **5. Discussion**

When it comes to investing in an energy project, the decision is not easy, and other than the technical restrictions and environmental considerations, a series of business and economic risks need to be taken into consideration. In real-life applications, a group of alternative scenarios needs to be appropriately developed to evaluate project risk, and promote measures to mitigate them. In this frame, the paper structure and the application results presented to support decisions and assess risks over the project lifecycle have taken into consideration changes driven by operational and external events. Based on the proposed analysis developed in this paper, the framework for developing alternative scenarios is given towards decisions on equity funding schemes and the added value on assets over time.

To mitigate the investment risk and to increase the financing leverage, an introduction of governmental or authorities' incentives are discussed. Taking into consideration the investment condition of the FSRU Alexandroupolis case study, a tax substitute incentive measure was discussed as a risk-sharing tool to boost the project implementation. The numerical application results could be essential drivers for decisions in similar cases, and illustrate comparisons with other case studies.

Finally, considering the geopolitical conditions and the new LNG energy strategy discussed in Europe, and the implementation of flexible offshore LNG terminal facilities in the European territory, it shall be extremely interesting to evaluate their economic performance, providing results on the viability and investment attractiveness of such projects in the European continent. The existence of real data also assists us to get an accurate economic result. Moreover, this framework application shall be useful to the research community, getting it compared to similar projects from all over the globe.

#### **6. Conclusions**

This paper deals with the feasibility appraisal framework from an economic point of view, dedicated to FSRU greenfield (new) energy projects. The decisions for such projects vary from a very short period up to many decades, due to the complexities driven by the stakeholders' expectations, the long-term payback period, the availability of capital, and the energy demand volatility raising business risks. The appraisal modelling structure provides an easy-to-handle tool to support energy project initiatives and prioritize energy investment decisions.

The modelling structure set a dynamic project appraisal modeling framework for the economic assessment of the FSRU projects that could support the involved parties to initiate project implementation. The proposed modelling structure is recommended to be used even for evaluating the project feasibility in the planning (initiative) stage even as a tool for monitoring project performance over time. The methodology section set an analytical, coherent, detailed, easy-to-use modeling framework dedicated for new FSRU projects. By breakdown, the project implementation, delivery, and operational characteristics of all the variables dealing with the project economic performance are taken into consideration. The modelling structure is depicted appropriately for planners and managers to support analysis in both: (a) to assess the investment, and comprehend and evaluate how each variable contributes to the project economic performance over time; and (b) to strategically appraise the future course of the project or to re-assess the performance into changed energy market (demand) or financial (investment) conditions.

Furthermore, the results of a sensitivity analysis are analyzed to discuss the level of dependency of FSRU projects to the demand and financing parameters, providing key messages regarding the FSRU projects risks, viability, and attractiveness. It is noteworthy that in terms of business risk mitigation, some key incentives are evaluated and highlighted, illustrating a risk-sharing mechanism that may increase the investment attractiveness, or share risk and benefits between involved parties.

The numerical case study is the Alexandroupolis FSRU Project, and we have pointed out the conditions and the frame in which this investment can become viable and profitable, considering the Project's internal and external conditions representing the appraisal modelling variables. The numerical application evaluates the project performance for a basic scenario, promoting key messages to decision makers regarding the project risks and estimations uncertainty. The sensitivity analysis evaluating the project outputs over time for different demand scenarios and project financing features (equity mix, currency and exchange volatility) highlighting key messages for similar projects. Nevertheless, the case study outputs could provide an essential background for comparisons in similar projects, and illustrate key messages to decision-makers for the feasibility of such a project due to energy market turbulences affecting Europe and other regions. The growing opportunities, and the investment attractiveness for implementing new LNG projects due to the North-East Europe conflicts, is an interesting issue, which is highly recommended for further investigation. The governmental incentives and the conditions to be affected to mitigate the investment risks should also be an interesting issue for dedicated research analysis.

This research received no external funding.

**Author Contributions:** Conceptualization, D.D.; methodology, D.D. and P.Z.; software, P.Z.; validation, D.D. and P.Z.; formal analysis, D.D.; investigation, P.Z.; resources, D.D.; data curation, P.Z.; writing—original draft preparation, P.Z.; writing—review and editing, D.D.; visualization, P.Z.; supervision, D.D.; project administration, D.D.; funding acquisition, D.D. 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:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The paper uses a conceptual analysis framework, case study inputs, and analysis outputs developed in the research project "ENIRISST–Intelligent Research Infrastructure for Shipping, Supply Chain, Transport and Logistics", implemented in the Action "Reinforcement of the Research and Innovation Infrastructure", funded by the Operational Programme "Competitiveness, Entrepreneurship and Innovation" (NSRF 2014–2020), and co-financed by Greece and the European Regional Development Fund.

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

#### **References**


#### *Review* **Oil Prices and the Hydrocarbon Markets: A Review**

**Ali Jadidzadeh 1,2,†, Mobin Mirzababaei 1,† and Apostolos Serletis 3,\* ,†**


† These authors contributed equally to this work.

**Abstract:** In this paper, we review the literature that investigates the impact of oil price shocks on the selected hydrocarbon prices. First, we present the empirical studies that presume, due to the global nature of the crude oil market, that the world oil price is an exogenous determining variable to the evolution of the local hydrocarbon markets such as natural gas or natural gas liquids (NGLs). Then, we present recent empirical studies that have improved our understanding of the source of oil price changes. They treat the real price of oil as an endogenous variable, identify the causes underlying oil price shocks, and then evaluate the impact of structural supply and demand shocks on the other hydrocarbon prices. The first strand of studies does not represent a consensus on the relationship between crude oil and other hydrocarbon prices—some demonstrate stable and asymmetric relationships, and some find no relationship or a very weak relationship. The second strand of studies shows that oil supply-side shocks have a transitory and temporary impact while oil demand-side shocks have a persistent and permanent impact on other hydrocarbon prices. In addition, it shows that the structural shocks in the global crude oil market explain approximately 50% of the variation in the other hydrocarbon markets in the long run.

**Keywords:** oil price shocks; hydrocarbon markets; transmission channels; asymmetries

#### **Citation:** Jadidzadeh, A.;

Mirzababaei, M.; Serletis, A. Oil Prices and the Hydrocarbon Markets: A Review. *Energies* **2022**, *15*, 6192. https://doi.org/10.3390/en15176192

Academic Editors: Bernard Zi˛ebicki and Edyta Bieli ´nska-Dusza

Received: 12 July 2022 Accepted: 22 August 2022 Published: 25 August 2022

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

**Copyright:** © 2022 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/).

#### **1. Introduction**

Since the second oil shock of 1979–1981 and the consequent increased oil price volatility, energy market stakeholders and researchers have paid more attention to how energy commodity prices are linked to oil price fluctuations. In the last two decades, the expansion of unconventional oil from shale and tar sands has also challenged many underlying beliefs that oil prices are the main driver of other hydrocarbon prices. For example, the crude oil and natural gas prices in the United States were moving together, but they have started to decouple episodically since 2000—see Jadidzadeh and Serletis [1] for more detail. Since then, the improvements in both theoretical and empirical studies modeling the sources of oil price shocks have evolved our understanding of the interaction between integrated *global* oil markets and *local* hydrocarbon markets.

Understanding the behavior of price volatility in hydrocarbon markets is significant for several reasons. Persistent changes in volatility can affect the risk exposure of energy producers and investors and alter their incentives to invest in natural gas and oil sectors and facilities for production and transportation. Likewise, volatility is a key determinant of the value of commodity-based contingent claims, whether financial or real. In addition, based on the recent trends toward increasing global trade in petroleum products, Zavaleta et al. [2] discuss that mainly due to a significant increase in global refining capacity utilization the global petroleum refining industry is moving from a primarily local to a global market structure. Thus, reviewing the studies that investigate the determinants of volatilities in hydrocarbon markets is important for derivative valuation, hedging decisions, decisions to invest in physical capital concerning the production or consumption of

hydrocarbons, and undertaking plausible policy decisions on key energy commodities by policy makers [3].

Therefore, this article aims to present a selective review of the studies on the effect of oil price shocks on selected hydrocarbon prices, primarily natural gas and natural gas liquids (namely, ethane, propane, normal butane, iso-butane, and natural gasoline). As most literature on the effect of oil price shocks on hydrocarbon prices has employed U.S. data that provide long historical time-series data to test and run econometric models, we focus mainly on the U.S. hydrocarbon markets. We first discuss the channels through which crude oil price shocks impact hydrocarbon prices and then present the empirical econometric methods that explore the relationship between oil price shocks and the hydrocarbon prices based on the economic theories.

The economic theories that explain the relationship between oil and hydrocarbon prices consider either substitutability between oil and hydrocarbons (a demand-side linkage) or the way that hydrocarbons are extracted from oil and gas (a supply-side linkage). With regard to the relationship between oil and natural gas liquids (NGLs), it should be noted that crude oil is not consumed directly, but it is the main source of hydrocarbons, such as methane, ethane, propane, butane, pentane, and various other paraffinic, naphthenic, and aromatic hydrocarbons which are derived from refining crude oil. Similarly, the raw natural gas extracted from gas wells is refined to produce the end products including methane and natural gas liquids (NGLs). Methane, delivered by pipelines, is the main source of heating and burning in the residential, commercial, and industrial sectors, and NGLs, stored and shipped in a liquid state, are used as feedstocks in petrochemical plants, space heating, motor gasoline components, and other fuel use. Therefore, it is the demand for refinery and petrochemical end-products that drives the demand for crude oil and natural gas. Therefore, the shocks in oil and gas markets lead to changes in the production costs and, consequently, the price of NGLs. In addition, Oglend et al. [4] argue that the oil and natural gas markets are also related through the state of the NGLs markets. The liquids' prices are determined in a market that is indexed to oil prices with long-term contracts. Therefore, an increase in oil prices might raise NGL prices.This might make natural gas producers increase their production, as NGLs are a key source of income for them, consequently lowering natural gas prices.

About the relationship between the crude oil market and the natural gas market, on the demand side, oil and gas compete indirectly with their derivatives. Methane (a natural gas derivative) and heating oil (a crude oil derivative) compete in the heating markets of residential and commercial sectors. Methane is also a substitute for the residual and distillate fuels in electricity generation and industrial plants. For example, Hartley et al. [5] argue that the installation of advanced combined cycle gas turbine (CCGT) power plants has increased the demand for natural gas relative to residual fuel oil in the United States. On the supply side, associated natural gas is a coproduct of crude oil as it occurs in oil reservoirs, whereas the non-associated natural gas is not in contact with crude oil and is found in natural gas reservoirs. Although associated natural gas accounts for a small fraction of natural gas production in the United States, it is also utilized to extract NGLs in oil refineries. In addition, recent technological innovations in horizontal drilling and hydraulic fracturing have increased the production of both shale oil and shale gas.

There are a large number of empirical studies that sought to investigate whether the price of crude oil was an important determinant of the natural gas price. However, less attention has been paid in the literature to the relationship between other hydrocarbon products and the oil market. Most of this literature employs time series models and can be categorized into two strands. The first strand of studies investigates the existence of correlation and long-run cointegrating relationships between oil and hydrocarbon prices using different types of time series models such as Vector Error Correction Models or Vector AutoRegressive (VAR) models. These studies mostly treat the price of oil as exogenous and investigate the response of other hydrocarbon prices to exogenous changes in the oil price. Basically, this literature shows that the relationship between the oil price and the natural gas or NGL prices is asymmetric in a way that the oil price predominantly drives the other hydrocarbon prices, but not the other way around. The second strand of studies treats the price of crude oil as endogenous and disentangles the causes underlying oil price shocks. This recent empirical literature, which is built on Kilian [6] employs a structural VAR model to identify the source of oil price fluctuations as a result of supply disruptions and demand changes in the integrated global oil market. They augment Kilian's structural VAR model to include the real price of another hydrocarbon and investigate the response of other hydrocarbon prices to the structural shocks in the oil market.

The paper is organized as follows. Section 2 discusses the channels through which oil price technically impacts other hydrocarbon prices. Section 3 reviews the literature that investigates the relationship between crude oil and other hydrocarbon markets, and the final section concludes the paper.

#### **2. Transmission Channels between Oil and Hydrocarbon Prices**

As illustrated in Figure 1, raw natural gas is extracted either from gas reservoirs or oil reservoirs. The former is called non-associated natural gas or wet gas, and the latter is called associated-dissolved natural gas. The associated natural gas is delivered to oil refineries to extract hydrocarbon gas liquids (HGLs). The HGLs from oil refineries are usually converted to olefins such as ethylene, propylene, normal butylene, and isobutylene to make resins, plastics, and adhesives in petrochemical plants. However, the non-associated gas is delivered to a gas processing plant and then to other facilities, namely, fractionators, to separate the majority of the hydrocarbon gas liquids (HGL) from the gas stream and to apply additional separation to meet certain commercial specifications. The end-products are natural gas liquids (NGLs), including ethane propane, normal butane, isobutane, and natural gasoline, which are abbreviated as C2, C3, NC4, IC4, and C5, respectively. The liquidation of NGLs makes possible their storage and delivery from remote producing areas, with oil-indexed prices and long-term contracts, to large consuming areas—see Jadidzadeh and Serletis [7], Wamsley [8], and the HGLs page by the U.S. Energy Information Administration [9] for more detail.

Crude oil is traded in a global marketplace where many international buyers and sellers are absent in a geographically proximate situation. However, markets for natural gas and refined petroleum products are typically considered national or even regional. Specifically, there are three major regional markets, including the North American, European, and East Asian trading hubs, where the prices of NGLs are determined. Three channels link NGL prices to oil prices. First, NGLs are either the direct product of crude oil or are used as blendstocks in petroleum products. For example, although a considerable share of NGLs is extracted from natural gas at the fractionators, a large amount originates from the crude oil at the refineries. Since 2000, the amount of U.S. HGLs production from petroleum refineries has remained almost unchanged at about 0.55 million barrels per day, while the production from natural gas processing plants has increased from 2.16 to 5.72 million barrels per day in 2020. Gasoline, which is the main product of crude oil, needs to be mixed with natural gasoline and normal butane to enhance combustion. So, the channel that the crude oil prices impact natural gasoline and normal butane through is the interplay of gasoline supply and demand. Second, the market structure of NGLs links the NGL contracts to oil prices. There are many trading gas hubs all over the world where NGL prices are determined by oil-indexation mechanisms that tie NGL prices to oil prices, compared to the gas-on-gas completion mechanism, where NGL prices are determined based on their supply and demand interactions [10]. Third, crude oil products such as heating oil compete with methane, propane, and butane through their utilization for fuel and heating. Furthermore, the development of the Liquefied Petroleum Gas (LPG) markets has strengthened the connection between the oil and gas markets. So, any change in oil or gas prices is expected to change the NGL prices either directly or indirectly.

**Figure 1.** Sources of hydrocarbon gas liquids (HGL) and Natural gas liquids (NGL).

#### **3. Empirical Models**

In this section, we first review the studies that presume that the real price of oil is exogenous with respect to economic fundamentals, and supply disruptions, such as political events in the Middle East, only drive the oil price shocks. The key insight in them is that the relationship between oil price and other hydrocarbons is evaluated on a *ceteris paribus* basis—i.e., the oil price varies while all other variables are held constant. Then, we present recent empirical studies that are mostly built on Kilian [6] and treat the real price of oil as an endogenous variable. They believe that oil price fluctuations do not occur *ceteris paribus*. As a result, the volatility in the integrated global oil market is driven by both supply and demand shocks. They disentangle the source of price changes and then evaluate the impact of supply and demand shocks on the other hydrocarbon prices.

#### *3.1. Oil Price as an Exogenous Variable*

This line of study relies on the global nature of the crude oil market and presumes that large variations in the price of oil are historically driven by supply disruption as a result of political events and OPEC decisions. It assumes that oil prices are exogenous to macroeconomics variables and discards the source of oil price shocks in their analyses. Among those, some studies, such as Wang et al. [11], Brown and Yücel [12], Panagiotidis and Rutledge [13], Pindyck [3], Serletis and Rangel-Ruiz [14], and Serletis and Herbert [15], show that there is a stable and asymmetric relationship between the crude oil price and natural gas price in a way that the oil price dominantly derives the natural gas price, but not the other way around. On the contrary, other studies such as Bachmeier and Griffin [16], Drachal [17], and Zhang and Ji [18] show that, in the long run, there is no relationship or a very weak relationship between the prices of crude oil and natural gas. Serletis and Rangel-Ruiz [14] show that there are no shared stochastic trends between West Texas Intermediate (WTI) crude oil prices and Henry Hub natural gas. They attribute this to the deregulations

of the natural gas market in the United States. Hailemariam and Smyth [19] argue that since the shale gas revolution in 2008 the price of natural gas has been determined based on its supply and demand in the natural gas market. In the short run, studies such as Hartley et al. [5] and Brown and Yücel [12] show that seasonal factors, such as weather and natural disasters, and oil inventories have a significant impact on natural gas prices, which makes them deviate from oil prices.

Yücel and Guo [20] is one of the first studies that investigated a cointegrating relationship between oil, natural gas, and coal prices from 1947 to 1990 in the U.S. They show that oil and natural gas are not cointegrated when using the full sample because the natural gas market was heavily regulated before 1974. However, their results verify that the cointegration exists episodically in sub-samples. For example, one can find a long-run relationship between oil and natural gas prices for the 1974–1990 time span in a way that the oil price drives natural gas prices—see also Hou and Nguyen [21], Hailemariam and Smyth [19], and Aruga [22], who verify Yücel and Guo's results over the 1974–1990 period. Similarly, Bachmeier and Griffin [16]'s study does not support the idea of a primary energy market as they show that oil, natural gas, and coal are very weakly integrated in the U.S. They argue that the end products of oil and natural gas are used in different sectors, while in the short-run, little possibilities of interfuel substitution exist. For example, oil and natural gas products, mostly heating oil and methane, can compete in residential and commercial heating or in power plants, while gasoline, which is mostly produced from oil, does not have a considerable successor to natural gas in the transportation sector. In contrast, Barcella [23] argues that fuel oil and natural gas are significant substitutes in the electric power sector and that is why crude oil and natural gas prices are highly correlated in the liberalized U.S. gas market. Therefore, the state of the transmission channels, discussed in the previous sections, is used to explain how the hydrocarbon prices are integrated with the crude oil prices. They attribute either changes in the state of market structure and regulations (i.e., oil-indexed versus gas-on-gas contracts), supply-side interruptions and developments, or demand-side changes in oil and gas derivatives to co-movements in the prices.

The liberalization of natural gas markets refers to the deregulation and structural reforms that lead to the gas-on-gas competition in liquid spot and futures markets rather than long-term contracts with prices determined in advance and frequently linked to the crude oil or heavy fuel price (a system known as oil-indexation). Some studies such as Barton and Vermeire [24] and Heren [25] believe that the liberalization of gas markets in the U.K. has weakened the linkage between crude oil and gas prices since the spot and futures markets were first established in 1995–1996. Asche et al. [26] employ cointegration models and show that the U.K. natural gas market and Brent oil are integrated in an interim period (1995–1998) when the U.K. gas market was liberalized in 1995 but not yet linked to the Continental gas markets through the U.K.–Mainland Europe Interconnector gas pipeline at the end of 1998. They do not find any evidence of cointegration for the period after opening the Interconnector between 1998 and 2002. In contrast to Asche et al. [26], Panagiotidis and Rutledge [13] and Drachal [17] show that a long-run equilibrium relationship between U.K. gas and oil prices exists over the period 1996–2003. Panagiotidis and Rutledge [13] believe that the differences in results with Asche et al. [26] could stem from the inclusion of the electricity price in their study. These studies do not represent a general consensus on the relationship between crude oil and natural gas prices. This lack of consensus across studies is mainly due to either methodological or some other factors such as data, time period, and location of the study.

In recent years, some studies employ new empirical techniques to address the spillover effects between natural gas and crude oil markets in different time frequencies—see Batten et al. [27], Ji et al. [28], Perifanis and Dagoumas [29], Zhu et al. [30], Wang et al. [11], Li et al. [31], and Lovcha and Perez-Laborda [32]. To investigate the information or volatility spillover effect of the markets, they use more explanatory variables (e.g., historical events or disasters) with high-frequency data (e.g., daily data) from the financial sector. Similar to the studies that employ cointegration techniques, there is no consensus on the results of

these studies. Many of them [27,29,30,32] argue that since the shale gas revolution in 2008, the spillover effect from the oil market to the gas market has disappeared or accounts for very small changes in the U.S., while before the shale gas revolution, the spillover effect was present episodically in some periods from the oil market to the natural gas market, or in some cases the effect was reversed. Wang et al. [11] show that the negative shock of the oil market did spill over sharply into the gas market. In addition, they argue that the effects of the shale gas revolution are not apparent in Europe.

Unlike the studies that focus on the relationship between oil and natural gas markets, there are limited studies that investigate the relationship between crude oil and other hydrocarbon markets. Oglend et al. [4] is the only study we are aware of in this strand. They investigate how the shale gas revolution has impacted the relationship between LPGs (namely, propane and butane), natural gas, and oil prices using a generalized cointegrated VAR model. They find that the expansion of shale gas has weakened the strong relationship between LPG and natural gas prices with crude oil prices in the United States. They argue that the state of the LGPs market affects the relationship between oil and gas prices. For instance, when high prices in the oil market increase the prices for liquids in LPG markets, then natural gas producers might increase the production of natural gas, aiming to gain more profits from the liquids, which, consequently, a higher supply of natural gas lowers its prices. However, there are other studies such as Westgaard et al. [33] and Myklebust et al. [34], who investigate the price dynamics of propane, butane, and naphtha traded in the north European market. They employ state-space models with unobservable components in a multivariate context to address the joint dynamics of these gas component prices. They do not incorporate natural gas or oil prices in their analysis to see how the prices might vary as the price of oil or gas fluctuates.

#### *3.2. Oil Price as an Endogenous Variable*

There has been considerable improvement in econometrics methods to identify the causes underlying oil price shocks in the last two decades. These studies are mostly built on Kilian [6]'s article, in which he shows how the price of oil is disengaged. Kilian [6] initiates an index to capture the changes in the real global economic activity and employs the structural Vector AutoRegressive (VAR) model to decompose the evolution of the real oil prices into three different sources: shocks to the global supply of crude oil, shocks to the global demand for all industrial commodities (including crude oil) that are driven by the global business cycle, and oil-market-specific demand shocks (also referred to as precautionary demand shocks). Since Kilian [6], the literature has departed from the earlier studies that mostly treat the price of oil as exogenous and incorporates supply and demand shocks to the real price of oil. Kilian's approach has resulted in the formation of new extensive literature to study the effects of the oil shocks. Some studies have examined the effect of oil price shocks on macroeconomic and financial variables. For more detail, see the review papers of Kilian [35] and Herrera et al. [36]. In this section, we review the studies that augment Kilian [6]'s structural VAR model to include the real price of hydrocarbons and investigate the response of the real price of hydrocarbons to the structural shocks in the crude oil market.

Jadidzadeh and Serletis [1] and Zamani [37] both augment Kilian [6]'s model and incorporate the real price of natural gas as the fourth variable to investigate the impact of those structural shocks in the oil market on natural gas prices. The former uses the monthly U.S. natural gas wellhead price from 1976:1 to 2012:12, and the latter uses the monthly U.S. natural gas imported price from 1989:1 to 2014:12. Both of these studies show that supply shocks in the oil market do not have a statistically significant impact on the real price of natural gas, but it increases the real price of natural gas after 12 months. The response of gas prices to the oil supply shocks measured by the impulse response functions are only statistically significant based on one-standard error bands. The demand shocks that are explained by an unexpected increase in the global demand for all industrial commodities have an immediate and sustained increase in the real price of natural gas. Finally, expectations in the oil market captured by the precautionary demand for oil has an immediate and sustained increase in natural gas prices.

Jadidzadeh and Serletis [1] employ the forecast-error-variance decomposition to quantify the share of the oil market structural shocks on the variation of natural gas prices over different time horizons. They find that in the short-run the oil market shocks do not demonstrate a considerable effect on the variations in the real natural gas prices. For example, all three structural shocks account for less than one percent fluctuations in the real price of natural gas in the first month, while the contribution of the structural shocks increases as the forecast horizon increases. However, in the long run, 45% of the variation in the real price of natural gas originates from the three structural shocks in the crude oil market. The largest contributor to the variation in the real natural gas prices is the aggregate demand shocks (which explain more than 16% of the variation), followed by the contribution of precautionary demand shocks and oil supply shocks (accounting for about 16% and 13% of the variation, respectively) to the long-run variation in the real price of natural gas. The remaining variation (about 55%) in the real price of natural gas is from the shocks that are specific to the natural gas markets or generally other shocks.

A recent study by Jadidzadeh and Serletis [7] investigates the impact of structural shocks in the crude oil market on the price of ethane, propane, normal butane, isobutane, and natural gasoline over the period from January 1985 to April 2020. The results are very similar to what was presented in Jadidzadeh and Serletis [1] and Zamani [37] for natural gas prices. Although the supply-side shocks in the crude oil market do not demonstrate statistically significant impulse responses, a negative (one standard deviation) shock in the supply of crude oil would tend to increase each NGL's price after six months temporarily. A positive (one standard deviation) demand shock, either in the aggregate demand shock or the precautionary demand shock, represents an immediate and persistent increase in NGL prices. The forecast-error-variance decompositions show that in the short-run (first three months horizon), the oil market's structural shocks explain 23%, 34%, 44%, 50%, and 67% of the variation in ethane, propane, normal butane, isobutane, and natural gasoline prices, respectively. In the long run, the share of structural shocks accounts for approximately 55% of fluctuations in the prices of NGLs.

Unlike the studies that treat oil prices as exogenous and employ cointegration models, this stream of studies distinguishes the transmission channels that the price of oil impacts the other hydrocarbon prices. They presume that the system of equations in the SVAR model is composed of two blocks. The first block consists of the three structural shocks that are specific to the oil market, including oil supply shocks, aggregate demand shocks, and precautionary demand shocks. The second block captures the volatilities in the natural gas or NGLs markets that are not driven by the shocks in the oil market. Then, the model investigates how the structural shocks that are specific to the oil market impact the natural gas or NGLs prices.

#### **4. Conclusions**

We have provided a survey of studies that investigated the effect of crude oil price shocks on the main hydrocarbon prices. First, we discussed the channels through which oil price technically impacts natural gas and hydrocarbon gas liquids. We then reviewed two strands of studies. The first strand relies on the global nature of the crude oil market and presumes that large variations in the price of oil are historically driven by supply disruption as a result of political events and OPEC decisions. It assumes that oil prices are exogenous to macroeconomics variables and discards the source of oil price shocks in their analyses. It investigates the existence of correlation and the long-run cointegrating relationship between oil and hydrocarbon prices using different types of time series models such as Vector Error Correction Models. Initially, this literature shows that the relationship between the oil price and the natural gas or NGLs prices is asymmetric in a way that the oil price predominantly drives the other hydrocarbon prices, but not the other way around.

The second strand of studies treats the price of crude oil as endogenous and disentangles the causes underlying oil price shocks. This recent empirical literature, which is built on Kilian [6], employs a structural Vector AutoRegressive (VAR) model to identify the source of oil price fluctuations as a result of supply disruptions and demand changes in the integrated global oil market. Other studies augment Kilian [6]'s structural VAR model to include the real price of another hydrocarbon and investigate the response of other hydrocarbon prices to the structural shocks in the oil market. They show that oil supply-side shocks have a transitory and temporary impact while oil demand-side shocks have a persistent and permanent impact on other hydrocarbon prices. In addition, they show that about 50% of the variation in the other hydrocarbon markets is explained by the structural shocks in the global crude oil market in the long-run.

**Author Contributions:** Conceptualization, A.S. and A.J.; validation, A.S., A.J. and M.M.; investigation, A.J. and M.M.; writing—original draft preparation, A.J. and M.M.; writing—review and editing, A.S. 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.

#### **Abbreviations**

The following abbreviations are used in this manuscript:


#### **References**


#### *Article* **An Exploratory Study on the Development of a Crisis Index: Focusing on South Korea's Petroleum Industry**

**Kyungbo Park <sup>1</sup> , Hangook Kim <sup>2</sup> and Jeonghwa Cha 3,\***


**Abstract:** Industrial crises exert considerable influence on a wide range of industries, national economies, and global economic landscapes. The primary objective of this study is to devise a crisis index specifically tailored for the petroleum sector—a vital component of South Korea's energy industry. An exhaustive analysis of the existing literature was conducted to extract pertinent elements and indicators, and indicator weights were determined using the analytic network process (ANP). Moreover, a combination of qualitative and quantitative methods was employed to rigorously evaluate the validity of the proposed crisis index. The implications derived from this study offer critical insights for stakeholders into the petroleum industry and demonstrate the potential applicability of a crisis index framework for other industries.

**Keywords:** energy industry; petroleum industry; crisis index

#### **1. Introduction**

Industrial crises are caused by specific and identifiable events [1]. As globalization spreads and technologies become more complex, inevitably, there is more potential to harm the industry. A crisis is defined as "a situation that threatens the high-priority goals of the decision making unit, restricts the amount of time available before the decision is transformed, and surprises the members of the decision making unit by its occurrence" [2] (p. 13). Thus, an industrial crisis threatens industry.

As the relationships among industries converge, a crisis in one industry is not just a crisis in a specific industry [3]. Crises in the energy industry that produces and sells energy have a great impact on the national economy and other industries [4]. The petroleum industry accounts for a large proportion of the energy industry, and, over several decades, it has become renowned for its substantial contribution to global economic development. The petroleum industry has facilitated socioeconomic growth and significantly influenced international politics. However, recent transformations in the energy market and technological advancements in manufacturing have led to a myriad of challenges within the oil industry, complicating the preservation of stability, and resulting in the emergence of multiple urgent crises. Factors such as geopolitical tensions, worldwide pandemics (e.g., the COVID-19 pandemic), shifting demands for energy resources, and the growing adoption of renewable energy to address climate change concerns have forged an unpredictable and volatile oil market.

Owing to these adversities, the oil industry has encountered various critical issues. These have included reduced energy demands, requisite transition to cleaner energy alternatives, and hazards associated with the industry's failure to adapt. Furthermore, the sector faces waning competitiveness because of outdated methods and practices coupled with insufficient investments. Thus, there is a necessity for proactive measures in response

**Citation:** Park, K.; Kim, H.; Cha, J. An Exploratory Study on the Development of a Crisis Index: Focusing on South Korea's Petroleum Industry. *Energies* **2023**, *16*, 5346. https://doi.org/10.3390/en16145346

Academic Editors: Bernard Zi˛ebicki and Edyta Bieli ´nska-Dusza

Received: 30 May 2023 Revised: 7 July 2023 Accepted: 11 July 2023 Published: 13 July 2023

**Copyright:** © 2023 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/).

to rapid market changes. Additionally, unrecognized risks may lie dormant, potentially exacerbating existing issues within the industry.

The oil industry is substantially important for South Korea's national economy. The petroleum industry accounts for the largest proportion of Korea's energy industry, and it accounts for 1.5% of Korea's gross domestic product (GDP). In addition, the world's second, third, and fifth oil refineries are located in Korea. Several studies have demonstrated that Korea's petroleum industry affects the global economy [5–7]. In summary, a crisis in Korea's petroleum industry could damage the global economy, along with the country's domestic economy and industry. Under such circumstances, conventional risk management methods are insufficient, which necessitate the exploration of inventive and all-encompassing approaches to regulate various crises. Therefore, it is essential to establish a crisis index that can guide informed decision making and promote effective crisis management approaches. The development of a petroleum industry crisis index is of paramount importance for several key reasons. It enables identification of risks and vulnerabilities within the industry, allowing stakeholders to address these challenges proactively. Additionally, understanding the factors contributing to crises facilitates the establishment of strategies to enhance preparedness and resilience in the face of potential adversities. Furthermore, a crisis index would serve as a guide for government and regulatory policy decisions, ensuring adequate support for the industry during challenging times. Investors could also employ the index to make informed investment decisions based on industry stability and risk factors. Lastly, an index would provide a quantifiable measure to monitor trends and changes within the petroleum industry, offering valuable insights and prompting stakeholders to adapt their strategies accordingly. Collectively, these reasons underscore the crucial role a crisis index would play in fostering industry stability, growth, and long-term success within the context of the petroleum sector. Devising an index specifically for the petroleum industry would help stakeholders to utilize a quantitative metric to appraise and track risk levels, subsequently permitting the implementation of targeted interventions and optimal allocation of resources. Moreover, the creation of an evidence-based framework to assess industry-related risks would bolster the overall resilience and sustainability of the sector.

Despite the importance of understanding the risks associated with Korea's petroleum industry, there continues to be few studies on gauging its crisis levels. The Korean government, local authorities, and companies have developed indicators for predicting crises. However, due to possible concerns, these indicators are not openly disclosed; however, the leading economic index and coincident economic index are disclosed, but these are indicators for measuring the overall economic situation and not for evaluating industries. Furthermore, most evaluation standards predominantly assess financial status, which presents another limitation. Consequently, this study develops a crisis index tailored to South Korea's petroleum sector. Consequently, the extant literature was reviewed to extract the factors that influence the petroleum industry. This information was used to establish indicators constituting these influencing factors.

Adjusting the weight of each indicator is crucial for developing a crisis index. Therefore, we employed the analytic network process (ANP) to adjust the weights. Numerous preceding studies have employed the analytic hierarchy process (AHP) to adjust the weights of indicators when developing indices [8,9]. However, the ANP overcomes the limitations of the AHP and has recently been utilized in several index development studies [10]. In the analytic hierarchy process (AHP), a hierarchical structure is assumed, and the analysis is dependent on a single factor. In contrast, the analytic network process (ANP) presumes a network structure, enabling simultaneous consideration of interrelated factors, thus providing a more comprehensive approach. Additionally, each indicator is normalized to account for varying scales. Through these steps, we established a crisis index for the South Korean petroleum industry. Finally, we employed the Johansen cointegration test to evaluate the validity of the crisis index.

The remainder of this paper is organized as follows: Section 2 reviews Korea's petroleum industry and crisis indices used in previous studies, Section 3 introduces the research framework and describes the composition of each phase, Section 4 presents the analysis of the results, Section 5 summarizes the results and presents the implications and future research directions.

#### **2. Research Background**

#### *2.1. The Petroleum Industry in Korea*

Petroleum coupled with refined products is of immense importance globally. In addition to being a central energy source, refined petroleum products serve as feedstock for numerous other industries. Hence, they perform a mounting and pertinent part in the life of human beings by generating numerous jobs and noteworthy tax revenues and royalties, including foreign earnings, for local, state, and national governments [11].

The Korean petroleum industry has depended on domestic markets to meet the energy demands that are essential for economic growth. In the early stages of development, major energy equipment and technologies were imported due to limitations in indigenous technology and funding difficulties. However, over the past 40 years, Korea's energy industry has progressively acquired expertise and technology in areas such as refined petroleum product manufacturing, oil storage facility construction, and refinery operation through domestic development efforts. Consequently, Korea has attained economic advancement via its energy sector.

Although Korea is not an oil producer, it plays a significant role in the global oil refining market. In 2020, among over 500 refineries worldwide, three refineries ranked within the top five based on daily oil-refining capacity. These refineries included SK Energy Ulsan Plant (second place), GS Caltex Yeosu Plant (fourth place), and S-Oil Ulsan Plant (fifth place). As of November 2022, Korea's total exports amounted to USD 629.1 billion, with petroleum product exports contributing USD 58.221 billion (9.3%). Thus, Korea's petroleum industry has a considerable impact on its economy and other sectors.

As Korea lacks natural resources, it relies primarily on the import of oil and other raw materials. Therefore, an early warning system for the oil and raw material markets is essential to monitor rapid increases in the prices of crude oil or raw material. The indicators used in the model include U.S. crude oil inventories, OPEC's crude oil production, variables related to economic fluctuations, and futures contracts for crude oil [12].

However, the existing early warning system merely focuses on oil prices and does not represent a crisis in the petroleum industry.

#### *2.2. Research Related to a Crisis Index*

Several studies have investigated crises in the petroleum industry and crisis indices in many areas. In addition, research has also been conducted on a resilience index. Table 1 summarizes the existing studies on crisis and resilience indices. The methodology used in these studies included natural networks, regressions, and formulas created by the researchers. In addition, in all the studies, the elements necessary for the index were set based on the existing literature and the factors and indicators were set based on research in the literature research. The existing research has analyzed data spanning from a minimum of one year to as long as 23 years, generating a crisis index as a result. The primary focus has been on financial risk indices, with recent studies shifting towards resilience. Furthermore, many investigations have endeavored to represent the crisis through a minimal number of variables. The methodology has ranged from utilizing established models to developing new formulas to more accurately depict the crisis index. In summary, risk index-related studies have primarily concentrated on financial aspects, while more recent research has explored resilience. Several studies have also attempted to measure the crisis by devising innovative formulas. Therefore, this study attempts to identify the factors affecting the Korean petroleum industry based on previous studies.


#### **Table 1.** Extant literature on a crisis index.

#### **3. Research Framework**

The purpose of this study is to produce a crisis index for the Korean petroleum industry and to compare it with the crisis index produced by major events. Therefore, we proposed a research framework (Figure 1) to achieve the purpose of this study. The research framework comprises three phases.

In Phase 1, factors affecting the Korean petroleum industry and the indicators included in these factors were identified. Subsequently, we constructed and administered a questionnaire. In Phase 2, the weight of each indicator was calculated based on the survey data. We also collected and normalized data related to the indicators. Through this process, we developed a crisis index for the Korean petroleum industry. In Phase 3, the crisis index's relevance was evaluated by comparing it with major events.

#### *3.1. Phase 1: Performance Measurements*

The initial phase employed a performance measurement (PM) extraction method in conjunction with a cause-and-effect diagram to reveal the causal relationships among critical crisis factors (CCFs). Most prior research pertaining to crisis indices within economic or industrial domains has focused on singular aspects of industrial crises, such as global or domestic economic conditions. To evaluate a crisis index that is integrated within the process instead of serving as a means of simplistic measurement, PMs should be extracted

from each economic condition, and CCFs should be correspondingly interconnected. Thus, a cause-and-effect relationship was employed to derive the perspectives and CCFs. *Energies* **2023**, *16*, x FOR PEER REVIEW 5 of 24

**Figure 1.** Research framework. **Figure 1.** Research framework.

*3.1. Phase 1: Performance Measurements*  The initial phase employed a performance measurement (PM) extraction method in conjunction with a cause-and-effect diagram to reveal the causal relationships among critical crisis factors (CCFs). Most prior research pertaining to crisis indices within economic Because the PMs extracted from Phase 1 exhibited different measurement units, it was important to apply a normalization method. Given the varying weight of each PM and their inherent causal relationships, the analytic network process (ANP) proved to be an appropriate choice. By utilizing normalization and aggregation techniques, it was feasible to compute normalized scores according to each perspective.

or industrial domains has focused on singular aspects of industrial crises, such as global or domestic economic conditions. To evaluate a crisis index that is integrated within the 3.1.1. Extracting the Perspectives for Developing a Crisis Index for the Energy Industry

process instead of serving as a means of simplistic measurement, PMs should be extracted from each economic condition, and CCFs should be correspondingly interconnected. Thus, a cause-and-effect relationship was employed to derive the perspectives and CCFs. Because the PMs extracted from Phase 1 exhibited different measurement units, it was important to apply a normalization method. Given the varying weight of each PM and their inherent causal relationships, the analytic network process (ANP) proved to be an appropriate choice. By utilizing normalization and aggregation techniques, it was feasible to compute normalized scores according to each perspective. Establishing perspectives for developing a crisis index is important for several reasons. First, perspectives provide a structured and well-defined framework that allows for a systematic approach to analyze the complexities associated with an industry's performance [21]. This well-organized methodology ensures that all relevant variables are considered, while assessing the potential risks and vulnerabilities faced by organizations within the industry. Second, by delineating critical perspectives, executives and decision makers can better understand the driving forces and external conditions that shape an organization's performance and susceptibility to crises. This comprehensive understanding enables them to identify the root causes of potential threats, and consequently, to develop more targeted and effective response strategies to mitigate the adverse impacts of such

3.1.1. Extracting the Perspectives for Developing a Crisis Index for the Energy Industry

mance [21]. This well-organized methodology ensures that all relevant variables are considered, while assessing the potential risks and vulnerabilities faced by organizations within the industry. Second, by delineating critical perspectives, executives and decision

Establishing perspectives for developing a crisis index is important for several reasons. First, perspectives provide a structured and well-defined framework that allows for events. Third, establishing well thought out perspectives facilitates the monitoring and tracking of changes within an industry and the various factors influencing the changes. Ongoing monitoring is essential for early identification and accurate forecasting of potential crises, which can help organizations to prepare and adapt proactively. Fourth, perspectives offer valuable insights into how different aspects of an industry are interconnected and mutually influenced by one another. A holistic understanding of the relationships among these factors enables decision makers to identify potential synergies and leverage points to maximize an organization's resilience and performance even amid crises. Finally, distinct perspectives for developing a crisis index promote the prioritization of critical factors according to their relative importance and impact on the industry. By focusing on the most pertinent and influential factors, stakeholders can strategically allocate their resources and efforts to effectively address the most pressing challenges, thereby enhancing an organization's overall robustness and crisis preparedness.

In conclusion, establishing the perspectives for developing a crisis index is essential to provide a structured approach for understanding the intricacies of managing industry crises, identifying and addressing potential threats, monitoring ongoing changes, analyzing interrelationships among factors, and prioritizing critical aspects to ensure an organization's resilience and success during crises. This approach can significantly improve the effectiveness of management decisions and strategies to address industry-related challenges. Numerous approaches have been used to develop industry-related risk indices. Specifically, the extant literature mentions three perspectives, i.e., global economic conditions, domestic economic conditions, and industry conditions.

#### 3.1.2. Developing the Candidates of Critical Crisis Factors (CCFs)

Prior research has identified three major categories of perspectives for developing an industry crisis index. However, previous studies have lacked a comprehensive survey of factors that affect a crisis index. Hence, this phase endeavored to achieve greater realism and precision by identifying critical crisis factors (CCFs) and mapping their interrelationships based on causal relationships. This phase utilized the existing research to ascertain CCFs while considering the three perspectives derived in the preceding step [22]. Relationships among CCFs were established using the causal connections among identified factors. Further, CCFs unrelated to the industrial crisis index were excluded from the study.

Rockart [22] defined CSF as "a limited number of components that can ensure an organization's ability to compete successfully if the outcome is satisfactory". They developed the "CSF method", which was intended to help executives to understand the factors that create essential and potential competitive advantages [22,23]. However, in this study, we extracted the factors that affected the crisis of the industry; thus, these factors were denoted as CCFs and not CSFs. CCFs are defined as "a limited number of components that can measure the risk of an industry".

#### 3.1.3. Drawing Cause and Effect Diagram of Critical Crisis Factors (CCFs)

Economic revitalization and industrial productivity display a virtuous cycle structure, emphasizing the interconnectedness and reciprocal nature of these elements in fostering growth and progress [24]. While there is notable heterogeneity between a state and its industry, most relationships between economic situations and industry-specific conditions exhibit positive correlations. This underscores the importance of carefully analyzing domestic economic indicators to accurately assess the situation in sectors such as Korea's petroleum industry [25].

When evaluating the prospects of a crisis within an industry, it is essential to consider domestic as well as global economic factors because of their far-reaching impacts on businesses and markets. However, focusing solely on these external macroeconomic aspects may not provide a complete understanding of the challenges faced by industries. Therefore, it is imperative to examine an industry's internal dynamics to gain a comprehensive perspective of its overall stability and potential risks [26]. This multifaceted approach

ensures that strategies developed to address potential crises are well informed and effective in addressing the unique challenges of each industry.

#### *3.2. Phase 2: Developing the Crisis Index (CI)*

The objective of this phase was to identify the factors and indicators influencing the domestic petroleum industry. Subsequently, weights were assigned to each indicator using the analytic network process (ANP) methodology. Finally, a crisis index was developed by incorporating the assigned weights. This index offers a comprehensive assessment of the industry's potential vulnerabilities.

#### 3.2.1. Drawing the Relative Importance of Weights Using the Analytic Network Process (ANP)

The analytic network process (ANP) [27] is a comprehensive decision-making technique that can include all relevant criteria that potentially affect decision making. The AHP serves as the starting point for ANP. The advantages of the AHP in group decision making are as follows [28]: First, tangible and intangible, that is, individuals and shared values, respectively, should both be included in the decision-making process. Second, in-group discussions may focus on objectives rather than alternatives. Third, the discussion should be constructed such that all factors related to the decision are considered. Fourth, in a structured analysis, the discussion should continue until each member of the group considers the relevant information and reaches an agreement. In addition to the advantages of AHP, the ANP provides a more generalized decision-making model without assuming the independence of lower- and within-self-level elements. Among the different criteria levels, bidirectional arrows or arcs graphically represent the interdependence of the ANP models. If interdependencies exist at the same level of analysis, loop arcs can be used to indicate these interdependencies. Further, a questionnaire was developed based on the extracted factors and indicators. In this study, the ANP was used to calculate the weight of each indicator. Table 2 indicates the ANP questionnaire used for calculating the weights. In this study, factors and indicators to be extracted through the subsequent process were compared.

**Table 2.** Example of the ANP questionnaire.


The AHP cannot consider the interrelationships among elements [29]. Consequently, the ANP is an advanced version of the AHP that considers the internal relationships among elements through additional steps [30].

The ANP does not require a hierarchy but rather a network of elements. In this network, the elements are considered to be nodes of a number of clusters, and the level of each node may both be dominant and be dominated in pairwise comparisons [31]. Assume that each cluster k, k = 1, . . . , m, includes *n<sup>k</sup>* elements eK1, . . . , eKnk. These elements can be used to build a supermatrix, Figure 1, and compared using Saaty's scale.

After comparing all the elements in the supermatrix, it is increased to any large constraint to obtain the cumulative effects of the elements on each other [32]. The supermatrix considers all possible relationships among elements.

3.2.2. Extracting the Normalized Performance Expression of Performance Measurements (PMs)

In this study, indicator data were collected from several sources. Therefore, the scale for each indicator was set differently. Therefore, normalization was performed for each indicator. The goal of normalization was to ensure that all the data points were reflected on the same scale (importance).

In several studies, normalization has been used to integrate indicators with different units and ranges of characteristics [33]. Therefore, this study normalized the different indicators.

The crisis index was generated based on the weights and data extracted using the above process. The crisis index (C.I.) was created using Formulas (1) and (2) below, where B.I. is the interim Business Index of the industry, *NI<sup>i</sup>* is the *i*th normalized indicator value, and *W<sup>i</sup>* is the weight of the *i*th indicator:

$$\text{B.I.} = \sum N I\_{\text{i}} \ast \mathcal{W}\_{\text{i}\prime} \tag{1}$$

$$\text{C.I.} = 100 - (\text{B.I.} - 100),\tag{2}$$

#### *3.3. Phase 3: Evaluating the Crisis Index (C.I.)*

Two methods were used to evaluate the developed crisis index. The first method evaluated the crisis index by comparing it with major events. We determined whether the C.I. proposed in this study was related to major events. Major events were identified from an article on the petroleum industry. The second method evaluated the crisis index by comparing it with the stock yield. For evaluation, the two indices were compared using the Johansen cointegration test [34]. Several studies have used stock returns as a representative index to determine the conditions of an industry [35–37]. The Johansen cointegration test is a method for verifying the existence of a stable and long-term balance between integrals through hypothesis verification of the vector autoregressive (VAR) model. The Johansen test can be considered to be a multivariate generalization of the augmented Dickey–Fuller test. Generalization refers to the examination of linear combinations of variables for unit roots. The Johansen test and estimation strategy—maximum likelihood—make it possible to estimate all cointegrating vectors when there are more than two variables [38].

The Johansen cointegration test is a pivotal statistical technique for discerning the cointegrating relationships among a multitude of nonstationary time series variables. Because of its efficacy in determining the existence of long-run equilibrium relationships among the scrutinized variables, this test has garnered considerable recognition among researchers engaged in economic and financial time series analysis [38]. By adopting a system-level approach, the Johansen cointegration test elaborates on the foundation of VAR models that simultaneously consider the interdependence of diverse variables [34]. The incorporation of VAR models enables the test to capture complex interrelationships among time series variables, thereby enabling researchers to unveil the fundamental patterns and associations embedded within the data. Moreover, the Johansen cointegration test encompasses two distinct test statistics: the trace test and the maximum eigenvalue test. Each of these tests examines the presence of cointegrating relationships at distinct vantage points. The use of these complementary tests enhances the reliability of the research findings and provides a comprehensive perspective on cointegration relationships. In conclusion, the Johansen cointegration test serves as a critical instrument for exploring the long-run equilibrium connections among non-stationary time series variables. Its foundation in the system-level approach, which is predicated on VAR models coupled with the provision of two supplementary test statistics, makes this method an invaluable resource for researchers seeking insights into the intricate dynamics and interdependencies that dictate the long-run trajectory of the variables under investigation. Because of its versatility and robustness, the Johansen cointegration test remains a highly acclaimed and indispensable tool in econometrics and applied time series analysis.

Thus, these two processes were used to validate the proposed index.

#### **4. Result**

#### *4.1. Phase 1: Extracting Performance Measurements (PMs)*

4.1.1. Extracting the Perspectives for Developing a Crisis Index for the Energy Industry

Based on the extant studies, the perspectives on the crisis in the Korean petroleum industry can be divided into three categories. First, global economic factors are classified into the first category, which includes the exchange rate, global supply chain management, international raw material prices, U.S. economic conditions, and the export/import status of Korea's major trading partners,.

Second, domestic economic factors comprise the second category, which includes the composite economic, sentiment, price, import/export, and industrial production indices. Finally, industrial economic factors comprise the third category, which includes the employment, shipment, and business survey indices. Table 3 summarizes the perspectives and factors for developing the crisis index.


**Table 3.** Perspectives for developing the crisis index.

#### 4.1.2. Extracting the Candidates of CCFs

In previous studies, many researchers have analyzed the variables that affect industries. Many variables affect the industry, but they are largely divided into global economic conditions [100,101], domestic economic conditions [102,103], and industrial conditions [104,105].

#### 1. Global economic conditions

Owing to recent globalization, the global economy has significantly influenced countries and companies [100]. The global economy is a factor for countries and companies with significant imports and exports [106]. Global economic conditions are significant to understand the industrial crisis because the Korean petroleum industry generates large profits from exports.

• Exchange rate

Exchange rate is an important macroeconomic variable used to determine international competitiveness [39]. Foreign exchange risk implies that changes in exchange rates affect a company's cash flow and values [40]. An oil-exporting (importing) country may experience currency appreciation (depreciation) in response to rising oil prices and currency depreciation (appreciation) in response to falling oil prices [41]. The exchange rate affects the price of imported crude oil, which, in turn, affects the price of petroleum product prices [42]. An increase in the exchange rate raises industrial profits related to petroleum in the short term; however, in the long term, the impact is reduced by the feather and rocket effect [43]. In summary, exchange rate fluctuations are important determinants of the petroleum industry's crisis.

#### • Baltic Dry Index

The international dry bulk shipping market is closely related to commodity and crude oil markets [44]. At the same time, the Baltic Dry Index (BDI) is usually considered to be a major economic indicator [45]. The Baltic Dry Index (BDI) comprises average prices paid for the transport of dry bulk materials across more than 20 routes [46]. Rising oil prices cause shipping companies to increase their expenditures on fuel costs, that is, crude oil is a fuel that directly affects the cost of shipping [45]. Thus, rising oil prices affect shipping fares. This, in turn, further increases the price of crude oil. An increase in the BDI negatively affects the petroleum industry. In Korea, crude oil is imported and refined petroleum products are exported. An increase in shipping costs negatively affects product price competitiveness. This would cause a crisis in the industry.

#### • Global Supply Chain Pressure

The Federal Reserve Bank of New York developed the Global Supply Chain Pressure Index (GSCPI) to measure the common factors of several cross-country and global indicators of supply chain pressure. The GSCPI indicates that supply chain disruptions are about a mismatch between demand and supply, indicating a lack of balance. A mismatch between demand and supply causes negative supply shocks. Supply shocks cause a decrease in production capacity, and therefore, increase the economic burden of companies [47]. In other words, an increase in the GSCPI will negatively impact the petroleum industry. In particular, Korea imports most of its crude oil supply from the Middle East; therefore, supply imbalances caused by supply chain pressure will negatively affect Korea's petroleum industry.

• Crude oil price

Several studies have suggested that rising oil prices shrink the economy [48–50]. An increase in the absolute crude oil price, and not an increase in price due to an exchange rate rise, may also negatively affect consumption, investment, and unemployment [51]. Additionally, increasing costs of firms affects the postponement of investment decisions [51]. A prolonged increase in oil prices could lead to changes in the production structure and have a deeper impact on unemployment [50]. Indeed, an increase in crude oil prices reduces the returns on sectors that are oil intensive and incites firms to adopt and construct new processes that are require lesser oil inputs [107,108]. This change generates capital and labor reallocations across sectors that can affect unemployment in the long run [52]. In summary, an increase in absolute crude oil prices will negatively affect the petroleum industry in the long and short terms.

• Commodity price

Commodity prices can provide instantaneous information on the state of the economy [53]. Moreover, because many primary commodities are important inputs for the production of manufactured goods, changes in commodity prices directly affect production costs [54]. Increasing production costs reduce a company's revenue and productivity [55]. This situation can lead to an industry-wide recession that reduces energy consumption [56]. The Bloomberg Commodity Index (BCOM) is a representative index of the price of international commodities. It is calculated on an excess return basis and reflects commodity futures' price movements.

• Gold price

In the past, gold was considered to be a means of risk hedging; however, now it is also a financial investment product. In addition, when the economy grows, each country's central bank develops a strategy to increase gold holdings [57]. Gold is a precious metal with a risk-hedging function that enables predictions of macroeconomic variables and economic activity [109]. In other words, gold is an indicator of the state of an economy, which represents energy consumption.

• U.S. economic situation

Many studies have shown that Korea's economy is subordinate to the U.S. economy [58]. Export-oriented economies such as Korea and Japan are closely related to the accumulation of foreign exchange reserves, which are affected by the USA [59]. In addition, because the USA is Korea's main export destination, its economy is affected by the U.S. economic situation [60,61]. Therefore, the U.S. economy is an important variable for the petroleum industry, which produces energy sources. A set of variables representing the U.S. economy includes the U.S. Industrial Production Index [62,63], U.S. standard interest [64], U.S. money stock [65,66], U.S. economic policy uncertainty [67], and the Dow Jones Index [68,69].

• Korea's main exporter and importer

Korea's ratio of exports and imports to GNI is relatively high compared to that of other countries. According to the OECD, Korea's ratio of exports and imports to GNI is 72.3%, compared to 31.4% for the USA, 37.5% for Japan, and 66.1% for France. This means that Korea has a trade-dependent economy [70]. The Korean economy shows a positive correlation with imports, which are a major factor in evaluating the revitalization of the economy [71]. Export performance and expansion represent economic growth and economic conditions [110]. First, higher exports ease the foreign exchange constraint, and higher imports of capital goods and intermediate goods are permitted [71]. Second, the home country can concentrate its investment in sectors in which it enjoys a comparative advantage [72]. Third, the addition of international markets provides scope for economies of scale in the export sector [71]. As of 2022, China (21.9%), the USA (13.5%), and Vietnam (6.1%) accounted for 41.6% of Korea's total imports and exports. Imports and exports to the top three countries mentioned above greatly influence the Korean economy [73,74]. Japan has been South Korea's primary trading partner. However, since 2013, there has been a notable shift in Korean exports to Vietnam. As of 2022, the income discrepancy has more than doubled, reflecting a significant change in trading dynamics. Furthermore, the European Union (EU) indicators do not accurately represent those of each individual member country, as considerable disparities exist within the EU. The export and import indicators of these EU member nations are notably lower than those of China, the United States, and Vietnam.

2. Domestic economic conditions

Economic vitalization and industrial productivity have a virtuous cycle structure [111]. There is heterogeneity between the state and industry; however, most of the relationships between the economic and industrial situations are positive [112]. In other words, the domestic economic situation reflects Korea's petroleum industry.

• Economic fluctuations

Many countries have created indices to evaluate their economic conditions [75], such as the business indicator index that represents Korea's real economy [76]. Several studies have used composite indices (CIs) to reflect Korea's economic situation [77,78]. CIs are summary measures designed to indicate changes in the direction of aggregate economic activity. Composite indices examine and predict the overall direction of economic change, speed of change, economic situation, and turning points in economic conditions by selecting, processing, and integrating major indices to reflect each part of the economy. Composite indices measure not only directions, phases, and turning points, but also amplitudes and levels in the economy. The CI consists of the composite leading index, composite coincidence index, and composite lagging index. The composite leading index indicates cyclical turning points in economic activity. Composite coincident index includes turning points that coincide with those of the business cycle. Subsequently, the composite lagging index confirms economic activity.

• Sentiment index

The effects of a survey-based consumer sentiment index (CSI) and economic sentiment index (ESI) as leading indicators of the actual economic situation have been confirmed, and CSIs and ESIs are actively used in policy judgment and decision making in many countries [79]. Previous studies have shown that a survey-based consumer sentiment index (CSI) is an effective leading indicator of real economic conditions [80,81]. A CSI offers valuable insights into the prevailing economic climate by monitoring the present state of the economy and signaling potential shifts in economic activity. In South Korea, the Bank of Korea (BOK) publishes a monthly CSI, aiming to capture both the emotional and the economic factors that influence consumer behavior. This is achieved by surveying households regarding their perceptions of current and future economic conditions, at the individual level as well as the national level. An economic sentiment index (ESI) is a survey-based index that provides information on the perceptions and expectations of economic agents, both from the demand (consumers) and the supply (producers) sides of the economy [82].

#### • Price-related index

A price index is the normalized average price of relatives for a given class of goods or services in a given region during a given time interval. Previous studies have used the benchmark interest rate [83,84], expected inflation rate [85,86], and producer price index [87,88] as indicators of the price index.

Increasing the benchmark interest rate increases interest rates in the market, thereby widening the gap between domestic and overseas interest rates and increasing the inflow of foreign capital into a country [87]. Consequently, the exchange rate and import prices fall, which further reduce domestic prices. In other words, the benchmark interest rate and inflation are negatively correlated. Interest rates alter firms' financing costs, affect the amount of loan interest and principal payments, and affect the cash flows of firms [90]. A decline in cash flows negatively affects an industry [91]. However, the effect of the benchmark interest rate varies across industries [90].

Real interest rate and expected inflation are two key economic variables [92]. Real rate variation should only affect the short end of the term structure, whereas the variation in long-term interest rates is primarily affected by shocks to expected inflation [36]. Therefore, the expected inflation rate was used to identify long-term changes [113]. Expected inflation rates also apply to perceived inflation rates [114]. However, the effect of the expected inflation rate varies across industries [115].

The producer price index (PPI) is an indicator that calculates the average cost of a collection of inputs frequently used by manufacturers. The PPI serves dual purposes: (1) offering price indices for deflating gross domestic product data and (2) presenting a comprehensive measure of inflation levels. Many economists believe that the PPI can function as a reliable predictor of upcoming consumer inflation since fluctuations in the prices paid by producers (i.e., shifts in production costs) typically occur prior to changes in consumer costs [87]. The effect of the PPI differs across industries [89]. However, the PPI is disadvantageous in that it includes all items corresponding to raw materials, intermediate goods, and final goods, because domestic shipments, including intermediate transactions between companies, are selected as populations and weighted. Therefore, Korea developed a domestic supply price index. The domestic supply price index was divided into raw materials, intermediate goods, and final goods according to the degree of processing, including not only domestic shipments but also import and export price index. Consequently, it is possible to understand the ripple process of prices step by step.

• All industry activity

Industry activity is fundamental to the economic value and situation [116]. Industrial activity is useful because it determines the current economic situation. Korea developed the Index of All Industry Production (IAIP) to understand industry activities. The IAIP is a single index that aggregates monthly production activity trends for goods and services across all industries in the entire Korean economy.

• Trade activities

Global and domestic economic situations are closely related [93]. Trade has a strong positive effect on economic growth, and trade activation confirms a positive national economic situation [94]. Several studies have confirmed that imports and exports are closely related to the economy. Active exports indicate that the overseas economy is booming, leading to vitalization of the domestic industry. Active imports mean that many raw materials are being invested in because of the vitalization of the domestic industry. In other words, export and import activities imply that the domestic economy is booming.

3. Industry conditions

To understand the crisis in the industry, it is necessary to examine the industry's internal situation [104] along with assessing the global and domestic factors.

• Employment

Employment growth is a useful indicator of an industry's location. If an industry's performance decreases, employment naturally decreases [95]. If the forecast for industrial production is good, then creation of employment and new employment increase [96]. In other words, employment and new employment are excellent indicators of the industry's situation.

• Industry shipments (manufacturing shipment)

Industrial shipments are attractive variables for monitoring industrial situations [97]. Industry shipments relate to the vitalization of an industry [98]. Korea developed an Index of Manufacturing Shipment (IMS) to indicate the level of shipment. The IMS represents the level at which goods are sold at the factory where they are produced. The IMS is a useful index for checking the inventory of industries. Rinsing inventory rates present a negative signal for the industry.

• Industry perception

These indices are considered to be the most representative of these industries and are quantitatively evaluated through quantified features. However, industry perceptions may differ from those used to evaluate current industries' performances.

Therefore, several countries use a business survey index (BSI) to examine the current situation in the industry. The business survey index (BSI) was compiled to understand the business conditions for the current month and their outlook for the following month by conducting surveys on entrepreneurs' perceptions. A BSI is commonly used in application forecasting. The existing academic literature also demonstrates that a BSI is a powerful tool for forecasting and tracking economic activity [99]. In Korea, 15 items, including business conditions, sales, and profitability, were surveyed among 3255 corporations that were selected using stratified systematic sampling and fall into the following classifications as per the 10th revision of the Korean Standard Industrial Classification (KSIC). The bulk of business surveys is useful because they provide leading indicators for the most obvious macroeconomic aggregates [99]. This information is summarized in Table 4. The data presented in Table 4 were obtained from the National Statistical Office of Korea and indexrelated organizations in the United States. The economic policy uncertainty index and the Dow Jones Index were sourced from previous studies, ensuring consistency and accuracy.

#### 4.1.3. Drawing the Cause-and-Effect Diagram of Critical Crisis Factors (CCFs)

We extracted three perspectives and 13 CCFs from the extant literature. Based on this information, we illustrated the cause-and-effect diagrams between the identified perspectives and CCFs. In the cause-and-effect relationship diagram, CCFs of the crisis index are categorized as leading and lagging factors, and the relationships are presented in ascending order. The cause-and-effect relationship between the leading and lagging CCFs is illustrated in Figure 2. The leading and lagging factors were developed through a comprehensive analysis of prior research.

Each perspective is directly associated with the crisis potential of an industry. Global and domestic economic conditions both influence an industry's economic performance. In the causal relationship diagram, CCFs are categorized as antecedent factors, and their relationships with lagging factors are depicted in ascending order to highlight the intertwined nature of these elements.


*4.2. Phase 2: Developing a Crisis Index (C.I.)*

4.2.1. Drawing the Relative Importance of Weights by Using the Analytic Network Process (ANP)

A total of 20 professors related to the industry were surveyed to establish weights. After explaining the crisis to 20 professors, they were asked to participate in the survey. The content of the questionnaire was set as "This questionnaire is related to the industrial crisis index. Regarding the crisis, you can fill in the position where you think item element-1 is 'very important' compared to item element-2.". We used Super Decisions Software (Ver.2.10) to calculate the ANP to produce an unweighted supermatrix and a weighted supermatrix. These two matrices were multiplied by a convergent dependency relationship to obtain the relative weight of each element. The correlations between the ANP assessment factors and indicators are shown in Figure 3, and the final results are listed in Table 5.

4.2.2. Extracting the Normalized Performance Expression of Performance Measurements (PMs)

The data used in this study were collected between January 2012 and April 2022. Data for each indicator were collected, and the units of each data point were normalized because they did not match. Because most of the indices released in Korea were based on 2015 = 100, other indices were also set to 100 in 2015. All indicators were derived as values relative to

the average value in 2015 and normalized. The normalized equation is as follows, where *I<sup>i</sup>* is the ith indicator:

$$NI\_i = \frac{I\_i}{\left(\sum\_{l=1}^{12} I\_{2015}\right) / 12} \* 100,\tag{3}$$

**Figure 2.** The cause-and-effect diagram of CCFs. **Figure 2.** The cause-and-effect diagram of CCFs.

(Ver.2.10) to calculate the ANP to produce an unweighted supermatrix and a weighted **Figure 3.** ANP assessment factors and indicators. **Figure 3.** ANP assessment factors and indicators.

**Table 5.** Weights of indicators evaluated using the ANP.

Table 5.

Economic fluctuations

Global Economic Condition

Domestic Economic Condition

Industry conditions

**Factor 1 Factor 2 Indicator Weight** 

supermatrix. These two matrices were multiplied by a convergent dependency relationship to obtain the relative weight of each element. The correlations between the ANP assessment factors and indicators are shown in Figure 3, and the final results are listed in

Exchange rate KRW/USD 0.075

U.S. Economic condition U.S. Industrial Production Index 0.009

Main exporter and importer Export/import to U.S. 0.027

 Global supply chain pressure 0.024 Commodity Crude oil price 0.021 Bloomberg Commodity Index 0.018 Gold price 0.004

 U.S. money stock M1 0.004 U.S. money stock M2 0.006 U.S standard interest rate 0.008 U.S. economic policy uncertainty 0.009 Dow Jones Index 0.011

 Export/import to China 0.024 Export/import to Vietnam 0.015

Sentiment Economic Sentiment Index 0.016 Consumer Sentiment Index 0.013 Price related Standard interest rate 0.009 Expected inflation rate 0.008 Domestic supply price index 0.017

All industry activity Index of All Industry Production 0.051 Trade activities Total export 0.074

Employment Employment 0.098 New employment 0.082

Industry shipments Export 0.034

Industry perception Performance 0.093

Total 1.000

Total import 0.022

Domestic 0.094

Forecast 0.039

Leading Index 0.023 Coincident Index 0.021 Lagging Index 0.018


2013.10 92.90 2016.05 105.46 2018.12 83.22 2021.07 102.13

**Table 5.** Weights of indicators evaluated using the ANP. *Energies* **2023**, *16*, x FOR PEER REVIEW 18 of 24

Table 6 presents the descriptive statistics for normalized indicators. In this study, 124 weeks of data were collected, and the crisis index was calculated through the collected data. 2013.11 86.27 2016.06 104.94 2019.01 81.38 2021.08 108.21 2013.12 94.75 2016.07 102.11 2019.02 93.68 2021.09 101.30

4.2.3. Creating the Crisis Index 2014.01 83.81 2016.08 104.16 2019.03 92.64 2021.10 95.36 2014.02 95.24 2016.09 102.95 2019.04 80.94 2021.11 95.21

A crisis index was created using the weight and normalization values, and the results of the crisis index are shown in Table 7 and Figure 4. Table 1 shows that the crisis in the Korean oil refining industry has increased since the COVID-19 outbreak. The highest point of the crisis index was in May 2020 and the lowest point of the crisis index was in October 2018. 2014.03 84.65 2016.10 105.81 2019.05 86.69 2021.12 92.29 2014.04 87.30 2016.11 103.75 2019.06 93.18 2022.01 93.00 2014.05 86.61 2016.12 98.38 2019.07 89.07 2022.02 89.21 2014.06 93.02 2017.01 98.99 2019.08 87.25 2022.03 75.39 2014.07 88.15 2017.02 97.34 2019.09 87.97 2022.04 80.78

**Figure 4.** Crisis index for January 2012–April 2022. **Figure 4.** Crisis index for January 2012–April 2022.


**Table 6.** Descriptive statistics for indicators.

#### *4.3. Phase 3: Verifying the Crisis Index (C.I.) Using Johansen Cointegration Test*

The crisis index was formulated using the limited information currently available based on historical data. Although the variables that constitute the index can impact stock prices, the crisis index derived from various techniques does not necessarily correspond with them. Stock prices are determined through a complex interplay of numerous factors, exhibiting randomness and irregularity, rendering trends statistically unpredictable. In the financial literature, stock prices are depicted as a multifaceted equation involving past information and future expectations. Statistically, stock returns in the capital market are influenced by a dynamic combination of factors that display randomness and irregularity. Factors such as market participants' behavior, economic conditions, political events, and corporate performance all contribute to stock prices and are inherently difficult to predict, exhibiting unpredictable volatility. Consequently, the presently unpredictable stock returns cannot be employed as risk information at the corporate or industrial level.

However, if the crisis index, calculated using various methodologies, shows a statistical correlation (not causality) with actual stock prices, it becomes meaningful as an indicator representing corporate risk. Therefore, the computed crisis index can serve as information reflecting a crisis at corporate or industrial levels, providing practical implications. To verify this correlation, the present study employed the Johansen cointegration test, a renowned statistical method for examining relationships between two-time series data. Specifically, the relationship between the stock return and the development index was verified using the Johansen cointegration test. It is important to note that, while the Johansen test confirms

cointegration, it does not guarantee causation. Thus, the research scope in this study was limited to the correlation between the crisis index and stock prices.


**Table 7.** Crisis index for January 2012–April 2022.

The results of the Johansen cointegration test are presented in Table 8.

**Table 8.** Results of the Johansen cointegration test.


\* Denotes the rejection of the hypothesis at the 0.05 level.

These findings shed light on the presence of a long-term correlation between stock returns and the developed crisis index, signifying that both variables share a notable connection. Although it remains uncertain whether there is a direct causal relationship between stock returns and the crisis index, the analysis demonstrates a statistically significant correlation between these two factors. As a result, the crisis index can be considered to be a reliable and valid measure, mainly because it does not provide information that is divergent from stock returns. Instead, the crisis index is in alignment with the representation of industry volatility, as evidenced by its strong, statistically significant correlation with stock returns. The extended relationship between the crisis index and stock returns further solidifies the relevance and importance of considering the crisis index in understanding market behavior and trends.

#### **5. Discussion**

This study derives a crisis index for the Korean petroleum industry. In previous studies, the index was created using limited indicators, and there was a limitation in that various situations could not be considered. In addition, only a limited number of people could check the crisis index. To derive the crisis index, the factors and indicators affecting the industry were identified from the extant literature. A questionnaire was constructed based on each factor and indicator; this questionnaire was used to survey ten experts to determine the factors and indicators having the greatest impact on the Korean oil refining industry. The expert group consisted of experts in management, the petroleum industry, and international trade. The relative importance of each indicator was calculated using ANP software. Additionally, the value of each indicator was normalized because the scale of each indicator was different. The Johansen cointegration test was used as the quantitative method. The analysis revealed a significant long-term relationship between stock returns and the developed index. Therefore, the developed index is validated.

The risk index calculated in this study and several events related to the industry were confirmed. The lowest crisis point occurred in October 2018. At that time, the unit price of exports increased because of rising international oil prices resulting from sanctions against Iran. Exports increased due to the increased global demand for petroleum products and the regular maintenance of oil refineries in Japan and China. The volume increased because of the normal operation of new facilities by domestic companies, and China, the ASEAN, and the EU imported many Korean oil refining products. Additionally, the industrial crisis index was high between August 2015 and November 2016. At that time, the demand for oil refining products decreased significantly due to the slowdown in the Chinese industry. The self-sufficiency rate of oil refining products in China and India increased significantly, which also caused this result. The crisis index continued to decrease from November 2017 to October 2018. At that time, a U.S. oil refinery facility was shut down because of a hurricane, and China implemented environmental regulations on the oil refining industry, resulting in a reflective profit due to reduced international supply.

#### **6. Conclusions**

In contrast to the indicators used in existing practices, this study selected variables to be used in the risk index by exploring the existing literature. In addition, weights were set using a scientific method, and the ANP was used to exclude subjectivity as much as possible. In addition, weights were provided so that decision makers related to the industry could utilize them, and validity was secured by comparing the derived index and stock return using the Johansen cointegration test.

Based on the process and results of this study, the academic implications are as follows: First, the extant literature was reviewed to identify factors and indicators that affect the industry. These factors were classified into global economic, domestic economic, and industrial conditions, and important indicators were selected from these factors. The factors and indicators are measures of the economic and crisis indices in other fields. Second, a survey was conducted with a group of experts to determine the weight of each indicator. We developed a crisis index using 32 indicators, and identified important indicators based on their weights. Third, the developed crisis index can be used to study the crisis indices of other industries.

In addition to the numerous academic implications, this study yields several noteworthy practical implications. Primarily, the vast majority of extant economic or crisis indices are generated on an annual basis. In contrast, we have devised a monthly crisis index to facilitate the expeditious dissemination of vital information among decision makers, a distinguishing attribute of our index compared to other indices. This feature paves the way for more nimble responses to economic fluctuations and crises in a time-sensitive manner. Secondly, by integrating a weighted methodology into the crisis index examined in this study, professionals within the industry can more effectively pinpoint potential crises utilizing an array of leading indicators. This capacity augments not only the rapid

detection of impending crises but also promotes the execution of proactive measures to mitigate adverse effects and foster more vigorous economic expansion. Ultimately, this research serves to bolster the ongoing advancement of pioneering approaches and tactics in crisis management and economic resilience.

Despite these implications, this study proposes follow-up research based on the limitations of the study process. This study considers three factors affecting the petroleum industry. However, other factors may also be involved. Therefore, future studies should explore other factors. In addition, the weight of each indicator was based on the petroleum industry. Therefore, it is necessary to develop indicators that cover all industries, and indicators that can be used in other countries should be utilized. Lastly, since the crisis index was developed through this study, further research should be conducted to predict the index. By utilizing methods such as data mining to anticipate the crisis index, early responses to potential crises can be facilitated, promoting proactive measures against emerging challenges.

**Author Contributions:** Conceptualization, H.K.; methodology, K.P.; software, K.P.; validation, K.P., J.C. and H.K.; formal analysis, K.P.; investigation, H.K.; resources, K.P.; data curation, J.C.; writing—original draft preparation, K.P.; writing—review and editing, K.P.; visualization, K.P.; supervision, J.C.; project administration, J.C.; funding acquisition, H.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Research Program of the Korea Institute of Science and Technology Information (KISTI) (No. K-22-L03-C03-S01 and No. K-23-L03-C03-S01).

**Data Availability Statement:** Not applicable.

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

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

MDPI St. Alban-Anlage 66 4052 Basel Switzerland www.mdpi.com

*Energies* Editorial Office E-mail: energies@mdpi.com www.mdpi.com/journal/energies

Disclaimer/Publisher's Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Academic Open Access Publishing

mdpi.com ISBN 978-3-0365-9605-1