*2.5. Scenario Development*

The methodological approach in the study is based on several forecasting scenarios, which are broken down into a sequence of time intervals of 2020–2025; 2025–2030; 2030–2035. Planning or scenario analysis is a consolidated and structured process of creating future opportunities that have socio-economic, environmental, and technological implications.

Scenario planning was based on an analysis of the Arctic's external environment, followed by the identification of the main factors affecting consumer development, electricity demand, and capacity. As a result of the analysis of the external environment, a list

of macro-environmental factors that have the greatest influence on the development of consumers in the Arctic in the period under consideration was compiled.

Based on the brainstorming method, the most significant factors were identified and optimized using cause-effect diagrams. This resulted in the selection of the most significant and independent factors. The brainstorming method has established itself as an effective way to generate creative and effective ideas when solving outstanding and complex problems that require the involvement of specialists from different specialties. The method is widely known and applicable, including in solving scientific and engineering problems [74–76]. The research work presented in this article implies a comprehensive study and the involvement of specialists and young scientists from different areas of the fuel and energy complex and MSC, which makes it necessary to organize work, including by the brainstorming method. Based on this method, the most significant factors were identified and optimized using cause-effect diagrams.

During the forecasting phase, several variants of different scenario outcomes were generated. The purpose of combining the most significant factors was to establish the interdependence between the predicted outcomes of the factors under consideration and to write scenarios.

The scenarios (Table 3) combine a variety of factors. For example, such as global economic growth, political factors, environmental issues, and technological development, that illustrate the relationship between the main driving forces. Scenario driving forces include various types of factors, some of them, such as the COVID-19 pandemic, arise spontaneously. Others represent sustained trends, such as Digitalization, Decarbonization, Decentralization, Dependence, and Decrease. Today's policies of companies and governments take into account the need to achieve sustainable development goals. Thus, the SDGs are also becoming one of the most important drivers for scenario planning.

> **Table 3.** Description of scenarios.



Survey design methodology and results of the survey. To form scenarios for the development of electricity and energy supply of the Arctic consumers, a survey was conducted on the basis of expert evaluations. A poll was developed to ge<sup>t</sup> expert data to forecast Arctic energy development and the priority set of fuel resources for three time intervals: 2020–2025, 2025–2030, 2030–2035, taking into account the political, economic, social, technological, environmental and legal risks of each resource's use in the Arctic zone. The risks were assessed on a nine-point scale, where 1 is the minimum impact of risk; 9 is the maximum impact of risk. An expert assessment was made of the development of types of consumers and demand for energy resources.

Each region in the Arctic zone is characterized by its own set of fuel resources, which at the moment is conditioned by the presence of hydrocarbon or coal deposits and developed transport infrastructure of resources.

Since hydrocarbons and coal occupy priority positions in the structure of the region's fuel and energy balance, it was decided to divide the Arctic zone into three groups following the affiliation of hydrocarbons (gas, oil, and oil products) and coal to these territories:


Figure 3 presents a map of the fuel and energy balance of the Arctic region.

The time frame for the survey was 1 month. The expert group was selected from various structures, scientific and social schools. During this time, 64 people took part in the survey: employees of 5 educational institutions, employees of 5 companies of mineral complex, employees of 7 companies of the fuel and energy complex.

Thus, the forecasting of the energy sector development is based on a survey of a large number of professional workers and teachers of specialized institutions, close to the subject. As a result of this research, key consumers and the most sought-after resources for energy supply were identified.

**Figure 3.** Structure of the fuel and energy balance of the Arctic region.

A complex step-by-step work was carried out: from assessment of the current situation in the energy sector of the Arctic to the choice of research method, the concept of questionnaire and experts, data analysis, and its use for forecasting in various scenarios.

Figure 4 shows a generalized methodology of building a mathematical model of scenario forecasting.

**Figure 4.** Methodology for constructing a mathematical model of scenario forecasting.


#### *2.6. Mathematical Model of Scenario Forecasting*

To formalize and establish the numerical values of the mutual influence of risks in different scenarios on the development of Arctic consumers, the interaction matrices compiled by the research participants were used [79]. The application of the method of interaction matrices makes it possible to identify the degree of mutual influence of the factors of the considered set and predict their behavior in the future. After analyzing and processing the results of the survey of the expert group, the risks were arranged on the plane (Figure 5) in accordance with the following axes:


**Figure 5.** Results of PESTEL-risk analysis under different scenarios:(**a**) neutral scenario with *K* = 0.095; (**b**) negative scenario with *K* = 0.21; (**c**) positive scenario with *K* = 0.585.

Risk assessment varies depending on the scenario. The brainstorming session analyzed how risks would manifest themselves in negative, neutral, and positive scenarios.

It is important to note that we propose a methodology and demonstrate approaches to its implementation. When implementing the methodology in the future, the results will depend on the expert group; in this case, the forecast will be adjusted not in terms of the methods used, but in terms of expert opinion. The team of authors allows such a dependence, since the study is devoted to strategic planning, and it is necessary to use the knowledge of experts. The scenario conditions proposed by the authors are based on the study of risks and their impact, as well as on the assumptions of the development of certain risks for the worse or for the better, which is described in Table 3. The assessment and application of this approach were based on the literature, but specific assessments in the works are always different since they are formed on the basis of various scenarios. We propose a specific sequence of actions that allows for a more comprehensive assessment of future needs and the development of their sustainable provision.

It is also worth emphasizing that the matrix is dynamic; over time, it is possible to reassess the risks, since they are not static in nature, but undergo changes over the entire time interval considered in the study.

Based on the data of the interaction matrix, each group of risks (political, economic, social, technological, environmental, legal risks) is presented in the form of a generalized impact factor *Kn*, which is derived from the Formula (1):

$$K\_{\rm ll} = \frac{(K\_S + K\_d)S}{2} \tag{1}$$

Based on these interaction matrices, each of the scenarios can be represented in the form of a generalized influence coefficient *Kn*.

Based on the results of the analysis and calculations for each of the scenarios, the final generalized impact factor (total) can be found, which determines the final degree of significance of the three indicators of each risk group for the development of Arctic consumers and, consequently, the growth of energy consumption and demand for the construction of energy sources and energy infrastructure.

The results of calculations of the generalized coefficients of influence of risk groups and scenarios are presented in Table 4 and Figure 5.


**Table 4.** Values of generalized coefficients of influence *Kn*.

2.6.1. Forecast Development of Consumer Types

The next stage of the study is to assess the impact of global challenges on the development of different types of consumers in the Arctic in order to further forecast the demand for energy by different consumers, taking into account the mutual influence of global challenges.

A total of nine types of consumers P1–P9 were identified (Table 5).

**Table 5.** Symbolic designation of Arctic consumers.


Based on the survey, the influence of risk groups on the development of consumers for different time ranges was determined. At the same time, the processing of questions about the impact of risks at different time intervals was carried out. Coefficients take values from 0 to 1 in increments of 0.01. Taking into account the ranges, three effects were identified: 0.3 characterizes a weak effect; the range 0.3–0.6 corresponds to the medium effect conditions; 0.61 is in the context of a strong effect. It should be noted that these values were averaged for all experts, excluding observations with incomplete information about the main types of consumers. Thus, we obtained weighting coefficients, which allow assessing the degree of risk impact on the development of energy consumers (Table 6).


**Table 6.** Impact of risks in consumer development in 2020–2025.

Further, the total weighting coefficients of the relationship between risks and types of Arctic consumers were summarized in Table 7.

**Table 7.** Total weighting coefficients of the relationship between risks and types of Arctic consumers.


Based on the final weight coefficients of the connection between risks and types of consumers, the forecast of energy demand for consumers at different time ranges is determined.

The value of the base vector of the probability of development of a certain type of consumers, obtained from a survey of experts, is used. This vector is normalized internally by Formula (2):

$$B\_{vpiu} = \frac{B\_{vpii}}{\sum\_{i=1}^{n} B\_{pi}} \cdot 100 \tag{2}$$

where *Bvpi* is the basic vector of the probability of development of the consumer species; *Bvpn* is the internal normalization of the basic vector of the probability of development of the consumer species.

The final forecast of energy demand growth for certain consumer types by years, taking into account scenarios, is calculated by Formula (3):

$$\mathcal{W}\_{ijn} = \frac{\mathcal{W}\_f(t) \cdot (1 \pm \mathcal{K}\_n \cdot \mathcal{R}\_{\mathbb{G}f}) \cdot B\_{vpm\_i}}{100} \tag{3}$$

where *Wf(t)* is the allocated trend of electricity consumption for the period preceding the forecast one. The available data on the consumption of the period from 2010 to 2020 on the territory of the Arctic were taken as the basis.

Table 8 shows the results of calculations of energy demand up to 2025. For other time intervals, similar calculations were made.


**Table 8.** Forecast of energy consumption by type of consumer by 2025.

On the basis of the expression, which takes into account the energy consumption change on the time interval and the normalized total risk influence, the distribution of the energy consumption increase (decrease) by consumer types was obtained.

Thus, at this stage, the prognosticated development of consumer types and the associated scenario change in energy consumption are justified.

#### 2.6.2. Forecast for Resource Use Development

For the scenario study, the main resource types were considered: fuel oil, gas, coal, LNG and compressed natural gas (CNG), nuclear, nontraditional, and renewable energy sources, associated petroleum gas, and hydrogen H1–H9 (Table 9).

**Table 9.** Symbolic designation of resources.


The energy requirements (Table 1) of consumers determine the strength of the connection with the types of resources based on their criteria (Table 2). The table of weight coefficients is compiled based on the results of processing the experts' evaluation of the connection. According to the results of the analysis and comparison of consumers' requirements and characteristics of resource types, the connection matrix is compiled (Table 10). The highest value indicates a more appropriate choice and compliance with the resource to provide the given type of consumers, taking into account the fullest satisfaction of requirements. The coefficients take values from 0 to 1 in increments of 0.001 and, with this in mind, range as 0.001–0.3, a weak relationship; 0.301–0.6, a medium degree relationship; 0.601–1, a strong relationship.


**Table 10.** Matrix of connection of weight coefficients of consumers with resources.

As a result of a survey of experts, it was proposed to put the strength of influence and the direction of the influence of risks on the development of bases, then the assessment was averaged and entered into the appropriate cell. The expert assessments in the questionnaire were processed in the standard way adopted for this approach [79].

Changes in demand for certain types of consumers in the Arctic will determine changes in the demand for energy resources. For this purpose, based on the table of weight coefficients of connection between consumers and resources, the forecast of energy demand from a certain type of resource was determined.

Forecasting the increase in demand for energy resources (4) was carried out taking into account the basic vector of the probability of an increase in demand for resources (*Hvi*) and its internal normalization ( *Hvn*). This vector was obtained on the basis of survey data. Then we modeled the connection matrix of electricity consumers and resources in the form of normalized vector (*Lvn*), which corrects the forecast of demand for certain types of resources and reflects the competitive distribution as a result of the scenario conditions but does not change the value of total demand.

$$H\_i = \Delta W\_{\text{iyo}} f(t) \cdot \frac{H\_{\text{vin}} \cdot L\_{\text{vin}} \cdot 100}{\sum\_{i=1}^{n} H\_{\text{vin}} \cdot L\_{\text{vin}}} \tag{4}$$

where Δ *Wiynf(t)*—changes in energy consumption, taking into account the scenario conditions in the allocated time range.

Table 11 shows the results of the calculations of scenario forecasting of an increase or decrease in demand for resources in the time interval 2020–2025. Similar calculations were conducted for other time intervals.

**Table 11.** Results of calculations of scenario forecasting of demand for resources by 2025.


## **3. Results and Discussion**

Figure 6 presents the results of the forecast of energy consumption by the Arctic consumers based on the scenario conditions and risks, where three scenarios of development of electricity consumption and the process of change in the fuel and energy balance in the Arctic for the period from 2021 to 2035 are considered.

The R<sup>2</sup> value—the coefficient of determination is the difference between the unit and the proportion of unexplained variance. This coefficient is applicable to determine the degree of correspondence between one random variable and many others. R<sup>2</sup> can be calculated automatically using, for example, standard MS Excel tools, as was done by the authors of the article.

The results of the calculated determination coefficients prove that the obtained mathematical model for predicting energy consumption using the developed hybrid method corresponds to the data from [80] sufficiently and does not contradict them. Relying on the available data on energy consumption by consumers in the Arctic, made it possible to carry out the initial iterations of the model tuning. In turn, this made it possible to impose on the collected database the influence of risks migrating over time.

Figure 7 shows the results for the distribution of the projected increase in energy consumption by type of consumers, which takes into account the energy consumption change on the time interval and the normalized total risk influence. The results indicate that the increase in electricity will be mainly due to data processing centers, hydrocarbon deposits and logistics clusters in neutral and positive scenarios. The growth will peak in the 2030s. Following the same scenarios there will be a degeneration of settlements based on industry. During the period under consideration, from 2021 to 2035, the volume of energy consumption will change mainly due to an increase or decrease in the demand for gas depending on the scenario variant of the Arctic zone development. The high potential of gas use is due to the large reserves of this resource in the Arctic and the developed infrastructure of gas pipelines in its territory. Thus, natural gas will occupy a leading position in the structure of the Arctic fuel and energy balance.

Accepting the positive development scenario as the best in terms of sustainability, a rather high contribution to the increase of energy consumption will be made by unconventional energy sources, such as atomic and hydrogen energy. The use of nuclear energy in the Arctic has proven its validity and effectiveness in the case of small and floating nuclear power plants. The results of the prognosis point to the development of hydrogen fuel as a new perspective energy source, which is planned to be produced based on the Kola NPP in the Murmansk Region and Yakutia, as well as on pilot sites in the Yamal-Nenets Autonomous District; and to start supplying LNG in nuclear tankers to remote Arctic areas. If the scenario is positive, there will be an increase in demand for LNG use, although this

increase will not have a significant impact on the fuel and energy balance of the Arctic. Following the trend of decarbonization, the development of RESs will continue, but the increase in these resources is not comparable in volume with other resources under consideration, so it is not reflected in the simulation results. The largest increase in energy consumption will be achieved in 2030 due to gas, which is due to the planned gasification of the regions. The share of petroleum products will decrease and become equal to the share of hydrogen fuel. Such diversification of the resource mix will allow not to disturb the sensitive ecosystem of the Arctic and ensure an innovative breakthrough in Russia's energy industry, as well as the country's competitiveness in the market of clean fuels.

**Figure 7.** Distribution of the projected increase in energy consumption by type of consumers.

The model allowed taking into account not only the growth but also the decline of certain types of consumers under different scenarios. There is a noticeable increase in energy consumption of data centers, which is due to the emergence of new strategic and economic facilities, which will appear by 2025–2030 (supply of the NSR, SSH, military bases, research bases). Similarly, the demand for hydrocarbon deposits will increase, as the Arctic has a large potential of natural resources, which are already being exploited, and the development and emergence of technologies specialized for the Arctic conditions will have a high demand for energy consumption. Investment in the NSR, SSH, and the development of ports, railways, and highways will also be an impetus to increase the energy consumption of logistics clusters in the Arctic. The model clearly shows a drop in energy demand in single-industry towns. This can be explained by their low attractiveness to their current state and the prospect of attracting residents to live in these cities permanently is not observed. Consequently, we can conclude that mono-cities with the existing infrastructure and economy are a dead end, and in the Arctic, such development of territories is ineffective. Figure 8 shows the results of scenario forecasting of an increase or decrease in demand for resources at the corresponding time intervals.

The results showed that demand for hydrogen will increase over time due to loyalty and investment in hydrogen by large companies, as well as the emergence of infrastructure and research centers adapted to hydrogen fuel. Also, demand for gas, APG, LNG, and CNG will increase as LNG and CNG transportation campaigns will continue to roll out. The emergence of small NPPs and the prospect and approval by the Russian governmen<sup>t</sup> of floating NPPs will be a step toward the growing demand for nuclear power in the Arctic. Demand for petroleum products will decrease due to the risks of spills leading to large fines, sanctions, and restrictions, which reduces the competitiveness of these fuels (fuel oil, diesel) compared to others. In the next 5–10 years, coal will still be in demand as a fuel for the Arctic, but gradually the demand for its use will start to fall.

**Figure 8.** Results of resource demand forecasting.

Thus, scenario modeling of the impact of risks due to global challenges on the development of consumers in the Arctic, and the assessment of changes in demand for energy resources will allow tracing the mutual influence of risks, taking into account the scenario conditions on energy supply technology. This, in its turn, makes it possible to identify, through resources, the key technologies necessary for energy supply to Arctic consumers, taking into account all efficiency requirements while reducing the negative impact on the environmental situation in the region to achieve the goals of sustainable development.
