Mapping the Energy Sector from a Risk Management Research Perspective: A Bibliometric and Scientific Approach

Abstract

This study aims to provide a comprehensive overview of risk management research developments in the energy sector by using bibliometric analysis techniques. We apply the SciMAT bibliometric analysis software to understand how the intellectual base of this topic has evolved over time and what the major themes are that have contributed to this evolution. We analyse 679 publications referenced in the Web of Science Core Collection and Scopus to map the content of publications on risk management research in the energy sector over a period of 30 years (1993–2022), following the methodical rigour of PRISMA (Preferred Reporting Items for Systematic and Meta-Analyses). Our results identify and support the evolution of risk management research in the energy industry, its interactions, its stability, and changes in its research network. Our work contributes to the current debate on identifying trends and enhancing understanding of the evolution in the energy sector from the perspective of risk management research. It can also be a reference point for those interested in deepening their knowledge in this field.

1. Introduction

The scope, complexity, and interdependencies of recently emerging risks (e.g., the rapidly growing coronavirus pandemic on a global scale, energy problems in Europe) have shown that a forward-looking approach to risk management is more important than ever. As a result, many companies, including those active in the energy sector, have developed their risk management approach to a more integrated and comprehensive level.

For companies operating in the energy sector, a critical risk is the reduced ability to meet energy demand. Their strategic importance is felt by a whole range of consumers, especially now, at a time of intensification of the energy crisis. As highlighted in the World Energy Outlook [1], the dynamic changes that currently take place in the energy market determine the need for changes in the future involving not only diversification of energy sources, but also “… changing the nature of the energy system itself while maintaining affordable and secure energy services” [1]. Undoubtedly, the current global energy crisis will have far-reaching consequences for households, businesses and entire economies. The short-term current responses of individual national governments are subjects of an in-depth debate on how to reduce the risk of future disruptions and promote energy security.

However changing the management of energy consumption and production is still challenging for many economies. The economies of European countries are particularly aware of these challenges. This is due to the lack of local energy resources (especially oil and natural gas), which results in their high dependence on energy imports [2]. Currently, the severity of this risk is particularly noticeable.

The energy market, including the renewable energy market, is largely shaped by regulation and legislation (environmental regulations, tax breaks, utility regulations, licensing rules, etc.) [3]. These have a huge impact on its potential, economics and propensity to enhance the use of clean technologies (e.g., [4,5]). Undoubtedly, as [6] point out, energy efficiency and conservation have become key to meeting national and global energy needs in the future. Achieving this requires a holistic view of energy management and an increased commitment to innovation in this area [7].

Questions about risks in the energy sector, and consequently how these risks are managed, seem to be (especially now) a high priority for both researchers and practitioners, including managers, shareholders, policy makers and other stakeholders [8,9]. Understanding the mechanisms of volatility and risk in the energy sector amid the complex interactions of networks of producers and traders is now becoming one of the leading challenges [10]. It is therefore important to provide innovative tools for forecasting changes in the energy sector to support the decisions of policy makers and managers [11]. Energy efficiency issues form one of the main sub-areas of research in this area. Improving processes sufficiently to achieve the required parameters of energy security with available resources in the economy is still one of the leading challenges. This is highlighted by [12], who studied the dynamics models of integral indices of energy security components. On the other hand, Jackson [13] demonstrated the need to incorporate value-at-risk energy efficiency analysis into these processes. Among the studies dealing with the issue of risk management in relation to investments or energy efficiency, we may distinguish a group of studies referring to renewable energy [14,15,16,17].

The ever-increasing demand for energy and the debate on the directions and methods of energy transition underline the importance of examining the various factors in this sector [18]. In the existing body of the literature, the research attention focuses on various aspects of the interrelationships and interactions of the area of risk management and problems of the energy sector. In our work, by applying a systematic review of these studies, we attempt to indicate their paths of evolution. In this regard, the prime aim of this study is to identify common leading research themes in these areas, identify their interrelationships (including interrelationships between internal sub-themes) and observe the evolution of their development over three decades.

The contribution of our study to the existing literature on risk management research in the energy sector is as follows. The main difference between this study and previous studies is that it attempts to take a comprehensive, holistic view of the evolution of risk management research in the energy sector. The majority of the existing works review this issue by covering some specific, detailed elements [9,19,20]. In previous studies, researchers have taken the siloed approach, analysing selected specific issues. The works that cover a broader view of the interconnectedness of these areas remain relatively scarce. In addition, this study determined the scientific structure of the issues on the basis of the analysis of longitudinal changes (over three decades) and detailed mapping of isolated thematic clusters. Previous studies have also not covered such a long period. In this regard, our work contributes by displaying evidence on the intersection of risk management issues with energy sector research over three decades. In addition, the range of data used in this work has allowed us to observe the evolutionary process of this research, which can complement the findings from earlier review studies.

In this study, we are primarily guided by the considerations of two dominant areas, namely risk management research and energy sector research. Our approach considers the area of risk management research as one of the key internal drivers of change (in terms of performance or value creation) in the energy sector. The high uncertainty about the way forward in the current energy crisis and what further disruption it may cause makes the debate about the place of risk management research in the energy sector more topical than ever. In this respect, our study is guided by the following research questions:

RQ1: 

How has the perspective of academic research in the field of risk management in the energy sector changed (how has it developed)?

RQ2: 

Which topics are addressed in the identified academic research in this field?

RQ3: 

What are the most important future research priorities in the field of risk management in the energy sector?

The remainder of this paper is organised as follows. In the Section 2, we discuss the adopted research design and method. We explain the methodical aspects of database collection, as well as the software used for bibliometric analysis. The results of bibliometric analysis are presented in the Section 3. The Section 4 discusses the results. The Section 5 summarizes the findings and concludes by specifying the further research routes in the field of risk management in the energy sector.

2. Research Design and Method

For the purposes of our study, we applied various bibliometric procedures and tools, as outlined in Figure 1. First of all, to perform the mapping of research on risk management in energy sector, we needed to identify the academic papers that cover these topics. For that purpose, we followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [21,22]. Further, for the identified dataset, we performed SciMAT software (v1.1.04)-based analysis. SciMAT is an open-source software tool that performs science mapping analysis within a longitudinal framework [23].

The research approach of knowledge mapping using dedicated software is an increasingly common way of discovering and exploring scientific research. The use of this type of software enables the researcher (using large datasets) to recognise the development and status of research in a particular field and to identify its hotspots and boundaries. Among the most commonly used bibliometric mapping tools are VOSviewer and CiteSpace [24,25]. A prerequisite for any type of research is correct and precise preparation (including the preparation of data for analysis and the ability to transform them). The SciMAT software used in our study appears to have a definite advantage in this respect. This is pointed out by, among others, Moral-Muñoz et al. [25], who emphasise that “SciMAT has great preprocessing and exporting capabilities and its visualisations allow the analyst to focus deeply on the research topic”.

Table 1. A comparison of our study with similar studies on approaches to mapping risk management issues in the energy sector—key factors.

Author	Key Factor
	Thematic Scope Include		Data Source	Sample	Time Period	Software	Data Preparation Methods				Analyses and Visualisations
	Risk Management Research	Energy Sector Research					De-Duplication	Plurals/Singulars	Time Slice	Filters	Scientific Productivity	Cited Papers	Cited Authors	Major Theme	Thematic Network	Dynamics of Change between Sub-Periods	Thematic Evolution
David et al. (2020) [26]	-	x	Scopus	153 articles	20 years (2000–2019)	VOS viewer, SciMAT	-	-	-	x	x	-	x	x	-	-	-
Díaz-López et al. (2021) [27]	-	x	Scopus, WoS	1440 academic papers	29 years (1993–2020)	SciMAT	x	x	x	x	x	x	x	x	x	x	x
Zhang et al. (2019) [28]	-	x	WoS	770 academic papers	19 years (1999–2018)	Cite Space	-	-	-	x	x	x	x	x	x	-	-
Ganbat et al. (2018) [29]	x	x	WoS	526 academic papers	10 years (2007–2017)	Cite Space	-	-	-	x	x	-	-	x	x	-	-
Zhu eta al. (2021) [30]	x	x	WoS	530 academic papers	20 years (2000–2019)	VOS viewer	-	-	-	x	x	x	x	x	x	-	-
Our Research	x	x	Scopus, WoS	679 academic papers	30 years (1993–2022)	SciMAT	x	x	x	x	x	x	x	x	x	x	x

Note: sign ‘-’ the selected key factor is not present in the survey; sign ‘x’ the selected key factor is present in the study.

compares the key factors in the proposed approach to mapping risk management research in the energy sector with other state-of-the-art research published in the last five years in the field (selected by relevance filter in SCOPUS database). A review of previous studies offering a cross-sectional view of risk management research in the energy sector clearly indicates the following:

-

There has been a lack of research covering a review over a period of three decades;

-

Research to date has generally focused on a siloed (specialised, focused) approach;

-

Research to date has implicitly linked risk management issues in the energy sector;

-

The most popular tools used for knowledge mapping in large databases (such as VOSviewer and CiteSpace) do not allow multivariate preparation of measurement data, measurement of the dynamics of change between sub-periods or visualisation of the evolution of research in the field.

The demonstrated research gaps in the recognition of risk management issues in the energy sector are complemented by our study. Namely, we show the following in this study:

-

A unique holistic approach to research in this area combines both the stage of database creation and its initial preparation for further analysis. The thorough preparation of the data and their cleaning (removal of duplicates, unification of records in terms of grammar, etc.) allow us to assume that the results obtained correspond much more closely to the actual state of knowledge;

-

Expert separation of sub-periods of analysis allows us to indicate how the research has changed and developed from the perspective of the separated sub-periods;

-

We identify the hot topics/issues in each sub-period and present the measurement and strength of connections from the perspective of both the individual hot topics and the sub-periods;

-

We present the evolution of the research over three decades, identifying themes that have been transformed, have been discontinued, or are future research priorities in the field of risk management in the energy sector.

In the next sections, we explain in greater detail the applied procedures and software.

2.1. Database Creation

The use of bibliometric methods and tools to map academic research is now one of the most widely used and accepted approaches to assess and measure scientific quality, productivity and development of research [31,32,33]. These methods require a rigorous procedure for the collection of dataset (the papers subject to review), as well as the choice of the right bibliometric software for the analysis of the content of these papers.

To achieve the robust results of our bibliometric analysis, we have followed the rigour of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [21,22]. In Figure 1, we present the details of our procedure as regards the data collection and selection of the sample (academic works subject to further analysis). We adopt the PRISMA model, as our main goal is to map the risk management studies in energy sector, within pre-defined criteria. As presented in Figure 2, we began the collection of data with a preliminary search of the bibliometric data in the Web of Science and Scopus databases. These databases are considered as the largest global scientific databases containing information on scientific publications and their citation scores. Although, as Harzing and Alakangas [34] point out, the simultaneous use of several databases does not increase the number of relevant papers subject to further analysis, in this study, we excluded duplicate records in the selected databases in order to avoid inadvertent omission of relevant studies. Data extraction from the Scopus and WoS databases was carried out in November 2022. For comparability of results, it is customary to limit the record set to full-year data. However, due to the dynamic recent changes in the energy market, we consciously decided to include the 2022 academic papers in the analysis (the years 2021–2022 are distinguished as a separate sub-period).

Prior to the final data collection, we carried out a preliminary observation of the data available in Scopus and Web of Science databases. In this initial search, we found that only a few academic papers on the subject were published before 2013 (less than 10 academic papers per year).

As bibliometric mapping gives better results for larger datasets, we did not restrict the search to a specific time period (all datasets corresponding to the selected search criteria were considered). The entry search request terms were defined as the phrases ‘risk management’ and ‘energy sector’. Under these request terms, we obtained initial reference collection of 3371 academic papers. Next, we narrowed our search to the titles, abstracts, and keywords of the academic papers in Scopus and Web of Science databases. After that, the identified observations were reduced by document type and publication language. Academic papers were limited to articles, conference materials and book chapters as productive scientific reports. In addition, we limited our database to English-language academic papers. We did not limit the search based on thematic categories. Using these search criteria, we obtained a total of 679 academic papers from both databases for analysis.

2.2. Science Mapping Analysis Software Tool—Main Characteristics

We further carried out the topic mapping by applying SciMAT version 1.1.04 to analyse the collected bibliometric data. In recent years, a number of tools have been developed for knowledge mapping and bibliometric analysis. VoSviewer and CiteSpace are among the most commonly used [25]. However, as indicated by Cobo et al. [23,35,36], SciMAT is a powerful tool that integrates most of the advantages of available scientific mapping tools.

Details on SciMAT modules and algorithms can be found in Cobo et al. [23]. SciMAT was designed as a tool for both performance analysis and scientific mapping. As indicated by Cobo et al. [35], SciMAT enables both the analysis of a research field and the detection and visualization of its conceptual subdomains (specific topics or general thematic areas) and its thematic evolution. In this study, we conduct a longitudinal analysis of conceptual science mapping based on the co-occurrence of keywords. We use the approach proposed by Cobo et al. [35,36]).

The research procedure adopted in this article begins with a longitudinal analysis of the selected publications based to the keywords dedicated to this research area (creation of a database of relevant publications according to the PRISMA procedure). It, in combination with the use of science mapping tools, allows us to reveal and assess the structure and dynamics of selected aspects of scientific research [23,37,38,39,40]. As a result of this, the separation of research sub-periods and the assessment of publication and author citability, among other things, were performed. Then, using the functionalities of the SciMAT software, strategy diagrams and dedicated thematic clusters were constructed. Analysis of the individual diagrams and clusters allowed research trends (for each sub-period) to be assessed. In turn, their identification enabled an evolutionary analysis of risk management research in the energy sector. Analysis of the structural changes in the identified network of links generated an alluvial diagram. As a result, the evolutionary paths of research over three decades were identified. Emerging thematic areas were also identified as directions for recent and future research.

Accordingly, the analysis of research on risk management in the energy sector was performed at three levels:

-

Level 1—detection of research themes;

-

Level 2—visualizing research themes network;

-

Level 3—visualizing research thematic evolution and level of research stability.

For detection of the research themes (level 1), we identified for each of the sub-periods in the study the corresponding research themes by applying a co-occurrence of keywords analysis. Similarity between keywords was assessed using the equivalence index [41].

$$E_{ij}=\frac{C_{ij}^2}{C_i*C_j}$$

where:

E_ij—the equivalence index;

i, j, …—keywords;

C_i—the number of occurrences of the keyword “i” (the number of times that it is used for indexing a document within the relevant document collection);

C_j—the number of occurrences of the keyword “j” (the number of times that it is used for indexing a document within the relevant document collection);

C_ij—the number of co-occurrences of the keywords “i” and “j” (the number of academic papers that are described by both keywords in the relevant academic papers collection).

For visualizing research themes network (level 2), we characterized the identified themes by two measures of the network: centrality and density, following [35]. To measure centrality, we applied the following algorithm:

$$c=10*\mathsf{\Sigma }{e}_{uv}$$

where:

c—centrality—measures the degree of interaction of a network with other networks, and it can be understood as the external cohesion of the network;

u—an item belonging to the cluster;

v—an item belonging to other clusters.

For measuring density, we applied the following algorithm:

$$d=100 \frac{\mathsf{\Sigma }{e}_{ij}}{n}$$

where:

d—density—measures the internal strength of the network, and it can be understood as the internal cohesion of the network;

i, j—items belonging to the cluster;

n—the number of items in the theme.

These measures (centrality and density) allow categorization of the detected clusters from each sub-period and visualization of their position in a two-dimensional strategy diagram. In the diagram, we distinguish four quadrants [36]. A two-dimensional strategy diagram with a description of its individual quadrants is presented in Figure 3.

To visualize research thematic evolution and level of research stability (level 3), we track the thematic evolution and its level of stability by computing “stability index” (number of keywords common to two consecutive sub-periods) between the different sub-periods identified in the study [23]. It should be noted that the information presented above on the interpretation of maps generated through SciMAT software is not exhaustive. This information is discussed extensively in studies by the authors of this solution [23,36]. Other researchers have also described it (e.g., [42,43]). In this paper, information on the interpretation of the different types of visualization generated in SciMAT software was limited to the key ones necessary for this study. A list of detailed parameters of the analysis performed in the SciMAT tool is presented in Appendix A.

3. Results

3.1. Overall Bibliometric Analysis—Publication Trends

In the first stage, the scientific productivity regarding research on risk management in the energy sector was assessed on the basis of the number of publications in this field (Figure 4). Changes in the number of academic papers on a specific topic are an important indicator of development trends in a research area. Analysis of the evolution of the number of publications over time allows conclusions to be made about the quantitative level of research and future developments in the field [44].

Several time sub-periods were defined to conduct the analysis. The division of the studied time period into five sub-periods was based on the observation of publication productivity over time, as well as practices and recommendations from previous studies of a similar nature (e.g., [31,39,40,43]).

The first sub-period (initial phase) was considered to be the years in which the number of publications was quite low and irregular (in our case, less than 10 publications per year). The second sub-period (emerging phase) included years in which a steady, slow increase in the number of publications per year was observed (in our case, to double the number of publications per year). The next two sub-periods were separated by adopting the same (in our case, five-year) time interval. The last sub-period was distinguished by the authors as a special sub-period in which particularly severe turbulence in the world economy occurred (COVID-19 pandemic, energy crisis caused by the war in Ukraine).

As can be seen, in the initial phase of the research on risk management in the energy sector (see Figure 4), the annual number of publications was less than 10. The first year in which the value of 10 was reached was 2003. The few publications in this sub-period means that a complete document system had not yet been formed. Thus, we considered the 1993–2003 sub-period as the initiating stage for the development of risk management research in the energy sector. In the following sub-period, the annual number of publications increases, although quite irregularly. A doubling of the annual number of publications in relation to the first sub-period (i.e., 20 publications per year) was observed in 2010. The sub-period 2004–2010 is taken (for the purposes of further analysis) as the second sub-period of the research. We divided the 2011–2020 sub-period into two five-year sub-periods. The first sub-period (2011–2015) saw a further systematic increase in publications. In the next sub-period highlighted (i.e., 2016–2020), we see a clear change. The annual number of publications more than doubles over the five-year period. The years 2021–2022 are grouped together. Research on risk management issues in the energy sector is clearly in the fast-growing stage. The annual increase in publications was almost 40. We analysed the defined sub-periods using the software SciMAT. The results of our analysis are presented in the following section.

3.2. The Most Cited Articles and Authors

The analysis of citations of the selected papers has shown that among the 10 top cited papers are those from the first 4 identified sub-periods. In

Table 2. List of the most cited papers.

Rank	Title of the Paper	Author	JOUR.	Year	TC	TCY
1	Measuring safety climate: Identifying the common features	Flin, R.; Mearns, K.; O’Connor, P.; Bryden, R. [45]	SOF	2000	962	43.7
2	Combining probability distributions from experts in risk analysis	Clemen, R.T.; Winkler, R.L. [46]	RA	1999	767	33.4
3	A new approach to risk evaluation and management: Risk-based, precaution-based, and discourse-based strategies	Klinke, A.; Renn, O. [47]	RA	2002	529	26.45
4	Risk analysis and assessment methodologies in the work sites: On a review, classification and comparative study of the scientific literature of the period 2000–2009	Marhavilas, P.K.; Koulouriotis, D.; Gemeni, V. [48]	JLPPI	2011	309	28.1
5	Overview and analysis of safety management studies in the construction industry	Zhou, Z.; Goh, Y.M.; Li, Q. [49]	SOF	2015	306	43.7
6	A survey on power system blackout and cascading events: Research motivations and challenges	Alhelou, H.H.; Hamedani-Golshan, M.E.; Njenda, T.C.; Siano, P. [50]	E	2019	230	76.7
7	Principles and criteria for assessing urban energy resilience: A literature review	Sharifi, A.; Yamagata, Y. [51]	RSER	2016	228	38.0
8	Measuring the security of external energy supply in the European Union	Le Coq, C.; Paltseva, E. [52]	EP	2009	187	14.4
9	Development of a risk framework for Industry 4.0 in the context of sustainability for established manufacturers	Birkel, H.S.; Veile, J.W.; Müller, J.M.; Hartmann, E.; Voigt, K.-I. [53]	S	2019	184	61.3
10	Measuring contagion between energy market and stock market during financial crisis: A copula approach	Wen, X.; Wei, Y.; Huang, D. [54]	EE	2012	177	17.7

Abbreviations: JOUR. = Journal; TC = Total citations; TCY = Total citations per year; SOF = Safety Science; RA = Risk Analysis; JLPPI = Journal of Loss Prevention in the Process Industries; E = Energies; RSER = Renewable and Sustainable Energy Reviews; EP = Energy Policy; S = Sustainability (Switzerland); EE = Energy Economics.

, we present the articles with more than 150 citations in the Scopus (TC) and their number of cites per article by year (TCY). The six most influential articles were published in the pre-expansion period (before 2016) and have a significant number of citations (between 962 and 177).

The most cited articles were by Flin et al. [45] in Safety Science, and by Clemen and Winkler [46] in Risk Analysis. The Flin et al. [45] study could be regarded as an important contribution to research on risk management in the energy sector. This article, published during the initial period of the research in the field, refers to the changing approaches to risk and safety in the energy sector. It was published during a period of significant change in the sector, when a process of reorientation in the approach to looking at safety was initiated. The reorientation was based on moving away from safety measures based on retrospective data related to lost time (accidents and incidents) and towards proactive (predictive) assessments of the safety climate of an organisation or workplace. In contrast, research by Clemen and Winkler [46] looked at the combination of experts’ probability distributions in risk analysis. They reviewed different methods for combining probability distributions in risk analysis and showed that, both normatively and empirically, combining can lead to improved probability quality.

The highest annual citation averages in Scopus were achieved by the articles that were published relatively recently by Alhelou et al. [50] in Energies and by Birkel et al. [53] in Sustainability. These two papers refer to the importance and complexity of risk management in the energy sector.

Finally, it should be noted that among the top 10 cited papers on the topic, there were 3 that are review papers [48,49,51]. This may suggest that researchers on this topic are constantly searching for and exploring new aspects of it.

Many of the authors have made individual and important contributions to the development of research in the area of risk management in the energy sector. In

Table 3. List of the highly cited authors.

R.	A.	TC	DOC	TCD	MCP	MCPS
1	Tan, R.R.	11,434	506	22.6	Klemeš, J.J.; Fan, Y.V.; Tan, R.R.; Jiang, P. (2020) [55]	Tan, R.R., et.al. (2015) [56]
2	Dey, P.K.	6245	206	30.3	Ho, W.; Xu, X.; Dey, P.K. (2010) [57]	Dey, P.K. (2010) [58]
3	Low, S.P.	5145	245	21	Pheng, L.S.; Chuan, Q.T. (2006) [59]	Zhao, X.; Hwang, B.; Low, S.P. (2013) [60]
4	Hwang, B.G.	4296	130	33	Hwang, B.; Tan, J.S. (2012) [61]	Zhao, X.; Hwang, B.; Low, S.P. (2013) [60]
5	Zhao, X.	3877	111	34.9	Hwang, B.; Zhao, X.; Gay, M.J.S. (2013) [62]	Zhao, X.; Hwang, B.; Low, S.P. (2013) [60]
6	Burgherr, P.	1796	106	16.9	Burgherr, P.; Meyer, E.I. (1997) [63]	Burgherr, P.; Eckle, P.; Hirschberg, S. (2012) [64]
7	Kirchsteiger, C.	750	68	11	Jones, S.; Kirchsteiger, C.; Bjerke, W. (1999) [65]	Kirchsteiger, C. (2007) [66]
8	Spada, M.	740	67	11	Sørensen, M.B. (2012) [67]	Spada, M.; Burgherr, P. (2014) [68]
9	Khoiry, M.A.	206	38	5.4	Hamzah, N.; et al. (2011) [69]	Bahamid, R.A.; et al. (2022) [70]
10	Eckle, P.	176	12	14.6	Eckle, P.; Burgherr, P.; Michaux, E. (2012) [71]	Burgherr, P.; Eckle, P.; Hirschberg, S. (2012) [64]

Abbreviations: R = Rank; A = Author; TC = Total citations; ND = Number of documents; TCD = Total citations per document; MCP = Most cited publication; MCPS = Most cited publication in the field of study.

, we provide the list and ranks of the most cited authors in the field, based on Scopus. The ranks cover the number of published articles by author (DOC) and the number of total citations by author (TC). These indicators were used as a measure to assess the total citations per individual academic paper (TCD). In addition, in Table 3, we provide the most relevant papers in an author’s research output (MCP) and the papers most relevant to the field studied in this article (MCPS).

The field of risk management research in the energy sector is characterised by continued growth (see Section 3.1) and an increasing number of authors addressing the issue. It should be noted that the authors are ranked according to their productivity in their overall area of research and not strictly on the research considered in this article. Most of the authors have published thematic articles related to the area of risk management in the energy sector after 2010.

3.3. Conceptual Analysis of Identified Research Sub-Periods

As we indicated in Section 2.2, SciMAT software is used to understand the structure and evolution of a field. In our case, the field is risk management research in the energy sector. Strategic diagrams for the individual time sub-periods (see Figure 2) are presented to analyse the most exposed aspects of this research field. It should be noted that the size of the clusters is proportional to the number of academic papers published on each research topic.

3.3.1. Works Published in 1993–2003 (the First Sub-Period)

As we have shown in Figure 5 and

Table 4. Cluster information and key parameters—sub-period (1993–2003).

Cluster	Centrality	Density	Core Documents			Internal Links
			Count	H Index	Average Citations	Node	Weight
RISK-MANAGEMENT	16.85	19.87	14	8	176.5	COSTS	0.08
						CULTURAL-FACTOR	0.08
						DEVELOPING-COUNTRIES	0.08
						ENERGY-SECTOR	0.12
						HUMAN	0.16
						RISK	0.09
						SOCIAL-ASPECTS	0.12

, the intellectual base in the field of risk research in the energy sector derives from the risk management research that forms its foundation (Figure 5a). The average citation count of the 14 leading papers in this cluster is almost 180, indicating their importance to the development of research in this area. An analysis of the indications of the weights of the links between the main keywords indicates that, in the initial phase, the research focuses on human security and social aspects (keyword weights of 0.16 and 0.12, respectively—Table 4). We note clearly that the research on the energy sector is related to cultural factors (Figure 5b—connecting with a thick line). This is a sub-period of reorientation in the approach to looking at security in the energy sector from a passive approach towards a proactive approach in building a safety climate [45]. Attention was also paid to research on the development of the market for home photovoltaic systems. These works focused mainly on eliminating key barriers to their widespread and accelerated diffusion [72]. This is also a sub-period of intense research development in the area of risk management.

Particularly noteworthy is the study by Klinke and Renn [47], in which they proposed a novel approach to improve the effectiveness, efficiency and political feasibility of risk management procedures. Among other things, they present a unique set of measurements distinguishing nine risk assessment criteria, six risk classes, a decision tree and three categories of management strategies.

3.3.2. Works Published in 2004–2010 (the Second Sub-Period)

In the sub-period 2004 and 2010, the new leading themes are emerging. Considering performance measures (such as number of academic papers and number of citations) and indications of the parameters of the strategy diagram, the following topics stand out: energy systems and risk management (Figure 6,

Table 5. Cluster information and key parameters—sub-period (2004–2010).

Cluster	Centrality	Density	Core Documents			Internal Links
			Count	H Index	Average Citations	Node	Weight
ENERGY-SYSTEM	63.67	107.74	6	3	5.67	ENERGY-TECHNOLOGIES	0.33
						EUROPEAN-COMMISSION	0.38
						NATURAL-HAZARD	0.5
						RELATED-RISK	0.33
						RELIABILITY-PERFORMANCE	0.33
						RISK-INDICATORS	0.33
						ENERGY-RISKS	0.6
						DECISION-SUPPORT-TOOLS	0.33
RISK-MANAGEMENT	66.9	12.16	34	11	19.59	ACCIDENTS	0.05
						DECISION-SUPPORT-SYSTEMS	0.09
						ENERGY-MARKET	0.09
						ENERGY-SECTOR	0.16
						ENERGY-SUPPLIES	0.07
						INVESTMENTS	0.16
						POWER-PLANTS	0.07
						SUSTAINABLE-DEVELOPMENT	0.07

). In addition to the risk management node (whose shift from quadrant one to quadrant two underscores that the issue of risk management in the energy sector is an important basic research area that still needs to be recognized), the “energy-systems”, “developing-countries” and “environmental-impact” nodes have emerged separately (Figure 6a). The thematic network “energy-systems” is characterized by high centrality and density. The density of internal linkages between topics (Figure 6b) is illustrated by the thickness of the lines. A comparative analysis of the parameters for the RISK-MANAGEMENT cluster in relation to the previous sub-period allows the following regularities to be formulated:

-

There is an increase in the importance of research on risk management in the energy sector, as indicated by a more than a doubling of the number of leading publications;

-

Changes in the levels of centrality and density indicators indicate an intensive search for common ground in the existing body of research on risk management and the energy sector (almost fourfold increase in the centrality indicator), with a relatively low decrease in the density indicator, resulting in the emergence of the second ENERGY-SYSTEM cluster.

An analysis of the indication of the relationship weights between the main keywords in the RISK-MANAGEMENT cluster indicates a clear focus on sectoral research. In contrast, the analysis of linkages in the ENERGY-SYSTEM cluster indicates that in this sub-period, the focus of research is on energy system risks arising from natural hazards (keyword weights are 0.6 and 0.5, respectively—Table 5).

The study of energy systems has included the creation of energy risk analysis methodologies. Such a study was undertaken by Colli, Serbanescu and Ale [73]. They made a comparative assessment of the risk issues associated with different energy systems. They proposed a methodology for grouping and ranking selected risk indicators to compare and rank the risks associated with different energy technologies. The research also looked at methodologies for developing a set of risk indicator characteristics to be used to compare risk expressions from different energy systems throughout their life cycle in order to obtain a reliable risk assessment [74]. The search for risk assessment methods also addressed the process of hydropower production in the context of climate change [75].

In contrast, with regard to the network centred around the risk management node, research has focused on security of energy supply (as one of the main goals of EU energy policy) [52]. Issues of environmental and health impacts of energy systems were also examined, with reference to major accident investigations. A study by Hirschberg et al. [76] provides an engineering perspective on energy risk issues in the context of social implications. During this sub-period, the Resource Conservation Challenge was initiated. This was reflected in research on building strategies that lead to less waste as well as more sustainable resource consumption in the energy sector [77].

3.3.3. Works Published between 2011 and 2015 (the Third Sub-Period)

Another 5 years of research on risk management in the energy sector result in the emergence of new leading themes (Figure 7,

Table 6. Cluster information and key parameters—sub-period (2011–2015).

Cluster	Centrality	Density	Core Documents			Internal Links
			Count	H index	Average Citations	Node	Weight
ACCIDENTS	21.22	26.1	13	5	34.08	ACCIDENT-RISKS	0.22
						COMPARATIVE-RISK-ASSESSMENT	0.22
						HEALTH-RISKS	0.25
						HUMAN	0.33
						RISK-INDICATORS	0.22
						SEVERE-ACCIDENT	0.25
						RISK-MANAGEMENT	0.1
						OIL	0.22
INVESTMENTS	13.17	20.52	13	8	32.38	COSTS	0.13
						ENERGY-MARKET	0.16
						ENERGY-RESOURCE	0.06
						ENERGY-SECTOR	0.06
						FOSSIL-FUELS	0.11
						RENEWABLE-ENERGY	0.06
						RENEWABLE-ENERGY-INVESTMENTS	0.14
						STOCHASTIC-MODELS	0.1

). Considering performance measures (such as number of academic papers and number of citations) and indications of the parameters of the strategy diagram, the following topics stand out: accidents and investments. With regard to the issue of risk management during this sub-period, studies on modelling and forecasting the price of fossil fuels, the price of CO₂ emission allowances and the price of electricity were highlighted. The analysis of clusters and keywords parameters for the following period shows the emergence of seven clusters. Among them, clusters concentrated around the keywords ACCIDENTS and INVESTMENTS stand out (Figure 5a). At the same time, the disappearance of the RISK-MANAGEMENT cluster can be observed. This indicates an integration of the previous approach to this issue with other issues common to both areas.

A comparative analysis of parameters for this sub-area (Table 6) allows the following regularities to be formulated:

-

In both distinctive clusters, an analogous number of leading academic papers and their average citations can be noted;

-

In both clusters, the centrality indexes are lower than the density indexes, which indicates the internal thematic cohesion of the clusters;

-

At the same time, a relatively low strength of linkages to other topics in the study areas can be noted.

The analysis of link weights between the main keywords in the ACCIDENTS cluster indicates a clear thematic coherence (link weights in the range of 0.22–0.33). On the other hand, the analysis of links in the INVESTMENTS cluster indicates that in this sub-area, the research focuses on the issue of investment in renewable energy versus fossil fuels and the associated costs (keyword weights of 0.14, 0.11 and 0.13, respectively—Table 6).

As pointed out by García-Martos, Rodríguez and Sánchez [78], the study of these issues will improve decision making by those involved in energy issues. The topics previously dedicated directly to energy systems and risk management are disappearing. In their place, seven new main topics have emerged, derived from studies of an earlier sub-period. Accidents becomes the leading theme. We can notice a group of three basic themes that require further research, in particular the issues of investments and sustainable development. The leading group of subjects in which this issue was addressed was construction firms. At this time, it is important to note the review and analysis of research on construction safety management by Zhou, Goh and Li [49]. They grouped research on construction safety by process, its individual/group characteristics and accident data. Sousa, Almeida and Dias [79,80], on the other hand (also referring to construction companies), proposed proprietary models of accident causes and studies of accident assessments. This has greatly influenced the shaping of safety and health risk management in the workplace. Among the specialized topics that stood out during this sub-period are the issues of greenhouse gases and the identification and management of critical factors. In this area, research by Eckle and Burgherr [81] indicates that the frequency, severity and trends of risk factors influence major differences between activities, highlighting the need for detailed risk analysis. On the other hand, the quantitative–probabilistic approach used by Iwańkowicz and Rosochacki [82], indicates the possibility of using the clustering technique to analyse an accident database and forecast process risks. A separate group of studies in this cluster are those related to risk management, which focused, among other things, on assessing the contagion effect between the energy market and stock markets during the financial crisis [54]). Research on energy sector and renewable energy investment risk was undertaken by Bhattacharya and Kojima [83]. They pointed out a number of findings that are important from the point of view of investors. Among other things, they recommended that when assessing the risk of investments in the energy sector, one should take into account (in addition to the absolute values of the costs of inputs such as fossil fuels) the fluctuations in their costs. They also showed that including renewable energy in an investment portfolio, despite seemingly increasing costs, can reduce the cost of hedging risk. They also pointed out that the international carbon price is not a sufficient incentive for investment in renewable energy.

3.3.4. Works Published between 2016 and 2020 (the Fourth Sub-Period)

Research on risk management in the energy sector falling into the sub-period of 2016–2020 brings further leading themes (Figure 8,

Table 7. Cluster information and key parameters—sub-period (2016–2020).

Cluster	Centrality	Density	Core Documents			Internal Links
			Count	H Index	Average Citations	Node	Weight
GASES	39.98	24.38	35	13	15.43	CONSTRUCTION-PROJECTS	0.21
						COSTS	0.05
						OIL-AND-GAS-PROJECTS	0.27
						RISK-FACTORS	0.09
						STOCK-MARKET	0.08
						STRUCTURAL-EQUATION-MODELLING	0.15
						RISK-MANAGEMENT	0.07
						OIL	0.14
INVESTMENTS	14.86	5.02	17	9	17	ENERGY-COMMODITY	0.05
						FINANCIAL-MARKETS	0.07
						INVESTMENT-DECISIONS	0.05
						INVESTMENT-OPPORTUNITIES	0.05
						MINERAL-RESOURCE	0.05
						PHOTOVOLTAIC-SYSTEM	0.05
						RENEWABLE-ENERGY	0.06
						VALUATION	0.07
SUSTAINABLE-DEVELOPMENT	7.54	5.12	12	9	45.75	CARBON-EMISSION	0.07
						ENVIRONMENTAL-IMPACT	0.07
						ENERGY-RESOURCE	0.04
						GREENHOUSE-GASES	0.07
						FOSSIL-FUELS	0.05
						LITERATURE-REVIEWS	0.06
						SUPPLY-CHAINS	0.04
						ENVIRONMENTAL-ECONOMICS	0.07

). Taking into account performance measures (such as number of academic papers and number of citations) and indications of strategy diagram parameters, the following topics stand out: gases, investments, sustainable development and risk. The base of research during this sub-period becomes the broad issue of gases and critical infrastructure projects (quadrant 1). The topic of sustainable development is gaining in importance. We can observe that research relating to investments and accidents continues. The analysis of cluster and keyword parameters for the following period shows the identification of eleven clusters. Among them, the clusters around the keywords GASES, INVESTMENTS and SUSTAINABLE-DEVELOPMENT stand out (Figure 8a). We still do not observe an individual cluster of RISK MANAGEMENT. This indicates a further integration of this issue with other issues common to both areas.

A comparative analysis of the parameters for this sub-area (Table 7) allows us to formulate the following findings:

-

The highest number of leading academic papers can be observed in the three distinctive clusters in the GASES cluster, with at the same time two/three times lower indications for the other clusters;

-

The average number of citations is highest in the cluster centred around the keyword SUSTAINABLE-DEVELOPMENT, which is placed in quadrant four. This indicates that this is a developing topic that is under-recognised in the context of risk management issues in the energy sector. It may provide suggestions for future research into common issues: risk management, energy sector, and sustainable development.

The linkage analysis between the main keywords in the GASES cluster shows a clear focus on OIL-AND-GAS and CONSTRUCTION-PROJECTS thematic issues (linkage weights in the range of 0.27–0.21). On the other hand, the analysis of links in the INVESTMENTS and SUSTAINABLE-DEVELOPMENT clusters indicate that research is not specialised in these sub-areas (keyword weights are no higher than 0.07—Table 7).

Researchers have highlighted problems related to improving the efficiency of energy projects [84] or identifying important risk factors that directly determine the success of projects (e.g., oil and gas construction projects) [85,86]. Considerable emphasis was placed during this sub-period on research that fit into the theme of sustainable development. Among the most cited works were cross-cutting studies on topics such as urban energy resilience [51] and electric vehicle integration policies [87]. Sustainability issues were mainly related to problems of carbon emissions, environmental impact, and energy resources (e.g., fossil fuels, greenhouse gases) and their diversification in the context of industrial safety risk and environmental management [56,88,89].

3.3.5. Works Published between 2021 and 2022 (the Fifth Sub-Period)

Dynamic changes in the recent sub-period have led us to isolate 2021–2022 as a separate sub-period (Figure 9,

Table 8. Cluster information and key parameters—sub-period (2021–2022).

Cluster	Centrality	Density	Core Documents			Internal Links
			Count	H Index	Average Citations	Node	Weight
RISK-MANAGEMENT	32.4	9.14	63	11	6.43	ENERGY-MARKET	0.07
						COSTS	0.06
						ENERGY-SECTOR	0.06
						HEALTH-RISKS	0.05
						HUMAN	0.05
						INVESTMENTS	0.18
						RISK	0.05
						SUSTAINABLE-DEVELOPMENT	0.05

). Despite the fact that the adopted sub-period is quite short compared to those previously adopted, our analysis showed that new topics have emerged in addition to those already identified in previous sub-periods. The analysis of the parameters of the last sub-period, distinguished by the authors due to particularly intense global events (COVID-19 pandemic, energy crisis caused by Russia’s aggression in Ukraine), does not follow the observed trend of the previous sub-periods. Once again, the RISK-MANAGEMENT cluster has emerged. However, its parameters, e.g., a density index that is relatively low, while at the same time more than three times the centrality index, allow us to conclude that this issue is still a coherent part of other themes. In all likelihood, a five-year analysis (2021–2025) will not show the presence of this individual cluster in the future. Issues related to the COVID-19 pandemic became the basis for consideration of risk management in the energy sector. It should also be noted that issues related to the energy transition, particularly in the context of investments in renewable energy, have emerged as a separate group of sub-themes [90,91,92].

3.4. Dynamics and Evolution of Quantitative Changes in Keywords

We also analysed the overlay graph to capture the changes between the highlighted sub-periods (Figure 10). It illustrates, on the one hand, the number of keywords common to the two consecutive sub-periods, supplemented by a stability index. On the other hand, it allows analysis of the number of keywords that have ceased to function in a given research area (arrow direction—from the circle) and the number of new keywords depicting new issues (arrow direction—to the circle).

Undoubtedly, the development of research on risk management in the energy sector over the past 30 years can be described as stable (the stability index over 5 sub-periods is about 0.6). During the period under review, the number of common keywords (sub-period I vs. sub-period V) has increased dynamically (as much as eight times). In each of the sub-periods, we can see that the number of new keywords is significantly higher (almost twice) than the number of fading words in the study area. This indicates that the field of research on risk management in the energy sector has expanded significantly over 30 years, covering a variety of new topics. This has also been demonstrated by the analysis carried out in the previous subsections. The large number of keywords added to the research area in each sub-period indicates the highly variable nature of risk management research in the energy sector. At the same time, this variability indicates that this area of research will evolve in the future.

The evolution map (Figure 11) highlights the importance of risk management research in the energy sector and allows tracing its paths of development over three decades. From sub-period to sub-period, this research has become increasingly diverse and is undoubtedly characterized by increasing dynamism. As already noted in earlier elements of the analysis, the foundational theme for research on risk management in the energy sector was risk management. Of the four newly identified topics in sub-period two, three originate from risk management issues. Initially, these referred more to research in developing countries and quite loosely to topics addressing aspects of environmental impact.

Topics that fit directly into the specifics of the energy sector appear in sub-period two (as a newly separated topic)—connecting in the next sub-period (i.e., the third sub-period) with the issue of accidents. It should be noted, however, that this link does not apply to the main themes present in both topics. In the next sub-period, we can see a clear link between the issue of risk management in the energy sector and the topics of investment, accidents and sustainability (which all appear as new topics in this sub-period). Of the seven identified topics in sub-period three, three are derived from risk management issues (accident, investment, sustainable development topics). Newly outlined topics include issues related to construction companies, greenhouse gases, algorithms and critical factors. The first of these (problems concerning construction companies) was a temporary topic (no continuation in subsequent sub-periods). On the other hand, of the fourteen themes identified in the fourth sub-period, six are a continuation of issues from the previous sub-period (investment themes, sustainable development, gases—derived from investment themes). Only five of them find continuity in the following fifth sub-period. In the last highlighted, fifth, sub-period, of the six topics, only two do not derive from previous research. These are studies relating to COVID-19 and decision support issues. However, it should be noted that the topic of risk management, in the last sub-period, has gained in importance, as evidenced by the size of the “risk management” node.

4. Discussion

Building on previous research on the issue of risk management in the energy sector, this study contributes to a better understanding of the trends and evolution of research in this area, indicating in particular that research on risk management in the energy sector has evolved over three decades from studying the issue in the context of cultural factors to a multidimensional perception. Investment issues, sustainable development and gases currently dominate among the leading topics in this area, while research on renewable energy and greenhouse gases, in particular for construction companies, can currently be considered as the most important future research priorities.

First of all, it should be emphasised that there were no review studies closely corresponding to the present one, which is due to the adopted thematic scope, the time span and the tool used. This is shown, inter alia, in Table 1. Initially, research focused on reorienting the approach to looking at security in the energy sector from a passive to a proactive approach to building a secure climate [45,71]. This was demonstrated in the first sub-period, which was characterised at the same time by a high interest of researchers in focusing on siloed, specialised approaches to the study of risk management in the energy sector. The second sub-period of the analysis showed that the research concentrated on finding methodological solutions. The findings of our study are consistent with and complement the studies by Le Coq i Paltseva [52], Colli, Serbanescu and Ale [73], Colli et al. [74] and Molarius et al. [75]). As shown in the next sub-section, the researchers’ considerations placed weight on the issues of decision-making problems. This was shown to be consistent with research by García-Martos, Rodríguez and Sánchez [78] and research in the construction industry by Zhou, Goh and Li [49], Sousa, Almeida and Dias [79,80], among others. It has also been shown that more extensive research on the issue of greenhouse gases was initiated during this time. This corresponds, for example, with studies by Eckle and Burgherr [81] and Iwańkowicz and Rosochacki [82]. There was also a wider interest in investment risk issues during this sub-period (e.g., [83]). The next sub-period of research indicates a reorientation of risk management research in the energy sector towards issues of improving the efficiency of energy projects [75,84,86] and sustainability [26,27,30,51,56,87,93].

A overlay and evolution analysis of risk management research in the energy sector should be considered a significant addition to the existing siloed review and mapping studies (see Table 1).

5. Concluding Remarks, Limitations and Ideas for Future Research

In this work, we reviewed the three-decade development of research in the energy risk management field divided into five sub-periods (1993–2003; 2004–2010; 2011–2015; 2016–2020; 2021–2022). We conducted the review by applying bibliometric analysis techniques. This study used bibliometric information on publications indexed in Scopus and WoS databases. Based on PRISMA’s recommendations, 679 academic papers extracted from both databases were finally analysed. The academic papers were retrieved by means of a query in which the search terms were ”risk management” and “energy sector” in the abstract, title or keywords. The analysis was carried out using SciMAT software. Although this study is not the first to analyse the topic of risk management using bibliometric methods, it is the first to analyse its scientific evolution in relation to the energy sector and the first attempt to map a structured conceptual framework of the topic. Therefore, it offers researchers a reference for deepening the indicated research trends in this area.

Research on risk management in the energy sector is clearly progressing at an ever-increasing pace (as shown in the analysis of publication trends). The current turbulence in the energy market (with geopolitical background) allows us to predict with certainty that the volume of research on this issue will increase significantly. Analyses of both the most productive authors and articles (based on citation criteria) have shown that researchers are constantly searching for and exploring new aspects of the topic (out of the 10 most cited papers, 3 were reviews).

We have shown that these studies have expanded steadily over the last 30 years (overlay graph analysis). In contrast, the number of keywords added to the research area in each sub-period (about 100 per sub-period on average) indicates the variable nature of this research. In each of the isolated sub-periods, new research topics emerge. The demonstrated thematic variability, combined with long-term stability, allows us to conclude that this area of study will evolve in the following years.

In addition, pathways for the development of risk management research in the energy industry are identified, and the links between the themes identified in each sub-period are demonstrated. Three main strands of research are identified, which concern the following topics: (1) accident issues; (2) investment issues, with a focus on renewable energy investments and greenhouse gas emissions issues; (3) sustainability issues, with a focus on energy transition issues.

Finally, we point out some limitations of the study and ideas for future research. Firstly, it is important to point out the limitations of the decision-making procedures when using the SciMAT software. The indications of the generated maps depended on the parameters chosen. The software’s functionalities allow for the implementation of different similarity measures and different clustering algorithms. The authors did their best by studying beforehand (i.e., prior to the research described in this article) the experiences of other researchers in the use of different SciMAT functionalities. Future studies may address similar issues by adopting different software parameters (e.g., comparative studies). They could also complement it by conducting a deeper content analysis (e.g., using dedicated content analysis software) of the articles in each extracted thematic set. This would allow a more detailed discussion of each identified topic. Another limitation is the method adopted for separating the sub-periods of analysis. The adoption of other criteria at this stage may also affect the dissimilarity of the thematic nodes created and the links between them. The next step in this study may be to conduct a comparative study, e.g., for sub-periods identified by an equal number of years in each sub-period.

We hope that the results presented in this article will encourage and facilitate new research on the topic of risk management in the energy sector.