Transformations in the European Gas Supply Network Due to the Russia–Ukraine Conflict

Abstract

The objective of this paper is to demonstrate the structural characteristics of the European gas supply system and changes in its network structure and the interaction and clustering among its nodes defined as countries, following the outbreak of the Russia–Ukraine conflict. The methodology relies on social network analysis, which employs mathematics of the graph theory to examine the state and dynamics of the given network structure. The impacts identified involve the reduced reliance of the system on Russian gas, a considerable reduction in the strength centrality of Russia and Germany, and a higher dispersion of gas flows, largely due to the increased import of LNG flows. After the conflict outbreak, countries such as Italy, Austria, the Slovak Republic, and Hungary retained their high influential position, in terms of the PageRank centrality, while the Balkan countries, together with the Middle East gas suppliers (Turkey and Iran), formed a common group with Russia. The estimated changes stress the challenges posed to the EU countries to enhance connectivity infrastructure investments and explore alternative ways of gas supply to support the objectives of strategic autonomy, while promoting resilience and the path toward a carbon-free transition.

1. Introduction

Natural gas is regarded as a cleaner form of energy, compared to oil fuels. The European Union (EU) has fostered the use of natural gas as a transition fuel in the shift from coal to cleaner energy sources in electricity generation and to promote its energy stability. Nevertheless, the EU is largely based on natural gas imports, which, until recently, had increased the energy dependency on Russia [1]. The “special military operation” of Russia against Ukraine launched on 24 February 2022, which escalated into a military conflict across multiple fronts and led most members of the Western alliance to impose sanctions on Russia. Among others, these sanctions targeted specific Russian energy companies and restricted Russian gas exports to the EU markets [2]. However, given that no direct EU-wide sanctions were imposed on imports of Russian natural gas, the Ukraine transit and imports via Turkey and liquefied natural gas (LNG) remained active during 2022–2024 [3]. At the same time, Russia increased gas exports to Asia to compensate for missing revenues from the European gas market [3].

Among others, the Russian–Ukrainian conflict caused disruptions in the gas supply volume and paths, and negative spillover effects on the world economy, contributing to shortages and increased energy prices. Higher oil and gas prices were observed in Europe, especially in the case that alternative sources are more expensive, with geoeconomic implications for both businesses and consumers [4,5]. The European market was found to experience the most substantial impact on natural gas prices due to its reliance on Russian gas, while the market in United States displayed the least price sensitivity [6]. The heightened risk and volatility in European gas prices during the conflict can be attributed to speculative strategies that create self-perpetuating dynamics and diminish market transparency, underscoring the need for regulatory intervention in gas markets [7].

Furthermore, European countries have been working to reinforce their sustainable energy transition, decrease gas consumption and diversify their energy sources to reduce dependence on any single supplier, including Russia [8], which was the main supplier of fossil energy to European countries. This situation accelerated investment in alternative energy sources, like renewables, as a more sustainable and independent option, and infrastructure projects, such as terminal facilities for importing LNG, pipelines, and storage facilities to enhance energy security and reduce uncertainty [9]. The adoption of climate policy strategies to achieve existing targets is expected to further diminish the dependence of European countries on Russian gas imports [3].

The LNG supply has profoundly enhanced the development of the EU as a macro hub and its integration with global gas flows, albeit weather-related conditions, global LNG market pressures and new threats and disruptions to European energy infrastructures significantly affect price uncertainty and the level of risks [10]. During the conflict, many countries have opened temporary direct LNG shipping routes, improving their access to the global LNG network and the overall network efficiency, while several European countries were integrated into the trade community of the United States of America [11,12]. In addition, after the onset of the Russia–Ukraine conflict, LNG transport variations were found to be highly correlated with temporary energy poverty and social conflicts [13]. This outcome signifies the inability to sustain continuous LNG transport in response to global price fluctuations and the need to focus on establishing market alliances and a comprehensive global credit system for energy transactions [13]. Therefore, the gas supply disruptions constitute part of broader geoeconomic changes and geopolitical tensions. These tensions can have long-term effects on international relations and may lead to geographical shifts in energy alliances and partnerships among European countries and their peripheral neighbours to find collective solutions to energy security challenges.

In this article, we use network analysis to identify changes made in structural and geographical patterns of European pipeline gas flows after the outbreak of the military conflict. Commodity transport systems have been long examined from the spectrum of complex network analysis to determine their diverse responses to exogenous shocks. These responses may greatly vary with the type of commodity and shock or failure, and their capacity, connectivity and operational characteristics, technology and control factors, management practices, and geographical scale. Nevertheless, the network analysis of gas pipeline flows is scarce in the current scholarly literature (see Section 2). The Russian–Ukrainian conflict offers the possibility of conducting a real-life experiment regarding the impact of such an exogenous shock on the structure and robustness of the cross-country European gas pipeline network.

The main results of the study verify the decreased reliance of the European gas supply system on Russian gas, the considerable reduction in the influence of Russia and Germany, and the higher dispersion of gas flows, largely due to the increased seaborne LNG flows, after the conflict outbreak. The clustering analysis indicates that, during the conflict, Balkan countries, together with the Middle East gas supplier countries (Turkey and Iran), formed a common group with Russia, mainly through transit pipeline flows. There is also evidence on the growing importance of smaller and peripheral countries in ensuring the system connectedness within and among the network clusters. The proposed approach offers a novel analysis and interpretation of the impact of the conflict spanning an almost three-year period, which renders the findings meaningful and reliable, compared to previous studies based on simulations rather than actual gas flow data. The results can contribute to the original understanding of the multiple scales at which changes in the gas network structure and its constituent clusters take place, and of the potential influence and varying role of individual countries to act both as hubs within their own cluster and/or bridges between different clusters.

As far as the organisation of this article is concerned, Section 2 provides a review of the existing scholarly literature on issues concerning the network analysis of gas (pipeline) supply systems and other studies investigating −earlier or preliminary− effects of the Russia–Ukraine conflict on natural gas flows. Section 3 presents the data sources and setup, and the methodology of the given study based on the use of suitable network metrics. Section 4 describes and discusses the empirical findings of the study and Section 5 summarises and concludes, including policy implications for the resilient and sustainable development of the natural gas supply system.

2. Literature Review

Network theory studies graphs as a representation of relationships between discrete objects. Network analysis techniques encompass a set of mathematical tools, including the modelling of relationships (connections) between discrete objects (nodes), community detection to identify clusters or communities of nodes, community-aware measures that exploit the network’s community structure to find influential nodes within and between groups, and node centrality measures to identify the most important, critical, or (potentially) influential nodes in the network. These tools allow the quantitative evaluation and interpretation of various types of networks, such as social, transportation, or technological networks. The network metrics used here principally come from social network analysis, which employs mathematics of the graph theory and applies them to social world problems. In the current context, the concept of a social network is reinterpreted so connections apply to natural gas transmission among nodes, which are here defined as countries rather than people.

The complex network analysis of pipeline systems can offer a range of valuable tools and insights to understand and interpret their structural, topological, geographical, and operational characteristics as well as strategic interrelationships between the constituent links and nodes, which may refer to various sovereign entities with spatial dimension, like countries and their regions. Specifically, network analysis may allow us to understand the efficiency and robustness of pipeline transport systems and draw conclusions about their vulnerability or resilience in the presence of exogenous shocks or failures (e.g., interruption of gas flow, pipeline closures, accidents, or sabotage). In this respect, it helps to identify critical components, such as the country–nodes in maintaining the stability of the pipeline transport system, and to suggest optimal mitigation measures and plans for the location of gas (or regasification) terminals/hubs, the diversification of routes and sources, the scheduling of maintenance activities, and the allocation of resources to maximise the efficiency and sustainability of gas transport and reduce risks of failure.

The examination of the topological structure of the trans-European gas pipeline network, with the complementary use of social network metrics (betweenness centrality) and the maximum flow method, showed features of considerable efficiency and robustness, in terms of error-tolerance to failures of high load links [14]. In addition, the complex network–analytic comparison of the topological structure of the European and North China natural gas pipeline transport systems demonstrated that the former has greater redundancy, transmission efficiency, and robustness than the latter [15]. Regarding the impact analysis of exogenous shocks, the use of infrastructure and dispatch model and simulations demonstrated that, during the 2009 Russian–Ukrainian gas conflict, the handling of the transit disruption by the European gas sector was efficient [16]. This is mostly due to the gas stocks for mitigating risks from supply disruptions as well as the increased flexibility offered by the reverse flow capabilities of the gas transport system. Nonetheless, the supply chain networks in Europe have long been disturbed due to the imposition of sanctions (against Russia) and counter-sanctions (against Western countries) in 2014, after the annex of Crimea [17]. These sanctions and counter-sanctions transformed the trade network between countries in several markets, such as in agriculture [18,19] and in the energy (oil) sector [20].

More recently, the changes that occurred in the global energy trade network, including that of natural gas, due to the COVID-19 pandemic were investigated by [21] with the use of complex network theory. The results showed that the trade in crude oil and natural gas became more dispersed during the pandemic, while changes in modularity were most noticeable in the natural gas network, showing a significant decrease and separation of energy trade clusters in the early stages of the pandemic, but a relatively fast recovery to the original level in the second year. Moreover, the network analysis of the world oil and gas trade patterns found that the natural gas trade mainly relies on pipeline transport, as it can avoid damage brought about by random attacks such as natural disasters and disease outbreaks to some extent [22]. Hence, the natural gas trade network exhibits relatively higher resilience when suffering random attacks.

Regarding the topic under investigation, the use of complex network analysis to simulate the preliminary cascade failure impact of the Russian–Ukrainian conflict on the global fossil fuel trade patterns demonstrated that the natural gas trade had the largest direct loss, compared to other energy (crude oil and coal) commodities [23]. In addition, the conflict reshaped the pattern of global fossil fuel trade, improving the trade efficiency of European countries. Similarly, cascading failure model simulations under different scenarios, using data referring to 2020, indicated that the gas (and coal) trade network was significantly influenced by the Russia–Ukraine conflict [24]. The results suggested that large countries with increased economic scale would face relatively fewer impacts compared to small and poor countries with limited economic scale, with energy substitution and diversification as effective ways to manage the risks. However, the existing findings are based on simulations using data for years before 2022, rather than actual commodity flow data. Last, the analysis of vessel movement data using various network metrics showed that the Russia–Ukraine conflict caused significant changes in the direction and volume of flows and the resilience/robustness of the global LNG shipping network, suggesting that countries heavily dependent on natural gas need to expand their LNG-receiving terminals and diversify their LNG-sourcing strategies [11,12,25,26].

The present article offers an original network–analytic evaluation of the implications of the military conflict between Russia and Ukraine into the structure of the European gas supply system, which is predominantly based on the cross-border gas pipelines. The analysis utilises actual gas flow data and spans a sufficient time period before and during the conflict. In this way, it allows a comprehensive multiple scale representation of the geopolitical shock impacts on the topological and spatial characteristics of the gas supply system at the level of the whole network, of the gas transmission clusters, and of the partner countries.

3. Materials and Methods

This article employs actual gas flow data originating from the International Energy Agency (IEA). This dataset reports the natural gas flow (mm³/h) through pipelines and the gas transported between Europe and the rest of the world (as Liquified Natural Gas or LNG) on monthly basis, including the cross-border and exit points, and LNG imports at the first landing point, as the aggregate of daily data provided by the European Network of Transmission System Operators for Gas (ENTSOG), i.e., the network of gas Transmission System Operators (TSOs) in Europe. The source of the data for the Netherlands is the Gas Transport Service website, while the source of the data for the Nord Stream is the Opal Nel’s website. The time period spans more than three years during 2020–2023, which allows for the identification of long-lasting impacts of the Russia–Ukraine conflict on the European gas supply network. Specifically, the total study period covers 38 months, i.e., from August 2020 to September 2023. For comparison purposes, this period is symmetrically separated into two subperiods of equal duration: the first one between August 2020 and February 2022 (referred to as “before conflict”), and the second one between March 2022 and September 2023 (referred to as “during conflict”).

Figure 1 illustrates the considerable gradual drop of the total amount of natural gas flow in the European gas supply network during the subperiod of Russia–Ukraine conflict (between March 2022 and September 2023), compared to the subperiod before the conflict (between August 2020 and February 2022). However, it is stressed that the natural gas flows in gaseous state exported from Russia took a few months to decline after the conflict began in February 2022 and started to significantly fall about one year later (February 2023). This lagged response can be attributed to the changing logistics needs and the time required for the EU countries to adjust to the sharp increase in prices and the new supply conditions, and to establish new agreements and partnerships regarding the gas supply from other countries and sources, mostly based on the LNG supply through the European ports. In addition, the conflict has considerably changed the mixture of natural gas supplied to the European countries, as the share of LNG increased from 11.1% before the conflict to 21% during the conflict period under study (Figure 2).

The European natural gas network is represented here as a complex network $G$ ∈ ($V$, $E$), where $V$ is the set of vertices (or nodes) and $E$ is the set of edges (or links) between them, whose structure is modelled as a directed graph composed of a number of $N$ nodes, which refer to the countries participating in the network, and a number of $L$ links that connect pairs of nodes and are weighted by the amount of natural gas transferred between node pairs. The European natural gas network includes 42 countries–nodes: 35 countries from Europe (including Russia), Turkey, Morocco, Tunisia, Algeria, Libya, Georgia, and Iran, plus the rest of the world (global suppliers of LNG).

Three plausible and suitable centrality measures are used here to measure the influential or critical position of each country–node in the European natural gas network: the total weighted degree (strength) centrality, the PageRank centrality, and the betweenness centrality.

Table 1. Technical description of country–node-specific centrality measures.

Metric	Description
Total weighted degree (strength) centrality	It refers to the normalised (in the range 0–1) measure of the total number of connections that a node * i possesses in the network, $d{c}_i^{total}={\displaystyle \frac{d_i^{in}+d_i^{out}}{{\sum }_{j=1}^N\left(d_j^{in}+d_j^{out}\right)}}$, $\mathrm{where}{d}_i^{in}={\sum }_i{a}_{ij}$ is column degree (in-degree) of node i, which denotes the number of edges ending in that node; $d_i^{out}={\sum }_j{a}_{ij}$ is the row degree (out-degree) of node i, which denotes the number of edges originating from that node; and $a_{ij}$ is the entry in the adjacency matrix A of network G, which takes the value 1 if there is an edge between the nodes i and j and 0 otherwise.
PageRank centrality	$\mathrm{It}\mathrm{denotes}\mathrm{the}\mathrm{probability}{\Pi }_r$ $\mathrm{of}\mathrm{any}\mathrm{random}\mathrm{walker}\mathrm{on}\mathrm{the}\mathrm{network}\mathrm{moving}\mathrm{to}\mathrm{a}\mathrm{node}r$ $\mathrm{as}\mathrm{a}\mathrm{first}\text{-}\mathrm{order}\mathrm{stochastic}\mathrm{process},{\Pi }_r=\left[\left(1-\theta \right)/N\right]+\theta {\sum }_i{\Pi }_i{\pi }_{ir}$$,\mathrm{where}1-\theta$ $\mathrm{is}\mathrm{the}\mathrm{probability}\mathrm{of}\mathrm{a}\mathrm{walker}\mathrm{from}r$ $\mathrm{making}\mathrm{a}\mathrm{random}\mathrm{move}\mathrm{to}\mathrm{any}\mathrm{other}\mathrm{node}i$$,{\pi }_{ir}$ $\mathrm{is}\mathrm{the}\mathrm{probability}\mathrm{for}\mathrm{a}\mathrm{walker}\mathrm{of}\mathrm{making}\mathrm{a}\mathrm{move}\mathrm{from}i$ $\mathrm{to}r$$,\mathrm{which}\mathrm{is}\mathrm{equal}\mathrm{to}{w}_{ir}/{\sum }_j{w}_{ij}$$,\mathrm{and}N$ $\mathrm{is}\mathrm{the}\mathrm{total}\mathrm{number}\mathrm{of}\mathrm{nodes}.\mathrm{The}\mathrm{parameter}\theta$ $\mathrm{is}\mathrm{considered}\mathrm{a}\mathrm{damping}\mathrm{factor}\mathrm{ranging}\mathrm{between}0\mathrm{and}1,\mathrm{which}\mathrm{is}\mathrm{typically}\mathrm{set}\mathrm{equal}\mathrm{to}0.85.\mathrm{The}\mathrm{state}\mathrm{probability}{\Pi }_r$ $\mathrm{is}\mathrm{computed}\mathrm{by}\mathrm{iteratively}\mathrm{solving}\mathrm{a}\mathrm{set}\mathrm{of}\mathrm{simultaneous}\mathrm{linear}\mathrm{equations}\mathrm{for}\mathrm{every}\mathrm{node}r$.
Betweenness centrality	$\mathrm{It}\mathrm{is}\mathrm{given}\mathrm{as}\mathrm{the}\mathrm{ratio}\mathrm{of}\mathrm{the}\mathrm{number}\mathrm{of}\mathrm{shortest}\mathrm{paths}\mathrm{that}\mathrm{pass}\mathrm{through}\mathrm{a}\mathrm{node}\mathrm{to}\mathrm{the}\mathrm{total}\mathrm{number}\mathrm{of}\mathrm{shortest}\mathrm{paths}\mathrm{between}\mathrm{any}\mathrm{two}\mathrm{connected}\mathrm{nodes}.\mathrm{By}\mathrm{denoting}\mathrm{it}\mathrm{as}{u}_{jri}$, $\mathrm{the}\mathrm{number}\mathrm{of}\mathrm{shortest}\mathrm{paths}\mathrm{that}\mathrm{pass}\mathrm{through}\mathrm{node}i$ $\mathrm{and}{u}_{jr}$ $\mathrm{the}\mathrm{total}\mathrm{number}\mathrm{of}\mathrm{shortest}\mathrm{paths}\mathrm{connecting}\mathrm{nodes}j-r$$,\mathrm{the}\mathrm{betweenness}\mathrm{centrality}b{c}_i$ $\mathrm{of}\mathrm{node}i$ $\mathrm{is}\mathrm{defined}\mathrm{as}b{c}_i={\sum }_{j,r\ne i}{u}_{jri}/u_{jr}$. $\mathrm{The}\mathrm{above}\mathrm{measure}\mathrm{is}\mathrm{adjusted}\mathrm{to}\mathrm{the}\mathrm{number}\mathrm{of}\mathrm{pairs}\mathrm{of}\mathrm{vertices}\mathrm{not}\mathrm{including}i$ $\mathrm{by}\mathrm{dividing}b{c}_i$ $\mathrm{by}\left[\right(N-1\left)\right(N-2\left)\right]$.

provides a detailed technical description of the definitions, formulae, and/or algorithms used in the present network analysis to calculate these node centrality measures.

The strength centrality of a node implies the probability that this node processes resources across the network. An increased strength centrality relates to higher capability for a country to exercise influence on, have access to, and exchange energy resources (natural gas) with other countries. Hence, it is more likely to be met in countries having large (economic or population) size, and being well developed, accessible, and trade-open markets. The measure of strength centrality can consider the responses of both in-degree (upstream providers) and out-degree (downstream consumers) of a country–node to how a shock (military conflict) spreads and affects −directly or indirectly− the integration of all other country–nodes in the natural gas flow network.

PageRank centrality, which was originally used for extracting information about important Internet nodes and link structures [27,28], can remove the effect of country (market) size, since it considers the constraints in the total gas volume which is sold/bought between supplying/purchasing countries. Being similar in concept to the commonly used measure of eigenvector centrality [29], it denotes how important (potentially leading or influential) a node as a hub is, in the sense that it is linked to other highly connected nodes in the network. However, in contrast with previous studies in the scholarly literature that did not consider the direction of gas flows in pipeline networks [14,15], PageRank centrality is used here to account for the directionality of weighted links, which is appropriate for directed graphs like the present one.

Betweenness centrality describes which nodes are the most connected to other parts of the gas network and, hence, how critical those nodes are for efficiently bridging country pairs in the network. In the setting of social network analysis, the measure of betweenness centrality relies on finding the minimum spanning tree along which gas flows are assigned (all-or-nothing assignment) between an origin–destination country pair. This tree denotes the shortest path connecting all the nodes with the minimum cost, defined as the total sum of constituent links. Country–nodes with an increased betweenness centrality may be regarded as potentially influential, in terms of their ability to mostly influence the diffusion of natural gas among other countries, by facilitating, mediating, hindering, or modifying the connections between them. Therefore, a country receiving high scores in these centrality metrics can be considered as having both an influential role in the transmission and flow distribution of natural gas as well as the connectedness among different countries. It is noted that all centrality values are normalised within the range 0–1 (so that the sum is unity) to express in a meaningful way the competitive position of country–nodes and facilitate comparisons among them.

Regarding the network-wide metrics, the reachable diameter $d_G$ of the graph $G$ can be defined as the maximum shortest path length between any two pairs ($i$,$j$), using only lengths of paths that exist in the weighted graph, and it could be interpreted as how spread out a network is. It can be defined as follows:

$$d_G=\mathit{max}\left\{\hspace{0.33em}{d}_{ij}\hspace{0.33em}\left|\hspace{0.33em}i,j\in V\right|\right\}$$

(1)

The average path length or distance ($\mathrm{AD}$) is derived by dividing the sum of the distances (shortest path lengths) $d_{ij}$ between any two country–nodes $i-j$ by the total number of pairs $P\left\{\left(i,j\right)\left|i,j\in V\right.\right\}$, provided that for any pair ($i$,$j$) in graph $G$ there exists a path through which $i$ can reach $j$ ($i$ → $j$), namely, the following:

$$\mathrm{AD}={\sum }_{i,j}{d}_{ij}/P$$

(2)

The global clustering coefficient $\mathrm{C}$ equals the sum of each country–node’s local clustering coefficient divided by the number of actors (countries) in the network [30,31,32]. It depicts the average probability of future linkage between any two unconnected country–nodes, which are both connected with a third country–node. The local clustering coefficient $\mathrm{C}\left(i\right)$ is measured as the ratio of the number of country $i$’s triangles with complete connections among each other, to the total number of country $i$’s triangles, which are formed with at least two linkages among them. It can be specified as the average probability of a country–node $i$ to have $\mathrm{E}\left(i\right)$ neighbours connected and is computed on the number of triangles configured by node $i$ to the number of total triplets $k_i·\left(k_i-1\right)$ shaped by this node [33].

$$C={\displaystyle \frac{\sum _{i}C\left(i\right)}{N}},\mathrm{where}C\left(i\right)={\displaystyle \frac{E\left(i\right)}{k_i\cdot \left(k_i-1\right)}}$$

(3)

The Newman’s method [34] is implemented for the accurate clustering of country–nodes into gas supply groups. This method relies on the structural characteristics of network topology, taking into account basic concepts of statistical mechanics, as it connects the strength centrality and interaction among individual nodes with the formation of communities (clusters) and partitions. Compared to other algorithms [35], which use the centrality measures themselves as optimisation criteria, this one determines highly influential nodes for both their own community and the whole network. Hence, it is regarded to provide meaningful results, in the sense that gas transmission strategies tend to create clusters with the largest possible interaction among suppliers and buyers within a group. The measure of modularity to be maximised is given as follows:

$$Q=\left(1/2m\right){\sum }_{ij}\left[w_{ij}-\left(s_i{s}_j/2m\right)\right]\delta (g_i,g_j),$$

(4)

where $w_{ij}$ denotes the weight of the edge between nodes $i$ and $j$; $s_i={\sum }_j{w}_{ij}$ and $s_j={\sum }_i{w}_{ij}$ are the weight sums of the edges attached to $i$ and $j$, respectively; $g_i$ and $g_j$ are the communities to which nodes $i$ and $j$ are assigned; $\delta$ function is 1 if $g_i$ = $g_j$ and 0 otherwise; and $m=(1/2){\sum }_{ij}{w}_{ij}$.

Regarding the community-aware measures to consider group characteristics, in addition to the number of links among constituent country–nodes, the group density and the external–internal link index ($EII$) are calculated. The (graph) density of a group can be defined as the ratio of the existing number of edges (cross-country links) to the maximum number of (or total number of possible) edges [33]. The maximum number of edges corresponds to the case where all countries of the European gas network connect with each other. Hence, the density values range from 0 to 1. A high value of density (close to unity) denotes that the gas network is denser and country–nodes are more cohesive, while a low density value (close to zero) signifies a less connected (sparse) network. A dense network structure can be regarded as more cost-effective and often more robust than that of a sparse network, as the natural gas can flow faster due to the reduction in average distance (or path length), facilitating gas transmission between any two country–nodes and increasing the number of feasible alternative connection paths. The $EII$ signifies the openness of a group to the whole network and how well it collaborates with the other groups [36]. It is defined for each group as the following ratio:

$$EII={\displaystyle \frac{EL-IL}{EL+IL}},$$

(5)

where the internal links ($IL$) join two country–nodes of the same group, while the external links ($EL$) join two country–nodes from different groups.

Furthermore, the measure of the modularity vitality ($MV$) index is calculated here to examine the structural features of each cluster. Specifically, $MV$ can depict potential changes in the partitioning quality of gas supply clusters when a single country–node is removed [37]. $MV$ can differentiate a hub in its own community from a bridge across different communities based on Newman’s modularity. It is defined as follows:

$$MV\left(\mathrm{i}\right)=Q\left(G\right)-Q\left(G\backslash \left\{i\right\}\right)$$

(6)

where $Q\left(G\right)$ is the network’s modularity and $Q\left(G\backslash \left\{i\right\}\right)$ is the network’s modularity after the removal of node $i$. By ordering country–nodes from positive to negative magnitude, the measure of $MV$ identifies hubs first and, as we move to lower (negative) values, bridges. According to this measure, distant hubs may have a larger impact on the community structure than those situated at the core part of the network. In the case of reduced resources, especially in networks with a medium or strong community structure, diffusion expands better through nodes actings as bridges rather than hubs [38].

Furthermore, centralisation indexes are calculated to illustrate the variability of centrality measures [39,40]. The network centralisation index ($\mathrm{N}\mathrm{C}\mathrm{I}$) is expressed as follows:

$$\mathrm{N}\mathrm{C}\mathrm{I}=\left[{\sum }_{1<\mathrm{i}<\mathrm{N}}\left({\mathrm{C}}_{\mathrm{m}\mathrm{a}\mathrm{x}}-{\mathrm{C}}_{\mathrm{i}}\right)\right]/\left(\mathrm{N}-1\right),$$

(7)

where ${\mathrm{C}}_{\mathrm{m}\mathrm{a}\mathrm{x}}{\left\{{\mathrm{C}}_{\mathrm{i}}\left|1\le \mathrm{i}\le \mathrm{N}\right.\right\}}_{\mathrm{m}\mathrm{a}\mathrm{x}}$ is the operator for the highest value of some centrality measure $\mathrm{C}$ (any of the three selected for calculation here) in a network composed of $\mathrm{N}$ nodes. The larger the value of $\mathrm{N}\mathrm{C}\mathrm{I}$, the higher the non-uniformity (skewness) of the distribution of each type of centrality measure with respect to the “most central” node in the European gas supply network. Additionally, more concentrated network structures, as implied by higher $\mathrm{N}\mathrm{C}\mathrm{I}$ values, are less balanced and more vulnerable to large shocks, as they rely on the existence and functioning of a few important country–nodes.

4. Results

4.1. Network-Wide Mapping and Metrics

Regarding the impact of the Russia–Ukraine conflict at the level of the whole European gas supply system, the values of several network-wide metrics are considered before and during the conflict (

Table 2. Values of European gas network metrics before and during conflict.

Key Measure	Pro Conflict	During Conflict	% Change
Links	118	124	5.08
Reachable Diameter	45,297.92	66,516.59	46.84
Average Distance	9713.18	12,465.85	28.34
Global Clustering Coefficient	0.22	0.23	2.89
Newman Modularity	0.50	0.47	−6.61
Strength Centralisation	0.14	0.10	−27.38
PageRank Centralisation	0.26	0.22	−15.37
Betweenness Centralisation	0.30	0.25	−16.67

). In addition, Figure 3 depicts the considerable impact of the conflict on the size of pipeline gas flows as well as on the distribution of flows across the European gas pipeline network, with the shift in the intensity of flows from the east to the west of the continent. By and large, the findings show that, during the conflict, the European gas network became more dispersed, but also more equitable in terms of the variance of the gas flows and access conditions among countries.

Both the average distance and the reachable diameter (using lengths of paths that exist in the network) were increased, signifying the expansion of the network size. Both the number of links and the global clustering coefficient were slightly increased, but the Newman’s modularity was reduced, denoting a weaker network partitioning during the conflict, compared with the period before it. All centralisation measures that denote the dispersion of the values of the corresponding centrality metrics across country–nodes were reduced after the outbreak of the conflict. The reduction in the total weighted degree (strength) and PageRank centralisation values during the conflict suggest a more dispersed network structure. Regarding the betweenness centralisation, which relies on the shortest paths between country pairs, its reduction also signifies a more dispersed network structure and longer shorter paths among country pairs.

These results can be attributed to the strategic responses of the European gas network agents (national governments and system operators) to multisource inputs and invest to make their energy trade relationships stronger, more secure, and less fragile during the energy crisis resulting from the Russia–Ukraine conflict. These actions led to greater dispersion or spatial diversification of the natural gas supply chains, through broadening the geographical scope of origin–destination markets, and increasing the importance given to peripheral countries in Europe and beyond it, through transient pipeline flows and regasification (LNG-receiving) hub facilities [4]. Such responses can be regarded as consistent with supply network formation strategies in crisis situations typically characterised by degraded connectivity and increased uncertainty of supply link operations [41].

4.2. Natural Gas Network Clusters and Their Characteristics

The results of community detection stress the existence of various clusters or country groups in the European gas supply network, whose formation can be explained by the geography, trade partnerships/agreements for gas supply, and the level of development and liberalisation of the gas industry in each country (see, e.g., [42] for a comprehensive review of the European gas hubs and market development). Figure 4 maps the cross-country groups of the European gas pipeline supply before and during the conflict.

Table 3. The composition and characteristics of the communities (cross-country groups) composing the European gas supply system before and during the Russia–Ukraine conflict.

Group	Size	Density	EII	Members
Before conflict
1	12	0.174	0	World, Algeria, Spain, Belgium, France, Portugal, United Kingdom, Luxembourg, Norway, Croatia, Ireland, Morocco
2	10	0.156	−0.120	Romania, Iran, Turkey, Bulgaria, North Macedonia, Greece, Albania, Serbia, Moldova, Georgia
3	9	0.264	0.050	Italy, Switzerland, Austria, Slovak Republic, Hungary, Ukraine, Tunisia, Libya, Slovenia
4	9	0.236	0.128	Estonia, Finland, Germany, Latvia, Netherlands, Czech Republic, Denmark, Sweden, Russia
5	3	0.333	0.636	Lithuania, Poland, Belarus
During conflict
1	15	0.133	0.082	World, Algeria, Spain, France, Portugal, Estonia, Finland, Latvia, Lithuania, Poland, Denmark, Sweden, Croatia, Belarus, Morocco
2	10	0.144	0.103	Romania, Iran, Turkey, Bulgaria, Russia, North Macedonia, Greece, Serbia, Moldova, Georgia
3	8	0.286	0.086	Belgium, Germany, Netherlands, United Kingdom, Czech Republic, Luxembourg, Norway, Ireland
4	6	0.267	0.200	Italy, Switzerland, Tunisia, Libya, Slovenia, Albania
5	4	0.750	0.379	Austria, Slovak Republic, Hungary, Ukraine

Note: The bold black letters denote countries for each group having the highest total degree (ignoring link values) using only links to/from nodes in their own group (internal links) and the bold blue letters denote countries for each group with the highest total weighted (strength) degree (if they are different from the former first ranking). The external/internal index ranges between −1 (all links are internal and the group is perfectly silo) to +1 (all links are external); a score of 0 means an equal number of internal and external links. The internal link count is the number of links that connect one group node to another group node. The external link count is the number of links that connect a group node to a non-group node.

describes the composition and main characteristics of these clusters in the European gas pipeline network before and during the conflict, which remained the same in number (five) but were considerably reorganised with varying sizes.

While the graph density of the two largest network groups (i.e., group 1 mostly connected with the world LNG suppliers, and group 2 encompassing southeastern Europe and Middle East) was reduced, denoting their dispersion, the remaining groups became denser and smaller in size after the outbreak of the conflict, compared to the earlier period. In addition, the external–internal link ($EII$) index increased in all but one groups of the European gas supply network during the conflict, signifying the growing extroversion and rising tendency of the gas network clusters to collaborate (exchange resources) with each other.

A notable change observed after the outbreak of the conflict concerns the formation of group 3 including both Germany and Norway, between which a significant amount of gas flow is transferred, compared to the period before the conflict, when Norway and Germany belonged to different gas clusters (group 1 and group 4, respectively). Specifically, before the conflict, the gas pipeline network link “Nord Stream” from Russia to Germany (both included in group 4) was the predominant one (Figure 3a), while during the conflict, the link transporting the largest amount of gas flow is from Belgium to Germany (Figure 3b), through the Eynatten cross-border interconnection point. The latter link allows access to the Trans Europa Naturgas Pipeline (TENP) and LNG transport to Germany.

In contrast with the period before the conflict, where the southeastern European (Balkan) countries, together with the Middle East gas supplier countries (Turkey and Iran), transported gas through a network of “transit” pipelines, forming an individual cluster (group 2), during the conflict, these countries were more closely connected and created a common group with Russia (group 2). Therefore, after the outbreak of the conflict, Russia moved its geographical sphere of economic influence, in terms of the natural gas transmission, from central and northern European countries (as shown in group 2 before the conflict) to Southern Europe and the Middle East.

The Iberian Peninsula (Spain and Portugal), together with part of the North Africa market (Morocco and Algeria), constitutes a subgroup of the largest natural gas cluster (group 1), which is heavily dependent on and linked to LNG. Before the conflict, this subgroup formed a single cluster with countries mostly belonging to western Europe (Belgium, France, the United Kingdom, Luxembourg, Ireland) and Norway. However, during the conflict, these close connections were shifted towards the north and northeastern European countries, especially with the inclusion of former group 5 (composed of Lithuania, Poland, and Belarus) in group 1.

In any case, the spatial contiguity of natural gas clusters is strong in the central–eastern and southeastern parts of Europe and beyond it, where natural gas is predominantly carried out by land transport (via pipelines). In contrast, in the western and northern parts of Europe, this spatial contiguity is relaxed, particularly after the outbreak of the conflict, due to the increased LNG supply by ships, mostly from the United States of America, which accounted for nearly half (48%) of total LNG imports to Europe in 2023, followed by Qatar (14%) and Russia (13%) [4]. Regarding the remaining clusters, group 3 (before the conflict) was separated (during the conflict) into two distinct clusters, i.e., group 4 and group 5, wherein Italy and the Slovak Republic constitute the most central nodes (country–hubs), respectively.

Furthermore,

Table 4. Countries with 10 largest modularity vitality values acting as group hubs before and during Russia–Ukraine conflict.

Rank	Country	Value
Before conflict
1	Belarus	0.040136
2	Austria	0.038875
3	Slovak Republic	0.028373
4	Italy	0.015490
5	Bulgaria	0.015009
6	Spain	0.013630
7	United Kingdom	0.013431
8	Ukraine	0.011346
9	Tunisia	0.010275
10	Greece	0.010154
During conflict
1	Turkey	0.055167
2	Italy	0.036047
3	Slovak Republic	0.030400
4	Spain	0.027553
5	Tunisia	0.024683
6	Bulgaria	0.022659
7	Ukraine	0.017177
8	Georgia	0.007948
9	Austria	0.007850
10	Iran	0.006313

Note: Modularity vitality measures the change in the cluster quality (measured by modularity) when a single node is removed. A positive score indicates that the node acts as a hub within its group by contributing to its community structure. If the node of interest has a higher than normal value (greater than 1 standard deviation(s) above the mean) the row is coloured red. The row is green if the node is within 1 standard deviation of the mean.

shows the countries with the largest modularity vitality ($MV$) scores during the conflict compared to the period before the conflict. These countries are considered as having a critical role as group hubs contributing to the connectedness of their own community structure. The results indicate that Italy has retained its position as a hub in its own group and increased its $MV$ score. Moreover, there are several countries which have considerably increased their position as group hubs during the conflict, compared to the period before the conflict. These countries are Turkey and Bulgaria (group 2), the Slovak Republic (group 5), Spain (group 1), and Tunisia (group 4). Georgia and Iran (group 2) are also countries beyond Europe that are included among those having the largest modularity vitality values after the outbreak of the conflict.

In contrast,

Table 5. Countries with 10 smallest modularity vitality values acting as group bridges before and during Russia–Ukraine conflict.

Rank	Country	Value
Before conflict
1	World	−0.021495
2	Norway	−0.014752
3	Germany	−0.012699
4	Belgium	−0.011309
5	Turkey	−0.009889
6	Russia	−0.008241
7	Netherlands	−0.008193
8	Switzerland	−0.006158
9	Lithuania	−0.001497
10	Croatia	−0.000886
During conflict
1	World	−0.033746
2	United Kingdom	−0.032369
3	Norway	−0.020948
4	France	−0.018300
5	Belgium	−0.013634
6	Denmark	−0.007303
7	Czech Republic	−0.004509
8	Germany	−0.004508
9	Switzerland	−0.001925
10	Russia	−0.001379

Note: Modularity vitality measures the change in the cluster quality (measured by modularity) when a single node is removed. A negative score indicates that the node acts a bridge by connecting different groups. The row is coloured green if the node is within 1 standard deviation of the mean. The row is coloured blue if the node has a lower than normal value (less than one standard deviation(s) below the mean).

indicates the countries with the smallest modularity vitality scores, representing country–nodes acting as group bridges in the whole European gas supply network. The world (suppliers of LNG) together with Norway and Belgium retained their influential position as group bridges, while countries such as the United Kingdom and France have upgraded their position among the top group bridges after the outbreak of the conflict. It is also stressed that Germany and, especially, Russia, which had an increased role as group bridges before the conflict, are found to lose their influential position, reflecting the strategic efforts of the European Union to enhance its energy security through reducing the pivotal role of Russia in the European gas market.

4.3. Results of Country-Level Centrality Metrics

Regarding the changes in network centrality metrics at the country–node level (see

Table 6. Results of network centrality metrics for countries in European gas supply network before and during Russia–Ukraine conflict.

Country–Node	Strength Centrality		PageRank Centrality		Betweenness Centrality
	Before	During	Before	During	Before	During
Albania	0.008	0.015	0.014	0.012	0.000	0.000
Algeria	0.006	0.007	0.000	0.001	0.000	0.000
Austria	0.053	0.027	0.168	0.238	0.023	0.000
Belarus	0.028	0.004	0.000	0.001	0.000	0.000
Belgium	0.038	0.090	0.010	0.007	0.015	0.009
Bulgaria	0.013	0.022	0.006	0.005	0.018	0.011
Croatia	0.002	0.003	0.042	0.008	0.063	0.049
Czech Republic	0.059	0.021	0.022	0.013	0.003	0.000
Denmark	0.003	0.008	0.002	0.002	0.004	0.009
Estonia	0.001	0.001	0.003	0.034	0.023	0.021
Finland	0.002	0.001	0.004	0.023	0.028	0.020
France	0.037	0.055	0.018	0.009	0.020	0.024
Georgia	0.007	0.008	0.000	0.001	0.000	0.000
Germany	0.164	0.126	0.034	0.023	0.022	0.027
Greece	0.013	0.023	0.006	0.004	0.000	0.000
Hungary	0.004	0.010	0.052	0.087	0.044	0.049
Iran	0.006	0.006	0.000	0.001	0.000	0.000
Ireland	0.003	0.003	0.003	0.001	0.000	0.000
Italy	0.051	0.053	0.277	0.195	0.040	0.032
Latvia	0.002	0.002	0.002	0.013	0.014	0.021
Libya	0.002	0.002	0.000	0.001	0.000	0.000
Lithuania	0.006	0.008	0.002	0.003	0.003	0.021
Luxembourg	0.001	0.000	0.002	0.001	0.000	0.000
Moldova	0.000	0.000	0.001	0.002	0.063	0.049
Morocco	0.004	0.000	0.000	0.001	0.000	0.000
Netherlands	0.030	0.033	0.008	0.005	0.000	0.000
North Macedonia	0.000	0.000	0.002	0.001	0.000	0.000
Norway	0.065	0.081	0.000	0.001	0.000	0.003
Poland	0.047	0.014	0.004	0.008	0.000	0.000
Portugal	0.004	0.005	0.050	0.005	0.000	0.000
Romania	0.002	0.003	0.014	0.008	0.000	0.047
Russia	0.064	0.028	0.003	0.003	0.007	0.005
Serbia	0.002	0.010	0.009	0.001	0.020	0.020
Slovak Republic	0.057	0.028	0.023	0.119	0.010	0.001
Slovenia	0.001	0.002	0.017	0.034	0.000	0.000
Spain	0.029	0.036	0.058	0.010	0.010	0.034
Sweden	0.001	0.000	0.002	0.001	0.034	0.024
Switzerland	0.007	0.014	0.123	0.040	0.015	0.013
Tunisia	0.015	0.018	0.000	0.001	0.479	0.489
Turkey	0.055	0.059	0.006	0.005	0.000	0.000
Ukraine	0.025	0.014	0.004	0.064	0.017	0.017
United Kingdom	0.029	0.053	0.003	0.002	0.000	0.000
World	0.056	0.105	0.006	0.002	0.024	0.006

Note: Blue (Red) coloured figures show 10 top (bottom) values for each centrality metric.

), on the one hand, the strength centrality was reduced in a few large countries after the outbreak of the conflict, particularly, in Russia (from 0.064 to 0.028, i.e., −56%) and, to a lesser extent, in Germany (from 0.164 to 0.126, i.e., −23%), which remained the country with the largest strength centrality in the European gas supply network. In addition, significant losses in strength centrality (more than −50%) took place in some eastern European countries, such as Belarus, Poland, the Czech Republic, and the Slovak Republic. On the other hand, the position of world LNG suppliers, Belgium, and Turkey was upgraded at the top five strength centralities, while Norway was maintained in this category and also improved its strength centrality. Several large countries, such as France, the United Kingdom, Italy, and Spain, also improved their strength centrality in the European gas supply network during the conflict compared to the period before the conflict.

Given that the European countries are basically gas importers (except Russia, Norway, and the United Kingdom) (Figure 3), this increase in strength centrality can be largely related to the higher in-degree centrality of the given countries, namely, their improved capacity to have access to and import (and store) natural gas compared to other countries in the network. Smaller-sized countries located in the southeastern part of Europe, such as Greece, Bulgaria, and Albania, also increased their strength centrality (more than 50%) during the conflict compared to the period before the conflict. These results show the increased capacity of Balkan states in supporting the continuity of natural gas supply, either as recipients and distributors of seaborne LNG flows or as facilitators of pipeline gas flows originating from areas outside Russia to the rest of Europe.

As far as the PageRank centrality is concerned, the countries which retained their top influential position during the conflict include Italy (though it has reduced PageRank centrality), Austria, the Slovak Republic, and Hungary (the latter three increased their PageRank centrality). Baltic countries such as Estonia, Latvia, Finland, as well as Ukraine and Slovenia also considerably increased their PageRank centrality. Thus, it is observed that several countries belonging to the eastern bloc of Europe enhanced their competitive position as gas hubs (connected with other highly connected countries) in the European gas supply system, despite the losses in strength centrality, given that PageRank centrality controls for the market size (total import and export gas volume). In contrast, the PageRank centrality was reduced for many countries of the “old” Europe, such as Switzerland, Germany, Spain, and France.

Furthermore, the countries which continued to function as critical nodes along the European gas transport corridors, possessing the highest betweenness centrality values after the outbreak of the conflict, include Tunisia, which steadily accounts for about 50% of the maximum betweenness centrality, Croatia (reduced by −22%), Moldova (reduced by −22%), and Hungary (increased by 11%). Large countries such as Germany, France, and Spain, as well as Romania and Baltic Sea countries (Lithuania, Latvia) increased their intermediacy role in maintaining the stability of the European gas supply network. In contrast, the betweenness centrality was mostly reduced for Sweden and Finland as well as the world LNG suppliers due to the geographical spreading of gas supplies by sea, making the European gas transmission system less vulnerable to supply shocks.

By considering information included in the measures of weighted degree centrality, PageRank centrality, betweenness centrality, group hubs, and group bridges, we can then identify which countries–nodes constitute the most significant hubs in the European gas supply system, as the most well-connected, transacting with other highly connected ones, and facilitating the interconnectedness among other countries in their own group, between different groups, and in the whole network. This approach is consistent with the fundamental concept of transport hub, which combines both features of geographical centrality and flow intermediacy [43]. On this basis, it is observed that there are countries in the European gas supply network that have a relatively high strength centrality (Germany, Belgium, Norway, Turkey, France, the United Kingdom, Italy, Spain) or high PageRank centrality (the Slovak Republic, Hungary) and, at the same time, possess an increased intermediacy role either in the whole network or in their own group (as group hubs) or between different groups (as group bridges).

The current findings amplify other preliminary ones in the existing scholarly literature about the impact of the Russian–Ukrainian conflict on the reduction in natural gas trade and the reorganisation of the gas trade patterns [23,24]. Nonetheless, the use of suitable network metrics in the current analysis offers novel and deeper insights into understanding the varying scales and types of impacts of the Russia–Ukraine conflict on different countries. Specifically, the analysis stresses the fact that, not only the economic size, but also the geographical location, topological position, and cluster participation of countries can play an important role in these impacts. Compared to the outcome of earlier conflicts between Russia and Ukraine, such as that in 2009 [16], the gas supply disruptions due to the ongoing military conflict can be regarded as more systematic and influential on the gas supply network structure. Although this conflict has not substantially affected the modularity of the gas supply system (see Table 2), as it was also found during the COVID-19 pandemic [21], the composition of network groups presented some considerable changes, which can be attributed to modifications in the gas flow paths, supply origins, gas composition in favour of LNG, and the cross-country energy partnerships shaped after the conflict outbreak. These changes are found to have a global impact on gas supply chains, leading to systematic shifts in network centralisation and clustering, compared to the case where the system would suffer from natural disasters or random attacks associated with localised impacts [22].

Summing up, the present study offers a comprehensive network analysis of the impacts of a geopolitical shock (the Russia–Ukraine conflict) on the topological and spatial distribution characteristics of the gas supply system at multiple scales, involving the levels of the whole network, the gas transmission clusters, and the partner countries. The results show an overall reduced reliance of the European gas supply system on Russian gas, a considerable reduction in the strength centrality of Russia and Germany, and a higher dispersion of gas flows, mainly due to the increased seaborne LNG flows. Based on the PageRank centrality, which controls for country size, countries such as Italy, Austria, the Slovak Republic, and Hungary retained their high influential position after the conflict outbreak. The clustering analysis demonstrates that, during the conflict, Balkan countries, together with the Middle East gas supplier countries (Turkey and Iran) formed a common group with Russia, mainly through transit pipeline flows. Therefore, it can be argued that the Russia–Ukraine conflict has brought about considerable changes in the network structure of the European gas supply system, particularly in terms of the cluster composition and the growing importance of mostly smaller and peripheral countries in ensuring the connectedness within and among network clusters. As it is further discussed in the next section, these findings can have useful implications for the design and evaluation of targeted investments and reforms to reinforce the gas supply system resilience and promote the global energy transition.

5. Discussion and Conclusions

This article investigated the changes in the natural gas supply across Europe due to the Russia–Ukraine conflict from a network analytic perspective. The consideration of all possible direct and indirect linkages and clusters in the gas transport system across the wider region of Europe can contribute to disentangling the varying impacts of such exogenous shocks, deepening the understanding of the role of individual countries and of cross-country partnerships in this system, making it more robust, sustainable, and equitable. The findings underline that the European gas transmission system is imbalanced, with most of the gas flows being concentrated in a few countries. However, the military conflict and subsequent actions of European countries to become less reliant on Russian gas rendered the gas flows further dispersed and more balanced, in terms of the access of countries to natural gas and, hence, less vulnerable to large shocks.

By and large, the present findings underline that the geography of the European gas supply system broadly follows various levels of gas hub development after the Russia–Ukraine conflict, albeit the geography and development of gas hubs also reflect different stages of evolution in gas price formation and liberalisation of the gas market. The findings here corroborate earlier ones [42,44,45], in the sense that the evolution of the European gas hubs does not correspond to one homogenous gas market, neither in terms of energy infrastructure nor of political partnerships, and that even within each geographical area there can be many levels of development. The gas flow network analysis highlighted the increasing role of smaller states and peripheral countries during the Russian–Ukrainian conflict, both in Europe and beyond it, through transit pipelines (e.g., the TurkStream passing through the Black Sea) and regasification hub facilities to recover the loss of gas flows previously exported by Russia through pipelines.

The European natural gas network is found to have the ability to adjust to the changing conditions by substantially maintaining its macro-performance, in terms of modularity, through reorganising its clusters, dispersing the influential position of the constituent country–nodes across smaller and peripheral ones, and enhancing the critical role of global LNG suppliers (at the expense of energy prices). The results can help us obtain more insights into the natural gas transport network, whose structural properties have been overlooked in comparison to other types of transport networks in the current literature. They can also support the identification of countries which have more influential position and/or critical role in maintaining the connectedness of their own cluster and of the whole system. Among others, they underscore the significance of international gas trade agreements or partnerships between Europe and the North African and Arab countries to retain the security and robustness of energy supply chains in the wider area [46,47].

Such a partnership refers to the Union for the Mediterranean (UfM) Gas Platform, which aims at the gradual development of a common Euro-Mediterranean gas market. More recently, in the same direction, the so-called “Piano Mattei” (Mattei Plan) sets long-term objectives to upgrade the role of Italy as an energy (natural gas and hydrogen) logistics hub through strengthening its connectivity with Africa [48]. This plan also includes a humanitarian dimension, as it may help to diffuse growth on African countries and reduce migration flows toward Europe. Nonetheless, reinforcing institutional capacity and green energy production in African countries can be regarded as important prerequisites to foster sustainable development [49].

Furthermore, the outcomes stress that the influence of Russia in the European gas supply system is still evident, albeit differentiated from the past, largely through its tightly knit connections with Balkan countries, which seem to be dependent on Russian and Middle East gas supplies, such as those arriving through the network of “transit” pipelines from Turkey, Iran, and Georgia. These close connections of Russia have replaced previous ones with Germany, Netherlands, the Czech Republic, and other countries bounding the Baltic Sea before the conflict with Ukraine. Recently, the EU adopted a new (14th) package of sanctions against Russia that prohibits the import of Russian LNG into specific terminals which are not connected to the EU gas pipeline network [50].

Nonetheless, the role of natural gas still exists and is considered important as a transient fuel to stabilise the energy supply, while addressing problems arising in periods of increased demand. The slowdown of decarbonisation process and the increased cost of global energy transition increase the need to promote natural gas system resilience. This tension varies among countries and relies on a multitude of factors which are global-wide and/or local-specific. These factors may encompass the capacity of each country to innovate, store and produce its own (alternative) energy resources, securely and economically exchange energy resources in its own territory and with other countries, align with international environmental agreements, and adjust its economy to a carbon-free growth model. Hence, natural gas is expected to remain in the foreseeable future as part of the energy mix in many countries.

The persistent focus of several European countries on supporting the import of natural gas (in either gaseous or liquid state), as a “cleaner fuel”, or even retaining the partial usage of coal or lignite, can also be attributed to the lack of proper financing and innovation in promoting renewable (solar, wind, geothermal, biofuel, and hydroelectric) energy and low-carbon alternatives, such as nuclear energy, and delays in developing strategic infrastructure projects to enhance grid integration among them and with peripheral non-European countries (e.g., around the Mediterranean region). In turn, these delays and shortages, together with speculative strategies and insufficient market regulation, tend to raise energy prices as well as price uncertainty, in the presence of increased geopolitical risk and changing global political coalitions. In addition, they hinder the achievement of global energy and environmental targets about the reduction in greenhouse gas emissions and environmental pollution, and the promotion of global energy security and independence from fossil fuel imports.

Given the multitude of factors influencing green transition, a pragmatic sustainability approach is necessary to ensure that the ambitious goals toward environmental protection are socioeconomically acceptable, without adverse impacts on economic performance and social progress [51,52]. In this respect, a more holistic decision-making process is required, involving the participation of all stakeholders and preservation of ecosystems to boost all dimensions of environmental, economic, and social sustainability. Recent moves of the EU toward this pragmatic sustainability approach involve the shift in the Green Deal to a more flexible and realistic implementation, through revising (too disruptive) targets and milestones, such as milder agricultural reforms, by providing more incentives and technological solutions to farmers. The Council Recommendation (as of 16 June 2022) on ensuring a fair transition towards climate neutrality (2022/C 243/04) stresses the need to accelerate the green transition in a just, equitable, and inclusive way, recognising that not all regions or industries can adapt equally, so as to treat energy poverty and mitigate regressive impacts of policy measures. Accordingly, the Just Transition Fund supports coal-dependent and industrial regions in shifting to greener economies [53] and the Social Climate Fund helps vulnerable social groups cope with rising energy costs and carbon pricing [54]. Additionally, the EU Emissions Trading System (ETS) has been expanded to include the maritime shipping industry, while there are plans to include new sectors, like road transport, buildings, and agriculture [55,56,57], but with phased implementation and social safeguarding.

As a conclusion, it can be argued that European countries should exploit the current trends in reorganisation of the natural gas logistics network, to support the attainment of strategic objectives with respect to energy self-sufficiency and stability, diversifying existing supply sources and paths, and reinforcing interconnectivity in the European market. Particularly, the countries that are heavily dependent on oil and gas need to focus on the deployment of alternative projects, such as renewable energy, energy storage and saving, smart grids, and green technologies. The role of renewable technologies is regarded as particularly important in reinforcing supply security and reducing the level of risks, while aligning with ambitious net-zero objectives of the EU countries. The EU is pushing for the integration of green and digital technologies, such as batteries, heat pumps, renewables, and grid modernization, to lower energy costs and risks, to fortify the European strategic autonomy and rare earth supply chains, and to support the circular economy, enhancing monitoring and efficiency in reusing and recycling materials [58]. Nonetheless, the impact of various renewable (and other green) technologies can considerably vary with the level of sectoral economic complexity and innovation [59] and the time required for these technologies to be adopted by firms and affect the energy mix in the long run.

Specific types of renewables, such as renewable gases (e.g., biogas and biomethane), and advances in related technologies have not yet adequately been examined to substitute fossil fuels and reduce the level of risks for investors. Specifically, biomethane from waste biomass can help to reduce greenhouse gas emissions from hard-to-decarbonise sectors and the reliance of EU countries from natural gas imports, but it could become more cost-competitive than natural gas [60]. Among others, policy coordination among sectors, support of feedstock plants and their efficiency with reduction in production costs, together with appropriate supply driven market incentives with emphasis on end-uses could reinforce the energy mix in favour of such renewables [61].

Green financing and the establishment of energy partnerships for supplying clean fuels (e.g., green hydrogen) would help to stabilise the system and achieve a smooth and affordable green transition [62]. In this respect, EU countries can employ policy packages, such as REPowerEU, the Connecting Europe Facility (CEF), and their national Recovery and Resilience Plans to implement suitable investments and reforms. In turn, they should focus on strengthening the resilience of the gas supply system against unforeseen events, like those originating from geopolitical shocks and military conflicts affecting energy supply and prices.

Regarding possible limitations of the study and directions for future research, several technical characteristics of the gas supply system could be incorporated in the analysis, such as gas storage and transmission capacities of the cross-border pipelines and LNG port facilities, to examine resilience to external shocks and provide a more holistic evaluation of the potential strength of each country. Although extending the analysis period could possibly capture some further structural shifts in gas flows, the current analysis spans a sufficient duration before and during the conflict to plausibly capture the clear tendency for reduction in pipeline flows and increase in LNG flows. Specifically, the findings can be regarded as meaningful and reliable compared to previous studies which rely on simulations rather than actual gas flow data. They depict the state of the European gas supply network and its structural characteristics and changes during the period (2020–2023) in which the given study was performed. The use of a longer timeframe, incorporating data about more recent months, does not necessarily imply a more stable network state and, hence, more meaningful or reliable results, given the high and persistent uncertainty pertaining to the changing energy market and geopolitical landscape. For example, we stress Ukraine’s refusal to renew the Russian gas transit deal, the tensions among EU countries on the continuation of the Russian gas transit, and the changing stance of the United States on Russia–Ukraine conflict, especially after the end of 2024. Further research could also encounter other modes (road, rail, maritime) for the transport of gas as well as other energy resources. In this way, the European gas supply network could be investigated as part of a multiplex of energy transport systems across the globe, considering simultaneously both substitution and complementarity relationships among modes and commodities.