Smart-Grid Technologies and Climate Change: How to Use Smart Sensors and Data Processing to Enhance Grid Resilience in High-Impact High-Frequency Events

Abstract

Smart-grid technologies are essential to achieving sustainable high-level grid resilience. Integrating sensors and monitoring devices throughout grid infrastructure provides additional data on weather-related parameters in real-time, enabling the smart grid to respond appropriately to inclement weather and its associated challenges. The recording of all these data associated with each extreme weather event helps in the study and development of methodological tools for decision-making on issues of restoration and modification of the electricity network, with a view to enhancing its resilience and consequently ensuring the uninterrupted supply of electricity, even during the occurrence of these weather phenomena. This article focuses on enabling the utilization of meteorological data archives of past events, which demonstrate that natural disasters and extreme weather phenomena nowadays require network designs that can cope with the more frequent occurrence (high frequency) of events that have a significant impact (high impact) on the smooth operation of the network.

1. Introduction

In recent years, the infrastructure of power systems has been increasingly impacted by extreme weather events such as storms, heatwaves, snowfall, and strong winds. For example, the Hellenic Electricity Distribution Network Operator (HEDNO) reported 3587 faults in 2021 in its Medium Voltage Network, with 18.5% directly attributed to adverse weather conditions [1]. The rising frequency and severity of such events—driven largely by climate change—highlight the vulnerability of legacy power grids [2,3], which were designed under assumptions of climatic stability.

This new climate reality has exposed a critical challenge: modern power systems must not only be robust but also adaptive, particularly in serving the increasingly complex and high-demand needs of today’s societies. This has intensified the urgency to develop resilient grid systems capable of withstanding repeated and severe disruptions [4].

To address this challenge, we introduce the High Impact High Frequency (HIHF) model as a novel framework for analyzing and responding to the growing threat of recurring extreme weather events. Unlike traditional High Impact Low Frequency (HILF) models—which focus on rare but catastrophic disasters [5,6]—the HIHF model shifts the focus toward frequent and destructive climate-driven disruptions. This shift is critical for designing real-time, regionally adaptive strategies for power grid resilience.

So far, the main scope of all methodologies to enhance the power system’s resilience aim to minimize or even eliminate the impact of extreme weather events on the power grid [7]. Smart-grid technologies have emerged as a promising solution to enhance the resilience and adaptability of power systems [8]. Smart appliances and operations, when implemented into power systems, help in early detection of extreme events and control the network’s functionality. For example, smart meter installation can help in energy risk management in cases of unplanned outages, using the necessary information collected by these devices [7,9]. Additionally, it is possible to use data extracted from weather forecasts for cost-effective upgrading of power systems by using smart technologies that could indicate susceptible parts of the network [9].

By integrating smart sensors and advanced data processing techniques, operators of power systems can monitor environmental conditions in real time, predict system vulnerabilities, and restore grid destruction appropriately [10]. Additionally, these techniques can help in the modification of the grid so that it becomes more resilient in a way that an uninterrupted power supply is guaranteed. Unlike conventional grid management approaches that rely on historical probability models, smart grids leverage continuous data collection to assess the frequency and impact of extreme weather events [11]. This shift is crucial as climate patterns evolve unpredictably, making past statistical models less reliable for future planning.

Despite the potential of smart grids, research in this field remains fragmented with ongoing debates regarding the most effective methodologies for improving grid resilience [11]. Some studies emphasize hardware-based solutions, such as underground cabling and reinforced infrastructure [12], while others highlight the role of software-driven approaches, including predictive analytics and decentralized control systems [13]. At the same time, cyberattacks can degrade the operation of power systems as the new technologies implemented for monitoring and management are vulnerable [14]. Therefore, the design and operation of a resilient network has become a critical priority for utilities, policymakers, and researchers, targeting the reduction of the impact of these phenomena on the network while ensuring quality and uninterrupted service for electricity consumers.

Most of the time, it is proven that building a more redundant and robust network is often costly [15]. Therefore, in order to enhance the resilience of the network, developing “smart” operational measures is more cost-efficient. In other words, today, operators of power systems are more focused on implementing operational measures into the network, which include “smart” technologies and seem to be inexpensive, rather than applying hardening measures which reinforce the network’s infrastructure [9].

The expansion of distributed generation, the development of microgrids, and the evolution of network protection devices undoubtedly benefit from the existence of smart-grid technologies [16]. Also, strategic plans designed for timely responses, as well as risk management in cases of extreme weather conditions, are planned easier thanks to the capabilities of smart grids, which can provide the manager with real-time data on the status of the network [17].

Our proposed methodology to deal with this challenge aims to bridge the existing gaps by demonstrating how smart sensors and data-driven decision-making can enhance grid resilience against HIHF events, by integrating three key components:

1.

Historical meteorological data (collected from 2020 to 2024) to assess the frequency and severity of weather-related threats;

2.

Machine learning algorithms (specifically Gradient Boosting) to model and predict regional vulnerability patterns;

3.

Multi-criteria decision-making tools (notably ELECTRE) to prioritize resilience-enhancing actions based on local risk profiles.

This approach allows for the statistical characterization of each prefecture in Greece according to its exposure to specific weather phenomena. Based on this analysis, we propose region-specific operational measures to enhance power system resilience, tailored to the most likely and impactful threats in each area. The results offer a scalable, data-driven framework that can be applied beyond Greece to other regions facing similar challenges.

Especially in Greece, a country with a great number of islands, we must take into consideration that many of them are not electrically interconnected with the mainland. Therefore, they are particularly vulnerable to power outages due to extreme weather events. This comes out through various projects which aim to upgrade the technology of the power systems of the islands so that smart grids can contribute to the resilience and operation maintenance of energy systems when extreme weather events occur [18].

Smart-grid technologies—such as smart sensors, predictive analytics, and real-time monitoring—play a central role in this framework. These tools support early detection of environmental hazards, automated response strategies, and more cost-effective maintenance planning compared to traditional hardening measures. While infrastructure reinforcement (e.g., underground cabling) remains important, our findings support a growing trend toward smart, adaptive operational strategies that are both scalable and financially sustainable.

In this context, the present study proposes a novel, integrated methodology that leverages five years of historical meteorological data, advanced machine learning (Gradient Boosting), and the ELECTRE multi-criteria decision-making tool to assess the weather-related risk across Greek prefectures. Each region is statistically characterized based on the severity and frequency of weather events affecting it. Using these insights, region-specific operational resilience measures are identified to guide strategic planning and resource allocation. This approach demonstrates how the HIHF model, when combined with smart-grid technologies and data-driven prioritization, can provide a scalable and cost-effective framework for improving distribution network resilience under the mounting pressure of climate change.

2. Greek Electricity Distribution Network: Risks and Challenges

Electricity distribution networks are a vital part of the power system infrastructure. Their primary role is to ensure that electrical loads are supplied with sufficient and reliable power to meet the energy demands of consumers. In Greece, the distribution network begins at High Voltage/Medium Voltage (HV/MV) substations, where voltage is typically stepped down from 150 kV to 20 kV. It ends at the point of consumption.

The distribution network consists of the following:

1.

Medium Voltage (MV) networks, mainly operating at 20 kV (though 15 kV and 22 kV also exist in Greece), which transmit energy from HV/MV substations to MV/LV substations.

2.

Low Voltage (LV) networks, operating at 230 V (phase voltage) or 400 V (line voltage), which deliver power from MV/LV substations to end consumers.

Both MV and LV networks, comprising overhead lines, underground cables, and substations, are essential for ensuring that power reaches final users. Depending on the type of consumer, industrial or residential, loads are supplied either through the MV or LV network.

In Greece, the Hellenic Electricity Distribution Network Operator (HEDNO S.A.) manages the distribution network. HEDNO oversees the operation, maintenance, and development of the MV and LV systems, and is also responsible for HV networks in Attica and the islands. Additionally, HEDNO operates the non-interconnected island networks, which are not connected to the mainland transmission or distribution systems.

Due to Greece’s varied geography—including mountains, coastlines, and numerous islands—approximately 88% of the distribution network is overhead.

2.1. Climate Risks—Extreme Weather Events and Power Systems’ Sensitivity

The distribution network is susceptible to a variety of failures depending on prevailing weather conditions. Key risks include the following:

1.

Thunderstorms: Lightning, high winds, and flooding can damage lines, transformers and substations. Mitigation strategies include lightning arresters, grounding systems, guy wires, reinforced concrete poles, and burying lines in storm-prone areas [19]. Flood-related protection involves elevating substations, enhancing drainage, and using watertight enclosures.

2.

Hurricanes and typhoons: These events can cause transmission towers and poles to collapse, salt corrosion, and inundation of substations. Solutions include underground grids, concrete poles—which are more wind-resistant than wooden, waterproofed equipment, anti-corrosive coatings, elevation of the substations, removal of salt deposits, and preemptive shutdowns in high-risk zones [2].

3.

Tornadoes: High winds and debris pose severe risks. Mitigation and quick restoration require underground systems, tornado-resistant towers, automated power rerouting, mobile substations, and rapid-response teams [20,21].

4.

Blizzards and ice storms: Ice accumulation and heavy snow can cause electrical infrastructure failure and stress the grid due to high demand. Measures include ice-resistant cables, de-icing technologies, robust towers, demand–response strategies, enhanced backup generation, and proactive vegetation management [2].

5.

Heatwaves: High temperatures increase electricity demand and reduce transformer and line efficiency. Droughts can impair hydroelectric power. Solutions include enhanced cooling systems, distribution upgrades, renewable integration, and demand-side management through dynamic pricing and efficiency programs [20].

6.

Floods: Water damage to underground infrastructure and erosion of pole foundations are major risks. Protective strategies include waterproof materials, improved drainage systems, and elevating critical infrastructure in flood-prone regions [2].

7.

Wildfires: Fires destroy power lines and poles, while smoke and heat damage insulation. Fire-resistant materials and vegetation management zones help mitigate these effects. Preemptive shutdowns are sometimes necessary to prevent fire spread [21].

8.

Hailstorms: Large hailstones can damage exposed equipment. Using hail-resistant materials, such as protective coatings or shields, can reduce vulnerability [19].

9.

Geomagnetic storms: Caused by solar activity, these storms induce currents in long conductors, overloading transformers and causing blackouts. Countermeasures include geomagnetic shielding, improved grounding, and real-time monitoring to allow preemptive equipment shutdowns [22].

2.2. Extreme Weather Events in Greece (2020–2024)

In recent years, Greece has faced numerous extreme weather events that severely impacted the electricity distribution network. Some of the most notable events include the following:

• Windstorms, cyclones, and hurricanes

• Cyclone Daniel (September 2023): This severely impacted Larisa, Karditsa, Magnesia, Sporades, Corfu, and Central Athens. Intense weather conditions and floods caused extensive damage to the power grid. In Magnesia, 12 MV lines and around 750 substations were affected, disrupting power in Volos and nearby settlements. Karditsa experienced damage to about 25 substations, causing extensive power supply issues [23].
• Cyclone Ianos (September 2020): This caused extensive damage in Magnesia, Karditsa, and the Ionian islands. Over 2500 faults were reported in Central Greece. About 17 km of MV and LV networks were affected, with 144 poles and 10 overhead transformers damaged [23,24].

• Snowfall and ice storms

• Snowstorm Barbara, (February 2023): This affected West and East Attica with 30 MV line faults. Evia, Magnesia, Voeotia, and Larissa were also impacted. Over 800 technicians in Attica and 150 in Evia worked with specialized equipment to restore power [23].
• Thundersnow Elpis, 2022: Extreme snowfall caused extended widespread outages in North and Central Athens, East Attica, Evia, and Voeotia, and islands like Crete, Naxos, Rhodes, and Kos. Utility poles broke, and both LV and MV conductors were severed [24].
• Snowfall Medea, 2021: Trees fell on two of the three MV lines supplying the Sporades island, cutting power to Skiathos. Widespread MV and LV damage occurred in mountainous regions of Central Greece, Epirus, and Peloponnese, and mountainous parts of regions like Karditsa, Lamia, Volos, Trikala, Thiva, Amfissa, Ioannina, Arta etc., as well as the North Aegean islands. Restoration was hindered by road access difficulties [24].

• Heatwaves and wildfires

• Varibobi, North Evia, and the Peloponese Wildfires (July, 2021): A 10-day heatwave with temperatures up to 44 °C stressed the power grid, causing a series of faults. Wildfires destroyed poles, overhead LV and MV lines, and substations. Load shedding in Attica was implemented to stabilize the grid. Workforce expansion was a key mitigation effort [24].
• Heatwave Cleon, July 2023: This lasted 15 days and affected areas in North, Central, and South Athens, as well as West Attica. Outages occurred in Ilioupoli, Argyroupoli, Elliniko, Ano Liosia, Agioi Anargyroi, and Kamatero. Restoration efforts were extensive and time-consuming [19].
• Wildfire in Evros and Samothraki, August 2023: Fires in Alexandroupoli damaged two main MV lines supplying the underwater cables that interconnected Samothraki to Alexandroupoli. Power was restored using generators installed on the island of Samothraki [23].

2.3. Grid Resilience

Power system resilience is a critical issue for engineers and designers, particularly as extreme weather events become more frequent. Although there is no universally accepted definition, resilience generally refers to a power system’s ability to withstand, adapt to, and recover from disruptions such as natural disasters, cyberattacks, and operational failures. Enhancing resilience minimizes economic and social impacts while ensuring service continuity [17].

Traditionally, distribution networks have been designed with a focus on reliability—the ability to maintain power supply under normal conditions. However, increasing weather-related disruptions have highlighted the need to prioritize the resilience of distribution networks. What needs to be achieved is the mitigation of the impacts of such events and enabling fast recovery from them [25].

Key distinctions among related concepts about the properties of an electric power distribution network include the following:

• Resilience: The network’s ability to operate through multiple failures (N-k criterion) and recover quickly from extreme events.
• Reliability: The network’s capacity to handle one or two simultaneous faults (N-1 and N-2 criteria) and frequent, mild disturbances (Low Impact High Frequency—LIHF) [26,27,28].
• Robustness: The system’s ability to resist operational stress using existing infrastructure without collapsing. However, it does not ensure the system’s recovery after a collapse, which is a core feature of resilience. Although increasing robustness may contribute to enhanced resilience, complete equipment replacement is often economically and practically unfeasible, especially during crisis conditions [26].
• Security: Security, especially cybersecurity, is a top priority in the design of modern Electric Power Systems (EPSs), as digitalization and internet connectivity make them vulnerable to cyberattacks. Security strategies focus on preventing and deterring malicious attacks [25,29]. Unlike resilience, security focuses on prevention rather than recovery.

Resilience, therefore, is a distinct and critical property aimed at the cost-effective management of extreme events and the rapid restoration of system functionality.

Improving resilience involves integrating smart-grid technologies, including real-time monitoring, automated control systems, and distributed energy resources (DERs) [30]. Predictive analytics and artificial intelligence (AI) help identify vulnerabilities and optimize responses [31]. A hybrid approach that combines physical infrastructure improvements with cybersecurity measures provides comprehensive protection. Finally, regulatory support and stakeholder collaboration are essential in building and maintaining resilient electricity distribution networks [31,32].

3. Weather Events and Risk Analysis in Power Distribution Networks

There is extensive literature analyzing weather phenomena that significantly impact power distribution networks. These events may be classified as either High Impact Low Probability (HILP) or High Impact High Probability (HIHP) events [33,34,35]. Traditionally, the resilience of distribution networks is evaluated based on the probability of extreme weather occurrences. However, with climate change causing a sharp increase in the frequency of such events, probability alone is no longer a reliable predictor of future impacts on network infrastructure.

Consequently, many studies assess the effects of extreme weather on power systems using the HILF model [14,36,37,38]. HILF events refer to rare but devastating phenomena—such as intense rainfall leading to floods, strong winds, tornadoes, cyclones, or severe storms. These events are irregular and unpredictable but can inflict significant damage (High Impact) on critical infrastructure, including power systems.

A key objective of the present analysis is to highlight the inadequacy of the HILF model in the face of recent trends, particularly in Greece. By analyzing weather data from 2020 to 2024, this study proposes a shift toward a new model: the High Impact High Frequency (HIHF) framework. This updated model reflects the increasing recurrence of extreme events much better and emphasizes the urgency of redefining resilience parameters in electrical distribution network research.

Studying HILF events remains vital, as it helps engineers design more resilient infrastructure, identify vulnerable regions, and develop disaster response strategies [37]. Climate change plays a central role in the emergence of these phenomena. Rising global temperatures, increased atmospheric moisture, and prolonged heatwaves have intensified the severity and frequency of weather-related disasters [39].

3.1. Weather Risk Analysis Based on HILF Events

Weather risk analysis is a systematic process that enables informed predictions about extreme weather threats in a region. An algorithmic methodology consisting of five key stages follows:

1.

Hazard assessment: This stage estimates the likelihood of HILF event occurrence, using historical data, meteorological records, General Circulation Models (GCMs) [40], and climate change scenarios.

2.

Vulnerability assessment: This evaluates how susceptible populations, infrastructure, and ecosystems are to adverse effects. It considers the structural resilience of buildings, power grids, drainage systems, and more. Socioeconomic and demographic factors—such as age, income, or health—are also assessed (e.g., elderly individuals are more vulnerable to heatwaves).

3.

Impact assessment: This involves assessing the severity of an event’s effects on humans, infrastructure, and the environment. It includes analysis of economic damage (e.g., property loss and insurance claims), public health consequences, and environmental degradation (e.g., soil pollution and biodiversity loss).

4.

Risk quantification: Risk is calculated as a function of likelihood (frequency of occurrence) and severity (impacts). Quantitative models use probabilistic tools and statistical data to produce numerical results and risk maps, identifying areas with high exposure to extreme events.

5.

Resilience and adaptation analysis: This final step evaluates a region’s ability to withstand and recover from disasters. It considers adaptation strategies such as improved infrastructure, early warning systems, afforestation, and sustainable water management.

3.2. Tools and Strategies for Efficient Risk Management

When weather risk analysis forecasts the likelihood of an extreme event, implementing effective risk management strategies is essential to mitigate potential damage to vital infrastructure such as power systems. Key tools and approaches include the following:

• Meteorological and climate models: General Circulation Models (GCMs) are used to study climate change on a global scale. They simulate large-scale atmospheric and oceanic processes, predicting changes in temperature, humidity, and storm patterns [41]. Regional Climate Models (RCMs) offer more localized insights [40], improving risk assessments. Specialized weather simulation tools also support disaster preparedness.
• Geographic Information Systems (GISs): GIS tools generate visual maps identifying regions frequently affected by specific extreme events, enhancing spatial analysis of vulnerabilities.
• Statistical models: Techniques like Monte Carlo simulations and Extreme Value Analysis (EVA) [42] estimate the probability and intensity of rare weather events. The EVA method creates probability curves from historical data, while Monte Carlo method simulations model a variety of risk scenarios.
• Satellites and radars: These tools facilitate real-time monitoring of weather conditions and detection of affected areas, enhancing situational awareness and response capabilities.

Risk management strategies that are applied to reduce the impact of events that fit the HILF model include early warning systems, which consist of meteorological stations networks, satellites, and sensors to detect events in early stages and provide timely alerts to authorities and citizens (e.g., flooding warnings). Such systems can be installed locally near rivers, for instance, to monitor rising water levels. Additionally, structural adaptation measures, such as infrastructure strengthening, enhance the resilience of buildings, bridges, power systems, and other critical facilities, allowing them to withstand extreme weather. On the other hand, non-structural adaptation measures involve designating risk zones (e.g., flood-prone areas), restoring natural ecosystems like wetlands and forests, and raising public awareness to foster community preparedness. Lastly, resilience planning involves developing recovery plans to restore infrastructure and support populations post-disaster, based on assessments of the societal and environmental recovery capacity.

3.3. Case Study: Weather Risk Analysis in Greece

Recent flood events in specific parts of Greece, like Mandra (2017) and Thessaly (phenomenon named Daniel, 2023), reveal the growing vulnerability of Greek regions to extreme weather. Factors influencing these events are climate change, the geographical location of the affected region, and the geomorphological conditions. Additionally, it should not be omitted that urban expansion and infrastructure deficiencies, such as uncontrolled city growth, destruction of natural watersheds, and inadequate drainage systems, amplify flood risk.

Applying the weather risk analysis model to these events yields the following insights:

1.

Hazard assessment: The occurrence of this phenomenon can be attributed to a combination of natural and anthropogenic factors. From a natural standpoint, climate change has resulted in rising global temperatures and elevated atmospheric humidity, both of which contribute to the increased intensity and frequency of heavy rainfall events. Greece’s geographical location within the Mediterranean basin further heightens its exposure to sudden and severe precipitation, often leading to flash floods. The country’s complex geomorphology—characterized by mountainous terrain—facilitates the rapid runoff of rainwater into low-lying areas, exacerbating flood risks. Anthropogenic factors also play a significant role. Unregulated urban expansion and the degradation of natural watersheds have diminished the soil’s capacity to absorb excess water. Moreover, existing drainage infrastructure is frequently insufficient to accommodate the volume of water generated during extreme weather events, thereby increasing the likelihood and severity of flooding.

2.

Vulnerability assessment: In several regions of Greece—particularly those located near rivers and in low-lying areas—flood protection infrastructure is either insufficient or deteriorating. This includes aging dams and underdeveloped or poorly maintained drainage systems. A prominent example is the region of Thessaly, where inadequate management of the Pinios River has significantly exacerbated flooding events. Moreover, socioeconomically disadvantaged or geographically isolated populations are disproportionately vulnerable, as they often lack access to early warning systems and the financial or logistical means necessary for effective disaster response and recovery [43]. Environmental degradation further intensifies vulnerability; widespread deforestation, frequently resulting from wildfires, reduces the soil’s water retention capacity, while the loss of wetlands eliminates crucial natural flood buffers. Collectively, these structural, social, and ecological factors contribute to the heightened vulnerability of the affected regions.

3.

Impact assessment: The 2023 floods in Thessaly resulted in extensive damage, with economic losses estimated in billions of euros. This includes the widespread destruction of agricultural land, residential buildings, and critical infrastructure [43]. These losses have enduring economic repercussions, particularly in rural areas that rely heavily on agriculture and local production. In addition to economic damage, the social impact has been significant: entire communities were displaced, and affected individuals endured both physical harm and psychological distress, with some incidents resulting in loss of life. The environmental consequences were also severe, as floodwaters carried debris, sewage, and industrial chemicals, leading to the contamination of rivers and soils. These pollutants contributed to the degradation of ecosystems and the loss of biodiversity, underscoring the multifaceted and far-reaching effects of such extreme weather events.

4.

Risk quantification: Building upon the preceding assessments, the calculated risk level for disruptions to the electricity distribution network in both the flood events mentioned above (Mandra (2017) and Thessaly (phenomenon named Daniel, 2023)) is found to be significantly high.

5.

Resilience and adaptation analysis: The response to these events revealed a lack of adequate pre-disaster planning and adaptation measures. Specifically, emergency action plans and resilience-enhancing interventions were not effectively implemented prior to the flooding events. This suggests a low level of preparedness and adaptive capacity within the electricity distribution infrastructure in the affected regions. Strengthening the resilience of the distribution network requires the adoption of both structural and non-structural strategies.

Key tools for risk mitigation include GISs, which facilitate the development of detailed flood hazard maps to identify high-risk zones. The application of EVA is also critical, enabling probabilistic estimation of flood events based on historical climatic data [42]. Furthermore, the use of real-time satellite monitoring supports dynamic assessment of flood extent, while specialized early warning systems enhance the ability to detect and respond to emerging threats. The installation of meteorological radars to monitor storm development in real time is recommended as an essential measure to further bolster adaptive capacity and minimize potential impacts.

4. The HIHF (High Impact High Frequency) Model

The HIHF (High Impact High Frequency) model is a comprehensive framework designed to analyze, predict, and manage weather-related events that occur with considerable frequency and carry severe consequences. These events—such as recurring storms, floods, heatwaves, and snowfalls—pose significant risks not only to critical infrastructure (particularly electricity distribution networks), but they also have significant social, economic, and environmental impacts.

What distinguishes the HIHF model from other risk assessment frameworks is its focus on events that combine high frequency with high intensity, as opposed to low-probability, high-impact events. The model integrates data from satellite systems, ground-based meteorological stations, and simulation platforms to improve forecasting precision. It employs a multidisciplinary approach, analyzing meteorological data alongside socioeconomic and environmental impact metrics. For example, it assesses how intense rainfall may disrupt agriculture or urban infrastructure.

The HIHF framework supports early warning systems (EWSs), facilitating prompt and informed responses by local authorities and utilities. It also enhances the operational resilience of infrastructure systems—especially energy, transport, and water networks—by identifying vulnerabilities and informing risk mitigation strategies. Ultimately, it plays a vital role in safeguarding human life and property through the development of proactive and responsive mechanisms.

4.1. Application Advantages and Challenges of the HIHF Model

The HIHF model presents several strategic advantages. One of the key advantages is its enhanced forecasting accuracy, by incorporating vast datasets and advanced modeling techniques [44]. Additionally, the model can be tailored to various topographic and climatic conditions, ensuring broad applicability. Its predictive capacity supports preventive measures that reduce service disruptions and reinforce grid reliability, and hence the resilience of the critical infrastructure.

However, implementation of the HIHF model is not without challenges. Deployment requires advanced technologies such as IoT-enabled sensors and automated monitoring systems. In many regions, including parts of Greece, IoT integration remains limited. High investment costs is another major obstacle, as infrastructure upgrades (e.g., underground cabling or installing weather-resilient materials) involve substantial financial commitments.

Moreover, climate change is increasing the unpredictability and severity of extreme events, complicating the model’s effectiveness. Finally, the absence of a unified operational framework limits the systematic application of the HIHF approach, although there are various fragmented practices that align with its principles.

4.2. HIHF Model Applications in Electricity Distribution Networks

The HIHF model plays a crucial role in enhancing the resilience of power distribution networks by identifying vulnerable network segments and forecasting high-impact weather events. By predicting periods of extreme heat or increased demand, the model helps detect areas at risk of cable overheating or overload. Similarly, it supports risk mitigation strategies against wind-, snow-, or flood-related outages (e.g., pruning, to avoid the falling of trees and branches on the lines) by analyzing environmental and network data.

HIHF’s primary objective is to aid short- and medium-term operational planning, enabling targeted preventive maintenance, microgrid formation, and infrastructure reinforcement in response to severe weather conditions. The model also underpins the development of early warning systems and backup measures, such as energy storage deployment, generator activation, and load rerouting.

Its integration with IoT sensors and artificial intelligence allows real-time grid monitoring, making smart grids a critical enabler of resilience and automation. Moreover, the use of big data analytics within the HIHF framework supports the identification of risk patterns and accelerates informed decision-making.

4.3. Comparison and Distinctions Between HILF and HIHF Events

The primary distinction between the HIHF and HILF models lies in the nature and recurrence of the events they address, as well as the corresponding risk management strategies employed. The HILF model focuses on rare but potentially catastrophic events, whereas the HIHF model targets frequent events that, while individually less extreme, can cause substantial cumulative impacts over time.

HILF events include major natural disasters such as earthquakes, tsunamis, hurricanes, and solar storms, as well as anthropogenic threats like large-scale cyberattacks or terrorist attacks targeting energy infrastructure. In contrast, HIHF events encompass recurrent phenomena such as severe storms, high winds, floods, prolonged heatwaves, and heavy snowfalls. While many of these were initially categorized under the HILF model, the increasing frequency of such events has necessitated the development of a distinct HIHF framework.

The HIHF model is primarily oriented toward the mitigation and management of frequent, high-impact events through strategies such as preventive maintenance, continuous monitoring, and rapid recovery. Its goal is to enhance operational resilience and minimize service disruptions. Conversely, the HILF model emphasizes long-term preparedness and infrastructure hardening to withstand and recover from rare but severe crisis events. These strategies typically involve significant investment and are tailored to mitigate large-scale systemic failures.

Table 1. Comparison table of the HILF and HIHF models.

Characteristic	HILF	HIHF
Main difference	Low frequency of extreme weather events (they rarely occur)	High frequency of extreme weather events (they often occur)
Impact size	Moderate to severe effects	Disastrous impact
Management	Long-term planning, resilient infrastructure	Preventive maintenance, regular monitoring, short-term plans
Strategies	Management of large-scale damages	Optimizing resilience on a daily basis
Planning	Long-term	Short-term and medium-term

provides a comparative overview of the key characteristics of both models.

In conclusion, the HIHF model is chiefly concerned with maintaining the reliability and resilience of systems in the face of recurring stressors, which can lead to the gradual degradation of infrastructure. The HILF model, by contrast, is designed to safeguard against low-probability, high-consequence events that demand robust contingency planning. Importantly, these two models are not mutually exclusive but rather complementary. An effective and comprehensive resilience strategy—particularly for critical infrastructure such as power distribution networks—should integrate both HIHF and HILF approaches to address the full spectrum of risks, from routine disruptions to extreme crisis scenarios.

4.4. Proposed Algorithm for Weather Risk Analysis Based on the HIHF Model

In Greece, the application of the HIHF model in the management of critical infrastructure—particularly power distribution networks—has not yet been widely adopted as a standardized and integrated practice. Nevertheless, various initiatives and strategies consistent with the principles of the HIHF model have begun to emerge. These developments are increasingly relevant as the frequency and severity of extreme weather events escalate, necessitating robust and data-driven approaches to infrastructure resilience. Several of these approaches incorporate data processing, machine learning, and optimization techniques.

This paper proposes an algorithm grounded in the HIHF model to assess weather-related risks and support decision-making aimed at enhancing the resilience of power distribution networks. The algorithm comprises the following key components:

1.

Data collection and processing: Effective implementation of the HIHF model requires the collection and processing of diverse data sources. These include historical weather data (e.g., precipitation, wind speed, and temperature), records of power outages and their causes (e.g., vegetation-related faults and overloading), and geospatial data detailing the locations of poles, substations, and transmission lines. Accurate network load data is also essential. This information constitutes the input for the proposed algorithm and must be appropriately pre-processed. In addition, data acquired from Internet of Things (IoT) sensors—measuring variables such as temperature, voltage, and humidity—can facilitate real-time monitoring. In high-risk regions, sensor deployment enables effective real-time analysis and rapid detection of anomalies.

2.

Application of machine learning models: Machine learning techniques are employed to estimate the probability of damage to specific network components based on weather conditions and equipment status. Predictive models such as Random Forests and Gradient Boosting are suggested for classification and regression tasks. For time-series forecasting of network load and the frequency of weather phenomena, Recurrent Neural Networks (RNNs) or Long Short-Term Memory (LSTM) networks are appropriate. Additionally, unsupervised learning techniques—such as Autoencoders or Isolation Forests—can be used to detect operational anomalies indicative of potential faults or overloads [45,46,47].

3.

Optimization models for maintenance planning: To ensure the optimal allocation of maintenance resources and personnel, optimization methods such as Linear Programming (LP) are proposed. These can support efficient scheduling and resource distribution. Moreover, heuristic algorithms such as Genetic Algorithms and Simulated Annealing are suitable for identifying critical maintenance points based on cost-effectiveness, event frequency, and damage severity. Maintenance strategies should be informed by comprehensive risk forecasting frameworks [48].

4.

Real-time decision making: Given the inherent unpredictability of weather conditions, dynamic and adaptive decision-making mechanisms are essential. Capabilities such as fault isolation, power flow reconfiguration, and service restoration must be supported in real time to ensure the uninterrupted operation of the network. Intelligent Decision Support Systems (DSSs) offer significant potential in this regard, enabling rapid responses to evolving conditions [6,49,50].

5.

Identification of preventive and corrective actions: In the final stage of the algorithm, actionable measures are proposed to mitigate the impact of extreme weather events. These include early warnings for severe weather, preventive isolation of vulnerable network segments, automatic activation of energy reclosers, and timely notifications to maintenance crews. Furthermore, long-term resilience can be enhanced through infrastructure upgrades, cable undergrounding, and equipment reinforcement in vulnerable zones, as identified through GIS analysis.

4.5. Application of the HIHF Model in Greece

The HIHF model offers a robust framework for addressing extreme weather events in the context of Greece’s electricity distribution system. For instance, during the “Medea” weather system in 2021, widespread damage was reported in overhead distribution lines, caused by falling trees and the collapse of lines and poles due to heavy snowfall. In response, HEDNO strengthened its response teams and modernized its network monitoring systems, actions that align with HIHF principles.

Similarly, during the summer of 2021, heatwaves caused high electricity demand and significant thermal stress on the grid. Concurrent wildfires severely damaged large sections of the network. Preventive measures such as replacing overhead lines with underground ones and substituting wooden poles with concrete alternatives, along with proactive equipment maintenance and enhanced load monitoring, all reflect the core practices of the HIHF framework.

In flood-prone regions such as Western Attica and Thessaly—both severely affected by storm “Daniel” in 2023—substantial protective measures were undertaken. These included constructing embankments around substations and installing pumping systems to prevent flooding, consistent with HIHF’s focus on hazard mitigation in vulnerable areas.

The systematic application of the HIHF model in Greece faces limitations due to infrastructure constraints. The integration of IoT technologies in the Greek distribution network still remains at an early stage. Moreover, costs associated with converting overhead lines to underground cables and implementing protective systems are extremely high. Forecasting and preparedness are also challenged by the growing intensity and frequency of extreme events, particularly in areas not previously considered vulnerable. Lastly, although scattered practices are consistent with HIHF philosophy, there is still no cohesive and institutionalized risk management strategy based on this model.

Nevertheless, there is significant potential for HIHF model implementation in Greece. Expanding the deployment of smart meters and IoT systems could enhance data collection and forecasting capabilities. Investments in resilient infrastructure, particularly in densely populated or high-risk zones, would yield long-term benefits. Additionally, developing maintenance programs based on risk data informed by risk analysis from frequent weather phenomena could serve both HIHF and HILF strategies.

Collaboration with academic and research institutions is highly recommended. Integration of advanced technologies such as artificial intelligence (AI) and GIS can further strengthen the decision-making processes and operational efficiency of the electricity grid. While Greece has not yet adopted the HIHF model comprehensively, the repeated occurrence of extreme weather events underscores the urgency of establishing a formalized framework grounded in its principles.

4.6. Gradient Boosting for Risk Assessment

Gradient Boosting is a widely adopted machine learning algorithm used for both classification and regression tasks. In the context of electric power networks, it can be effectively employed to predict the probability of damage to specific components caused by extreme weather events. As such, it may be integrated into the HIHF model to enhance the overall resilience and operational reliability of the electric power distribution system.

The algorithm leverages multiple data sources as input, including historical weather records, structural attributes of the power network, records of past failures, geographical features, and energy consumption patterns during peak demand periods. Through this integration, the model can identify network segments with elevated failure risk. Based on the algorithm’s outputs, network operators can implement targeted preventive and corrective measures to mitigate potential disruptions.

The recommended model architecture is based on Gradient Boosted Decision Trees (GBDTs), with widely used implementations including XGBoost, LightGBM, and CatBoost. Critical hyperparameters to be optimized during model training include the learning rate (which governs the incremental improvement of the model), number of trees (determined by the size and complexity of the dataset), maximum tree depth, and subsampling rates (to prevent overfitting). Additionally, L1/L2 regularization techniques are applied to limit over-adaptation to the training data [51,52].

Furthermore, real-time application of the Gradient Boosting algorithm is feasible through the deployment of IoT sensors across the network. These sensors enable the continuous collection of environmental and operational data, facilitating immediate detection of high-risk segments. Coupled with GIS, the outputs can be visualized spatially, allowing operators to monitor vulnerable areas and respond proactively.

4.7. Application of Gradient Boosting Algorithm for HEDNO

To implement a Gradient Boosting-based model for enhancing the resilience of the Greek electricity distribution network, a structured data processing pipeline is required.

For training and deployment of the algorithm, the following data categories are essential:

a.

Historical failure data: includes failure types, location, timing, causes, restoration time, and maintenance actions taken.

b.

Meteorological data: sourced from meteorological stations or in situ sensors, and include wind speeds, temperature extremes, rainfall intensity (mm/h or mm/24 h), lightning density, and frost characteristics.

c.

Network infrastructure data: such as location and specifications of distribution lines (i.e., MV/LV), location of substations, technical characteristics of conductors and poles (e.g., material, diameter, and length), and installation dates.

d.

Geospatial data: encompassing terrain features, land use classification (urban, forest, rural), and proximity to natural hazards such as rivers or dense vegetation.

e.

Real-time operational data: gathered via installed IoT sensors to monitor line temperature, voltage, current, load, and short-term weather forecasts.

The algorithm processes these datasets, by initially cleaning and merging them, then training a predictive model to classify network segments by failure probability.

The output can be visualized as probability maps of failure-prone areas, risk rankings of network segments, or recommended preventive actions (such as strengthening transmission lines with wind-resistant conductors, preemptively isolating vulnerable segments near forested areas before storms, and performing targeted maintenance on aging infrastructure).

During severe weather events, the system can issue early warnings to maintenance teams and recommend alternative routing strategies to maintain service continuity and minimize outage durations.

5. Description and Objective of the Study

The primary objective of this study is to leverage post-event meteorological data to enhance the resilience of electrical power distribution systems in Greece. Historical evidence suggests that natural disasters and extreme weather events are occurring with increasing frequency and severity, necessitating proactive network planning. These events, characterized by both their high frequency and high impact (HIHF), pose substantial risks to the stability and functionality of power infrastructure.

To address these challenges, the HIHF model has been developed to classify weather events according to their impact on the electrical network and suggest appropriate mitigation measures. The present analysis is designed to serve as a decision-support tool, facilitating the planning and prioritization of reinforcement projects across various Greek regions.

Specifically, the study aims to do the following:

1.

Classify the need for resilience restoration into High, Medium, or Low categories, based on the combination of event impact and frequency—categorized as HIHF, HILF, LIHF (Low Impact High Frequency), or LILF (Low Impact Low Frequency). Risk levels are calculated based on the type, duration, intensity, and consequences of each event.

2.

Select suitable categories of reinforcement projects for both overhead and underground networks.

While this analysis relies on historical data, its conclusions underscore the importance of real-time data acquisition through smart-grid technologies. Such data enhances the precision and responsiveness of resilience strategies by supporting targeted, evidence-based interventions.

5.1. Results of the Weather Risk Analysis Using MATLAB

Methodological tools can significantly aid in timely and informed decision-making regarding the modification and restoration power systems need in response to extreme weather events. This study developed such a tool using MATLAB (version: R2024b) to analyze historical weather data, identifying the areas of the distribution network most frequently impacted by severe conditions. This allows for strategic planning to bolster system resilience.

5.1.1. Data Collection (2020–2024)

Data regarding extreme weather events and their impacts on the Greek power distribution network between 2020 and 2024 were gathered from several sources: the National Observatory of Athens (www.meteo.gr (accessed on 26 of February 2025)), HEDNO press releases, and publicly accessible meteorological databases. The collected dataset includes the following:

• Event occurrence dates.
• Affected regional units (by each extreme weather event).
• Named events (where applicable).
• Event duration (in days).
• Intensity classification (Very strong–Strong–Medium).
• Impact severity (Limited–Several–Extended).
• Event type (e.g., tornado, rainfall, storm, wind, snow, lightning, hail, heatwave, drought, or frost).
• Description of effects on the power distribution network.

These data are organized in

Table A1. Extreme weather events in Greece (2020).

Date	Regions Affected	Intensity	Implications	Type of Weather Event	Network Influence
6 January	West and East Attica, North Athens, Veotia, Lesbos, Attica Islands, Andros, Tinos, Chania	Very strong	Extended	Snow, frost, wind, rainfall	Damages to overhead Medium Voltage lines
15 February	Rhodes	Very strong	Several	Storm	Damages to overhead Medium Voltage lines, power outage, network damages
2 April	Imathia, Chalkidiki, Serres, Xanthi	Very strong	Extended	Snow, wind	Network damages
6 April	Xanthi, Kavala, Evia	Very strong	Extended	Wind, rainfall	Network damages
22 May	Thessaloniki	Strong	Several	Storm, lightning	Power outage
28 June	Larissa	Strong	Limited	Rainfall	Tree falling on overhead lines, damages to overhead Medium Voltage lines
5 July	Larissa	Medium	Limited	Rainfall, wind	Power outage
6 August	Pieria	Very strong	Extended	Rainfall, wind	Power outage
9 August	Evia	Very strong	Extended	Rainfall	Power outage, dropped poles
20 September	Kefalonia, Ithaca, Zakynthos, Magnesia, Karditsa	Very strong	Extended	Storm, wind	Network damages, substation damages, power outage
23 September	Corfu	Very strong	Several	Storm, wind, hail	Tree falling on overhead lines
24 September	Elis	Very strong	Several	Tornado, storm	Tree falling on overhead lines, dropped poles
29 September	Lesbos	Strong	Several	Wind, rainfall, hail	Network damages
13 October	North and South Athens	Very strong	Several	Rainfall, wind	Power outage
21 October	Heraklion	Very strong	Extended	Lightning	Network damages
22 October	Chania	Very strong	Extended	Rainfall	Power outage
28 October	Chania	Very strong	Limited	Rainfall, wind	Power outage
7 December	Thessaloniki	Very strong	Several	Rainfall, wind	Power outage

,

Table A2. Extreme weather events in Greece (2021).

Date	Regions Affected	Intensity	Implications	Type of Weather Event	Network Influence
2 February	Evros	Very strong	Extended	Storm	Power outage
8 February	Serres	Very strong	Limited	Tornado, rainfall	Power outage, cable damages
13 February	Trikala, Karditsa, Evia, Voeotia, Sporades, Magnesia, North, Central and West Athens, East Attica, Fokida, Ioannina, Arta, Fthiotida	Very strong	Extended	Snow, frost	Tree falling on poles, power outage, pole damages
5 April	Ikaria	Strong	Limited	Wind	Power outage, tree branch falling on poles
21 May	Thessaloniki	Very strong	Several	Wind	Tree falling on poles, power outage
21 June	Karditsa	Strong	Limited	Hail, rainfall	Tree falling on poles, power outage
18 July	Karditsa	Medium	Limited	Hail, rainfall	Power outage, dropped poles
1 August	Rhodes	Strong	Extended	Heatwave, drought	Damages to overhead Medium Voltage lines, power outage
3 August	Kos, East Attica, Arcadia, Laconia, Aetoloacarnania, Messenia, Elis, Fokida, Evia	Very strong	Extended	Heatwave	Power outage, damages to overhead Medium Voltage lines, pole damages, substation damages
8 October	Corfu	Very strong	Extended	Wind, rainfall	Power outage
14 October	Rethymno	Very strong	Extended	Rainfall	Power outage
26 November	Corfu	Very strong	Extended	Storm	Tree falling on poles, dropped poles
28 November	Elis, Messenia	Very strong	Extended	Tornado, rainfall	Network damages, tree falling on poles, power outage
3 December	Aetoloacarnania	Strong	Several	Storm	Power outage
12 December	Serres, South Athens, East Attica	Very strong	Extended	Wind	Power outage, tree branch falling on poles, tree falling on poles, dropped poles

,

Table A3. Extreme weather events in Greece (2022).

Date	Regions Affected	Intensity	Implications	Type of Weather Event	Network Influence
11 January	Magnesia	Very strong	Several	Wind, frost	Power outage
12 January	Chalkidiki	Very strong	Several	Wind, frost	Tree falling on poles, power outage
24 January	Kos	Very strong	Extended	Snow, frost	Power outage, line faults, network damages
25 January	East Attica, North and Central Athens, Naxos, Veotia, Evia, Rhodes	Very strong	Extended	Snow, frost	Power outage
26 January	Chania	Very strong	Several	Snow, frost	Power outage
2 April	Chania	Strong	Several	Wind	Power outage
19 May	East Attica, Central Athens	Very strong	Several	Wind	Power outage
10 June	Aetoloacarnania, Achaea	Very strong	Extended	Wind, storm, hail	Power outage, substation damages
11 June	Kavala	Very strong	Extended	Rainfall	Power outage, streetlight damage
18 June	Larissa	Strong	Limited	Tornado, hail, rainfall	Power outage, tree falling on poles
26 June	Serres	Strong	Limited	Wind, rainfall	Tree falling on poles, power outage, dropped poles
9 July	Rethymno	Very strong	Extended	Storm, wind	Power outage
20 July	West Attica, North Athens	Very strong	Extended	Wind, drought, heatwave	Power outage, substation damages, line faults, network damages
21 August	Kozani	Very strong	Extended	Storm	Power outage
22 August	Chalkidiki	Very strong	Extended	Storm	Power outage
23 August	Trikala	Very strong	Extended	Storm	Power outage
5 September	Thessaloniki	Very strong	Several	Storm	Power outage
29 November	Rhodes	Very strong	Limited	Rainfall	Power outage
1 December	Thessaloniki	Strong	Several	Rainfall, wind	Power outage
11 December	Thesprotia, Corfu	Very strong	Extended	Rainfall, wind	Power outage
12 December	Lesvos	Very strong	Extended	Rainfall	Power outage, tree falling on poles

,

Table A4. Extreme weather events in Greece (2023).

Date	Regions Affected	Intensity	Implications	Type of Weather Event	Network Influence
20 January	Thesprotia	Very strong	Several	Tornado	Power outage, dropped poles
26 January	Zakynthos	Very strong	Extended	Rainfall, wind	Power outage, dropped poles
5 February	Evia, Magnesia, Veotia, Fthiotida, West and East Attica, Larissa	Strong	Extended	Snow, frost	Power outage, damages to overhead Medium Voltage lines
24 June	Kozani	Strong	Limited	Rainfall, wind, lightning	Tree falling on poles, power outage, Medium Voltage lines cut, fire on pole from lightning
25 June	Achaea	Strong	Several	Hail, rainfall	Power outage
14 July	North, Central and South Athens, West Attica	Very strong	Extended	Heatwave	Power outage
22 August	Evros	Very strong	Extended	Heatwave, drought, wind	Power outage, damages to overhead Medium Voltage lines
5 September	Magnesia, Larissa, Corfu, Central Athens, Sporades, Karditsa	Very strong	Extended	Rainfall, wind, tornado, lightning	Power outage, damages to overhead Medium Voltage lines, voltage drop, substation damages
28 September	Magnesia	Very strong	Extended	Rainfall	Power outage
4 November	Trikala, Karditsa, Chalkidiki	Very strong	Extended	Rainfall, tornado, wind, hail	Power outage, fire on poles, dropped poles
10 November	Arkadia	Strong	Limited	Wind	Tree falling on poles, power outage, Medium Voltage lines cut, fire on pole from lightning
22 November	Rhodes	Very strong	Several	Rainfall	Power outage
25 November	Chania	Very strong	Several	Rainfall	Power outage
17 December	East Attica	Very strong	Limited	Tornado	Damages to overhead Medium Voltage lines

and

Table A5. Extreme weather events in Greece (2024).

Date	Regions Affected	Intensity	Implications	Type of Weather Event	Network Influence
7 January	Elis, North Athens	Very strong	Extended	Tornado, rainfall	Power outage, tree falling on poles
12 February	Limnos	Very strong	Extended	Hail, rainfall	Tree falling on poles
5 March	Thessaloniki	Medium	Several	Hail, rainfall	Power outage
23 April	Sporades	Very strong	Several	Wind, storm	Pole damages
14 June	Imathia	Strong	Limited	Rainfall	Power outage, tree falling on poles
4 July	Evia	Very strong	Extended	Storm, lightning	Power outage
20 July	Piraeus, West central and South Athens	Very strong	Extended	Heatwave	Power outage, damages to underground Medium Voltage lines
11 August	North and East Athens	Very strong	Extended	Drought, wind	Power outage, damages to overhead Medium Voltage lines
21 August	Larissa, Magnesia	Strong	Several	Wind, rainfall	Power outage, dropped power cables, tree falling on poles
11 September	Chios	Very strong	Several	Storm	Streetlight damage, network damages
5 October	Corfu	Strong	Extended	Storm, lightning	Power outage
15 November	Aetoloacarnania	Strong	Several	Storm	Power outage
23 November	Trikala, Preveza	Very strong	Several	Storm, wind	Power outage, tree falling on poles
1 December	Kozani, Pieria, Chalkidiki, Thessaloniki	Very strong	Extended	Rainfall, storm, wind	Power outage, network damages
27 December	Chania	Very strong	Several	Lightning	Network damages

in Appendix A and serve as inputs for further analysis using MATLAB. Specifically, this part of the study aims to classify the Greek regions based on the frequency of extreme weather events during the period 2020 to 2024 that had a severe impact on infrastructure and operation of the power distribution system.

5.1.2. Analysis and Results

Figure 1 presents a color-coded map of Greece, illustrating the frequency with which extreme weather events have impacted the regional units of the national power distribution network. The classification is based on the HIHF–HILF–LIHF–LILF framework. Specifically, red and orange areas correspond to regions affected by HIHF events. Yellow denotes regions impacted by HILF events, while green represents areas experiencing LIHF events. Blue regions indicate LILF events, suggesting minimal or infrequent weather-related disruptions.

The results of this spatial analysis highlight specific regions—East Attica, the North Sector of Athens, Evia, Magnesia, Thessaloniki, Corfu, and Heraklion—as the most frequently affected by extreme weather conditions. These findings underscore the necessity of implementing enhanced resilience measures in these areas to ensure the reliable operation of the power distribution network.

To inform the development of targeted mitigation strategies, a risk assessment was conducted to evaluate the vulnerability of the distribution network to different types of extreme weather events. This assessment considers three key parameters: event intensity (weighted at 40%), event duration (30%), and the extent of impact on the power infrastructure (30%). The results of this analysis can be used to support strategic planning and operational decision-making when extreme weather events are forecasted.

Figure 2 presents the results of a weather-related risk analysis, illustrating the percentage risk associated with each type of extreme weather event during the 2020–2024 period. For example, 25.86% of snow events have resulted in severe impacts on the power distribution network. These risk percentages provide critical insight for system operators, enabling them to tailor their preparedness and response strategies based on the nature of the impending event. Notably, the heightened risk level associated with heatwaves indicates the need for more robust preventive measures and contingency planning relative to other event types.

Figure 3 displays the average risk level to the power system by region across Greece, calculated based on the occurrence of extreme weather events during the 2020–2024 timeframe. This regional risk assessment incorporates key parameters, including event duration, intensity, frequency, and the severity of associated adverse effects. The outcomes of this analysis identify regions that require strengthened resilience and targeted interventions to enhance the robustness and operational reliability of the power distribution infrastructure in the face of future extreme weather events.

5.2. Suggested Actions Using the ELECTRE Method

Following the risk assessment of potential distribution network disruptions due to extreme weather events, it is imperative to identify appropriate strategic interventions to mitigate these impacts. The ELECTRE (ELimination Et Choix Traduisant la REalité) multi-criteria decision-making method is applied in this context to analyze and prioritize potential resilience-enhancing solutions. The core objective of this analysis is to determine the most suitable strategy among the proposed alternatives to support long-term planning efforts aimed at strengthening the resilience of power system infrastructure across Greece. Specifically, region-specific recommendations are developed to address vulnerabilities arising from the distinct extreme weather phenomena each region encounters.

It is important to underscore that this methodological framework emphasizes infrastructure planning and long-term resilience enhancement, rather than short-term or immediate operational responses to power system disruptions. Strategies include the integration of Internet of Things (IoT) technologies to facilitate real-time monitoring and control within a digitalized power distribution network.

The ELECTRE method utilizes a set of weather-related criteria representing various extreme weather phenomena, each of which poses distinct risks to the distribution network. These criteria are as follows:

• C1—Cyclone: High-impact weather event involving strong winds and flooding. While it minimally affects underground cables, it can flood underground substations and significantly damage overhead lines, poles, and transformers due to wind and debris.
• C2—Frost: Ice formation on power lines adds weight, while the mechanical stress on power lines increases, causing sagging and breakage. It can also impair the insulation of electrical components.
• C3—Hail: Hail causes mechanical damage to transformers, substations, and overhead lines, potentially resulting in outages.
• C4—Heatwave: Elevated temperatures lead to increased energy demand (for cooling), often overloading transformers and causing thermal expansion and sagging of power lines.
• C5—Lightning: Direct strikes can destroy transformers, substations, and distribution lines, causing surges, fires, and widespread blackouts.
• C6—Rainfall: Heavy rain can result in flooding, damaging both overhead and underground infrastructure. It may compromise utility pole stability and cause insulator flashovers and short circuits due to excessive moisture. Often, technical crews find difficulties in accessing the network to perform operations and restore power.
• C7—Snow: Snow accumulation can cause line breakage due to extra weight and tree collapse onto lines, resulting in outages. Accessibility to the network is extremely difficult during this event.
• C8—Wind: Strong winds can damage overhead networks through falling trees and debris. Poles may be uprooted, and lines snapped. There is minimal impact on underground systems.

The proposed strategic actions for overhead power systems—evaluated using the ELECTRE method—include the following:

• A1—Undergrounding overhead power lines: Conversion to underground systems in high-risk regions (e.g., those affected by cyclones, snow, or strong winds) can significantly improve network reliability. Partial undergrounding, where feasible, offers a cost-effective compromise. However, this method is less effective against phenomena such as heatwaves or heavy rainfall.
• A2—Cable/conductor replacement: Deploying insulated or covered conductors reduces the risk of short circuits caused by falling debris, particularly during lightning, heatwaves, or high winds. It demonstrates moderate effectiveness for other weather events.
• A3—Vegetation management: Regular trimming of trees near power lines minimizes the likelihood of outages due to falling branches during storms, snow, or wind events.
• A4—Concrete poles: Replacing wooden poles with concrete alternatives enhances resilience to high winds, cyclones, snow, and fire.
• A5—Compact line design: Implementing compact configurations helps reduce line sag and sway during wind, hail, or rainfall events, thereby enhancing network stability.
• A6—Substation design improvements: In flood-prone areas, elevating substations and installing waterproof equipment is critical. Fire-resistant equipment is also mandatory in areas with dense vegetation. Modern designs can incorporate fire-resistant coatings and improved structural resilience.
• A7—Smart-grid technologies: The integration of smart technologies is vital to the ongoing digitalization of the Greek electricity distribution network. Accelerating the deployment of smart systems can enable more effective inspection, control, and real-time monitoring.
• A8—Proactive inspections (using drones and satellites): Routine inspections using aerial and satellite technologies can identify system vulnerabilities and enable rapid responses. Drones are especially effective in post-event assessments, reducing fault detection times and aiding in system restoration.

The above mentioned strategies and criteria are consolidated into a decision matrix (

Table 2. Decision matrix for resilience enhancement in overhead power systems.

Actions	Power System Resilience Criteria
	C1	C2	C3	C4	C5	C6	C7	C8
A1	7	4	4	2	7	3	6	7
A2	4	5	4	5	6	4	5	4
A3	4	3	5	3	4	7	6	7
A4	7	4	4	3	4	4	6	7
A5	6	4	6	4	5	5	4	6
A6	7	4	4	6	5	7	4	5
A7	6	4	3	7	7	6	4	5
A8	6	5	5	5	6	5	5	6
Weights	0.12	0.14	0.09	0.16	0.12	0.11	0.14	0.12

), facilitating the systematic application of the ELECTRE method to identify the most effective regional resilience strategies.

To improve the resilience of underground power infrastructure against extreme weather events, a range of strategic actions have been identified and evaluated using the ELECTRE (ELimination Et Choix Traduisant la REalité) multi-criteria decision-making method. These actions address specific vulnerabilities associated with underground systems and support long-term planning to ensure the continuity and reliability of the power supply across Greek regions.

The following measures are proposed for enhancing the performance and robustness of underground power networks:

• B1—Installation of pumps and flood control systems: Installing drainage pumps and flood mitigation infrastructure in vulnerable areas allows rapid water removal during heavy rainfall or cyclonic events. These systems help protect underground assets from flooding, reducing the likelihood of short circuits, equipment failure, and prolonged outages.
• B2—Water-resistant insulation and sealed cables: Using cables with water-resistant insulation and sealed joints enhances network resilience to moisture and submersion. This is particularly important in areas exposed to flooding caused by rainfall or cyclones.
• B3—Elevation of substations and critical infrastructure in flood-prone areas: Raising substations and vital electrical equipment above flood levels mitigates water damage and ensures operational continuity. This proactive approach supports rapid recovery during and after extreme weather events and enhances grid resilience to climate-induced challenges.
• B4—Heat-resistant cable insulation: Deploying thermally resilient insulation materials prevents overheating and cable degradation in high-temperature environments. This reduces the risk of thermal failure, prolongs equipment life, and ensures uninterrupted power delivery during heatwaves and high-demand periods.
• B5—Use of high-quality, resilient cables (XLPE): Cross-linked polyethylene (XLPE) cables exhibit high resistance to electrical stress, moisture, and corrosion. They improve network durability and reliability, especially in harsh conditions. While widely deployed in Greece, some regions require upgrades to their existing underground cabling.
• B6—Underground sensors and active cooling systems: Integrating sensors to monitor temperature, moisture, and insulation integrity in real time, combined with active cooling systems, enhances early fault detection and operational efficiency. These systems are instrumental in regulating the underground environment and preventing thermal-induced failures.
• B7—Regular inspections: Employing advanced inspection technologies, such as drones for aerial monitoring, cable fault detection vans (for underground cables), thermal imaging (to detect overheating), and partial discharge systems (to detect insulation breakdown), facilitates the early identification of issues. This comprehensive inspection regime is effective across all weather conditions, regardless of geomorphology, significantly reducing outage risk.
• B8—Smart-grid technologies: Incorporating smart-grid solutions enables automated fault detection, system reconfiguration, real-time monitoring, and load balancing. These technologies expedite outage recovery, maintain grid stability, and enhance the overall responsiveness to disruptions. For example, smart sensors and automated switches detect and isolate the faults quickly, supporting faster restoration and minimizing outage duration. Additionally, these systems can swiftly reroute power and optimize load distribution to maintain grid stability during disruptions.

The above actions and their associated criteria are summarized in the decision matrix presented in

Table 3. Decision matrix for resilience enhancement in underground power systems.

Actions	Power System Resilience Criteria
	C1	C2	C3	C4	C5	C6	C7	C8
B1	7	3	3	2	4	7	4	3
B2	5	5	4	3	6	6	5	4
B3	7	2	2	2	3	7	3	2
B4	3	3	3	7	6	3	3	4
B5	4	5	4	4	7	5	5	5
B6	4	3	3	7	6	4	3	5
B7	5	5	5	5	5	5	5	5
B8	6	4	4	7	7	6	4	5
Weights	0.08	0.21	0.01	0.27	0.13	0.20	0.10	0

.

Each alternative is assessed against a consistent set of weather-related criteria, using the same evaluation framework applied to overhead systems. A seven-point scale is used to score the effectiveness of each action per criterion. This scale strikes a balance between granularity and interpretability, avoiding the ambiguity of a three-point scale and the complexity of a nine-point one. For instance, using a nine-point scale risks increasing complexity and makes it harder to identify clear and distinct solutions, while a three-point or five-point scale might fail to capture subtle differences between alternatives.

The criteria used in the ELECTRE method remain identical for both overhead and underground systems to ensure methodological consistency. Although certain weather phenomena—such as wind—have negligible or no impact on underground networks, they are retained in the evaluation for uniformity. To ensure accurate representation in the decision-making process, weights derived from the weather risk analysis are applied to each criterion. Accordingly, the weight assigned to the wind criterion in the evaluation of underground systems is set to zero, effectively excluding it from influencing the outcome.

The decision matrices developed for the ELECTRE method reflect the resilience level achieved by each alternative. This technically focused analysis seeks to identify the most effective resilience-enhancement strategies without taking implementation costs into account at this phase. Nevertheless, the integration of a techno-economic assessment in future applications of the ELECTRE method could offer more comprehensive insights and strengthen the decision-making process.

As noted, the criteria used for ELECTRE are consistently applied across both overhead and underground systems, even though some phenomena (e.g., wind) exert little influence on the latter. Maintaining a unified set of evaluation criteria facilitates comparative analysis, while the strategic use of criterion-specific weights—such as assigning zero weight to wind in underground systems—ensures relevance and fairness in the evaluation process.

The ELECTRE method produces dominance tables that establish a hierarchy of alternatives based on their effectiveness in enhancing power system resilience to extreme weather phenomena. These rankings are derived separately for overhead and underground networks.

For overhead networks, the dominance table resulting from the application of the ELECTRE method indicates that the alternative “proactive inspections (drones and satellites)” is more suitable when compared to the other methods:

• Substation design improvement;
• Proactive inspections (drones and satellites);
• Undergrounding;
• Concrete poles;
• Compact line design;
• Smart-grid technologies;
• Vegetation management;
• Cable/conductor replacement.

For underground networks, the dominance table resulting from the application of the ELECTRE method indicates that the alternative “smart-grid technologies” is more suitable when compared to the other methods:

1.

Smart-grid technologies;

2.

Water-resistant insulation and sealed cables;

3.

Using high-quality resilient cables (XLPE);

4.

Regular inspections;

5.

Installation of pumps and flood control systems;

6.

Heat-resistant cable insulation;

7.

Underground sensors and active cooling systems;

8.

Elevation of substations and critical infrastructure in flood-prone areas.

The alternative that is indicated as suitable in each case outperforms the others.

Subsequently, the outcomes of the ELECTRE method are used to inform region-specific strategies for enhancing power system resilience in areas identified as high-risk. This targeted approach supports the selection of the most appropriate measures among the proposed alternatives and provides a foundation for the development of strategic plans at the local level.

Alternative resilience enhancement measures are categorized according to the risk rate associated with each area:

• Low resilience restoration: risk rate of 0–33%;
• Moderate resilience restoration: risk rate of 34–66%;
• High resilience restoration: risk rate of 67–100%.

The higher the identified risk level, the more extensive the recommended interventions. Based on the types and frequencies of extreme weather events observed during the 2020–2024 period, the most suitable resilience measures for each Greek region are proposed in Appendix B.

A summarized overview of these region-specific strategies is provided in the consolidated

Table 4. Best power system enhancement methods for each Greek region.

Greek Regions Whose Grid Is Affected by HILF Events	Best Method for Enhancing Overhead Power Systems	Greek Regions Whose Grid Is Affected by HILF Events	Best Method for Enhancing Underground Power Systems
Corfu, Elis, Kozani, Rhodes	Substation design improvement	Aetoloacarnania, Corfu, Evros, North Athens	Smart-grid technologies
Central Athens, Evia	Proactive inspections (drones, satellites)	Arta, Central Athens, Chania, East Attica, Evia, Fokida, Fthiotida, Imathia, Ioannina, Karditsa, Kefalonia, Kos, Naxos, Rhodes, Sporades, Voeotia, West Athens, West Attica	Water-resistant insulation and sealed cables
Aetoloacarnania, Arcadia, Arta, Chalkidiki, Chania, Chios, Evros, Fthiotida, Heraklion, Ikaria, Imathia, Ioannina, Ithaca, Karditsa, Kavala, Kefalonia, Larissa, Naxos, North Athens, Pieria, Preveza, Rethymno, Serres, Sporades, Thesprotia, Thessaloniki, Trikala, West Athens, Xanthi, Zakynthos	Undergrounding	Achaea, Heraklion	High-quality resilient cables (XLPE)
East Attica	Concrete poles	Chalkidiki, Trikala	Regular inspections
-	Compact line design	Andros, Attica islands, Chios, Elis, Ithaca, Kavala, Kozani, Larissa, Lesvos, Magnesia, Pieria, Preveza, Rethymno, Serres, Thesprotia, Thessaloniki, Tinos, Xanthi, Zakynthos	Installation of pumps and flood control systems
Laconia, Piraeus	Smart-grid technologies	Laconia, Piraeus, South Athens	Heat-resistant cable insulation
Achaea, Andros, Attica islands, Lesvos, Magnesia, South Athens, Tinos, Voeotia	Vegetation management	-	Underground sensors and active cooling systems
Fokida, Kos, West Attica	Cable/conductor replacement	-	Elevation of substations and critical infrastructure in flood-prone areas

.

5.3. Discussion of ELECTRE’s Output

Table A6. Suggested methods and actions for enhancing grid resilience in Greek regions affected by HIHF events.

Region	Risk Score	Need of Resilience Enhancement	HIHF Events (Times Occurred)	Suggested Actions (in Priority Order) for Enhancing Overhead Power Systems	Suggested Actions (in Priority Order) for Enhancing Underground Power Systems
Achaea	42.29	Medium	Hail (1) Rainfall (1) Wind (1)	1. Vegetation management 2. Compact line design 3. Proactive inspections (drones, satellites)	1. High quality resilient cables 2. Smart-grid technologies 3. Regular inspections
Aetoloacarnania	50.21	Medium	Cyclone (3) Hail (1) Heatwave (1) Wind (1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Substation design improvement 4. Compact line design	1. Smart-grid technologies 2. Regular inspections 3. Water-resistant insulation and sealed cables 4. Installation of pumps and flood systems
Andros	53.33	Medium	Wind (1) Rainfall (1)	1. Vegetation management 2. Compact line design 3. Proactive inspections (drones, satellites) 4. Concrete poles	1. Installation of pumps and flood systems 2. Flood barriers 3. Water-resistant insulation and sealed cables 4. Smart-grid technologies
Arcadia	43.13	Medium	Wind (1)	1. Undergrounding of the overhead power systems 2. Vegetation management 3. Concrete poles	-
Arta	56.67	Medium	Frost (1) Snow(1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Cable/conductor replacement 4. Proactive inspections (drones, satellites)	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies
Attica Islands	53.33	Medium	Wind (1) Rainfall (1)	1. Vegetation management 2. Compact line design 3. Proactive inspections (drones, satellites) 4. Concrete poles	1. Installation of pumps and flood systems 2. Flood barriers 3. Water-resistant insulation and sealed cables 4. Smart-grid technologies
Central Athens	71.04	High	Frost (2) Heatwave (2) Rainfall (1) Snow (2) Wind (1)	1. Proactive inspections (drones, satellites) 2. Substation design improvement 3. Undergrounding of the overhead power system 4. Concrete poles 5. Cable/conductor replacement	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies 5. Installation of pumps and flood control systems
Chalkidiki	62.00	Medium	Cyclone (3) Frost (1) Hail (1) Snow (1) Wind (2)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Proactive inspections (drones, satellites) 4. Substation design improvement	1. Regular inspections 2. Water-resistant insulation and sealed cables 3. High quality resilient cables (XLPE) 4. Smart-grid technologies
Chania	59.94	Medium	Frost (1) Lightning (1) Rainfall (2) Snow (1) Wind (2)	1. Undergrounding of the overhead power system 2. Proactive inspections (drones, satellites) 3. Concrete poles 4. Smart-grid technologies	1. Water-resistant insulation and sealed cables 2. High quality cable insulation (XLPE) 3. Regular inspections 4. Smart-grid technologies
Chios	39.17	Medium	Cyclone (1)	1. Undergrounding of the overhead power systems 2. Concrete poles 3. Substation design improvement	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies
Corfu	67.80	High	Cyclone (4) Hail (1) Lightning (1) Rainfall (3) Wind (5)	1. Substation design improvement 2. Undergrounding of the overhead power system 3. Proactive inspections (drones, satellites) 4. Compact line design 5. Smart-grid technologies	1. Smart-grid technologies 2. Water-resistant insulation and sealed cables 3. Installation of pumps and flood control systems 4. Regular inspections 5. High quality resilient cables (XLPE)
East Attica	73.98	High	Cyclone (1) Frost (3) Heatwave (2) Rainfall (1) Snow (3) Wind (4)	1. Concrete poles 2. Undergrounding of the overhead power system 3. Proactive inspections (drones, satellites) 4. Vegetation management 5. Compact line design or substation design improvement	1. Water-resistant insulation and sealed cables 2. Regular inspections 3. High quality resilient cables (XLPE) 4. Smart-grid technologies 5. Installation of pumps and flood control systems
Elis	55.97	Medium	Cyclone (3) Heatwave (1) Rainfall (1)	1. Substation design improvement 2. Smart-grid technologies 3. Concrete poles 4. Undergrounding of the overhead power system	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables or regular inspections
Evia	67.43	High	Cyclone (2) Frost (3) Heatwave (2) Lightning (1) Rainfall (1) Snow (3) Wind (1)	1. Proactive inspections (drones, satellites) 2. Undergrounding of the overhead power system 3. Concrete poles 4. Substation design improvement 5. Smart-grid technologies	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies 5. Installation of pumps and flood control systems
Evros	55.63	Medium	Cyclone (1) Heatwave (1) Wind (1)	1. Undergrounding of the overhead power systems 2. Concrete poles 3. Substation design improvement 4. Smart-grid technologies	1. Smart-grid technologies 2. Underground sensors and active cooling systems 3. Regular inspections 4. Installation of pumps and flood control systems
Fokida	58.96	Medium	Frost (1) Heatwave (1) Snow (1)	1. Cable/conductor replacement 2. Proactive inspections (drones, satellites) 3. Smart-grid technologies 4. Substation design improvement	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies
Fthiotida	52.71	Medium	Frost (2) Snow (2)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Cable/conductor replacement 4. Proactive inspections (drones, satellites)	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies
Heraklion	53.33	Medium	Lightning (1)	1. Undergrounding of the overhead power system 2. Smart-grid technologies 3. Cable/conductor replacement 4. Proactive inspections (drones, satellites)	1. High quality resilient cables (XLPE) 2. Smart-grid technologies 3. Underground sensors and active cooling systems 4. Heat-resistant cable insulation
Ikaria	14.17	Low	Wind (1)	1. Undergrounding of the overhead power systems	-
Imathia	50.00	Medium	Rainfall (1) Snow (1) Wind (1)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles 4. Proactive inspections (drones, satellites)	1. Water-resistant insulation and sealed cables 2. Installation of pumps and flood control systems 3. High quality resilient cables (XLPE) 4. Regular inspections
Ioannina	56.67	Medium	Frost (1) Snow (1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Cable/conductor replacement 4. Proactive inspections (drones, satellites)	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies
Ithaca	55.00	Medium	Cyclone (1) Wind (1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Compact line design 4. Proactive inspections (drones, satellites)	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables
Karditsa	48.33	Medium	Cyclone (2) Frost (1) Hail (2) Lightning (1) Rainfall (3) Snow (1) Wind (2)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Substation design improvement	1. Water-resistant insulation and sealed cables 2. Smart-grid technologies 3. Installation of pumps and flood control systems
Kavala	54.79	Medium	Cyclone (1) Rainfall (1) Wind (1)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles 4. Substation design improvement	1. Installation of pumps and flood control 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables
Kefalonia	55.00	Medium	Cyclone (1) Wind (1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Compact line design 4. Proactive inspections (drones, satellites)	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables
Kos	58.13	Medium	Frost (1) Heatwave (1) Snow (1)	1. Cable/conductor replacement 2. Proactive inspections (drones, satellites) 3. Compact line design 4. Substation design improvement	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies
Kozani	50.69	Medium	Cyclone (2) Lightning (1) Rainfall (2) Wind (1)	1. Substation design improvement 2. Undergrounding of the overhead power system 3. Smart-grid technologies 4. Proactive inspections (drones, satellites)	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables
Laconia	55.00	Medium	Heatwave (1)	1. Smart-grid technologies 2. Substation design improvement 3. Proactive inspections (drones, satellites) 4. Cable/conductor replacement	1. Heat-resistant cable insulation 2. Underground sensors and active cooling systems 3. Smart-grid technologies 4. Regular inspections
Larissa	40.21	Medium	Cyclone (2) Frost (1) Hail (1) Lightning (1) Rainfall (3) Snow (1) Wind (1)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Substation design improvement	1. Installation of pumps and flood control systems 2. Flood barriers 3. Water-resistant insulation and sealed cables
Lesvos	49.58	Medium	Hail (1) Rainfall (2) Wind (1)	1. Vegetation management 2. Substation design improvement 3. Compact line design	1. Installation of pumps and flood control systems 2. Flood barriers 3. Water-resistant insulation and sealed cables
Magnesia	65.65	Medium	Cyclone (2) Frost (2) Lightning (2) Rainfall (6) Snow (1) Wind (5)	1. Vegetation management 2. Substation design improvement 3. Smart-grid technologies 4. Compact line design 5. Proactive inspections (drones, satellites)	1. Installation of pumps and flood control systems 2. Flood barriers 3. Water-resistant insulation and sealed cables 4. Smart-grid technologies 5. High quality resilient cables (XLPE)
Messenia	58.96	Medium	Cyclone (1) Heatwave (1) Rainfall (1)	1. Substation design improvement 2. Smart-grid technologies 3. Compact line design 4. Proactive inspections (drones, satellites)	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables
Naxos	53.33	Medium	Frost (1) Snow (1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Cable/conductor replacement 4. Proactive inspections (drones, satellites)	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies
North Athens	68.51	High	Cyclone (1) Frost (1) Heatwave (3) Rainfall (3) Snow (1) Wind (4)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles 4. Proactive inspections (drones, satellites) 5. Compact line design	1. Smart-grid technologies 2. Water-resistant insulation and sealed cables 3. Regular inspections 4. Installation of pumps and flood control systems 5. High quality resilient cables (XLPE)
Pieria	58.13	Medium	Cyclone (1) Rainfall (1) Wind (2)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles 4. Substation design improvement	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables
Piraeus	75.00	High	Heatwave (1)	1. Smart-grid technologies 2. Substation design improvement 3. Proactive inspections (drones, satellites) 4. Cable/conductor replacement	1. Heat-resistant cable insulation 2. Underground sensors and active cooling systems 3. Smart-grid technologies 4. Regular inspections
Preveza	37.50	Medium	Cyclone (1) Wind (1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Compact line design	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies
Rethymno	55.63	Medium	Cyclone (1) Rainfall (1) Wind (1)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles 4. Substation design improvement	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables
Rhodes	51.00	Medium	Cyclone (1) Frost (1) Heatwave (1) Rainfall (2) Snow (1)	1. Substation design improvement 2. Smart-grid technologies 3. Proactive inspections (drones, satellites) 4. Vegetation management	1. Water-resistant insulation and sealed cables 2. Smart-grid technologies 3. Installation of pumps and flood control systems 4. Regular inspections
Serres	44.58	Medium	Cyclone (1) Rainfall (2) Snow (1) Wind (3)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles	1. Installation of pumps and flood control systems 2. Flood barriers 3. Water-resistant insulation and sealed cables
South Athens	63.75	Medium	Heatwave (2) Rainfall (1) Wind (2)	1. Vegetation management 2. Substation design improvement 3. Smart-grid technologies 4. Compact line design 5. Proactive inspections (drones, satellites)	1. Heat-resistant cable insulation 2. Underground sensors and active cooling systems 3. Smart-grid technologies 4. Regular inspections 5. High quality resilient cables (XLPE)
Sporades	58.96	Medium	Cyclone (1) Frost (1) Lightning (1) Rainfall (2) Snow (1) Wind (2)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Proactive inspections (drones, satellites) 4. Substation design improvement	1. Water-resistant insulation and sealed cables 2. Smart-grid technologies 3. Installation of pumps and flood control systems 4. High quality cable insulation
Thesprotia	49.38	Medium	Cyclone (1) Rainfall (1) Wind (1)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies
Thessaloniki	51.88	Medium	Cyclone (3) Hail (1) Lightning (1) Rainfall (2) Wind (3)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Vegetation management 4. Substation design improvement	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resistant insulation and sealed cables
Tinos	53.33	Medium	Rainfall (1) Wind (1)	1. Vegetation management 2. Concrete poles 3. Substation design improvement 4. Compact line design	1. Installation of pumps and flood control systems 2. Flood barriers 3. Water-resistant insulation and sealed cables 4. Smart-grid technologies
Trikala	59.17	Medium	Cyclone (2) Frost (1) Rainfall (1) Snow (1) Wind (1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Proactive inspections (drones, satellites) 4. Substation design improvement	1. Regular inspections 2. Water-resistant insulation and sealed cables 3. High quality resilient cables (XLPE) 4. Smart-grid technologies
Voeotia	60.83	Medium	Frost (3) Rainfall (1) Snow (3) Wind (1)	1. Vegetation management 2. Concrete poles 3. Undergrounding of the overhead power system 4. Proactive inspections (drones, satellites)	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Installation of pumps and flood control systems
West Athens	68.96	High	Frost (1) Heatwave (1) Snow (1) Wind (1)	1. Undergrounding of the overhead power system 2. Concrete poles 3. Proactive inspections (drones, satellites) 4. Cable/conductor replacement 5. Vegetation management	1. Water-resistant insulation and sealed cables 2. High quality resilient cable (XLPE) 3. Regular inspections 4. Smart-grid technologies 5. Installation of pumps and flood control systems
West Attica	59.58	Medium	Frost (2) Heatwave (2) Snow (2) Wind (1)	1. Cable/conductor replacement 2. Proactive inspections (drones, satellites) 3. Compact line design 4. Substation design improvement	1. Water-resistant insulation and sealed cables 2. High quality resilient cables (XLPE) 3. Regular inspections 4. Smart-grid technologies
Xanthi	53.13	Medium	Cyclone (1) Snow (1) Wind (2)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles 4. Proactive inspections (drones, satellites)	1. Installation of pumps and flood control systems 2. Water-resistant insulation and sealed cables 3. Regular inspections 4. Flood barriers
Zakynthos	57.29	Medium	Cyclone (1) Rainfall (1) Wind (2)	1. Undergrounding of the overhead power system 2. Vegetation management 3. Concrete poles 4. Compact line design	1. Installation of pumps and flood control systems 2. Flood barriers 3. Smart-grid technologies 4. Water-resilient insulation and sealed cables

in Appendix B presents the results obtained through the application of the ELECTRE method, as outlined in the preceding sections. The primary aim of this analysis is to identify the most appropriate resilience enhancement measures for each Greek region classified as high-risk, based on the weather risk assessment conducted in this study. The recommendations provided are tailored to reflect both the level of climate-related risk and the specific extreme weather events affecting each region.

Each region is assigned a prioritized list of actions, ranked in descending order of effectiveness. The top-ranked action in each list is considered the most suitable resilience measure for that particular region. The type of distribution network—whether overhead or underground—was also a key consideration in the selection process, ensuring that the proposed measures align with the physical and operational characteristics of the existing infrastructure.

In regions where strong wind was identified as the sole weather-related threat, no measures were proposed for underground systems, given that such systems are largely immune to wind-related impacts. This underscores the importance of adapting resilience strategies not only to regional risk levels but also to network typology and the nature of the hazard.

The results demonstrate that vulnerabilities differ significantly across regions, emphasizing the necessity for targeted, region-specific approaches to power system resilience. These findings suggest that further refinement of the methodology—such as conducting sub-regional analyses within each geographic area—could yield more granular, accurate, and actionable insights to guide infrastructure planning and investment.

Finally, integrating this type of decision-support framework into national and regional infrastructure resilience planning could enhance evidence-based policymaking, strengthen preparedness, and optimize the allocation of resources. This is particularly critical in the context of climate change, which is expected to increase the frequency, intensity, and unpredictability of extreme weather events affecting energy infrastructure across Greece.

6. International Applications of HIHF-Type Frameworks

While the HIHF framework presented in this study is tailored to the Greek distribution network and its unique geographical and climatic challenges, similar methodologies emphasizing climate-driven resilience and frequency-based event assessment have been implemented in other countries with notable success.

In Australia, regulatory authorities have adopted formal resilience obligations for electricity transmission and distribution networks, emphasizing scenario-based planning and real-time monitoring. These frameworks closely align with the HIHF philosophy, as they require utilities to account for weather patterns that may no longer be rare but are increasingly recurrent and disruptive. Tools like ESKIES (Enabling Resilient, Affordable, and Equitable Community Energy Solutions) incorporate both meteorological data and community vulnerability profiles to enhance grid robustness and social equity in disaster preparedness [53].

Similarly, Japan has integrated high-frequency disaster modeling into national grid planning through its initiatives led by the National Research Institute for Earth Science and Disaster Resilience. Their approach focuses not only on the probability of events but also on infrastructure vulnerability and exposure, closely mirroring the objectives of the HIHF framework. Japan’s model is further enhanced by a deep integration of IoT sensors and localized data systems to enable swift restoration and adaptive planning across regional networks [54].

In the United States, the Department of Energy’s Grid Deployment Office has invested in regional risk assessment frameworks that examine the compound effects of climate events on the grid. These efforts emphasize the need for flexible, data-driven planning strategies that can handle high-frequency disturbances, moving beyond traditional probabilistic models. Notably, these frameworks often integrate geospatial risk mapping and decision-support tools to prioritize grid hardening investments under constrained budgets—an approach that parallels the combined use of HIHF and ELECTRE models proposed in this paper [55].

These international cases underscore a growing consensus: traditional risk assessment methods that prioritize HILF scenarios may no longer be adequate in the face of climate variability. The HIHF model offers a complementary and increasingly essential tool for resilience planning, capable of being adapted to diverse national contexts. Its integration with tools like the ELECTRE method enhances its applicability not only in the energy sector but potentially across other infrastructure systems as well.

7. Conclusions—Future Work

As extreme weather events become both more frequent and intense, traditional risk assessment frameworks—which have long treated such phenomena as rare outliers—are increasingly inadequate. In this evolving context, the High Impact High Frequency (HIHF) model emerges as a critical tool, reflecting the new operational reality of power grids under accelerating climate change. The HIHF model prioritizes the need to design and operate energy infrastructure capable of withstanding frequent, high-impact disruptions, which are no longer exceptions but growing norms.

By leveraging historical meteorological data specific to each region and integrating smart-grid technologies for real-time monitoring and adaptive response, the HIHF model facilitates proactive planning, rapid restoration, and uninterrupted power supply even under extreme conditions. This paradigm shift is essential not only for grid resilience but also for protecting public health, safety, and the continuous operation of critical infrastructure and services.

The HIHF framework provides tailored strategies to mitigate the impacts of extreme events by enhancing the resilience of distribution networks to each specific weather phenomenon. Energy companies that adopt the HIHF model are better equipped to protect their infrastructure, reduce service disruptions, and maintain high levels of customer satisfaction through reliable energy delivery.

Despite its relevance, the HIHF model has not yet been formally implemented in Greece as a standardized approach for strategic grid restoration and resilience planning. However, the recurring weather-related disruptions experienced in recent years underscore the urgent need for a cohesive national framework that incorporates HIHF principles. Systematic adoption of this model would significantly strengthen the capacity of Greek energy infrastructure to withstand modern climate-related challenges.

Other established models, like HILP and HIHP, are commonly used to analyze weather phenomena with significant impacts on the distribution network with either low or high probabilities of occurrence. However, it is important to also consider the frequency of these events, as their occurrence has been increasing rapidly in recent years due to climate change. As a result, the probability of extreme weather events is no longer a reliable predictor for anticipating future impacts on the network. Low probability does not necessarily mean low frequency. This is the main reason this study focuses on the transition of the HILF model to the HIHF model as a tool for strategic resilience planning.

Historical data have traditionally been used to assess the impact of weather events on distribution networks. However, recent trends suggest that the probability-based classification of such events is no longer sufficient, as their frequency and severity are both rising. The distinction between low probability and high frequency is increasingly blurred—making probability alone an unreliable predictor of future impacts. This study, therefore, emphasizes a necessary shift from HILF frameworks toward HIHF-based resilience planning.

In this context, the study integrates the ELECTRE method—a multi-criteria decision-making (MCDM) approach—with the HIHF model. The ELECTRE method is particularly well-suited for complex decision environments where multiple, often conflicting criteria must be considered simultaneously. Rather than providing a simple ranking or score, ELECTRE enables pairwise comparisons of alternatives to identify those that outperform others under defined thresholds. This structured evaluation supports robust decision-making in infrastructure planning, energy systems, and climate resilience enhancement.

While the methodology presented here focuses on analyzing climate risks to power distribution infrastructure, it has broader applicability. It can be adapted to improve the resilience of other critical infrastructure systems and implemented in different countries or regions, provided the relevant criteria and network characteristics are appropriately calibrated to local conditions.

The weather risk analysis conducted in this study is based on historical meteorological data. The results confirm a notable increase in extreme weather events aligning more closely with the HIHF model than the HILF paradigm. Nevertheless, the accuracy of risk assessments could be significantly improved through the integration of real-time weather data from smart-grid sensors deployed throughout the distribution network. This highlights the imperative need to modernize Greece’s electrical grid into a fully functional smart-grid system. Moreover, adopting IoT technologies would allow infrastructure to operate in real time, further enhancing the system’s adaptability and resilience.

As discussed, the current application of the ELECTRE method is intended primarily to identify optimal solutions from a technical perspective. It serves as a prototype for evaluating and selecting the most effective strategies to improve grid resilience under extreme weather conditions. The approach can be further refined by integrating additional parameters such as construction costs, implementation timelines, or regulatory constraints, thus evolving into a techno-economic decision model.

Furthermore, the integration of ELECTRE with GIS and the development of customized algorithms could enable the generation of region-specific solutions, ranked by performance indicators that consider local geomorphological conditions. In such cases, impractical or ineffective options could be automatically excluded, streamlining the selection process and enhancing the relevance of proposed interventions.

In conclusion, the combined application of the HIHF framework and the ELECTRE method represents a comprehensive, forward-looking approach to climate-resilient infrastructure planning. It offers substantial potential to inform policy, guide investment, and safeguard power system reliability in the face of an increasingly volatile climate.