Identifying and Prioritizing Climate-Related Natural Hazards for Nuclear Power Plants in Korea Using Delphi

Abstract

Climate change is projected to increase the intensity and frequency of natural hazards such as heat waves, extreme rainfall, heavy snowfall, typhoons, droughts, floods, and cold waves, potentially impacting the operational safety of critical infrastructure, including nuclear power plants (NPPs). Although quantitative indicators exist to screen-out natural hazards at NPPs, comprehensive methodologies for assessing climate-related hazards remain underdeveloped. Furthermore, given the variability and uncertainty of climate change, it is realistically and resource-wise difficult to evaluate all potential risks quantitatively. Using a structured expert elicitation approach, this study systematically identifies and prioritizes climate-related natural hazards for Korean NPPs. An iterative Delphi survey involving 42 experts with extensive experience in nuclear safety and systems was conducted and also evaluated using the best–worst scaling (BWS) method for cross-validation to enhance the robustness of the Delphi priorities. Both methodologies identified extreme rainfall, typhoons, marine organisms, forest fires, and lightning as the top five hazards. The findings provide critical insights for climate resilience planning, inform vulnerability assessments, and support regulatory policy development to mitigate climate-induced risks to Korean nuclear power plants.

1. Introduction

Climate change has increased the frequency of natural hazards such as heat waves, extreme rains, heavy snowfall, typhoons, droughts, floods, and cold waves [1,2,3,4]. Globally, the number of climate change-related events resulting in human and economic losses continues to rise [5,6]. For example, Europe experienced unseasonably high temperatures from late 2022 to early January 2023, whereas Southeast Asia experienced a severe heat wave in April 2023 [1,7]. Furthermore, in April 2023, Chicago experienced a record-high temperature, followed four days later by a sudden drop to subzero temperatures, resulting in heavy snowfall [7]. In East Africa, below-average rainfall during the rainy season over the last few years has resulted in widespread and prolonged droughts. However, in 2023, extreme rains caused widespread flooding, displacing a large number of people [7,8].

According to the Intergovernmental Panel on Climate Change (IPCC) report, climate change is expected to increase the intensity and frequency of natural disasters [9]. As a result, severe natural disasters can pose both direct and indirect threats to the safety and operation of critical infrastructure, such as nuclear power plants [10,11,12,13]. Nuclear power plants must be designed to operate reliably in the event of natural hazards [14,15]. However, changes in the frequency and intensity of natural hazards caused by climate change may affect the safety and operation of nuclear power plants [16,17,18,19,20]. For example, in Texas—a region typically known for its warm climate—a nuclear power plant was tripped due to a sudden cold snap that froze sensing lines [17,21]. Meanwhile, in France, heatwaves and abnormally high temperatures have frequently caused nuclear power plants to close or reduce production [11,22,23]. In Korea, changes in sea temperature have accelerated the appearance of marine organisms, necessitating the shutdown of several nuclear power plants [11]. Additionally, in several countries, nuclear power plants have been halted due to off-site power outages caused by typhoons and hurricanes [11]. Therefore, effective protective designs and countermeasures against natural hazards associated with climate change must be implemented to ensure the operation and safety of nuclear power plants [17,24].

Research on nuclear power plants in response to climate change is underway [11,20,25,26,27]. For example, Kopytko and Perkins analyzed the impact of sea-level rise, coastal erosion, storms, floods, and heatwaves on nuclear power plants [25]. Ahmad empirically confirmed that the frequency of nuclear power plant shutdowns has significantly increased due to climate change [26]. Meanwhile, Linnerud et al. pointed out that rising temperatures can decrease the thermal efficiency of nuclear power plants and increase the risk of operational shutdowns [27]. Furthermore, the guidelines of various nuclear power agencies specify that climate change should be considered to ensure the safety of nuclear power plants [16,28,29,30,31,32,33,34]. These guidelines emphasize the assessment of the impacts of climate change on nuclear power plant operation and protection design. Although several guidelines and studies have addressed assessing external events, including climate-related events, quantitative assessments of hazards reflecting climate change are minimal. Probabilistic safety assessments (PSA) for natural hazards other than earthquakes are also minimal [11,35,36]. It is practically impossible to thoroughly assess all natural hazards associated with climate change due to their diversity and limited resources. Therefore, it is critical to prioritize natural hazards related to climate change that affect nuclear power plants to maximize resource utilization and respond effectively.

This study identified natural hazards associated with climate change for nuclear power plants in Korea, and priorities were established using the Delphi method. When prior research or standardized data is insufficient, the Delphi method is a qualitative evaluation approach that gathers expert intuition and opinions to reach consensus [37,38,39,40]. The Delphi method is characterized by respondent anonymity, iterative feedback surveys, and statistical processing of responses [41,42]. It effectively prevents distorted communication during committee or expert discussions and issues in which opinions may be skewed due to group pressure or conflict. However, the Delphi method has limitations. Because it relies heavily on experts’ subjective and intuitive judgment, it can lead to bias in experts’ opinions. This study assembled a panel of 42 nuclear experts to ensure a diverse range of perspectives. After each round of the Delphi survey, we calculated the coefficient of variation—a measure of the stability of expert opinions—to determine whether sufficient consensus had been reached and thus conclude the survey. Additionally, we employed the best–worst scaling (BWS) method to cross-validate the priority rankings derived from the Delphi process, thereby offsetting the uncertainty inherent in qualitative evaluations.

2. Research Method

This study used literature and historical data from Korean sources to identify external hazards, whereas climate change-related natural hazards affecting nuclear power plants were identified through expert elicitation. These climate change-related natural hazards were prioritized using two rounds of Delphi surveys. The Delphi surveys evaluated selected natural hazards using two criteria: frequency of occurrence (likelihood) and impact (severity of consequences). Both likelihood and impact were rated on a five-point Likert scale: 1 = very low; 2 = low; 3 = moderate; 4 = high; and 5 = very high. In the second round, experts were presented with summary statistics from the first round, including the mean, median, and interquartile range for each criterion, allowing them to reflect on the group-level responses. Based on the results, a final priority score was calculated for each hazard using Equation (1) [40], where n represents the total number of participating experts and k denotes each climate change-related natural hazard evaluated. In addition, the coefficient of variation (CV) and interquartile range (IQR) for likelihood and impact were analyzed to assess the stability and validity of expert assessments.

$$\mathrm{Priority} \mathrm{Score}={\sum }_{i=1}^n(Likelihoo{d}_{i,k}\times Impac{t}_{i,k})$$

(1)

Because the Delphi method relies on experts’ subjective and intuitive judgments to produce rational results, expert selection is crucial [37,38,39,40,43,44]. Therefore, a panel of 42 nuclear power plant experts drawn from academia, research institutions, industry, and regulatory agencies was formed to evaluate the priorities of natural hazards related to climate change. The 42 experts specialize in nuclear power plant systems and safety.

Table 1. Summary of survey expert panels.

Panel Group	Institution	Number of Panels	Ratio (%)
Research Institute (include University)	Korean Atomic Energy Research Institute (KAERI)	18	43
	Pusan National University Chung-Ang University
Industries	Korea Hydro & Nuclear Power Co., Ltd. (KHNP)	17	40
	Korea Electric • Power Corporation Engineering & Construction (KEPCO E & C) CENITS Corporation INC. FNC Technology Co., Ltd. NESS Nuclear Engineering Services & Solutions Co., Ltd. Innose Tech Co., Ltd.
Regulatory institute	Korea Institute of Nuclear Safety (KINS)	7	17
Total		42	100

summarizes the affiliations of the panel experts who participated in the survey. Figure 1 depicts the panelists’ professional backgrounds and educational qualifications. The composition of this expert panel, with its extensive experience, improved the evaluation’s reliability.

3. Identification of Natural Hazards Related to Climate Change

Identifying hazards that may affect the operation and safety of nuclear power plants, while considering Korea’s climate characteristics, is essential for assessing the prioritization of climate change-related natural hazards. For this purpose, 69 external hazards—excluding seismic hazards—were identified based on operational records from the Operational Performance Information System (OPIS) for Nuclear Power Plants and various literature sources [45,46,47,48,49,50,51].

Table 2. Identification of external hazards [45,46,47,48,49,50,51].

Group	Natural Hazards	Anthropogenic Hazards
Atmospheric hazards	• Extreme air pressure (high/low/gradient) • Extreme winds and tornadoes • Extreme Rain • Fog/Mist • Drought • Frost • Hail • High air pollution • High air temperature • Hurricane/typhoon • Ice storm/freezing rain/sleet • Lightning • Low air temperature • Meteorite • Salt storm • Sandstorm • Snow (including snowstorms and drifts) • Solar storms • Storm surge • Strong winds (other than hurricanes or tornado) • Heavy load drops	• Aircraft impacts • Electromagnetic Interference/disturbance • Externally generated missiles • Satellite • Release of chemicals from storage on site • Toxic gas/chemical release/radioactive release • Pipeline accident • Transportation accidents (including explosion, fire, and smoke from fire)
Ground hazards	• Animals • Avalanche • Forest fire • Land rise • Landslide • Soil shrink–swell, sinkholes • Volcanic activity	• Eddy currents in the ground • Excavation work • External fire other than forest fire (including impact from smoke)—includes transportation accidents, pipeline accidents, etc. • Ground contamination (e.g., from chemicals) • Ground vibration (e.g., due to nearby explosions) • Industrial or military facility accident • Internal fire spreads from another plant • Internally generated missile • Turbine-generated missiles
Water hazards	• Biological events • Coastal erosion • Corrosion (e.g., from salt water) • Erosion • External flooding • Frazil ice • Groundwater (too much or too little) • High water temperature • High tide • Ice barriers • Ice cover • Lake or river-borne material that plugs water intakes/organic material in water • Low lake or river level • Low water temperature • Other extraordinary waves • River diversion • Seiche • Strong currents (underwater erosion) • Tsunami • Underwater landslide (impact on the soil; that is, not tsunami) • Waves • Sediment or silting of intake	• Chemical releases into water • Impurities in water from ship release (solids and liquids)

shows that the identified hazards were first classified into natural and anthropogenic categories, and then subdivided into atmospheric, terrestrial, and hydrological hazards [46,47]. The excluded hazards include animals, high air pollution, meteorites, soil shrink-swell, sinkholes, and heavy load drops.

Table 3. Natural hazards related to climate change.

Number	Natural Hazards Related to Climate Change
1	Avalanche
2	Biological events/Lake- or river-borne material plugging water intakes/organic material in water
3	Coastal erosion
4	Drought
5	Erosion
6	External flooding
7	Extreme air pressure (high/low/gradient)
8	Extreme rain
9	Extreme winds and tornadoes/hurricane/typhoon/strong winds (other than hurricane or tornado)
10	Fog/Mist
11	Forest fire
12	Frazil ice
13	Frost
14	Groundwater (too much or too little)
15	Hail
16	High air temperature
17	High water temperature
18	High tide/Other extraordinary waves/Storm surge/Waves
19	Ice barriers
20	Ice cover
21	Landslide
22	Lightning
23	Low air temperature
24	Low lake or river level
25	Low water temperature
26	River diversion
27	Salt Storm
28	Sandstorm
29	Seiche
30	Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Sleet
31	Strong currents (underwater erosion)
32	Underwater landslide (impact on the soil; that is, not tsunami)

lists related wind hazards, including extreme winds and tornadoes, hurricanes/typhoons, and other strong winds not classified as hurricanes or tornadoes.

Table 3 lists natural hazards associated with climate change, some of which are unlikely to occur or may have little or no impact on nuclear power plants. Therefore, it is critical to identify and eliminate hazards that have a minimal impact on nuclear power plant operations. This filtering process involved six experts with over 25 years of experience in the safety and systems of nuclear power plants (from KAERI and KINS). Before the screening process, the experts were thoroughly briefed and engaged in a structured review of extensive reference materials to ensure consistent judgment and a solid understanding of the technical background. These materials encompassed a wide range of information relevant to evaluating climate-related natural hazards. Specifically, the experts reviewed the qualitative screening criteria outlined in EPRI 102997 (

Table 4. Recommended qualitative screening criteria [47].

Screening Criterion No.	Description
QL-1	The hazard has a lesser damage potential than other similar hazards for which the plant has been designed.
QL-2	The hazard has a significantly lower mean frequency of occurrence than another hazard that has been screened, and the hazard could not result in worse consequences than the other hazard screened.
QL-3	The hazard cannot occur at the site or close enough to the site to affect the plant.
QL-4	The hazard is included in the definition of another hazard.
QL-5	The hazard develops slowly, so it can be demonstrated that there is enough time to eliminate the source of the threat or provide an adequate response.
QL-6	The hazard does not cause an initiating event (including the need for a controlled shutdown) or loss(es) of safety system function needed for the event.
QL-7	The consequences to the plant do not result in a reactor trip or shutdown and do not require the actuation of front-line systems.

), a comprehensive list of 69 external event types, and operational experience data sourced from the Operational Performance Information System (OPIS). In addition, they examined global climate change impact assessments, including those published by the Intergovernmental Panel on Climate Change (IPCC), as well as national-level climate change reports issued by Korea’s Ministry of Environment and Ministry of Science and ICT. Furthermore, by reviewing the studies presented by Kim et al., 2024 [11], SKI Report 02:27 2003 [46] and Shanley and Miller 2015 [47], the experts understood how climate-induced natural hazards can affect the safety and operation of nuclear power plants. Through this preparatory process, the experts were equipped with a consistent and informed basis for identifying and excluding low-priority hazards in the context of nuclear power plant safety.

Table 5. Reasons for screening for natural hazards related to climate change.

Natural Hazards Related to Climate Change	Reason for Screening
Avalanche	QL-2, QL-3, QL-4
Coastal erosion	QL-5
Drought	QL-3, QL-6
Erosion	QL-2, QL-5
Extreme air pressure (high/low/gradient)	QL-4, QL-7
Fog/Mist	QL-6
Frazil ice	QL-1, QL-3, QL-6
Frost	QL-6, QL-7
Groundwater (too much or too little)	QL-6
High air temperature	QL-5, QL-6
High water temperature	QL-3, QL-5, QL-6
Ice barriers	QL-1, QL-2, QL-3
Ice cover	QL-1, QL-2
Low air temperature	QL-6
Low lake or river level	QL-3, QL-6
Low water temperature	QL-6
Sandstorm	QL-3
Seiche	QL-2

presents the excluded natural hazards related to climate change and the reasons for their exclusion. Ultimately,

Table 6. List of climate change-related natural disasters that could affect nuclear power plants (priority assessment targets).

Number	Natural Hazards Related to Climate Change
1	Biological events/Lake or river-borne material that plugs water intakes/organic material in water
2	External flooding
3	Extreme rain
4	Extreme winds and tornadoes/Hurricane/Typhoon/Strong winds (other than hurricane or tornado)
5	Forest Fire
6	Hail
7	Landslide
8	Lightning
9	High tide/Other extraordinary waves/Storm surge/Waves
10	River diversion
11	Salt storm
12	Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Sleet
13	Strong currents (underwater erosion)
14	Underwater landslide (impact on the soil; that is, not tsunami)

shows the final list of climate change-related natural hazards for priority evaluation.

4. Priorities of Natural Hazards Related to Climate Change

4.1. Priority Results

To determine the priorities of natural hazards related to climate change that affect nuclear power plants, 42 nuclear experts participated in a Delphi survey. The Delphi survey’s first and second rounds ran from 2 July to 25 July 2024. The statistical results from the first round of the Delphi survey were presented to the experts in the second round.

Table 7. Prioritization of natural hazards related to climate change using the first Delphi method.

Natural Hazards Related to Climate Change	Likelihood		Impact		Priority Score
	Point	Rank	Point	Rank	Point	Rank
Biological events/Lake or river-borne material that plugs water intakes/Organic material in water	137	3	142	4	477	3
External flooding	131	5	144	3	459	4
Extreme rain	163	1	146	2	583	1
Extreme winds and tornadoes/Hurricane/Typhoon/Strong winds (other than hurricane or tornado)	153	2	149	1	546	2
Forest fire	133	4	139	5	455	5
Hail	91	12	92	14	211	12
Landslide	115	7	124	7	347	7
Lightning	106	8	104	9	278	8
High tide/Other extraordinary waves/Storm surge/Waves	124	6	135	6	416	6
River diversion	95	9	101	10	241	10
Salt storm	94	10	106	8	254	9
Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Sleet	92	11	94	13	219	11
Strong currents (underwater erosion)	88	13	96	11	207	13
Underwater landslide (impact on the soil; that is, not tsunami)	86	14	96	11	200	14

and

Table 8. Prioritization of natural hazards related to climate change using the second Delphi method.

Natural Hazards Related to Climate Change	Likelihood		Impact		Priority Score
	Point	Rank	Point	Rank	Point	Rank
Biological events/Lake or river-borne material that plugs water intakes/Organic material in water	139	4	142	4	481	4
External flooding	140	3	154	2	519	3
Extreme rain	166	1	152	3	609	1
Extreme winds and tornadoes/Hurricane/Typhoon/Strong winds (other than hurricane or tornado)	161	2	157	1	606	2
Forest fire	126	5	130	5	395	5
Hail	86	10	85	14	175	12
Landslide	109	7	111	7	302	6
Lightning	96	8	97	8	228	7
High tide/Other extraordinary waves/Storm surge/Waves	118	6	124	6	361	8
River diversion	88	9	93	10	197	10
Salt storm	83	13	95	9	198	9
Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Sleet	84	11	87	12	175	12
Strong currents (underwater erosion)	84	11	89	11	178	11
Underwater landslide (impact on the soil; that is, not tsunami)	80	14	87	12	165	14

show the results of the Delphi survey’s likelihood, impact, and priority scores for both rounds. The likelihood and impact presented in Table 7 and Table 8 are the sum of scores evaluated by 42 experts. Additionally, the prioritization results for climate change-related natural hazards are shown in terms of likelihood, impact, and priority scores.

In the first round of the survey, “extreme rain” had the highest likelihood score of 163 points and the second highest impact (Impact) score of 146 points, resulting in the highest total priority score of 583 points. Meanwhile, “extreme winds/tornadoes/hurricanes/typhoons” came in second with a likelihood score of 153 points and first with an impact score of 149 points, for a total priority score of 546 points, placing it second overall. “Biological events” and “external flooding” ranked third and fifth, respectively, with likelihood scores of 137 points and 131 points. In terms of impact, they received 142 points (fourth place) and 144 points (third place), resulting in total priority score of 477 points (third place) and 459 points (fourth place), respectively. The average, quartiles, and median values from the first round of the Delphi survey were distributed to experts in the second round.

In the second round of the Delphi survey, extreme rain came in first with a likelihood score of 166 points and third with an impact score of 152 points, maintaining the highest total priority score of 609. Meanwhile, extreme winds/tornadoes/hurricanes/typhoons ranked second with a likelihood score of 161 points and first with an impact score of 157 points, resulting in a total priority score of 606 points and second place overall. External flooding ranked third with a likelihood score of 140 points and second with an impact score of 154 points, resulting in a total priority score of 519 points and third place. Biological events came in fourth place with a likelihood score of 139 points and an impact score of 142 points, resulting in a total priority score of 481 points. Extreme rain and extreme winds/tornadoes/hurricanes/typhoons consistently ranked as the most dangerous, and the top five items remained the same in both rounds. This consistency in expert opinions demonstrates that natural hazards related to climate change are recognized as significant risks to the operation and safety of nuclear power plants.

Table 9. Prioritization of natural hazards related to climate change by institute.

Natural Hazards Related to Climate Change	Research Institute University		Industry		Regulatory Institute
	Priority Score	Rank	Priority Score	Rank	Priority Score	Rank
Biological events/Lake or river-borne material that plugs water intakes/Organic material in water	218	4	187	4	76	4
External flooding	231	3	188	3	100	3
Extreme rain	276	1	221	2	112	2
Extreme winds and tornadoes/Hurricane/Typhoon/Strong winds (other than hurricane or tornado)	239	2	254	1	113	1
Forest fire	180	5	148	5	67	5
Hail	77	12	64	14	34	9
Landslide	139	7	121	6	42	7
Lightning	102	8	91	8	35	8
High tide/Other extraordinary waves/Storm surge/Waves	177	6	121	6	63	6
River diversion	96	9	68	11	33	10
Salt storm	85	10	85	9	28	11
Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Sleet	73	13	74	10	28	11
Strong currents (underwater erosion)	84	11	66	12	28	11
Underwater landslide (impact on the soil; that is, not tsunami)	71	14	66	12	28	11

compares natural hazards priorities related to climate change by schools/research institutions, industries, and regulatory institutes. Except for the ranking differences for “extreme rain” and “extreme winds/tornadoes/hurricanes/typhoons,” the survey findings show that each item had a general similarity across the institutions. This suggests that experts from various institutions have a consistent view of the severity of natural hazards related to climate change.

4.2. Verification and Discussion

The validity, consistency, and stability of expert opinions in prioritizing natural hazards associated with climate change were evaluated using CV and IQR.

Table 10. Results of the first Delphi survey.

Natural Hazards Related to Climate Change	Likelihood					Impact
	SD	Mean	Median	IQR	CV	SD	Mean	Median	IQR	CV
Biological events/Lake or river-borne material that plugs water intakes/Organic material in water	1.23	3.26	3	1.75	0.37	1.02	3.38	3.5	1	0.30
External flooding	1.15	3.11	3	2	0.37	0.97	3.42	4	1	0.28
Extreme rain	1.15	3.88	4	2	0.29	1.15	3.47	4	1	0.33
Extreme winds and tornadoes/Hurricane/Typhoon/Strong winds (other than hurricane or tornado)	0.86	3.64	4	1	0.23	0.90	3.54	4	1	0.25
Forest fire	1.02	3.16	3	2	0.32	1.01	3.30	3	1	0.30
Hail	0.94	2.16	2	2	0.43	0.90	2.19	2	1	0.41
Landslide	1.13	1.13	3	2	0.41	0.94	2.95	3	2	0.32
Lightning	0.87	0.87	2.5	1	0.34	0.85	2.47	3	1	0.34
High tide/Other extraordinary waves/Storm surge/Waves	1.19	2.95	3	2	0.40	1.08	3.21	3	2	0.33
River diversion	0.97	2.26	2	1	0.43	0.95	2.40	2	1	0.39
Salt storm	0.92	2.23	2	1.75	0.41	0.98	2.52	2	1	0.38
Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Sleet	0.98	2.19	2	1.75	0.44	0.86	2.23	2	1	0.38
Strong currents (underwater erosion)	0.83	2.09	2	1.75	0.40	0.88	2.28	2	1	0.38
Underwater landslide (impact on the soil; that is, not tsunami)	0.75	2.02	2	1.75	0.36	0.98	2.28	2	1	0.42

and

Table 11. Results of the second Delphi survey.

Natural Hazards Related to Climate Change	Likelihood					Impact
	SD	Mean	Median	IQR	CV	SD	Mean	Median	IQR	CV
Biological events/Lake or river-borne material that plugs water intakes/Organic material in water	0.77	3.30	3.5	1	0.23	0.65	3.38	3	1	0.19
External flooding	0.74	3.33	3	1	0.22	0.64	3.66	4	1	0.17
Extreme rain	0.65	3.95	4	0	0.16	0.57	3.61	4	1	0.15
Extreme winds and tornadoes/Hurricane/Typhoon/Strong winds (other than hurricane or tornado)	0.53	3.83	4	0	0.13	0.53	3.73	4	1	0.14
Forest fire	0.81	3	3	2	0.27	0.47	3.09	3	0	0.15
Hail	0.61	2.04	2	0	0.30	0.40	2.02	2	0	0.20
Landslide	0.72	2.59	2	1	0.27	0.75	2.64	3	1	0.28
Lightning	0.50	2.28	2	0.75	0.21	0.55	2.30	2	1	0.24
High tide/Other extraordinary waves/Storm surge/Waves	0.76	2.80	3	1	0.27	0.72	2.95	3	1	0.24
River diversion	0.52	2.09	2	0	0.25	0.55	2.21	2	0	0.25
Salt Storm	0.59	1.97	2	0	0.30	0.58	2.26	2	0	0.25
Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Sleet	0.43	2	2	0	0.21	0.40	2.07	2	0	0.19
Strong currents (underwater erosion)	0.37	2	2	0	0.18	0.44	2.11	2	0	0.21
Underwater landslide (impact on the soil; that is, not tsunami)	0.42	1.90	2	0	0.22	0.33	2.07	2	0	0.16

show the standard deviation, mean, median, IQR, and coefficient of variation (CV) for the impact and likelihood of these hazards in the first and second rounds of the Delphi survey. Figure 2 compares CV and IQR values between the first and second Delphi survey rounds, demonstrating increased stability and consensus in expert opinions. A CV value less than 0.5 typically indicates that expert opinions are stable [52]. When applying a five-point scale, an IQR of one or less suggests that respondents have reached a consensus [53]. The first Delphi survey results show that all items had CV values less than 0.5, indicating stable opinions [52]. The second Delphi survey resulted in a decrease in standard deviations, IQR, and CV values, indicating increased stability and validity in expert evaluations. The IQR and CV confirm that the prioritization of natural hazards associated with climate change was consistently based on expert opinions, thus enhancing the reliability of prioritization.

The BWS method validated the prioritization of natural hazards related to climate change. BWS evaluates the relative importance of alternatives by identifying the most and least important options, providing a simple and consistent response method [54,55]. The BWS survey involved 42 experts.

Table 12. Comparison of Delphi method results and BWS results.

Rank	Delphi Methods Result	BWS Results
1	Extreme rain	Extreme winds and tornadoes/Hurricane/Typhoon/Strong winds (other than hurricane or tornado)
2	Extreme winds and tornadoes/Hurricane/Typhoon/Strong winds (other than hurricane or tornado)	Extreme rain
3	External flooding	Biological events/Lake- or river-borne material plugging water intakes/Organic material in water
4	Biological events/Lake- or river-borne material plugging water intakes/Organic material in water	Forest fire
5	Forest fire	External flooding
6	High tide/Other extraordinary waves/Storm surge/Waves	Landslide
7	Landslide	High tide/Other extraordinary waves/Storm surge/Waves
8	Lightning	Lightning
9	Salt storm	Hail
10	River diversion	Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Snow
11	Strong currents (underwater erosion)	Strong currents (underwater erosion)
12	Hail	River diversion
12	Snow (including snowstorms and drifts)/Ice storm/Freezing rain/Snow	Underwater landslide (impact on the soil; that is, not tsunami)
14	Underwater landslide (impact on the soil; that is, not tsunami)	Salt storm

shows the results of comparing the priorities of natural hazards related to climate change using both the Delphi and BWS methods. Despite differences in rankings produced by the two methods, the top eight items remained consistent. This suggests that the Delphi and BWS methods complement each other in assessing the importance of natural hazards related to climate change, implying a common perception among experts.

This study collected expert opinions and identified typhoons, biological events, heavy rain, external flooding, and wildfire as the top-ranked hazards to Korean nuclear power plants. These top-ranked hazards, as confirmed by OPIS (Operational Performance Information System), have affected Korean nuclear power plants [11,45]. A journal that screened extreme natural hazards for Korean nuclear power plants evaluated internal flooding due to heavy rain, wind pressure due to strong winds, and extreme weather conditions as potential threats encountered at Korean nuclear power plant sites [56]. A Delphi-based study that analyzed the causes of large-scale power outages in the Korean transmission network identified strong winds/typhoons and heavy rain/floods as the top-ranked hazards [40]. Thus, historical records and previous studies show similar trends to the priorities derived in this study.

The impacts of climate change and the hazards faced by nuclear power plants can vary considerably across different regions due to their unique environmental and climatic conditions. For example, heat waves and elevated water temperatures have affected European nuclear power plants [11]. In contrast, hurricanes, floods, and periods of extreme cold have affected facilities in the United States [11]. This highlights the importance of tailoring risk assessments and adaptation strategies to specific regional contexts, rather than adopting a one-size-fits-all approach. The IPCC research in Figure 3 illustrates that projected climate impact indicators differ by region, including East Asia, Europe, and North America. In particular, average precipitation, river floods, fire weather, and typhoons are projected to increase in East Asia, showing a trend similar to the priority results of this study. These suggest that hazard prioritization may need to consider region-specific environmental and climatic conditions.

In Korea, periodic safety reviews and post-Fukushima stress tests have assessed the impacts of external hazards, including those associated with climate change. However, these assessments have relied on historical data and records. Due to climate change, relying on past data may be insufficient because historical records often fail to capture new hazards or shifts in their frequency and intensity. Therefore, safety assessments should integrate future climate scenarios, such as those based on SSPs or RCPs, to account for potential changes in the frequency and intensity of climate-related hazards. In addition, given the potential for future climate change, proactive and forward-looking assessments should be considered to reassess continually and, if necessary, supplement existing safeguards.

The results can be a reference for future risk assessments and response planning. The prioritization outcomes offer a practical foundation for strengthening risk-informed decision-making in the safety planning of nuclear power plants in Korea. They can also support future vulnerability assessments, the development of adaptation strategies, and the formulation of regulatory policies that account for evolving climate-related risks.

5. Conclusions

Climate change is driving increases in atmospheric temperature, sea surface temperature, and sea level, intensifying the frequency and severity of natural hazards such as typhoons, extreme rainfall, and forest fires. These climate-related hazards pose significant risks to nuclear power plants’ safe and reliable operation. Therefore, identifying and prioritizing these hazards is essential for effective risk management and safety planning.

This study identified climate change-related natural hazards affecting nuclear power plants in Korea by gathering expert opinions and evaluating the likelihood and impact of each hazard. The results showed that extreme rainfall, typhoons, external flooding, biological events, and forest fires were consistently classified as high-risk hazards. The rankings were generally consistent across experts from different sectors, including research institutes, industry, regulatory agencies, and academia. This consistency reinforces the credibility of the prioritization results and demonstrates a shared understanding of key risks across institutional boundaries.

The prioritization outcomes offer practical value for enhancing preparedness measures and supporting risk-informed decision-making. They can guide the efficient allocation of resources and serve as a reference for developing climate adaptation strategies for nuclear power plant operations. While the Delphi method is qualitative, consensus among expert opinions was confirmed using the coefficient of variation (CV), and applying the best–worst scaling (BWS) method provided cross-validation, thereby enhancing the robustness of the results.

This methodology relies on expert judgment and does not incorporate complementary quantitative modeling or scenario-based simulations. Future research can enhance analytical rigor and support a more evidence-based hazard prioritization by integrating probabilistic risk models and climate projection data. Moreover, since the findings reflect expert input and climate conditions specific to Korea, their applicability to other geographical contexts may be limited. When applying this methodology internationally, researchers should consider regional differences in hazard profiles, regulatory frameworks, and expert perspectives. For example, while Korea prioritizes extreme rainfall and typhoons, several nuclear power plants in Europe have shut down due to high ambient temperatures. These differences highlight the importance of accounting for regional climatic and environmental conditions in hazard prioritization. Future studies should adapt and apply this approach across regions through international collaboration to enable comparative analysis and improve the generalizability of the results.

Nuclear power plants must integrate climate risk into their safety and operational frameworks as climate change accelerates. This includes updating design bases to reflect projected hazards due to climate change, considering multi-hazard and cascading scenarios, and continuously improving hazard prediction models. Strengthening the predictive capabilities of climate impact assessments and adopting forward-looking policies will be essential for ensuring the long-term resilience of the nuclear sector under evolving environmental conditions.