1115

1 INTRODUCTION

In recent years, there is a concern that Natech (Natural-

hazard triggered technological accidents) risk may

increase the damage not only from primary disasters

such as frequent typhoons and tsunamis caused by

major earthquakes, but also from secondary disasters

caused by such natural disasters. Therefore, it is

necessary to conduct research that comprehensively

considers disaster prevention and mitigation against

such risks. Although various factors can be taken into

account for this Natech risk, one of its origins can be

attributed to the evacuation of vessels if we focus on

situations where large industrial areas and important

ports are located. This is because vessels are

responsible for transporting goods to industrial areas

and port facilities, and a large number of vessels

navigate in the waters surrounding these areas. In the

event of a typhoon or tsunami, these vessels evacuate

to ensure their own safety, but the evacuation depends

on the judgment of individual vessels, and there is

currently no controlled evacuation in the sea area.

Therefore, this research aims to create measures to

reduce Natech risk in the maritime field by establishing

a safe and optimal evacuation guidance method for

ships in the event of a natural disaster and identifying

potential risks that may cause secondary disasters.

This paper presents a time-series visualization

analysis of the evacuation behaviour of a group of

vessels in Osaka Bay when Typhoon No. 14 (Named:

NANMADOL) of 2022 hit the Kansai region of Japan

on September 18, 2022, using actual navigation record

data. The existence of potential risks in ship evacuation

was also clarified by extracting “crossing points,”

which are dangerous situations during ship

navigation, calculating the risk of collision between

vessels, and identifying the sea area.

Clarification of Potential Risks in the Evacuation

Process of Ships to Reduce the Risk of Maritime -

NATECH

H. Makino, A. Tokuyama, G. Kusuma, N. Mohammadi & S. Fujimoto

Kobe University, Kobe, Japan

ABSTRACT: This study aims to "reduce the risk of Natech (industrial disasters caused by natural disasters) in the

maritime sector" and aims to develop a safe and optimal evacuation guidance method for ships that can be a

source of risk, especially in coastal areas where large-scale industrial zones and important ports are located. In

this paper, a time-series visualization analysis was performed based on data transmitted from automatic

identification systems to grasp the overall picture of ship evacuation behaviour for ships in Osaka Bay when

Typhoon No. 14 of 2022, which made landfall in Japan on September 18, 2022, approached Kansai. In addition,

the "Crossing Point," which is a dangerous situation during navigation that can also be a factor in the collision

risk of ships in usual situation, was calculated, and the potential risk factors in the process of evacuation

behaviour outside the port were also analyzed.

http://www.transnav.eu

the International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 19

Number 4

December 2025

DOI: 10.12716/1001.19.04.08

1116

2 OUTLINE OF TYPHOON NO. 14 IN 2022 AND

THE SURVEYED SEA AREA

2.1 About Typhoon No. 14 of 2022[1]

According to the Japan Meteorological Agency,

Typhoon No. 14, which originated near the Ogasawara

Islands at 3:00 a.m. on September 14, 2022, moved

north-westward south of Japan, and by 3:00 a.m. on

September 17, it had developed into a large typhoon of

extreme strength. The typhoon made landfall in

Kagoshima Prefecture at 19:00 on September 18 with

large size and very strong force, and traversed Kyushu

Island until the morning of September 19. After that, it

changed its course to the east and moved over the Sea

of Japan from the Chugoku region, making landfall

again in Niigata Prefecture after 4:00 a.m. on the 20th,

and then changed to an extratropical cyclone east of

Japan at 9:00 a.m. on the 20th (Fig. 1). The approach,

passage, and landfall of the typhoon caused stormy

winds over a wide area from western Japan to northern

Japan, centering on Kyushu, and caused severe

droughts and heavy gales at sea. Storm surges

exceeding the warning criteria were observed in some

areas. Wind speeds reached a maximum of 50.4 meters

per second in Saiki City, Oita Prefecture, and many

locations in the Kyushu, Chugoku, and Kinki regions

recorded the highest maximum wind speeds ever

recorded. Heavy rainfall occurred in the Kyushu and

Shikoku regions due to the long duration of rain clouds

around the typhoon and its main body, and total

precipitation from September 17 to 20 was about twice

the normal for the month of September at several

locations. The heavy rainfall caused landslides,

swollen and overflowing rivers, inundation of low-

lying areas, wind storms, and storm surges mainly in

the Kyushu and Shikoku regions.

Figure 1. 2022 Typhoon No. 14 path map (JMA

announcement)

2.2 Outline of the surveyed sea area

The survey area introduced in this paper is Osaka Bay,

and Figure 2 shows Osaka Bay, each port, and each

harbour area. The Hanshin Industrial Zone, one of

Japan's three major industrial zones, is located within

Osaka Bay. Heavy industries, particularly

petrochemicals, are concentrated in the eastern part of

the bay, which also covers a large area. In addition,

there are several international strategic ports, so many

large ships are constantly sailing through the bay. In

addition, there are many fishing grounds, making it a

congested sea area with a wide variety of ships

coexisting. Furthermore, on September 4, 2018,

Typhoon No.21 (Jebi) caused strong winds and high

tides in the bay, damaging port facilities and

disrupting industrial and economic activity in various

ports, including the suspension of container terminals

due to gantry cranes not functioning due to power

supply being submerged. In addition, the disaster

caused extensive damage in various locations, such as

ships running aground and containers spilling out,

disrupting ship navigation. In particular, at Kansai

International Airport, in addition to damage such as

the runway not functioning due to flooding and power

outages at the passenger terminal, a ship that had been

anchored for evacuation dragged its anchor due to

strong winds and collided with the bridge connecting

the airport to the opposite shore, resulting in the

closure of the airport, leaving approximately 7,800

people inside the airport stranded for five days. In

order to prevent the recurrence of large-scale disasters

caused by such marine accidents, the 5th Regional

Coast Guard Headquarters, which has jurisdiction

over Osaka Bay, has issued a “recommendation to

refrain from anchoring” for ships of 100 gross tons or

more within 3 nautical miles from each of Kansai

International Airport, Kobe Airport, and Sakai-

Semboku Port Pier (shaded area in Figure 2) as a

navigation rule to prevent evacuation vessels from

dragging anchor during rough weather. In addition, if

the typhoon is expected to reach the target sea area

with a central wind speed of 40 m/s or more, (1) car

carriers, container ships, gas tankers, tankers with a

total length of 160 m or more, (2) passenger ships and

ferry ships with a total length of 200 m or more. A

system for out-of-bay evacuation recommendations

and orders was established by the amendment to the

Maritime Traffic Safety Act and came into effect on July

1, 2021, targeting ships such as ships, cargo ships, (3)

dangerous goods ships with a gross tonnage of 50,000

tons or more, and (4) liquefied gas ships with a gross

tonnage of 25,000 tons or more. In Osaka Bay, this

evacuation recommendation outside the bay was first

issued during Typhoon No. 14 in 2022. In the next

chapter, we will introduce the results of an analysis of

the evacuation situation of ships under such

circumstances.

Figure 2. Major ports in Osaka Bay and areas where

anchoring is not recommended

1117

3 ANALYSIS OF ACTUAL EVACUATION

BEHAVIOUR OF SHIPS AND IDENTIFICATION

OF POTENTIAL RISKS

3.1 Time series visualization analysis using actual

navigation record data

In this study, we used data transmitted from the

Automatic Identification System (AIS) installed on

ships as the actual navigation record data used in the

evacuation behaviour analysis. AIS is a type of

broadcast-type automatic dependent surveillance, and

is required by the International Maritime Organization

(IMO) to be installed on target ships (passenger ships

and ships of 300 gross tons or more on international

voyages, and ships of 500 gross tons or more on non-

international voyages) [3]. The transmitted data can be

broadly divided into 1) static information (IMO

number, call number, ship name, captain, overall

width, ship type, etc.), 2) dynamic information (time,

ship position, ship speed, relative position, heading,

navigation status, etc.), and 3) navigation-related

information (draft, destination, estimated time of

arrival, cargo, etc.). Figure 3 shows a map of the tracks

of ships in Osaka Bay during normal times using this

AIS data. The yellow dashed lines in this figure show

the ship tracks. The main routes for maritime traffic

within Osaka Bay are: (1) the route connecting A-B in

this figure from the Seto Inland Sea through the Akashi

Strait to Osaka Port, Kobe Port, and Sakai-Senboku

Port (or vice versa), and (2) the route connecting B-C in

the figure passing from the Pacific Ocean through

Tomogashima Channel and entering the bay to Osaka

Port, Kobe Port, and Sakai-Senboku Port. (or the

reverse route), (3) A route connecting A-C in this figure

from the Seto Inland Sea, passing through the Akashi

Strait, navigating within the bay, passing through the

Tomogashima Suido to the Pacific Ocean (or the

reverse route), and it can be confirmed from the track

map that there are especially many ships navigating

the route connecting A-B as explained in (1).

Figure 3. Map of ship tracks in Osaka Bay

Next, a time-series visualization analysis was

conducted on the behaviour of ships in Osaka Bay

using AIS data from September 17th to September 20th,

2022, the day Typhoon No. 14 of 2022 came closest to

Osaka Bay, focusing on that day and the period before

and after (September 17th to September 20th). The

results are shown in Figure 4. Each black triangle in the

figure indicates a ship and the accompanying numbers

indicate the ship's overall length. Figure 4 (a) shows the

status of ships in Osaka Bay at 15:00 on September 17th,

2022, when a "Bay Evacuation Advisory" was issued by

the Chief of the 5th Regional Coast Guard

Headquarters, which has jurisdiction over Osaka Bay.

At this point, it was confirmed that many ships had

already taken refuge at anchor in the southwest and

northeast of Kansai International Airport. Regarding

ships within the bay that were recommended to

evacuate outside the bay, 12 ships were identified

within Osaka Bay. Among them, it was confirmed that

there were 10 ships anchored within each port area and

one ship outside the port area. Figure 4(b) shows the

status of ships in the bay at 0:00 on September 18, 9

hours after the "Out of Bay Evacuation Advisory" was

issued. The number of ships anchored for evacuation

had increased since Figure 4 (a), and it was confirmed

that the area northeast of Kansai International Airport

was in a state of congestion. Ships anchored for

evacuation were also confirmed within each port

district. Furthermore, five large ships were confirmed

to be anchored within the berths, and three outside the

port districts. Figure 4 (c) shows the status of ships in

the bay at midnight on September 19th, 36 hours after

the issuance of the "Evacuation Advisory for Outside

the Bay." At this point, it was confirmed that port

district or outside Osaka Bay (there were six ships

subject to the evacuation advisory that had anchored in

the bay). However, it was confirmed that the ships that

had anchored in the bay were overcrowded, as they

were anchored in limited waters outside the port

districts and outside the areas where anchoring was

recommended to be avoided. In Osaka Bay, a

maximum instantaneous wind speed of 25 m/s was

observed at the Japan Meteorological Agency's Kansai

Airport Island Observatory at around 8:20 p.m. on the

same day, but the wind gradually weakened thereafter

as the typhoon passed. Figure 4 (d) shows the status of

ships in Osaka Bay at 0:00 on September 20th, after the

typhoon had passed through the bay. At this point,

several ships were observed drifting and taking refuge

in the deep waters of the central-western part of Osaka

Bay. It was also confirmed that the number of ships

anchored for evacuation had further increased. The

"advisory to evacuate outside the bay" was lifted at 6:00

the same day. Figure 4 (e) shows the status of ships in

the bay at 7:00, one hour later. At this point, many of

the ships that had anchored for evacuation within the

bay were observed to have completed their evacuation

actions and were sailing out of the bay.

(a) Status of ships in Osaka Bay at 15:00 on September 17th,

2022

1118

(b) Status of ships in the bay at 0:00 on September 18th, 2022

2022

(d) Status of ships in Osaka Bay at 0:00 on September 20t, 2022

(e) Status of ships in the bay at 7:00 on September 20t, 2022

Figure 4.Transition in ship evacuation behaviour in Osaka

Bay

3.2 Understanding potential risks in ship evacuation

In order for ships to evacuate more safely, it is

necessary to avoid any maritime accidents such as

collisions during the evacuation process. Especially

during an emergency evacuation, even experienced

ship operators may find it difficult to judge the

situation appropriately and may not be able to perform

the same actions as normal ship operations [4].

Therefore, in this paper, we extracted the "Crossing

Point", which is the intersection of the course with

other nearby ships, which is a dangerous situation

during navigation that is a factor in the risk of ship

collision during normal times, and analyzed the

potential risk factors in the process of evacuation action

outside the port.

This Crossing Point indicates a dangerous situation

where there is a possibility of a collision because the

vessel that is obliged to give way has not taken any

action to give way, even though the ships are in a

position to avoid collision. The procedure for

extracting crossing points is as follows. First, it is

determined whether the ship is underway based on

dynamic data contained in AIS data. The determining

factor for this judgment is whether the ship's speed is 2

knots or more and continues for 30 seconds or more.

Next, for each extracted ship underway, the predicted

navigation line 5 minutes later from the latest ship

position was calculated, and any intersections with

other predicted navigation lines that continued for

more than 1 minute were extracted as crossing points.

The reasons for extracting crossing points using the

predicted navigation line for 5 minutes at this ship

speed are as follows. If a ship with a ship speed of 12

knots sails for 5 minutes, it will travel a distance of 1

nautical mile, and depending on the positional

relationship, there will be two ships within a range of

2 nautical miles at the farthest distance, considering

other ships around. At that point, it can be determined

that some time has passed since the two vessels were

in an arranged relationship, and that they are not

taking any action to avoid danger, so the closer the

distance between the two vessels becomes, the more

abrupt a change of course is required to avoid the

danger. In addition, such a situation could create a

dangerous situation in the surrounding sea area, as the

vessel suddenly forms an arranged relationship with

other ships that do not have an arranged relationship.

In other words, the crossing point extracted this time is

an extremely dangerous situation that poses danger

not only to the two ships that created the situation, but

also to other ships sailing in the vicinity. Regarding this

Crossing Point extraction, we analyzed the period from

15:00 on September 17, 2022, when the "Out of Bay

Evacuation Advisory" was issued, when many

evacuation vessels were navigating, to 23:59 on the

same day. Figure 5 shows the crossing points

calculated at 15:02:56, about three minutes after the

issuance of the "Evacuation Advisory Outside the Bay."

The yellow triangles in the figure indicate ships that

were in Osaka Bay. Also, the upper number indicates

the total length of the ship, and the lower number

indicates the ship's speed. In addition, the straight line

extending from the tip of each triangle represents the

predicted 5-minute navigation line. Furthermore, the

red stars in the figure indicate the crossing points.

1119

Figure 5. The crossing points calculated at 15:03 on

September 17

Figure 6 shows the results of a detailed analysis of

the movements of both ships at this crossing point. This

Crossing Point was caused by an arrangement between

a 143m long cargo ship (green triangle in the diagram)

that departed from Kobe Port and a 56m domestic

tanker (red triangle in the diagram) that sailed

westward from Hanshin Port's Amagasaki. Figure 6 (a)

shows the situation when the crossing point was first

calculated. At this point, the distance from each ship to

the crossing point was about 0.8 miles. In this situation,

the ship that must give way is the ship indicated by the

red triangle. However, at this stage, no avoiding action

was taken, and three minutes later, as shown in Figure

6 (b), both ships began sudden manoeuvring to avoid

a collision. At this point, the distance between the two

ships and the crossing point was 0.45 miles for the

vessel shown in red and 0.4 miles for the ship shown in

green. Figure 6 (c) shows the situation one minute later,

when both ships took action to avoid a collision and the

crossing point disappeared. Both ships then attempted

to correct their course and return to their original

course, but this resulted in both ships creating a

Crossing Point with another ship. This situation is

shown in Figure 6 (d). It is predicted that these two

secondary crossing points would not have occurred if

the ship had taken early avoidance action when the

crossing point occurred as shown in Figure 6(a). In this

way, it was confirmed that one dangerous event has

the possibility of leading to a later, larger danger. The

first crossing point occurred near the entrance to the

main sea route of Kobe Port, which has one of the

largest container terminals in Osaka Bay, and if a

marine accident occurred at this point, it is possible

that entry and exit would be blocked, affecting the

evacuation of many ships. It was also revealed that the

ship that was at risk at this time was a high-speed

passenger ship, and that even more serious damage

was expected. Figure 7 shows a map of the crossing

points calculated as above that occurred during the

target period. It was confirmed that multiple crossing

points have occurred in the sea areas indicated by A to

D in the figure. Sea area A is near the main sea route

entrance to the main container terminal of Kobe Port,

as explained above. Sea area B is near the entrance/exit

of the main shipping route to Osaka Port, and area C is

near the main shipping route entrance/exit to the large

passenger ship terminal at Kobe Port. Sea area D is the

eastern entrance to the navigational route of Akashi

straight, which connects Osaka Bay and the Seto Inland

Sea. Both points are very busy, with many ships

passing through. If a marine accident occurs even in

one place in such a sea area, it will have a huge impact

on maritime traffic within Osaka Bay even in normal

times, but during evacuation situations such as this one

during a natural disaster, it is a risk factor that must be

prevented, as it not only obstructs the safe evacuation

of ships in the bay, but can also be a factor that directly

leads to secondary disasters. The results of this analysis

have identified these potential risks in the bay and

identified the process by which they occur, so these

results must be utilized for the safe and smooth

evacuation of ships in the bay in the event of a natural

disaster.

(a) Situation when the crossing point was first calculated.

(b) Status of both ships 3 minutes after Crossing point was

calculated

(d) Emergence of new Crossing points

Figure 6. Analysis results of Cross Point occurrence

1120

Figure 7. Occurrence of multiple cross points

4 CONCLUSION

In this paper, for the purpose of disaster prevention

and mitigation in ports in the event of a super-large

natural disaster, we focused on evacuation behaviour

analysis of a group of ships responsible for maritime

transportation, which is the main function of a port,

and introduced the results of understanding the details

of the evacuation situation of ships through a ship

evacuation behaviour analysis using actual navigation

data of a group of ships when Typhoon No. 14 of 2020

passed through. It was also confirmed that the

occurrence of the Crossing Point was ascertained and

that its continued existence included secondary

dangers to other ships navigating around it. The results

of this investigation showed that there were no

reported cases of serious accidents resulting in loss of

life as a result of the evacuation actions taken by each

vessel, and as a result, it can be said that there were no

major problems with the evacuation response of most

vessels. However, these are all cases in which the

evacuation was possible within the expected range of

each ship operator and those involved in the operation,

and if an unexpected situation such as dragging anchor

or drifting occurs after berthing, it could not only be

dangerous for the ship itself, but could easily cause

harm to other ships and damage to nearby port

facilities. Therefore, in the future, it will be necessary to

clarify these situations and consider measures for

vessel evacuation that can ensure not only the safety of

one's own vessel, but also the safety of the entire sea

area.

REFERENCES

[1] Japan Meteorological Agency, Weather Cases Causing

Disasters,

https://www.data.jma.go.jp/stats/data/bosai/report/inde

x.html (Last accessed on April 8, 2025)

[2] 5th Regional Coast Guard Headquarters, Navigation

Rules for Preventing Running Anchor Marine Accidents,

https://www.city.kobe.lg.jp/documents/13621/20210915s

oubyoukainannboushi.pdf (Last accessed on April 8,

2025)

[3] SOLAS：International Convention for the Safety of Life at

Sea. London, International Maritime Organization (2012)

[4] Japan Marine Casualty Prevention Association: Interim

report on research study on prevention of disasters

caused by large tankers, 1967