Analysis on a database of ship accidents in port areas

In the last 15 years, the number of vessels in the world fleet has grown by around 53% and their gross tonnage has increased by 47%, with a consequent significant increment of marine accidents worldwide. Accident database are the basic resource for risk assessment methods to help decision-makers to enact strategies and undergo hazard and vulnerability mitigation measures. Understanding ship accidents distribution in terms of involved GT, typical age of the affected vessels, category of the ships, as well as distribution of underlying causes and consequences is the first necessary step to improve accident mitigation actions to be implemented for future assessments. In the present work, the results of an analysis on a database of vessel accidents in Mediterranean and worldwide port areas developed within the framework of the project ISY PORT (Integrated SYstem for navigation risk mitigation in PORTs) is herein presented. The distribution of accidents was analyzed in terms of relevant vessel characteristics i.e. Grosse Tonnage (GT), age at the time of the accident, ship's category, causality event, weather conditions and number of fatalities/injuries/lost at sea. The database can be used as a basis for maritime risk assessment methods and for calibration of real-time ship collision avoidance scenarios.


a b s t r a c t
In the last 15 years, the number of vessels in the world fleet has grown by around 53% and their gross tonnage has increased by 47%, with a consequent significant increment of marine accidents worldwide. Accident database are the basic resource for risk assessment methods to help decision-makers to enact strategies and undergo hazard and vulnerability mitigation measures. Understanding ship accidents distribution in terms of involved GT, typical age of the affected vessels, category of the ships, as well as distribution of underlying causes and consequences is the first necessary step to improve accident mitigation actions to be implemented for future assessments. In the present work, the results of an analysis on a database of vessel accidents in Mediterranean and worldwide port areas developed within the framework of the project ISY PORT (Integrated SYstem for navigation risk mitigation in PORTs) is herein presented. The distribution of accidents was analyzed in terms of relevant vessel characteristics i.e. Grosse Tonnage (GT), age at the time of the accident, ship's category, causality event, weather conditions and number of fatalities/injuries/lost at sea. The database can be used as a basis for maritime risk assessment methods and for calibration of real-time ship collision avoidance scenarios.
© 2023 The Author(s The data were collected from the SeaSearcher and IHS databases.These two databases were compared considering IMO Number, accident date and location of the accident (at sea or in port), in order to create a unified database with consistent data and as much reliable as possible. In the event of common accidents found within the system (meaning with the same IMO number and the same accident date), the location was compared. If the two locations in the databases were not matching, the incident was not included in the unified database for the analysis. The database used to conduct the analysis (after the comparison) comprises of 13,846 incidents worldwide, of which 2,799 occurred in port area. Data source location: Mediterranean port areas within SeaSearcher and IHS database of accidents

Value of the Data
• The dataset provide insights on ship collision accidents in the Mediterranean and worldwide, obtained from the analysis of the IHS and SeaSearcher merged datasets. • The provided data are intended to support calibration of ship collision risk assessment methodologies (e.g. Bayesian networks, neural networks, swarm intelligence), with a specific focus on waterways and port areas. • The dataset addresses different type of accidents (heavy weather, fire…) to support risk assessment due to different type of threats (environmental, economic, climatic).

Objective
Understanding ship accidents causes and consequences is the first necessary step to improve accident mitigation strategies [ 1 ]. This can help to identify patterns and trends in accidents, which can be used to inform prevention measures and response plans. The aim of this dataset is to provide a statistical analysis of vessels accidents that have occurred in port areas. The dataset investigates the causes and consequences of these accidents, with a focus on worldwide and Mediterranean trends. By analyzing these data, it is possible to gain insights into the factors that contribute to accidents and evaluate their impacts.

Data Description
The dataset [ 2 ] includes the results of an analysis on a database of ship collision accidents in port areas occurred worldwide, from 1990 to Q1 of 2021. The analysis was also conducted only on Mediterranean port areas, in order to provide a comparison with worldwide statistics. The marine accidents were classified based on the following accident characteristics:  The results obtained with the segmentations reported above are reported in Tables 1-6 .       Table 6 Number of fatalities, injured and lost at sea caused by accidents in the port areas.

Mediterranean Worldwide
Fatalities Accidents without fatalities 2753 619 Accidents with at least 1 fatality 46 15 Accidents with more than 2 [1] fatalities 24 6 Maximum number of fatalities for a single accident 10 4 Total number of fatalities 108 24 Injured Accidents without injured 2745 614 Accidents with at least 1 injured 54 20 Accidents with at more than 3 [1] injured 28 12 Maximum number of injured for 1 accident 55 55 Total number of injured 337 131 Lost at sea Accidents without lost at sea 2795 633 Accidents with at least 1 lost at sea 4 1 Accidents with at more than 3 [1] lost at sea 2 1 Maximum number of lost at sea for 1 accident 15 3 Total number of lost at sea 22 3 Accidents with human life consequences [2] 31 95 [1] The value is defined as the median value [2] The value is the number of the accident with human life consequences. In some cases, the accident has caused fatalities, injured and lost at sea at the same time. This value is not the sum of accident with at least 1 fatalities/injured/lost at sea.

Experimental Design, Materials and Methods
Thedatabase used in the analysis was created starting from the following two databases: • SeaSearcher [ 3 ], which recorded 79,592 vessel accidents between 1967 and Q2 of 2021; • IHS [ 4 ], which recorded 23,897 vessel accidents between 1990 and Q2 of 2021 These two databases were compared considering IMO Number, accident date and location of the accident (at sea or in port), in order to create a unified database with consistent data and as much reliable as possible. In the event of common accidents found within the system (meaning with the same IMO number and the same accident date), the location was compared. If the two locations in the databases were not matching, the incident was not included in the unified database for the analysis. The database used to conduct the analysis (after the comparison) comprises of 13,846 incidents worldwide, of which 2,799 occurred in port area.
The focus of the overview and analysis of the marine incidents is the Mediterranean area, but the analyses have also been conducted on the worldwide incidents database, for comparison and are presented in the article.
The incidents occurred in the Mediterranean ports, considering ship with an IMO number and with a gross tonnage (GT) above 500 GT from 1990 to Q2 of 2021 are 634, and the analyses presented in this article are based on this dataset. The incidents occurred in the ports worldwide, when considering the same constrains of vessels GT and IMO number as in the Mediterranean analysis, are 2,799 from 1990 to Q2 of 2021.
Each accident includes the following information: IMO, Accident date, Name of the vessels, Flag, Deadweight (DWT), Grosse Tonnage (GT), Age of the vessel, Class, Vessel type, Cause of accident, Loss of vessel/cargo, Pollution, Number of fatalities, Number of injured, number of lost at sea, Port of accident, location of accident, weather at time of accident.

Ethics Statements
The present work meets the publisher ethical requirements ( https://www.elsevier.com/ authors/journal-authors/policies-and-ethics ), and does not involve studies with animals and humans.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data Availability
Analysis on a database of ship accidents in port areas (Original data) (Mendeley Data).