Analysis of Accidents of Fishing Vessels Caused by Human Elements in Korean Sea Area

Abstract

With an estimated 32,000 annual fatalities, fishing vessel accidents are 100-fold deadlier than those involving merchant ships. Despite ongoing safety training, accident rates remain high. Since most fishing vessel accidents occur in small fishing vessels (<12 m) and are primarily attributable to human elements, this study focuses on small fishing vessel accidents where the human element is the primary cause, exploring preventive measures for major accident types and performing a type-specific risk assessment. First, we performed a quantitative analysis of frequently occurring accidents and the indirect factors influencing the human element using maritime accident statistics and surveys, respectively. Next, we employed the fault tree analysis technique proposed by the International Maritime Organization in its Formal Safety Assessment to quantitatively assess the rate of accidents caused by the human element attributable to various indirect factors. The primary indirect factors most significantly impacting the human element were ship factors (22.8%), people factors (18.9%), and organization on board (17.4%). Secondary factors included personal negligence (14.1%), aging equipment and poor maintenance (10.3%), and harsh natural conditions such as rough waves (9.6%). Eliminating the top three secondary indirect factors reduced accidents due to the human element by 15.4% (64.5−49.1%).

1. Introduction

An estimated 39 million fishermen are engaged in capture fisheries on 4.56 million fishing vessels worldwide [1]. Fishing is one of the most dangerous occupations globally [2], with annual fatalities reported to exceed 32,000 [1], 100 times the death toll on merchant vessels [3]. This disparity is further highlighted when comparing the international safety regulations between merchant and fishing vessels. Merchant ships are subject to a comprehensive set of treaties, including SOLAS 1974 (safety), STCW 1978 (training), LL 1996 (stability), COLREGs 1972 (collisions), MLC 2006 (labor standards), and MARPOL 1973/78 (environmental protection). In contrast, fishing vessels are governed by only three treaties: STCW-F 1995 (training), C188 2007 (labor standards), and PSMA 2009 (environmental protection), two of which came into effect only after 2000.

The International Maritime Organization (IMO), in cooperation with the Food and Agriculture Organization and the International Labor Organization, is actively working to improve fishing vessel safety and protect fishermen’s lives. Recently, efforts have intensified to bring the Cape Town Agreement into force, with cooperation from member states and non-governmental organizations [3].

Even more concerning is that these treaties apply only to vessels exceeding 24 m in length, leaving smaller vessels entirely unregulated. Paradoxically, 81% of all fishing vessels fall under 12 m, and they experience the highest incidence of accidents and injuries [1]. These small-scale fishing vessels, often operated by precarious businesses or individuals, typically lack adequate navigation and communication equipment, and proper safety training for captains and fishermen is frequently absent [1]. The exclusion of these smaller vessels from international safety regulations is a significant obstacle to reducing fatalities.

Building on this background, numerous studies have investigated maritime accidents on fishing vessels, analyzing their types, occurrence patterns, and causes with the goal of reducing fatalities. Initial research employed quantitative analyses of statistical data from accredited organizations to identify frequently occurring accident types and causes, focusing on high-risk industries and their contributing factors [4,5]. One study quantitatively analyzed fatal fishing vessel accidents using official data and emphasized the need for preventative measures [6]. Another study explored the impact of preventive measures on accident rates for fatal fishing vessel accidents [7]. Additionally, researchers have proposed various probability models to estimate accident probabilities and suggest preventive measures based on statistical analyses of fishing vessel accidents [8,9,10,11,12].

Human error is widely recognized as the primary cause of ship accidents [4,13,14,15,16], with mechanical failures due to aging machinery and equipment [11] and environmental factors such as worsening maritime weather conditions also playing significant roles [17,18,19].

However, most studies focus solely on direct accident causes and their outcomes, leaving the underlying indirect factors uninvestigated. For example, if a fisherman gets his finger caught in equipment (direct cause), it is difficult to determine if negligence or aging equipment is the root cause (indirect factor) without directly questioning the fisherman.

Given the difficulties associated with going through such a process for every accident, most research results converge on simple direct causes like human error or mechanical failure. Consequently, safety trainings offered to fishermen have been limited to superficial preventive measures, which might be one of the reasons for the persistently high accident rate on fishing vessels.

In this study, based on the hypothesis that analyzing which indirect factors have a greater impact, even for accidents caused by the same direct cause, could enable selective prevention, the probability of occurrence of indirect factors was quantitatively analyzed and their risks were assessed. The analysis procedure followed the Formal Safety Assessment (FSA) methodology presented by the IMO [20]. FSA is a methodology aimed at scientifically analyzing accident causes before marine accidents occur to prevent them, consisting of five stages: Identification of Hazards, Risk Assessment, Risk Control Options, Cost–Benefit Assessment, and Decision Making. However, since this methodology was developed for large merchant ships, this study was only carried out up to the Risk Control Options stage applicable to fishing vessels, and the process is as follows:

• Considering that most accidents on fishing vessels occur during fishing operations [21], we quantitatively analyzed the major accident patterns associated with the human element using maritime accident statistics.
• Expert surveys were conducted to identify the indirect factors influencing the human element. These factors were categorized according to the “Guidelines for the Investigation of Human Factors in Marine Casualties and Incidents” proposed in the IMO resolution (A/Res.884) [22].
• Using marine accident statistics and survey data to identify the types of accidents and associated indirect factors, this study quantitatively analyzed the probability of marine accidents caused by these indirect factors using the Fault Tree Analysis (FTA) technique presented in the IMO Formal Safety Assessment (FSA). Although the FSA presents various analytical techniques such as Fault Tree Analysis (FTA), Event Tree Analysis (ETA), Failure Mode and Effect Analysis (FEMA), and Hazard and Operability Studies (HAZOP), this study utilized the FTA technique. It was determined that FTA could systematically and quantitatively analyze the contributing indirect (basic) factors of accidents using a Top-Down approach and AND/OR logic gates.
• Using the results of the FTA, a risk assessment was conducted on the indirect factors that contribute the most to accident occurrences.

2. Materials and Methods

2.1. Target Vessel Selection

Table 1. Global distribution of fishing vessels registered in major countries by continent.

	Length	<12 m	12–24 m	>24 m		Length	<12 m	12–24 m	>24 m
Country					Country
Africa					Asia
Angola		4694	186	256	Bangladesh		32,859	33	205
Benin		578	10	17	Cambodia		85,724
Malawi		2493			Republic of Korea		54,832	8866	1289
Mauritius		3605	33	6	Lebanon		1852	90
Senegal		8844	4656	152	Myanmar		13,141	3806	1141
Sudan		1120		60	Oman		23,678	1400	128
Tunisia		5469	1198	303	Sri Lanka		28,625	2455	24
Total		26,803	6083	794	Taiwan Province of China		14,514	6140	800
					Total		255,225	22,790	3587
America					Europe
Bahamas		1220	45	3	Iceland		1656	168	168
Chile		10,545	1902	105	Norway		4763	781	313
Guatemala		79	30	2	Poland		656	112	49
Guyana		712	806		Total		7075	1061	530
Mexico		61,294	1690	246	Oceania
Saint Lucia		482			New Caledonia		752	16	3
Suriname		418	587	63	New Zealand		571	415	69
Total		74,750	5060	419	Vanuatu		191	30	66
					Total		1514	461	138

shows the global distribution of powered fishing vessels in major countries where fishing operations take place, categorized by continent [23].

This table indicates that countries adjacent to seas or rivers are actively engaged in vessel fishing, with vessels under 12 m in length (length overall, LOA) accounting for the highest proportion at 89.9% (Figure 1).

By continent, the distribution is highest in Asia (69.9%), followed by the Americas (20.5%), Africa (7.3%), Europe (1.9%), and Oceania (0.4%), with Asia accounting for a markedly higher proportion compared to other continents (Figure 2).

By country, Cambodia ranks first (23.5%), followed by Mexico (16.8%) and Korea (15.0%) (Figure 3).

In this study, based on the number of registered fishing vessels by continent and country, Korea in Asia was selected as the target country, and powered fishing vessels under 12 m LOA were chosen as the target vessels.

2.2. Fishing Vessel Accident Statistics (Insurance Payment Data)

In general, the economic and environmental damage caused by fishing vessel accidents is less severe compared to that of merchant ships. Additionally, personal accidents on fishing vessels often go unreported or are excluded from official statistics [24]. Only major accidents involving collisions or capsizing that result in multiple casualties tend to be formally investigated [25]. Consequently, accurately estimating the number of accidents is challenging, and estimates are typically derived from compiling statistical data from various countries.

In this study, to obtain statistical data on the number and types of accidents attributable to the human element, insurance coverage records for 1093 accidents involving fishing vessels under 12 m registered in Korea over the past 5 years (2018–2022) were analyzed [26]. These data serve as a valuable source of information when quantitative verification of accident types and counts is not otherwise possible. It includes details such as the nature of the accident and the amount of insurance payment received based on severity, allowing for the quantitative analysis of accidents not recorded in formal adjudications.

Despite potential limitations arising from using data from a single country, Figure 3 highlights Korea’s position as the third-highest holder of fishing vessels under 12 m LOA, following Cambodia and Mexico. This suggests that the analysis, while based on a single country, can offer valuable insights globally applicable to vessels in the same length category.

2.3. Selection of Indirect Factors

2.3.1. Updated Domino Theory

To determine the indirect factors that contribute to direct causes, it is necessary to examine the disaster theory followed in this study as its theoretical framework. Initially, H.W. Heinrich introduced the domino theory, which states that accidents occur as a result of a chain of events consisting of five elements, the social environment, fault of a person, unsafe act and/or unsafe condition, accident, and injury, arguing that removing unsafe acts and conditions prevents disasters. In contrast, Frank E. Bird Jr. proposed an updated domino theory consisting of lack of control by management, basic causes, immediate causes, accident, and injury, arguing that it is more important to eliminate basic causes (incident, near accident, nearly hurt) underlying immediate/direct causes [27].

This study utilized Bird’s theory for the chain of events leading to accidents associated with the human element. However, to align with the study’s objectives, “lack of control by management” was excluded, and basic causes were classified into six factors as proposed by the IMO. The chain of events involving various factors and the human element is depicted in Figure 4.

2.3.2. Selection of Indirect Factor Items

Traditionally, education and training have been emphasized as preventive measures for the human element in fishing vessel accidents. However, as shown in Figure 5, the IMO considers the human element to be complexly related to various factors, not just human factor [22].

For this reason, Resolution A.884(21) stipulates that factors affecting the human element in maritime accident investigations should be classified into six categories: people factors, organization on board, working and living conditions, ship factors, shore-side management, and external influences and environment. The major items and content related to these factors are listed below:

• People factors:
  • Ability, skills, knowledge (outcome of training and experience);
  • Personality (mental condition and emotional state);
  • Physical condition (medical fitness, drugs and alcohol, and fatigue);
  • Actual behavior at time of accident/occurrence.
• Organization on board:
  • Composition of the crew (nationality/competence);
  • Workload/complexity of tasks;
  • Teamwork, including resource management.
• Working and living conditions:
  • Level of automation;
  • Ergonomic design of working, living and recreation areas and equipment;
  • Adequacy of living condition;
  • Level of ship motion, vibrations, and heat and noise.
• Ship factors:
  • Design;
  • State of maintenance;
  • Equipment (availability and reliability);
  • Cargo characteristics, including securing, and handling and care.
• Shore-side management:
  • Policy on recruitment;
  • Safety policy and philosophy (culture, attitude, and trust);
  • Management commitment to safety;
  • General management policy.
• External influences and environment:
  • Weather and sea conditions;
  • Port and transit conditions (such as VTS and pilots);
  • Traffic density;
  • Regulations, surveys and inspections (including international, national, port, and classification societies).

However, IMO’s six-factor classification, designed primarily for large merchant ships, includes many items not applicable to small fishing vessels. Therefore, we selected the items most relevant to maritime accidents occurring on fishing vessels. Additionally, we provided a wider range of options for respondents by including various examples for each item. Figure 6 illustrates the indirect factors selected for this study based on the IMO classification scheme. In this hierarchical diagram, the six factors classified by the IMO were defined as primary indirect factors, and their respective subfactors were defined as secondary indirect factors.

2.4. Indirect Factor Survey

While fishing vessel accident data facilitate the quantitative analysis of accident patterns, a survey enables the qualitative analysis of indirect factors affecting the human element. This study conducted a survey using items derived from Figure 6. The survey was carried out as part of the “Fishing Operations Safety and Accident Prevention Manual” project overseen by the National Institute of Fisheries Science (NIFS) [28]. It targeted 201 fishermen engaged in fishing on vessels under 12 m. To ensure analytical accuracy, incomplete surveys (those with even one unanswered question) were excluded, resulting in a final sample of 192 respondents.

Most fishermen were aboard multipurpose vessels capable of engaging in various fishing activities, such as gillnetting, longlining, and trap fishing, depending on the season and location (Figure 7). The years of onboard experience among respondents were distributed as follows: less than 10 years (n = 16), 10–20 years (n = 37), 20–30 years (n = 58), and more than 30 years (n = 81).

Although the sample size was relatively small compared to the total number of maritime accidents involving fishing vessels, the wide spectrum of experience and the fact that nearly half of the respondents had over 30 years of experience on board were considered sufficient to provide representativeness. Furthermore, the use of multipurpose vessels for diverse fishing activities added to the sample’s representativeness.

The survey comprised multiple-choice items, seeking experts’ opinions on major accident patterns derived from maritime accident statistics and the indirect factors most influencing the human element leading to accidents. To avoid biased responses, the primary indirect factors were broadly categorized using a widely accepted classification scheme (man, management, machine, environment), and the secondary indirect factors were mixed appropriately (Figure 8).

2.5. Fault Tree Analysis (FTA)

In this study, we analyzed maritime accidents using the FTA method, one of the FSA STEP-2 Risk Analysis Methods proposed by the IMO, as a methodology for taking preventive measures before a disaster occurs [20]. FTA is a graphical modeling technique used to investigate the root causes of system failures and malfunctions at various system levels [29]. Specifically, it creates a systematic diagram using logic gates to represent the relationship between potential accident events in plants or industries and their causative faults or errors. By quantitatively and probabilistically analyzing these relationships, it expresses the causes of accidents in terms of probabilities, enabling selective and systematic accident prevention activities. Moreover, by detailing the relationship between direct causes and indirect factors schematically, it provides a comprehensive overview of the chain relationship from accident occurrence patterns to underlying factors [30]. Since it quantitatively analyzes the causes of accidents, the results can be used to create safety checklists.

2.5.1. Analysis Procedure

While the procedure for conducting an FTA may vary slightly depending on the analysis target, it generally follows these key steps:

• Top event selection: The significance of events is evaluated based on their type and pattern. The most critical undesired event, referred to as the top event, is selected as the starting point for the risk analysis.
• Cause identification: A thorough investigation is conducted to identify the root causes that could lead to the top event. This may involve analyzing data from accident reports, incident databases, or expert interviews.
• FT diagram construction: The relationships between the top event and its potential causes, including direct causes, indirect factors, and basic events, are depicted graphically using an FT diagram. Logic gates connect these elements to show how various contributing factors can combine to cause the top event.
• FT quantification: The event occurrence rates are calculated by calculating the frequencies of individual factors (basic events) within the FT diagram, starting from the most basic level and working upwards.
• Risk assessment: Based on the calculated occurrence rates obtained in step 4, a risk assessment is conducted.

2.5.2. Symbols and Terms

The FTA technique employs various symbols and terms. However, using overly complex symbols can reduce the ease of calculation and visibility. Therefore, this study constructed an FT diagram using only basic symbols and terms, as outlined in

Table 2. Fault tree (FT) symbols and descriptions.

Symbols	Name	Description
	Intermediate event	An event that occurs as a result of one or more preceding events within an FT.
	External event	An event that is outside the control of the system being analyzed and is assumed to occur independently.
	Basic event	A fundamental event that does not need further development within the fault tree.
	AND Gate	A logic gate in FT analysis (FTA) where the output event occurs only if all input events occur simultaneously.
	OR Gate	A logic gate in the FTA where the output event occurs if at least one of the input events occurs.

.

2.5.3. Probability Calculation

The FTA technique employs Boolean Algebra as a tool for logical calculation. By representing subsets of sets with AND gates and OR gates, the probability of events, i.e., the probability of accidents, can be calculated. The AND gate symbol used in the FT diagrams indicates the situation in which a result event occurs as a result of the fulfillment of all entry conditions, and the OR gate symbol indicates the situation where an outcome event occurs as a result of at least one of the entry conditions [31]. The probability of events found using analytical methods is as follows (Figure 9 and Figure 10). Here, the letters A, B, and C are used to describe the structure of the FT diagram.

• AND Gate (Conjunction)

$${\mathrm{F}}_{\mathrm{A}}=\mathrm{F}\left({\mathrm{B}}_{ \mathrm{A}\mathrm{N}\mathrm{D}} \mathrm{C}\right)=\mathrm{F}\left(\mathrm{B}\cap \mathrm{C}\right)$$

(1)

If B and C are independent events,

$${\mathrm{F}}_{\mathrm{A}}={\mathrm{F}}_{\mathrm{B}}·{\mathrm{F}}_{\mathrm{C}}$$

(2)

or

$${\mathrm{F}}_{\mathrm{A}}=1-{\mathrm{R}}_{\mathrm{A}}=\left(1-{\mathrm{R}}_{\mathrm{B}}\right)\left(1-{\mathrm{R}}_{\mathrm{C}}\right)={\mathrm{F}}_{\mathrm{B}}·{\mathrm{F}}_{\mathrm{C}}$$

(3)

When n basic events are combined with an AND gate, the probability of a successful event (F_S) is calculated using Equation (4):

$${\mathrm{F}}_{\mathrm{S}}={\mathrm{F}}_1·{\mathrm{F}}_2\cdots {\mathrm{F}}_{\mathrm{n}}=\prod _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{F}}_{\mathrm{i}}$$

(4)

where F is the probability of event, R is the reliability of event, and n is the number of events.

2.

OR Gate (Disjunction)

$${\mathrm{F}}_{\mathrm{A}}=\mathrm{F}\left({\mathrm{B}}_{ \mathrm{O}\mathrm{R}} \mathrm{C}\right)=\mathrm{F}\left(\mathrm{B}\cup \mathrm{C}\right)$$

(5)

If B and C are independent events,

$${\mathrm{F}}_{\mathrm{A}}={\mathrm{F}}_{\mathrm{B}}+{\mathrm{F}}_{\mathrm{C}}-{\mathrm{F}}_{\mathrm{B}}·{\mathrm{F}}_{\mathrm{C}}$$

(6)

or

$${\mathrm{F}}_{\mathrm{A}}=1-{\mathrm{R}}_{\mathrm{A}}=1-{\mathrm{R}}_{\mathrm{B}}·{\mathrm{R}}_{\mathrm{C}}=1-\left(1-{\mathrm{F}}_{\mathrm{B}}\right)\left(1-{\mathrm{F}}_{\mathrm{C}}\right)={\mathrm{F}}_{\mathrm{B}}+{\mathrm{F}}_{\mathrm{C}}-{\mathrm{F}}_{\mathrm{B}}·{\mathrm{F}}_{\mathrm{C}}$$

(7)

When n basic events are combined with an OR gate,

$${\mathrm{F}}_{\mathrm{S}}=1-\left(1-{\mathrm{F}}_1\right)\left(1-{\mathrm{F}}_2\right)\cdots \left(1-{\mathrm{F}}_{\mathrm{n}}\right)=1-\prod _{\mathrm{i}=1}^{\mathrm{n}}\left(1-{\mathrm{F}}_{\mathrm{i}}\right)$$

(8)

3. Results and Discussion

3.1. Analysis of Fishing Vessel Maritime Accident Statistics

The insurance payment data for maritime accidents involving fishing vessels in this study categorizes accident types as trip/slip, bump/hit, stuck, fall, and others. This categorization aligns with the classification method used for industrial accidents by the Korea Occupational Safety and Health Agency [32]. Consequently, this study employed the same accident categories to classify the types of accidents that occurred.

From 2018 to 2022, a total of 1093 accidents involving fishing vessels under 12 m were recorded. Over these five years, trip/slip accidents were the most frequent (41.4%), followed by bump/hit (12.6%), stuck (8.6%), and fall (4.3%). The “others” category, encompassing all other accident types such as illnesses and unknown collisions or groundings, was excluded from the frequency ranking (

Table 3. Types of fishing vessel accidents.

	Year	2018	2019	2020	2021	2022	Total
Accident Type
Trip/Slip		67 (36.0)	106 (46.3)	100 (43.3)	94 (40.9)	85 (39.2)	452 (41.4)
Bump/Hit		26 (14.0)	25 (10.9)	32 (13.9)	35 (15.2)	20 (9.2)	138 (12.6)
Stuck		22 (11.8)	18 (7.9)	19 (8.2)	19 (8.3)	16 (7.4)	94 (8.6)
Fall		7 (3.8)	13 (5.7)	12 (5.2)	9 (3.9)	6 (2.8)	47 (4.3)
Others		64 (34.4)	67 (29.3)	68 (29.4)	73 (31.7)	90 (41.5)	362 (33.1)
Total		186 (100.0)	229 (100.0)	231 (100.0)	230 (100.0)	217 (100.0)	1093 (100.0)

Unit: number (%).

).

3.2. Survey Results

3.2.1. Top Event Selection

The survey results (

Table 4. Survey results on accident types.

	Onboard Experience	Less than 10 Years	10 to Less than 20 Years	20 to Less than 30 Years	30 Years or More	Total
Accident Type
Stuck		9 (56.3)	17 (45.9)	25 (43.1)	34 (42.0)	85 (44.3)
Bump/Hit		3 (18.8)	1 (2.7)	16 (27.6)	32 (39.5)	52 (27.1)
Trip/Slip		3 (18.8)	14 (37.8)	15 (25.9)	7 (8.6)	39 (20.3)
Fall		1 (6.3)	5 (13.5)	2 (3.4)	8 (9.9)	16 (8.3)
Total		16 (100.0)	37 (100.0)	58 (100.0)	81 (100.0)	192 (100.0)

Unit: number (%).

) deviated slightly from the ranking of accident types based on maritime accident statistics. Despite minor variations due to years of onboard experience, experts generally perceived the frequency of accidents to be stuck (44.3%), bump/hit (27.1%), trip/slip (20.3%), and fall (8.3%).

This discrepancy between perceived (survey results) and actual (statistical data) accident frequency might explain why the rate of fishing vessel accidents stemming from the human element has not decreased despite extensive safety training. Notably, the accident type ranked third by experts (trip/slip) was actually the most frequent, while the type they ranked first (stuck) was only the third most frequent. This suggests that fishermen might be more cautious about “stuck” accidents, believing them to be the most common. This heightened awareness of “stuck” accidents may have served as a preventative measure, potentially reducing their occurrence.

However, safety training should prioritize the accident types that actually occur most frequently, with safety checklists established accordingly. In this study, “trip/slip” emerged as the top event, despite not being recognized as such by the experts.

3.2.2. Indirect Factor Occurrence Rates

Table 5. Survey results.

Primary Indirect Factors	Secondary Indirect Factors (Survey Items)	Answer
People factors (C1)	Personal negligence (D11)	22 (14.1)
	Failure to wear personal protective equipment (D12)	2 (1.3)
	Incorrect working posture (D13)	6 (3.8)
	Unskilled fishing operations (D14)	1 (0.6)
Organization on board (C2)	Inappropriate fishing methods (D21)	1 (0.6)
	Poor communication due to mixed foreign fisher men (D22)	2 (1.3)
	Reduced concentration due to fatigue, such as lack of sleep (C23)	7 (4.5)
	Insufficient pre-education on risk factors/situations (D24)	7 (4.5)
	Fishing practices that do not prioritize safety (D25)	12 (7.7)
Working and living conditions (C3)	Cluttered and untidy deck (D31)	1 (0.6)
	Confined working space with no place to escape (D32)	13 (8.3)
	Dangerous fishing gear with high tension (D33)	2 (1.3)
	Impaired communication due to severe noise and vibration (D34)	6 (3.8)
Ship factors (C4)	Structural defects in fishing machinery (D41)	8 (5.1)
	Lack of warning devices for abnormal operations (D42)	3 (1.9)
	Unsafe arrangement of fishing equipment (D43)	3 (1.9)
	Absence of emergency stop devices (D44)	7 (4.5)
	Aging equipment and poor maintenance (D45)	16 (10.3)
	Defective safety barriers such as no-entry barriers (D46)	2 (1.3)
Shore-side management (C5)	Lack of guidelines on work regulations (D51)	2 (1.3)
	Failure to post safety rules (D52)	4 (2.6)
	Absence of hazard warning signs (D53)	5 (3.2)
	Inadequate management of health examinations for workers (D54)	9 (5.8)
External influences and environment (C6)	Harsh natural conditions such as rough waves (D61)	15 (9.6)
	Total	156 (100.0)

Unit: number (%).

presents the survey results on secondary indirect factors identified by experts, who considered “trip/slip” accidents to be the most frequent. Among the people factors, “personal negligence” (14.1%) emerged as the most prominent secondary indirect factor. It was followed by “aging equipment and poor maintenance” (10.3%) among the ship factors and “harsh natural conditions like rough waves” (9.6%), the single factor in the category “external influences and environment.”

It is noteworthy that factors categorized under “ship factors” and “external influences and environment” were among the top three factors (second and third, respectively), despite the presence of diverse secondary indirect factors within the category “people factors”. This indirectly suggests that multiple factors beyond human error contribute to the human element in maritime accidents.

4. FT Diagram Configuration

Accidents can generally be categorized into those caused by human error and those caused by unavoidable events, often referred to as “Acts of God” such as natural disasters. A prime example of an Act of God in a maritime context would be a powerful typhoon causing a ship to sink. In this study, also, both human error and Act of God events are represented as “events.” While “Acts of God” was used for FT diagram construction (Figure 10), it was not analyzed separately in this study.

The most frequently occurring type of accident (slip/trip) identified from the quantitative analysis of insurance payment data was designated as A, and the human element constituting the direct cause was represented as B. Additionally, the six indirect factors influencing the human element are represented as C (primary indirect factors), and their subfactors are represented as D (secondary indirect factors) (Table 5 and Figure 11). Here, the letters A, B, C, D, and the numbers are simply used to represent the structure of the FT diagram.

Since the aim of this study is to establish selective preventive measures based on the occurrence rates of each sub-factor and their impact on the higher-level factors, the logical gates in Figure 10 are constructed as OR gates. Specifically, secondary indirect factors are represented using OR gates because they are unlikely to occur simultaneously and were classified as separate factors in IMO Resolution A.884(21). Similarly, primary indirect factors are also represented with OR gates to establish selective preventive measures based on the extent to which each factor contributes to the direct cause, aligning with the study’s objective.

FT Quantification and Risk Assessment

Table 6. Occurrence rates of secondary indirect factor.

D11: 0.141	D12: 0.013	D13: 0.038	D14: 0.006
D21: 0.006	D22: 0.013	D23: 0.045	D24: 0.045	D25: 0.077
D31: 0.006	D32: 0.083	D33: 0.013	D34: 0.038
D41: 0.051	D42: 0.019	D43: 0.019	D44: 0.045	D45: 0.103	D46: 0.013
D51: 0.013	D52: 0.026	D53: 0.032	D54: 0.058
D61: 0.096

Unit: number.

presents the calculated occurrence rates of all secondary indirect factors based on the survey data from Table 5. Here, to maintain the same format as the calculation results using logical AND and OR, the probabilities in Table 5 were divided by 100.

Since secondary factors are considered to be inherently within primary factors, they have the effect of triggering primary factors. Therefore, applying an OR gate (disjunction) to the occurrence rates of the secondary factors, those of the primary indirect factors can be obtained as follows:

$$\mathrm{C}1=1-\left\{\left(1-0.141\right)\times \left(1-0.013\right)\times \left(1-0.038\right)\times \left(1-0.006\right)\right\}=0.18928\left(18.9\mathrm{\%}\right)$$

$$\mathrm{C}2=1-\left\{\left(1-0.006\right)\times \left(1-0.013\right)\times \left(1-0.045\right)\times \left(1-0.045\right)\times \left(1-0.077\right)\right\}=0.17413\left(17.4\mathrm{\%}\right)$$

$$\mathrm{C}3=1-\left\{\left(1-0.006\right)\times \left(1-0.083\right)\times \left(1-0.013\right)\times \left(1-0.038\right)\right\}=0.13454\left(13.5\mathrm{\%}\right)$$

$$\mathrm{C}4=1-\left\{\left(1-0.051\right)\times \left(1-0.019\right)\times \left(1-0.019\right)\times \left(1-0.045\right)\times \left(1-0.103\right)\times (1-0.013)\right\}=0.22782\left(22.8\mathrm{\%}\right)$$

$$\mathrm{C}5=1-\left\{\left(1-0.013\right)\times \left(1-0.026\right)\times \left(1-0.032\right)\times \left(1-0.058\right)\right\}=0.1234\left(12.3\mathrm{\%}\right)$$

$$\mathrm{C}6=1-(1-0.096)=0.096\left(9.6\mathrm{\%}\right)$$

The calculated results show the following order of occurrence rates of the primary indirect factors affecting the human element associated with trip/slip: ship factors (22.8%) > people factors (18.9%) > organization on board (17.4%) > working and living conditions (13.5%) > shore-side management (12.3%) > external influences and environment (9.6%).

It has been a common research trend to attribute accidents solely to the human element. Consequently, preventive measures have often focused on safety training for fishermen. However, this study reveals that ship factors have a greater impact on human error than people factors. This suggests that safety improvements should address not just the crew, but also the physical environment they work in.

Since all primary indirect factors can lead to the human element, we can calculate the occurrence rates of the direct causes of trip/slip accidents leading to the human element, applying OR gates (disjunction) similar to determining those of primary indirect factors as follows:

$$\mathrm{B}=1-\left\{\left(1-0.189\right)\times \left(1-0.174\right)\times \left(1-0.135\right)\times \left(1-0.228\right)\times \left(1-0.123\right)\times (1-0.096\right\}=0.64535\left(64.5\mathrm{\%}\right)$$

The probability of the human element occurring due to indirect factors was calculated at 64.5%.

Subsequently, we examined the occurrence rate of the human element when a selective prevention of specific indirect factors was implemented. First, the occurrence rate of the human element (B′) was determined, with the occurrence rate of “personal negligence”, the most frequent secondary factor, set to zero, as follows:

$$\mathrm{B}\mathrm{’}=1-\left\{\left(1-0.056\right)\times \left(1-0.174\right)\times \left(1-0.135\right)\times \left(1-0.228\right)\times \left(1-0.123\right)\times (1-0.096\right\}=0.58719\left(58.7\mathrm{\%}\right)$$

The elimination of “personal negligence” reduced the overall occurrence rate of the human element by 5.8%.

Finally, we calculated the occurrence rate of the human element (B″) by setting the three most frequent secondary factors (personal negligence, aging equipment and poor maintenance, and harsh natural conditions such as rough waves) to zero:

$$\mathrm{B}\mathrm{’}\mathrm{’}=1-\left\{\left(1-0.056\right)\times \left(1-0.174\right)\times \left(1-0.135\right)\times \left(1-0.139\right)\times \left(1-0.123\right)\times (1-0.0\right\}=0.4907\left(49.1\mathrm{\%}\right)$$

Eliminating these three most-frequent secondary indirect factors yielded a decrease in the occurrence rate of the human element by 15.4%.

However, since the top event (trip/slip) involves several other contributing factors and the focus of this study is on identifying the role of indirect factors and deriving selective prevention measures, the top event itself was not included in this process.

5. Conclusions

Most accidents occurring on fishing vessels result in casualties, and research has shown that these accidents are primarily caused by human error, leading to a focus on crew education. However, despite ongoing training for the crew, the number of marine accidents on fishing vessels has not decreased; in some cases, it has even increased.

Researchers have concluded that this is due to fishermen not implementing the training content, prompting them to propose enforceable safety policies such as marking hazardous zones and requiring the use of safety helmets. However, without identifying the underlying causes inherent in accident occurrences, safety policies aimed at reducing only the surface-level causes did not lead to significant changes.

Small fishing vessels often operate in a regulatory blind spot, making maritime accidents a significant concern. This study aimed to explore ways to selectively prevent maritime accidents on small fishing vessels. This focus arose from three key observations: most fishing vessel accidents occur in small vessels under 12 m; these accidents are caused by the human element; and the occurrence rate of these accidents has not decreased despite the ongoing safety training of fishermen.

The analysis identified “personal negligence”, one of the people factors, as the secondary indirect factor with the greatest impact on the human element and “ship factors” as the most influential primary indirect factor. Additionally, “working and living conditions” and “shore-side management” also significantly influenced the human element. Eliminating the three most-frequent secondary indirect factors—personal negligence, aging equipment and poor maintenance, and harsh natural conditions (such as rough waves)—decreased the occurrence rate of the human element by over 15%, from 64.5% to 49.1%.

These results indicate that focusing solely on fishermen’s behavior through safety training may no longer be the most effective approach. Instead, prioritizing improvements in high-occurrence indirect factors, such as repairing aging vessels and improving working conditions, is considered to be a more effective method to selectively prevent fishing vessel accidents.

Future research plans include creating a safety checklist for the selective prevention of frequently occurring indirect factors. Additionally, a quantitative analysis will be conducted to determine the extent of actual accident rate reductions depending on the implementation of the checklist.

On the other hand, as mentioned in the Introduction, obtaining official statistics on the number of maritime accidents involving fishing vessels is challenging, as no comprehensive reports have been published by the ILO since their 2002 report. This limitation is also a constraint for this study. In other words, while this study aims to systematically prevent maritime accidents involving fishing vessels under 12 m worldwide, there is uncertainty regarding whether the sample used in this study accurately represents such incidents globally.

Nonetheless, this study was conducted using the most official and diverse samples available, including accident occurrence data published by international organizations, crew experience, and types of fisheries. Therefore, it is hoped that the findings of this study will aid in preventing maritime accidents involving vessels under 12 m, which often lie in regulatory blind spots.