Original Article

Proc IMechE Part M:

J Engineering for the Maritime Environment
1–11
A human reliability assessment Ó IMechE 2020
Article reuse guidelines:
through enclosed space entry sagepub.com/journals-permissions
DOI: 10.1177/1475090220979867

operation onboard ships journals.sagepub.com/home/pim

Omer Soner1 and Metin Celik2

Abstract
An enclosed space poses a serious risk to workers since its atmosphere can be oxygen-deficient, polluted or contain
flammable and/or toxic gases or vapours. Hence, it is important to keep in mind that an enclosed space should not be
entered unless it is absolutely necessary. However, many processes often need the worker to be present in an enclosed
space. At this point, great attention should be paid to create and maintain safe working conditions, when entering
enclosed space is required. Even though studies have investigated several aspects of enclosed space entry, further
research is required to explore the human element role. This is urgently needed, considering that human errors are one
of the main reasons for enclosed space entry accidents. Hence, this study has investigated the human element role dur-
ing an enclosed space entry operation via the enhanced SOHRA method. Three scenarios have been developed to assess
HEP values. Beyond that, systematic control measure assignment is provided to reduce HEP values. Hierarchy of control
measures are considered to improve entering enclosed space operational safety. Elimination (drone-assisted survey),
substitution (IoT), engineering (permit control system-PCS), and administrative control measures (electronic permit)
have been proposed. Further, this study results can also provide useful insight for other industries (petrochemical, oil,
refinery, gas, agriculture, construction etc.), where the accidents in enclosed spaces taking place. Therefore, the study
provides not only a theoretical insight but also practical application to improve the safety of seafarers.

Keywords
Enclosed space entry, confined space, human reliability, ship safety, ship operation

Date received: 12 March 2020; accepted: 11 November 2020

Introduction attempts may lead to more casualties on board ships.4,5

Thus, it is not surprising that enclosed spaces entry is
Enclosed space is a high-risk working environment the primary cause of accidental deaths on board ships.6
which endangers workers life not just on-board ship The International Maritime Organization (IMO) has
but also in many other industries including agriculture1 implemented the Recommendations for Entering
and construction.2 Limited openings for entry and exit, Enclosed Spaces Aboard Ships to prevent accidents
inadequate ventilation, and not being designed for con- related to entry into enclosed spaces onboard ships.3
tinuous worker occupancy are the main characteristics Moreover, IMO includes enclosed spaces entry and res-
of the enclosed spaces.3 Considering the characteristics
cue drill on a regular basis (once every 2 months) and
of the enclosed spaces and a ship structure, a great
has amended the Revised Recommendations for
number of working areas on board ships might identify
Entering Enclosed Spaces Aboard Ships and Safety of
as enclosed spaces (including cargo spaces, double bot-
toms, fuel tanks, ballast tanks, cargo pump-rooms,
including cargo compressor rooms.).3 However, it is 1
Department of Maritime Transportation and Management Engineering,
not always easy to identify an enclosed space on ships, Van Yuzuncu Yıl University, Tusba, Van, Turkey
especially where entry into enclosed spaces happens 2
Department of Basic Science, Istanbul Technical University, Tuzla,
rarely. It should also be noted that pursuing the rescue Istanbul, Turkey
plan and correct use of rescue equipment is challenging
Corresponding author:
when attempting to rescue a colleague who collapsed Omer Soner, Department of Maritime Transportation and Management
within an enclosed space.4,5 At this point, impulsive Engineering, Van Yuzuncu Yıl University, Tusba 65090, Van, Turkey.
reactions of the seafarer and inappropriate rescue Email: [email protected]
2 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

Life at Sea (SOLAS) Regulation III/19.7,8 SOLAS reg- spaces.24 A comprehensive framework has also been
ulation XI-1/7 also offers guidelines to facilitate the developed to prevent and control risks in confined
selection of portable atmosphere testing instruments spaces25,26 have been investigating the contributing fac-
for enclosed spaces, and has been implemented.9 tors of confined space through semi-structured accident
Despite the above-mentioned safety measures being investigation reports and interviews. In addition to risk
implemented, enclosed space entry related accidents assessment studies, confined space emergency
and detentions rates still remain at a higher than accep- response,27 technologies for confined space,28 identifi-
table level. According to the Marine Accident cation of confined spaces,29 and a confined space res-
Investigators’ International Forum (MAIIF) survey, cue procedure30,31 have been investigated, as well.
101 enclosed space accidents recorded resulted in 93 Even though the above studies have investigated sev-
deaths and 96 injuries, between 1998 and 2009.10 In eral aspects of enclosed space entry, further research is
accordance with the Paris MoU (Memorandum of required to explore the human element role. Since
Understanding) Concentrated Inspection Campaign human errors have been identified one of the main
(CIC) on Familiarization for Enclosed Space Entry causes of enclosed space entry accidents, this is needed
report, enclosed space entry is the most frequently urgently. To remedy mentioned gap in the literature,
encountered deficiency that leads to ship detentions (49 this study aims to perform a human reliability assess-
out of 71, or 69%).11 It is also mentioned in the same ment for enclosed space entry on board ships.
report that Paris MoU and Black Sea MoU CIC results Therefore, the study is expected to contribute at the fol-
are almost equally high (40% and 40.82% respectively). lowing points: (i) systematic human error prediction
Moreover, according to the Marine Accident and assessment on enclosed space (ii) in addition to
Investigation Branch (MAIB) safety bulletin, three procedural requirements, a holistic approach can be
crew members of the cargo vessel SUNTIS lost their recommended for enclosed spaces (iii) after required
lives in the main cargo hold on 26 May 2014.12,13 adjustment, the model is applicable for other industries
Another fatal accident, resulting in death of a crew (petrochemical, oil, refinery, gas, agriculture, construc-
member, involving enclosed space entry was reported tion etc.) to conduct human error assessment (iv)
in 2017.14 proactive risk modelling is exemplified for shipping
While many causal factors for confined space inci- operations (v) advanced risk assessment modelling has
dents have been identified, the function of human fac- been implemented for the enclosed space entry. The
tors has been emphasized.15 According to the research study is organized in this context as follows. Section 1
that have been investigated human errors in shipping explains the motivation behind the study and literature
accidents reported the major contribution of skill-based review about enclosed space entry. Section 2 presents
errors, personal factors, and organizational process for the theoretical background of the utilized method.
shipping accidents.16 Studies that have focused on the Section 3 demonstrates the model application with
enclosed space entry accidents reveal that most of the three accidents/incident scenarios. Section 4 provides
accidents and detentions occur due to human concluding remarks.
errors.4,6,11,13,17–19 In particular, three deficiencies are
highlighted: inadequate training and knowledge of the
proper entry procedures, disregard for them, and the Methodology
need to take precautions.4,7 Since enclosed space entry
SOHRA
is required for continuous care to create and maintain
safe working conditions, the human element is an indis- Human error assessment and reduction technique
pensable factor along with the procedures and safety (HEART) is used to measure human error probability
equipment.4,20 Therefore, decision-makers and during the execution of a given task. HEART theory is
researchers should be focused on human related factors based on the idea that any time a task is done there is a
to avoid enclosed space entry related accidents. risk of failure and that the probability of a loss is influ-
According to one study, which is a literature review enced by one or more error-producing conditions
conducted by,21 existing risk assessment tools are (EPCs) / performance shaping factors (PSF). Thus,
imprecise, and more systematic, multidisciplinary, and EPC/PSF values are essential to be used to obtain HEP
specific characteristics and tools are needed to support values. Since EPC affects human performance nega-
managers, practitioners and decision makers22 propose tively when an EPC is assigned a value, which is less
a risk assessment tool for confined spaces. The study than 1.0, the influence on the HEP value has become
focused on atmospheric, chemical, biological, falling, insignificant.32 Shipboard Operation Human
mechanical, physical and ergonomic risk factors and Reliability Analysis (SOHRA) is a modified version of
human error was not taken into consideration,23 on the the HEART to measure human error and predict
other hand, compared checklists, risk scales, and risk human reliability in shipping operations.33,34 The
matrices-based risk analysis tools to assess the effective- method takes advantage of the m-EPC (marine specific
ness of these methods. A multi-criteria decision tool error producing condition) values which specify the
was found to quantitatively determine the most impor- operational factors that affect the crew performance in
tant characteristics to avoid hazards in confined operational condition on aboard a ship32 have
Soner and Celik 3

Table 1. GTT and GEP values.

Generic task type (GTT) Generic error probability (GEP)

(5th–95th percentile Bounds)

A Totally unfamiliar; performed at speed with no real idea of likely consequences 0.55 (0.35–0.97)
B Shift or restore system to a new or original state on a single attempt without 0.26 (0.14–0.42)
supervision or procedures
C Complex task requiring high level of comprehension and skill 0.16 (0.12–0.28)
D Fairly simple task performed rapidly or given scant attention 0.09 (0.06–0.13)
E Routine, highly practiced, rapid task involving relatively low level of skill 0.02 (0.07–0.045)
F Restore or shift a system to original or new state following procedures with 0.003 (0.0008–0.007)
some checking
G Completely familiar, well-designed, highly practiced, routine task occurring 0.0004 (0.00008–0.009)
several times per day, performed to highest possible standards by highly
motivated, highly trained, and experienced personnel, with time to correct
potential error, but without the benefit of significant job aid
H Respond correctly to system command even when there is an augment or 0.00002 (0.000006–0.0009)
automated supervisory system providing accurate interpretation of system
state
M Miscellaneous task for which no description can be found 0.03 (0.008–0.11)

produced the m-EPC values by reviewing a hundred

marine accidents caused by human error. In addition, Table 2. List of m-EPC.
SOHRA provides remarkable consistency in its No Error producing Max. affect
assessed proportion of effect (APOA) calculation, condition
which is the key aspect of the m-EPC equation of
human error weighting.34 On the other hand, perfor- EPC1 Unfamiliarity 17.00
mance shaping factors (PSF) of crew should be defined EPC2 Time shortage 14.01
EPC3 Low signal–noise ratio 3.31
with respect to the m-EPCs.34–36 Since SOHRA is con- EPC4 Features over-ride allowed 8.72
sidered a consistent and practical model to measure EPC5 Spatial and functional 5.76
human error, it has been used for calculating the incompatibility
human error probability (HEP) in the bunkering opera- EPC6 Model mismatch 2.64
EPC7 Irreversibility 2.23
tions on a ship.34 The model consists of the two main
EPC8 Channel overload 14.45
parameters, which are generic task type (GTT) and m- EPC9 Technique unlearning 5.29
EPC.32 While the GTT makes it possible to identify EPC10 Knowledge transfer 11.00
how the task would be performed in perfect conditions, EPC11 Performance ambiguity 8.60
m-EPC indicates the operational conditions that influ- EPC12 Misperception of risk 12.51
EPC13 Poor feedback 12.55
ence the generic error probability (GEP) value.32,34 EPC14 Delayed/incomplete feedback 6.72
GTT and GEP values is presented in the Table 1.33 The EPC15 Operator inexperience 10.03
m-EPC list and its full impact is shown in Table 2.32 EPC16 Impoverished information 8.42
The SOHRA quantifies human error within seven main EPC17 Inadequate checking 2.79
steps. Figure 1 shows a flow chart for the enhanced EPC18 Objectives conflict 2.15
EPC19 No diversity 2.74
SOHRA methodology.34 In this study, enhanced EPC20 Educational mismatch 2.88
SOHRA model has been derived from.32 Further, the EPC21 Dangerous incentives 3.62
SOHRA model is combined with the hierarchy of con- EPC22 Lack of exercise 1.64
trol measures to systematically reduce the probability EPC23 Unreliable instruments 5.69
of human error. Detailed information for each step is EPC24 Absolute judgements required 1.17
EPC25 Unclear allocation of function 1.22
given as follows. EPC26 Progress tracking lack 3.28
EPC27 Physical capabilities 4.35
EPC28 Low meaning 2.56
The process of SOHRA EPC29 Emotional stress 1.59
EPC30 ill-health 0.89
EPC31 Low morale 3.00
Step 1. Task analysis: The objective of the first step is EPC32 Inconsistency of displays 9.43
to determine the tasks in accordance with the scenario. EPC33 Poor environment 9.90
EPC34 Low mental workload 2.63
A task is an activity or step to be performed by crew EPC35 Sleep cycles disruption 10.30
members within a specified period to accomplish the EPC36 Task pacing 3.85
work-related objectives. The main tasks should be sub- EPC37 Supernumeraries 4.14
divided and each of them must be conducted with EPC38 Age 3.61
4 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

Figure 1. Flow diagram of the advanced SOHRA.

respect to the hierarchical task analysis (HTA).32,37 In (AHP) to sensitive and consistence calculation of the
this manner, HEP values for each sub-task are calcu- weighting process. The AHP is one of the multi-criteria
lated in order to capture the human error changes.34 decisions making (MCDM) approaches that aimed to
Step 2. Scenario definition: After the task analysis step, quantify subjective assessment.38 It allows a user to create
scenario details are determined in order to specify the a pair-wise matrix ratio scale. The AHP enables the user
environmental conditions of the scenario. To illustrate, to perform consistent evaluation via consistency and ran-
stress level, time availability, noise level, weather condi- dom index and thus it is considered a reliable
tions, fatigue, familiarity with a task, and time of day approach.34,39 The method process consists of three steps.
might all be considered environmental conditions. I. Composing a pair-wise comparison matrix: First of all,
Next, respective GTT and m-EPC are also selected to obtain relative weight of each m-EPC, a pair-wise
based on the conducted scenario.34 comparison matrix is established. In this regard, the
Step 3. GTT selection: For each sub-task, the appropri- linguistic relative importance scale of Saaty 1–9 is
ate GTT is selected by a responsible officer in relation used.40 A comparison matrix A is an actual matrix of
to the task analysis. It is critical to match GTT with nxn where n denotes the number of m-EPCs are
each sub-task, because GEP values are determined assessed. The relative importance of the ith versus
based on these matches.34,35 the jth is indicated for each m-EPC by inserting aij (i,
Step 4. m-EPC selection: The responsible officer selects j = 1, 2, 3, . . . , n) in Matrix A. This means ith m-EPC is
the most suitable m-EPC for each subtask based on the more than jth when the aij . 1. In the case of aij \ 1,
defined scenario. However, it is feasible to choose more ith m-EPC is less essential than jth . If there is the same
than one m-EPC for each sub-task. In such circum- relative weight of two m-EPCs then aij =1. For com-
stances, the APOA calculation should also be con- posing the matrix A, the following equation (1) is used.
ducted to determine weight priorities.34
Step 5. APOA calculation: APOA calculation is
required to reach the portion effect of each m-EPC. aij xaji ð1Þ
The SOHRA adopts the Analytic Hierarchy Process
Soner and Celik 5

II. Calculating criteria weights: This step is designed to aimed at promoting the implementation of safety pro-
calculate relative (w) weight of each m-EPC. cedures. This is because inadequate awareness or mis-
understanding of the need to follow precautionary
measures is responsible for most enclosed space-related
1 Xn a accidents.7 An enclosed space might become a threat to
wi = Pn ij ð2Þ
n j=1
k = 1 akj human life when its atmosphere is oxygen-deficient,
oxygen-enriched, flammable and/or toxic.7,20 Effects of
III. Checking consistency rate: In this stage, the data
exposure to lower oxygen concentration include mental
inserted in a comparison matrix are monitored to
confusion, loss of consciousness, permanent brain
ensure consistency. The rate of consistency is measured
damage, and even death.19,20,45 Therefore companies,
according to equations (3)–(5), respectively.41,42
related shareholders, and researchers should consider
the procedures of entering enclosed spaces as one of
lmax  n the key shipboard operations.7 At this point, systematic
CI = ð3Þ and robust methods should support the creation and
n1
Xn maintenance of safe working conditions to prevent
aijwj = lmax wi ð4Þ
j=1 casualties to ships’ personnel entering enclosed spaces.
CR = CI=RI ð5Þ Therefore, the study takes advantage of the enhanced
SOHRA to investigate the IMO’s revised Enclosed
Step 6. HEP calculation: The HEP value is calculated Space Entry Permit procedures using an extensive
for each of the sub-tasks after a sensitive APOA calcu- human reliability analysis with three scenarios. IMO’s
lation has been performed. Equation (6) is used to for revised Enclosed Space Entry Permit checklist is pro-
this objective.33 The formula is based on EPCi , which vided in Table 3.7 The checklist indicates all task and
is ith (i = 1, 2, 3, :::n; n = 38) m-EPC, and APOApi , sub-task that expected to be performed prior to any
which is (0 \ pi \ 1) ith , and thus the proportional enclosed space entry. The checklist should be com-
impact measured in APOA is evaluated. pleted prior to entering any enclosed space onboard
ships.
( )
Y 
HEP = GEPx (EPCi  1)APOApi + 1 ð6Þ Task analysis and scenario details
i
Different methods have been presented in the literature
for the risk assessment. Quantitative risk analysis,
Step 7. Applying the hierarchy of control measures: probabilistic safety analysis, worst-case methodology
The National Institute for Occupational Safety and for risk assessment, and optimal risk analysis are
Health (NIOSH) defines five steps of the hierarchy of among the most notable ones.46,47 The analysis of acci-
controls.43 The hierarchy is a systematic way of devel- dent scenarios is at the core of all these methods; a
oping the most effective and feasible method of reduc- range of scenarios are developed and then evaluated
ing hazard-related risks. The hierarchy is based on the for comprehensive consequences.46,48,49 In addition,
idea that control measures at the top are potentially recent statistics indicate that activities conducted within
more effective and powerful than those at the bottom the enclosed space dramatically increase the risk level
(see Figure 1). In this top-down process, elimination of the enclosed space and contribute to the generation
implies removing the hazard physically, while substitu- of accidents.50 Therefore, different accident scenarios
tion means replacing/aiming to replace the hazard. are used in the study to develop extensive and efficient
Where these measures are not possible or not enough, accident prevention and management strategies. The
engineering controls should take into consideration iso- details of the scenarios are given as follows.
lating people from the hazard.44 Beyond that, the
administrative controls, which are the next step at the Scenario I: A chemical tanker is on the way to load
hierarchy, are intended to change the way people work. 23,000-DWT methanol from San Francisco to Los
Even if the least effective means of controlling mea- Angeles. The weather is overcast with rain, winds are
sures is Personal Protective Equipment (PPE), it should 17 to 21 knots, and waves are 8 to 10 feet. During the
be used together with the all the previous measures.43,44 voyage, it is intended that cargo tanks be cleaned from
This step aims at taking safety measures to minimize the last cargo of benzene and be prepared for loading.
human errors. Experts have determined HEP’s effect A pre-cleaning meeting is conducted, the crew members
on vital ship operations. The control measures are involved are identified and their roles are explained.
aimed at reducing the highest HEP values. The chief officer is authorised to sign the entry permit,
also assigned as supervisor of overall operations and
takes responsibility for testing and monitoring the
Case study
cargo tanks’ atmosphere. The bosun is instructed to
Requirements and standard-setting regulations con- remain at the entrance during the operation to take
cerning the entering enclosed space onboard ships are necessary actions in the event of an emergency and is
6 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

Table 3. HTA (Hierarchical Task Analysis) of an enclosed space entry operation.

1 Pre-entry preparation
1.1 Has the space been thoroughly ventilated by mechanical means?
1.2 Has the space been segregated by blanking off or isolating all connecting pipelines or valves and electrical power/equipment?
1.3 Has the space been cleaned where necessary?
1.4 Has the space been tested and found safe for entry?
1.5 Pre-entry atmosphere test readings:
- oxygen % vol (21%)*....
- hydrocarbon % LFL (less than 1%) ...
- toxic gases ppm (less than 50% OEL of the specific gas) ...
1.6 Have arrangements been made for frequent atmosphere checks to be made while the space is occupied and after work
breaks?
1.7 Have arrangements been made for the space to be continuously ventilated throughout the period of occupation and during
work breaks?
1.8 Are access and illumination adequate?
1.9 Is rescue and resuscitation equipment available for immediate use by the entrance to the space?
1.10 Has an attendant been designated to be in constant attendance at the entrance to the space?
1.11 Has the officer of the watch (bridge, engine room, cargo control room) been advised of the planned entry?
1.12 Has a system of communication between all parties been tested and emergency signals agreed?
1.13 Are emergency and evacuation procedures established and understood by all personnel involved with the enclosed space
entry?
1.14 Is all equipment used in good working condition and inspected prior to entry?
1.15 Are personnel properly clothed and equipped?
2 Pre-entry checks
2.1 I have received instructions or permission from the master or nominated responsible person to enter the enclosed space.
2.2 Section 1 of this permit has been satisfactorily completed by the master or nominated responsible person.
2.3 I have agreed and understand the communication procedures.
2.4 I have agreed upon a reporting interval of ..... minutes.
2.5 Emergency and evacuation procedures have been agreed and are understood.
2.6 I am aware that the space must be vacated immediately in the event of ventilation failure or if atmosphere tests show a
change from agreed safe criteria.
3 Breathing apparatus and other equipment
3.1 Those entering the space are familiar with any breathing apparatus to be used
3.2 The breathing apparatus has been tested as follows:
- gauge and capacity of air supply .......
- low pressure audible alarm if fitted ........
- face mask - under positive pressure and not leaking ........
3.3 The means of communication has been tested and emergency signals agreed
3.4 All personnel entering the space have been provided with rescue harnesses and, where practicable, lifelines.

in direct contact with persons within the space and the anchoring operation, the anchor chain is not able to be
navigating bridge or Chief Officer as appropriate, and let the anchor down due to a tight chain.
two able seamen (A/B) are prepared for entrance into Consequently, the vessel drops another anchor down
the enclosed space for mopping up. The tanks are safely but both anchors should be ready before a possi-
washed and all made gas-free as soon as the enclosed ble sudden berthing operation. Therefore, crew mem-
space entry permit has been issued at daylight. bers have to enter into that chain locker to remedy the
Scenario II: A 35,000-DWT tanker is sailing to the port situation as soon as possible. The chief officer and two
of Vancouver, British Columbia from the port of crew members are selected to enter the chain locker for
Louisiana in empty / ballast condition. The forecast is the repairing operation.
for worsening weather conditions, wind speed is 6
knots with waves of up to 2 feet. During the voyage,
crew members realize that the fuel oil settling tank is GTT and m-EPC selection
leaking. Therefore, the fuel oil in the settling tank has The experts’ judgements can provide a quick evaluation
to be transferred to the double bottom fuel tank so that of the state of knowledge about any particular
the settling tank may undergo repairs. Enclosed space aspect.51,52 An expertise group, consisting of members
entry shall be carried out upon the completion of the who each have more than 7 years of experience,
fuel oil transfer. The chief engineer and three crew (marine superintendent, Hsqe manager and Technical
members are instructed to perform the operation. superintendent) have selected relevant GTT and m-
Scenario III: A bulk carrier 180,000-DWT is scheduled EPC/s for each sub-task in accordance with the details
to anchor to await the loading of coal in Norfolk, of the scenarios. Since the GTT makes it possible to
Virginia. The wind is 7 knots with waves to 2 feet, and find out the appropriate task in the process, and EPC
the weather conditions are favourable. During the displays the performance shaping factors, thus, both
Soner and Celik 7

Table 4. Selected task types and m-EPC for each scenario.

Steps Scenario-1 Scenario-2 Scenario-3

Selected task types Selected m-EPC Selected task types Selected m-EPC Selected task types Selected m-EPC

1. Pre-entry preparation
A1.1 F m-EPC 5 M m-EPC 1 D m-EPC 2
A1.2 C m-EPC 17 D m-EPC 1 M m-EPC 17
A1.3 F m-EPC 5 E m-EPC 5 E m-EPC 22
A1.4 C m-EPC 11 C m-EPC 5 C m-EPC 15
A1.5 C m-EPC 11 C m-EPC 6 C m-EPC 17
A1.6 E m-EPC 2 C m-EPC 2 C m-EPC 2
A1.7 F m-EPC 5 D m-EPC 22 D m-EPC 16
A1.8 D m-EPC 21 F m-EPC 5 C m-EPC 12
A1.9 D m-EPC 2 C m-EPC 5 C m-EPC 2
1.10 E m-EPC 8 E m-EPC 20 E m-EPC 13
1.11 E m-EPC 10 D m-EPC 14 D m-EPC 14
1.12 G m-EPC 3 F m-EPC 3 F m-EPC 1
1.13 G m-EPC 1 F m-EPC 15 A m-EPC 22
1.14 F m-EPC 19 C m-EPC 15 C m-EPC 17
1.15 F m-EPC 20 F m-EPC 15 A m-EPC 5
2. Pre-entry checks
2.1 A m-EPC 12 D m-EPC 12 D m-EPC 11
2.2 F m-EPC 29 C m-EPC 15 C m-EPC 15
2.3 G m-EPC 16 M m-EPC 16 M m-EPC 15
2.4 G m-EPC 14 M m-EPC 16 M m-EPC 15
2.5 G m-EPC 1 M m-EPC 9 M m-EPC 9
2.6 F m-EPC 19 M m-EPC 9 M m-EPC 21
3. Breathing apparatus and other equipment
3.1 G m-EPC 20 C m-EPC 9 C m-EPC 20
3.2 C m-EPC 1 C m-EPC 20 C m-EPC 20
3.3 G m-EPC 3 F m-EPC 3 F m-EPC 1
3.4 E m-EPC 20 F m-EPC 10 F m-EPC 5

can increase HEP values in the associated task.32 The HEP results show that scenario-3 has the high-
Table 4 shows the assigned GTT and m-EPC/s. est overall human error probability, compared with the
other scenarios. Since the scenario-3 includes both a
time shortage and unexpected navigational safety-
HEP calculation related incident, it is reasonable to identify this scenario
It is highly critical to generate HEP values to identify as the riskiest scenario. In this scenario; A1.15, A1.6,
risky sub-tasks within the safety-critical operations. A1.9, A1.8, A1.4, A2.2 and A1.1 have been revealed as
Thus, in this sub-section, equation (6) is applied to cal- a risky sub-task, respectively. This indicates that all
culate the HEP value for each sub-task in enclosed risky sub-tasks except A2.2 in scenario-3 are conducted
space entry operation onboard ships. Accordingly, under the first part of the checklist (1. Pre-entry pre-
paration). Scenario-1 has the second highest overall
HEP results for each scenario are presented in Table 5.
human error probability. In scenario-1, A2.1, A3.2,
Since the HEP values for enclosed space entry opera-
A1.4, A1.5 and A1.9 have been found to be the pri-
tion onboard ships is derived, controls measures should
mary risky sub-tasks, respectively. Scenario-2, on the
be developed accordingly.
other hand, has the minimum overall human error
probability. Sub-tasks that stand out in the scenario-2
are A1.6, A1.14, A2.2, A1.2, A1.4, A1.9 and A3.1,
Results and discussion
respectively. In addition to these, A1.4 and A1.9 has
According to the HEP result, human error probability been identified as a risky sub-task in all three scenarios.
intervals are unique for each scenario. HEP values are A1.6, and A2.2 have been taking a part in both scenar-
in the range of [7,760E-04 – 3,710E + 00] for scenario- ios-2 and scenario-3. The sub-task which has maximum
1, [6,000E-03 – 1,200E + 00] for scenario-2, and human error probability is the A2.1 in the scenario-
[2,600E-02 – 1,860E + 00] for scenario-3. Since the 1[3,71E + 00].
enclosed space entry procedures have a hierarchical Figure 3 specifies the risky sub-task for all scenar-
structure, a tree map is one of the best ways of visualiz- ios. According to Figure 3, A2.1, A1.6, A1.9, A1.4,
ing attributes of the scenarios using the size and colour A3.2, A1.15 and A2.2 are the sub-tasks that have
coding. Figure 2 represents the HEP results and enables the highest overall human error probability,
the user to spot patterns and exceptions. respectively.
8 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

Table 5. HEP results for each scenario. while the space is occupied, and for after work breaks?)
and A2.2 (Section 1 of this permit has been satisfacto-
Step Scenario-1 Scenario-2 Scenario-3 rily completed by the master or nominated responsible
1.1 1,01E-02 2,70E-01 6,75E-01 person.). Thus, human error is expected when the
1.2 3,03E-01 8,10E-01 5,70E-02 intended task is complex, time-consuming, and requir-
1.3 1,01E-02 6,80E-02 2,60E-02 ing a high level of comprehension and skill.
1.4 7,68E-01 5,41E-01 8,80E-01 Additionally, there are also relatively simple tasks
1.5 7,68E-01 2,91E-01 3,03E-01 that may be carried out quickly with inadequate atten-
1.6 1,50E-01 1,20E + 00 1,20E + 00
1.7 1,01E-02 1,19E-01 4,24E-01 tion. However, misperception of related risk in these
1.8 2,08E-01 1,00E-02 1,08E + 00 tasks would increase the human error probability. To
1.9 6,75E-01 5,41E-01 1,20E + 00 illustrate, subtask A2.1 (I have received instructions or
1.10 1,55E-01 3,90E-02 1,36E-01 permission from the master or nominated responsible
1.11 1,20E-01 3,47E-01 3,47E-01 person to enter the enclosed space.) is a fairly simple
1.12 8,62E-04 6,00E-03 2,70E-02
1.13 3,60E-03 1,70E-02 7,26E-01 task, but it has been identified as the riskiest sub-task
1.14 5,61E-03 8,80E-01 3,03E-01 since it is generally performed at speed with no real
1.15 5,82E-03 1,70E-02 1,86E + 00 idea of likely consequences. Operator experience and
2.1 3,71E + 00 6,08E-01 4,32E-01 familiarity are other crucial factors that affect the
2.2 3,89E-03 8,80E-01 8,80E-01 expected human error, regardless of the task require-
2.3 1,88E-03 1,41E-01 1,65E-01
2.4 1,54E-03 1,41E-01 1,65E-01 ments. Both sub-task A3.2, which is identified complex
2.5 3,60E-03 9,40E-02 9,40E-02 task, and A1.2, which is considered a fairly simple task,
2.6 5,61E-03 9,40E-02 6,90E-02 could be risky when the operator is unfamiliar with
3.1 7,76E-04 5,03E-01 3,10E-01 these sub-tasks (m-EPC1). Moreover, sub-task A1.14
3.2 1,44E + 00 3,10E-01 3,10E-01 (Is all equipment used in good working condition and
3.3 8,62E-04 6,00E-03 2,70E-02
3.4 5,82E-03 1,80E-02 1,00E-02 inspected prior to entry?) can be risky when the opera-
tor is inexperienced (m-EPC15). Lastly, performance
ambiguity also increases the human error probability in
sub-task A1.5 (Pre-entry atmosphere test readings).
Since the sub-tasks A1.4 (Has the space been tested Consequently, time shortage, operator inexperience,
and found safe for entry?) and A1.9 (Is rescue and spatial and functional incompatibility, misperception
resuscitation equipment available for immediate use by of risk, unfamiliarity, and performance ambiguity are
the entrance to the space?) are complex tasks requiring most effective m-EPCs that can affect human error
a high level of comprehension and skill, they have been during the enclosed space entry procedures.
identified as risky in all three scenarios. This is because
such tasks are more complicated and require more time
to perform correctly. Similar tasks are also found risky Hierarchy of control measures
in two scenarios. These are A1.6 (Have arrangements In the hierarchy of controls measures, the ways of con-
been made for frequent atmosphere checks to be made trolling risks range from maximum safety and reliability

Figure 2. A tree map for HEP results.

Soner and Celik 9

Figure 3. HEP result for each scenario.

to the lowest. However, the essential aim is always to tasks in the enclosed space entry procedures are simpli-
eliminate the risk, which is the most effective control fied and modified with an electronic permit, the human
measure. If this is not possible, it becomes essential to error probability will decline with this administrative
reduce the risk by working through other hierarchical control.
alternatives. Elimination and substitution as a control
measure would be practical when the advantage of
highly specialized tools and processes are integrated
Conclusion
into the enclosed space entry procedures. A drone- An enclosed space poses a serious risk to workers since
assisted survey is a pioneering example that can its atmosphere can be oxygen-deficient, polluted or con-
enhance the safety of seafarers.53,54 Sampling and mea- tain flammable and/or toxic gases or vapours. Hence, it
surement can be done while seafarers are outside an is important to keep in mind that an enclosed space
enclosed space in a drone-assisted survey. Thus, a should not be entered unless it is absolutely necessary.
drone-assisted survey is crucial to eliminate the related However, many processes often need the worker to be
risk, considering the fact that entering into enclosed present in an enclosed space. At this point, great atten-
space is required even before the atmosphere is checked. tion should be paid to create and maintain safe working
The Internet of Things (IoT) is a further technological conditions, when entering an enclosed space is required.
development that may improve the safety of sea- Despite implemented standards, regulations, and con-
farers.50,55 IoT can provide continuous atmosphere ducted research, accidents still occur in enclosed spaces
monitoring for an enclosed space and this would reduce onboard ships. To improve the safety of seafarers in an
the related risk. The human error probability associated enclosed space, this study has conducted a human reli-
with sub-tasks A1.1, A1.4, A1.5 and A1.6 would be ably analysis. The human errors have been identified as
reduced with these control measures. In addition to one of the main causes of enclosed space entry acci-
these, developing an electronic permitting system could dents. Hence, the study has investigated the human fac-
reduce the probability of human error. Since the elec- tor during an enclosed space entry operation via the
tronic permitting system will automate the whole pro- enhanced SOHRA method. Three scenarios have been
cess, it can be combined with the IoT.24,50 Accordingly, developed to assess HEP values. According to the
Permit Control System (PCS) can be developed. In the results, scenario-3 has the highest overall human error.
PCS, the measurements of an enclosed space atmo- Especially, A2.1, A1.6, A1.9, A1.4, A3.2, A1.15 and
sphere can be transferred to the electronic permit auto- A2.2 are the sub-tasks that have the highest overall
matically and master or a nominated responsible officer human error probability. Since these sub-tasks involve,
will be signed off on the permit electronically.56 In this for example, misperception of risk, and performance
manner, as an engineering control measure, seafarers ambiguity (A2.1), and temporal measurements (A1.6),
can be isolated from the hazard. Since complex sub- they pose significant challenges for crew members and
10 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

continuous attention is needed. In addition, these sub- 7. International Maritime Organization. Revised recommen-
tasks are required a high level of comprehension and dations for entering enclosed spaces aboard ships. London:
skills to perform without any error. Therefore, the International Maritime Organization, 2011.
related HEP values of these sub-tasks are high and 8. Safety of Life at Sea. Amendments to the international
expected to increase when the intended task is time- convention for the safety of life at sea (SOLAS), 1974,
Resolution MSC.350(92). London: International Mari-
consuming and complex. Accordingly, hierarchy of
time Organization, 2013.
control measures have been considered to improve the 9. International Maritime Organization. Early implementa-
level of operational safety in entering enclosed space. tion of SOLAS regulation XI-1/7 on atmosphere testing
Elimination (drone-assisted survey), substitution (IoT), instrument for enclosed spaces. London: International
engineering (electronic permit combined with IoT), and Maritime Organization, 2015.
administrative controls (electronic permit) are also rec- 10. Marine Accident Investigators International Forum. Pre-
ommended. Therefore, the study provides not only a liminary survey report on enclosed space incidents. Bram-
theoretical insight but also practical application to ley: Marine Accident Investigators International Forum,
improve the safety of seafarers. 2009.
Future research can be extended to deal with the 11. Paris MoU. Report of the 2015 concentrated inspection
campaign (CIC) crew familiarization for enclosed space
uncertainty of experts’ judgment in HEP calculation in
entry. The Hague, The Netherlands: Paris MoU, 2015.
the enhanced SOHRA. In this respect, fault tree analy- 12. Marine Accident Investigation Branch. Safety bulletin:
sis or fuzzy set theory can be considered to combine fatalities in enclosed spaces. Southampton: Marine Acci-
with the present model to determine the HEP values. dent Investigation Branch, 2008.
The proposed model is also applicable to other indus- 13. Marine Accident Investigation Branch. Safety bulletin:
tries (petrochemical, oil, refinery, gas, agriculture, con- entry of a confined space on board the cargo ship SUNTIS.
struction etc.) to conduct human error assessment in Southampton: Marine Accident Investigation Branch,
entering enclosed space operation when any required 2014.
adjustment is performed. 14. Indian Register of Shipping. Technical Circular (No.:
121/2017). 2017. Mumbai: Indian Register of Shipping
(IRCLASS).
Declaration of conflicting interests 15. Nano G and Derudi M. Evaluation of workers accidents
The author(s) declared no potential conflicts of interest through risk analysis. Chem Eng 2012; 26: 495–500.
with respect to the research, authorship, and/or publi- 16. Celik M and Cebi S. Analytical HFACS for investigating
human errors in shipping accidents. Accid Anal Prev
cation of this article.
2009; 41(1): 66–75.
17. Roberts SE, Nielsen D, Kot1owski A, et al. Fatal acci-
Funding dents and injuries among merchant seafarers worldwide.
Occup Med 2014; 64(4): 259–266.
The author(s) received no financial support for the
18. Mohana Krishnan G and Menon M. Risk assessment for
research, authorship, and/or publication of this article. entering enclosed spaces on board ships. ARPN J Eng
Appl Sci 2017; 12(4): 1092–1097.
ORCID iD 19. Sundal MK, Lilleng PK, Barane H, et al. Asphyxiation
death caused by oxygen-depleting cargo on a ship. Foren-
Omer Soner https://orcid.org/0000-0002-8100-7874 sic Sci Int 2017; 279: e7–e9.
20. International Chamber of Shipping (ICS). Tanker safety
References guide (Chemicals), 4th ed. 2014.
21. Burlet-Vienney D, Chinniah Y and Bahloul A. The need
1. Riedel SM and Field WE. Summation of the frequency,
for a comprehensive approach to managing confined
severity, and primary causative factors associated with space entry: summary of the literature and recommenda-
injuries and fatalities involving confined spaces in agri- tions for next steps. J Occup Environ Hyg 2014; 11(8):
culture. J Agric Saf Health 2013; 19(2): 83–100. 485–498.
2. Dorevitch S, Forst L, Conroy L, et al. Toxic inhalation 22. Burlet-Vienney D, Chinniah Y, Bahloul A, et al. Design
fatalities of US construction workers, 1990 to 1999. J and application of a 5-step risk assessment tool for con-
Occup Environ Med 2002; 44(7): 657–662. fined space entries. Saf Sci 2015; 80: 144–155.
3. International Maritime Organization. Recommendations 23. Burlet-Vienney D, Chinniah Y, Bahloul A, et al. Risk
for entering enclosed spaces aboard ships. London: Inter- analysis for confined space entries: critical analysis of four
national Maritime Organization, 1997. tools applied to three risk scenarios. J Occup Environ Hyg
4. Spencer C. A master’s guide to enclosed space entry. Lon- 2016; 13(6): 99–108.
don: The Standard P&I Club, 2012. 24. Botti L, Bragatto P, Duraccio V, et al. Adopting IOT
5. The Shipowners’ Club. Loss prevention – enclosed space technologies to control risks in confined space: a multi-
entry. Luxembourg: The Shipowners’ Mutual Protection criteria decision tool. Chem Eng Trans 2016; 53: 127–132.
and Indemnity Association, 2016. 25. Botti L, Duraccio V, Gnoni M, et al. An integrated holis-
6. Nielsen D and Panayides PM. Causes of casualties and tic approach to health and safety in confined spaces. J
the regulation of occupational health and safety in the Loss Prev Process Ind 2018; 55: 25–35.
shipping industry. WMU J Marit Affairs 2005; 4(2): 26. Naghavi Z, Mortazavi SB and Hajizadeh E. Exploring
147–167. the contributory factors of confined space accidents using
Soner and Celik 11

accident investigation reports and semi structured inter- 42. Saaty TL. How to make a decision: the analytic hierarchy
views. Saf Health Work 2019; 10(3): 305–313. process. Eur J Oper Res 1990; 48(1): 9–26.
27. Wilson MP, Madison HN and Healy SB. Confined space 43. American Industrial Hygiene Association (AIHA). Occu-
emergency response: assessing employer and fire depart- pational health and safety management systems. ANSI/
ment practices. J Occup Environ Hyg 2012; 9(2): 120–128. AIHA Z10-2005. Virginia, ABD: American Industrial
28. Botti L, Ferrari E and Mora C. Automated entry technol- Hygiene Association, 2005.
ogies for confined space work activities: a survey. J Occup 44. Manuele F. A. Risk assessment & hierarchies of control.
Environ Hyg 2017; 14(4): 271–284. Prof Saf 2005; 50(5): 33–39.
29. Botti L, Mora C and Ferrari E. A methodology for the 45. European Industrial Gases Association. Hazards of oxygen
identification of confined spaces in industry. In: 4th inter- deficient atmospheres (Doc 44/18; Revision of Doc 44/09).
national conference on sustainable design and manufactur- Brussels: European Industrial Gases Association, 2018.
ing, SDM, Bologna, Italy, 26–28 April 2017, pp.701–709. 46. Khan FI. Use maximum-credible accident scenarios for
30. Suruda AJ, Pettit TA, Noonan GP, et al. Deadly rescue: realistic and reliable risk assessment. Chem Eng Prog
the confined space hazard. J Hazard Mater 1994; 36(1): 2001; 97(11): 56–65.
45–53. 47. Arunraj N and Maiti J. Risk-based maintenance—
31. Selman J, Spickett J, Jansz J, et al. Confined space rescue: techniques and applications. J Hazard Mater 2007;
a proposed procedure to reduce the risks. Saf Sci 2019; 142(3): 653–661.
113: 78–90. 48. Khan FI and Abbasi SA. A criterion for developing cred-
32. Akyuz E, Celik M and Cebi S. A phase of comprehensive ible accident scenarios for risk assessment. J Loss Prev
research to determine marine-specific EPC values in Process Ind 2002; 15(6): 467–475.
human error assessment and reduction technique. Saf Sci 49. Nivolianitou ZS, Leopoulos VN and Konstantinidou M.
2016; 87: 63–75. Comparison of techniques for accident scenario analysis
33. Williams JC. A data-based method for assessing and in hazardous systems. J Loss Prev Process Ind 2004;
reducing human error to improve operational perfor- 17(6): 467–475.
mance. In: Proceedings of IEEE 4th conference on 50. Botti L, Duraccio V, Gnoni MG, et al. A framework for
human factor and power plants. Monterey, CA, 1988, preventing and managing risks in confined spaces
pp.436–453. through IOT technologies. In: Proceedings of the 25th
34. Akyuz E, Celik M, Akgun I, et al. Prediction of human European safety and reliability conference, Glasgow, Scot-
error probabilities in a critical marine engineering opera- land, 25–29 September 2016, pp.3209–3217.
tion on-board chemical tanker ship: the case of ship bun- 51. Riahi R, Bonsall S, Jenkinson I, et al. A seafarer’s relia-
kering. Saf Sci 2018; 110: 102–109. bility assessment incorporating subjective judgements.
35. Kirwan B. Human reliability analysis of an offshore Proc IMechE, Part M: J Engineering for the Maritime
emergency blowdown system. Appl Ergon 1987; 18(1): Environment 2012; 226(4): 313–334.
23–33. 52. Riahi R, Li K, Robertson I, et al. A proposed decision-
36. Kirwan B and Gibson H. CARA: a human reliability making model for evaluating a container’s security score.
assessment tool for air traffic safety management— Proc IMechE, Part M: J Engineering for the Maritime
technical basis and preliminary architecture. In: Proceed- Environment 2014; 228(1): 81–104.
ings of the fifteenth safety-critical systems symposium (ed 53. Rao B, Gopi AG and Maione R. The societal impact of
Felix RedmillTom Anderson), Bristol, UK, 13–15 Febru- commercial drones. Technol Soc 2016; 45: 83–90.
ary 2007, pp.197–214. London: Springer. 54. Subchan M. FSO cargo tank inspection with Drone, a
37. Shepherd A. Hierarchial task analysis. 1st ed. London: lesson learned for future implementation. In: Proceeding,
Taylor & Francis, 2001, p.115. Indonesian petroleum association, forty-third annual con-
38. Saaty TL. The analytical hierarchy process, planning, pri- vention & exhibition, Jakarta, Indonesia, September 2019.
ority, resource allocation. New York: RWS Publication, 55. Riaz Z, Arslan M, Kiani AK, et al. CoSMoS: a BIM and
1980. wireless sensor based integrated solution for worker
39. Forman EH. Facts and fictions about the analytic hierar- safety in confined spaces. Autom Constr 2014; 45:
chy process. Math Comput Model 1993; 17(4–5): 19–26. 96–106.
40. Saaty TL. Axiomatic foundation of the analytic hierarchy 56. Wagner DD. Confined space entry goes wireless: testing,
process. Manag Sci 1986; 32(7): 841–855. monitoring & permitting will never be the same. Indus-
41. Vargas LG. Reciprocal matrices with random coeffi- trial Safety & Hygiene News (ISHN), https://www.ishn.
cients. Math Model 1982; 3(1): 69–81. com (2018, accessed 10 November 2019).