close

Вход

Забыли?

вход по аккаунту

?

j.oceaneng.2017.09.044

код для вставкиСкачать
Ocean Engineering xxx (2017) 1–8
Contents lists available at ScienceDirect
Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng
The use of a virtual environment in managing risks associated with human
responses in emergency situations on offshore installations
Norafneeza Norazahar a, c, Jennifer Smith b, Faisal Khan a, *, Brian Veitch b
a
Centre for Risk, Integrity and Safety Engineering, Faculty of Engineering & Applied Science, Memorial University of Newfoundland, A1B 3X5, St. John’s, Newfoundland
and Labrador, Canada
Faculty of Engineering & Applied Science, Memorial University of Newfoundland, A1B 3X5, St. John's, Newfoundland and Labrador, Canada
c
Centre of Hydrogen Energy, Institute of Future Energy, Universiti Teknologi Malaysia, 81310 Johor Bahru, Malaysia
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
Emergency situation
Harsh environment
Human responses
Offshore installation
Risk management
This paper presents the use of a virtual environment for investigating risks associated with human responses in
emergency situations on offshore installations. The virtual environment and the risk management can be used to
analyze risks associated with human responses by investigating the effectiveness of existing safety measures,
identifying areas of improvement, and proposing new designs for safety measures.
Problems: Dealing with emergency scenarios on offshore installations is a tremendous undertaking. The interaction
of personnel using the safety measures in emergency situations can be affected by hazards, environment conditions, malfunctioning equipment, and inadequate emergency preparedness. Such factors have the potential to
prevent personnel from arriving at a safe area, increase the level of risks, and consequently, cause injuries or
fatalities to personnel. Risks associated with human responses in emergency scenarios are often unforeseen due to
difficulties with modeling realistic emergency scenarios.
Objective: The objective of the research is to study demonstrate the use of virtual environments for investigating
and managing risks associated with human responses in emergencies.
Method: Risk management is employed to assess and manage risks associated with human responses in emergency
scenarios. The risk management is tested using experimental data collected from past studies of human responses
in virtual environments.
Results: Risks associated with human responses during emergency scenarios are determined by safety measures,
the environment, and the egress route choices that have been taught. Participants’ performance and interaction
with improved safety measures are better than the performance of participants without safety measures.
Conclusion: This paper provides a demonstration of the utility of virtual environments to assess risks associated
with human responses in emergency situations.
Application: The findings of this study may be useful for offshore energy operations.
1. Introduction
The organization or operator of offshore installations should prioritize the emergency response plan and safety barriers for escape in
emergency situations. Safety measures for escape can include an alarm
system, primary and alternative escape routes, muster stations, and
personal protective equipment (e.g., Health and Safety Executive [HSE],
1997; Canadian Association of Petroleum Producers [CAPP], 2010). For
offshore installations operating in harsh environmental conditions, both
escape routes and muster stations can be heat traced and fully enclosed to
protect individuals when performing escape activities in cold
temperatures (Eikill et al., 2007).
The organization must ensure that personnel practice emergency
drills to familiarize themselves with the equipment and procedures, and
identify limitations, potential hazards, and risks in performing escape
from hazardous areas. The challenges and risks of performing escape
depend on an individual's skills and experience, teamwork, procedures,
roles and responsibilities, communication, as well as the emergency
response plan, environment conditions, and reliability of emergency
equipment (HSE, 2007; CAPP, 2010a, 2010b). All of these factors influence the effectiveness of safety barriers, the success of escape operations,
and the safety of individuals should an emergency occur.
* Corresponding author.
E-mail address: fikhan@mun.ca (F. Khan).
https://doi.org/10.1016/j.oceaneng.2017.09.044
Received 9 April 2017; Received in revised form 27 August 2017; Accepted 24 September 2017
Available online xxxx
0029-8018/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Norazahar, N., et al., The use of a virtual environment in managing risks associated with human responses in
emergency situations on offshore installations, Ocean Engineering (2017), https://doi.org/10.1016/j.oceaneng.2017.09.044
N. Norazahar et al.
Ocean Engineering xxx (2017) 1–8
probability based on a normal distribution (Modarres, 2006). This paper
assumes a normal distribution to simplify the calculations.
The probability is calculated considering a normal distribution
function. A normal density function in Equation (1).
The potential of hazards cannot be eliminated totally. Emergency
situations in the presence of hazards can worsen when personnel fail to
interact with emergency equipment and follow procedures consistently.
The effects of fires, explosions, and poor environmental conditions can
cause failures of both personnel responses and the escape. The probability of failure will increase further when escape is performed in adverse
weather conditions or harsh environments, such as in the presence of ice
and snow, cold temperatures, darkness, or strong winds (Mould, 2001;
Kjellen, 2007).
There are many studies that have introduced or proposed effective
tools and techniques as safety measures in emergency situations on
offshore installation. DiMattia et al. (2005) and Deacon et al. (2010)
proposed prevention and mitigation barriers in risk management
focusing on personnel performing escape, evacuation and rescue (EER)
activities. Andersen and Mostue (2012) presented integrated operations
(IO) on risk management approaches using real-time data, collaborative
techniques, and multiple expertise in making better decisions and
implementations for the Norwegian oil and gas industry.
Simulation-based studies have been used before to investigate human
performance in emergency scenarios. Doheny and Fraser (1996)
modelled an individual making decisions in an emergency scenario on
offshore installations while considering the offshore facility relatively
sparsely populated. In a study of individuals' behaviours conducted by Ha
et al. (2014), walking direction was determined based on destination,
obstacles, and visibility during emergency scenarios. Kim et al. (2004)
integrated both individuals’ behaviours and a dynamic model in the
study of evacuation during marine accidents. Joo et al. (2013) investigated dynamic individual actions interacting with emergent hazards,
such as fire, using an agent-based simulation. Duarte et al. (2014) used a
virtual environment to investigate dynamic versus static signs on human
behaviour during emergency evacuation. Augustijn-Beckers et al. (2010)
studied pre-evacuation behaviour and exit choice during emergency
scenarios involving a group of individuals. Using a virtual environment of
an offshore installation, Bradbury-Squires (2013) assessed efficacy of
active and hybrid learning modes on task performance during emergency
training. Smith (2015) and Smith et al. (2015) studied the effectiveness
of simulation training on individual’s competency and learning during
emergency situations. Colombo and Golzio (2016) introduced a
simulation-based approach to training teams, including operators and
managers, in making decisions and increasing their competencies as a
team in critical situations.
Poor performance or lack of response in emergency situations can
result in injuries and fatalities to personnel. There is a need to reduce and
manage risks associated with personnel performance in emergency situations on offshore installations. The objective of the present work is to
demonstrate the use of virtual environments for investigating and managing risks associated with human responses in emergencies. Engineered
and procedural safety measures are studied and discussed in this paper
(see Sections 3.1.2 and 3.2.2). The approach uses experimental data of
human responses obtained from previous studies that used virtual environments (Duarte et al., 2014; Smith, 2015).
Section 2 describes the risk calculation and its formulation. Section 3
explains experimental studies of human responses using virtual environments. Sections 4 and 5 present and conclude the risk management study.
2
1
ðx μÞ
f ðx; μ; σÞ ¼ pffiffiffiffiffiffiffiffiffiffi exp 2σ 2
2πσ 2
(1)
where
x is the passing score;
μ is the average of overall scores;
σ is the standard deviation of scores;
σ2 is the variance in the distribution.
The next step is to estimate the risks associated with the human responses by comparing two different failure probabilities; one is the
emergency situation equipped with safety measures, and the other is the
emergency situation without safety measures.The calculation of the
change in risk is formulated as shown in Equation (2).
ΔRISK ¼ Probability without safety measures
Probability with safety measures
(2)
Confidence intervals are employed to verify the calculation of probabilities and risks (Kumamoto and Henley, 1996). Equation (3) is a formula for determining confidence interval. The sample size, mean, and
standard deviation are used to calculate the confidence interval of the
mean. The calculation is based on a normal distribution and a 95 percent
confidence interval. The results are presented in Section 4.4.
pffiffiffi CI ¼ μ ±α σ n
(3)
where
α is the desired confidence level, and
n is the sample size.
3. Case studies of emergency scenarios in virtual environment
Data from two published experimental studies of different virtual
environments have been selected to provide data for this risk management study. The first experimental study, entitled ‘The effect of virtual
environment training on participant competence and learning in offshore
emergency egress scenarios’, is the source of data on human responses in
an emergency scenario on an offshore installation (Smith, 2015). The
second experimental study, entitled ‘Behavioural compliance for dynamic versus static signs in an immersive virtual environment’, is the
source of data on behavioural compliance with signage (Duarte
et al., 2014).
Both experimental studies were conducted using virtual environments (VE) with the purpose to observe human responses and behaviours
during emergency conditions (Duarte et al., 2014; Smith, 2015; Smith
et al., 2015). Simulating emergency conditions in the VE can provide a
safe medium for participants to acquire artificial experience, which is
otherwise impractical and risky to obtain in a real situation. Details of
emergency scenarios in the VE are explained in Sections 3.1 and 3.2.
2. Calculation of risks
3.1. Offshore emergency egress scenario on an offshore installation
In this paper, risk is assessed with regard to the probability of failures
only. In case studies used here, we treat consequences of failure as neutral
(i.e. risk is proportional to probability only).
The probability of failures is calculated by considering the performance score in the emergency scenarios from two experimental studies,
as further explained in Section 3. The performance score is analysed to
determine the mean and standard deviation (Duarte et al., 2014; Smith,
2015). Information on mean and standard deviation is used to calculate a
Smith (2015) studied the effectiveness of simulation training on individual’s competency and learning during emergency situations using
the All-hands Virtual Emergency Response Trainer (AVERT) software.
The layout in AVERT includes accommodations, and a muster station and
lifeboat station, both located on the main deck. Three routes were provided as egress routes: the primary route characterized as an interior
2
N. Norazahar et al.
Ocean Engineering xxx (2017) 1–8
knowledge of the platform, specifically their understanding of the layout
and egress routes. In this paper, the wayfinding is assessed by considering
a) the route selection (primary, secondary, and tertiary routes), b) the
arrival at the correct muster or lifeboat station, and c) incorrect deviations along the route. The participants’ performance is compared to
responses during two different emergency scenarios, which are a) in a
normal lighting condition (TA1) and b) a blackout scenario (TA3).
route with an inside stairwell, and secondary and tertiary routes characterized as exterior routes with an outside stairwell.
In the study (Smith, 2015), the simulation scenarios begin with the
activation of an emergency alarm (General Platform Alarm) that requires
all personnel onboard to move to a muster station using one of the
designated escape routes. In the case of an escalating event, an evacuation alarm (Prepare to Abandon Platform Alarm) is triggered to notify
personnel to muster directly at the lifeboat station. Hazards such as
blackouts, fire, and smoke were implemented in AVERT to create credible emergency scenarios. The emergency scenario with the presence of
hazards required participants to find a safe route to the muster or lifeboat
station by avoiding the escape routes blocked by hazards (Smith, 2015).
ii) Competency of participants
Smith (2015) defined competence as the demonstration of knowledge
related to alarm recognition, routes and mapping, and hazards avoidance, which participants gained in the training tutorials. The participants
were evaluated based on their performance in recognizing types of alarm,
re-routing and taking the safest routes, avoiding hazards on route, and
arriving at the correct muster or lifeboat station (Smith, 2015). Participant’s competency is considered as procedural safety. This paper uses
evaluation of participants’ competency in two emergency scenarios (TH1
and TH2, as presented in Table 1) with escalating events that required
them to re-route (Smith, 2015).
The criteria in the study performed by Smith (2015) are selected as a
subset for assessing and managing risks associated with human responses. Table 2 lists the subset criteria according to emergency scenario
i) equipped with the engineered safety measures, and ii) required
knowledge of procedural safety.
3.1.1. Participants of study
In the study conducted by Smith (2015), thirty-six volunteers
participated. The participants’ ages ranged between 19 to 55 years old.
Smith (2015) distributed a video game experience questionnaire to
participants prior to dividing them into two groups that were balanced in
terms of video game experience. In the experiment, the amount of
practice the participants received differed according to the group. Participants in Group 1 had repeated training. Participants in Group 2 had a
single exposure to training (Smith, 2015; Smith et al., 2015). The 17
participants in Group 1 reviewed training tutorials and repeated AVERT
practice scenarios in preparation for the test scenarios. The 19 participants in Group 2 received the initial tutorial training and no AVERT
practice scenarios.
3.1.3. Score for performance in test scenarios
This paper uses a different scoring scheme than that used by Smith
(2015), because only a subset of Smith’s criteria are used here. Performance scores for the test scenarios in TA1 and TA3 are based on route
selection and arrival at the correct location according to the type of
alarm. Both TA1 and TA3 have 15 points as the total performance scores
(see Table 3). Test scenarios in TH1 and TH2 have performance scores
based on route selection, re-routing when the route has been blocked,
hazard avoidance, and arrival at the correct location. Both TH1 and TH2
have 35 points as the total performance scores.
3.1.2. AVERT emergency response test scenarios
Of 12 emergency scenarios in the study done by Smith (2015), four
(4) emergency scenarios in AVERT were selected for this paper, as listed
in Table 1. The four scenarios provide information related to engineered
and procedural safety measures. Normal emergency lighting is categorized as an engineered safety measure for emergency scenarios. Emergency response procedures are the procedural measures used in the
study. These relate to the participants’ competence to respond to emergency scenarios. This paper uses different performance scoring criteria
than presented in Smith (2015). Details of performance scores are
described in Section 3.1.3.
Emergency lighting provided in the event of emergency is considered
as an engineered safety measure. The participants’ performance is
compared to responses during two different emergency scenarios: a) in a
normal lighting condition (TA1) and b) a blackout scenario (TA3). According to Smith (2015), wayfinding reflects the participants' spatial
3.1.4. Mean and standard deviation of emergency response test scenarios
The performance scores are used to calculate mean and standard
deviations for each test scenario. This is followed by calculating the
probability of failure based on the mean and standard deviations
considering a normal distribution. The probability of failure shows the
numbers of actions failed according to the subset criteria in Table 3. The
probability of failure is shown in the range between 0 and 1 in this paper.
Table 4 presents the mean, standard deviation, and probability of failure
for the test scenarios of TA1 and TA3. The mean, standard deviation, and
probability of failure in test scenarios TH1 and TH2 are listed in Table 5.
The mean and standard deviation in Tables 4 and 5 are scores out of the
total performance scores.
Table 1
Description of emergency scenarios designed in the AVERT (Smith, 2015).
i) Wayfinding in normal lighting conditions and blackout scenario
Scenario
label
Scenario description
TA1
The participants are required to respond to a general platform alarm
(GPA) and find a way from their accommodation to their primary
muster station.
The participants are required to respond to a prepare to abandon
platform alarm (PAPA) and find a way from their accommodation to
their lifeboat station in a blackout scenario due to equipment failure.
The participants are required to respond to a GPA because there is fire
in the galley. The emergency scenario escalates and causes a PAPA
activation. In response to the GPA, participants must go to a primary
muster station from their accommodation. When the alarm changes to
PAPA, the participants must change their route and head to a lifeboat
station. Both the primary route and muster station have been blocked
by the effects of the fire.
The participants are required to respond to a GPA because there is a fire
on the helideck. The emergency scenario escalates due to explosion and
heavy smoke, causing a PAPA activation. The task required the
participants to go to their primary muster station from their
accommodation and change their route to head to the lifeboat station.
The secondary route has been blocked by the effects of the heavy
smoke.
TA3
TH1
TH2
Table 2
Subset criteria for assessing risks associated with human reponses (Smith, 2015).
Criteria
Scenarios
Criteria for
calculating
risks
Types of scenarios
Wayfinding in different
conditions
Competency
- TA1 and TA3
- Take
primary,
secondary, or tertiary
route,
- No change of route
from one to another
route, and
- Arrive at the correct
location.
- TH1 and TH2
- Take primary, secondary, or tertiary
route,
- Re-route when the route has been
blocked or affected by the effects of
fires and explosions,
- Avoid hazards, and
- Arrive at the correct location.
3
N. Norazahar et al.
Ocean Engineering xxx (2017) 1–8
Table 3
Performance scores based on the subset criteria (Smith, 2015).
Criteria
Scenarios
Route selection
TA1
TA3
TH1a
TH2b
Re-route
TH1a
TH2b
Avoid hazard
exposure
TH1a
TH2b
Reach correct
location
TA1
TA3
TH1
TH2
a
b
Criteria for behavioural measure
Score
i) Follows a designated route (selecting the primary, secondary or tertiary route and remaining on the same route until arriving at the
correct location)
OR
ii) Minor off route (selecting the primary, secondary or tertiary route but experiencing minor deviations from one of the three route
options)
OR
iii) Major off route, lost behaviour, and/or fails to arrive at the correct location.
iv) Follows the safest designated route for the situation (selecting secondary or tertiary route and remaining on the same route until
arriving at correct location)
OR
v) Minor off route (selecting the secondary or tertiary route but with minor deviations from the route)
OR
vi) Major off route, lost behaviour, and/or fails to arrive at the correct location.
vii) Follows the safest designated route for the situation (selecting the primary route and remaining on the same route until arriving at
the correct location).
OR
viii) Minor off route (selecting the primary route but with minor deviations from the route).
OR
ix)Major off route, lost behaviour, and/or fails to arrive at the correct location.
x) Changed route selection based on new information from listening to the alarm change and PA announcement.
OR
xi) Not changing route selection after hearing the alarm change but changing route selection after encountering the hazard blocking
the route.
OR
xii) Not changing route selection after encountering the hazard blocking the route.
xiii) Participant does not encounter a hazard along their route or experience exposure to any hazards.
OR
xiv) Participant encounters a hazard and is exposed to the hazard along their egress route.
xv) Arrives at the correct location for the situation (muster station or lifeboat station).
OR
xvi)Fails to arrive at the correct location for the situation (muster station or lifeboat station).
15
points
7.5
points
0 point
10
points
5 points
0 point
10
points
5 points
0 point
15
points
10
points
0 point
10
points
0 point
Pass
Fail
End of primary route has been blocked by hazards.
Beginning of secondary and tertiary routes have been blocked by hazards.
3.2.1. Participants of study
A total of 90 participants consisting of university students were
involved in the experimental study. Thirty (30) participants were
assigned to each of the following groups according to the different types
of exit signs: a) a minimal design, b) in a static configuration, and c) in a
dynamic configuration (Duarte et al., 2014).
Table 4
Mean, standard deviation, and probability of failure in TA1 and TA3.
Test
scenarios
Mean (out of 35
points)
Standard deviation (out of
35 points)
Probability of
failures
TA1
TA3
13.23
11.91
4.22
5.96
0.66
0.70
3.2.2. Types of exit signage
In the study conducted by Duarte et al. (2014), available egress routes
were marked by exit signs consisting of an arrow and a running figure in a
doorway. The experiment varied the number of exit signs available and
the type of exit signs (static and dynamic signs). Three different types of
exit signs are described in Table 6.
The objective of the study was to investigate human behaviour in
complying with exit signs. The participants were expected to move toward the exit door following the exit signs in order to evacuate the
building safely. They were given scores for the performance of safe
3.2. Behavioural compliance for dynamic versus static signs in a building
evacuation
Researchers (Duarte et al., 2014) used a virtual environment known
as ErgoVR to investigate dynamic versus static signs on human behaviour
during emergency evacuation. ErgoVR simulated a building consisting of
four (4) rooms: meeting room, laboratory, cafeteria, and warehouse. The
walls of the rooms and hallway have safety signs and exit signs. The
experiment required participants to go to every room and look for instructions for the given tasks in the scenario. There was an emergency
scenario involving an explosion followed by a fire when the participants
entered a warehouse. The fire alarm was triggered due to the explosion
and fire in the VE. All corridors except the exit route were affected by the
hazard and blocked by flames and smoke. The emergency scenario
required participants to follow the exit signs in order to safely evacuate
the building.
Table 6
Description of exit signs used in the virtual environment (Duarte et al., 2014).
Type of
signage
Description
Minimal exit
signs
No exit sign provided in the VE. The evacuation scenario has labels
for the functions of buttons only. One example of buttons is fire
alarm siren.The scenario with minimal design of exit signs is
assigned as a baseline with the purpose to assess the impact of exit
signs on behavioural compliance (Duarte et al., 2014).
The exit route in the VE is equipped with color printed exit signs.
The color of the exit signs is designed according to the standard in
the International Organization for Standardization for safety colours
and safety signs.
The exit signs in the VE are designed to have five (5) flashing lights
in an orange color and an alarm ‘beep’ sound activated or deactivated by sensors.
Table 5
Mean, standard deviation, and probability of failure for TH1 and TH2.
Test
scenarios
Groups
Mean (out of
35 points)
Standard deviation (out
of 35 points)
Probability of
failures
TH1
1
2
1
2
12.35
18.68
26.18
16.32
15.52
16.65
15.16
17.70
0.93
0.84
0.72
0.85
TH2
Static exit
signs
Dynamic exit
signs
4
N. Norazahar et al.
Ocean Engineering xxx (2017) 1–8
evacuation (Duarte et al., 2014).
3.2.3. Scores for behavioural measures
The behaviour and performance of participants were assessed according to specific criteria (Duarte et al., 2014). This paper uses
complying with exit signage as one criteria to give six (6) points as the
total performance score. Table 7 describes the criteria and scores for
participants’ behavioural compliance in the simulation.
3.2.4. Mean and standard deviation of behavioural simulation
Using the results reported by Duarte et al. (2014), the participants’
performance scores were calculated to get the values of the mean and
standard deviation for each type of exit sign provided in the simulation.
This was followed by calculating the probability of failure based on the
mean and standard deviation assuming a normal distribution. The mean,
standard deviation, and probability of failure are listed in Table 8.
Fig. 1. Risks of wayfinding error in blackout scenario.
4. Risk management of in emergency scenarios
4.2. Competence
Sections 4.1, 4.2, and 4.3 present the results of data analysis for the
following factors: wayfinding in normal lighting conditions compared to
blackout scenario, competency of participants in emergency offshore
evacuation, and behavioural compliance with exit signs, respectively.
The risk management in emergency scenarios are categorized into i)
engineered safety measures, which are normal lighting and exit signage,
and ii) procedural safety, which reflects participants’ competence.The
results in Sections 4.1–4.3 are supported by confidence intervals in
Section 4.4.
This paper assigns scenario TA1 as a baseline for assessing the performance of participants in emergency situations. The probability of
failure in TA1 is 0.66.
4.2.1. Emergency scenario requiring hazards avoidance and re-routing
(TH1)
Based on performance scores in this paper, the probability of failure
in Group 1 in TH1 is 0.93. Of 17 participants in Group 1, four (4) participants re-routed from the primary to the secondary egress route after
hearing the evacuation alarm (PAPA) and three (3) participants changed
their route from the primary to the secondary egress route after
encountering the hazards.
The difference in probabilities between TA1 and TH1 is denoted as
ΔRiskGroup1 with a value of 0.27. The risk is shown in Fig. 2. The performance in TA1 and TH1 show that participants preferred to use the
primary route as their main means to the muster and lifeboat stations.
Referring to performance scores in Table 3, the probability of failure
in Group 2 in TH1 is 0.84. The comparison of the probabilities of failure
between TH1 and TA1 results in a ΔRiskGroup2 of 0.17, as shown in Fig. 2.
One potential factor contributing to the risk is a small number of participants who did not change the route from the primary to the secondary
egress route even after they heard the evacuation alarm (PAPA) and
failed to arrive at the lifeboat station.
4.1. Impact of lighting condition and blackout scenario on wayfinding
Two scenarios from Smith’s study (2015) were used to look at the
effect of visibility on the risk associated with making wayfinding errors.
TA1 represents a normal lighting condition. TA3 represents a blackout
condition. The change in risk represents the difference between the
probabilities of failure for the blackout and normal lighting scenarios
(denoted as ΔRiskBlackout). As indicated in Fig. 1, the risk is 0.04. The risk
of making wayfinding errors is high due to the difference between participants’ route selection in the two scenarios. Participants showed a
tendency to select the primary route in normal lighting conditions.
However, during the blackout scenario, participants seemed to use any of
the available routes (i.e. primary, secondary, and tertiary).
4.2.2. Emergency scenario requiring hazards avoidance and re-routing
(TH2)
Based on the performance scores used in this paper, the probability of
Table 7
Scores for the performance of participants in the simulation (Duarte et al., 2014).
Signage
Criteria for behavioural measures
Scores
Exit
signs
i) Number of times the participant
moved in the direction shown by
the sign or toward the exit way
correctly.
ii) A decision is determined by
participants crossing a square of
2 2 m.
iii) Any small movement is considered
a hesitation and no score is given.
Make six (6) correct decisions
will give a score of 100 percent
success.
Table 8
Mean, standard deviation, and probability according to types of signage.
Types of
signage
Mean (out of 6
points)
Standard deviation (out of
6 points)
Probability of
failures
Minimal
sign
Static sign
Dynamic
sign
2.90
1.40
0.25
4.60
5.23
1.69
1.38
0.06
0.01
Fig. 2. Risks of emergency situations in TH1.
5
N. Norazahar et al.
Ocean Engineering xxx (2017) 1–8
failure in Group 1 in test scenario TH2 is 0.72. Comparing the data of TA1
(as a baseline) to assess performance in TH2, it is found that ΔRiskGroup1 is
0.06 as presented in Fig. 3. The potential factor contributing to the risk
value is four (4) participants who failed to re-route after hearing the PAPA
alarm or after encountering the hazards. There was also one participant
who became lost and therefore failed to reach at the lifeboat station.
The probability of failure in Group 2 in TH2 is 0.85. The difference in
probability of failure between TA1 and TH2 is denoted as ΔRiskGroup2 ¼
0.19 as shown in Fig. 3. Some participants (Smith, 2015) used the secondary egress route as a way to the lifeboat station no matter what the
circumstances. These participants did not re-route even after they heard
the PAPA alarm and after they encountered the hazard.
4.3. Behavioural compliance with exit signs (Duarte et al., 2014)
4.3.1. Minimal signs as a baseline
Exit signs with a minimal design were used as a baseline in the case
study (Duarte et al., 2014). The probability of failure to follow minimal
design of exit signs is 0.25, the highest probability of failure, as shown
in Table 8.
Fig. 3. Risks of emergency situations in TH2.
4.3.2. Static exit signs and dynamic exit signs
In Fig. 4, ΔRiskStatic represents the difference in probability of failure
between minimal exit signs and static exit signs. It has a value of 0.19.
This change in risk shows that the influence of static exit signs resulted in
better performance than a group of participants using minimal exit signs
during emergency (Duarte et al., 2014).
In the experimental study, the probability of failure of participants to
take egress routes following the direction shown by a dynamic exit signs
is 0.01. Using the minimal design of exit signs as a baseline (with the
probability of 0.25), the comparison of participants' behavioural
compliance with dynamic exit signs is denoted as ΔRiskDynamic. As shown
in Fig. 4, ΔRiskDynamic has a value of 0.24. The dynamic exit signs had
more impact on participants’ decision to take egress routes during
emergency than the static exit signs.
Fig. 4. Risks of minimal design of exit signage.
4.4. Confidence interval
The confidence interval can show the uncertainty level. The confidence interval is expressed as the upper and lower limits for the mean
performance scores. A small difference between the upper and lower
limits indicates the less uncertainty.
4.4.1. Confidence interval for test scenarios in TA1 and TA3
Fig. 5 shows the confidence interval for AVERT emergency responses
in TA1 and TA3. Table A1 lists the mean and both the upper and lower
limits for the test scenarios TA1 and TA3. The differences between the
upper and lower limits of TA1 and TA3 are 4.01 and 5.67, respectively.
This indicates that the mean performance scores in TA1, based on the
performance scores used in this paper, is better than the mean performance scores in TA3.
4.4.2. Confidence interval for TH1
Referring to Table A1, the differences between the upper and lower
limits of Groups 1 and 2 are 14.76 and 14.98, respectively. As illustrated
in Fig. 6, the uncertainty levels of of the means for Groups 1 and 2 are
considered high.
Fig. 5. Upper and lower limits for the performance of participants in TA1 and TA3.
of exit signs is demonstrated in Fig. 8. Referring to Table A2, the difference between the upper and lower limits of the minimal, static, and
dynamic signs are 1.00, 1.21, and 0.99, respectively. Dynamic signage
has the highest mean value and the lowest confidence interval compared
to the other two types of exit signs. This reflects that many participants
took the exit route as directed by the dynamic signs.
4.4.3. Confidence interval for TH2
Fig. 7 presents the confidence interval of the mean for TH2. As shown
in Table A1, the difference between the upper and lower limits of Groups
1 and 2 are 14.41 and 15.92, respectively. The big gap between the upper
and lower limits indicates a high level of uncertainty in Groups 1 and 2.
5. Discussion and conclusion
Based on the experimental studies done by Smith (2015) and Duarte
et al. (2014), this paper analysed risks of emergency situations with and
4.4.4. Confidence interval for behavioural compliance in exit signs
The confidence interval for behavioural compliance to different types
6
N. Norazahar et al.
Ocean Engineering xxx (2017) 1–8
emergency scenarios in a virtual environment (VE) to allow participants
to practice emergency operations. In the other study (Duarte et al., 2014),
participants were exposed to emergencies involving building evacuation.
In both studies, the performance of participants was measured.
5.2. Designing and implementing new safety measures
The use of virtual environments in a risk management study is a
starting point for decision making on the design and implementation of
new or existing structures or safety measures. A virtual environment can
be used to experiment with and implement new safety measures for
emergency situations. The decision to modify safety measures can be
based on the performance of individuals interacting with the safety
measures in a virtual environment.
5.3. Providing evidence to improve safety measures
Fig. 6. Upper and lower limits of Groups 1 and 2 in TH1.
Data on performance in emergency scenarios using virtual environments could provide support and evidence to improve engineered and
procedural safety measures. Risks associated with performance in
emergency scenarios using virtual environments can demonstrate the
need for improvement of the existing safety measures.
without engineered and procedural safety measures. Sections 5.1 to 5.3
discuss the use of virtual environments for assessing and managing risks
in emergency scenarios.
Key points
5.1. Introducing individuals to emergency scenarios on offshore
installations
Virtual environment can be used to assess human responses and to
provide data on risks associated with human responses during
emergency situations.
Human behaviour during emergency scenarios is dependent on the
egress route choices and the training provided.
Additional safety measures can improve performance during an
emergency. The new safety measures can be tested in a virtual environment prior to implementation in real life.
Further testing is required to establish the long-term effects of virtual
environments on participants' performance.
Two experimental studies related to performance in emergency scenarios were used in this paper. Smith (2015) developed credible offshore
Acknowledgements
The authors would like to thank the Atlantic Innovation Fund (AIF)
and Natural Sciences and Engineering Research Council of Canada
(NSERC) for the financial support necessary for this research study.
Special thanks to Virtual Marine Technology for assisting the researchers
with the AVERT technology for this study. We thank to the reviewers for
their insightful and thorough review.
Appendix
Fig. 7. Upper and lower limits of Groups 1 and 2 in TH2.
Table A1
Mean, upper limit, and lower limit emergency scenarios.
Test scenarios
Mean
(out of 35 points)
Upper limit
(out of 35 points)
Lower limit
(out of 35 points)
TA1
TA3
TH1
TH2
TH1
TH2
13.24
11.91
12.35
26.18
18.68
16.32
15.24
14.75
19.73
33.38
26.17
24.28
11.23
9.08
4.97
18.97
11.20
8.36
(Group
(Group
(Group
(Group
1)
1)
2)
2)
Table A2
Mean, upper limit, and lower limit of different type of exit signs.
Fig. 8. Upper and lower limits for the performance of participants following exit signs.
7
Type of signage
Mean
(out of 6 points)
Upper limit
(out of 6 points)
Lower limit
(out of 6 points)
No/Minimal sign
Static sign
Dynamic sign
2.90
4.60
5.23
3.40
5.20
5.72
2.40
4.00
4.74
N. Norazahar et al.
Ocean Engineering xxx (2017) 1–8
References
Ha, S., Ku, N., Roh, M.–I., Li, X., 2014. Evacuation, escape and rescue (EER) analysis of
offshore plant based on human behavior model. In: Proceedings from ASME 2014:
33rd International Conference on Ocean, Offshore and Arctic Engineering. ASME, San
Francisco, CA.
Health and Safety Executive, 1997. Prevention of Fire and Explosion, and Emergency
Response on Offshore Installations. Retrieved from. http://www.hse.gov.uk/pUbns/
priced/l65.pdf.
Health and Safety Executive, 2007. Managing Competence for Safety-related Systems.
Retrieved from. http://www.hse.gov.uk/humanfactors/topics/mancomppt1.pdf.
Joo, J., Kim, N., Wysk, R.A., Rothrock, L., Son, Y.–J., Oh, Y.G., Lee, S., 2013. Agent-based
simulation of affordance-based human behaviors in emergency evacuation. Simul.
Model. Pract. Theory 32, 99–115.
Kim, H., Park, J.H., Lee, D., Yang, Y.S., 2004. Establishing the methodologies for human
evacuation simulation in marine accidents. Comput. Ind. Eng. 46, 725–740.
Kjellen, U., 2007. Safety in the design of offshore platforms: integrated safety versus
safety as an add-on characteristic. Saf. Sci. 45, 107–127.
Kumamoto, H., Henley, E.J., 1996. Probabilistic Risk Assessment and Management for
Engineers and Scientists. IEEE Press, Piscataway, NJ.
Modarres, M., 2006. Risk Analysis in Engineering: Techniques, Tools, and Trends. Taylor
& Francis Group, Boca Raton, FL.
Mould, G.I., 2001. Assessing systems for offshore emergency evacuation. J. Oper. Res.
Soc. 52, 401–408.
Smith, J., 2015. The Effect of Virtual Environment Training on Participant Competence
and Learning in Offshore Emergency Egress Scenarios. Published master’s thesis.
Memorial University of Newfoundland, St. John’s, NL.
Smith, J., Veitch, B., MacKinnon, S., 2015. Achieving competence in offshore emergency
egress using virtual environment training. In: Proceedings of the ASME 2015 34th
International Conference on Ocean, Offshore and Arctic Engineering.
Andersen, S., Mostue, B.A., 2012. Risk analysis and risk management approaches applied
to the petroleum industry and their applicability to IO concepts. Saf. Sci. 50,
2010–2019.
Augustijn-Beckers, E.W., Flacke, J., Retsios, B., 2010. Investigating the effect of different
pre-evacuation behavior and exit choice strategies using agent-based modeling.
Procedia Eng. 3, 23–35.
Bradbury-Squires, D., 2013. Simulation Training in A Virtual Environment of an Offshore
Oil Installation. Published master’s thesis. Memorial University of Newfoundland, St.
John’s, NL.
Canadian Association of Petroleum Producers, 2010a. Guide: Atlantic Canada Offshore
Petroleum Industry Escape, Evacuation and Rescue. Retrieved from. http://www.
capp.ca/library/publications/atlanticCanada/pages/pubInfo.aspx?DocId¼173688.
Canadian Association of Petroleum Producers, 2010b. Standard Practice for the Training
and Qualifications of Personnel. Retrieved from. http://www.cnsopb.ns.ca/pdfs/raw.
pdf.
Colombo and Golzio, 2016. The Plant Simulator as viable means to prevent and manage
risk through competencies management: experiment results. Saf. Sci. 84, 46–56.
Deacon, T., Amyotte, P.R., Khan, F.I., 2010. Human error risk analysis in offshore
emergencies. Saf. Sci. 48, 803–818.
DiMattia, D.G., Khan, F.I., Amyotte, P.R., 2005. Determination of human error
probabilities for offshore platform musters. J. Loss Prev. Process Ind. 18, 488–501.
Doheny, J.G., Fraser, J.L., 1996. MOBEDIC - a decision modelling tool for emergency
situations. Expert Syst. Appl. 10 (1), 17–27.
Duarte, E., Rebelo, F., Teles, J., Wogalter, M.S., 2014. Behavioral compliance for dynamic
versus static signs in an immersive virtual environment. Appl. Ergon. 45, 1367–1375.
Eikill, G.O., Oftedal, B.C., Statoil, A.S.A., 2007. Rig Design Philosophy Key for Harshenvironment Drilling in Barents Sea, Haltenbanken Area, pp. 84–87. Drilling
Contractor, March/April 2007.
8
Документ
Категория
Без категории
Просмотров
2
Размер файла
936 Кб
Теги
044, 2017, oceaneng
1/--страниц
Пожаловаться на содержимое документа