Actual safety performance of the Malaysian offshore oil platforms: Correlations between the leading and lagging indicators

Abstract

Introduction:
This study establishes the correlations between performance of a set of key safety factors and the actual lagging performance of oil platforms in Malaysia, hence the relevance of the key safety factors in evaluating and predicting the safety performance of oil and gas platforms. The key factors are crucial components of a safety performance evaluation framework and each key safety factor corresponds to a list of underlying safety indicators.

Method:
In this study, participating industrial practitioners rated the compliance status of each indicator using a numbering system adapted from the traffic light system, based on the actual performance of 10 oil platforms in Malaysia. Safety scores of the platforms were calculated based on the ratings and compared with the actual lagging performance of the platforms. Safety scores of two platforms were compared with the facility status reports' findings of the respective platforms.

Results:
The platforms studied generally had good performance. Total recordable incident rates of the platforms were found to show significant negative correlations with management and work engagement on safety, compliance score for number of incident and near misses, personal safety, and management of change. Lost time injury rates, however, correlated negatively with hazard identification and risk assessment. The safety scores generally agreed with findings of the facility status reports with substandard process containment found as a contributor of hydrocarbon leaks.

Conclusions:
This study proves the criterion validity of the safety performance evaluation framework and demonstrates its usability for benchmarking and continuous improvement of safety practices on the Malaysian offshore oil and gas platforms.

Practical applications:
This study reveals the applicability of the framework and the potential of extending safety reporting beyond the few conventional lagging safety performance indicators used. The study also highlights the synergy between correlating safety factors to streamline safety management on offshore platforms.

Full text

Actual safety performance of the Malaysian offshore oil platforms:
Correlations between the leading and lagging indicators
Daniel Kuok Ho Tang,
a,
⁎Siti Zawiah Md Dawal,
b
Ezutah Udoncy Olugu
c
a
Department of Mechanical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
b
Department of Mechanical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
c
Faculty of Engineering, Technology & Built Environment, UCSI University, Kuala Lumpur, Malaysia
abstractarticle info
Article history:
Received 4 J anuary 2018
Received in revised form 27 March 2018
Accepted 8 May 2018
Available online 19 May 2018
Introduction: This study establishes the correlations between performance of a set of key safety factors and the
actual lagging performance of oilplatforms in Malaysia, hence therelevance of the keysafety factors inevaluating
and predicting the safety performance of oil and gasplatforms. The key factors are crucial components of a safety
performance evaluation framework and each key safety factor corresponds to a list of underlying safety
indicators. Method: In this study, participating industrial practitioners rated the compliance status of each indica-
tor using a numbering system adapted from the traffic light system, based on the actual performance of 10 oil
platforms in Malaysia. Safety scores of the platforms were calculated based on the ratings and compared with
the actual lagging performance of the platforms. Safety scores of two platforms were compared with the facility
status reports' findings of the respective platforms. Results: The platforms studied generally had good perfor-
mance. Total recordable incident rates of the platforms were found to show significant negative correlations
with management and work engagement on safety, compliance score for number of incident and near misses,
personal safety, and management of change. Lost time injury rates, however, correlated negatively with hazard
identification and risk assessment. The safety scores generally agreed with findings of the facility status reports
with substandard process containment found as a contributor of hydrocarbon leaks. Conclusions: This study
proves the criterion validity of the safety performance evaluation framework and demonstrates its usability for
benchmarking and continuous improvement of safety practices on the Malaysian offshore oil and gas platforms.
Practical applications: This study reveals the applicability of the framework and the potential of extending safety
reporting beyond the few conventional lagging safety performance indicators used. The study also highlights the
synergy between correlating safety factors to streamline safety management on offshore platforms.
© 2018 National Safety Council and Elsevier Ltd. All rights reserved.
Keywords:
Safety performance
Framework
Oil platforms
Injury
Near misses
1. Introduction
Safety performance measurement of offshore oil and gas platforms
in Malaysia is conventionally fragmented covering entities such as tech-
nical integrity, structural integrity, process safety, and occupational
safety (Hassan & Abu Husain, 2013). There is a lack of integrative
approach in the offshore platform safety performance measurement
by combining the major safety and health entities using both leading
and lagging indicators for proactive and reactive monitoring (Tang,
Leiliabadi, Olugu, & Md Dawal, 2017). An integrative safety performance
measurement provides indication of the “health”of an offshore oil and
gas platform, which is crucial for timely actions or rectifications should
the “health”status fall below satisfactory level.
Nonetheless, it is of interest to know how a safety performance mea-
surement framework correlates with the actual performance. A safety
performance measurement framework that is predictive of actual per-
formance is desirable as it enables more effective accident prevention,
hence death and injury reduction (Martinovich, 2013). Such framework
relies on specific and relevant indicators measuring crucial safety facets
of offshore oil and gas platforms, hence the overall “health”of the
platform. Generally, a “healthy”safety system is one with high compli-
ance to the safety performance targets or standards set. A healthy
system is commonly associated with lower incident rates, be it fatality,
injury, or near-miss (Shannon, Mayr, & Haines, 1997). Auditing the ef-
fectiveness of safety management is vital as higher accident rates are as-
sociated with higher safety management failings (Kawka & Kirchsteiger,
1999; Reason, 1997). Mearns, Whitaker, and Flin (2003) reported
connection between certain safety climate scales as well as proficiency
in some safety management practices, and official accident statistics.
This highlights that different safety practices and aspects exert varying
Journal of Safety Research 66 (2018) 9–19
⁎Corresponding author.
E-mail addresses: daniel.tang@curtin.edu.my,(D.K.H.Tang),
sitizawiahmd@um.edu.my, (S.Z. Md Dawal), Olugu@ucsiuniversity.edu.my (E.U. Olugu).
https://doi.org/10.1016/j.jsr.2018.05.003
0022-4375/© 2018 National Safety Council and Elsevier Ltd. All rights reserved.
Contents lists available at ScienceDirect
Journal of Safety Research
journal homepage: www.elsevier.com/locate/jsr

influence on incident occurrence. An understanding of the influence
permits more effective safety management on the offshore installations.
Currently, there is no database in Malaysia to collect and share data
on safety performance of offshore oil and gas platforms beyond the com-
mon parameters such as fatality and injury rates, as well as hydrocarbon
leaks (Petronas, 2015). An integrative safety measurement that is well-
accepted may set the path for performance benchmarking and informa-
tion sharing. However, the success of performance benchmarking using
the same sets of safety indicators depends on transparency in reporting
and performance standards. Performance reporting has been collected
on a voluntary basis (Petronas, 2015) and it could be difficult to impose
performance reporting on all oil and gas companies in Malaysia due to
challenges in data collection, hence cost consideration. In addition,
target-setting for the performance indicators may present certain chal-
lenges as different platform operators have different priorities and levels
of safety management (Podgorski, 2015).
According to the United States Transportation Research Board
(2016), theoffshore industry is fragmented with diverse workforce hav-
ing different safety attitudes and practices. The industry consists of large
and small companies differing in their commitment of resources for
safety and safety culture. The complex relationship among operators,
contractors, and subcontractors on offshore oil and gas installations
complicate safety-related roles and responsibilities. This presents signif-
icant challenge for industry-wide goal-setting in the effort of perfor-
mance benchmarking. This study examines the potential use of a set
of safety factors for performance evaluation and reporting on offshore
oil and gasplatformsusing a safety performance evaluation framework,
without attempting to achieve industry-wide goal setting for safety per-
formance. It also depicts the correlationsbetween actual performance of
leading and lagging safety factors.
2. Literature review
Safety performance measurement forms a crucial component of the
safety management system. The safety management system is the fru-
ition of systemic approach in safety, which can be linked to the concept
of safety system or safety system engineering made popular by
Bertalanffy (1971),Johnson (1980),andHammer (1989) in the 1970s.
Safety system integrates safety management techniques and perceives
safety as comprising inter-related components whose respective per-
formance contribute to the overall systemic performance. This approach
does not single out any component as the sole determinant of safety
(Hammer, 1989). The list of components in the system is subject to con-
tinuous review, with substitution and addition of components based on
relevance, for instance focus was placed on technical personnel such as
control room operators and maintenance works in the late 1970s
(Bertalanffy, 1971) but shifted to other areas as knowledge related to
safety advanced. The safety system is therefore a dynamic system that
evolves and improves in light of new knowledge, technology, and
experience.
In the mid-1980s, after the Chernobyl accident, focus was placed on
safety culture due to several safety deficiency of the Chernobyl power
plant such as ambiguous operating procedures, flawed designs, and
safety features, breaching of safety rules by operating staff, lack of
competence, and pressures to meet production goals (Hammer, 1989).
Nonetheless, Rentch (1990) and Witt, Hellman, and Hilton (1994)
pointed out that safety culture is not the entirety of safety system but,
rather, an aspect of safety system and promulgated a more holistic
view of safety management with safety being an emergent property
subject to continuous improvement in lifecycle of an installation.
Safety management has been given multiple definitions. Gupta and
Edwards (2002) defined safety management as “the management
process to ensure that risks are reduced to a level as low as reasonably
practicable via hazard identification, risk assessment and monitoring.”
Cox and Tait (1991) interpreted it as “the process whereby informed
decision are taken to meet safety criteria”while HSE (2013) construed
it as an intervention mechanism to prevent accidents. From the defini-
tions, safety management aims to reduce risks, meet safety criteria, and
prevent accidents via management process adapted from business-like
approach. Typical elements of a management system consist of policy
setting, organizing, planning, and implementation, evaluation, and action
for improvement (ILO, 2001).
Performance monitoring and measurement fits into the evaluation
stage of a safety system where safety performance is continuously and
systematically monitored, measured and recorded, and procedures of
performance measurement is consistently reviewed (ILO, 2001).
Reliable performance monitoring necessitates adoption of relevant
safety indicators. Safety indicators have historically developed parallel
to advancement of safety approach from lagging indicators such as fatal-
ities and injuries rates, organizational indicators such as work arrange-
ment, operational indicators, to resilience based indicators (Reiman &
Pietikainen, 2012).
Safety indicators have also been categorized based on safety domains.
On an offshore installation, the major safety domains encompass process
and personal safety. Process safety stemmed from the occurrences of in-
dustrial major accidents (e.g., the Flixborough explosion [Kletz, 1999]
and the Seveso disaster resulting in dioxin leakage). Process safety man-
agement, therefore, aims to prevent, minimize, and control industrial
major accidents such as fires and explosions, which cause not only inju-
ries and fatalities but property and environmental damage. Process
safety is frequently equated to asset integrity and both terms have
been used interchangeably (Lauder, 2012; Ratnayake, 2012). By defini-
tion, the latter covers the breadth of management of people, systems,
processes, and resources to minimize operational risks of an asset to
employees, the public, and the environment (Hassan & Khan, 2012). In
practice, asset integrity management on an offshore facility closely re-
sembles process safety focusing on the safety critical elements (SCEs),
which form crucial barriers of a system to prevent accidents and escala-
tion of the accidents once they occurred (Frens & Berg, 2014). The focus
is placed on the hard barriers of a system such as piping and instrumen-
tation as well as the corresponding design and operational parameter
(Vinnem, 1998).
The KP3 Asset Integrity Program by the HSE UK initiated in 2008
tracked the progress of participant oil and gas companies in indicators
related to maintenance and management of safety critical elements on
offshore facilities. Participating companies adopted a set of SCEs recom-
mended by the program and tracked compliance of the SCEs (HSE,
2008), usually via facility status reporting that captures the performances
of the SCEs (Frens & Berg, 2014). For comparison and benchmarking,
companies participating in the KP3 Asset Integrity Program report
hydrocarbon releases, verification non-compliance, and safety-critical
maintenance backlog to the HSE (HSE, 2009). HSE then tracked the
asset integrity performance of the participating companies based on the
few key indicators over a fixed duration.
Personal safety, on the other hand, emphasizes the health, safety,
and wellbeing of individual employees via minimization of their expo-
sure to occupational risks (ILO, 2001). In the offshore context, personal
safety focuses on reducing workers' exposure to radiations, chemicals,
noise, vibration, extreme temperatures, and ergonomic hazards via
measures such as industrial hygiene monitoring, chemical health risk
assessment, job safety analysis, medical surveillance, safety awareness
program, and work arrangement to reduce fatigue and increase alert-
ness (Venkataraman, 2008). Personal safety also looks into competence
building and safety behavior promotion (Arezes & Miguel, 2008). The
reporting of personal safety performance in terms of number of injuries
or fatalities caused by slip and trip, falling from height, electrical expo-
sure, struck-by, caught between and burns, and so forth (IOGP, 2016)
is more common than process safety performance. Personal safety is
also reported to a significantly larger extent than process safety in
corporate safety reporting.
Though conventionally managed separately, process and personal
safety are not mutually exclusive. Studies have pointed to their
10 D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

inter-relations. Olsen, Naess, and Hoyland (2015) revealed that
work climate factors are negatively correlated with incidences of hydro-
carbon leakand can be usedto predict leaks of different severity. Bergh,
Ringstad, Leka, and Zwetsloot's (2014) also reported positive correla-
tion between psychological risk scores and incidences of hydrocarbon
leaks. Both safety domains ultimately aim to reduce fatality, injury
rates, and near-misses. They converge at the organizational level
where leadership and organizational culture serve as the driving force
of both safety domains (Guldenmund, 2000; Morrow, Koves, & Barnes,
2014).
In order to effectively capture the overall safety performance of an
offshoreplatform, it has been promulgated that themajor areas of safety
should be included with the use of both lagging and leading indicators
(Baker, 2007; CSB, 2012). Facility status reporting undermines soft
barriers and the leading aspects of safety management such as organiza-
tional factors, work arrangement, safety culture and documentation, as
well as personal safety. Personal safety reporting, however,overempha-
sizes the lagging performance such as injuries and fatalities without
examining the leading aspects that are often also the root causes of
major accidents, for instance incompetence and operational shortcuts
(Williams, Hamid, & Misnan, 2017).
As proactive and integrative safety monitoring has received much
attention in recent years, inclusion of leading safety factors in
performance reporting could present a greater opportunity for mutual
learning of best practices in offshore platform safety management
(CSB, 2012). This study tests the usability of a framework proposed
earlier (Tang et al., 2017) combining crucial aspects of personal and
process safety, using both leading and lagging indicators, in attempt to
shed a more holistic insight into how the leading and lagging aspects
of both safety domains are connected.
3. Method
3.1. Testing of safety performance measurement framework
This study tested the predictive ability of a safety performance
framework by drawing a correlation between the actual performance
of 10 offshore oil platforms and the evaluated safety performance of
the platforms using the framework proposed. The actual performance
centered around the number of fatality, fatal accident rate, total record-
able incident rate, lost time injury rate, and reported near-misses in
2016.
Fatality is defined as death, either immediate or within one year of
the date of injury, of an employee or a contractor's employee due to
work, while fatal accidents are accidents resulting in fatality. Total
recordable incidents encompass all fatalities, lost time injuries, illnesses,
and medical treatment cases occurring at work but do not include
first-aid injury. Lost time injury results in inability of an employee to
continue work, hence a loss of productive work time. Near-misses, on
the other hand, are unintended occurrences that could potentially
harm human, the environment, and properties (IOGP, 2016; Petronas,
2015). Incident and injury rates are counted as occurrences per million
man-hours worked. In the case of total recordable incident rate for
instance, it is counted as number of total recordable cases per million
man-hours worked (HSE, 2015; IOGP, 2016; Petronas, 2015).
A questionnaire consisting of a list of 63 safety indicators grouped
under 14 safety indicators identified from a previous study by Tang
et al. (2017), as well as a section for actual performance reporting
with the parameters mentioned above were disseminated to the partic-
ipating industrial practitioners in three oil and gas companies in
Malaysia. The safety factors comprised: (1) Inspection and maintenance,
Idenficaon
of indicators
• Literature review
• Consultation with health and safety practitioners
• Grouping of indicators under relevant safety factors
Quesonnaire
• 70 survey items
• Pilot study to validate and improve survey items
• Adminstration of questionnaire
• 172 complete responses
Stascal
analysis
• Mean perceived importance ratings and perceived risk ratings of
safety indicators, hence the corresponding safety factor
• Derivation of weight of safety indicators
Final list of
indicators
• Final review of indicators reduced the number to 63
• Form the basis of safety performance measurement framework for
offshore oil and gas platforms in Malaysia
Fig. 1. Development of indicators for offshore safety performance measurement (Tang et al., 2017).
11D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

(2) Emergency management, (3) Management and work engagement,
(4) Number of incidents and near misses, (5) Personal safety, (6) Contrac-
tors' safety, (7) Management of change, (8) Operation and operating pro-
cedures, (9) Competence, (10) Hazard identification and risk assessment,
(11) Plant design, (12) Instrumentation and alarm, (13) Documentation,
and (14) Start-ups and shutdown. A summary of earlier indicators
development is shown in Fig. 1.
The industrial practitioners were involved in collection of safety
performance data for safety reporting and were best-suited to provide
data for the questionnaire. The respondents were requested to indicate
the compliance status of each of the safety indicator using a numbering
system modified from the traffic light system of the HSE (2008) in its
Asset Integrity Key Program, where red indicates non-compliance,
amber indicates isolated failure or incomplete system, and green
indicates compliance. In the numbering system, “3”was assigned for
compliance, “2”for isolated failure or incomplete system, and “1”for
non-compliance. To account for indicators without data, “0”was
assigned. The compliance status was determined based on comparison
of the indicators' performance against the respective performance tar-
gets or standards. The study participants needed to extract the existing
performance data of the platforms from multiple sources and interpret
the data based on the format of this safety measurement framework.
The score of an indicator was obtained by multiplying the compli-
ance status of the indicator with its corresponding weight identified in
a previous study (Tang et al., 2017). To compute the score of a safety
factor, the score obtained for each indicator categorizedunder the safety
factor was summed up, as shown in the mathematical expression
below:
Score of safety factor;j¼∑n
i¼1WijCij ð1Þ
where,
W
ij
,Weightofi
th
indicator under safety factor j
C
ij
, Compliance status of i
th
indicator under safety factor j
Taking a safety factor (i.e., inspection and maintenance (Table 1)for
instance), there are four indicators categorized under the safety factor
with weights derived from a survey conducted in a previous study
(Tang et al., 2017)asshowninTable 1. One part of the survey required
the respondents to rate their perceived importance of the indicators on
a scale of 1 to 5, which represents increasing degree of perceived impor-
tance. The other part required the respondents to rate their perceived
risk due to failure of observing the indicators, also on a 1-to-5 scale of
increasing risk. The weights of the respective indicators were subse-
quently calculated by multiplying the mean perceived importance rat-
ings and the mean perceived risk ratings of the respective indicators.
The score of this safety factor is the sum of scores of the underlying in-
dicators calculated as shown in Eq. (1).
Total safety score of a platform is the sum of scores of all 14 safety
factors as expressed below:
Total safety score of platform;a¼∑14
i¼1SFia ð2Þ
where,
SF
ia
,Scoreofi
th
safety factor for platform a
The total safety score of a platform corresponds positively to its
safety performance. The higher the score, the higher the safety perfor-
mance is. Descriptive statistics including range, means, standard
deviation, and variance were calculated for each factor to give an over-
view of how the performance of each safety factor varied among the
platforms. Correlation between the scores of safety factors and the
actual performance of the offshore oil and gas platforms was analyzed
using Pearson correlation (IBM, 2014). Fig. 2 provides an overview of
the safety performance measurement framework.
3.2. Validation of safety performance measurement framework
The safety performance measurement framework was validated
against the facility status reports of safety critical elements (SCE) groups
of two offshore oil platforms located on the shallow water of Miri,
Malaysia. Facility status reporting of oil and gas platform using the SCE
was promulgated by HSE's Key Programme 3 (KP3) to ensure platform
operators manage risk related to structure, plant, and equipment effec-
tively in the prevention of major accidents and report performance of
SCE in standardized format, thus facilitating sharing of best practices
and benchmarking (HSE, 2008).
Table 2
Score demarcation of safety factors.
Safety factor Maximum
score
Middle
score
Low
score
Minimum
score
Inspection and maintenance 114.3 76.2 38.1 0
Emergency management 92.3 61.5 30.8 0
Management and work engagement
(MWE)
120.9 80.6 40.3 0
Number of incidents and near misses 255.8 170.6 85.3 0
Personal safety 225.8 150.5 75.3 0
Contractors' safety 52.4 35.0 17.5 0
Management of change 160.7 107.1 53.6 0
Operation and operating procedures 280.8 187.2 93.6 0
Competence 94.4 62.9 31.5 0
Hazard identification and risk assessment 88.7 59.1 29.6 0
Plant design 83.1 55.4 27.7 0
Instrumentation and alarm 121.7 81.1 40.6 0
Documentation 68.6 45.7 22.9 0
Start-ups and shutdown 71.0 47.3 23.7 0
Total 1,830.4 1,220.3 610.1 0
Determine
compliance
status of each
indicator
underpinning
the 14 safety
factors
Calculate
scores of
indicators
Calculate
scores of
safety factors
Calculate total
safety score of
a plaorm
Fig. 2. Safety performance measurement.
Table 1
Indicators related to inspection and maintenance.
Safety factor: inspection and maintenance
Indicators Weight
Number of safety critical plant/equipment that performs within
specification when inspected
10.017
Number of maintenance actions identified that are completed to the
specified timescale
9.303
Number of hours of critical maintenance backlog 9.143
Number of all hydrocarbon leaks 9.637
12 D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

SCE, according to the Offshore Installations (Safety Case) Regulations
2005 of the UK, is defined as parts of an installation or plant that play a
crucial role in prevention or minimization of impacts of major accidents
(HSE, 2006). Failure of SCE could therefore escalate into the occurrence
of major accidents. Facility status reporting is part of safety performance
measurement currently practiced on offshore oil and gas platforms and
bear much resemblance to the proposed framework, particularly in
indicating the compliance status of each SCE. The reporting system
employs the traffic light method recommended by the HSE with red
indicating non-compliance, amber indicating isolated failure or incom-
plete system, and green indicating compliance (Frens & Berg, 2014).
Both SCE and non-SCE are monitored in facility status reporting.
Safety scores obtained from the proposed framework are compared
against the findings of facility status reports of two oil platforms in
Malaysia to examine how close the findings from both instruments
are in terms of the platform safety performance.
4. Results
Table 2 shows the maximum possible score for each safety factor, as-
suming all indicators of each safety factor have a status of compliance.
The middle score represents the score of a safety factor if all its indica-
tors assume a deviated status that can be interpreted as isolated failure,
or incomplete system. The low score of a safety factor is the sum of
scores of the underlying indicators with all having a status of non-
compliance. Minimum score is only possible when no indicators are
available for each of the safety factor, which indicates that a system to
evaluate safety performance is not in place or no indicators comparable
to the framework were used on the platforms.
All scores of safety factors calculated using the framework proposed
(Table 3) fall between the maximum and middle scores (Table 2)of
the respective safety factors with most of the scores above the mid-line
of the range, except the scores of start-ups and shutdown for two
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Plaorm
1
Plaorm
2
Plaorm
3
Plaorm
4
Plaorm
5
Plaorm
6
Plaorm
7
Plaorm
8
Plaorm
9
Plaorm
10
Pla orm Safety Scores
Inspecon and Maintenance Emergency Management
Management and Work Engagement Number of Incidents and Near Misses
Personal Safety Contractors' Safety
Management of Change Operaon and Operang Procedures
Competence Hazard Idenficaon & Risk Assessment
Plant Design Instrumentaon & Alarm
Documentaon Start-ups and Shutdown
Fig. 3. Platform safety scores for year 2016.
Table 3
Safety factor scores of offshore oil and gas platforms for year 2016.
Safety factor Platform
12345678910
Inspection and maintenance 104.7 104.7 104.7 104.7 95.0 94.7 114.3 114.3 85.5 114.3
Emergency management 92.3 92.3 82.3 82.3 92.3 82.3 81.6 81.6 81.6 92.3
Management and work engagement 101.3 120.9 120.9 120.9 110.5 91.5 100.7 101.3 100.7 100.8
Number of incidents and near misses 197.5 246.1 236.8 236.8 246.5 227.0 236.3 255.8 226.6 236.4
Personal safety 198.5 225.8 216.6 216.6 225.8 188.6 207.7 216.6 198.7 216.8
Contractors' safety 52.4 52.4 52.4 52.4 52.4 43.5 43.9 52.4 52.4 43.5
Management of change 137.9 160.7 160.7 160.7 153.2 150.5 151.8 160.7 144.3 153.2
Operation and operating procedure s 252.1 280.8 280.8 280.8 280.8 271.0 251.8 241.2 261.5 260.1
Competence 84.1 94.4 94.4 94.4 94.4 94.4 94.4 84.2 94.4 84.2
Hazard identification and risk assessment 88.7 88.7 88.7 88.7 88.7 88.7 88.7 58.8 88.7 88.7
Plant design 83.1 83.1 83.1 83.1 83.1 83.1 75.2 83.1 72.5 75.2
Instrumentation and alarm 121.7 121.7 121.7 121.7 121.7 110.6 121.7 110.6 91.7 121.7
Documentation 68.6 68.6 68.6 68.6 68.6 68.6 68.6 68.6 68.6 68.6
Start-ups and shutdown 48.1 71.0 59.6 71.0 48.1 59.6 59.6 71.0 71.0 59.6
Total score 1,630.8 1,811.1 1,771.1 1,782.5 1,761.0 1,653.8 1,696.1 1,700.1 1,638.2 1,715.1
13D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

platforms (Platforms 1 and 5) just slightly above the middle score. Score
of inspection and maintenance of Platform 9 was the lowest of all
platforms surveyed, though still above the middle score of 76.2. Platform
2 had the highest safety score of 1,811 followed by Platform 4 and
Platform 5 (Table 3). Platform 1 had the lowest safety score, though it
was still significantly higher than the middle score (refer Table 2).
Fig. 3 shows the performance of the platforms studied in terms of
their total safety scores. Referring to Table 4, number of incident and
near misses had the largest range of variation with a variance of 249.0,
followed by operation and operating procedures with a variance of
218.5. All the platforms had uniform score for documentation indicating
that documentation is generally an established practice, hence full
compliance. Spreads of scores for plant design and contractors' safety
from the respective means were low, as indicated by the variances
and standard deviations in Table 4.
Actual performance of offshore oil and gas platforms is commonly
captured in lagging indicators, particularly in terms of fatality and inju-
ries (Morrow et al., 2014).Table 5 shows the actual lagging performance
of the platforms. There was no fatality reported on all the platforms in
2016. Platforms 1, 6, and 9 reported a total recordable incident rate of
more than two, higher than the other platforms. Platform 8 recorded
the highest lost time injury rate, while Platform 1 had the highest
reported near-misses.
Table 6 shows correlation analysis between scores of the safety
factors and the actual lagging performance of the platforms. Total
recordable incident rate significantly and negatively correlated with
management and work engagement (MWE), number of incidents and
near misses, personal safety, and management of change. Lost time
injury rate demonstrated significant negative correlation with the
scores of operation and operating procedures, hazard identification
and risk assessment, as well as instrumentation and alarm. Nearmisses,
however, negatively correlated with number of incident and near mis-
ses, personal safety, management of change, hazard identification and
risk assessment, as well as start-ups and shutdown. Significant positive
correlation can be observed between near misses and total recordable
incident rate.
Correlations between the safetyfactors based on Pearson correlation
are also demonstrated in the dendrogram below (Fig. 4). Increasing
horizontal distance from the left to the right shows increasing dissimilar-
ity between the clusters. The dendrogram depicts the correlations in a
pairwise manner with the closer pairs of safety factors and clusters of
safety factors being merged successively in an agglomerative manner
where the clusters are paired in a “bottom”up approach with more
related elements or clusters more closely paired. The rescaled distance
represents the degree of dissimilarity (Liu, Zhu, Qiu, & Chen, 2012).
Facility status report of Platform 2 for year 2016 showed compliance
for all groups of safety critical elements (SCE) except process contain-
ment with a deviated status due to isolated failure in meeting the
performance target (Table 7). Each SCE group consists of a list of SCE
that represent the barriers to prevent and contain accidents likely to
occur on offshore platforms (Jager, 2013). Structural integrity, for
instance, consists of subsea structures, topside/surface structures,
heavy lift cranes and mechanical handling, ballast systems, mooring
systems, and drilling systems (Table 8). Performance goal was set for
each SCE as indicated in Table 9 and was evaluated based on the under-
lying functional criteria. There are two types of performance goals
(i.e., design performance standards and operations performance
standards). Design performance standards specify the design features,
capacity, or loads of the safety critical elements, while operational
performance standards specify the operational envelops, conditions,
and efficiency the elements are expected to demonstrate (Jager,
2013). As the platforms studied have been in operation, operational
performance standards are the major concerns.
Facility status report of Platform 2 (Table 7) shows good agreement
with the results of the safety performance framework proposed
(Table 3). Based on the framework, all the safety scores of Platform 2
achieved the status of full compliance except inspection and mainte-
nance, and number of incidents and near misses. Under the category
of inspection and maintenance, the indicator that was rated deviated
was the number of all hydrocarbon leaks, which correlates with the
finding of its facility status report on processcontainment. Facility status
of Platform 3 shows deviation for process contaminant, shutdown sys-
tems, and emergency response.
5. Discussion
The safety performance evaluation framework presents an integra-
tive approach in measuring the safety performance of offshore oil and
gas platform based on the 14 safety factors identified in a previous
study (Tang et al., 2017). The number of indicators under each safety fac-
tor varies, hence the maximum score for each safety factor. All the plat-
forms had scores closer to the side of the maximum score (Table 2)
with Platform 2 topping the list (refer to Table 3 for the platforms' safety
scores), indicating that the platforms were generally well managed.
Table 4 shows higher variation, hence variance of number of inci-
dents and near misses, as well as operation and operating procedures
among the platforms. The former was attributed primarily to difference
in number of incidents and near misses caused by contractors or visi-
tors, during start-ups and shutdown, and where operational shortcuts
were identified. The latter was due to missing indicators under the
Table 5
Actual safety data of offshore oil and gas platforms for year 2016.
Indicator Platform
12345678910
Fatality 0000000000
Fatal incident rate 0000000000
Total recordable incident rate 2.43 1.21 1.69 1.71 1.75 2.26 1.88 1.84 2.37 1.92
Lost time injury rate 0.38 0.22 0.21 0.35 0.19 0.47 0.25 1.07 0.55 0.33
Reported near-misses 13 5877108296
Table 4
Descriptive statistics of safety factor scores.
Safety factor Range Mean Standard
deviation
Variance
Inspection and maintenance 28.8 103.7 9.6 92.1
Emergency management 10.7 86.1 5.3 28.5
Management and work engagement 29.4 106.9 10.6 112.8
Number of incidents and near misses 58.3 234.6 15.8 249.0
Personal safety 37.2 211.2 12.4 153.8
Contractors' safety 8.97 49.8 4.3 18.2
Management of change 22.8 153.4 7.8 60.9
Operation and operating procedures 39.6 266.1 14.8 218.5
Competence 10.3 91.3 4.9 24.3
Hazard identification and risk assessment 29.9 85.7 9.5 89.3
Plant design 10.6 80.4 4.3 18.7
Instrumentation and alarm 29.9 116.5 9.8 96.7
Documentation 0.0 68.6 0.0 0.0
Start-ups and shutdown 22.8 61.8 9.0 81.1
14 D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

safety factor in few platforms. Thecompliance statusfor number of inci-
dents and near misses wasin agreement with the actual lagging perfor-
mance of the platforms, which were based on fatalities, injuries, and
near misses. Higher total recordable incident rate and near misses of
Platform 1 were confirmed by its lowest score in this category among
the platforms (refer to Tables 3 and 5). The same was observed for Plat-
forms 6 and 9. This demonstrates that the score of number of incidents
and near misses in the framework should inversely correlate with the
actual number of injuries and near misses reported where a higher
score indicates higher compliance. As the score of number of incidents
and near misses were dependent on the actual number of injuries and
near misses, a higher compliance score indicates lower number of inju-
ries and near misses. The correlation is captured by Pearson correlation
showing significant negative correlation between number of incidents
and near misses, and total recordable incident rate as well as number
of reported near misses (Table 6).
A previous study by Tang et al. (2017) demonstrated that MWE on
safety, number of incidents and near misses, contractors' safety, man-
agement of change, plant operations and operating procedures, plant
design, personal safety, as well as hazard identification and risk assess-
ment were more connected than the other safety factors via perceived
importance of the safety indicators rated in a survey among oil and
gas safety professionals in Malaysia. Correlation analysis in Table 6
shows that MWE was significantly positively correlated with personal
safety, contractors' safety, management of change (also see Fig. 4), and
plant operations and operating procedures, in line with the previous
findings. Personal safety in this study correlated significantly with
MWE on safety, number of incidents and near misses as well as manage-
ment of change, while contractors' safety correlated significantly with
MWE on safety (Fig. 4,Table 6), confirming the previous clustering
(Tang et al., 2017). Table 6 also reveals new correlations such as that
between inspection and maintenance, and competence, as well as
between inspection and maintenance, and instrumentation and alarm,
which were not identified in previous study (Tang et al., 2017). MWE
is a facet of safety culture related to leadership and commitment. Few
meta-analyses studies (Beus, Payne, Bergman, & Arthur Jr., 2010;
Christian, Bradley, Wallace, & Burke, 2009) demonstrated that safety
culture correlated with accidents and injuries, as well as employee's
self-reported safety behaviors. Morrow et al. (2014) demonstrated
that higher management commitment led to lower human perfor-
mance error rate. These studies have invariably pointed to correlations
between safety culture and safety performance.
Significant positive correlation between total recordable incident
rate and near misses confirms the accident triangle where higher num-
ber of near misses leads to higher number of recordable incidents in-
cluding injuries and fatality (Williamsen, 2003). Based on the accident
triangle, near misses are always more than recordable injuries though
different ratios between the two had been reported (Bellamy, 2015).
The same is shown in Table 5. It is established that better safety culture
leads to better safety performance, particularly accident and injury rates
(Feng, Teo, Ling, & Low, 2014; Morrow et al., 2014). MWE in this study,
being a measure of safety culture, was shown to significantly and posi-
tively correlate with total recordable incident rate. Mearns et al. (2003)
reported a significant association between management commitment
and accident rate, which supports the correlation between MWE and
incident rate.
Both total recordable incident rate and near misses were inversely
correlated to personal safety score. Personal safety encompasses noise
exposure, cases of occupational diseases and occupational poisoning,
fatigue,overtime, extended shifts, andchemical exposure. Fatigue,over-
time, and extended shifts are aspects of psychosocial risks. Bergh et al.
(2014) demonstrated that psychosocial risks including work schedule,
workload, and work pace correlated with number of hydrocarbon
leaks on platforms. Fatigue was linked to extended work hours, night
shifts, and rotating shifts (IPIECA-OGP, 2012), and was shown to
negatively impact human performance leading to higher risks, hence
Table 6
Correlations between safety factor scores and actual platform lagging performance.
Ins Emer MWE Incident Personal Con Change Ops Com HIRA Design Inst Startup TRIR LTIR NM
Ins 1 0.082 0.072 0.259 0.370 −0.285 0.384 −0.422 −0.531 −0.389 0.065 0.619 0.031 −0.368 0.124 −0.437
Emer 0.082 1 0.133 −0.169 0.385 0.081 −0.220 0.173 −0.344 0.293 0.161 0.476 −0.499 −0.192 −0.421 0.087
MWE 0.072 0.133 1 0.333 0.726 0.605 0.624 0.685 0.380 0.188 0.397 0.447 0.241 −0.766 −0.437 −0.278
Incident 0.259 −0.169 0.333 1 0.697 0.062 0.845 0.175 0.210 −0.473 0.094 0.063 0.466 −0.730 0.208 −0.926
Personal 0.370 0.385 0.726 0.697 1 0.381 0.692 0.360 0.033 −0.153 0.237 0.513 0.125 −0.861 −0.221 −0.693
Con −0.285 0.081 0.605 0.062 0.381 1 0.137 0.231 0.054 −0.218 0.414 −0.102 0.175 −0.219 0.132 −0.117
Change 0.384 −0.220 0.624 0.845 0.692 0.137 1 0.391 0.248 −0.331 0.364 0.323 0.515 −0.849 0.047 −0.796
Ops −0.422 0.173 0.685 0.175 0.360 0.231 0.391 1 0.698 0.592 0.363 0.274 −0.007 −0.499 −0.672 0.059
Com −0.531 −0.344 0.380 0.210 0.033 0.054 0.248 0.698 1 0.507 −0.002 −0.107 0.178 −0.298 −0.501 0.112
HIRA −0.389 0.293 0.188 −0.473 −0.153 −0.218 −0.331 0.592 0.507 1 −0.215 0.211 −0.356 0.063 −0.895 0.654
Design 0.065 0.161 0.397 0.094 0.237 0.414 0.364 0.363 −0.002 −0.215 1 0.469 −0.173 −0.338 0.020 −0.069
Inst 0.619 0.476 0.447 0.063 0.513 −0.102 0.323 0.274 −0.107 0.211 0.469 1 −0.438 −0.526 −0.543 −0.044
Startup 0.031 −0.499 0.241 0.466 0.125 0.175 0.515 −0.007 0.178 −0.356 −0.173 −0.438 1 −0.331 0.438 −0.572
TRIR −0.368 −0.192 −0.766 −0.730 −0.861 −0.219 −0.849 −0.499 −0.298 0.063 −0.338 −0.526 −0.331 1 0.290 0.676
LTIR 0.124 −0.421 −0.437 0.208 −0.221 0.132 0.047 −0.672 −0.501 −0.895 0.020 −0.543 0.438 0.170 1 −0.413
NM −0.437 0.087 −0.278 −0.926 −0.693 −0.117 −0.796 0.059 0.112 0.654 −0.069 −0.044 −0.572 0.676 −0.413 1
Ins: Inspection and maintenance
Emer: Emergency management
MWE: Management and work engagement
Incident: Number of incidents and near misses
Personal: Personal safety
Con: Contractors' safety
Change: Management of change
Ops: Operation and operating procedures
Com: Competence
HIRA: Hazard identification and risk assessment
Design: Plant design
Inst: Instrumentation and alarm
Startup: Start-ups and shutdown
TRIR: Total Recordable Incident Rate
LTIR: Lost Time Injury Rate
NM: Near misses
: Correlation is significant at the 0.05 level (1-tailed)
: Correlation is significant at the 0.01 level (1-tailed).
15D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

injuries and near misses. Parkes (2010) reported that increased injury
rates of offshore day work were related to circadian disruption as a result
of night work and “rollover”shift patterns involving shift change usually
from nights to days in the middle of an offshore tour. Such shift pattern
adversely affected sleep, performance, and alertness. Nonetheless,
Parkes (2010) also highlighted that shift extension from 8 h to 12 h did
not cause significant negative impact on performance or health, in an on-
shore setting. Demographically, Reiner, Gerberich, Ryan, and Mandel
(2016) reported increased risk of machinery-related injury among
Fig. 4. Dendrogram of safety factors based on Pearson correlations.
Table 8
Example of safety critical elements.
SCE group Safety critical element
Structural integrity •Subsea structures
•Topside/surface structures
•Heavy lift cranes and mechanical handling
•Ballast systems
•Mooring systems
•Drilling systems
Process containment •Pressure vessels
•Heat exchangers
•Rotating equipment
•Tanks
•Piping systems
•Pipelines
•Relief system
•Well containment
•Fired heaters
•Gas tight floor/walls
•Tanker loading
•Helicopter refuel
•Wireline equip
•Oil-in-water control
Shutdown system •Emergency shutdown and depressurization systems
•Depressurization systems
•High-integrity pressure protection systems
•Operational well isolation
•Pipelines isolation valves
•Process emergency shutdown valves
•Drilling well control
•Utility air
Table 7
Compliance of SCE groups in facility status report of platforms 2 and 3.
SCE groups Platform 2 Platform 3
Compliance
status
Remark Compliance
status
Remark
Structure integrity Green Compliant Green Compliant
Process containment Amber Deviated
a
Amber Deviated
a
Ignition control system Green Compliant Green Compliant
Detection systems Green Compliant Green Compliant
Protection systems Green Compliant Green Compliant
Shutdown systems Green Compliant Amber Deviated
a
Emergency response Green Compliant Amber Deviated
a
Life saving systems Green Compliant Green Compliant
Non-SCE items Green Compliant Amber Compliant
Competency deviation Green Compliant Green Compliant
Standards Green Compliant Green Compliant
a
Due to isolated failure. However, the performance sti ll falls within acceptable zone below
the corresponding performance target.
16 D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

male workers, workers of greater age, workers with a history of injury,
and increased work hours in the U.S. agricultural sector.
This study reveals that lost-time injury rate was negatively corre-
lated with the scores of operation and operating procedures, hazard
identification and risk assessment, as well as instrumentation and
alarm. While lost-time injury is a subset of total recordable incidents,
correlation between these two entities in this study is not sufficiently
significant. A relatively weaker negative correlation was also shown
between lost-time injury and competence, which often relates to train-
ing and experience (HSE, 2017). McVittie, Vi, and Eng (2009) reported
that density of trained supervisor in the Canadian construction sector
was negatively correlated with lost-time injury rate. The inverse rela-
tion between injury rates and level of experience was also highlighted
in a study among seafarers and rig workers in the Western Australia off-
shore oil and gas industry (Martinovich, 2013). Though no established
correlation was reported between lost-time injury and operating
procedures, hazard identification and risk assessment has been identi-
fied as a crucial element of occupational injury and illness prevention
program (Liu et al., 2010). Studies on the effect of various safety factors
on injury rates, including lost time injury, and near misses focus primar-
ily on organizational and workplace factors such as workforce empower-
ment, delegation of safety activities, and top management commitment
(Shannon et al., 1997). It would also be of interest to look at how other
safety factors, as identified in this framework, affect injury rates and
near misses particularly in the offshore sector.
Results of the framework were also compared against results of
facility status reporting (Table 7), which is currently practiced on the
platforms of an established oil and gas company. A major difference
between the framework and the facility status report is that the latter
are oriented towards hard barriers on the platforms comprising struc-
tures, equipment, vessel, physical system, and so forth (Table 8). The
framework on the other hand, monitors both hard and soft barriers
using both lagging and leading indicators covering multiple facets
such as organizational, resilience-based, and personal (Tang et al.,
2017). Soft barriers (i.e., competency deviation and standards) are also
included in facility status report but to a lesser extent than the frame-
work proposed. Significant agreement is shown between results of the
framework and facility status report. Platform 2 had the highest safety
score, which was also reflected by the compliance of all the SCE groups
except process containment (Table 7). Deviation in process contain-
ment was in fact captured by the framework via the indicator of
‘number of all hydrocarbon leaks’under the safety factor ‘inspection
and maintenance’which had a “deviated”status.
Similarly for Platform 3, deviation of process containment in the
facility status report is associated with deviated state for ‘number of
all hydrocarbon leaks’. In addition, Platform 3 also registered a deviated
state for shutdown system and emergency response in the facility status
report due to isolated failure of certain hardware related to the two sys-
tems. The findings correspond to deviation in the ‘number of elements
of emergency procedure that fail to function to performance standard’
under the safety factor of emergency management, as well as deviation
in the ‘number of deferred start-ups & unplanned shutdown’under the
safety factor of start-ups and shutdown, in the framework. Isolated
failures in the hardware might have contributed to increased failure of
emergency procedure to meet performance standard as well as higher
deferred start-ups and unplanned shutdown.
Sub-standard performance of number of all hydrocarbon leaks was a
common problem in most of the platform studied. Hydrocarbon leaks
is a major risk to accidents on offshore platform, hence a crucial safety
performance indicator (Olsen et al., 2015). Based on a review by Sklet
(2006), hydrocarbon leaks were mainly caused by operational error
during routine production, maintenance-related latent failure, dissem-
bling of hydrocarbon system for maintenance, technical/physical
failures, process upsets, and design related failures. These factors were
largely captured by the safety factors used in the framework. The effect
of process upsets and hydrocarbon system integrity on hydrocarbon
leaks was also demonstrated by the facility status report (Table 7)show-
ing isolated failure of process containment. Nonetheless, the facility status
report, as mentioned, is hardware-oriented and the performance stan-
dards set are in relation to the functional criteria of the hardware, such
as fire water pumps under the protection system (Table 9). Unlike facility
status reporting, the framework proposed unites various crucial aspects of
safety in determining the overall safety performance of a platform.
The study is limited in the number of platforms that were tested
using the safety framework. The reason is that the evaluation process
was elaborate where the study participants needed to gather the infor-
mation related to compliance of the indicators from multiple sources.
Integrating the information into a single framework presented a
tremendous challenge because safety performance monitoring in the
offshore sector was fragmented into personal, process, and asset while
the use of leading indicators monitoring documentation, organizational
factors, and competence especially is limited. Facility status reports of
platforms are not readily accessible and securing a full report even
from one single platform presents a great challenge due to corporate
considerations and legal process.
6. Conclusion
The findings of the framework provide important insight into
understanding recordable incidents, injuries, and near misses on the
platforms and show satisfactory agreement with the facility status
report, indicating the applicability of the framework for safety perfor-
mance measurement of the Malaysian offshore oil and gas platforms.
Grouping the safety indicators into 14 major safety factors and
employing a scoring system of the safety factors enable performance
comparison between the Malaysian offshore oil and gas platforms,
thus, contributing to benchmarking and continuous improvement. The
framework also enables correlations between the safety factors to be
Table 9
Example of operational performance standard.
SCE
group
Protection system
SCE Fire water pumps
SCE goal Provision of adequate water to extinguish or contain, hence reduce impact of fire.
Function
no.
Functional criteria Minimum assurance task Assurance
measure
Assurance
value
1Eachfire pump shall operate as per its design specifications Test performance of fire pump as per design pump
curve to ensure delivery of largest firewater demand
Firewater discharge pressure
Firewater flow rate
≥Pressure
≥Flow
2Eachfire pump shall be activated by initiation signals and run
withoutinterruptioninthespanofadefined emergency
event
Each fire pump shall be activated by pressing:
•Local panel pushbutton
•Fire & Gas panel pushbutton
•Fire main pressure switch
A testing plan should be in place for equal testing of all
start signals above.
Fire pump starts on demand Y/N
17D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

captured and improves understanding on how the factors influence
actual lagging performance of the platforms.Future study can therefore
focus on continuous improvement of the framework as safety practices
on the offshore platforms evolve and improve. Investigation on the
correlation of the safety factorsand their effects on lagging performance
of the platforms can also be a potential area of study.
7. Practical application
The study shows that the framework proposed has the ability to
change the current safety reporting practice for offshore platforms by
utilizing a wider range of leading and lagging indicators covering process
and personal safety to provide a more well-rounded representation of
platforms' safety status. It puts forth a scoring system that enables perfor-
mance comparison while enabling masking of sensitive information
where necessary. The study also reveals important correlations between
safety factors that contributes to better understanding of the synergy
between the safety factors, hence more efficient safety management.
Acknowledgement
The authors wish to acknowledge the participating industrial practi-
tioners who had voluntarily rendered great effort to provide safety
performance data of offshore oil platforms in the format required.
Conflict of interest
The study was not subject to the funding of any grant. No association
had been made between the data used and any oil and gas companies.
References
Arezes, P. M., & Miguel, A. S. (2008). Risk perception and safety behavior: A study in an
occupational environment. Safety Science,46(6), 900–907.
Baker, J. (2007). The report of the BP U.S. refineries independent safety review panel. Re-
trieved from http://www.csb.gov/assets/1/19/Baker_panel_report1.pdf.
Bellamy,L. J. (2015). Exploring the relationship between major hazard, fatal and non-fatal
accidents through outcomes and causes. Safety Science,71,93–103.
Bergh, L. I. V., Ringstad, A. J., Leka, S., & Zwetsloot, G. I. J.M. (2014). Psychosocial risks and
hydrocarbon leaks: An exploration of their relationship in the Norwegian oil and gas
industry. Journal of Cleaner Production,84(2014), 824–830.
Bertalanffy, L. (1971). General safety theory. UK: Penguin Press.
Beus, J. M., Payne, S. C., Bergman, M. E., & Arthur, W., Jr. (2010). Safety climate and inju-
ries: An examination of theoretical and empirical relationships. Journal of Applied
Psychology,95(4), 713–727.
Christian, M. S., Bradley, J. C., Wallace, J. C., & Burke, M. J. (2009). Workplace safety: A
meta-analysis of the roles of person and situation factors. Journal of Applied Psychology,
94(5), 1103–1127.
Cox, S., & Tait, R. (1991). Reliability, safety and the human factor. Stoneham, MA: Butterworth-
Heineman.
CSB (2012). Offshore safety performance indicators –Preliminary findings on the Macondo
incident [PowerPoint Slides]. Retrieved from http://www.csb.gov/UserFiles/file/
MacKenzie%20Presentation.pdf.
Feng, Y., Teo, E. A. L., Ling, F. Y. Y., & Low, S. P. (2014). Exploring the interactive effects of safety
investments, safety culture and project hazard on safety performance: An empirical
analysis. International Journal of Project Management,32(6), 932–943.
Frens, A., & Berg, W. V. (2014). Asset integrity and process safety management [PowerPoint
Slides]. Retrieved from http://euoag.jrc.ec.europa.eu/files/attachments/01-nam-shell-
euoag_-_independent_verification_in_uio-t.pdf.
Guldenmund, F. W. (2000). The nature of safety culture: A reviewof theory and research.
Safety Science,34(2000), 215–257.
Gupta, J. P., & Edwards, D.W. (2002). Interently safer design: Present and future.Chemical
Engineers Journal,80(B), 115–125.
Hammer, W.(1989). Handbook of system and product safety. New Jersey:Prentice-Hall Inc.
Eaglewood Cliffs.
Hassan, Z., & Abu Husain, M. K. (2013). Risk-based asset integrity management for off-
shore floating facilities in Malaysia. Proceedings of the 2nd international conference
on engineering business management 2013 (ICEBM 2013). Kuala Lumpur, Malaysia:
Universiti Teknologi Malaysia.
Hassan, J., & Khan, F. (2 012). Risk based ass et integrity indic ators. Journal of Loss
Prevention in the Process Industries,25(2012), 544–554.
HSE (2006). A guide to the offshore installations (safety case) regulations 2005. United
Kingdom: Health and Safety Executive.
HSE (2008). Key programme 3 –Asset integrity programme. United Kingdom: Health and
Safety Executive.
HSE (2009). Key programme 3 –Asset integr ity: A review of industry's progress. United
Kingdom: Health and Safety Executive.
HSE (2013). Managing for health and safety. Retrieved from http://www.hse.gov.uk/
pubns/priced/hsg65.pdf.
HSE (2015). Injury frequency rates. Retrieved from http://www.hse.gov.uk/statistics/
adhoc-analysis/injury-frequency-rates.pdf.
HSE (2017). What is competence? Retrieved from http://www.hse.gov.uk/competence/
what-is-competence.htm.
IBM (2014). IBM SPSS Statistics 22 brief guide. Retrieved from http://www.umass.edu/
statdata/software/spss/manuals/IBM_SPSS_Statistics_Brief_Guide.pdf.
ILO (2001). Guidelines on occupational safety and health management system. Geneva: Interna-
tional Labour Organization.
IOGP (2016). Safety performance indicators –2016 data. United Kingdom: International
Association of Oil & Gas Producers.
IPIECA-OGP (2012). Performance indicators for fatigue risk management systems –Guid-
ance document for the oil and gas industry. Retrieved from http://www.iogp.org/
pubs/488.pdf.
Jager, R. (2013). Health, safety & environment in shell: The role of leadership [PowerPoint
Slides]. Retrieved from http://www.zeroharm.org.nz/assets/docs/events/2013/Peer-
Learning-Event-STOS-Shell-13-Aug-2013.pdf.
Johnson, W. G. (1980). MORT safety assurance system. New York: Marcel Dekker Inc.
Kawka, N., & Kirchsteiger, C. (1999). Technical note on the contribution of sociotechnical
factors to accidents notified to MARS. Journal of Loss Prevention in the Process
Industries,12(1), 53–57.
Kletz, T. A. (1999). The origins and hist ory of loss preventi on. Process Safety and
Environmental Protection,77(B), 109–116.
Lauder, B. (2012). Major hazard(asset integrity) key performance indicators in use in the
UK offshore oil and gas industry. Retrieved from http://www.csb.gov/UserFiles/file/
Lauder%20(OGUK)%20-%20Paper%20-%20printed.pdf.
Liu, H., Burns, R. M., Schaefer, A. G., Ruder, T., Nelson, C., Haviland, A. M., ... Mendeloff, J.
(2010). The Pennsylvania certifie d safety committee program: An evaluation of
participation and effects on work injury rates. American Journal of Industrial Medicine,
53(8), 780–791.
Liu, X., Zhu,X. H., Qiu, P., & Chen, W. (2012). A correlation-matrix-based hierarchical clus-
tering method for functional connectivity analysis. Journal of Neuroscience Methods,
211(1), 94–102.
Martinovich, T. (2013). Factors influencing the incidence rates of injuries and accidents
among seafarers and rig workers providing support to the WA offshore oil and gas
industry. Retrieved from http://ro.ecu.edu.au/theses/1084.
McVittie, D. J., Vi, D. P., & Eng, M. (2009). The effect of supervisory training on lost-time
injury rates in construction. Construction Safety Association of Ontario.
Mearns, K., Whitaker, S., & Flin, R. (2003). Safety climate, safety management practice,
and safety performance in offshore environments. Safety Science,41(2003), 641–680.
Morrow, S. L., Koves, K. G., & Barnes, V. E. (2014). Exploring the relationship between
safety culture an d safety performan ce in U.S. nuclear pow er operations. Sa fety
Science,69(2014), 37–47.
Olsen, E., Naess, S., & Hoyland, S. (2015). Exploring relationships between organizational
factors and hydroc arbon leaks on offshore platform. Safety Science,80(2015),
301–309.
Parkes, K. R. (2010). Offshore working time in relation to performance, health and safety: A
review of current practice and evidence. Report RR772. London, UK: Health and Safety
Executive.
Petronas (2015). Sustainability report 2015 –Thriving in tough times. Kuala Lu mpur,
Malaysia: Petroliam National Berhad.
Podgorski, D. (2015). Measuring operational performance of OSH management –Adem-
onstration of AHP-based selection of leading key per formance indicators. Safety
Science,73(2015), 146–166.
Ratnayake, R. M. C. (2012). Modelling of asset integrity management process: A case
study for computing operational integrity preference weights. International Journal
of Computational System Engineering,1(1), 3–12.
Reason, J. (1997). Managingthe risks of organizational accidents. England: Ashgate Publish-
ing Limited.
Reiman, T., & Pietikainen, E. (2012). Leadingindicators of system safety –Monitoring and
driving the organizational safety potential. Safety Science,50(2012), 1993–2000.
Reiner, A. M., Gerberich, S. G., Ryan, A. D., & Mandel, J. (2016). Large machinery-related
agricultural injuries across a five-state region in the Midwest. Journal of Occupational
and Environmental Medicine,58(2), 154–161.
Rentch, J. R. ( 1990). Climate and culture:Interaction and qualitative differences in organi-
zational meeting. Journal of Applied Psychology,75(1990), 668–681.
Shannon, H. S., Mayr, J., & Haines, T. (1997). Overview of the relationship between
organisationa l and workplace facto rs and injury rates. Safety Science,26(1997),
201–217.
Sklet, S. (2006). Hydrocarbon releases on oil and gas production platforms: Release sce-
narios and safet y barriers. Journal of Loss Prevention in the Process Indu stries,19
(2006), 481–493.
Tang, D. K. H.,Leiliabadi, F., Olugu, E. U., & Md Dawal, S. Z. (2017). Factors affecting safety
of processes in the Malaysian oil and gas industry. Safety Science,92(2017), 44–52.
Transportation Research Board US (2016). Strenthening the safety culture of the offshore oil
and gas industry (technical report). Retrieved from https://www.nap.edu/catalog/
23524/strengthening-the-safety-culture-of-the-offshore-oil-and-gas-industry.
Venkataraman, N. (2008). Safety performance factor. International Journal of Occupational
Safety and Ergonomics,14(3), 327–331.
Vinnem, J. E. (1998). Use of performance indicators for monitoring HS E operating
achievement. In S. Lysersen, G. K. Hansen, & H. Sandtorv (Eds.), Safety and reliability:
Proceedings of the European conference on safety and reliability (IESRE' 98). Trondheim,
Norway: CRC Press.
18 D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19

Williams, O. S., Hamid,R. A., & Misnan, M. S. (2017). Analysisof fatal building construction
accidents: Cases and causes. Analysis,4(8), 8030–8040.
Williamsen, M. (2003). Revisiting Heinrich's “accident triangle”.ISHN,37(2), 27.
Witt, L. A., Hellman, C., & Hilton, T. F. (1994). Management influences on perceived safety.
Paper presented at the Annual Meeting of the Ame rican Psychologi cal Society, San
Francisco, CA.
Daniel Tang Kuok Ho is a PhD student under the Department of Mechanical Engineering
in the Faculty of Engineering of University of Malaya, Malaysia. He is also a lecturer of
Curtin University Malaysia. His research interest includes health and safety management,
offshore safety, and sustainability.
Siti Zawiah Md Dawal is an associate professor under the Department of Mechanical
Engineering in the Faculty of Engineering, University of Malaya, Malaysia. She has over
160 publications. Her area of expertise cover s work design, ergonomics, and flexible
manufacturing systems.
Ezutah Udoncy Olugu is currently engaged as an assistant professor in the Faculty of
Engineering, Technology & Built Environment, UCSI University, Malaysia. Dr. Olugu's
research interest includes green manufacturing, sustainable production, green supply
chain management, reverse logistics, and totalquality management.
19D.K.H. Tang et al. / Journal of Safety Research 66 (2018) 9–19