Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Contents lists available at ScienceDirect

Process Safety and Environmental Protection

journal homepage: www.elsevier.com/locate/psep

A risk-based shutdown inspection and

maintenance interval estimation considering
human error
Abdul Hameed, Faisal Khan , Salim Ahmed
Safety and Risk Engineering Group, Faculty of Engineering & Applied Science, Memorial University, St Johns,
Canada, A1B 3X5

a r t i c l e

i n f o

a b s t r a c t

Article history:

This paper presents a risk-based methodology to estimate shutdown inspection and main-

Received 27 May 2015

tenance interval by integrating human errors with degradation modeling of a processing

Received in revised form 14 October

unit. The methodology presented in this paper addresses to identify the number of shut-

2015

down intervals required to achieve a target reliability over a goal period. The proposed

Accepted 30 November 2015

methodology is the extension of the previously published work by the authors to deter-

Available online 17 December 2015

mine the shutdown interval considering the systems desired availability. The proposed
work is novel in the sense that a concept of human error during shutdown inspection and

Keywords:

maintenance is introduced while modeling the system failure. Selection of critical equip-

System reliability

ment is the most important aspect in obtaining the shutdown interval to minimize overall

Risk-based inspection and

operational risk. In order to achieve this, a risk criticality matrix is proposed to select the

maintenance

critical equipment for shutdown inspection and maintenance. Probability of human error

Human error and shutdown

induced during shutdown inspection and maintenance is estimated using Success Likeli-

inspection and maintenance

hood methodology (SLIM). The proposed methodology is composed of three steps namely,

Failure probability

equipment selection considering criticality of operation, system failure modeling consid-

Consequence

ering human error and nally a risk-based shutdown inspection and maintenance interval

Failure model

estimation. The proposed methodology is applied to a gas chilling and liquefaction unit of
a hydrocarbon processing facility. The methodology is used to ensure the practicality of the
proposed formulation to the real industry. The proposed methodology can be applied to any
plant (process or non-process) such as those for LNG processing, petrochemicals, reneries
or manufacturing plants. The key elements for the success of the proposed methodology are
the identication and selection of critical equipment, breakdown of activities to estimate
human error probability and plant-specic data for modeling system failures.
2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1.

Introduction:

Due to continuous production demands, processing facilities are getting not only bigger and bigger but also more
complex in nature. The increase in complexity and size is
inviting maintenance and reliability engineers to put more
emphasis on system inspection and maintenance optimization to minimize unplanned downtime, overall cost and risk

exposure. Effective inspection and maintenance is one of

the critical elements for operating facilities. The core objective of inspection and maintenance is to make sure that the
facilities or equipment are optimized in a way, which does
not only increase the reliability and availability of the plant
but also minimizes the overall operational risk. Taking the
unit or facility out of the service, generally termed as shutdown, performs inspection and maintenance on some of the

Corresponding author. Tel.: +1 709 864 8939.

E-mail address: [email protected] (F. Khan).
http://dx.doi.org/10.1016/j.psep.2015.11.011
0957-5820/ 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

10

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

equipment. Duffua and Daya (2004) and Lawrence (2012) have

stated that a planned periodic shut down is carried out to
perform maintenance and to inspect, test and replace process materials and equipment. Inspection and maintenance
strategies of the equipment, which do not require facility
to be taken in shutdown mode, can be developed based on
individual equipment. Shutdown interval is one of the most
important factors in determining an effective inspection and
maintenance policy. In case if the shutdown inspection and
maintenance interval is too short, facility shutdown time and
production loss along with the inspection and maintenance
cost will be too high, vice versa if the shutdown interval
is too long, the production loss and inspection and maintenance cost will be low but the risk exposure will be high.
This leads to nd an optimal solution for shutdown inspection and maintenance interval. Failure of equipment may
lead to signicant consequences due to improper planning.
Understanding the facilities system from operation and safety
is the most important faucet when selecting and designing
a shutdown inspection and maintenance model. A typical
processing facility consists of hundreds of equipment, which
works, in rigorous environment. One of the key aspects, which
should be covered and included when modeling for shutdown inspection and maintenance optimization, is to include
human error and its impact on the equipment or system
failure. Integration and design of the systems such as acting in series, parallel, combination of seriesparallel, 50%
load capacity or 100% load capacity dictate the development
of shutdown inspection and maintenance strategy for the
processing plant. Inspection and maintenance operation is
one of the key links in the process chain for achieving the
required production and management goals. While performing inspection and maintenance, a minor failure and omission
in following a clear guideline or process not only minimizes all
of the inspection and maintenance benets but also increases
and changes the failure rate or behavior of the equipment or
system due to introduction of human error. Despite technological advancement in equipment design and consideration
given for maintainability, manmachine interface cannot be
eliminated. In general, any inspection and maintenance process involves disassembly, reassembly and/or replacement
of components. These processes require human interaction
and, under various circumstances, create potential to include
human error by installing or replacing a wrong part or assembling the part in wrong sequence despite all technological
enhancement. In this paper, the focus is on the group of equipment which cannot be inspected or maintained and requires a
shutdown of the facility. Thus, in order to develop an optimal
inspection and maintenance strategy, attention must be paid
while selecting these critical equipment.

2.

Past studies

Inspection and maintenance optimization has gained huge

momentum and dynamic changes over the last couple of
decades due to the realization of potential benets in plant
availability, reliability, scheduling, cost and risk minimization.
Risk, reliability and availability are the three facet of facility
operation and are interminably linked together. A high risk is
generally an indication of facility lower reliability and availability, while higher availability means higher reliability and
lower risk. Operation risk is associated with the probability
of equipment or component failure and the consequences

of failure such as loss of revenue due to production loss,

asset damages, safety and health issues and inspection and
maintenance costs. Obiajunwa (2012) reported that typically,
power plant turnaround maintenance is planned for every
four years, oil renery and petrochemical plant shutdown
maintenance is planned for every two years, and chemical,
steel, glass and food and beverage plant shutdown maintenance is planned for every year. Alsyouf (2007) presented a
model enabling the decision-makers to identify how an effective maintenance policy could inuence the productivity and
protability through its direct impact on quality, efciency and
effectiveness of operation. Backlund and Hanu (2002) reported
that while doing the risk analysis, focus must be put on the
function required of the subsystem and equipment. Fujiyama
et al. (2004) proposed a risk-based maintenance system for
steam turbine plants which is coupled with an inspection system. Ghosh and Roy (2009), Rusin and Wojaczek (2012), Vaurio
(1995), Khan and Haddara (2003, 2004a,b), Krishnasamy et al.
(2005), Tan and Kramer (1997), Duarte et al. (2006) and Vatn
et al. (1996) have presented methods to estimate the optimal maintenance and inspection interval considering cost,
risk, availability and reliability for individual equipment and
have not considered the impact of facility shutdown. Neil and
Marquez (2012) proposed a hybrid Bayesian network (HBN)
framework to model the availability of renewable systems
considering corrective repair time, logistics delay time and
scheduled maintenance. These were combined with timeto-failure distributions using HBN. Mannan and Yang (2010)
proposed a dynamic risk assessment (DORA) methodology
considering various process variables such as level, ow rate,
temperature, pressure and chemical concentration and their
impact to guide and improve the process design and optimize failure probability. However, the proposed methodology
is not considering whether a sequence of component failure
will lead to the system failure. The uniqueness of the presented methodology is that it helps to optimize the shutdown
inspection and maintenance interval to minimize the overall system failure which will lead to reduce the un-necessary
shutdown. Jacob and Amari (2005) presented a binary decision
diagram to calculate system reliability and availability. Pil et al.
(2008) proposed a redundancy optimization and maintenance
strategies based on a time-dependent Markov approach. Khan
and Haddara (2003) proposed a comprehensive and quantitative methodology for risk-based maintenance. Dey et al.
(1998) and Dey (2001) have applied risk-based approach to
the maintenance of oil pipelines. Khan et al. (2008) have presented a risk-based methodology to maximize a systems
availability by considering risk-based inspection and maintenance program to reduce the risk of failure and enhance
the overall availability of the system. Sarkar and Behra (2012),
Bertolini et al. (2009), Kumar and Chaturvedi (2008), Zhaoyang
et al. (2011) and Wang et al. (2012) proposed that selecting
a maintenance strategy based on risk reduces the overall
risk. However, most of these studies are concerned with optimizing equipment inspection and maintenance cycles based
on perfect (AGAN) as good as new or minimal (ABAO) as
bad as old repair. AGAN strategy holds the assumption that
after the maintenance intervention, the system starts its life
under the same failure rate as if it were new. On the other
hand, ABAO holds that the equipment or system is maintained with minor action, which has not changed the failure
rate behavior, and after the maintenance activity, the failure
rate remains the same as it was before the maintenance. In
order to overcome the short fall of AGAN or ABAO strategy,

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

researchers have introduced concept of imperfect maintenance. Nguyen and Murthy (1981) introduced the concept of
imperfect maintenance considering that the failure characteristic of the system is different (worse) from that of correctively
maintained system to minimize the mean cost rate. Block et al.
(1985) extended Brown and Proschan model and proposed an
age-dependent repairable model considering a probability of
p(t) a complete repair or with probability q(t) = 1 p(t) is a minimal repair, where, t is the age of the equipment in use at
the failure time. Ben-Daya and Alghamdi (2000) presented two
sequential preventive maintenance model considering age
reduction model, an extension of Nakagawas model, while
in the second model, PM intervals are dened such that the
integrated hazard rate over each interval is the same for all
intervals. Levitin and Lisnianski (1999) have used Genetic Algorithm to generalize a preventive maintenance optimization
to multi-state system considering the effective age of equipment. Nakagawa et al. (2012) considered system damages due
to shock (a shock number N) and a damage level k and proposed a preventive model considering imperfect maintenance.
Rangan and Grace (1989) extended Brown and Proschan model
to develop a replacement policy for a deteriorating system
with imperfect maintenance. Li and Shaked (2003) extended
Brown and Proschan (1983) imperfect maintenance approach
to model preventive maintenance and obtained stochastic
maintenance comparisons for the number of failures under
different policies via a point process approach. Malik (1979)
introduced the concept of improvement factor assuming that
maintenance action changes the system time of the failure
rate curve and the failure rate post maintenance lies between
as good as new and as bad as old. Brown and Proschan (1983)
reported that some possible causes for imperfect, worse or
worst maintenance due to the maintenance performer, such
as repair the wrong part, only partially repair the faulty part,
repair the faulty part but damage the adjacent part, are the
true contributor. Nakagawa and Yasui (1987) reported that
hidden failures which are not detected during the maintenance, human errors such as wrong adjustment and further
damage done during maintenance and replacement with
faulty parts. Dhillon (1986) reported that operation, assembly design, inspection, installation and maintenance are all
prone to human errors. These errors are due to poorly written
maintenance procedures, complex maintenance tasks, harsh
environment, fatigue, outdated maintenance manual, inadequate training and experience. Dhillon and Liu (2006) reported
that reasons for the occurrence of human errors including
inadequate lightning in the work area, inadequate training
or skills of the manpower involved, poor equipment design,
high noise level, an inadequate work layout, improper tools
and poorly written equipment maintenance and operating
procedure. They further classied human errors into six categories: (1) operating error (2) assembly error (3) design error
(4) inspection error (5) installation error and (6) maintenance
error. Dhillon and Yang (1995) cited that failure of repairable
system can occur not only due to hardware failure but also
due to operating human error or maintenance error. Factors
such as temperature, dust, fatigue, incomplete or inappropriate maintenance tools, errors in inventory and personal
problems may be the contributor for these errors. Noroozi et al.
(2014, 2013) used Human Error Assessment and Reduction
Technique for evaluating human error risk assessment and
applied to pre- and post-maintenance procedure of a process
facility and has also applied Success Likelihood method (SLIM)
to perform human factors analysis in pre- and post-pump

11

maintenance activities for offshore facility. The above-referred

literature covers inspection and maintenance interval modeling considering cost, reliability, availability and risk, imperfect
maintenance and human error. However, majority of these
have discussed inspection and maintenance based on equipment by equipment basis. In reality, a process plant, a unit
or system consists of hundreds of pieces of equipment that
run in continuous mode. Further, possible human errors in
the inspection and maintenance activities can impact equipment performance. For example, poor repair can play an
instrumental role in increasing the number of equipment
breakdowns or failure pattern which in turns can signicantly
increase the risk associated with equipment. Understanding
the probability of human error while performing inspection
and maintenance and including in the modeling will provide
a better sight to reduce the overall risk and increasing reliability, availability and safety. Developing an inspection and
maintenance strategy without considering the impact of these
inspection and maintenance due to the shutdown of the plant
and the associated risk due to human errors will not produce
an optimum interval. This problem could be solved by considering a risk-based critical equipment selection and then by
developing an optimized shutdown inspection and maintenance interval considering the imperfectness due to human
error for the system or unit. The objective of this paper is
to develop a risk-based shutdown inspection and maintenance optimization methodology by integrating human error
probability in the system failure model for a continuous
processing facility. This proposed methodology will provide
a rational basis to make a shutdown inspection and maintenance decision-making considering human error contribution
in inspection and maintenance and the overall risk exposure.

3.
A risk-based inspection and
maintenance modeling considering human
error
Although several studies have been published on inspection
and maintenance interval modeling and optimization in the
literature as discussed above, a majority of these works deals
with addressing the individual equipment. Consideration of
shutdown inspection and maintenance modeling is found to
be limited. Hameed et al. (2014) and Khan and Hameed (2014)
has proposed a risk-based shutdown interval modeling for
continuous operating facilities. These modelings are based on
the assumption of the perfect maintenance. In reality, majority of the time inspection and maintenance does not meet the
perfect conditions and fall short. The result of imperfectness
is induced due to the contribution of human errors. Errors
induced during inspection and maintenance may be realized
immediately and result in the premature failure of the equipment or system or in some cases may lie dormant with the
equipment or system for some period of time, until a combination of other factor accelerate the degradation mechanism and
lead to a failure. Sometime these have resulted in serious accidents in process industries. The proposed model is developed
considering the human error which may be induced during the
shutdown inspection and maintenance and is the extension of
Hameed et al. (2014) and Khan and Hameed (2014). The objective of this paper is to integrate human errors in the system
failure model and to interlink this with the operational risk.
Higher the probability of failure of inducing the human error
in inspection and maintenance, the degradation or the system failure probability will be more impacted. The proposed

12

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Fig. 1 Risk-based inspection and maintenance interval optimization model considering human error.
shutdown inspection and maintenance methodology is broken down in three modules and shown in Fig. 1. Assumption
in developing the model as well as detail of each module is
described below.

3.1.

Assumptions for model development

A group of equipment which cannot be inspected or maintained without taking the plant into non-operational mode
(shutdown).
Failure as a time-dependent process.
Material, labor, specialist and production cost are available.
The degree of imperfectness is assumed to be a number
between 0 and 1 due to human error while performing
inspection and maintenance.
Sufcient manpower and equipment are available to execute the planned job.
Inspection and maintenance durations are non-negligible.

3.2.
Module 1: equipment selection considering
criticality of operation
A processing facility consists of a large number of interacting systems or equipment. When these system or equipment

performs together, the facility delivers its intended function.

These systems or equipment are subject to periodically or
non-periodically planned inspection and maintenance during
its life cycle. The inspection and maintenance management is
one of the key decisions for continuously operating plants to
improve the plant reliability, availability and integrity. Some
of these system or equipment cannot be isolated to perform
inspection and maintenance, and requires the plant to be
taken in shutdown mode. The remaining systems and units
do not impact the overall facility operation and can be taken
out for inspection and maintenance without causing facility
operation shutdown. Thus, it is very critical to identify system or equipment for their inspection and maintenance to
minimize the impact due to operation loss and to achieve
required reliability and availability. The success to achieve
an optimal inspection and maintenance plan for the facility
depends on identifying and selecting these critical equipment.
In this paper, a risk criticality matrix is proposed to select
these equipment. The uniqueness of the proposed matrix is
to identify these critical equipment and system which will not
only help to reduce the utilization of inspection and maintenance resources but also help to increase the reliability and
availability of the unit. The process consists of the following
steps:

13

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

3.3.1.

Table 1 Five level of probability of failure (Hameed

et al., 2014).
Possible qualitative rank

Annual likelihood of failure

Frequent
Probable
Occasional
Remote
Extremely unlikely

>0.1
0.0010.01
0.00010.01
0.000010.0001
<0.000001

(1) Developing a boundary diagram by breaking down units or

plant into manageable system.
(2) Reviewing all equipment in the selected boundary using
a risk matrix to establish the criticality in relation to the
operation of facility, production loss and impact on asset
damage. The proposed criticality matrix is shown in Fig. 2.
(3) Estimating the qualitative risk criticality number and comparing with the acceptable criteria.
(4) If the risk number does not meet the acceptable criteria,
include the equipment in shutdown inspection and maintenance planning.
(5) Continue the process until all the equipment in the
selected boundary has been analyzed.
The advantage of this qualitative risk assessment helps to
categorise the equipment that will require a facility shutdown.
For qualitative risk analysis, a criticality risk number may be
developed depending on the company operational considerations. According to API (2009), risk-based inspection, for a
qualitative risk analysis, the likelihood of failure may be categorized from one through ve. However, it is appropriate to
associate an event frequency with each likelihood category
to provide guidance to determine the probability of failure as
shown in Table 1 and consequences are represented in monetary terms ($) as shown in Table 2.
For the case of several competing consequences for
equipment/components, the highest observed risk (criticality
number) among the consequences should be considered to be
the most critical component. Overall risk, R, can be selected
using below equation:
RISK = MAX (RISKAD ,RISKPL ,RISKOC , )

(1)

3.3.
Module 2: component/system failure modeling
considering human error
Failure characteristics of the component or system are one of
the most important parameters used to analyze and model the
system failure or behavior. Since failure of a system is a probability, the rule of probability theory may be applied to compute
the system failure probability from knowledge of component
or system characteristics.

Table 2 Five level consequence table (Hameed et al.,

2014).
Possible qualitative rank
Negligible
Minor
Moderate
Major
Catastrophic

Economic loss range

EL < $10,000
$10,000 < EL $100,000
$100,000 < EL $1,000,000
$1,000,000 < EL $10,000,000
EL > $10,000,000

System failure modeling

A unit or system may constitute pieces of equipment arranged

in series, in active redundancy or in standby. Very often, these
redundancies or standby is designed to take full load in case of
a functional failure to avoid unplanned shutdown. Operational
relationships and knowledge regarding various system elements is required to develop a system failure model. System
failures cannot be evaluated and improved until it is known
that how these various elements affect system operation.
A true representation of these relationships is required for
prediction and assessment based on either cost or risk. A functional block diagram is developed once the critical equipment
are identied to represent their relationships and provide an
indication of the element which must operate successfully for
the system to accomplish its intended function. Depending
on the functional block diagram, system reliability or failure
model equation can be developed using either standard series
or seriesparallel equations. Further, most often, inspection
and maintenance modeling is considered based on a xed
time interval between two consecutive inspection and maintenance. In this paper, it is assumed that the inspection and
maintenance interval will be dictated by the risk considered
and will vary between two consecutive shutdown inspection
and maintenance as shown in Fig. 3. Considering the variability in shutdown interval, the reliability of the system over the
life cycle can be written as follows:

Rsys (t) =

n


R (i ) R

i=0


t

n



i

(2)

i=0

where,

(1) t >

n


i

i=0

(2) i = 1, 2, 3 . . .. . . ..n is shutdown inspection and maintenance

interval
(3) 0 = 0
(4) R (0) = 1
(5) 1 > 2 > 3 >. . .. . .> n

3.3.2. Human error modeling in inspection and

maintenance
Inspection and maintenance activities are critical to improve
the reliability and availability of equipment. These activities are performed not only under immense pressure to
bring the facility up and running in shortest period of time
but also under difcult and hazardous conditions. Even
with all kind of technical advancement while designing the
equipment or system, human involvement needs to be considered when performing inspection and maintenance. Human
interactions with machines or systems are prone to introduce error while performing inspection and maintenance
due to various factors. Human errors during inspection and
maintenance activities have already produced disastrous outcomes (in millions of dollars) such as Flixborough, Three
Mile Island, Piper Alpha and Bhopal accident. Sometime
human errors in inspection and maintenance may lie in
dormant mode for a longer period of time before leading the equipment or system to failure. An example of a
loosely secure nut during maintenance may produce vibration and results in a fatigue crack over a period of time.
Human errors such as misinterpretation of engineering drawings and maintenance manuals, inadequate training, poor

14

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Fig. 2 Qualitative criticality risk ranking matrix.

working environment, time constraint and working environment as well as processing hazards are some factors which
impact human performance. Swain and Guttmann (1983) have
dened these factors as performance shaping factor (PSF) and
listed several PSF which are linked with internal, external, or
stress related for complex manmachine interface for nuclear
power plants. Wilson and McCutcheon (2003) reported that
facility layout, workstation conguration, controls, hand tools,
control systems, noise, vibration, lighting, temperature, force,
repetition, posture, work schedule/workload, behavior based
safety, labels/signs, communications, training, stress/fatigue,
motivation, tness/body size) are the areas where human
factor should be considered. Toriizuka (2001) evaluated the
importance of each PSF from the viewpoint of work efciency, workload and human reliability. Any inspection and
maintenance activity whether non-shutdown or shutdown
for a complex system consists of three major steps, namely
preparation activities, inspection and maintenance activities
and boxing up/lining up activities before it is taken in operation. The above three steps consist of several sub-sets of
task. Since these actions are performed by human, there is
always a probability of introducing an error. If the probability
of inducing such human error is analyzed considering various PSF and represented by p, it can be integrated in to the
system failure function of the equipment or system. Various techniques have been presented in literature to estimate
the human error probability. Some of the major techniques
are Success Likelihood Indexing Method (SLIM), Technique for
Human Error Rate Production (THERP), Justied Human Error
Data Information (JHEDI) and Human Error Assessment and
Reduction Technique (HEART). Kirwan (1996, 1997) and Kirwan
et al. (1997) have discussed and validated these techniques in
TSD1

tup

td

TSD2

tup

TSDn

TSD3

td

tup

td

tup

Td

Fig. 3 System operational availability.

details. These techniques have been applied in nuclear, air and

process industries. In this paper, SLIM technique is utilized
to estimate human error probability and later integrated in
the system failure probability. Fig. 4 represents the SLIM process. SLIM process is based on developing the Performance
Shaping Factors (PSF) and their impacts on human behavior.
Generally, an expert judgment is used to quantify PSF which is
used to derive a Success Likelihood Index (SLI) for each activity. A detailed set of task needs to be developed. Each task is
reviewed in view of the considered PSF. Field and Technical
expert judgments are used to assign a ranking and weighting
to each PSF in terms of the inuence on the success of a task.
Eq. (9) is used to convert estimated SLI to estimate human
error probability.
SLIk =

Rjk Wj

(3)

LOG (HEP) = a SLI + b

(4)

Where, SLIk is the SLI for each activity, Wj is the importance

weight for the jth PSF, Rjk is the scaled rating of task k on the jth

Select set of Performance

Shaping Factor

Calculate PSF ranking

Determine PSF Weight

Determine SLI for each task

Predict HEP for each acon

td

Calculate overall HEP

Fig. 4 SLIM methodology to calculate HEP.

15

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Fig. 5 Impact of human error in survival function.

PIF. The overall probability of human error in inspection and
maintenance can be calculated using Eq. (5):


Fsys (t) = 1

n


R (i ) R

i=0

p=

m


(1 HEPk )

(5)

k=1

Incorporating inspection and maintenance human error

probability in Eq. (6) will result in:

Rsys (t) =

n


R (i ) R

i=0


t

n



i

i=0

n


(1 p)

(6)

i=1

Fig. 5 represents the impact of human-induced error in

the survival function of the system and is compared with no
inspection and maintenance.

3.4.
Module IIIrisk-based shutdown interval
estimation
Too frequent inspection and maintenance of equipment or
system will increase not only the loss in revenue but also the
overall operational risk exposure due to possible human error
introduction in the system. On the other hand, an optimal
inspection and maintenance interval will not only help to have
better utilization of the inspection and maintenance resources
but also reduce the risk of increased failure. It is necessary to
express the probability of failure for a piece of equipment or a
system as a function of time for risk-based shutdown inspection and maintenance interval estimation. In order to develop
a risk-based shutdown inspection and maintenance interval,
system failure probability considering human error and consequences are required to generate the overall risk exposure
and are discussed below.

3.4.1.

Estimation of system failure probability

In general, operating companies keenly focus on targeting to

maintain the desired reliability to achieve their operational
goal. Since the system failure probability is a function of failure rate and is complementary to the reliability of the system,
mathematically it can be written as
Fsys (t) = 1 Rsys (t) = Pr{T < t}

(7)


t

n

i=0


i

n



(1 p)

(8)

i=1

where, t is the failure time of the system on or before a time

Tt
A system or equipment failure can be modeled using exponential, Weibull, normal or lognormal probability distribution.
Plant-specic inspection and maintenance data are the best
source to identify the model. However, sometime due to the
limited availability of plant-specic data, test data, data bank
or expert judgment is also used. Weibull distributions due
to its inherent exibility such as normal (for = 3.4) and
exponential (for = 1) distributions is most commonly used
to model system failure probability. In this study, the Weibull
model with the parameters and  is used to model the timedependent reliability of the equipment involved in the system.
A decreasing failure rate (1) corresponds to an early life failure
or infant mortality. A constant failure rate ( = 1) suggests that
items are failing from random events. An increasing failure
rate ( > 1) suggests that wear out is occurring and that parts
are more likely to fail over time (Ghosh and Roy, 2009).

3.4.2.

Estimation of the economic consequences

In general, process plants such as reneries, chemical plants,

natural gas processing facilities equipment operate under
extreme conditions (pressure and temperatures). It is due to
these extreme operating conditions, failure consequences are
not limited to the localized failure of component or equipment. Further, these facilities process hazardous substances
(hydrocarbons), failures due to leak of hydrocarbons may produce sever impact on the nearby assets, safety and health
issues of the people and operational losses in the presence
of an ignition source. For example, in a gas processing plant,
hydrocarbon release may result in ash res, jet res, pool
res, boiling liquid expanding vapor explosion (BLEVE), re
balls, vapor cloud explosions or toxic dispersion. In order to
estimate the impact in the effected zones, operating pressure,
temperature, physical and chemical properties and atmospheric conditions are required. Economic consequences due
to these may be estimated based on the effects of thermal
radiation and overpressure on surrounding equipment and
personnel in monetary ($) terms for effected zones. Failure
to consider these consequences when developing inspection
and maintenance strategies for the facility may increase the

16

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Table 3 Consequence estimation (Hameed et al., 2014).

Economic consequence of asset loss

Due to heat ux
Due to Overpressure
Due to heat ux
Due to overpressure

Economic Consequence of Human Health

Loss (ECHHL)
Economic Consequence of Production Loss (ECPL)
Economic Consequence of Inspection and
Maintenance (ECI&M)

The total economic consequences of shutdown (ECT )

(ECAL)x = (PDHF)x AD
ECAL = PDOP EZ AD
(ECHHL)x = (PIHF)x PD CF
ECHHL = PDOP EZ PD CF
ECPL = SDT PL SP
Cp(i) = Clp tp ,
CI (i) = ClI ti + CIe ti ,
Cm (i) = Clm tm + Csp ,
Cts(i) = Clts tts ,
ECSI&M(i) = Cp(i) + CI(i) + Cm(i) + Cts(i) .
ECT = Max (ECAL+ECHHL)i +
n


(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)

ECSI&M+ECPL,

i=1

operational risk. Hameed et al. (2014) has described various

equations to estimate economic consequences in monetary
($) terms. Table 3 lists the summary of these equations which
has been used here.
where, x represents the various possible scenarios, PDHF
is the probability of damage due to heatux in the effected
zones and AD is the asset density, PDOP is the probability of
damage due to overpressure, EZ is the effected zone due to
overpressure, PDHF is the probability of injury or death due
to heatux in the effected zones, PD is the population density
and CF is the cost of injury or fatality, PDOP is the probability of
damage due to overpressure, PL is production loss volume per
day and SP is the selling price of the product per unit volume,
SDT is the shutdown duration, Cp is the preparatory cost, Clp
is the cost of preparatory maintenance labor per hour ($/h),
tp is the duration of the preparatory work, CI is the inspection
cost, ClI is the cost of skilled inspection labor per hour ($/h), CIe
is cost of inspection equipment per hour ($/h), ti is the duration of inspection work, Cm is the maintenance cost, Clm is the
cost of skilled maintenance labor per hour ($/h), Csp is the cost
of spare parts consumed, tm is the duration of maintenance
work, Cts is the technical support cost, Clts is the cost of a technical support specialist per hour ($/h) and t is the duration of
technical support work in hours

3.4.3.

Risk estimation

Since the risk is a combination of probability and consequence, one needs to consider all possible consequences,
including safety and health, operational and non-operational
consequences. In this paper, risk is estimated in terms of ($)
value due to all possible combination and can be written as:

RISK = Fsys (t) Max (ECAL+ECHHL)i +

n


n


meet the energy requirement. This trend is due to the fact

that natural gas (NG) is the cleanest of the fossil fuels. LNG is
a temporarily converted form of natural gas for storage and
shipping because it occupies 600 times less volume. The natural gas liquefaction process begins when the natural gas is
extracted from the underground reservoirs and is sent to a
liquefaction facility, where natural gas is liqueed at 160 o C.
The liquefaction plants are asset intensive which operates on
a continuous basis. Unavailability of major equipment or system due to any failure may have severe consequences and
produce signicant risk to the operating companies due to production loss, asset damage, safety and company perception.
A typical onshore liqueed natural gas processing plant consists of several units as shown in the following Fig. 6. Raw gas
is received from either an onshore or offshore reservoir to the
inlet/receiving area where condensate and water is separated
from hydrocarbon. Hydrocarbon is then pre-treated to remove
corrosive and hazardous contents. These include H2 S, CO2 ,
mercury, helium and water. The dry sweet gas is then cooled
to separate heavier hydrocarbon such as C3 , C4 etc. Finally, it
is cooled in the cryogenic units to liquefy natural gas for storage and/or shipping. In this paper, a section of gas chilling
and liquefaction unit of a LNG processing facility is selected
to develop shutdown inspection and maintenance interval for
a targeted goal time to achieve a desired reliability (0.95). The
general schematic and the block diagram of gas chilling and
liquefaction unit section is shown in Fig. 7 and is taken from
Khan and Hameed (2014) previous work. The main purpose
of the Gas Chilling and Liquefaction Unit is to condense the
sweetened, dry, lean feed gas into LNG in the Main Cryogenic
Heat Exchanger, and then sub-cool it in the Sub-cooling Heat
Exchanger.

ECSI&M

4.1.
Module 1: equipment selection considering
criticality of operation

SDi=1

ECPL

(20)

SDi=1

4.
The application of RBSIM to a LNG
processing facility
Snmez et al. (2013) reported that energy plays a fundamental role in both manufacturing and services, and natural gas is
rapidly becoming a key energy source worldwide. Since the last
few decades, liqueed natural gas (LNG) plants are increasing in number due to the growing demand for natural gas to

Equipment criticality analysis helps to evaluate how failure

of these equipment impacts the performance of asset and is
use to prioritize equipment selection for inspection and maintenance. Module I of the presented methodology is applied
to select critical equipment using the risk criticality matrix.
The risk criticality matrix helps to estimate the risk criticality number for all equipment in the unit for all considered
consequences. Using the risk criticality matrix, equipment
can have a criticality number ranging from 1 to 25. Criticality number 1 indicates that the equipment has least impact
while a criticality number 25 indicate signicant impact on the
risk. For example, if an equipment is having high probability

17

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Fig. 6 Typical LNG plant process ow.

(frequent = 5) of failure and the unavailability of the equipment

having moderate impact on production, 50% loss (moderate = 3), the risk number will be 15, compared to an equipment
which upon failure will result total loss of operation will get
a criticality number of 25. Higher criticality number indicates
that the equipment will have higher risk on operational facilities due to the unavailability, and should be considered for
a shutdown inspection and maintenance. Other risk factors
can also be estimated to obtain the criticality number. In the
considered unit, hydraulic turbines are designed to take fullload capacity, in case if one unit fails, other unit will take the
load, hence it does not get a higher criticality number and
are not considered for shutdown inspection and maintenance
interval optimization. These can be inspected and maintained
any time without taking the unit in shutdown. Equipment E01,
E02, E03, E04, E05 and E06 are estimated to be having higher
critical number and are selected for shutdown inspection and
maintenance interval for the studied unit.

4.2.
Module 2: integrating human error in reliability
modeling
A functional block diagram is developed for the selected
critical equipment and is shown in Fig. 8. System failure
probability is calculated using the block diagram from the

Table 4 The failure characteristics of considered

equipment.
Equipment No.
E01
E02
E03
E04
E05
E06

Characteristic life (, h)

Shape parameter ()

282,000

4.38

450,000

2.0

selected critical equipment. Since these equipment (in this

unit) operate in series, reliability equation in series is appropriate for this scenario. In order to estimate the risk-based
inspection and maintenance interval, failure data is a key. For
this study, failure and repair data for the selected equipment
are taken from the previously published work by Khan and
Hameed (2014) and is listed in Table 4.
All inspection and maintenance activities whether it is
minor services, such as tightening the bolt or loose parts,
cleaning and removing dust and rust, repairing and/or replacing degraded components such as bearing and seals, or
performing welding to strengthen the integrity of the equipment require human interactions. As discussed in Module
II, various factors may introduce human error and result in

18

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Fig. 7 Gas chilling and liquefaction unit.

Table 5 List of considered PSF.

External PSF
Environment
Working hours, work breaks, shift rotations
Availability and adequacy of special equipment, tools and supplies
Method, policies and Procedure
Criticality and complexity of task

E01

E02

E03

E04

E05

Internal PSF

Stressors

Training
Experience
Knowledge of performance standards
Stress (mentally or bodily)
Work memory
Physical condition

Fatigue, pain or discomfort

Temperature and radiations
Oxygen insufciency
Vibration
Movement limitation
Risk and threats

E06

Fig. 8 Functional block diagram of selected equipment.

accelerating the degradation or failure of the system. Table 6
list the considered PSF which may have impacts on the shutdown inspection and maintenance activities. Expert judgment
from the experienced plant personnel are used to identify
the signicant Performance Shaping Factors and are listed in
Table 5. These PSF are then ranked to select the top ve ranked
PSF. Selected top ve PSF were then assigned a weightage
and is listed in Table 6. Table 7 list the general and common
shutdown inspection and maintenance activities which were
developed using eld engineers experience. This information
was used to estimate the HEP for individual activities.

Table 6 Performance shaping factor, rank and weight.

Number
1
2
3
4
5

Performance shaping factor

Rank

Training
Experience
Time pressure
Work memory
Work environment

5
4
3
2
1

Weight
0.20
0.25
0.20
0.15
0.20

Table 7 Common shutdown inspection and

maintenance activities and estimated HEP.
Task
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

Task description

HEP

Receive I&M work permit

Perform system blinding and isolation
Open entry points
Install ventilation system
Install internal/external lighting
Remove manways
Install internal/external scaffolding for I&M
Remove/Clean demisters, (if required)
Clean weld Joints, drains, nozzles etc.
Carry out inspection
Carry out repairing/maintenance works,
Re-x the demisters (if required)
Remove internal scaffoldings
Re-x internal manways
Perform nal inspection
Final internal inspection
Close entry points
De-blinding and hydra tightening
Close I&M work permit

0.001594
1.47E-05
0.00029
0.00029
1.47E-05
1.47E-05
1.47E-05
0.000359
0.010839
1.47E-05
1.47E-05
1.47E-05
1.47E-05
0.010839
5.07E-06
0.010839
1.47E-05
1.47E-05
1.47E-05

19

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Table 8 Estimated failure consequences.

Equipment No.

E01
E02
E03
E04
E05
E06

Consequences of asset damage and

human health loss in terms of
(million dollars)
344.98
307.29
256.40
88.34
284.18
377.76

4.3.
Module IIIrisk-based shutdown inspection and
maintenance interval
Module III of the proposed methodology consists of estimating system failure probability, all possible consequences due to
process unit shutdown and nally developing a risk prole to
nd the optimal shutdown inspection and maintenance intervals. Since system failure probability is the function of time,
it is modeled here using Weibull distribution. To calculate the
system failure probability, the Weibull distribution parameters of the equipment are used as discussed in Module II, thus
the system failure probability function can be written using
Eq. (8). Considered gas chilling and liquefaction unit is used
to process hydrocarbons and operates at higher pressure and
cryogenic temperature conditions, the consequences of failure will be very high. Failure scenario considered to estimate
the asset damage and human health loss for these equipment
is the release of hydrocarbon which upon nding an ignition
source may result in explosion which could generate shock
waves. Calculated shock waves are then transformed in to
probability of damage due to shock wave for various effected
zones. This shock wave may cause serious asset damage as
well as human injury or fatality. In this particular scenario,
estimated consequence is based on a radius of 200 m and is
listed in Table 8. It is important to note that the loss of life
or injury suffered by the people is hard to estimate in dollar value; however, the cost associated with the compensation
and corporate liability needs to be taken into account in consequence analysis. In this regard, published note from Judycki
(1994) of the US Department of Transportation and the Federal
Highway Administration published relating the injury scale
(in terms of severity) to the comprehensive costs in policereported crashes is used. In this paper, the production loss
estimated using Eq. (13) for 10 days of shutdown is considered.
Plant-specic data and eld engineer supports to estimate
number of manhour, labor and equipment cost is considered
to estimate economic consequence of each shutdown inspection and maintenance interval.
The total calculated consequence along with Eq. (20) is used
to generate the expected risk prole to determine the total
number of shutdown intervals. Fig. 9 show the obtained risk
prole to achieve optimal shutdown inspection and maintenance interval, which would enable a level of 95.0% system
reliability for a desired goal time while meeting the risk
exposure to the lowest level. The shutdown inspection and
maintenance interval is obtained from Fig. 9. A sensitivity
analysis is performed to study the effect of the number of
shutdown days on the shutdown interval model. The results
are plotted in Fig. 10. It is evident from Fig. 10 that operational risk prole is increasing or decreasing with the increase
or decrease of shutdown duration, due to the impact of production losses; however, the number of shutdown required

Fig. 9 Risk-based shutdown inspection and maintenance.

Fig. 10 Risk-based shutdown inspection and

maintenance.

Table 9 Estimated shutdown cumulative time.

Shutdown number
1
2
3
4
5
6

Cumulative time (h)

43,222
72,014
96,487
1,17,289
1,34,971
1,50,000

in achieving desired system reliability over the goal time does

not have signicant impact. Table 9 shows the estimated shutdown cumulative time. It is evident from Fig. 9 that the overall
operational risk is low when a total of six shutdowns with
unequal intervals are performed under the given conditions
to achieve minimum operational risk ($) per hour. The presented methodology does not only provide optimal number
of the facility shutdown interval but also suggest when these
shutdowns should be planned to achieve target reliability, to
achieve target availability and to minimize the overall risk for
the considered life of the facility.

5.

Conclusion

Shutdown inspection and maintenance activities are performed to increase the availability and reliability of facility by
selecting the critical equipment to optimize the overall risk
prole. Risk assessment integrates the system failure probability and consequences. A risk-based shutdown inspection
and maintenance helps to select the critical equipment and
systems which cannot be inspected or maintained without

20

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

taking the plant out of operation. This paper presents a

risk-based shutdown inspection and maintenance intervals
optimization for a processing facility unit considering human
errors which may be introduced during these activities. The
unit considered in this study is a gas chilling and liquefaction
unit of LNG facility. In this study, SLIs were calculated for each
activity and converted into human error probability using
SLIM methodology. Estimated human error probability is then
integrated in the probability of system failure. Shutdown
inspection and maintenance consequences of selected equipment were determined and expressed in nancial values ($).
The cost of inspection and maintenance were calculated using
the plant data and/or using engineers experiences. System
probability of failure couples with estimated consequences
is used to generate operational risk prole. By including the
human error probability in system failure modeling, it is safe
to say that shutdown inspection and maintenance planning
will help to obtain a risk prole which will allow achieving
not only an optimal shutdown intervals but also the desired
system reliability over the goal time by minimizing the overall
risk. As the system operating life increases, the reliability
decreases, which require shorter duration of inspection and
maintenance to avoid excessive operational consequences.
Duration of shutdown and the number of activities depends
on the number of equipment selected which will go for
inspection and maintenance and the scope of shutdown. The
proposed methodology is an extension of Khan and Hameed
(2014) and Khan et al. (2014) to estimate shutdown interval,
where in the previous studies, the impact of human error was
not considered. Further, these methodologies was considered
to nd a risk-based shutdown interval considering the equal
interval, whereas the proposed methodology helps to nd
optimal shutdown inspection and maintenance considering
unequal interval over a goal time to minimize the overall
operational risk. Risk-based shutdown inspection and maintenance interval methodology proposed in this paper can be
applied and used for any processing or manufacturing facility.
By using the risk criticality matrix, it is possible to select the
critical equipment which should be considered for shutdown
to optimize the overall risk.

References
Alsyouf, I., 2007. The role of maintenance in improving
companies, productivity and protability. Int. J. Prod. Econ.
105 (1), 7078.
API, 2009. Risk Based Inspection; Recommended Practice 580,
Second ed. American Petroleum Institute, Washington, D.C.,
USA.
Backlund, F., Hanu, J., 2002. Can we make maintenance decision
on risk analysis? J. Qual. Maint. Eng. 8 (1), 7791.
Ben-Daya, M., Alghamdi, A.S., 2000. On an imperfect preventive
maintenance model. Int. J. Qual. Reliab. Manag. 17 (06),
661670.
Bertolini, M., Bevilacqua, M., Ciarapica, F.E., Giacchetta, G., 2009.
Development of risk-based inspection and maintenance
procedures for an oil renery. J. Loss Prev. Process Ind. 22,
244253.
Block, H.W., Borges, W.S., Savits, T.H., 1985. Age dependent
minimal repair. J. Appl. Probab. 22 (2), 370385.
Brown, M., Proschan, F., 1983. Imperfect repair. J. Appl. Probab. 20,
851859.
Dey, K.P., Ogunlana, S.O., Gupta, S.S., Tabucanon, M.T., 1998. A
risk-based maintenance model for cross-country pipelines.
Cost Eng. 40 (4), 2431.

Dey, P.M., 2001. A risk-based model for inspection and

maintenance of cross-country petroleum pipeline. J. Qual.
Maint. Eng. 7 (1), 2541.
Dhillon, B.S., Liu, Y., 2006. Human error in maintenance: a review.
J. Qual. Maint. Eng. 12 (1), 2136.
Dhillon, B.S., Yang, N., 1995. Probabilistic analysis of a
maintainable system with human error. J. Qual. Maint. Eng. 1
(2), 5059.
Dhillon, B.S., 1986. Human Reliability with Human Factor.
Pergamon Press, New York, USA.
Duarte, J.A.C., Craveiro, J.C.T., Trigo, T.P., 2006. Optimization of the
preventive maintenance plan of a series components system.
Int. J. Press. Vessel Pip. 83, 244248.
Duffua, S.O., Daya, M.A.B., 2004. Turnaround maintenance in
petrochemical industry, practice and suggested improvement.
J. Qual. Maint. Eng. 10 (3), 184190.
Fujiyama, K., Nagai, S., Akikuni, Y., Fujiwara, T., Furuya, K.,
Matsumoto, S., Takagi, K., Kawabata, T., 2004. Risk-based
inspection and maintenance system for steam turbine. Int. J.
Press. Vessel Pip. 81, 825835.
Ghosh, D., Roy, S., 2009. Maintenance optimization using
probabilistic cost-benet analysis. J. Loss Prev. Process Ind. 22,
403407.
Hameed, A., Khan, F., Ahmed, S., 2014. A risk-based methodology
to estimate shutdown interval considering system availability.
Proc. Safety Prog., http://dx.doi.org/10.1002/prs.11722.
Jacob, D., Amari, S., 2005. Analysis of complex repairable systems.
Inst. Electr. Electron. Eng. 54, 517529.
Judycki, D.C., 1994. Technical Advisory on Motor Vehicle Accident
Costs. US Department of Transportation, Federal Highway
Administration, USA,
http://safety.fhwa.dot.gov/facts stats/t75702.
Khan, F., Hameed, A., 2014. A framework to estimate the
risk-based shutdown interval for a processing plant. J. Loss
Prev. Process Ind. 32, 1829.
Khan, F.I., Haddara, M., 2004a. Risk-based maintenance (RBM): a
new approach for process plant inspection and maintenance.
Am. Inst. Chem. Eng. Process Saf. Prog. 23 (4), 252265.
Khan, F.I., Haddara, M., 2004b. Risk-based maintenance of
ethylene oxide production facilities. J. Hazard. Mater. A108,
147159.
Khan, F.I., Haddara, M., 2003. Risk-based maintenance (RBM); a
quantitative approach for maintenance/inspection scheduling
and planning. J. Loss Prev. Process Ind. 16, 561573.
Khan, F.I., Haddara, M., Krishnasamy, L., 2008. A new
methodology for risk-based availability analysis. IEEE Trans.
Reliab. 57 (2008), 103112.
Kirwan B. (1996); The validation of three Human Reliability
Quantication techniques THERP, HEART and JHEDI; Part 1
technique descriptions and validation issues.
Kirwan, B., 1997. The validation of three human reliability
quantication techniquesTHERP, HEART and JHEDI; part
IIIpractical aspects of the usage of the techniques;. Appl.
Ergon. 28 (1), 2739.
Kirwan, B., Kennedy, R., Taylor-Adams, S., Lambert, B., 1997. The
validation of three human reliability quantication
techniquesTHERP, HEART and JHEDI; part IIresults of
validation exercise. Appl. Ergon. 28 (1), 1725.
Krishnasamy, L., Khan, F.I., Haddara, M., 2005. Development of
risk-based maintenance (RBM) strategy for a power
generating plant. J. Loss Prev. Process Ind. 18, 6981.
Kumar, E.V., Chaturvedi, S.K., 2008. Prioritization of maintenance
tasks on industrial equipment for reliability. Int. J. Qual.
Reliab. Manag. 28 (1), 109126.
Lawrence, G., 2012. Cost estimating for turnarounds. Pet. Technol.
Q. Q1, 3743, www.eqtq.com.
Levitin, G., Lisnianski, A., 1999. Optimization of imperfect
preventive maintenance for multi-state system. Reliab. Eng.
Syst. Saf. 67 (2000), 193203.
Li, H., Shaked, M., 2003. Imperfect repair models with preventive
maintenance. J. Appl. Probab. 40, 10431059.
Malik, M.A.K., 1979. Reliable preventive maintenance policy. AIIE
Trans. 1 (l/3), 221228.

Process Safety and Environmental Protection 1 0 0 ( 2 0 1 6 ) 921

Mannan, M.S., Yang, X., 2010. The development and application of

dynamic operational risk assessment in oil/gas and chemical
process industry. Reliab. Eng. Syst. Saf. 95 (2010), 806815.
Nakagawa, T., Yasui, K., 1987. Optimum policies for a system with
imperfect maintenance. IEEE Trans. Reliab. R-36 (5), 631633.
Nakagawa, T., Qian, C., Zhao, X., 2012. Optimal imperfect
preventive maintenance policies for a used system. Int. J. Syst.
Sci. 43 (9), 16321641.
Neil, M., Marquez, D., 2012. Availability modelling of repairable
systems using Bayesian networks. Eng. Appl. Artif. Intell. 25,
698704.
Nguyen, D.G., Murthy, D.N.P., 1981. Optimal maintenance policy
with imperfect preventive maintenance. IEEE Trans. Reliab.
R-30 (5).
Noroozi, A., Khakzad, N., Khan, F., MacKinnon, S., Abbassi, R.,
2013. The role of human error in risk analysis: application to
pre and post-maintenance procedures of process facilities.
Reliab. Eng. Syst. Saf. 119, 251258.
Noroozi, A., Khan, F., MacKinnon, S., Amyotte, P., Deacon, T., 2014.
Determination of human error probabilities in maintenance
procedures of a pump. Process Saf. Environ. Prot. 92, 131141.
Obiajunwa, C., 2012. A framework for the evaluation of
turnaround maintenance projects. J. Qual. Maint. Eng. 18 (4),
368383.
Pil, C.K., Rausand, M., Vatn, J., 2008. Reliability assessment of
reliquefaction systems on LNG carriers. Reliab. Eng. Syst. Saf.
93, 13451353.
Rangan, A., Grace, R.E., 1989. Optimal replacement policies for a
deteriorating system with imperfect maintenance. Adv. Appl.
Probab. 21, 949951.
Rusin, A., Wojaczek, A., 2012. Optimization of power machines
maintenance intervals taking the risk into consideration.
Eksploat. i Niezawodn.Maint. Reliab. 14 (1), 7276.

21

Sarkar, A., Behra, D.K., 2012. Int. J. Adv. Res. Eng. Appl. Sci. 1 (2),
2038.
Snmez, E., Kekre, S., Scheller-Wolf, A., Secomandi, N., 2013.
Strategic analysis of technology and capacity investments in
the liqueed natural gas industry. Eur. J. Oper. Res. 226 (1),
100114.
Swain A.D., Guttmann H. E., (1983); Handbook of Human
Reliability Analysis with Emphasis on Nuclear Power Plant
Applications; NUREC/CR-1278, U.S. Nuclear Regulatory
Commission, August 1983, with Addendum #1 to
NUREC/CR-1278, August 1983, September 1, 1985, by A. D.
Swain.
Toriizuka, Takashi, 2001. Application of performance shaping
factor (PSF) for work improvement in industrial plant
maintenance tasks. Int. J. Ind. Ergon. 28 (2001), 225236.
Tan, J.S., Kramer, M.A., 1997. A general framework for preventive
maintenance optimization in chemical process operations.
Comput. Chem. Eng. 21 (12), 14511469.
Vatn, J., Hokstand, P., Bodsberg, L., 1996. An overall model for
maintenance optimization. Reliab. Eng. Syst. Saf. 51,
241257.
Vaurio, J.K., 1995. Optimization of test and maintenance intervals
based on risk and cost. Reliab. Eng. Syst. Saf. 49, 2336.
Wang, Y., Cheng, G., Hu, H., Wu, W., 2012. Development of a
risk-based maintenance strategy using FMEA for a continuous
catalytic reforming plant. J. Loss Prev. Process Ind. 25,
958965.
Wilson, L., McCutcheon, D., 2003. Industrial Safety and Risk
Management. The University of Alberta Press, Edmonton,
Alberta, Canada.
Zhaoyang, T., Jianfeng, L., Zongzhi, W., Jianhu, z., Weifeng, H.,
2011. An evaluation of maintenance strategy using risk based
inspection. Saf. Sci. 49, 852860.