http://hfs.sagepub.com/ Ergonomics Society of the Human Factors and Human Factors: The Journal http://hfs.sagepub.com/content/early/2013/11/26/0018720813512328 The online version of this article can be found at: DOI: 10.1177/0018720813512328 November 2013 published online 26 Human Factors: The Journal of the Human Factors and Ergonomics Society Alireza Noroozi, Rouzbeh Abbassi, Scott MacKinnon, Faisal Khan and Nima Khakzad Gas Facilities Effects of Cold Environments on Human Reliability Assessment in Offshore Oil and Published by: http://www.sagepublications.com On behalf of: Human Factors and Ergonomics Society can be found at: Society Human Factors: The Journal of the Human Factors and Ergonomics Additional services and information for http://hfs.sagepub.com/cgi/alerts Email Alerts: http://hfs.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: What is This? - Nov 26, 2013 OnlineFirst Version of Record >> at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014 hfs.sagepub.com Downloaded from at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014 hfs.sagepub.com Downloaded from
Transcript
http://hfs.sagepub.com/Ergonomics Society
of the Human Factors and Human Factors: The Journal
http://hfs.sagepub.com/content/early/2013/11/26/0018720813512328The online version of this article can be found at:
DOI: 10.1177/0018720813512328
November 2013 published online 26Human Factors: The Journal of the Human Factors and Ergonomics Society
Alireza Noroozi, Rouzbeh Abbassi, Scott MacKinnon, Faisal Khan and Nima KhakzadGas Facilities
Effects of Cold Environments on Human Reliability Assessment in Offshore Oil and
Published by:
http://www.sagepublications.com
On behalf of:
Human Factors and Ergonomics Society
can be found at:SocietyHuman Factors: The Journal of the Human Factors and ErgonomicsAdditional services and information for
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
Objective: This paper proposes a new methodology that focuses on the effects of cold and harsh environments on the reliability of human performance.
Background: As maritime operations move into Arctic and Antarctic environments, decision makers must be able to recognize how cold weather affects human performance and subsequently adjusts management and operational tools and strategies.
Method: In the present work, a revised version of the Human Error Assessment and Reduction Technique (HEART) methodology has been developed to assess the effects of cold on the likelihood of human error in offshore oil and gas facili-ties. This methodology has been applied to post-maintenance tasks of offshore oil and gas facility pumps to investigate how management, operational, and equipment issues must be con-sidered in risk analysis and prediction of human error in cold environments.
Results: This paper provides a proof of concept indicating that the risk associated with operations in cold environments is greater than the risk associated with the same operations performed in temperate climates. It also develops guidelines regarding how this risk can be assessed. The results illustrate that in post-maintenance procedures of a pump, the risk value related to the effect of cold and harsh environments on oper-ator cognitive performance is twice as high as the risk value when performed in normal conditions.
Conclusion: The present work demonstrates significant differences between human error probabilities (HEPs) and associated risks in normal conditions as opposed to cold and harsh environments. This study also highlights that the cogni-tive performance of the human operator is the most impor-tant factor affected by the cold and harsh conditions.
Application: The methodology developed in this paper can be used for reevaluating the HEPs for particular scenarios that occur in harsh environments since these HEPs may not be comparable to similar scenarios in normal conditions.
Keywords: cold regions, human error, maintenance, risk analysis, offshore oil and gas industry
IntroductIonThe study of the human factor is an important
area of process engineering that includes the systematic application of information related to human characteristics and behavior to improve the performance of human-machine systems (McSweeney, de Koker, & Miller, 2008). According to Dhillon and Liu (2006), poor design factors in equipment, maintenance, and work layout; difficulties faced by workers such as improper work tools; and overstress-induced fatigue are the main factors that contribute to error occurrence in maintenance procedures. Other contributing factors include environmen-tal factors such as humidity, lighting, and tem-perature. Improper training, use of outdated maintenance manuals, and lack of proper expe-rience also cause a high number of maintenance errors.
On the other hand, there are a few factors that improve the work environment, such as person-nel training, ensuring emotional stability, hiring workers with a greater aptitude for their environ-ment, improving team work, and boosting morale.
Nelson (1996) argued that accident occurrence due to maintenance and inspection activities should be taken into consideration in the industry. Balkey (1996) also asserted that risk-based inspection procedures and human error proce-dures must be considered in inspection proce-dures. Further data were provided by Eves (1985) on accidents in the chemical manufacturing industry during maintenance. Since collecting samples of different types of human errors and interactions can be helpful in preventing such errors in future, exhaustive work has been carried out to investigate the role of human error in main-tenance, inspection, and system performance (Carr & Christer, 2003; Dhillon & Yang, 1993; Gramopadhye & Drury, 2000; Ramalhoto, 1999).
512328 HFSXXX10.1177/0018720813512328Human FactorsEffects of Cold Environments2013
Address correspondence to Faisal Khan, Professor, Safety and Risk Engineering Group, Faculty of Engineering and Applied Science, Memorial University of Newfoundland, St. John’s, NL A1B3X5, Canada; e-mail: [email protected].
Effects of Cold Environments on Human Reliability Assessment in Offshore Oil and Gas Facilities
Alireza Noroozi, Memorial University of Newfoundland, St. John’s, Canada, Rouzbeh Abbassi, Princeton University, Princeton, New Jersey, Scott MacKinnon, Faisal Khan, and Nima Khakzad, Memorial University of Newfoundland, St. John’s, Canada
Several methodologies have been developed to estimate human error in maintenance proce-dures. Also, numerous challenges related to the operation of equipment, systems, structure, and safety equipment performed under cold and harsh environments have previously been pointed out by researchers (e.g., Parsons, 2003; Strauch, 2004). However, there is a lack of methodology to quantify the human error prob-abilities (HEPs) of maintenance activities in Arctic conditions. Unique characteristics of Arc-tic regions and their effect on human perfor-mance during maintenance procedures demand for a methodology to account for the effect of cold and harsh environments in the final estima-tion of HEPs.
Some effects of cold temperature and harsh environments on human performance are listed in Table 1.
When core body temperatures begin to fall below the normal resting values, hypothermia starts (Makinen, 2006). Body metabolism increases to produce more body heat, and as cooling contin-ues, a person will start to shiver; this is a visible sign that body cooling has continued beyond a
comfortable level. By increasing metabolic rates, the amount of time a person can sustain will be reduced (Legland, Conachey, Wang, & Baker, 2006). Motor control becomes impaired as a body cools, making an operator vulnerable to physical injuries. Extremely cold conditions adversely affect mental skills and cognition (Bourne & Yaroush, 2003). As operational tem-peratures decrease, the frequency of cognitive error increases. Operations at cold temperatures coupled with physical distracters such as noise or moving environments will affect the quality of perception, memory, and reasoning, further increasing the risk of error in decision making (Legland et al., 2006). Specific effects of extreme environments on human performance are high-lighted in Karwowski (2001) and Hoffman (2002) and must be considered when assessing task per-formance, operating procedures, and equipment design.
Physical performance decrements resulting from exposure to cold weather can have profound effects on how a task is completed. Direct deficits include loss of strength, mobility, and balance. Whereas thermal protective clothing may mitigate
TAblE 1: General Cold Environmental Factors Affecting Human Performance (HP)
Stressors Details
Cold temperature Breathing difficultyMuscular stiffnessFrostbiteLowered metabolismHypothermiaBulky clothingStiffness of suits impairing movement
the neurophysiologic responses, it can indirectly affect manual performance due to a decrease in mobility and inability to perceive external ele-ments or cues. Investigations have reported mini-mal decreases in simple reaction time except in the most extreme conditions (Enander, 1987; Hoff-man, 2002). However, for more complex tasks, cold environments have resulted in substantially poorer performances. It is reported that reaction times were increased among subjects exposed to an ambient temperature of −26°C with a wind speed of 10 mph or greater (Hoffman, 2002). Out-comes also showed an increase in the number of errors, incorrect responses, and false alarms and a decrease in the ability to inhibit incorrect responses. When a person fully dressed in Arctic clothing is exposed to extremely cold air tempera-tures, a significant reduction in performance is observed when compared with that in normal tem-peratures (Parsons, 2003). Extreme cold stress may produce confusion and impaired conscious-ness. Researchers demonstrated an increase in the number of errors when performing at the tempera-ture of 5°C, compared with the performance in 22°C ambient temperature (Hoffman, 2002; Orden & Benoit, 1996; Pilcher, Nadler, & Busch, 2002; Wright, Hull, & Czeisler, 2002). One of the major consequences of working in cold and harsh environments includes fatigue, both physical and cognitive. Fatigue continues to be either a main cause or a contributory factor to casualties and damage to the environment and property. Fatigue impacts the individual’s skills to react and recog-nize and interpret stimuli in the work environ-ment. Fatigue also encourages the apathy status and decreases motivation at work, consequently contributing toward poor performance (Xhelilaj & Lapa, 2010).
This paper develops a methodology that takes into consideration the effects of cold operating conditions on various features of human perfor-mance by particularizing the Human Error Assessment and Reduction Technique (HEART) methodology for cold environments. The pro-posed methodology will help the assessors in investigating the probabilities of human error more accurately in cold conditions, the under-standing of which will help in improving the overall reliability of offshore oil and gas facili-ties in cold and harsh environments.
A developed Methodology ApplIed In cold envIronMentsHEART is a technique widely used in human
reliability assessment to compare HEPs, based on the degree of error recovery. In a standard HEART methodology, the specification of a par-ticular scenario based on the present conditions of a facility (or a part of the facility) is required. Thus, observing the specific conditions such as cold temperature, high speed wind, lack of vis-ibility, and slippery conditions is required for describing an accurate scenario to be applicable in cold and harsh environments. Considering the aforementioned factors is the cornerstone of the methodology developed in this work, distin-guishing it from the standard HEART methodol-ogy. HEART methodology has previously been used to estimate the HEPs in normal operating conditions (Casamirra, Castiglia, Giardina, & Tomarchio, 2009; Kirwan & Gibson, 2007; Noroozi et al., 2012). All subtasks that are to be completed by the operator within each task in a considered scenario should be investigated to properly utilize this methodology in any par-ticular scenario. Subsequently, a nominal human unreliability score is determined for the particu-lar task (Kirwan, Kennedy, Taylor-Adams, & Lambert, 1996).
In the standard HEART, an analyst usually uses values ranging from 5th to 95th percentile boundaries of nominal human unreliability for a particular task (typically the mean values). Because of the harsh and cold conditions, the values of the 95th percentile in the modified methodology are used, which is considered as a worst-case scenario.
By identifying the particular scenario, the factors that influence the HEP, known as error producing conditions (EPCs), will be deter-mined.
According to the scenario and for illustration purposes, only three to four EPCs of higher nominal amounts are selected to estimate the final HEPs. In the developed methodology, the EPCs are divided into four different categories: physical, cognitive, instrumentation, and man-agement (Table 2). These four major categories have been derived based on previous work on the effect of cold and harsh conditions on pro-ducing errors in human performances (Bourne
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
TAblE 2: Error-Producing Conditions (EPCs) in HEART Methodologya
Error-Producing Condition
Maximum Predicted Nominal Amount by
Which Unreliability Might Change Going From
“Good” Conditions to “Bad”
1 Unfamiliarity with a situation that is potentially important but that only occurs infrequently or is novel
17
2 A shortage of time available for error detection and correction (P) 11 3 A low signal-to-noise ratio (C) 10 4 A means of suppressing or overriding information or features that is
too easily accessible9
5 No means of conveying spatial and functional information to operators in a form that they can readily assimilate
8
6 A mismatch between an operator’s model of the world and that imagined by the designer (C, M)
8
7 No obvious means of reversing an unintended action 8 8 A channel capacity overload, particularly one caused by simultaneous
presentation of nonredundant information6
9 A need to unlearn a technique and apply one that requires the application of an opposing philosophy
6
10 The need to transfer specific knowledge from task to task without loss (C)
5.5
11 Ambiguity in the required performance standards 512 A mismatch between perceived and real risk 413 Poor, ambiguous, or ill-matched system feedback (C, I) 414 No clear direct and timely confirmation of an intended action from the
portion of the system over which control is to be exerted3
15 Operator inexperienced (e.g., a newly qualified tradesman, but not an “expert”)
3
16 An impoverished quality of information conveyed by procedures and person-person interaction
3
17 Little or no independent checking or testing of output (P, I, M) 318 A conflict between immediate and long-term objectives 2.519 No diversity of information input for veracity checks 2.520 A mismatch between the educational achievements level of an
individual and the requirements of the task2
21 An incentive to use other more dangerous procedures (P, C) 222 Little opportunity to exercise mind and body outside the immediate
confines of the job1.8
23 Unreliable instrumentation (I, M) 1.624 A need for absolute judgments that are beyond the capabilities or
experience of an operator (C)1.6
25 Unclear allocation of function and responsibility 1.626 No obvious way to keep track of progress during an activity 1.427 A danger that finite physical capabilities will be exceeded (P) 1.4
(continued)
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
& Yaroush, 2003; Forsius et al., 1970; Hoffman, 2002; Mekjavic, Banister, & Morrison, 1988; Orden & Benoit, 1996; Staal, 2004). The EPCs related to each category in the modified method-ology are added to the main EPCs, which are similar to the normal conditions and then used in the final estimation of HEPs.
Each EPC has a maximum nominal amount, which should be inserted in Equation 1 as the maximum effect. The next step is to assess the proportion of affect (APOA), which is weighted for each chosen EPC based on its importance. Accordingly, each EPC is individually weighted from 0 to 1 (Williams, 1988).
Equation 1 can be applied to calculate the effect of each EPC and its relevant APOA on the
HEP. The HEP of each task is calculated by mul-tiplying the selected HEP with the nominal amount of APOA related to each EPC (Williams, 1988). Figure 1 illustrates the modified HEART methodology applicable to cold and harsh environments.
ApplIcAtIon of the developed Methodology
The developed methodology was applied to postmaintenance procedures of a condensate pump in an offshore oil and gas facility to demonstrate the variation in HEPs in both cold and normal environments. This is a particular type of pump applied to the condensate water produced in an HVAC (heating or cooling), condensing boiler furnace, or steam system. The regular maintenance activities of a pump in an offshore oil and gas facility can be divided using three different categories: premaintenance,
Error-Producing Condition
Maximum Predicted Nominal Amount by
Which Unreliability Might Change Going From
“Good” Conditions to “Bad”
28 Little or no intrinsic meaning in a task 1.429 High-level emotional stress 1.330 Evidence of ill health among operatives, especially fever (P) 1.231 Low workforce morale (C, M) 1.232 Inconsistency of meaning of displays and procedures 1.233 A poor or hostile environment (below 75% of health or life-
threatening severity) (P)1.15
34 Prolonged inactivity or highly repetitious cycling of low mental workload tasks
×1.1 for first half hour×1.05 for each hour there
after35 Disruption of normal work-sleep cycles (C, M) 1.136 Task pacing caused by the intervention of others 1.0637 Additional team members over and above those necessary to perform
task normally and satisfactorily× 1.03 per additional man
38 Age of personnel performing perceptual tasks 1.02
Note. P = physical; C = cognitive; I = instrumentation; M = management.aThe following variables were assessed for impact due to operations in cold environments. These variables were considered to influence operator physical or cognitive performance and/or affect management decision making.
TAblE 2: (continued)
(1)
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
maintenance, and postmaintenance. The focus of this paper is, however, on the post-main-tenance activities. These activities have been developed in conjunction with the Single Buoy Moorings (SBM) Company, Nova Scotia, Can-ada. The scenarios were developed based on maintenance reports provided by the offshore platform and were selected based on the most frequent occurring scenarios.
selected scenario for postmaintenance Activities
After each maintenance service, the opera-tors and engineers must continue with post-maintenance activities, since production must not be halted for extended periods of time. These activities focus on returning the system to normal operation. The following information provides characteristics of the selected scenario in this work:
1. Some junior operators have logged insufficient training hours. Because of the high amount of work undertaken, the pressure is high, leading to intense fatigue for the workers.
2. An inexperienced workforce engineer is respon-sible for ensuring site readiness for reinstatement. The site engineer is using a poorly written report to perform an inspection and to ensure that the site and equipment are in safe conditions.
3. The responsible incoming assistant lacks ade-quate information regarding returning keys and supplying certificates.
4. The system feedback is unreliable.5. The supervisor is too busy to provide complete
supervision for the procedure.6. There is insufficient time, due to the urgency of
starting operations to prevent extra costs.
Generally, there are time constraints related to further extending the shutdown activities.
The aforementioned scenario is considered in the calculation of the HEPs in normal conditions. A similar scenario is applied to calculate the HEPs in cold and harsh environments. The par-ticular specifications of these regions are listed in Table 1. postmaintenance work requires sequen-tially executed activities. Factors such as opera-tor experience, time constraints, administrative procedures, and high work demands can lead to task error. The postmaintenance scenarios for the
Identifying the full range of subtasks in apart of offshore facility
Considering the particular scenario includingthe specific conditions of cold and harshenvironments
Determining a nominal humanunreliability score
Identifying Error Producing Conditions(EPC) considering the cold conditions
Assessing proportion of effect of eachEPC on HEP
Calculating the final value of the HEPs for the subtasks occurred cold conditions
Applying the 95th nominal amount to beconsidered the adverse effect of harsh andcold environments
Dividing the EPCs to four different categoriesand considered each category in the finalcalculation of HEPs
Estimating the effect of the each EPCbetween 0 to 1 on final value HEP based onthe particular conditions in specified scenario
Evaluating the final HEP values to increasethe overall system's reliability at coldenvironments
Figure 1. A modified methodology developed to calculate the human error probabilities (HEPs).
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
considered procedures of a condensate pump are indicated in Tables 3 and 4.
hep calculationThe developed methodology is applied to
calculate the HEPs for all the aforementioned activities for a considered scenario. As an exam-ple, a detailed calculation of Subactivity 1.1, “Check lines and equipment for obstructions,” is presented in Table 4. This calculation refers to the effect of cold on the physical performances of employees during maintenance procedures of a pump.
The first step is to determine a generic task (GT). The upper-bond value of GT considered for Subactivity 1.1 is “E.” For the type “E” task, the nominal unreliability is 0.045. Using Table 2, the EPCs and their maximum predicted nomi-nal amounts related to this subactivity are selected based on the scenario illustrated previ-ously. Considering Table 2, the EPCs related to the effect of cold and harsh environments on the major human performances (cognitive and phys-ical) as well as management and instrumentation are adopted for each category and added to the
related EPCs. A proportionate weight factor is applied when an EPC is considered. This is dem-onstrated in the column labeled “Assess Propor-tion of Effect” in Table 4 for Subactivity 1.1. As illustrated in Table 4, the values of 0.01 for the EPCs of 2 and 11 and 0.05 for the other EPCs are selected based on the degree of effectiveness of each EPC on human error.
Based on Equation 1, the assessed effect of each EPC is calculated. Finally, the HEP is cal-culated based on the effect of cold and harsh environments on physical performances for Subactivity 1.1 as 6.17E-02. The similar proce-dure is performed to estimate the HEPs of differ-ent subactivities in normal condition (Table 5) and also by considering the effect of cold and harsh environments on human performances (cognitive and physical), decision making (man-agement), and instrumentation (Table 6) used in post-maintenance activities of a pump.
HEP is normally calculated using an empiri-cal formula, whereas more sophistication is required. Statistical approaches such as Markov models have been proposed for human error quantification (Sridharan & Mohanavadivu, 1997). However, as human behavior and actions
TAblE 3: Activities Required During Postmaintenance
Subactivity Activity
1. Reconnect pump 1.1 Check lines and equipment for obstructions
1.2 Remove mechanical isolation/connect lines to pump
1.3 Remove locks and tags from valves, leaving valves closed
2. Workforce supervisor (WFS) ensure site and equipment left in safe state
3. WFS return keys and certificates 4. Permit to work coordinator (PTWC) ensure site
is ready for reinstatement4.1 Return lock-out keys4.2 Give worksite authority back to Area
are highly variable and unpredictable, the use of an empirical formula is preferred to a statistical technique.
hep comparisonStatistical comparison. To examine the influ-
ence of cold and harsh environments on HEPs, the normalized relative differences in HEP val-ues calculated in cold and normal environments are presented in Table 7, and appropriate statisti-cal analyses are applied. To this end, the normal-ity of the HEPs in each column is checked. As the HEPs in each column are not distributed nor-mally, the Wilcoxon signed-rank test is applied to define significant differences between each column of the HEPs resulting from harsh and cold environments and those resulting from nor-mal conditions.
Wilcoxon signed-rank test is a nonparametric statistical hypothesis test used to compare two related groups of data to assess whether their
population mean ranks differ (Vaughan, 2001). To define the difference between the columns of HEPs that have been obtained by considering the effect of cold on human performance (cogni-tive and physical), instrumentation, and man-agement, the Friedman test (considering post hoc tests) is used. The Friedman test compares the mean ranks between the related groups and indicates how the groups differ, although not demonstrating exactly where those differences lie. Thus, to examine where the differences occur, the Wilcoxon signed-rank test is used to combine the considered groups (cognitive with physical, cognitive with management, etc.)
The results obtained by applying the Wil-coxon signed-rank test to the Z-scores and p-values (Devore, 2008) demonstrate that there are statistically significant differences between the HEPs received from the effect of cold on physical performance and normal condition (Z = −3.180, p = .001), cognitive performances
TAblE 4: The Human Error Probabilities (HEPs) Calculation for Subactivity 1.1 Due to the Effect of Cold and Harsh Environments on Physical Performances
Activity: 1.0 Reconnect Pump
Subactivity 1.1
Check lines and equipment for obstructions (Arctic conditions: effect on physical performances; EPCs)
Generic Task
Generic Error Probability Number EPCs
Total HEART Effect
Assess Propor-tion of Effect
Assessed Effect
E 0.045 2 Time shortage 11 0.01 1.1 11 Ambiguity in standards 5 0.01 1.04 17 Little or no independent
checking3 0.05 1.1
21 An incentive to use other more dangerous procedures
2 0.05 1.05
27 A danger that finite physical capabilities will be exceeded
1.4 0.05 1.02
30 Evidence of ill-health among operatives
1.2 0.05 1.01
33 A poor or hostile environment
1.15 0.05 1.0075
Total assessed EPC effect 1.371433HEP 6.17E-02
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
and normal conditions (Z = −3.181, p = .001), management and normal conditions (Z = −2.691, p = .007), and instrumentation with normal con-ditions (Z = −3.181, p = .007). These results highlight the necessity of reevaluating the human errors due to the effect of cold and harsh condi-tions, increasing the overall reliability of mainte-nance procedures in Arctic and sub-Arctic regions. The results obtained from the Friedman test show that there is a statistically significant difference in HEPs depending on which type of effects are considered due to the cold and harsh conditions, χ2(2) = 32.908, p = .00. Post hoc anal-ysis with Wilcoxon signed-rank test is conducted while a Bonferroni correction is applied, result-ing in a significance level set at p < .0125. Median (IQR) perceived effort levels for the physical, cognitive, instrumentation, and management cat-egories include 7.32E-02 (3.81E-02 to 1.79E-01), 1.54E-01 (7.41E-02 to 4.56E-01), 7.48E-02 (3.98E-02 and 1.95E-01), and 8.79E-02 (4.66E-02 and 2.46E-01), respectively. There are no sig-nificant differences between HEPs received from the effect of cold on physical performance and management (Z = −2.272, p = .023), physical performance and instrumentation (Z = −2.064,
p = .039), and instrumentation and management (Z = −2.273, p = .023). However, there are statis-tically significant differences between the HEPs received from the effect of cold on physical and cognitive performances (Z = −3.180, p = .001), cognitive performances and management (Z = −3.182, p = .001), and cognitive performances and instrumentation (Z = −3.185, p = .001).
Risk-based comparison. Quantitative risk analysis (QRA) has played an important role in identifying major risks and maintaining safety in process facilities. QRA includes several steps such as hazard identification, accident model-ing, consequence analysis, and risk estimation. The results of QRA can either be used in assist-ing decision makers with risk levels of different plans or to improve the safety measures of facili-ties. Event tree is a technique widely used in QRA to explore and calculate the probabilities of the consequences of an undesired event (Khakzad, Khan, & Amyotte, 2012, 2013). In this study, to investigate the effect of cold and harsh environments on HEPs, a risk assessment is conducted to compare the values of risks resulting from human-error-induced accidents both in normal and cold environments. It should
TAblE 5: Human Error Probabilities (HEP) in Normal Conditions
Activities HEP
1.0 reconnect pump 1.1 Check lines and equipment for obstructions 8.01E-03 1.2 Remove mechanical isolation/connect lines to pump 1.45E-01 1.3 Remove locks and tags from valves, leaving valves closed 8.93E-032.0 Workforce supervisor (Wfs) ensure site and equipment
left in safe state8.98E-04
3.0 Wfs return keys and certificates 6.55E-024.0 permit to work coordinator (ptWc) ensure site ready for reinstatement 4.1 Return lock-out keys 7.06E-02 4.2 Give worksite authority back to Area Authority (AA) 8.90E-04 4.3 (Supervisors) Document reinstatement 1.57E-015.0 ptWc and AA finalize ptW 6.73E-026.0 open valves and reinstate pump 6.1 Test pressure 8.01E-03 6.2 Remove air from lines and pump 9.16E-04 6.3 Open valves, fill pump, and test for leaks 7.93E-03 6.4 Start pump 6.30E-02
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
be noted that for cold conditions, separate risk analyses are performed for the categories physi-cal, cognitive, instrumentation, and manage-ment, which result in four different values, respectively.
Considering the pump postmaintenance as the initiating event, the event tree in Figure 2 is developed. Based on field studies and expert opinions, the most probable accident scenario following a human error in the postmaintenance
TAblE 6: Human Error Probabilities (HEP) in Cold Conditions
procedure of the pump is determined as a release of flammable liquid. Meeting an ignition source, a pool fire would occur, which can be extin-guished only if a water sprinkler system is acti-vated by a flame/smoke detector (Figure 2).
The probabilities of the components of the event tree are indicated in Table 8 (Khakzad et al., 2012, 2013). However, it is worth noting that the probability of the top event “human error,” HEP, in Figure 2 for normal and cold conditions is derived using Tables 5 and 6, respectively, assuming that the activities and subactivities are independent and act like a
series system (the worst-case scenario). Thus, HEP can be calculated using Equation 2:
HEP Pii
n
= − −=∏1 11
( )
where Pi is the probability of each activity (or subactivity).
Having the probabilities of the event tree’s top events, the probabilities of the consequences can be calculated as listed in the last five columns of Table 9. Using the consequence severity matrix (Appendix A), the severity of each consequence
Figure 2. Event tree of pump postmaintenance.
TAblE 8: Probabilities of Event Tree’s Components
Top Event Description Probability
F Frequency of pump maintenance 0.3Human error
probabilityHuman error probability for normal
conditions and cold conditions including physical, cognitive, instrumentation, and management
Normal: 0.472Cold: 0.913, 0.997, 0.926,
and 0.982
X2 Occurrence probability of release given a human error
0.1
X3 Occurrence probability of ignition given a release
0.1
X4 Failure probability of flame detector given a fire
0.01
X5 Failure probability of water sprinkler given the operation of flame detector
0.04
Source. Khakzad, Khan, and Amyotte (2012, 2013).
(2)
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
is determined (Column 3 of Table 9) based on the extent of its adverse effects such as causali-ties and environmental and property damages. Assigning corresponding monetary values to each consequence (Column 4 of Table 9), the total amounts of the envisaged risks of the acci-dent scenario for normal and cold conditions are estimated (last row of Table 9).
It is worth noting that although C5 and C6 are of a similar severity, their probabilities are dif-ferent. According to the event tree in Figure 2, C5 results if the flame detector works, trying to activate the water sprinkler. However, the water sprinkler most likely may not work due to its failure modes. On the other hand, C6 results if the water sprinkler does not work due to a failure of the flame detector. Thus, in either C5 or C6, a major accident can occur due to unsuccessful fire extinguishment.
These results confirm that the cold and harsh conditions may have significant effects on pro-ducing human errors due to the effects on peo-ple’s cognitive performance (Enander, 1987; Makinen et al., 2006; Pilcher et al., 2002).
conclusIonInvestigation of the attributes of people in
cold regions requires accurate calculation of the probability of error in human activities. A new methodology is developed in this paper to estimate the HEPs in Arctic environments. In the new methodology, the upper-bond values of human unreliability can be applied for the extreme environmental conditions. Also, the existence of specific EPCs related to Arctic conditions, such as high-level emotional stress and a poor hostile environment, may add more value to the methodology to calculate the HEPs.
TAblE 9: Risk Analysis of Pump Postmaintenance Accident in Normal and Cold Conditions
Application of the developed methodology to postmaintenance procedures of a pump demon-strated that the HEPs in arctic conditions are in the higher ranges as opposed to the normal con-ditions. Statistical analysis indicated that there exist significant differences between the HEPs in cold and harsh conditions and normal condi-tions. This is more evident for tasks for which cold temperatures, wind, ice, and visibility are able to decrease human performance. Further, statistical analysis showed the effect of cold on cognitive attributes such as attention, decision
making, diagnosis, memory, and problem solv-ing. This study confirmed that reevaluation of the HEPs is required for any scenario that occurs in harsh environments since the HEPs calculated in normal conditions are not compatible with similar scenarios in harsh and cold conditions. Comparing the risk of the normal and cold con-ditions including physical, cognitive, instrumen-tation, and management categories, the cogni-tive category is shown to have the highest risk values. Cognitive impairment can increase the HEPs and subsequently the risk.
AppendIx A
TAblE A1: Consequence Severity Matrix
Severity Class
Dollar Value Equivalent Asset Loss Human Loss
Environmental Loss
Confidence or Reputation Loss
1 0 No significant asset loss
Minor mishap, no injury
No remediation required
Get noticed by operating unit only
2 0.01K–10K Short-term production interruption
Minor injury, first aid attention required
Around the operating unit, easy recovery and remediation
Get noticed in the operation line/line supervisor
3 10K–500K Equipment damage of one unit requiring repair/medium-term production interruption
One injury requiring hospital attention however no threat to life
Around the operating line, easy recovery and remediation
Get noticed in plant
4 500K–5M Equipment damage of more than one unit requiring repair/long-term production interruption
More than one injury requiring hospital attention however no threat to life
Within plant, short-term remediation effort
Get attention in the industrial complex; information shared with neighboring units
5 5M–50M Loss of one operating unit/product
Multiple major injuries, potential disabilities, potential threat to life
Minor offsite impact, remediation cost will be less than $1 million
Local media coverage
6 50M–500M Loss of major portion of equipment/ product
One fatality and/or multiple injuries with disabilities
Community advisory issued, remediation cost remain below $5 million
Regional media coverage, a brief note on national media
7 >500M Loss of all equipment/products
Multiple fatalities Community evacuation for longer period, remediation cost in excess of $5 million
National media coverage, brief note on international media
Source. Kalantarnia (2009). at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
Key poInts • A new methodology is developed to investigate
the effects of cold and harsh environments on the reliability of human performance.
• Results confirmed that reevaluating the human error probabilities (HEPs) is required for any sce-nario that occurs in harsh environments.
• Results of this study illustrate that the cold and harsh conditions may have significant effects on producing human errors due to its effect on the cognitive performances of humans.
referencesBalkey, J. P. (1996). Human factors engineering in risk-based
inspections. Safety Engineering and Risk Analysis, 6, 97–106.Bercha, F. G., Cerovsek, M., & Abel, W. (2003). Human perfor-
mance in arctic offshore escape, evacuation and rescue. Paper presented at the Thirteenth International Offshore and Polar Engineering Conference, Honolulu, HI.
Bercha, F. G., Cerovsek, M., & Abel, W. (2004). Reliability assess-ment of arctic EER systems. Paper presented at the 17th Inter-national Symposium on Ice, St. Petersburg, Russia.
Bourne, L. E., & Yaroush, R. A. (2003). Stress and cognition: A cognitive physiological perspective. Boulder, CO: University of Colorado.
Carr, M. J., & Christer, A. H. (2003). Incorporating the potential for human error in maintenance models. Journal of the Opera-tional Research Society, 53, 1249–1253.
Casamirra, M., Castiglia, F., Giardina, M., & Tomarchio, E. (2009). Fuzzy modeling of HEART methodology: Application in safety analyses of accidental exposure in irradiation plants. Radiation Effects & Defects in Solids, 164, 291–296.
Devore, J. L. (2008). Probability and statistics for engineering and the sciences. Belmont, CA: Thomson Higher Education.
Dhillon, B. S., & Liu, B. S. (2006). Human error in maintenance: A review. Journal of Quality in Maintenance Engineering, 12, 21–36.
Dhillon, B. S., & Yang, N. F. (1993). Human error analysis of a standby redundant system with arbitrarily distributed repair times. Microelectronics and Reliability, 33, 431–444.
Enander, A. (1987). Effects of moderate cold on performance of psychomotor and cognitive tasks. Ergonomics, 30, 1431–1445.
Eves, D. C. T. (1985). Deadly maintenance: A study of fatal acci-dents at work. London: HMSO.
Forsius, H., Eriksson, A. W., & Luukka, H. (1970). Ophthalmologi-cal characteristics of Eskimos in Augpilagtok. Arctic Anthro-pology, 8, 9–16.
Gramopadhye, A. K., & Drury, C. G. (2000). Human factors in aviation maintenance: How we get to where we are. Interna-tional Journal of Industrial Ergonomics, 26, 125–131.
Hoffman, R. (2002). Human psychological performance in cold envi-ronments. Medical aspects of harsh environment. Falls Church, VA: Office of the Surgeon General United States Army.
Kalantarnia, M. (2009). Dynamic risk assessment using accident precursor data and Bayesian theory (Unpublished master of engineering thesis). Memorial University, St. John’s, Canada.
Karwowski, W. (2001). International encyclopedia of ergonomics and human factors. London: Taylor and Francis.
Khakzad, N., Khan, F., & Amyotte, P. (2012). Dynamic risk analy-sis using bow-tie approach. Reliability Engineering and System Safety, 104, 36–44.
Khakzad, N., Khan, F., & Amyotte, P. (2013). Dynamic safety analysis of process systems by mapping bow-tie into Bayes-ian networks. Process Safety and Environmental Protection, 91, 46–53.
Kirwan, B., & Gibson, H. (2007). CARA: A human reliability assessment tool for air traffic safety management—Technical basis and preliminary architecture. In The safety of systems (Part 5, pp. 197–214). New York, NY: Springer.
Kirwan, B., Kennedy, R., Taylor-Adams, S., & Lambert, B. (1996). The validation of three human reliability quantification techniques—THERP, HEART and JHEDI: Part 1—Technique descriptions and validation issues. Applied Ergonomics, 27, 359–373.
Legland, E., Conachey, R., Wang, G., & Baker, C. (2006). Win-terization guidelines for LNG/CNG carriers in Arctic environ-ments. Houston, TX: American Bureau of Shipping.
Makinen, T. M. (2006). Human cold exposure, adaption and per-formance in a Northern climate, Oulu, Finland: Faculty of Medicine, Department of Physiology, Centre for Arctic Medi-cine, Thule Institute, University of Oulu.
Makinen, T. M., Palinkas, L. A., Reeves, D. L., Paakkonen, T., Rintamaki, H., Leppaluoto, J., & Hassi, J. (2006). Effects of repeated exposures to cold on cognitive performance in humans. Physiology and Behavior, 87, 166–176.
McSweeney, K. P., de Koker, T., & Miller, G. (2008). A human factors engineering implementation program (HFEIP) used on offshore installations. Naval Engineers Journal, 120(3), 37–49.
Mekjavic, I. B., Banister, E. W., & Morrison, J. B. (1988). Environ-mental ergonomics: Sustaining human performance in harsh environments. London: Taylor and Francis.
Nelson, W. E. (1996). Steam turbine over speed protection. Chemi-cal Processing, 59(7), 48–54.
Noroozi, A., Khan, F., Amyotte, P., & Deacon, T. (2012). Determi-nation of human error probabilities in maintenance procedures of a pump. Process Safety and Environmental Protection, 91, 333–422.
Orden, K. F. V., & Benoit, S. L. (1996). Effects of cold stress on the performance of a command and control task. Human Factors, 38, 130–141.
Parsons, K. (2003). The effects of hot, moderate, and cold environ-ments on human health, comfort and performance. London: Taylor & Francis.
Pilcher, J. J., Nadler, E., & Busch, C. (2002). Effects of hot and cold temperature exposure on performance: A meta-analytic review. Ergonomics, 45, 682–698.
Ramalhoto, M. F. (1999). Research and education in reliability, maintenance, quality control, risk and safety. European Jour-nal of Engineering Education, 4, 233–237.
Sridharan, V., & Mohanavadivu, P. (1997). Reliability and avail-ability analysis for two non-identical unit parallel systems with common cause failures and human errors. Microelectronics and Reliability, 37, 747–752.
Staal, M. A. (2004). Stress, cognition, and human performance: A literature review and conceptual framework. Moffett Field, CA: Ames Research Center.
Strauch, B. (2004). Investigating human error: Incidents, accidents, and complex systems. Hants, UK: Ashgate Publishing Limited.
Vaughan, L. (2001). Statistical methods for the information profes-sional: A practical, painless approach to understanding, using, and interpreting statistics. Medford, NJ: Information Today, Inc.
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
Williams, J. C. (1988). A data-based method for assessing and reducing human error to improve operational performance. Paper presented at the IEEE, 4th Conference on Human Fac-tors, Monterey, CA.
Wright, K. P., Hull, J. T., & Czeisler, C. A. (2002). Relationship between alertness, performance, and body temperature in human. American Journal of Physiology, 283, 1370–1377.
Xhelilaj, E., & Lapa, K. (2010). The role of human fatigue factor towards maritime casualties. Maritime Transport & Naviga-tion Journal, 2(2), 23–32.
Alireza Noroozi recently graduated with a PhD in the area of human factors importance and applica-tion in oil and gas engineering operations from Fac-ulty of Engineering and Applied Science at Memo-rial University. He completed his BSc and MSc in mechanical engineering in Iran.
Rouzbeh Abbassi is a Postdoctoral Fellow at Princ-eton University. He finished his PhD in 2009 at Memorial University and is currently undertaking research on industrial application of risk and reli-ability assessment techniques.
Scott MacKinnon is former Dean of School of Human Kinetics, Memorial University of New-foundland. His research focuses on health and safety in marine and coastal occupations, understanding the
relationship between physical work exposure and health risk and the impact of participatory ergonom-ics in the establishment of safe and healthy work environments. He is leading a multimillion-dollar research initiative on human factor performance evaluation in emergency condition.
Faisal Khan is Chair of Process Engineering in Memorial’s Faculty of Engineering and Applied Sci-ence and also Vale Research Chair in Process Risk and Safety Engineering. His research focuses on developing advanced methodologies and models for risk assessment and design of safety measure for processing facilities. He is leading a safety and risk engineering research group at Memorial University (www.engr.mun.ca/research/sreg/).
Nima Khakzad joined Memorial University in 2009 as a PhD student. Currently, he is a Postdoctoral Fel-low at Faculty of Engineering and Applied Science of Memorial University, conducting research on developing advanced risk and reliability assessment techniques in processing facilities.
Date received: November 21, 2012Date accepted: October 13, 2013
at MEMORIAL UNIV OF NEWFOUNDLAND on February 25, 2014hfs.sagepub.comDownloaded from
