Process accident prediction using Bayesian network based on IT2Fs and Z-number: A case study of spherical tanks.

This study aimed to propose a novel method for dynamic risk assessment using a Bayesian network (BN) based on fuzzy data to decrease uncertainty compared to traditional methods by integrating Interval Type-2 Fuzzy Sets (IT2FS) and Z-numbers. A bow-tie diagram was constructed by employing the System Hazard Identification, Prediction, and Prevention (SHIPP) approach, the Top Event Fault Tree, and the Barriers Failure Fault Tree. The experts then provided their opinions and confidence levels on the prior probabilities of the basic events, which were then quantified utilizing the IT2FS and combined using the Z-number to reduce the uncertainty of the prior probability. The posterior probability of the critical basic events (CBEs) was obtained using the beta distribution based on recorded data on their requirements and failure rates over five years. This information was then fed into the BN. Updating the BN allowed calculating the posterior probability of barrier failure and consequences. Spherical tanks were used as a case study to demonstrate and confirm the significant benefits of the methodology. The results indicated that the overall posterior probability of Consequences after the failure probability of barriers displayed an upward trend over the 5-year period. This rise in IT2FS-Z calculation outcomes exhibited a shallower slope compared to the IT2FS mode, attributed to the impact of experts' confidence levels in the IT2FS-Z mode. These differences became more evident by considering the 10-4 variance compared to the 10-5. This study offers industry managers a more comprehensive and reliable understanding of achieving the most effective accident prevention performance.

Determining performance shaping factors to assess human error in the emergency response team in chemical process industries: a case study.

Objectives. Human error (HE) plays a crucial role in the occurrence of accidents in chemical process industries (CPIs). Emergency response team (ERT) members are predisposed to HEs due to the nature of their work. The HE potential is influenced by the performance shaping factors (PSFs). Managing PSFs can diminish the human error probability (HEP) and consequently increase the emergency response success chance. This article aimed to determine the PSFs for ERT members in CPIs. Methods. First, an initial list of PSFs was searched and classified within human reliability analysis methods and studies. Then, an expert panel of the emergency management system was utilized to identify, classify and weight the initial PSFs. The fuzzy Delphi method and content analysis technique were applied to summarize and categorize the PSFs. Results. The results of the study showed that 11 PSFs had greater impacts on the ERT members' error potential. Findings revealed that stress and physiological stressors, competency, and team and organization were the three most important PSFs. Conclusion. The most important and relevant PSFs can be effectively used in accurate HE assessment of ERT members in CPIs.

Fatigue profile among petrochemical firefighters and its relationship with safety behavior: the moderating and mediating roles of perceived safety climate.

The present study aimed to investigate fatigue among firefighters and its relationship with their safety behavior. The moderation and mediation effects of perceived safety climate on this relationship were also investigated. Firefighters from 10 petrochemical companies in Iran were investigated. Moderation analysis was performed based on hierarchical regression analyses and mediation effects were studied using structural equation modeling. A total number of 261 firefighters participated in this study. The mean scores of general fatigue, physical fatigue, reduced activity, reduced motivation and mental fatigue were 10.5, 9.0, 8.2, 7.6 and 8.7, respectively. Fatigue had a negative significant effect on firefighters' safety behavior. Perceived safety climate had a mediation but not moderation effect on the relationship between fatigue and safety behavior of firefighters. Generally, the fatigue level among firefighters was fairly low, but had a significant negative effect on firefighter's safety behavior. Perceived safety climate can mediate such a negative effect.

Prediction of human error probability during the hydrocarbon road tanker loading operation using a hybrid technique of fuzzy sets, Bayesian network and CREAM.

Objectives. The hydrocarbon road tanker loading operation is vulnerable to human error. The present study aimed to develop a methodology for predicting human error probabilities (HEPs) in various subtasks of this operation. Methods. First, task analysis was performed using hierarchal task analysis. Then, HEP was calculated using a hybrid technique of fuzzy set theory (FST), Bayesian network (BN) and cognitive reliability and error analysis method (CREAM). FST was used for handling uncertainties regarding common performance conditions (CPCs) and the BN was employed for modeling the interrelationships among CPCs and HEPs. The weighted sum algorithm was used for quantifying conditional probability tables in the network. Results. Twenty-six subtasks were required for completing the road tanker loading operation. Investigating the internal parts of the tanker before the loading operation and attaching the ground rode clamp were the subtasks with highest HEPs. Working conditions and crew collaboration were the CPCs with the highest contribution to these errors. HEP was most sensitive to crew collaboration. Conclusion. Improving collaboration among the driver, site operators and control room operators, as well as increasing the knowledge of the road tanker driver regarding the hazards of incompatible chemicals, are the best practices for reducing HEPs in this operation.

Need for recovery scale in the firefighting context: Examination of the Persian culture-adapted version.

BACKGROUND: Work-related fatigue can cause disruption in mental and physical activities of the workers. Considering the importance of recovery from fatigue in emergency services, as seen in firefighting duties, the preparation and validation of the Need For Recovery Scale (NFRS) questionnaire can be useful and effective. OBJECTIVE: This study assessed the reliability and validity of the Persian version of Need For Recovery Scale among industrial firefighters. METHODS: Need for recovery scale contains 11 questions. First, a questionnaire was translated using backward-forward method. Then, for measuring validity of the questionnaire, the opinions of 10 experts in related fields were obtained. Content Validity Index (CVI) and Content Validity Ratio (CVR) were calculated. The reliability was determined through 107 questionnaires completed by the industrial firefighters. RESULTS: The average age of the firefighters in this study was 37.5 years and they had an average of 10 years of work experience. The mean Need For Recovery Index from fatigue was 33.03 for the study population. Of the first 11 questions, 9 questions in the original NFRS showed a satisfactory level of CVR. The Content Validity Index and Cronbach's α were 0.89 and 0.77 respectively. CONCLUSION: Based on the results, the Need For Recovery Scale is a reliable tool for assessing fatigue and revealed an acceptable level of validity and reliability that makes it usable for fatigue assessment in the industrial firefighters.

Application of Social Network Analysis to Major Petrochemical Accident: Interorganizational Collaboration Perspective.

OBJECTIVE: Crisis management in major accidents requires the collaboration among different organizations. One of the most important problems of crisis management is the lack of coordination between executive organizations. The aim of this study was to examine the structural characters and problems of interorganizational network during crisis in the petrochemical industry and provide solutions to achieve the highest performance in crisis management. METHODS: The organizations involved in crisis management were identified through interviews and questionnaires. Gephi (0.9.1) software was used to examine interorganizational relationships. RESULTS: In this study, the crisis management team consisted of 25 public and private organizations and non-governmental organizations. The highest betweenness centrality was observed in Crisis Management of Provincial Government (CMPG) (142.16) and Fire Department of Petrochemical Complex (FDC) (89.3). The highest closeness centrality was observed in FDC (0.77), CMPG (0.7), Shazand Governorate (0.7), and Crisis Management of University of Medical Sciences (0.7). CONCLUSIONS: Coordination between organizations plays an important role in crisis and emergency management, and social network analysis helps identify strengths and weaknesses of organizations involved in crisis management, overcome those weaknesses, and consequently achieve the best performance in crisis management.

Dynamic Human Error Assessment in Emergency Using Fuzzy Bayesian CREAM.

BACKGROUND: Human error is one of the major causes of accidents in the petrochemical industry. Under critical situation, human error is affected by complex factors. Managing such a situation is important to prevent losses and injury. This study aimed to develop a dynamic model of human error assessment in emergencies in the petrochemical industry. STUDY DESIGN: A cross-sectional study. METHODS: Fuzzy Bayesian network was used to improve the capabilities of the method for determining the control mode. Fuzzy-AHP-TOPSIS method was also used to prioritize emergency scenarios and human error assessment was applied for the most important emergency condition. RESULTS: Fire in a chemical storage unit was recognized as the most important emergency condition. Common Performance Conditions (CPCs) were determined based on the opinions of a panel of 30 experts and specialists and 7 CPCs were selected for emergencies; then, based on the results of AHP method the relative weights were determined. Finally, membership functions, inputs, and outputs of fuzzy sets, CPC values in 8 emergency response tasks, and the probability of control modes were determined using Bayesian Cognitive Reliability and Error Analysis Method (CREAM) method. CONCLUSION: This method could be applied to overcome the weaknesses of traditional methods, provide a repeatable method for human error assessment, and manage human error in an emergency.

Analysis of human and organizational factors that influence mining accidents based on Bayesian network.

Purpose. The present study aimed to analyze human and organizational factors involved in mining accidents and determine the relationships among these factors. Materials and methods. In this study, the human factors analysis and classification system (HFACS) was combined with Bayesian network (BN) in order to analyze contributing factors in mining accidents. The BN was constructed based on the hierarchical structure of HFACS. The required data were collected from a total of 295 cases of Iranian mining accidents and analyzed using HFACS. Afterward, prior probability of contributing factors was computed using the expectation-maximization algorithm. Sensitivity analysis was applied to determine which contributing factor had a higher influence on unsafe acts to select the best intervention strategy. Results. The analyses showed that skill-based errors, routine violations, environmental factors and planned inappropriate operation had higher relative importance in the accidents. Moreover, sensitivity analysis revealed that environmental factors, failed to correct known problem and personnel factors had a higher influence on unsafe acts. Conclusion. The results of the present study could provide guidance to help safety and health management by adopting proper intervention strategies to reduce mining accidents.

Effects of human and organizational deficiencies on workers' safety behavior at a mining site in Iran.

OBJECTIVES: Throughout the world, mines are dangerous workplaces with high accident rates. According to the Statistical Center of Iran, the number of occupational accidents in Iranian mines has increased in recent years. This study investigated and analyzed the human and organizational deficiencies that influenced Iranian mining accidents. METHODS: In this study, the data associated with 305 mining accidents were analyzed using a systems analysis approach to identify critical deficiencies in organizational influences, unsafe supervision, preconditions for unsafe acts, and workers' unsafe acts. Partial least square structural equation modeling (PLS-SEM) was utilized to model the interactions among these deficiencies. RESULTS: Organizational deficiencies had a direct positive effect on workers' violations (path coefficient, 0.16) and workers' errors (path coefficient, 0.23). The effect of unsafe supervision on workers' violations and workers' errors was also significant, with path coefficients of 0.14 and 0.20, respectively. Likewise, preconditions for unsafe acts had a significant effect on both workers' violations (path coefficient, 0.16) and workers' errors (path coefficient, 0.21). Moreover, organizational deficiencies had an indirect positive effect on workers' unsafe acts, mediated by unsafe supervision and preconditions for unsafe acts. Among the variables examined in the current study, organizational influences had the strongest impact on workers' unsafe acts. CONCLUSIONS: Organizational deficiencies were found to be the main cause of accidents in the mining sector, as they affected all other aspects of system safety. In order to prevent occupational accidents, organizational deficiencies should be modified first.

The effect of artificial smoke compound on physiological responses, cognitive functions and work performance during firefighting activities in a smoke-diving room: an intervention study.

This study investigated the combined effect of firefighting operations and exposure to smoke on physiological responses, cognitive functions and physical performance among firefighters. Each firefighter was engaged in two experimental conditions: (a) without artificial smoke, control (NS); (b) with artificial smoke (WS). The results of data analysis revealed that heart rate and body temperature increased significantly after the firefighting activities in the two conditions relative to baseline. In addition, reaction time (RT) and correct response (CR) decreased after the activity relative to baseline and there was a significant difference in RT and CR between the NS and WS conditions. There was also a significant difference in performance time between the NS and WS conditions. These findings demonstrated that a smoke compound has an insignificant effect on physiological responses during firefighting activity. On the contrary, the smoke compound has a detrimental influence on cognitive and physical performance.

A path analysis model for explaining unsafe behavior in workplaces: the effect of perceived work pressure.

BACKGROUND: Unsafe behavior is closely related to occupational accidents. Work pressure is one the main factors affecting employees' behavior. The aim of the present study was to provide a path analysis model for explaining how work pressure affects safety behavior. METHODS: Using a self-administered questionnaire, six variables supposed to affect safety employees' behavior were measured. The path analysis model was constructed based on several hypotheses. The goodness of fit of the model was assessed using both absolute and comparative fit indices. RESULTS: Work pressure was determined not to influence safety behavior directly. However, it negatively influenced other variables. Group attitude and personal attitude toward safety were the main factors mediating the effect of work pressure on safety behavior. Among the variables investigated in the present study, group attitude, personal attitude and work pressure had the strongest effects on safety behavior. CONCLUSION: Managers should consider that in order to improve employees' safety behavior, work pressure should be reduced to a reasonable level, and concurrently a supportive environment, which ensures a positive group attitude toward safety, should be provided. Replication of the study is recommended.

The Effect of Various Hot Environments on Physiological Responses and Information Processing Performance Following Firefighting Activities in a Smoke-Diving Room.

BACKGROUND: Fire service workers often implement multiple duties in the emergency conditions, with such duties being mostly conducted in various ambient temperatures. METHODS: The aim of the current study was to assess the firefighters' physiological responses, information processing, and working memory prior to and following simulated firefighting activities in three different hot environments. Seventeen healthy male firefighters performed simulated firefighting tasks in three separate conditions, namely (1) low heat (LH; 29-31°C, 55-60% relative humidity), (2) moderate heat (MH; 32-34°C, 55-60% relative humidity), and (3) severe heat (SH; 35-37°C, 55-60% relative humidity). It took about 45-50 minutes for each firefighter to finish all defined firefighting activities and the paced auditory serial addition test (PASAT). RESULTS: At the end of all the three experimental conditions, heart rate (HR) and tympanic temperature (TT) increased, while PASAT scores as a measure of information processing performance decreased relative to baseline. HR and TT were significantly higher at the end of the experiment in the SH (159.41 ± 4.25 beats/min; 38.22 ± 0.10°C) compared with the MH (156.59 ± 3.77 beats/min; 38.20 ± 0.10°C) and LH (154.24 ± 4.67 beats/min; 38.17 ± 0.10°C) conditions (p < 0.05). There was no significant difference in PASAT scores between LH and MH (p > 0.05). Nonetheless, there was a measurable difference in PASAT scores between LH and SH (p < 0.05). CONCLUSION: These consequences demonstrate that ambient temperature is effective in raising the physiological responses following firefighting activities. It is therefore argued that further increase of ambient temperature can impact firefighters' information processing and working memory during firefighting activity.

Determining the appropriate strategies for emergency planning through AHP-SWOT.

During an unexpected incident, companies should demonstrate appropriate behaviour based on predetermined and rehearsed emergency strategies. This paper describes how to select the proper strategies for emergency situations via means of the AHP-SWOT tool, where the initial SWOT analysis is conducted for the emergency management system, and the final strategies are selected via the analytic hierarchy process (AHP). Alternative response strategies comprise internal, shared and external responses. Most of the scenarios investigated for this study demanded a shared response. These findings highlight the importance of mutual aid agreements, cooperative exercises and the improvement of communication systems. Organisations can take advantage of integrated approaches to select the best strategies and tactics for normal situations in general and emergency situations in particular.

Selecting Strategies to Reduce High-Risk Unsafe Work Behaviors Using the Safety Behavior Sampling Technique and Bayesian Network Analysis.

BACKGROUND: High-risk unsafe behaviors (HRUBs) have been known as the main cause of occupational accidents. Considering the financial and societal costs of accidents and the limitations of available resources, there is an urgent need for managing unsafe behaviors at workplaces. The aim of the present study was to find strategies for decreasing the rate of HRUBs using an integrated approach of safety behavior sampling technique and Bayesian networks analysis. STUDY DESIGN: A cross-sectional study. METHODS: The Bayesian network was constructed using a focus group approach. The required data was collected using the safety behavior sampling, and the parameters of the network were estimated using Expectation-Maximization algorithm. Using sensitivity analysis and belief updating, it was determined that which factors had the highest influences on unsafe behavior. RESULTS: Based on BN analyses, safety training was the most important factor influencing employees' behavior at the workplace. High quality safety training courses can reduce the rate of HRUBs about 10%. Moreover, the rate of HRUBs increased by decreasing the age of employees. The rate of HRUBs was higher in the afternoon and last days of a week. CONCLUSIONS: Among the investigated variables, training was the most important factor affecting safety behavior of employees. By holding high quality safety training courses, companies would be able to reduce the rate of HRUBs significantly.

The effect of practical cooling strategies on physiological response and cognitive function during simulated firefighting tasks.

Background: Firefighters often perform multiple tasks during firefighting operations under unknown and unpredictable conditions in hot and hostile environments. Methods: In this interventional study each firefighters engaged in 4 conditions: namely (1) no cooling device; control (NC), (2) cooling gel (CG), (3) cool vest (CV), and (4) CG+CV. Cooling effects of the employed interventions were evaluated based on heart rate (HR), temporal temperature (TT), reaction time (RT), and the correct response (CR). Results: HR and TT values for use of CG+CV (147.47 bpm [SD 4.8]; 37.88°C [SD 0.20]) and CV bpm (147.53 [SD 4.67]; 37.90°C [SD 0.22]) were significantly lower than the CG (153.67 bpm [SD 4.82]; 38.10°C [SD 0.22]) and NC (154.4 bpm [SD 4.91]; 38.11°C [SD 0.23]) at the end of the activity. RT and CR for use of CG + CV (389.87 ms [SD 6.12]; 143.53 [SD 1.24]) and CV (389.53 ms [SD 6.24]; 143.47 [SD 1.18]) were significantly higher than the CG (385.73 [SD 7.25] ms; 143.07 [SD 0.88]) and NC (385.67 ms [SD 7.19]; 143.00 [SD 0.84]) at the end of the activity. Conclusion: It is concluded that CV was more effective than the CG in attenuating physiological responses and cognitive functions during firefighting operations. Furthermore, combining CV with CG provides no additional benefit.

Constructing a Bayesian network model for improving safety behavior of employees at workplaces.

INTRODUCTION: Unsafe behavior increases the risk of accident at workplaces and needs to be managed properly. The aim of the present study was to provide a model for managing and improving safety behavior of employees using the Bayesian networks approach. METHODS: The study was conducted in several power plant construction projects in Iran. The data were collected using a questionnaire composed of nine factors, including management commitment, supporting environment, safety management system, employees' participation, safety knowledge, safety attitude, motivation, resource allocation, and work pressure. In order for measuring the score of each factor assigned by a responder, a measurement model was constructed for each of them. The Bayesian network was constructed using experts' opinions and Dempster-Shafer theory. Using belief updating, the best intervention strategies for improving safety behavior also were selected. RESULTS: The result of the present study demonstrated that the majority of employees do not tend to consider safety rules, regulation, procedures and norms in their behavior at the workplace. Safety attitude, safety knowledge, and supporting environment were the best predictor of safety behavior. Moreover, it was determined that instantaneous improvement of supporting environment and employee participation is the best strategy to reach a high proportion of safety behavior at the workplace. CONCLUSION: The lack of a comprehensive model that can be used for explaining safety behavior was one of the most problematic issues of the study. Furthermore, it can be concluded that belief updating is a unique feature of Bayesian networks that is very useful in comparing various intervention strategies and selecting the best one form them.

Major accident prevention through applying safety knowledge management approach.

OBJECTIVE: Many scattered resources of knowledge are available to use for chemical accident prevention purposes. The common approach to management process safety, including using databases and referring to the available knowledge has some drawbacks. The main goal of this article was to devise a new emerged knowledge base (KB) for the chemical accident prevention domain. DESIGN: The scattered sources of safety knowledge were identified and scanned. Then, the collected knowledge was formalized through a computerized program. The Protégé software was used to formalize and represent the stored safety knowledge. RESULTS: The domain knowledge retrieved as well as data and information. This optimized approach improved safety and health knowledge management (KM) process and resolved some typical problems in the KM process. CONCLUSION: Upgrading the traditional resources of safety databases into the KBs can improve the interaction between the users and knowledge repository.

Developing an ontological explosion knowledge base for business continuity planning purposes.

Industrial accidents are among the most known challenges to business continuity. Many organisations have lost their reputation following devastating accidents. To manage the risks of such accidents, it is necessary to accumulate sufficient knowledge regarding their roots, causes and preventive techniques. The required knowledge might be obtained through various approaches, including databases. Unfortunately, many databases are hampered by (among other things) static data presentations, a lack of semantic features, and the inability to present accident knowledge as discrete domains. This paper proposes the use of Protégé software to develop a knowledge base for the domain of explosion accidents. Such a structure has a higher capability to improve information retrieval compared with common accident databases. To accomplish this goal, a knowledge management process model was followed. The ontological explosion knowledge base (EKB) was built for further applications, including process accident knowledge retrieval and risk management. The paper will show how the EKB has a semantic feature that enables users to overcome some of the search constraints of existing accident databases.