The relationship between human factors and plant maintenance reliability in a petroleum processing organisation

Abstract

Despite the considerable emphasis on improving maintenance reliability in the petroleum industry by adopting an engineering approach (International Standards Organization, 2006b), production losses, ineffective maintenance, and major disasters continue to occur (Urbina, 2010; Pidgeon, 2000). Analyses of these events have indicated that a failure to consider the human factors in the design (Taylor, 2007), operation (Øien, 2001a), or maintenance (Bea, 1998) of hazardous process technologies is often an important contributor. Based on research to evaluate the influence of these human factors on organisational performance, various models (Rasmussen, 1982; Dekker 2005) and taxonomies (Reason, 1998) for analysing organisational processes at the individual-, group- and organisational-level have been developed.By using these models, the current research was designed to determine the influence of human factors on maintenance reliability in petroleum operations. Three studies were conducted in petroleum operations with the objective in the first two studies of identifying the most-frequent contributors to maintenance-related failures, and in the third study, determining if group differences between higher and lower reliability work areas could be differentiated on the basis of these human factors.In Study 1, the First Priority incident database of the target organisation was used to determine the most frequently reported human factors in maintenance-related, lost-production failures. The most-frequent factors in the incidents (N=194) were found to be Violations, Design & Maintenance, Detection, and Decision-making. These results accorded with earlier studies in the field of human factors (Hobbs & Williamson, 2003; Lawton 1998), which frequently identified human error and violations as the causes of failures. Study 2 provided a more rigorous investigation of the organisational contributors to failures through structured interviews with maintenance personnel. The results of these interviews (N=38) using the Human Factors Investigation Tool (HFIT) (Gordon, 2005) demonstrated that Assumption, Design & Maintenance, and Communication were the most frequent contributors to maintenance-related failures.Based on the predominant factors identified in Study 2, a survey of the perceptions of maintenance personnel (N=178) was conducted for Study 3. Scales measuring Problem-solving (Morgeson & Humphrey, 2006) and Vigilance (Mann, Burnett, Radford, & Ford, 1997) were used to measure the processes that provoke assumptions. Design & Maintenance items from HFIT (Gordon, 2001), and scales from Wiio’s (1978 a&b) Organisational Communication Development questionnaire (OCD/2) were used to test the factors identified in Study 2. Exploratory Factor Analysis indicated that the responses to the Design & Maintenance items loaded onto a single variable, while the Communication items loaded onto two variables, which were named Job-related feedback and Information about change.The perceptions of personnel in lower and higher reliability work areas across the target organisation were compared using these scales, with reliability level ranked according to the monthly Mean Time Between Deferments of petroleum production. Significant between-group differences were found between work areas on Design & Maintenance and Problem-solving. These results suggest that better maintainability in the design of plant is predictive of higher reliability level. In addition, greater requirements for Problem-solving were associated with lower reliability level. There were no significant effects of reliability on Vigilance or either communication measure.The quantitative data was triangulated with comments in response to an open-ended question asking about factors that help or hinder maintenance activities. Respondent’s comments indicated that Communication was not significantly associated with reliability at the group-level. The reason appeared to be that Communication was an organisation-level property of the employing company. Many comments indicated that access to information was difficult, explaining the high occurrence of assumptions reported in Study 2. In addition, although maintenance personnel generally agreed in the survey that they were vigilant in decision-making, personnel in lower reliability facilities provided a higher proportion of comments indicating that the decision-making of supervisors and management had a negative impact on their work.The results of the three studies support past research demonstrating that problem-solving skills (Tucker, 2002) and the design of socio-technical facilities (Reiman, Oedewald & Rollenhagen, 2005) have an important influence on organisational performance. The findings further extend research in the field of human factors by demonstrating a significant relationship between these two factors and group-level performance. The findings also demonstrated the importance of organisational communication, but as an organisational-level dimension that might not influence group-level measures. This research has implications for organisations that operate complex, hazardous technologies and that are attempting to improve organisational processes by utilising a human factors approach.