The Impacts of Total System Design Factors on Human Performance in Power Plants

The objective of this study is to evaluate the impact of total system design factors (TSD) on human performance in a power plant. The TSD factors are defined as design factors, which have impact on overall performance of the power plants in context of total human engineering or macroergonomy. The systems being studied are the control rooms and maintenance departments of a 2000 MW thermal power plant? To achieve the above objective the TSD factors were addressed and assessed through a detailed questionnaire. The relationships between TSD factors and human performance were then examined through non-parametric correlation analysis (Kramer’s Phi) and Kruskal-Wallis test of means. The selected TSD factors are related to procedures, work assessment, teamwork, self-organization, information exchange and communication. The results of this study show that the TSD factors such as organizational and safety procedures, teamwork, self-organization, job design and information exchange influence human performance in the power plant. The findings also suggest that the selected TSD factors are correlated to human performance and must be considered, designed and tested concurrently with the engineering factors at the design phase of the system developmental cycle. Consequently, total system’s faults and organizational errors are reduced to an acceptable level and human performance is significantly increased. This is a challenging task for designers of power plants but is required if we are to face unforeseen and complex issues of such systems in twenty-first century. The methodology discussed here may be easily extended to all types of


INTRODUCTION
Total System Design (TSD) is an integrated developmental process, which is based on a series of well-defined phases. Frequently in the past, designers used other approaches without giving much attention to human performance. TSD requires equal consideration to all major components of the system such as human, hardware, software and organizational structures. Indeed, it is quite important to pay serious attention to human and organizational aspects of the TSD process from early design phase.
Total system design factors in context of human performance are referred to as socio-technical factors in context of system design. It should be noted that the engineering design process is often perceived as mainly technical activity, yet within engineering design organization it really only coheres as a social activity. This study introduces the socio-technical factors as essential and vital part of the design process in power plants and because they are related to overall management and organization structures, they are referred to as total system design (TSD) factors in context of human performance [1][2][3] .
TSD factors in context of human performance define the macroergonomics features of the system design and human performance engineering, whereas, the conventional system design factors in context of human performance define the ergonomics features of the system design and human performance engineering. Macroergonomic and the concept of total human factors were developed by Hendrick and Meshkati and have been elaborated by other researchers [4][5][6][7][8][9][10] .
Ergonomic attempts to optimize the interaction between human operator and machine. It considers those factors of machine, design and work posture that affect the user interface and working conditions related to the job or task deign. In a macroergonomics study, the ergonomics factors are considered in parallel to organizational and managerial aspects of working conditions in context of a total system design. Moreover, it attempts to create equilibrium between, organization, operators and machines. It focuses on total "people-technology" systems and is concerned with the impacts of technological systems on organizational, managerial and personnel subsystems [11][12][13] .
Macroergonomic adopts a more holistic approach to human factors' problems of manufacturing systems. It considers the whole and avoids the trap of dealing with specialties with which we feel comfortable. A macroergonomics program optimizes interface between operators, machines and organization by using teamwork, on-the-job training, well-defined procedures and total management.

MATERIALS AND METHODS
TSD factors in context of human performance are defined as factors influencing total system's performance such as rules and procedures and information exchange between personnel/departments. To measure the impacts of TSD factors on human performance, a quesionnair was designed and handed out to all control room and maintenance operators. It was designed based on total system design aspects of human performance in power plants. Moreover, key macroergonomics factors were included to evaluate human performance. The selected TSD factors are related to procedures, work assessment, teamwork, selforganization, information exchange and communication. They were inputted to the questionnaire and their statistical relationships to the human performance were examined through two nonparametric statistical (namely, Cramer's Phi and Kruskasl-Wallis) approach. The selected TSD factors in context of human performance were tested in the following format: 1. Degree of familiarity with rules and procedures 2. Supervisors' monitoring and assessment at work 3. Reward for teamwork by supervisors 4. Ease of contact with supervisors 5. Problems with co-workers due to interorganizational relationship 6. Quality of perceived information from supervisors 7. Quality of perceived information from co-workers 8. Usefulness of informal information exchange 9. Freedom for self-organized and individual decision-making As mentioned, a set of non-parametric test of hypothesis is conducted to foresee if human performance is independent of the selected TSD factors. Furthermore, job pressure is selected as the factor representing human performance since it is identified as one of the most important human shaping factors. The sources of job pressure in the power plants are classified as 1) workload, 2) stress and 3) time considerations. Because workload is identified as the most influential source of job pressure, it is selected as the measure of human performance in this study ( Fig. 1). It is tested whether job pressure due to workload is influenced by the TSD factors. Also, the difference between mean ratings of operators in respect to selected TSD factors are examined through Kruskal-Wallis test. For example, the operators who can easily communicate with supervisors are compared with the ones who can't easily communicate with supervisors in respect to the level of job pressure.

RESULTS
The Cramer's Phi statistic tests the null hypothesis (H 0 ) of no correlation between the two variables against alternative hypothesis (H 1 ) of correlation between the two variables [14] . The results of the non-parametric Cramer's Phi correlation between human performance (job pressure) and the nine TSD factors are presented in the Table 1. The test of hypothesis is in the following general format: H 0 : The TSD factors are not correlated with job pressure due to workload H 1 : Otherwise As shown there is strong evidence that the nine TSD factors are correlated with the job pressure at work. Furthermore the job pressure at work is influenced by familiarity with organizational rules and procedures and information flows between co-workers and co-workers and supervisors. Also, job pressure is positively correlated with teamwork. Operators who are rewarded for teamwork report lower level of job pressure and consequently produce higher performance. The freedom for self-organization is positively correlated with human performance. In summary, these findings suggest the positive impacts of TSD factors on human performance.
To further our investigation, series of comparative studies are performed between various groups of operators in the next section. It is examined if TSD factors influence the human performance in particular and the system in general. To achieve this objective, two groups of operators are examined on the selected response variables. The selected response variables are the quality of information perceived from supervisors and co-workers and job pressure. The Kruskasl-Wallis test performs an analysis that is very similar to an  analysis of variance (ANOVA) on the ranks. The test is conducted when the assumptions for the parametric ANOVA cannot be made [15] . Furthermore, it assumes independence between subjects in conditions. This test also acts as verification and validation process of the previous test and almost the same types of results are reported in different format. The general format for the test is as follows: Ho: The two groups of operators have the same performance with respect to the response variable, where the response variables are the quality of perceived information from supervisors and coworkers and job pressure.
Operators who can't easily communicate with supervisors report higher level of job pressure. Operators who can easily communicate with supervisors report higher quality of perceived information from supervisors. Operators who believe that there could be a better job design reported higher level of job pressure. This is an important finding which reveals the current system of job design is partially rather than totally optimized. This is due to lack of considerations of the TSD factors of when the current system of job design was designed and implemented. This means the existing system of job design must be re-engineered.
The significant levels of the tests (P-Value) on the quality of perceived information from supervisors and co-workers (TSD factors) and human performance (job pressure) are summarized in Table 2 and 3, respectively. The last column in Tables 2 and 3 define the relative advantage of group 1 over group 2 in relation to the quality of information perceived from supervisors and co-workers, respectively. Furthermore, the relative statistical advantage of group 1 over group 2 is tabulated by the percent increase in quality of information perceived from supervisors and coworkers, respectively. The last column in Table 3 defines the relative advantage of group 1 over group 2 in relation to the job pressure. The significant difference between the groups of operators who are utilizing the TSD factors and the groups who are not with respect to the response variables reveal that TSD factors extensively influence the human performance in particular and the system in general.
The Kruskal-Wallis test of comparison between the two groups verifies and validates the previous results obtained from the test of correlation between TSD factors and job pressure. It can be concluded that TSD factors significantly influence human performance and therefore they must be considered and designed concurrently with other conventional hardware and software factors in order to optimize human performance in particular and the system in general.

DISCUSSION
The conventional design approach in power plants considers the engineering design parameters and ergonomics factors (in some cases). However, the total system design (TSD) approach of this study in context of human performance considers the engineering design parameters and macroergonomics factors. The impacts of TSD factors on human performance are shown in this paper. This is shown through design and evaluation of a detailed survey containing information about TSD factors and human performance. It has been shown that a total system design approach in context of human performance is much more efficient than a conventional design approach. This is shown through introduction of the TSD model, applying the model in a power plant and showing its advantage through statistical analysis.
Non-parametric statistical analyses are used to show positive correlation between human performance and TSD factors and also to highlight the impact of TSD factors on human performance. Furthermore, it is noted that by designing and implementing a TSD approach, the system and its human element are totally rather than locally optimized in context of human performance.
It should be noted that the conventional design approach in context of human factors is only capable of identifying local or stationary human performance issues. This study shows that the employment of a TSD approach is superior to conventional design approach.
The findings of this study have several design implications. Rules and procedures, information exchange between personnel (operators and supervisors) teamwork and self-organization may be designed and accommodated through standardization of the documentation process and automated tracking systems. This may be achieved through: * Implementation of ISO 9000 series of standards to promote standardization of documentation (rules, procedures, guidelines and communications) process. * Implementation of ISO 14000 series of standards to promote standardization of documentation process for environmental management systems * Implementation of OHSAS 18000 to develop standardization of documentation process for safety management and occupational hygiene systems. * Design and implementation of automated information exchange in context of information technology. This would facilitate and enhances the existing information structure.
Design and implementation of the re-engineering concept may enhance organizational relationships and surveillance. Re-engineering is the collection of activities and mechanisms required changing from hierarchical to horizontal, flat and cross-functional structures based on teamwork within an organization. The main goal in such program is customer's satisfaction. More elaboration on the scientific tools for implementation of TSD factors in context of human performance is left for a full research paper in the future [16,17] .