Article Preview
TopIntroduction And Background
Maintenance management of an industrial plant consists of several activities, to ensure levels of availability proper to its production capacity.
The overall availability depends on performance and connections of machines that constitute it. From this point of view, an industrial plant can be considered as complex system, whose reliability target depends on the performance of its components. In this case the target depends on the quantities set out in the master production schedule and the service level established in supply contracts (Falcone et al., 2011).
Maintenance design activities are based on information collected by monitoring the condition of machines and processes. Information can be classified into direct information, and indirect information.
For collection of direct information, the parameter fault or wear condition parameters are measured (for example thickness of brake linings). On the other hand, indirect information give indications on the dynamics of failure, but they are not a direct measure (Christer & Wang 1995; Raheja et al., 2006).
The traditional literature in this field can be divided into three groups. The first group of research is mainly focused on the determination of effective inspection times, in order to ensure the correct functioning of the machines for predetermined times (Chen & Trivedi 2002; Wang 2003; Kalleh & van Noortwijk 2006; Wang & Jia 2007).
A common feature of the previous research is the use of direct information. Consequently, the condition of a system can be verified by monitoring it and then, after an inspection, it is possible to provide appropriate maintenance action. The aim is to determine the best time of inspection, in order to optimize an identified parameter (maximum availability or the minimum cost of production).
A second group of research is focused on the dynamic determination of the inspection time and the time of maintenance or replacement (Castanier et al., 2003; Chen & Trivedi 2005; Ghasemi et al., 2007; Wang et al., 2009). In this case future inspections and the maintenance policies are estimated according to specific performance criteria, such as the “long run system availability” and “long-run expected maintenance cost”. For this purpose, information on the wear level of the system result from non-periodic inspections.
The third group of research are intended to determine the optimal level of maintenance or replacement (Banjevic 2001; Chen & Wu 2007; Lu et al., 2007). A common assumption is that the information collected by monitoring the conditions of the system are indirect. The inspections are carried out periodically and not, but with a predetermined schedule. The aim is to determine a level that optimizes a given performance criterion.
Makis and Jardine (1992) give a definition of the limit level of a system, for a random fault. In their model, the checks are carried out at fixed interval. The equipment is replaced each time it fails. After each inspection and according to the results of the inspection, if the fault cost reaches or exceeds a predetermined limit, a preventive replacement is provided.
Another important issue is the choice of the right maintenance strategy. The most used model for this purpose and therefore for an efficient use of economic resources is the Failure Mode Effect and Criticality Analysis (FMECA) (Bowles 2003). However, it is important to note that this method has some limitations and weaknesses due to the fact that it does not take into account some important factors, such as the issues of economics. Furthermore, the considered factors are of equal importance, regardless of the context of application. In addition, the goodness of the results is influenced by the analyst's knowledge.
To overcome this limitations, other recent studies focused on maintenance decision models propose decision making approaches for maintenance strategy selection such as Analytic Hierarchy Process (AHP), Fuzzy set theory, Genetic Algorithm (GA) and Mathematical Programming (Zaim et al., 2012).