The objective of this paper is to compare the performance of three on-line test and maintenance strategies (corrective maintenance, preventive maintenance and predictive maintenance) for standby kk-out-of-nn safety systems. Each channel of the kk-out-of-nn system is modelled by an age-dependent unavailability model to reflect the effect of maintenance on the aging process. The system unavailability, the probability of spurious operation and the overall cost under the above maintenance strategies are analyzed and compared to obtain the optimal maintenance strategy. Sensitivity analyses are performed to reveal the effect of different model parameters on the system performance. A standby safety system in Canadian Deuterium–Uranium (CANDU) Nuclear Power Plants (NPPs), the Shutdown System Number One (SDS1), is used to illustrate the proposed analysis and the procedure. It is concluded that maintenance should neither be performed too frequently nor too rarely. When the system deteriorates very slowly, the corrective maintenance is more preferable than the preventive and predictive maintenance. When the failure rate of the system is high, the preventive maintenance results in the best system performance.
For safety-critical control systems, such as those in Nuclear Power Plants (NPPs), chemical processes, and aerospace systems, it is important to have standby safety (shutdown) systems in addition to on-line regulation systems. During severe conditions, the standby system can take actions automatically to lead the system to a safe state and to prevent catastrophic consequences. However, standby safety systems may fail to operate on demand as well because of latent failures. Therefore, it is important to reveal and fix these problems through test and maintenance.
Normally, the test and maintenance are carried out following the recommendations from the system manufacturers, and they generally tend to be conservative. In safety-critical applications, the maintenance interval is often dictated by regulatory bodies. For example, in the field of NPPs, NRC1 (US) and CNSC2 (Canada) impose strict rules on the frequency of maintenance to keep the unreliability below a certain level. In general, more frequent tests can increase the likelihood of disclosing a failure. However, it may also deteriorate the system faster, incur unnecessary expenditure on resources, and can lead to a higher rate of spurious operations. Therefore, a reasonable and effective test and maintenance strategy is of considerable importance for standby safety systems.
The kk-out-of-nn systems are widely used in standby safety systems [1] and [2]. Basically, a kk-out-of-nn system consists of nn parallel channels, as long as kk of them are operational, the system will be able to perform its intended function. Each channel might have two failure modes: (1) channel unavailability, and (2) spurious operation. Channel unavailability means that the channel fails to function when there is a request. Spurious operation means that the channel takes action when there is actually no need. According to the logic of a kk-out-of-nn system, if more than nn–kk channels are unavailable, the kk-out-of-nn system will become unavailable. If kk or more channels function spuriously, the system will function spuriously.
In this paper, a general procedure to determine the optimal test and maintenance strategy for a kk-out-of-nn system has been proposed. The ultimate goal is to determine the relationships among the unavailability, and the probability of spurious operation, and the overall cost of maintenance. In order to achieve this goal, the analysis of the unavailability, the probability of spurious operation and the overall cost for a kk-out-of-nn system under different test and maintenance strategies have to be performed. It is assumed that each channel in a kk-out-of-nn system can be modelled in terms of an age-dependent model as described in [3]. This model evaluates, in a quantitative manner, how test and maintenance affects the system performance and the overall cost. The model explicitly considers how maintenance affects the life of a component. Reliability parameters such as demand-failure probability, standby failure rate and aging rate have all been considered. Component unavailability as a function of test and maintenance interval is obtained analytically. The analysis has revealed the relations among system parameters and provided insights into system characteristics [4]. Based on the analysis results, the optimal test and maintenance strategy can be determined.
This paper is organized as follows: in Section 2, three test and maintenance strategies are introduced. Then, the age-dependent model for a single component is presented in Section 3. Based on this model, the system unavailability, the probability of spurious operation and the overall cost for a kk-out-of-nn system under the three test and maintenance strategies are analyzed in Section 4. Subsequently, the general procedure of maintenance decision making is presented. Following this, in Section 5, a standby safety system in Canadian Deuterium–Uranium (CANDU) NPPs [5] and [6], the Shutdown System Number One (SDS1), is used to illustrate how the proposed process of maintenance strategy is determined. Finally, in Section 6, sensitivity analyses are performed on several parameters to assess their effects on the optimal maintenance strategy.
This paper analyzes three on-line test and maintenance strategies for a standby k-out-of-n safety system by examining the system unavailability, the probability of spurious operation and the overall cost. It is shown that during on-line test and maintenance, a k-out-of-n configuration will become a (k 1)-out-of-(n 1). The later has a lower level of unavailability but a higher probability of spurious operation. A general framework of maintenance decision making for standby safety systems is proposed.
The Shutdown System Number One (SDS1) in CANDU nuclear power plants has been selected to illustrate the concept developed. It is concluded that for all three test and maintenance strategies considered, neither too frequent nor too rare test and maintenance are desirable. An optimal test and maintenance interval can be determined following the proposed procedure in this paper. The effect
of system parameters on maintenance decision making has also been investigated. It is shown that a high failure rate renders the three test and maintenance strategies function similarly. When the system deteriorates very slowly, the corrective maintenance is the most cost-effective choice.
