Hybrid preventive maintenance of competing failures under random environment

Competing failures are extensively observed in complex industrial systems, which may result in tremendous economic losses and safety hazards. In this article, we study a system subject to two typical failure modes, degradation-based failure and sudden failure. The system is operating under a random environment where external shocks arrive according to a Poisson process. The impact of shock damage on system failure is two-fold: (a) increase the hazard rate of sudden failure; (b) cause abrupt degradation increment. The system is preventively replaced when its age attains a pre-determined threshold (age-based replacement), and undergoes a finite number of condition monitoring (CM) before this replacement. At a CM, the control limit of preventive replacement varies with the number of CM and is determined by a reliability criterion. The objective of this paper is to jointly optimize the replacement interval, monitoring interval and reliability criterion such that the expected cost per unit time is minimized. A case study on oil pipeline is provided to illustrate the applicability of the maintenance strategy.