Maintenance scheduling under age replacement policy using proportional hazards model and TTT-plotting

The failure characteristics of a sytem may depend on the total operating time, operating time since the last repair, failure history, operating conditions or on the values of monitored variables. A reliability based approach which takes into consideration values of monitored variables is suggested for estimating the optimum maintenance (or replacement) time interval for a system or threshold values of monitored variables under the age replacement policy. The maintenance cost equation is formed on the basis of the planned and unplanned maintenance costs and the values of monitored variables. The proportional hazards model is used to identify the importance of monitored variables. The reliability function is estimated considering the values of monitored variables. A total time on test plot based on this estimate of the reliability function is used to estimate the optimum maintenance (or replacement) time interval for a system or threshold values of monitored variables. This approach is illustrated with an example.