Abstract

Power distribution systems are physically very large exposed entities with little redundancy. They contain a multiplicity of feeder lines of varying configuration, age and maintainability and have very long service lives. This work describes repairable system and life data methods for assessing preventive maintenance of such systems. It is postulated that a power line can be treated as a system which is minimally repaired upon failure by emergency maintenance and subsequently restored to a as old state by preventive maintenance. The arrival of failure events over time can be treated as a stochastic point process modelled as a nonhomogeneous Poisson process with nonconstant intensity function to determine if the system is deteriorating or improving and when to apply maintenance relative to emergency and preventive maintenance costs. The failure event may mark the loss of a nonrepairable component. Providing suitable life and cost data is available, a distribution can be derived and utilised to determine if batch replacement of similar components of comparable operational age should take place or left until failure. Where nonrepairable components initiate an outage, replacement is rarely drawn from the original population although functionality remains unchanged. The research associated with this work related to a very large widespread rural-provincial system and a portion of an urban network. This paper confines itself to an urban feeder based on six years of interruption data and observations made of a particular manufacture of a 33 kV insulator which had a history of unreliability. A specific objective of the feeder study was to determine underlying trends by discounting various weather states, and whether interruptions can be characterised by a simple Poisson process, the conventional approach adopted in power system reliability analysis. Critical to any evaluation method is an understanding of how line systems and their components perform over time within a localised operating environment. A power line like any piece of equipment can never be any better than that dictated by its inherent maintainability. If it is a bad design, it will be unreliable and remain so whatever the maintenance regime.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call