Validity of using a global repair service model in power system reliability studies

Traditionally, repairs has been modeled in power system reliability as part of the component reliability models; an alternative way is to represent them by means of queuing theory concepts. One of the advantages of this approach is that it allows assessing the performance of crews, a feature not offered by any other method. However, the application of this modeling approach arises the following question: ¿what is the correct way to generate the times to repair, using a global model for all components or using one for each component or component class? To reply this question, in this paper reliability of a power transmission system is assessed using both approaches and considering for power transformers the cases of repair, replacement with a local spare and replacement with a spare taken from a remote location. The conclusion is that if repair durations of the components are very different is not valid to use a global model for generating the repair times for all components, i. e. the approach commonly used in QT is not valid here; instead, it is necessary to generate them using a model for each component or component class and this demands more computational time.