Rigorous maintenance practices for overhead lines not only extend an asset’s life, but can also detect damage conditions early to avoid a premature failure, thus avoiding forced outages. Traditional time-based maintenance practices collect inspection data at a fixed time interval and execute maintenance decisions based on the processing of asset condition data, which are primarily deterministic.
This report uses a stochastic model (Markov chain model) to apply asset condition data; this model considers the uncertainty in the deterioration rate, which is determined from a transitional probability matrix (TPM) of the condition data. The optimum maintenance interval is then determined by using a survival curve derived from the deterioration rate and by balancing the cost of maintenance against the cost of replacement (cost of failure). The optimization model uses a Weibull distribution in which the two key parameters are determined from the survival curve developed from the Markov chain model. Two example problems are presented: the first illustrates the methodology while the second presents a non-dimensional chart of cost-ratio (cost of maintenance to cost of replacement/failure) against the maintenance interval. The second example problem shows not only that the maintenance activity is highly cost-effective when the cost of maintenance is considerably lower than the expected cost of failure, but also that the determined interval is in line with the industry’s best practices.
inspection of line, deterioration model, asset condition rating, Markov chain model, transitional probability, Weibull analysis, optimum maintenance interval