A comprehensive methodology for the optimization of condition-based maintenance in power transformer fleets

•There are limited methodologies for asset management in power transformer fleets in the state-of-the-art.•With the adequate amount of data, k-means clustering technique is able to accurately evaluate the risk of power transformers units.•Continuous times Markov chains can provide a realistic representation of the life-cycle of power transformers.•The deviation from the optimum inspection's interval could lead the important economic losses or longer unavailability times. Power transformers (PT) are critical assets in power systems. Electric utilities and power companies that manage PT fleets require a large amount of information to inform decision-making processes related to acquisition, maintenance, operation, and unit availability. Processing such information poses a significant challenge, as misinterpretation could lead to incorrect decisions that compromise unit condition, ultimately impacting power system operation. Considering the above-mentioned, in this paper, a methodology for asset management on PT fleets is proposed. The first step of the methodology involves determining the technical condition of individual equipment and its strategic importance. PT units are then clustered according to related risk, and for each cluster, the status of the units and economic factors are analyzed to determine the optimal inspection strategy for condition-based maintenance. The proposed methodology was applied to a case study of a fleet of fifty-eight PT. The obtained results demonstrate the suitability of the methodology for practical application. Specifically, the methodology enables the asset manager to apply condition-based maintenance strategies effectively to PT fleets.