Abstract
Transformers are key components in the power system and transformer failures can cause long power outages with high costs to society. Transformer failures are rare, and each case is unique with respect to its consequences. This shapes the data and statistics we have available to predict future failures and related consequences. Models to support risk assessments and asset management decisions for these critical assets should rely on practical approaches to include both available data as well as expert judgements. This paper looks at outage duration, an important parameter in risk evaluation and asset management decisions. It presents a transformer outage duration model which can be conditioned on relevant asset management input variables. A use case is constructed to exemplify the usage of the model in an asset management decision context.