Probabilistic Management of Power Delivery Based on Dynamic Transformer Rating

The smart grid paradigm pushes for intelligent operation of the transformers that are already installed in the networks, to cope with peak load and/or to enable more intense exploitation of renewables. However, transformer loading is affected by several factors that should be considered, among which the thermal stress is recognized as the most influencing one. In this context, the dynamic thermal rating concept is of great interest and it allows fixing the maximum allowable current in different operating conditions, still maintaining acceptable risk levels based on the consequences of loading the transformers beyond the nameplate ratings. A probabilistic procedure for managing the delivery of power to load by the dynamic transformer rating is presented in this paper. The procedure is based on the risk analysis related to the thermal stress introduced by the transformer (over)load, which determines loss of life and potential dielectric failure. Numerical experiments based on actual data are performed for several scenarios, and several cases are presented to support the procedure for the intelligent exploitation of the potentialities of transformers.