Enhanced Thermodynamic Modeling of Converter Transformer Influenced by DC Bias

Abstract

DC bias in high-voltage DC transformers, arising from converter operations and DC transmission, poses significant challenges to their performance. The detrimental effects of DC bias primarily manifest in increased temperature, jeopardizing the safe operation of the transformers. This article presents a novel approach by extending the utilization of the Thermal Equivalent Circuit (TEC) to accurately predict temperatures at different elements of a converter transformer under DC bias conditions. Specifically designed for a 240 MVA converter transformer, the TEC incorporates capacitances and dynamic resistances as model parameters. Additionally, an electro-thermal finite element model is implemented to comprehensively analyze the transformer's behavior under varying levels of DC bias. To estimate the TEC parameters, a hybrid GWO-CS (Grey Wolf Optimization – Cuckoo Search) algorithm is employed based on measured values. Furthermore, the paper highlights the impact of DC bias on the converter transformer's life expectancy, considering the aging acceleration factor.