An Integrated Fuzzy based Online Monitoring System for Health Index and Remnant Life Computation of 33 kV Steel Mill Transformer

Electric arc furnace transformers are of prime importance in steel mill transformers that are designed inherently to supply high magnitude currents. Traditional maintenance strategy for steel mill transformers involves performing an offline diagnostic test and computing health index and remnant life using mathematical aggregations. Stresses in a transformer can be due to an individual or multiple factors which may lead to its deterioration. In order to avoid premature failure, computation of health index and remnant life by online monitoring of arc furnace transformers is imperative. Based on the optimum selection of single and multi-parameter sensors, an architecture for online monitoring is discussed in this paper. Subsequently, dedicated fuzzy models for computation of criticalities of various components, the overall health index, and remnant life are developed in this research. Also, the decision logic is established to identify faulty components based on information collected by online sensors. Weighted fuzzy rules and color code for health indication are some of the additives which enhance the significance of the model. This research will be helpful for maintenance engineers and researchers involved with the maintenance and overhaul of furnace transformers.