Aging State Assessment of Transformer Oil Based on UV-Vis Spectrum and Fuzzy Algorithm

Insulation oil is used as insulation and cooling medium in transformers. With the operation of the transformers, insulation oil gradually ages, leading to decline in insulation strength, blockage in oil channels, and difficulty in heat dissipation, which brings potential risks to the equipment and power system. Therefore, timely and accurate aging state assessment of insulation oil is significant to guide the timing of oil filtering or oil replacement. In this paper, the degradation laws of mineral oil and 3-element mixed insulation oil were studied through accelerated thermal aging tests, and the parameters indicating aging state of oil were measured. Test results showed that the characteristic value of UV-Vis (ultraviolet-visible) spectrum was closely related to the conventional aging parameters, such as acid value, interfacial tension, and resistivity. A mathematical model was established between the characteristic value and conventional aging parameters of insulation oil, which showed high accuracy verified by oil-replacement aging test. Finally, an aging state evaluation model of insulation oil was established with fuzzy algorithm based on the abovementioned mathematical model, which is expected to be used for online monitoring of oil aging state.