Fast demagnetization method for power transformers combined with residual flux measurement

• A fast demagnetization method for power transformers combined with residual flux measurement is proposed. • The relationship between residual flux density and measurable parameter permeability is established through experiments. • The Jiles-Atherton model with variable parameters is used to characterize the magnetic characteristics of transformer cores, improving simulation accuracy. • A numerical simulation model of transformers has been developed, which can simulate transformers under different residual flux conditions. When power transformers are re-energized, the inrush current may be generated and damage electrical equipment. In order to suppress the generation of serious inrush current, the residual flux (RF) in the power transformer core needs to be eliminated. However, existing methods require first magnetizing the core to saturation to calibrate the magnetic flux due to the unknown residual flux density ( B r ), which increases demagnetization power and time. In addition, the lack of effective B r measurement methods makes it impossible to evaluate the demagnetization effect. Therefore, this paper proposes a demagnetization method that considers the amount of B r , which achieves accurate measurement and quantitative elimination of B r in the transformer core. The measurement is first performed by applying DC voltages of different polarities, and then a specific demagnetization DC voltage is applied based on the measurement results. The magnetic flux in the transformer core is reduced directly to zero without calibrating flux throughout the entire process. Experimental studies are performed on core-type transformers and 10 kV, 250 kVA power transformers, and the experiment results show that the B r is reduced to within 1 % of the knee-point flux and the measurement error is within 4 %, which proves the effectiveness of the proposed method. Compared with existing methods, the demagnetization time can be reduced to less than 1 s and the demagnetization power can be reduced by 95 % using the proposed method.