Capacity Reduction of Distribution Transformer by Harmonic Effect

This article actual picks up the load, the harmonics and the temperature with distribution transformer of the high voltage commercial building by two underground 22.8 kV feeders supplied. Train the ANN using the MATLAB training data to form the ANN until converged. Construct the recall sets by total voltage harmonic distortion \( V_{thd} \) and total current harmonic distortion \( I_{thd} \) and so on. That investigates the \( V_{thd} \) or \( I_{thd} \) mapping to the core temperature of the distribution transformer to fit by artificial neural networks and two order polynomials to reach load capacity reduction rate calculation. It is setting \( I_{thd} \) is 6 P.U., \( V_{thd} \) is 1 P.U, Q is 0.4 P.U, P is 0.95 P.U for first case, its capacity reduction rate is 26.40%. Second case, P is 0.9 P.U, the capacity reduction rate is 26.62%. Third case, P is 0.85 P.U, its reduction rate is 24.65%, and the forth-case P is 0.8 P.U, its reduction rate is 23.08%.