Evaluation of Errors of the Wattmeter Method for Determining Magnetic Losses in Transformers

We examine the problem of the influence of the accuracy of dividing magnetic losses into components in transformers on the efficiency of minimizing these losses. Various methods of studying and evaluating the magnetic properties of electrical steels and the power of magnetic losses are described. It is shown that the use of the wattmeter method for determining the components of magnetic losses in a specific transformer, in contrast to the currently used methods for determining the power of specific magnetic losses, does not require special equipment and software. An assessment of two specific errors of the wattmeter method for determining the magnetic losses in the transformer core is carried out. The first error is associated with the errors of the measuring devices (wattmeter and frequency meter); the second concerns electrical losses during a no-load experiment. The dependences of these electrical loss errors on the accuracy class of the devices are obtained. It was found that for a dry single-phase transformer, the error of hysteresis and eddy current losses is 21 and 35 times higher, respectively, than the device accuracy class. For low-power transformers (up to 5 kVA), taking into account electrical losses in the no-load experiment underestimates the losses due to hysteresis and overestimates the losses due to eddy currents, which can significantly exceed the total error in determining the magnetic losses. It is shown that for transformers with power exceeding 25 kVA, the error which includes electrical losses in the no-load test is negligible.