New method for determination of steady-state temperature rises of transformers

For transformers whose thermal time constant exceeds two hours, unacceptably large systematic errors occur if recommended methods for determination of steady-state temperature rise are used. Therefore a new criterion for heat run duration and a new method for determination of steady-state temperature rise are proposed, which ensure reliable determination of steady-state temperature rise independent of transformer thermal time constant. According to this proposal, the heat run should be continued until the temperature rise exceeds 95% of the steady-state temperature rise, which is estimated by a three-points method. Then, by the method of least squares, a more accurate value of steady-state temperature rise is obtained by extrapolating only those measured points that are higher than 60% of the previously estimated steady-state temperature rise. For those instrument transformers whose largest thermal time constant of the system (transformer) is governed by the time constant of the winding, a universal intermittent heat run independent of the transformer time constant is proposed: ton/toff = 120 min/2 min. Estimation errors of the steady-state temperature rise and thermal time constant when applying the three-points method and the intermittent heat run are analysed.