ABSTRACT
By means of measurements on a small core type oil filled transformer the possibility of the following methods of determining winding temperature have been examined: (a) Use of lumped constant coefficients for steady—state prediction; (b) Use of the superposition of open circuit and short circuit temperaturerises to give the rises for the corresponding load test.It is seen that (a) gives reasonable results but that (b) gives higher temperatures than those measured on the load test. Finally the assumption of all heat absorbed in the windings over a short circuit is seen to be valid.