TRANSFORMERS 449 



between 500 Kv.A. and 1000 Kv.A., to forty-eight hours; be- 

 tween 1000 Kv.A. and 2000 Kv.A. to sixty hours; for all larger 

 capacities the process should be carried on for at least seventy-two 

 hours. In case there is no evidence that the transformer is unduly 

 moist, discretion may be used in slightly decreasing the limits 

 given for the voltage. A transformer of 20,000 to 30,000 volts, 

 for instance, having a capacity of 200 Kv.A. or less, may be dried 

 in only twenty-four hours. The limits given for the capacities, 

 however, should be rigidly adhered to, and in no case should the 

 process be carried on for less than twenty-four hours. 



While the insulation resistance of a transformer cannot be 

 relied upon as a sure indication of its condition at any one time, the 

 general trend of megger readings as a drying run proceeds is a 

 fairly accurate indication of the progress of drying. The drying 

 process should be continued until the curve becomes approxi- 

 mately flat at an elevation considerably above the low point of 

 the curve. Variation in temperatures causes wide variation in 

 resistance, the values varying inversely. If the megger shows a 

 short circuit, that is, an insulation resistance too low to be read, 

 it is very likely due to an excessive amount of moisture. Low 

 readings also sometimes indicate the presence of moist spots in the 

 insulation. Widely different megger readings may be obtained 

 on different transformers, but average readings should be approx- 

 imately alike for transformers of the same capacity and design. 

 Shell-type transformers have, in general, a lower insulation resist- 

 ance than core-type. 



Oil Drying. Oil, whether shipped in sealed barrels or in 

 special tank cars direct from the manufacturer, may require 

 drying at its destination before it is suitable for use in high-volt- 

 age transformers. All oil should be tested before using, but, if 

 it is absolutely necessary to use a part of oil from barrels before 

 tests can be made, the barrels should be allowed to settle for sev- 

 eral hours and then the oil pumped from the top to within 4 inches 

 of the bottom; i.e., do not use the oil which settles in the bottom 

 until it can be tested and dried if necessary. Oil drums should be 

 stored lying on their sides. 



The best method for drying and filtering oil consists of forcing 

 it under pressure through several layers of blotting paper, which 

 removes all moisture and solid matter held in suspension in the oil. 

 A filter press, such as shown in Fig. 283, has been developed for 



