Abstract
This paper describes the method used at the Electrical Testing Laboratories for measuring the dielectric power losses in 10-foot samples of three-conductor cables with three-phase potential applied to the cable. The difficulties encountered and the methods employed to overcome them are discussed in considerable detail. Typical results are given in the form of data for two specimens of cable, one having a low power loss in the dielectric and one having a high power loss in the dielectric. The data are also presented in the form of curves. The discussion includes: (a) The theory of excessive internal dielectric loss as accounting for cable failures at local ``hot spots.'' (b) The advantages of plotting data with logarithmic scales. (c) A comparison of results obtained by computation from single-phase measurements with those obtained by direct measurement with three-phase potential. The conclusions drawn are: (a) The power loss in the dielectric in a three-conductor cable under actual three-phase conditions can be readily measured in the laboratory with specimens ten feet long. (b) No special apparatus is necessary for such measurements other than a reflecting high sensitivity wattmeter. (c) Apparently the power loss in the dielectric cannot in all cases be accurately calculated from data obtained in single-phase tests although it is highly probable that for all practical purposes the discrepancy would not be serious.

This publication has 2 references indexed in Scilit: