Abstract :
This report deals with a two years investigation in Gothenburg, where a large part of the 11 kV paper insulated cables have been tested by the use of PD-mapping, a method to measure the partial discharges in the cables to size and location. The method provides a good picture of the ongoing processes in the cable, and may be of large value when making a decision about the replacement of a cable length. But the possibility to prevent faults by the use of PD-mapping is much more difficult. Such a possibility must be based on the assumption, that a fault in a joint, in a termination or in the cable itself is foregone by partial discharges, which will continuously increase until the cable insulation is completely broken down, and a fault will occur. The whole process must also take time enough, to provide possibility for the future fault to be located beforehand at a diagnostic control. Our own experience from measurements during a two years period have not confirmed that these conditions are at hand. In contrast, some faults have occurred on cables where no partial discharge activity at all could be seen in advance at the location where the fault later happened. In other cables, however, it has been possible to see partial discharges in joints, which have later failed. The levels of partial discharges have in these cases not been exceptionally high, but they have been numerous and located in the same place, often in all the three phases. It is our opinion, that more basic research must be done, in order to learn how a fault develops in paper insulated cables, and how it should be diagnosed in advance
