Nondestructive Detection of Smallest Defects Inside Power Cable Insulation Systems

Since the partial-discharge (PD) testing during the routine test of the cable is not able to safely detect defects in the range of 0.2mm, other testing procedures are necessary for detection. So, the hot-oil-test is performed to exclude those small defects, which can lead to breakdown of the cable after indefinite time. Unfortunately, the hot-oil-test gives only response about a short part of the produced cable length. Moreover, the measurement process is quite dependent on the experience of the employee. Hence there is a large interest to find new - in best case nondestructive - methods to find and locate small defects inside cables insulation. Former investigations focusing on ultrasonic monitoring of the cross-linking of insulation material in power cables led to the question of the minimal detectable defect size, which can be detected with ultrasonic technology. Therefore, a first study was carried out to measure defects in extruded insulation, which normally are detected by the hot-oil-test. For this article bottom holes of specified sizes down to 0.1mm were made from the inside of a medium voltage (MV) and a high voltage (HV) cross-linked poly ethylene (XLPE) cable core. Afterwards the scanning of the sample is performed. An automated micro positioning system is used, so position and depth of the defects inside the insulation layer can be exactly measured. The results of the measurements are post processed to determine the minimal detectable defect size. After analyzing the signal quality and using of different frequency filters to reduce the noise, the spectrum of the inner semicon layer (ISL) shows a valid graph with a peak at 2.1 MHz. The visualization of the defects is done with the time signal of the ultrasonic impulse (A-scan), the depth map along one axis (B-scan) and the maximum amplitude map (C-scan). The evaluation of the data showed that defects larger than 0.4 mm are clearly visible in the C-scan, which is the fastest way of visualization. Defects down to 0.1 mm are visible characteristically in the B-scan. Furthermore the theoretical minimal detectable defect size of an poly propylene (PP) flat sample is calculated.