Abstract This paper provides information on aging of URD cable insulated with tree-retardantcrosslinked polyethylene (TR-XILPE)compound, installed in a typical utility environment. Numerous evaluations were performed on samples of the cables aged up to 7 years in-service. AC and impulse voltage breakdown data are compared with data for similar 35 kV ethylene propylene rubber (EPR) and crosslinked polyethylene (XLPE) insulated cables removed from the same ultility system. The data show that, to date, the rate of degradation of TR-XLPE cables is less than that of the EPR and XliPE insulated cables. Extrapolation of the data, assuming lhe same rate, indicates TR-XLPEcable will have the longest life. I. INTRODUCTION Beginning in 1985 Houston Lighting & Power Co. (HL&P), undertook a comparative study to establish the rate at which cables with diffeIent extruded insulations age. The results of the laboratory evaluations were published in [l]. As a consequence of these evaluations, HL&P adopted the use of TR-XLPE as the preferred insulation for their Underground Residential Distribution (URD) cabl!e systems. TR-XLPE is a modified XLPE insulation and is claimed to provide an improved performance with respect to water treeing. EPR insulated cable has and continues to be used in the HL&P feeder system [2] [3]. The present paper deals with the rate of aging during the first seven years of operation of the TR-XLPE cable. For comparison, voltage breakdown curves also are provided for EPR and XLPE cables, installed in the same systems. TR-XLPE has been in commercial us,e since 1983. HL&P has been specifying this type of insulation for their URD cables since 1987. 1370 km (850 miles) of TR-XLPE cables have been installed from 1987 to 1995 without any premature failure to daБайду номын сангаасe. Approximately 920 km (570 miles) of the TR-XLPE installed was due to a 35 kV URD cable replacement program. A number of papers on experience with TR-XLPE cables during laboratory aging and testing have been published [4,5, 61. This is the first comprehensive report on in-service performance. The aim is to characterize changes in properties that impact field performance during early service years, and to attempt, by extrapolation, to predict the expected life of these cables.
Michael Walker - Member IEEE Houston Lighting & Power Company Houston, Texas
11. CHARACTERISTICS OF THE SERVICE ENVIRONMENT All cables had been installed under similar conditions in individual 0.15 m diameter polyvinyl chloride (PVC) ducts, at a depth of 1.2 m. The ducts were partially filled with water during the time the cables were in service. The load on these cables is relatively low; therefore, it is fair to assume they operated at relatively low temperatures. The service area is basically flat and located at about 10 to 20 m above sea level. The Houston metropolitan area has a very high lightning (isokereunic) level. MOV riser pole arresters are installed at the open point of the URD loop to limit the effects of voltage impulses due to lightning. The average temperature at the underground level where the cables are installed is approximately 10°C with an average year round ambient mean temperature of approximately 28°C. 111. CABLES EVALUATED Six (6) in-service aged 35 kV cables, having similar construction (Table 1)were evaluated. These cables were aged but had not failed during service. The cables were made by four manufacturers during the period 1987 to 1989. Except for one sixyear old cable, the conductor was 1/0 AWG (54 mm2), 19 strand compressed aluminum, extruded semiconducting copolymer XLPE conductor shield, 8.76 mm (nominal) TR-XLPE copolymer insulation, semiconducting copolymer XLPE insulation shield, 16 No. 14AWG (2.08 mm2)copper concentric wires and an insulating polyethylene (PE) jacket overall. The exception had a 2/0 AWG (67 mm2) conductor. About 90 to 120 m of each cable were made available for testing. All cables were made with HFDA 4202 TR-XLPE copolymer insulation compound. According to literature from Union Carbide, the compound manufacturer, the tree retardant characteristics are attributed to a permanent polymeric additive. Identification of the semiconducting shielding compounds is not available to the authors of this paper. It is known that all are crosslinked, with the insulation shield having strippable characteristics. The cables, in service for five to seven years, were identified by printings on their respective jackets. For comparison purposes, also data are given on a 1988 TR-XLPE cable (made by manufacturer A) that was tested after being in storage for four to six months. This cable had HFDA 4202 insulation and the semi-conducting shields had similar characteristics as the other six (6) cables. IV. EVALUATION OF THE CABLES The cables, as removed from service were placed on reels and transported to the testing laboratory. Upon arrival they were cut into approximately 9.2 m long lengths and immersed in water after removing their jackets and injecting water between the