R. Ross

Composition and growth of water trees in XLPE

On the basis of experimental results, a model describing the composition and the structure of water trees is derived. It is concluded that water trees are in fact salt trees, i.e., they consist of tracks of hydrophilic salt groups. These tracks most likely run through the amorphous phase of polyethylene. In water trees in cables, the salts are carboxylate salts and/or inorganic salts. In needle test objects the concentration of the inorganic salts is much higher. The differences in chemical composition appear to correlate with the aging conditions. The degradation products in water trees reflect the processes which took place during aging. In this way the occurrence of oxidation, ion diffusion, and chain scission is confirmed. A water tree growth hypothesis is formulated, which incorporates an electrochemical and an electrophysical effect. The first effect is very important in service-like aging, whereas the latter predominates in needle tests. >

Graphical methods for plotting and evaluating Weibull distributed data

Aiming at making Weibull plots, three types of plotting positions are reviewed: median probability, mean probability and mean plotting position. It is important whether the (log)data are regarded as median or as mean values. Median type plotting gives a biased plot on the average and, if used to identify outliers, can be somewhat misleading at the low end of the curve. In contrast, the mean type gives unbiased estimators by linear regression estimation, thus granting the best Weibull plot on the average. For use of Weibull paper in combination with a calculator, adequate formulas approximating probabilities are discussed. A somewhat improved formula for expected plotting position is given.<<ETX>>

Dielectric properties of water trees

The present study looks into the dielectric behavior of water trees. Wet water trees have much higher conductivities than dry water trees and than PE. This is assigned to the dependence of ion mobility on the presence of water. Voltage dependent effects are found. These results are related to two important issues. Firstly, a voltage dependent conductivity explains why DC-testing of water trees insulation is dangerous. Secondly, the voltage dependent ion conductivity of water trees opens gates to successful diagnostics for water tree detection in cables.

The chemical and physical composition of water trees in XLPE

The arguments supporting a model of the composition of water trees are updated with additional information concerning inorganic salts in water trees. It is pointed out that water trees are in fact salt trees. Hydrophilic tracks (rather than channels) run through the amorphous phase of polyethylene. The hydrophilic groups are carboxylate salts chemically bonded to the polymer matrix and trapped (on the removal of the electrical field) inorganic salts. The polymer matrix may still be largely intact, but (severe) mechanical damage is possible.<<ETX>>

Water tree growth processes in XLPE

Water trees as degradation products contain information about the aging process (es). The likelihood of oxidation, ion diffusion, and chain scission occurring during water tree growth in polyethylene insulation is discussed. The possible roles of the three processes are discussed. They cooperate in water tree growth, but may compete with respect to reaction products. It is also pointed out that antioxidants may play a dual role if they can be easily oxidized to ionic species. It is concluded that, with testing, care must be taken to accelerate the right process, because many parameters influence the growth rate of water trees.<<ETX>>

Aging of interfaces by discharging

The paper gives an outline of ongoing work on interface reliability. It gives an outline of progress made within Cigre as well as at KEMA. The sample cell is found to perform quite well. Pressure is found to compensate for a certain roughness, which is in agreement with common conviction. However, the presence of a lubricant is not always found to improve the performance of the interface. One of the reasons can be that the reference interfaces are already very good. On the other hand, there are more publications reporting that lubricant does not always improve an interface.

Pressure effects and moisture penetration in mass-impregnated cable

Mass-impregnated cable or PILC (paper insulated lead cable) exhibits pressure effects under load variations. The pressure build-up in a cable is related to materials and design of the cable as well as to temperature and load. As one of the results, in some cables moisture can travel easily. This can cause early degeneration of the cable insulation and accessories. It thus enhances the risk of failures. The present paper describes a part of a larger research project that aimed at assessing the pressure effects due to load cycling on full size cables and the relation to moisture ingress. The experiment included a full short circuit test on five cables.

The relation between expected life estimation and amount of data

Small data sets from e.g. limited tests yield (far) too optimistic impressions of product reliability. This paper explains the background and illustrates this effect numerically.

PD diagnostics on cables and terminations for CBM

Condition Based Maintenance (CBM) aims at maintenance as required by the actual condition of networks. This paper discusses the background and the implementation of CBM. It largely follows the line of a paper on CBM presented at ISH this year. A number of practical results are given to demonstrate and illustrate CBM in projects that are carried out by KEMA.

Recent developments on Weibull statistics

The paper describes a selection of matters that have been discussed during recent years: parameter estimation and plotting. The paper discusses the relevance of applying corrections and countermeasures by evaluating the effects that occur with usual inferences. It is recommended, for cases with small test sample sets, to apply the available correction techniques.