Off-Line Diagnostic and Withstand Testing

Guidance for developing a medium-voltage cable system aging management program for power plants is presented in the EPRI report Aging Management Program Guidance for Medium-Voltage Cable Systems for Nuclear Power Plants (1020805) [5]. The report describes attributes of cable assessment and testing and provides guidance on those tests most applicable to plant cables. A cable testing regimen should be chosen based on the particular cable design, the adverse environment factors, and operating history.

8.6.2.1 Dissipation Factor (Tan δ) Testing

Tan δ testing is a global test that evaluates the ratio of the resistive current divided by the

capacitive current in the insulation layer (see Figure 8-9). In 1981, Bahder et al. were among the first to report on this technique [84]. Later, Bach et al. and Hvidsten et al. reported a correlation between the decreasing ac breakdown level at power frequency and increasing dissipation factor at 0.1 Hz [85, 86]. Baur claims a strong correlation between the off-line 0.1 Hz tan δ value and the amount of water tree damage of the cable insulation [87]. The measurement of the 0.1 Hz tan δ provides a cable-aging assessment method that differentiates between good, defective, and highly deteriorated cable insulation.

Figure 8-9

Derivation of Dissipation Factor (Tan δ) Measurement in Insulation

As with other testing methods, tan δ testing can be performed using line frequency, VLF, or variable-frequency methods. EPRI report 1020805 recommends the VLF method and provides assessment criteria for this method [5].

VLF tan δ measurements are desirable because small portable test sets can be used. VLF tan δ measurements are made at several predetermined voltage level steps, starting at 0.5 V0 and proceeding in 0.5 V0 increments up to a maximum of 2.0 V0, provided no unacceptably high values are encountered. Preprogrammed equipment can be used or test operators can change the voltages at specific times. In a 0.1-Hz test, one cycle takes 10 seconds. Accordingly, the test duration must be long enough to take sufficient measurements to give valid results. Step durations vary from 30 seconds to 5 minutes. The test equipment automatically calculates and records the average and standard deviation of the readings. Although the absolute value of tan δ

Testing: Manufacturing, Installation, and Maintenance or In-Service Tests

provides some indication of aging, the difference in the tan δ readings at 2 V0 and 1 V0, the degree of degradation, is considered a stronger indicator. Table 8-4 provides criteria for XLPE insulation as being good, aged, or highly deteriorated. EPRI report 1020805 provides assessment criteria for EPR and butyl insulations [5].

Table 8-4

IEEE Standard 400 Criteria for Assessment for Cross-Linked Polyethylene Insulated Cables

Tan δ at 2V0

10^-3 Tan δ Increment at 2 V0 vs. V0

10^-3 Assessment

< 1.2 < 0.6 Good cable

≥ 1.2 ≥ 0.6 Aged cable

≥ 2.2 ≥ 1.0 Highly degraded cable

Figure 8-10 shows a typical portable test device that is suitable for testing the relatively short cables in power plants.

Figure 8-10

Typical Variable-Frequency, Very-Low-Frequency Portable Test Equipment for Performing 0.1-Hz Dissipation Factor (Tan δ) Testing

Courtesy of HV Diagnostics, Inc.

Testing: Manufacturing, Installation, and Maintenance or In-Service Tests

Figure 8-11 shows the voltage dependence of the dissipation factor at 0.1 Hz for new and service-aged XLPE-insulated cables.

Figure 8-11

Voltage Dependence of Dissipation Factor for New and Aged Cross-Linked Polyethylene Cable

Another way of viewing the data is shown in Figure 8-12. Here the increase of dissipation factor with voltage stress is shown as a bar (the color code is the same as that used in Figure 8-11).

Figure 8-12

0.1-Hz Dissipation Factor of Cross-Linked Polyethylene-Insulated Cables

Testing: Manufacturing, Installation, and Maintenance or In-Service Tests

This test method is global and, therefore, determines the general condition of the cable under test rather than identifying and localizing weakened sites. One drawback is that numerous small water-aging degradation sites can respond in the same manner as one or a few severe (long) water-aging degradation sites, with the latter being the condition of most importance. It is believed that the severe degradation sites will cause the differential measurement between V0

and 2 V0 to be larger, rather than just causing the overall measurement to be larger but stable with increasing voltage. Advantages of 0.1-Hz dissipation factor testing include the following:

 The test identifies the existence of water trees and water-related degradation before the point of conversion to electrical trees.

 Test equipment is portable and does not require the use of a van.

Disadvantages of 0.1 Hz dissipation factor testing include the following:

 Because an overvoltage is applied, a failure of a severely aged insulation system is possible during testing.

 Because tan δ is an off-line test with elevated voltage, the cables must be disconnected from their loads (such as motors and transformers).

 Testing of dissimilar, interconnected cable types, such as XLPE and EPR, will require separation to determine the state of both insulations correctly. In this case, normal EPR results could mask problems in the XLPE segment.

8.6.2.2 Dielectric Spectroscopy

Measurements of capacitance and dielectric losses at power frequency have been used for many years as a method of characterizing cable insulation systems. More recently, measurements over a range of frequencies have been studied. Frequency domain dielectric spectroscopy reveals information about the degree of degradation (water treeing) of aged XLPE cable systems;

although cable insulation materials should ideally have an infinite insulation resistance, in practice they demonstrate small conduction currents. As the insulation ages with time, the conduction current tends to increase as a result of aging-induced changes, including oxidation.

These changes affect results at different frequencies, depending on the severity of aging.

This test method involves measuring dissipation factor over a range of frequencies at several voltage levels. At each voltage level, “swept” frequency measurements are performed at frequencies such as 0.1, 0.2, 0.5, and 1.0 Hz. This was the procedure used in the EPRI report Advanced Diagnostics: Estimation of Life of Extruded Cables (1001727) [88]. The development team for this technology [89] measured in the frequency domain from 0.0002 Hz to 100 Hz (at 20-kV peak).

Testing: Manufacturing, Installation, and Maintenance or In-Service Tests

The data developed have been characterized in several manners in terms of dielectric response of cables without water trees, and several categories for XLPE with water trees XLPE [89], as follows:

 Low-loss linear permittivity. Characterized by an almost frequency-independent

capacitance. A linear response across a wide frequency band indicates that little or no aging has occurred.

 Voltage-dependent permittivity. Characterized by increases in both capacitance and dissipation factor with increasing voltage, but essentially, the increase is independent of frequency. This response is characteristic of cables in which water tree deterioration is significant, but the trees have not penetrated the insulation wall.

 Transition to leakage current. At low voltage levels, the response is similar to the voltage-dependent permittivity response, but at higher voltage levels, the dissipation factor losses increase with voltage and higher leakage current occurs at both high and low voltages.

Numerous dielectric spectroscopy curves are presented by Werelius et al. [89].

EPRI report 1001727 [88] refers to responses defined as different types:

 Type 1. A voltage-dependent increase of dissipation factor and capacitance, almost