Partial discharge tests

There are several methods to determine and compare the resistance to partial discharges. Some tests are done on finished cables, such as the U-bend test, and others are done on laboratory samples molded from the insulation, that are subjected to partial discharges created by sharp objects, such as needles under high voltages. The tests compare either the energy required or the length of time required to erode or fail (short circuit) samples of similar thickness. 

Accelerated tests usually require the application of a higher electrical stress (voltage) coupled sometimes with an increase in temperature as a second accelerating factor. As in all accelerated testing, care must be taken to ensure that this does not introduce ageing mechanisms different to those that occur in service. The most common relation used to predict insulation life in the presence of partial discharges is the power law ... 

So, partial discharge measurements have become a well-established method for type and routine testing as a non-destructive material test and quality control [23], [26], [53]. Apart from optical and acoustical detection methods of partial discharges, the high frequency detection and measuring method preferably covers a wide range of applications. It will be considered in more detail in the following. 

Part 3 Test methods for partial discharge measurement on lengths of extruded power cable. 

IEC61294 Insulating liquids - determination of the partial discharge inception voltage (PDIV) Test procedure. 

In addition to the dielectric-strength test at power frequency, which remains the basic test, other criteria have been adopted, such as impulse breakdown voltage, partial-discharge voltage and electrical stability under partial discharges as characterized by gassing. 

Figure 2. Test cell for measuring the partial-discharge threshold voltage.

Haidara and Denat (1991) investigated the onset of partial discharge pulses in a test cell with 50 D impedance consisting of a fine tip as cathode and a plane metal disk as anode. By comparing the threshold voltage in the vapor and in the liquid they were able to estimate values for a in the liquid phase as a function of applied field. Their data and for comparison the ionization rates reported for silicon are included in Figure 1. A crude estimation of a in liquid n-hexane obtained from time lag to breakdown measurements with thin liquid layers by Arii et al. (1979) is also included. 

Test Method for Voltage Endurance of Solid Electrical Insulating Materials Subjected to Partial Discharges (Corona) on the... 

Pressure Development Overpressure in a UVCE results from turbulence that promotes a sudden release of energy. Tests in the open without obstacles or confining structures do not produce damaging overpressure. Nevertheless, combustion in a vapor cloud within a partially confined space or around turbulence-producing obstacles may generate damaging overpressure. Also, turbulence in a jet release, such as may occur with compressed natural gas discharged from a ruptured pipehne, may result in blast pressure. 

The test gas, arriving at the measuring electrode (cathode) either by diffusion or by pumping, is electrochemically converted. The resulting ions pass the electrolyte and are discharged at the anode the measurable voltage is proportional to the partial pressure of the test gas. 

A major fallacy is made when observations obeying a known physical law are subjected to trend-oriented tests, but without allowing for a specific behaviour predicted by the law in certain sub-domains of the observation set. This can be seen in Table 11 where a partial set of classical cathode polarization data has been reconstructed from a current versus total polarization graph [28], If all data pairs were equally treated, rank distribution analysis would lead to an erroneous conclusion, inasmuch as the (admittedly short) limiting-current plateau for cupric ion discharge, albeit included in the data, would be ignored. Along this plateau, the independence of current from polarization potential follows directly from the theory of natural convection at a flat plate, with ample empirical support from electrochemical mass transport experiments. 

Fig. 12.22. Cell for partial H30+ discharge current (/H) measurements. A, Anodic compartment B, cathodic compartment C, capillary (diameter 0.3 mm) W, water jacket Fe, iron test electrodes E, reference electrode X, T, Z, Teflon taps G, taps for connection with Nichrome furnaces. (Reprinted from J. O M. Bockris, A. Despic, and D. M. Drazic, Electrochim Acta 32 p. 329, copyright 1961, Fig. 1 with permission from Elsevier Science).

---