High Voltage Testing and Partial Discharge In GIS
Overview
Insulation integrity testing in GIS is realized using Power Frequency over voltage withstand and Partial Discharge(PD) measurements. PD testing has become a widely accepted method for insulation diagnostics and a required part of the acceptance testing for most HV equipment.
GIS insulation integrity can be affected by the following:
- Protrusions or fixed particles
- Moving particles
- Voids in insulators
- Poor contact
- Loss of insulation integrity due to loss of Gas pressure/density


High Voltage Connection
Application of HV potential to the GIS bus is done in three fashions:
- Air connection to the GIS bushing
- Direct connection using SF6 bus
- Induced voltage through secondary of GIS VT
Test Power Requirements
Test power requirements for external energization of the GIS is derived from:
S = P + Q = I2R + I2X,
where X = XC -XL, R = resistance of the primary circuit and XC = capacitive reactance and XL = inductive reactance.
It can be seen that in order to reduce the test power needs, it is best to minimize or eliminate the reactive power. To accomplish this, most test supplies used on GIS are of “resonance” type. Meaning they either introduce inductance to counter the capacitance of the circuit or they vary the frequency to match the inductance and capacitance reactances.
GIS Test Transformers
GIS Test Transformers typically either use oil or SF6 as their insulating medium. Oil transformers have significantly higher power outputs and tuning ranges, however they cannot be connected directly to the GIS bus and must utilize external air connections. SF6 transformers with direct SF6 bus to the test object are the preferred method of connection for PD measurements.



Test Levels
The application of test voltage and durations is defined by the GIS standards, mainly IEC 62271 and IEEE C22.122. It should be noted that despite being harmonized, the two standards differ in terms of requirements for field Partial Discharge testing. IEEE has eliminated this requirement, likely due to the difficulty of execution and lower level of industry awareness in the region. Since Partial Discharge testing at site is crucial it is often advised to follow the more stringent of requirements where and when possible. Impulse testing can be utilized and substituted for PD testing however it is not a common practice.

Partial Discharge Measurements
Partial Discharge(PD) is a mini break down in/around the insulation, which speaks directly to the quality of the insulation system. PD testing has long been covered under standard IEC60270. This method is a well-established technique in which PD is quantified in terms of ‘apparent charge’. In this method, the test object is connected in parallel to a coupling capacitor. Using a coupling device (quadripole) that converts input current to output voltage, the instruments measure apparent charge. The obtained PD patterns, inception and extinction voltages are key to differentiating the PD type. This type of PD analysis is by far the most common and is often a mandatory requirement for factory testing and site commissioning of most types of switchgear. The IEC 60270 setup is relatively easy to be implemented in the factory where test object capacitances are limited and shielding can easily be achieved. However in field conditions, and depending on the voltage class of the GIS as well as presence of bushings or large bay lineups, conventional PD measurements may not be practical. Hence an exception to deviate form the standard should be considered. In these cases Ultra High Frequency (UHF) PD measurements which are sometimes referred to as non-conventional PD measurements, are utilized. UHF PD is “non-coupled” which means it is detected via electromagnetic emissions. The inherent downfall principles of radiation are accuracy and decay, hence the UHF measurement acceptance criteria is a contentious topic. The rule of thumb is that all observed anomalies with UHF or conventional measurements should be investigated.



UHF Vs Conventional PD measurement
A simplified comparison of conventional and UHF PD measurement techniques is given below:
Since most of the components and bay assemblies undergo PD testing the factory, site PD testing is more targeted to identifying workmanship related issues during assembly. By far the most common issues observed during PD testing are free moving metallic particles. Protrusions, floating/ungrounded objects and poor contact are other frequency anomalies
Directly Coupled IEC 60270 PD Measurement | UHF PD Measurement |
Simpler test setup | High accuracy of detecting and localizing defects |
Single point of measurement | Can be used on systems with high background noise and air connections |
Clear pass/fail criteria | UHF sensors can be used for permanent or periodic online monitoring |
Hard to implement on large or complex circuits | Requires many integrated ( higher accuracy) or retrofit(lower accuracy) sensors and may have blind zones of measurement if sensor count is not sufficient |
Hard to implement on air connection HV setups due to high noise | Unclear pass/fail criteria since it is difficult to correlate to pC level |
Requires a PD free test transformer and connection | Not all systems can produce PRPDs |
Acceptance Levels
PD acceptance test levels, are defined in IEC62271. It is normal to maintain a brief conditioning period to burn/move any free moving particles and/or condition or stabilize insulation PD. After this period, voltage is increased to the withstand level for a duration of one minute, after which the voltage is lowered to the PD acceptance voltage. The PD inception and extinction voltages are often monitored and the rate of rise and fall of voltage follows the IEC60270 guidelines. Minimum accepted PD extinction voltage is often 1.2x nominal phase to ground voltage and a sensitivity of 5pC or better is expected. Apparent charge pC values are calibrated against test object capacitance using a pulse injector.


Interpretation of PD patterns
Many studies have been done on the interpretation of Partial Discharge readings, their correlation with time under voltage and their superposition placement on the applied AC waveform. The physics of electric field and electron flow allow for good understanding of the patterns produced under different geometrical conditions, hence by evaluating the Phase Resolved PD (PRPD) patterns, an experienced test engineer can often identify the source, type and severity of the observed anomaly.
