The Basics of Partial Discharge Testing
Partial discharge (PD) is evidence of a degrading insulation system, which could lead to very costly repairs and can predictively lead to an electrical breakdown of high voltage apparatus. The phenomenon is of great practical interest in the electric power industry, as the presence and magnitude of PD are important criteria to measure for the early detection of degrading insulation quality and the assessment of manufactured, installed, or repaired product quality. One of the most difficult tasks is interpreting the PD data and determining the time an insulation system may remain in service before damaging and costly failure can occur. HV TECHNOLOGIES, Inc. has many years of experience in the field of PD testing and supplies equipment with different measurement technologies for different PD testing applications.
What is Partial Discharge?
When speaking of partial discharge, the most important standard that every expert will refer to is IEC 60270: High-voltage test techniques – Partial discharge measurements. This standard applies to the measurement of PD in electrical apparatus or systems when testing with AC voltage up to 400 Hz or with DC voltage. In this blog post we will focus on measuring PD by applying AC voltages.
IEC 60270 defines the term “partial discharge” as:
“a localized electrical discharge that only partially bridges the insulation between conductors and which can or can not occur adjacent to a conductor”.
In general PD occurs in areas of the insulation or on the surface of the insulation that is subjected to higher local electric field stress concentrations, such as in a cavity in a solid dielectric, a sharp protrusion, or in low density regions of a liquid dielectric. PD initiates once the electric field is high enough, or the mean-free path is short enough, to cause impact ionization to create a “streamer”, otherwise known as an electron avalanche. Furthermore, a free electron must be present in order to start the discharge process, which can be generated by cosmic rays or naturally occurring radiation. Due to this, PD initiation is subject to a statistical time lag. The figures below show some typical sources of PD in air (e.g., insulators, outer cable terminations), liquid (e.g., transformers), and solid insulation (e.g., cable systems).
Typical Partial Discharge Measurement Circuits
IEC 60270 recommends specific PD measuring circuits to ensure reproducible and comparable results while conducting partial discharge measurements. The figure below shows an overview of the 4 basic measuring circuits according to IEC 60270, where Z is the blocking impedance placed between the source and the test object to reduce high frequency interference from the voltage source and limit propagation of transient signals into the voltage source, Ca corresponds to the capacitance of the device under test (DUT), Ck is a coupling capacitor together with a measuring impedance Zm to decouple the PD signals and convert into voltage pulses for analysis with a partial discharge detector M.
The most common circuit for measuring PD is the top left in the above image, in which the measuring impedance Zm is in series with the coupling capacitor Ck, both of which are in parallel to the DUT. Detection sensitivity can be increased if the measuring impedance is placed in series with the ground of the DUT (top right). However, in this configuration the measuring impedance could be seriously damaged from an unexpected breakdown. Even higher measurement sensitivity can be achieved by utilizing a bridge impedance (lower left) by connecting to the low voltage side of the DUT and coupling capacitor. This set-up significantly helps to reduce external electromagnetic influences. In transformer testing, if capacitively-graded bushings are available, these can be used as the coupling capacitor for the PD measurement circuit (bottom right).
What Parameters are Typically Measured During a Partial Discharge Test?
Partial discharge pulses characteristically have rise times that are < 1µs in liquid dielectrics and < 1 ns in solid dielectrics. The pulse parameters not only depend on the medium of the insulation, such as type of gas, gas pressure, or size of the cavity according to Paschen’s Law, but also on the shape of the applied voltage. The most convenient parameter to measure the intensity of a PD pulse is its pulse charge, which is determined by the integration of current over time within the frequency bandwidth of the measuring system. The image below shows a typical PD pulse as viewed on an oscilloscope, in which the charge is determined by integrating the current signal going into a 50 Ω input.
When measuring PD via galvanic coupling to the terminals of the test object according to the circuits recommended in IEC 60270, high frequency attenuation and distortion of the PD pulse as the signal propagates from the source to the measurement location will cause the pulse peak, fall time, and pulse duration to vary over a wide range. However, since the signal is being integrated over time, this will always result in the same measured charge regardless of the shape of the pulse. Therefore, pulse charge is an extremely important and useful parameter to measure.
A typical wideband PD measurement according to IEC 60270 is characterized by a transfer impedance having fixed values of the lower and upper limit frequencies of 100 and 500 kHz, respectively (Note: an amendment to IEC 60270 in 2015 has allowed new frequency parameters to be 100 kHz and 1 MHz, respectively). In this frequency range a quasi-integration is conducted to determine the pulse charge.
Considering that a PD source is generally not directly accessible in HV apparatus, unless PD couplers are installed at time of manufacturing or “free space” PD detection (detection of the electromagnetic waves) is conducted, this means that the charge from the discharge cannot be measured directly. As a result, the PD measuring circuit needs to be “calibrated” in order to be able to measure apparent charge, which IEC defines as:
“Apparent charge q of a PD pulse is the charge which, if injected within a very short time between the terminals of the test object in a specified test circuit, would give the same reading on the measuring instrument as the PD current pulse itself. The apparent charge is usually expressed in picocoulombs (pC).”
Before any PD test is conducted with high voltage applied, pulses of known charge values must be fed into the test circuit for calibration of the circuit and for determination of the scale factor of the PD measuring system. When conducting this, the partial discharge calibrator should be as close as possible to the high voltage terminals of the DUT to reduce any possible errors due to stray capacitance.
Phase Resolved PD Patterns:
In an energized insulation system containing cavities, the voltage appearing across the cavity will follow the waveform of the applied voltage on the DUT. As was described above, if the electric field within the cavity is greater than the breakdown strength of the gas within the cavity and an initiating electron is present with sufficient energy to cause impact ionization, a streamer or electron avalanche will grow. The conductivity of the streamer will cause the electric field across the cavity to collapse and the streamer will stop growing once the field is below that required to sustain its growth. An increase in capacitance between the electrodes will be seen from the collapse of the electric field, which will result in a transient voltage drop between them. This sequence can also be portrayed using an equivalent circuit diagram, as shown below.
The voltage appearing across a cavity UC by applying an alternating voltage UA to the DUT is given by:
where CA is the capacitance of the DUT, CB is the capacitance of the healthy insulation in series to the cavity and CC is the capacitance of the gas within the cavity. This also allows the apparent charge qA to be measured in picocoulombs (pC), which is given by qA = CB*ΔUC = CA*ΔUA. Partial discharge signals occur in the nanosecond time scale so that the discharges appear as pulse like currents that can be superimposed over one cycle of the AC waveform (e.g. 16.7 ms for 60 Hz), also known as a phase resolved PD pattern.
The characteristic PD signatures can be a useful tool in determining what type of defect is present in the insulation system being tested. The pulses tend to appear in pulse trains with polarity similar to the applied AC voltage. In general, the number of discharges will increase in frequency with applied test voltage. The amplitude of the PD signals will tend to remain constant in cavities within solid insulation due to space limitations. On the other hand, PD amplitudes will increase with increasing applied voltage in liquids or during tracking discharges in air where there is less restricted volume of gas for the streamers to grow. Most discharges will superimpose on the zero-crossings of the AC sinewave (0-90° and 180-270°) due to the electric field enhancement caused by the switching of the AC voltage polarity at these locations. The figure below shows some typical PD patterns of void (cavity) or surface discharges.
Corona discharges, which IEC 60270 defines as “a form of partial discharge that occurs in gaseous media around conductors which are remote from solid or liquid insulation”, typically have discharges that occur at the maximum of the AC voltage sinewave. There is also a polarity effect that when a sharp protrusion is at high voltage potential, the discharges will superimpose on the maximum of the negative sinewave (270° phase). When the sharp protrusion is at ground potential, the discharges will superimpose on the maximum of the positive sinewave (90° phase).
Radio Interference Voltage (RIV):
RIV is a typical partial discharge measurement that is conducted on transformers and surge arrestors. The technique is based on the relatively outdated standard of NEMA 107-1987, which based the technique on measurement receivers to estimate the disturbance of communication lines. The measurement is similar to IEC 60270, yet the quasi-integration is conducted at a narrow bandwidth of 9 kHz with a tunable center frequency between 10 kHz to 10 MHz. The RIV value is given in “interference voltage” in units of μV, which is based upon the transferred charge value and the repetitive rate of the PD pulses.
The Different Types of PD Measurement Technologies
HVT offers a long line of different PD measurement technologies depending on the desired application. In general, the technologies are divided between “Conventional”, or electrical, and “Non-Conventional” detection methods.
Conventional PD Detection:
Everything that has been written in this blog post pertains to the conventional or electrical PD detection method according to IEC 60270. These methods pertain to PD testing while “offline”, which means the excitation voltage is provided via an external source.
We can provide you with accurate and reliable systems for PD measurements of HV equipment such as transformers, generators, switchgear and cable systems using applied voltage in laboratory or on-site testing environments.
The DDX 9121b is a highly sophisticated detector for PD and RIV testing. Traditional partial discharge measurements according to IEC 60270 or RIV measurement or PD under DC are covered, as well as PD site location (localization) in power cables. The DDX9121b increases the laboratory sensitivity as it is equipped with digital filters allowing the measurement frequency band to be shifted into a less noisy range and suppressing frequency dependent noise. Determining the exact location of PD signals along the length of cable with PD site location is also possible.
The DDX 9161 is the newest laboratory optimized PD detector from Haefely for AC, DC, and RIV partial discharge measurements. The fully digital state-of-the-art detector can accommodate up to four simultaneous PD/RIV and Voltage inputs. The modular design fits a wide range of PD detection applications. Conventional AC & DC PD measurements according to IEC 60270 or RIV measurements according to NEMA and CISPR standards are covered. A brand new and intuitive software includes all the required PD analysis tools such as information display, pulse diagram, PRPD pattern (fingerprinting), data logger, and more.
The DDX 9160 is a lightweight and portable PD detector from Haefely for AC, DC, and RIV partial discharge measurements. Together with the optional integrated measuring impedance, the DDX 9160 is ideal for on-site PD measurements when an existing coupling capacitor is available or for power transformer PD measurements using the bushing tap method. A brand new and intuitive software includes all the required PD analysis tools such as information display, pulse diagram, PRPD pattern (fingerprinting), data logger, and more.
The HTP-2 Digital Partial Discharge Detector is a simplified and cost-effective solution for measuring PD according to IEC 60247. There are 3 versions available:
- HTP-2/D – alphanumeric LCD is situated on the front panel of the detector showing the measuring values in pC and the frequency of the applied test voltage.
- HTP-2/W – measuring data access is given via web browser with calibration, test voltage, charge, and various diagrams.
- HTP-2/S – PD detector is supplied with HVT Software Suite for PC operation and with advanced functionality with recording of tests and replay of results (analyze PD inception and extinction voltages and phase resolved PD patterns).
When wanting to conduct PD testing of medium voltage cables using very low frequency (VLF, 0.1 Hz) voltages, we offer the PD-TaD 62 Portable PD Diagnostics System, which allows for the exact location of PD sources along the length of a cable using the “Time of Flight” PD measurement technique.
Non-Conventional Partial Discharge Detection:
The presence of PD can also be detected by other means other than the offline approach taken in IEC 60270. These unconventional technologies rely on sensors and are commonly used for on-site PD testing, in which an “online” test is conducted. This has the advantage that the voltage supplied to the DUT does not need to be shut off for testing, as the measurement does not require galvanic coupling to the high voltage conductors.
The HTP-2/UHF PD Detector offers flexibility of either single or multiple channel measurements for multiple apparatus testing at a single location. The PD analyzer can be configured with any standard decoupling package such as UHF sensor for cable sealing end, UHF drain valve & hatch sensor for transformers or internal sensor for GIS.
The Liona Online Partial Discharge Spot Tester is a portable online PD detector that uses capacitive TEV (“transient earth voltage”) sensors that can be used to measure PD signals in switchgear or other grounded surfaces. Furthermore, the use of sensitive HFCT sensors allows for the equipment to measure PD signals in cables that are inductively coupled from the cable sheath or ground.
UHF PD Sensors are a great tool to use to decouple high-frequency PD signals from joints and terminations of medium and high voltage cables, transformers, and motors or from within a transformer by using a Drain valve sensor. Installation of these sensors is conducted at ground potential allowing for safe operation. The stronger attenuation of distant signals in higher frequency ranges allows for very favorable signal-to-noise ratios and locally selective tests.
We also offer accessories, including special calibrators for calibrating standard PD measurement circuits to determine the apparent charge. Measuring impedances and coupling capacitors are offered for the decoupling of PD signals in standard measuring circuits according to IEC 60270.
- Coupling Capacitors with various voltage and capacitance ratings
- Measuring Impedances
- PD, RIV, and UHF Calibrators
- UHF PD Sensors
- PD Filters for noise suppression
- Multiplexers for switching between multiple measuring impedances