“Time of Flight” Partial Discharge Measurements

“Time of Flight” partial discharge (PD) measurements are based upon the principle of time domain reflectometry (TDR). The PD pulse signal generated at the PD source travels along the length of the cable allowing for localization of the pulse source. By analyzing the time it takes for the PD signal to travel a certain length of the cable, direct allocation of PD activity in cable segments and accessories, such as joints and terminations, is enabled. This measurement technique can detect defects in joints, terminations, electrical trees in polymeric cables, insufficient oil insulation in PILC cables, and mechanical damage of the cable sheath.

The PD-TaD 62 and PD-TaD 80 partial discharge diagnostic systems allow for the measurement of PD according to IEC 60270 and for source localization of PD activity along the length of the cable using a very low frequency (VLF) sinewave voltage waveform. Together with a VLF generator, such as the Frida or Viola, a complete portable VLF partial discharge measurement solution can be created. Furthermore, when combining the PD-TaD system with a VLF generator that has integrated tan delta measuring function, such as the Frida TD or Viola TD, a full monitored withstand test can be conducted. This means that tan delta and partial discharge can be measured simultaneously (see a 2018 T&D paper written on this here).

In a previous blog post, The Basics of Partial Discharge Testing, detailed information on PD measurements according to IEC 60270 was provided. Essentially a high voltage source (VLF generator) is used to energize the cable to be tested and a coupling capacitor, in series with a measuring impedance (quadripole), is placed in parallel to the cable under test to decouple the PD signals and convert into voltage pulses for analysis with a partial discharge detector. The frequency bandwidth for the measurement of pulse charge in pC according to IEC 60270 is between 100 and 500 kHz (100 kHz to 1 MHz according to the 2015 IEC 60270 amendment). In order to measure the distance where a PD signal originates according to the principle of TDR, the PD detector must also be designed as a digital storage oscilloscope (DSO). For the PD-TaD detectors the frequency bandwidth for this is given as 100 MHz. Below is a typical set up for PD measurements using field-portable VLF equipment.

During a partial discharge event the PD signal will split and propagate down the length of the cable with one signal propagating towards the near end, where the diagnostics system is connected, and another signal propagating towards the far end of the cable. Once the first signal reaches the near end, the PD detector will be triggered and a reflection will occur so that the signal then propagates towards the far end of the cable. In the meantime, the second signal from the source of PD has reflected from the far end and is propagating towards the near end. Once it reaches the detector, a second pulse appears on the oscilloscope. The initial signal that triggered the PD detector is still propagating down the cable and reflects from the far end. Once this signal returns to the near end a third and final pulse will appear on the oscilloscope, which concludes the PD measurement. A graphical representation of this process is shown below.

In the above sequence you will notice that the PD signal loses energy as it propagates due to attenuation and dispersion, causing the pulse peak, fall time, and pulse duration to become smaller. However, since the signal is being integrated over time, the measured charge of the PD pulse does not change, regardless of its shape (see The Basics of Partial Discharge Testing).

Analyzing the oscilloscope diagram after the measurements have completed can also help to interpret the general location of the PD source. For example, if 2 pulses with generally the same pulse shape, amplitude, and duration are very close to one another at the near end of the oscilloscope display, this is a clear indication that the PD source is very close to the far end of the cable. This is because the signal propagating towards the far end is almost immediately reflected and closely follows the signal propagating towards the near end.

Similarly, if 2 pulses are very close to one another at the end of the measurement display, then the PD source originates close to the near end of the cable. The initial PD pulse that triggers the measurement has a very high amplitude and is very narrow, as the signal originates close to the detector. This signal is then reflected and quickly follows the signal propagating towards the far end. The subsequent 2 pulses are highly attenuated and damped, as they have traveled twice the cable length.

It is important to note that when PD originates from the body of the cable, 3 PD pulses will be registered during the time of flight measurement, as shown in the sequence above. If only 2 signals are visible, this is a clear indication that the PD originates from the cable termination either at the near or far end. To determine at which end the signals originate, the pulse shape can be analyzed. A very large and narrow initial pulse with an attenuated second pulse will indicate a near end termination source. An attenuated initial pulse and severely attenuated second pulse will indicate a far end termination source. The initial pulse will have traveled the entire cable length and second pulse will have traveled a distance three times the cable length. The image below shows the difference in pulse shape when comparing near end PD (top) and far end PD (bottom).

Before starting a PD measurement, the PD circuit must be calibrated with pulses of known charge to determine the scale factor of the PD measuring system for pulse charge measurements according to IEC 60270. Moreover, during the calibration of the system, the signal propagation velocity of the cable system is also determined. Below is an example of a 500 pC calibration signal being injected into a cable under test. The calibration signal allows for charge and distance calibration of the PD measuring circuit.

As the time of flight measurement technique is based off the principle of TDR, the PD detector measures the time it takes for the pulses to return after the system has been triggered, and therefore, by knowing the exact length of cable, the source of the PD signal can be determined. In a cable system with many sources of PD, the time of flight method allows for a simple overview of the location of the PD sources along the length of a cable. This provides helpful information to determine which areas of a cable should be replaced before in-service failures could occur.