Kingstone,
The key to online testing is to understand its limitations.
The levels you have been given by the manufacturer are completely arbitrary. What are you using the unit for? It is really is not practical to test cable insulation with an online PD tester. If you are measuring PD in transformers, busbars, cable accessories, switchgear or other hardware components you might learn something from your measurement. Such components lend themselves to online PD testing because they can sometimes handle PD for an extended period of time allowing you time to notice them and take corrective action.
The unit’s sensitivity is not calibrated for each test:
The only way to determine actual presence of partial discharge (PD) is to calibrate the instrument you are using. To calibrate the instrument, according to AEIC, IEC 60270 or any other international standard you must inject a calibrated pulse (e.g. 5pC-pico Coulombs) into the apparatus being monitored and measure the response your receive from the measurement device. In this way you can determine the losses incurred by a signal traveling to your sensor. There will be losses in due to the traveling path (e.g. the cable), at joints, disruption in the cable shield, non-linear losses in the semi conducting shield, loss coupling into the air (on it way to your sensor) and losses due to the inefficiency of your signal coupler. All the standard require that you take the cable off-line and calibrate your sensor. Without calibrating you don’t know how much of the PD signal is actually getting to you! You may walk away from test think that no PD was detected, when really the PD just never arrived at your sensor!
The unit probably can’t tell the difference between noise and PD:
Many times in medium voltage cable systems there is harmonic (communication signals) and pulse type noises. Some pulse type noises correspond with the typical phase angles of the PD signals that you are looking for! You may read a high intensity which could be due to other PD sources near your sensor, for example other terminations, and switchgear components.
Sensitivity and Cable insulation defects:
What is the smallest detectable signal in pC that your unit can detect? I doubt it is better than 50pC. If you are looking for PD in joints and terminations you might have a chance of finding something, but 70% of all PDs in cable insulation are below 50pC!
Limitation of On-line testing with cable insulation defects:
There is another problem with the levels you have been given. Say for instance you do calibrate your sensor, you find a good sensitivity (e.g. 5-20pC), and you can distinguish between noise and PD, only 3% of all cable insulation defects have PD at operating voltage! The ones that do appear at operating voltage go to failure in minutes because they are under conditions with constant PD!
Again, the key to successfully using an on-line PD test is for you to understand its limitations.
Recommendations for use:
If you find PD:
Try to determine the location of the source is coming from. Take the system off-line and inspect the PD location. In the case of a switch cabinet and a termination with PD disconnect your cable and energize the cabinet to determine if the PD source is the cable or the cabinet. If you repair the cable accessory or the hardware component and the PD still exists the test may be beyond a simple online test and you may have to call in someone to perform an off-line PD test up to 2-3Uo depending on age and voltage class.
If you do not find PD:
Since you are performing an uncalibrated test you should never make the claim there is no PD. The cable or device may still fail due to undetected PD. If you want to know more about your system’s future performance you will have to call in someone to perform an off-line PD test up to 2-3Uo depending on age and voltage class.
-cheers
