of 44

Addressing Noise Sources During Online PD Testing

4 views44 pages

Download

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Addressing Noise Sources During Online PD Testing Timothy Erwin National Sales Manager, EA Technology LLC March 11, 2019 Agenda Review: What is PD? Example 1: Noise (?) Example 2: Power inverters and different
Addressing Noise Sources During Online PD Testing Timothy Erwin National Sales Manager, EA Technology LLC March 11, 2019 Agenda Review: What is PD? Example 1: Noise (?) Example 2: Power inverters and different technologies Example 3: Dry-type transformers What can we do to avoid these problems? What did we learn? EA Technology Group Originally established as R&D Center for the UK Electricity Industry. Privatized in the early 1990s. Provides research, strategic engineering consultancy, HV asset condition assessment services, specialized instrumentation and asset management software and consulting. Discovered the TEV effect in 1978 and developed the first practical detection method in 1982 Instrumental in the development of PAS-55 and ISO Maintains a large forensic investigation lab equipped with the latest technology and staffed by highly experienced scientists. Is Partial Discharge Real? 4 PD failure process What is Partial Discharge? Multiple causes Starts small ALWAYS gets worse Leads to FLASHOVER PD is the inability of a portion of the insulation to withstand the electric field applied to it Partial Discharge - A flashover of part of the insulation system due to a localized electric field greater than the dielectric withstand capability of that part where the overall insulation system remains capable of withstanding the applied electrical field. C1 C2 C3 Conductor Void Ground HF Current Pulse V withstand of C2 applied field V withstand of C1&C3 applied field One effect of this flashover is a high frequency current pulse that travels through the capacitance of the insulation (C1 & C3) Phase Resolved Plots Partial Discharge tends to occur on the rising edge of the voltage sine wave. As such, PD impulses tend to be synchronized to the AC waveform and 180 degrees apart. Phase Resolved plots show PD impulses on a power system cycle so groupings 180 degrees apart can be seen. 180 Degrees Online test classifications Ultrasonic/Acoustic testing through louvers, vents, contact sensors, and parabolic dishes (Audio over 20KHz) TEV testing Makes use of the Transient Earth Voltage phenomenon to safely detect internal discharge from outside cabinets RFCT testing* - By attaching RFCT to cable ground straps, the PD current can be safely measured on live cables RF Testing Specifically designed directional and non-directional radio receivers can pickup the EMI generated by PD * installing RFCT on live cables requires opening the HV compartment and appropriate safety measures Types of PD Internal discharges occurring in defects, voids or cavities within solid insulation (TEV, UHF, RFCT, IR) Surface discharges occurring across the insulation surface (Ultrasonic, UHF, RFCT) Corona discharge occurring in gaseous dielectrics in the presence of inhomogeneous fields (Corona Camera, Ultrasonic, UHF,) Contact discharge occurs on floating metal in high field conditions (TEV, Ultrasonic, UHF) Surface PD Flexible microphone 11kV HV Cable phases too close to insulated plastic boot Can be just heard with naked ear Surface PD, Crossed phases Contact Probe 11kV HV Cable phases too close crossed phases Can see 180 degree clustering Difficult to determine from single reading but obvious on site as only difference in switchboard Corona / Surface PD Flexible Microphone 11kV connection at rear of CB spouts, poor insulation interface Note white build up, also rust and green verdigris Corona with some surface PD can be seen in phase resolved pattern Corona tends to have a deeper sound Corona, 132kV Isolator - Ultrasonic 132kV isolator, sharp point into air High ultrasonic Note one peak of activity on one half of sine wave Magnetic Transformer Vibration, Contact Probe 11kV Transformer magnetic vibration noise detected with contact probe no problem detected Sound goes away when metal plate is physically leant on stops vibrating Note 4 clusters of activity in phase resolved plot, not indicative of PD Noise, Industrial- TEV Industrial noise from factory load PPC count No PD signals can be detected Example 1: NOISE(?) Typical switchgear Breakers, busbars No motors or rotating equipment PD Indication Low certainty No Phase Resolve Plot Let s take a listen That is something Not really electrical but close Sounds a little more mechanical Neighboring cell: Similar levels and pattern Classified as noise Fairly high % certainty Very similar sound Located midway on 9 section lineup Classified as Noise Levels similar. (All sections) Rotated microphone, levels stayed consistent End section of gear Near large floor circulating fan Audio is high pitched, well above 120 Hz Not like sections 7 and 8. Most likely from fan, possibly bearing noise High levels of background noise most likely caused by circulation fans and AC units within an isle's width of the gear. Because of mid summer temps we were unable to turnoff auxiliary fans. There is something at sections 7 and 8. Not necessarily PD, possibly loose metal or panel vibrating. Our work permit did not allow opening of energized equipment. Investigation is ongoing. UHF 800 MHz probably would have eliminated a lot of background noise giving better indication of PD Example 2: Power inverters and different detection technologies Electronic switching inverters Locations with switching inverters are problematic as they tend to be very noisy in the spectrums where we look for PD. Common in windfarms, solar farms and scientific manufacturing facilities that need very clean power These locations also tend to be tight on space and utilize dry type transformers that also radiate their own noise signatures. Operate in the range of 10 s to 100 s KHz Example 1: Scientific manufacturing facility in room with switching inverters Ultrasonic Background levels in room high Significant noise level Gain could be masking lower level PD Example 1: Scientific manufacturing facility in room with switching inverters TEV: High levels 43 Db High Pulse Per Cycle count Within background levels of 39Db and PPC Example 2: Wind Turbine with known PD in basement mounted transformer Transformer had been tested with off-line equipment and it was determined that there was PD on at least one winding Testing with multiple online technologies was performed including TEV Ultrasonic RFCT on transformer incoming, tail and chassis ground straps Radiometric (RF) PD Detector Different test equipment was used for the TEV and Ultrasonic, RFCT on cable straps and the RF PD detector. Tests were run with the turbines generating and with them off and transformer back fed from the 33Kv side. This allowed a comparison of the affect of the inverters when generating. TEV: Sample Average amplitude and PPC over 24 hours Measurements are recorded on 3 types of probes- internal and external types and an antennae at multiple locations near each winding When turbine is on amplitudes are high When turbine is off (windspeed dropped) amplitudes drop. From dbmv to 0-16 dbmv. Alarm threshold is 29 dbmv Full time monitoring or planned interruption of turbine operation would be necessary. TEV: Phase resolved plot- Probe on transformer panel. Measurements with inverters operating will not allow TEV detection But with inverters stopped the PRP shows PD. TEV: Phase resolved plot- Probe on transformer panel. Clear PD pattern with turbine stopped, inverts off. Ultrasonic: Average amplitude over same 24 hours as TEV Similar trend as TEV Low levels of U.S. activity when inverter was off suggest it is not possible to use U.S. as a method of PD detection in transformers Because the transformers have solid resin insulation there is no airpath out, so ultrasonic internal PD detection is not possible with airborne sensors. Cable testing- RFCT RFCT Incoming 33 kv interconnect cables Transformer Tails Incoming cable winding #1 Turbine off Inverter noise No PD Pulse Incoming cable winding #3- Turbine off Inverter noise Good PD cluster PD Pulse (6K pc) Reflected 12.7 ms Transformer tail winding #3- Turbine off No inverter noise Good PD cluster PD Pulse (9K pc) Reflected 12.7 ms Transformer chassis earth- Turbine off No significant PD Clusters Waveforms hazy with significant ringing Possible reflection at about 22.5ms Transformer tail winding #3- Turbine generating Note levels of noise introduced by the turbine generating. No sign of grouping or clustering Waveforms characterized by bursts around the trigger point and significant noise on the rest of the waveform Filtering does not help significantly. UHF (RF) Test Turbine generating Center Freq 777 MHz Inverter operating freq. usually 10 s to 100 s khz Not effected by inverter noise Gain Center Freq UHF (RF) Test Turbine not generating Little difference between turbines on and off Little apparent interference from inverters PRP was slightly unstable and not exactly 180 o This may have been caused by inception voltage different on positive and negative half cycles When the HV transformer was de-energized the pulses stopped and the audio output went quiet. Sensor placement Whip antennae US Microphone Plate antennae Telescopic antennae Dry type transformers Challenges: Vibrate at 2x system frequency (120Hz)- noisy Lack of dampening cause vibrations throughout transformer panels etc. Significant source of ultrasonic noise. Resin cast windings often dampen PD Ultrasonic emissions to the point its undetectable Ultrasonic Test High background noise Cannot hear anything on audio Cant really make a valid assessment n TEV Clusters at 90 deg. Not synchronous to 60 Hz (0-PPC) What have we learned? Know the equipment to be tested and understand the environmental impact. The use of different technologies, TEV, US, UHF, can provide different results, or supporting data for your conclusions Background measurements are valuable for determining if a measurement is real or noise. Please Remember to Complete Your Evaluation Form Thank You!
Advertisement
MostRelated
View more
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks