PARTIAL DISCHARGE MEASUREMENT AS CONDITION MONITORING TOOL FOR HV EQUIPMENT

1. PARTIAL DISCHARGE MEASUREMENT AS CONDITION MONITORING TOOL FOR HV EQUIPMENT Transformer Services Dept./TCB/BHEL

2. 1.INTRODUCTION Partial discharges (PD) occurs within electrical equipment whenever the voltage stress exceeds in some regions of the dielectric (insulation). Importance of partial discharges (PD) measurement -For quality control of high-voltage equipment - established test during development and final testing at supplier’s works. - Now, attempts are being made to carry out such PD measurements on HV equipment on site. -aims at a better assessment of the condition of high-voltage insulation to allow for scheduled maintenance in due time. - sudden, unexpected breakdown of equipment can be prevented with this condition-monitoring tool avoiding consequential damage. Even in the course of acceptance tests on new installation, a possible lack in quality can be recognized by PD measurements early stage . Transformer Services Dept./TCB/BHEL

3. But PD measurements under on-site conditionshave limitations due to • An adequate electromagnetic shielding of interference fields, extensively as used in the laboratory or test bay, cannot be implemented site at reasonable costs. • Customary PD measurement techniques are based on the de-coupling of PD signals via a HV coupling capacitor. For connection to the HV terminals of the unit under test, however, parts of the installation must be disconnected so that a service interruption cannot be avoided. • To overcome on-site PD measurement problems, PD probe technology based on measurement of pulse-shaped electromagnetic signals and acoustic detection technology based on ultrasonic signals developed. Transformer Services Dept./TCB/BHEL

4. 2. PD FUNDAMENTALS A partial discharge (PD) is defined as an electric discharge that partially bridges the Insulation between two conductors. Fig. 1: PD inside an insulation block The terms “PD” and “Corona” apply to the same phenomena of electrical discharges. However, the term “Corona” is preferably reserved for external partial discharges, for example, on the bushing terminal or on the transmission line. For internal discharges, the term “PD” is used. Transformer Services Dept./TCB/BHEL

5. Reasons for PD • Insufficient design margins • Improper choice of insulating materials • Bad processing • Bad workmanship • Faulty manufacturing procedures • Loose connections • Improper shielding inside & outside the equipment Transformer Services Dept./TCB/BHEL

6. Damaging effects of PD • Erosion on the surfaces / watts • Chemical degradation • Formation of pits • Tracking / Treeing • Loss of mechanical strength • Local heating • Interference with radio communication • PDs of high magnitude (> 10,000 pC) may lead to accelerated aging causing a complete damage of the insulation within a short period of time. Transformer Services Dept./TCB/BHEL

7. PD Quantities 1. Apparent Charge, q Fig. 2: Insulation with cavity and its Equivalent circuit Fig. 3: Voltage curves for applied voltage and PD in cavity Figure 2 shows a cavity inside an insulation block. Figure 3 shows its equivalent circuit in which the PD defect is simulated by a capacitor C1. A partial breakdown of C1 will occur as soon as a critical inception voltage u10 across C1 is exceeded. This PD will cause a voltage step (PD pulse) u1 across C1 which will be transmitted to the terminal A via the series capacitor C2. This results in to a step change in the measured voltage at terminal A as u (t) = u1. C2/C3, where C3 is the capacitance of the unit. Consequently, the charge transmitted to the terminals of the test object will be Transformer Services Dept./TCB/BHEL

8. C3 = u(t). C3 = u1. C2/ C3, C3 = u1. C2 (1) However, in the PD source, the charge Q1 drawn from the capacitance C3 will be Q1 = u1 (C1+C2) (2) Form the equation (1) & (2) Q3 = Q1. C2 / C1+C2 (3) From this it can be seen that the charge Q3 measured at the terminal A is always less than the charge Q1 converted in the PD defect. But since the ratio C2 / C1 + C2 is unknown, Q3 is defined as apparent charge. 2) Energy of individual PD, W W = 1/2 1 x (V)2 3) Repetition Rate, n – no. of PD pulses/sec. 4) Average Discharge Current, I I = 1/Tq1 + q2 + ...qn coulomb/sec 5) Quadratic Rate, D q2+q22+ ….qn2 coulomb/sec D = • Phase Angle 1 and time t1 of occurrence of a PD pulse , •  = 360 (t1/T) Where t1 is the time measured between the preceding positive going transition of the test voltage and the PD pulse. This the period of the test voltage. The phase angle is expressed in degrees. Transformer Services Dept./TCB/BHEL

9. Where PD can occur in electrical equipment? • Cavities in an insulation layer • Metal electrodes at boundaries of these insulation layers (surface discharges) • Electrode edges of a foil in an insulation layer (e.g. bushings, capacitors) • Electrodes in the air • Metal parts at free potential • Loose contact at high voltage points • Bad earthing • Solid particles in oil (DC equipment) Transformer Services Dept./TCB/BHEL

10. 3.RECOGNITION OF PARTIAL DISCHARGES PD in mineral-oil Impregnated paper PD at High voltage point (Line-end corona) PD at earth against High voltage (earth-end corona) PD caused by metallic Parts at free potential PD due to bad contacts Surface discharges Transformer Services Dept./TCB/BHEL

11. PD TEST CIRCUITS • Most circuits in use for Partial discharge measurements can be derived from one or other of the basic circuits, which are shown in figures 1a to 1d. Each of these circuits consists mainly of • A test object, which may usually be regarded as a capacitor Ca • A coupling capacitor Ck, which shall be of low inductance design, or a second test object Ca1, which shall be similar to the test object Ca. Ck or Ca1 should exhibit a sufficiently low level of partial discharges at the specified test voltage to allow the measurement of the specified partial discharge magnitude. • A higher level of partial discharges can be tolerated if the measuring system is capable of separating the discharges from the test object and the coupling capacitor and measuring them separately; Transformer Services Dept./TCB/BHEL

12. A measuring system with its input impedance A high-voltage supply, with sufficiently low level of background noise to allow the specified partial discharge magnitude to be measured at the specified test voltage; - An impedance or a filter may be introduced at high voltage to reduce background noise from the power supply. U high-voltage supply Zm input impedance CC connecting cable OL optical link Ca test object Ck coupling capacitor CD coupling device MI measuring instrument Z filter Fig. 4: PD Test Circuits Transformer Services Dept./TCB/BHEL

13. PD MEASURING SYSTEMS • The main purpose of PD measurement is following: • - To verify that test object is free from PD • above a specified magnitude (guaranteed value) at a specified voltage. • To determine the voltage level at which such PD starts (Inception) with increasing voltage and stops with decreasing voltage (Extinction) • - To determine PD level at the test voltage • PD meters, generally available, measure the magnitude q, the largest discharge (PD) signal during any short time interval. • The PD meters measure the signal with a peak discharge meter having a scale calibrated in pico-coulombs with display of PD pulses on a CRO using a rotating time base or a linear time base. Transformer Services Dept./TCB/BHEL

14. i) Wide-band PD instruments In combination with the coupling device, this type of instrument constitutes a wide-band PD measuring system which is characterised by a transfer impedance Z (f) having fixed values of the lower and upper limit frequencies f1 and f2, and adequate attenuation below f1 and above f2. Recommended values for f1, f2 and f are 30 kHz  f1 100 kHz f2 500 kHz; 100 kHz f  400 kHz The response of these instruments to a partial discharge current pulse is in general a well-damped oscillation. Both the apparent charge q and polarity of the PD current pulse can be determined from this response. The pulse resolution time Tr is small and is typically 5 s to 20 s. Transformer Services Dept./TCB/BHEL

15. ii) Narrow-band PD instrument (RIV Meters) These instruments are characterized by a small bandwidth f and a midband frequency fm, which can be varied over a wide frequency range, where the amplitude frequency spectrum of the PD current pulse is approximately constant. Recommended values for f and fm are 9 kHz f  30 kHz 50 kHz  fm  1 MHz However, normally wide-band detectors are preferred for the following reasons: - Better resolution - No resonance - PD signal magnitude measured do not change with PD repetition rate -Slow rising pulses i.e. Low frequency pulses can also be measured. This slow rising pulses may be possibly more damaging being erosion type discharges. Today, digital PD detectors are commonly used because of their ability to process the individual PD signal and record its instantaneous value of the test voltage, phase angle of occurrence and the time instant. The post-processing of recorded data facilitates reduction in noise level, objective evaluation of parameters and in-depth analysis of insulation quality. Transformer Services Dept./TCB/BHEL

16. 6. CALIBRATION OF A PD MEASURING SYSTEM The object of calibration is to verify that the measuring system will be able to measure the specified PD magnitude correctly. The calibration of a measuring system in the complete test circuit is made to determine the scale factor k for the measurement of the apparent charge. As the capacitance Ca of the test object affects the circuit characteristics, calibration shall be made with each new test object, unless tests are made on a series of similar objects with capacitance values within 10% of the mean values. The calibration of a measuring system in the complete test circuit is carried out by injecting short-duration current pulses of known charge magnitude qo (= Uo.Co) into the terminals of the test object, as shown in Figure. Calibration procedure Calibration of measuring systems intended for the measurement of apparent charge q, should be made by injecting pulses by means of a calibrator, across the terminals of the test object, as shown in figure 4. The calibration should be performed at one Transformer Services Dept./TCB/BHEL

17. magnitude in the relevant range of the magnitudes expected, to assure good accuracy for the specified PD magnitude. The relevant range of magnitude should, in lieu of other specifications, be understood to be from 50% to 200% of the specified PD magnitude. As the capacitor Co of a calibrator is often a low voltage capacitor, the calibration of the complete test arrangement is performed with the test object de-energised. For the calibration to remain valid, the calibration capacitor Co should not be larger than 0.1 Ca. If the requirements for the calibrator are met, the calibration pulse is then equivalent to a single event discharge magnitude qo - UoCo. Consequently, Co must be removed before energizing the test circuit. If, however, Co is of high voltage type, and has a sufficiently low level of background noise to allow the specified PD level to be measured at the specified test voltage, it can remain connected in the test circuit. In case of tall objects several meters in height, the injection capacitor Co should be located close to the high-voltage terminals of the test objects as the stray capacitance could cause unacceptable errors. Transformer Services Dept./TCB/BHEL

18. The connection cable between the step voltage generator and capacitor Co should be shielded and be equipped with appropriate termination to avoid distortion of the voltage step. Fig. 6: PD calibrator circuit Fig. 7: Bandwidth of calibrator signal Transformer Services Dept./TCB/BHEL

19. 7. PD TESTING UNDER LABORATORY CONDITIONS / AT WORKS • Precautions during testing • Surroundings must be clean • All metallic objects in the vicinity shall be earthed • - All capacitive objects like capacitors, insulators and bushings etc. shall be shorted • and earthed • Any high voltage tests in the vicinity shall be stopped • - No switching operations shall be done - No crane operations shall be done • Check for faulty tube lights & switch then off • - Test object shall be properly earthed - High voltage terminals shall be properly • shielded with discharge free corona shields • - The screen or coaxial cables shall be properly earthed. • In order to obtain reproducible results in PD testing, careful control of all relevant factors is necessary. The PD system shall be calibrated as above prior to testing. • Before being tested, a test object should undergo a conditioning procedure as specified below. Transformer Services Dept./TCB/BHEL

20. Under no circumstances, however, shall the voltage applied exceed the rated withstand voltage applicable to the apparatus under test. • Determination of the partial discharge magnitude at a specified test voltage • b1) Measurement without pre-stressing on capacitive test objects • For example on – Bushings, capacitors, instrument transformers etc. • The partial discharge magnitude in terms of the specified quantity is measured at a specified voltage, which may be well above the expected partial discharge inception voltage. The voltage is gradually increased from a low value to the specified value and maintained there for the specified time. As the magnitudes may change with time, the specified quantity shall be measured at the end of this time. • The magnitude of the discharges may also be measured and recorded while the voltage is being increased or reduced or throughout the entire test period. • Measurement with pre-stressing – (on inductive test objects) • For example on – Transformers, reactors, motors, generators etc. Transformer Services Dept./TCB/BHEL

21. The test is made by raising the test voltage from a value below the specified partial discharge test voltage up to a specified voltage exceeding this voltage. The voltage is then maintained for the specified time and, thereafter, gradually reduced to the specified partialdischarge test voltage. At this voltage level, the voltage is maintained for a specified time and at the end of this time the specified PD quantity is measured in a given time interval or throughout the specified time. Measuring uncertainty and sensitivity The magnitude, duration and pulse repetition rate of PDpulses may be greatly affected by the time voltage application. Also, the measurement of different quantities related to PD pulses usually presents larger uncertainties than other measurements during high voltage tests. Consequently it may be difficult to confirm PD test data by repeating tests. This should be taken into consideration when specifying partial discharge acceptance tests. Disturbances Transformer Services Dept./TCB/BHEL

22. The measurements are affected by disturbances which should be low enough to permit a sufficiently and accurate measurement of the PD quantity to be monitored. As disturbances may coincide with PD pulses and as they are often superimposed on the measured quantities, the background noise level should preferably be less than 50% of a specified permissible partial discharge magnitude. High readings clearly known to be caused by external disturbances may be neglected. An electro-magnetically shielded hall (Faraday cage) will solve many of the interference problems from radio transmissions and other external sources there by reducing the overall disturbance level. However, shielded laboratory is not a pre-requisite for PD testing. It is necessary in the cases where PD testing. It is a regular activity and disturbance level is very high due to crane movements, commutators and other electrical operations. In order to obtain reproducible results in partial discharge tests, careful control of all relevant factors is necessary. The partial discharge measuring system shall be calibrated prior to testing . Fig. 8: PD measurement on a tapping of a bushing Transformer Services Dept./TCB/BHEL

23. CONDITION MONITORING TOOLS FOR ELECTRICAL EQUIPMENT The primary purpose of condition monitoring is to detect the first sign of an incipient fault, ageing development or other problems and monitor their evolution to enable the operator to take appropriate action and avoid major outages. With modern technology, it is possible to monitor a large number of parameters of the installed equipment at site, by both electrical and non-electrical condition monitoring tools. In order to get effective monitoring equipment to a moderate cost; it is necessary of focus on a few important parameters based on failure statistics and to the estimated consequences of the respective failure. The condition monitoring tools, which are in use for many decades, are listed below. i) Traditional - Insulation resistance measurement - Capacitance and tan delta measurement - Oil characteristics & indicators - Dissolved Gas Analysis (DGA) for oil- filled electrical equipment Transformer Services Dept./TCB/BHEL

24. ii) Non-traditional - Detection of audible & visible corona - Infra-red thermography - Hydrogen in oil (DGA-updated) iii) Special tools (based on modern technology) - Fibre-optic hot spot sensor - Degree-of polymerization of cellulose - Detection and location of PD by electrical and acoustic sensors - Mechanical vibration & stress measurement - Signature analysis Presently, capacitance and tan delta (DDF) measurement for all types of electrical equipment and DGA for oil-filled electrical equipment have proved to be quite reliable and established technique for assessing dielectric performance of the electrical equipment in service. These conventional methods show a general condition of the equipment and provide evidence of an incipient fault or ageing but they do not indicate the location of the fault or the faulty zone within the equipment. Transformer Services Dept./TCB/BHEL

25. PD as condition monitoring tool One of the valuable methods to assess and locate faults is the non-conventional on-line PD measurement on the equipment. This avoids system shutdown and provides fault indication at very early stage. PD measurement under on-site conditions encounters crucial problems because there is no electromagnetic shielding from the interference fields, which causes disturbance during PD measurements. Today, development of sophisticated digital measuring and analysing systems have made PD detection and location possible on equipment in service, these systems are based on digital counting of discharge pulses and the evaluation of their number, amplitude and phase related to the applied voltage. This allows discriminating different types of discharges and their location. The two techniques normally employed for PD detection and location without the disconnection of test object are i) Acoustic technique based on ultrasonic signals emanated from PD source and ii) Electrical technique based PD probing by capacitive and inductive couplers. Transformer Services Dept./TCB/BHEL

26. In general, these methods are not suitable for quantitative measurement of PD quantities, but they are essentially use to detect and/ or to locate the discharges. Fig. 10: Acoustic detection technique 1. ACOUSTIC DETECTION Acoustical measurements are usually made with microphones or ultrasonic transducers in combination with amplifiers and suitable display units. The technique has been found to be more useful for locating the PD site with high sensitivity in the ultrasonic range. Here for locating corona discharges in air, normally, directional selective microphones are used. For locating discharges in GIS and oil-immersed equipment such as transformers, acoustic transducers in the ultrasound frequency rage are useful. Transformer Services Dept./TCB/BHEL

27. 1.1 Acoustic detection on the metallicwall of the test equipment a.For detection of PD activity The sensor is mounted on the tank wall and the hand-held detector enables to sense PD activity within the equipment. Example, PDD-1000 detector of Babcock &Wilcox, UK b.For location of PD source (for transformers, reactors) Minimum three sensors are mounted on the tank wall and PD source detected and located by geometrical triangulation. Though time consuming, it gives precise location of discharge site. See figure 1.2 Acoustic wave-guides (for transformers, GIS etc) In this method, the sensor is coupled to one end of the wave-guide, with the other end of the wave-guide immersed in the transformer through a packing gland on the side wall or top of the transformer tank. The wave-guide picks up acoustic signal in the range of 100-300 kHz and transmits the converted electrical pulses to a monitor with preset alarm for indication presence of harmful PD signal. Example, WECAS Acoustic System by Westinghouse, USA. Transformer Services Dept./TCB/BHEL

28. 1.3 Corona detectors (for cables, connectors, insulators, bushing terminals, cable, joints) The directional corona detectors measure the acoustic emission from air discharges (corona) with a resonant frequency between 30 to 50 kHz. Generally, with this type of detector, the acoustic emissions from a PD source of 100 pC can be detected at distance of 5 m. Example, Corona Leak detector by Biddle Instruments, USA. Transformer Services Dept./TCB/BHEL

29. 2. ELECTRICAL PD PROBE (ELECTRO-MAGNETIC) DETECTION This non-conventional technique in comparison to the standard PD measurement is based on PD probing by field coupling of the test object with the sensor and then taking the advantage of latest digital PD pulse processing techniques in the wide-band range to detect the PD signal. 2.1 PD Probe measurement technique In this technique, PD detector in MHz range is used as an active wide-band aerial to pick up the pulse-shaped signals in the nano-seconds range. The detection of the PD signals occurs via the electro-magnetic radiation field by either potential-free field coupling (C-mode) or potential-referenced field coupling (L-mode). The technique is being used for on-site PD diagnosis on cables, GIS, transformers, generators, bushings etc. with limited success due to poor modes of coupling. It only provides comparative PD activity as preventive measure. Example, LEMKE Probe by LDIC, Germany Transformer Services Dept./TCB/BHEL

30. 2.2 PD detectors for stator-winding problems This detector uses two types of sensors, which are permanently mounted on the motor or generator to enable detection of stator insulation problems, while eliminating all types of electrical interference. One type of sensor called stator slot coupler (SSC) is a wide-band antenna and is permanently installed in the stator windings under the wedges or between the top and bottom bars in a slot. The other type of sensor called Bus couplers are radio-frequency current transformers (RFCT) mounted on the terminals of a machine. The PD signals picked by these sensors are analysed and processed by a wide-band (0.1 to 800 MHz) PD detection system to assess the condition of stator winding. Example, TGA by IRIS Power Eng. Inc., Canada. Fig. 14: Electrical probe for PD detection Transformer Services Dept./TCB/BHEL

31. C B D A E U 1 U U 2 2 1.1 / 3 1.1 U / 3 m m U < U start start A = 5 min B = 5 min C = test time in seconds D = 5 min for ACSD and 30/60 min for ACLD E = 5 min Fig. 12 Time sequence for the application of test voltage with respect to earth Here Um=Highest voltage for equipment U1= Test voltage U2= Partial discharge evaluation level The detailed procedure and specific test requirements are addressed in IEC-60076-3 Transformer Services Dept./TCB/BHEL

32. Induced AC voltage test Transformer Services Dept./TCB/BHEL