In-Service Testing and Diagnosis of Gapless Metal Oxide Surge Arresters According to IEC

Size: px

Start display at page:

Download "In-Service Testing and Diagnosis of Gapless Metal Oxide Surge Arresters According to IEC"

Andrea Burns
6 years ago
Views:

1 In-Service Testing and Diagnosis of Gapless Metal Oxide Surge Arresters According to IEC

2 Overview of presentation Motivation for condition monitoring of metal oxide surge arresters (MOSA) The Surge arrester life Service experience Examples of arrester failures Characteristic properties of MOSA (ZnO)- arrester Aging and causes of failure Consequences of failure transformer failures Title 2 2

3 Overview of presentation contd. Surge arrester condition assessment IEC about Diagnostic indicators of metal oxide surge arresters in service Monitoring equipment and field application for third harmonic analysis with compensation Testing strategy and risk assessment Case studies Title 3 3

4 Background and Motivation The Metal Oxide Surge Arrester (MOSA) is a cheap and passive component, but protecting crucial apparatus. Overlooked despite severe consequences if it fails. MOSAs can age and fail due for a number of different reasons. May offer inadequate overvoltage protection, especially if the rated voltage is selected too low. Diagnostic indicator: Resistive leakage current increases with time increasing risk for failure. Title 4 4

5 Power System Overview Typical Location of Surge Arresters Typical location of surge arresters: In substations At the end of transmission lines At cable ends At transformers, generators, capacitors etc Location depending on voltage level, equipment and local conditions Title 5 5

6 The Surge Arrester Life The normal destiny of the surge arrester is to be: specified, purchased, installed - and forgotten Most common maintenance practice: No testing of surge arresters Only replacement after breakdown. surge arresters are inexpensive no big deal to replace!!! Is this really an acceptable practice? Title 6 6

7 The Surge Arrester Life Why care about surge arresters? 1. The arrester is your bodyguard for protecting important apparatus against the overvoltage terrorists 2. You cannot see if an arrester is bad, but you can measure it. The big question is: Are the arresters fit for fight? Title 7 7

8 Surge Arrester Service Experiences Failure rate depending on arrester quality, dimensioning and local conditions Typical failure effects on arresters: Explosion and external damages visual detection Puncturing and causing earth fault - indicated by earth fault relay, can be difficulty to locate Aged arrester with reduced protection level cannot be found without checking the arrester Title 8 8

9 Failure of 400kV Surge Arrester Title 9 9

10 Failed Arrester hanging with Bus Pipe Title 10 10

11 Shattered Pieces of Surge Arrester Stacks Title 11 11

12 Damaged Surge Monitor and shattered Pieces of Arrester stack Title 12 12

13 Another failed Surge Arrester Title 13 13

14 More Failure of Surge Arrester Title 14 14

15 Surge Arrester Properties Main objectives: Protect important apparatus against dangerous overvoltages Low resistance during surges so that overvoltages are limited High resistance during normal operation, to avoid negative effects on the power system Sufficient energy absorption capability for stable operation Title 15 15

16 Equivalent Circuit Diagrams SiC Arrester MO Arrester Series spark gap and RC control MO discharge resistor SiC discharge resistor Title 16 16

17 Voltage Current Characteristics MOSA (ZnO) and SiC Arresters Title 17 17

18 Thermal Instability Thermal instability and arrester failure can occur at operating voltage in case the temperature of the blocks is too high. MOSA must be correctly selected with respect to: o continuous operating voltage o different kinds of overvoltages o ambient temperature o pollution o ageing Title 18 18

19 Total Power Dissipation Accumulated of Sequence of Incidents in the Network 1. A lightning strike causes a discharge in a MOSA 2. The lightning causes an earth fault in the network 3. Single line to earth fault causes voltage increases on the two healthy phases 4. The earth is disconnected by a circuit breaker 5. Disconnection of the fault can cause increased TOV due to load dropping 6. Circuit breaker reclosing cause additional arrester energy due to switching overvoltages Title 19 19

20 Critical factors to avoid failure Rated voltage must be chosen high enough based on: - Normal operation conditions - Ambient temperature - Continuous voltage - Surface contamination - Ageing - Accumulated energy from previous discharges Title 20 20

21 The choice of MOSA is always a compromise Increased nominal/rated voltage: Possibility that the MOSA will withstand the stress increases Reduced protection margin Arresters with higher energy class: reduced risk for arrester failure Price increases The choice of MOSA is a compromise between protection level, voltage withstand and energy absorption Title 21 21

22 Design of Porchelain-MOSA eks. Cooper Power Systems Title 22 22

23 Design of Polymeric-MOSA (ABB) Title 23 23

24 Ageing of MOSA Normal operating voltage causes ageing Pollution and overvoltage surges can cause ageing from overloading of all or some of the blocks Moisture entry through sealing gaskets, may lead to shorting of ZnO discs and overstressing of healthy ZnO blocks. Degree of ageing depends on the nature/ quality of the granular layer. Increase in resistive leakage current may bring the arrester to thermal instability and complete arrester breakdown. Title 24 24

25 Metal Oxide Surge Arresters - Causes for Failure Incorrect arrester specification corresponding to actual system voltage and overvoltage stress Overloading due to: Temporary overvoltages (cracking, puncturing). Switching overvoltages (cracking, puncturing, flashover). Lightning overvoltages (change of characteristic/ageing, flashover, puncturing). External pollution or moisture penetration. Consequence of aging: Increase in the continuous resistive leakage current. This is a good indicator of the arrester condition. Title 25 25

26 Consequences of Arrester Failures Reduced overvoltage protection - Increased risk of equipment failure and outages for instance breakdown in transformer, bushings, switchgear Possible break-down of porcelain housings: - Risk of personal injury - Risk of damage to other equipment Title 26 26

27 Reasons for Transformer Failures The US insurance company HSB reports, as reasons for transformer failures: Electrical disturbances: 29% Lightning: 16% Has the arrester done its job? Title 27 27

28 YES -Transformers Do Fail Title 28 28

29 YES -Transformers Do Fail Title 29 29

30 Possibilities for Surge Arrester Condition Assessment SiC - Arresters with spark gaps: No reliable in-service method available off-line tests: Spark-over test and grading current measurement Dielectric loss (the Doble test) Metal oxide surge arresters without gaps: In service tests are possible On line tests: Continuous leakage current during normal service. Available in-service methods discussed in Amendment 1 to IEC : Diagnostic indicators for metal oxide surge arresters in service. Title 30 30

31 Why Condition Assessment of Surge Arresters? Condition check performed on regular basis will: Increase the safety for the operational and maintenance staff. Give early warning signals utilize life time and take aged arresters out of service before they fail. Prevent costly arrester failures and service interruptions. Prevent damages to other equipment, e.g. transformer bushings. Title 31 31

32 IEC Part 5: Selection and Application Recommendation Title 32 32

33 Methods for Monitoring of degradation of MOSA Visual inspection Locating external abnormalities on the arrester and gives practically no information about the internal of the arrester Surge counters Frequently installed on MOSA, but has no practical use for diagnosis of condition of the arresters Temperature measurements Thermo Vision Frequently used method. Detects the increased block temperature on the housing surface of the arrester. Leakage current measurements Most used diagnostic method. For in-service testing, the method with indirect determination of the resistive leakage current with compensation for harmonics in the voltage (THRC) is providing the best available information quality with respect to diagnostic efficiency. Title 33 33

34 Conventional Surge Counters Title 34 34

35 Modern Surge Counter ABB EXCOUNT II SURGE COUNTING: - Number - Time stamp - Current amplitude classif. Mod. 1 Mod. 2 CONDITION MONITORING: - Total leakage current - Resistive leakage current (Method B2 - IEC ) Title 35 35

36 Monitoring Spark Gaps, from TriDelta Title 36 36

37 IEC : Leakage Current Measurements I t I c: ma U I r: A Title 37 37

38 IEC : Leakage Current Measurements Measurement of the total leakage current example: I t I c = 100 I c = 100 I c: ma U I r: A I t = 100,5 I t = 104,5 I r = 10 I r = 30 Title 38 38

39 IEC : Leakage Current Measurements Measurement of the total leakage current example: I c = 100 I c = 100 I t = 100,5 I t = 104,5 The total leakage current increases with only 4% when the resistive part is triple. This small change in I t is difficult to read on the ma meter. I r = 10 I r = 30 Title 39 39

40 IEC : Leakage Current Measurement IEC , clause : At given values of voltage and temperature, the resistive component of the leakage current is a sensitive indicator of changes in the voltagecurrent characteristic of non-linear metal-oxide resistors. The resistive current can, therefore, be used as a tool for diagnostic indication of changes in the condition of metal-oxide arresters in service. Title 40 40

41 Equivalent Circuit of MOSA I c in the same size as I t. I t I r is nonlinear and depends on voltage level and temperature. I c: ma U I r: A U sinusoidal (fundamental component only): I 1c, I 1r, I 3r Harmonics in the operating voltage U: I 1c, I 1r, I 3r, I 3c I 3r (and I r ) is generated by the arrester itself and can be used as a diagnostic indicator. Title 41 41

42 Typical Voltage - Current Characteristics The resistive current component: is typically 5-20% of the total leakage current under normal operating conditions. is a sensitive indicator of changes in the voltagecurrent characteristic. depends on the voltage and temperature. Title 42 42

43 IEC : Leakage Current Measurements Method B1: 3rd harmonic analysis of leakage current: IEC says: Error range for third harmonic leakage current without compensation for different phase angles of system voltage third harmonics: Includes various voltage-current characteristics of no linear metal-oxide resistors. 1% third harmonic in voltage may give ±100% measurement error. (Norway: 0,1 0,9% harm.) Title 43 43

44 IEC : Leakage Current Measurements Method B2: Harmonic analysis of leakage current using third harmonic with compensation: 3 rd harmonic analysis chosen is used as a basis to obtain feasibility/reliability measurements in three-phase applications onsite. Presences of harmonics in the operating voltage generate harmonic capacitive leakage currents that is indirectly measured and compensated for. The key for compensation is application of field probe for indirect measurement of the 3rd harmonic capacitive leakage current generated by the operating voltage. The total and true resistive leakage current I r is calculated from I 3r and arrester data (incl. correction for temperature and voltage). Title 44 44

45 IEC : Leakage Current Measurements Weakness with Method B1: The presence of harmonics in the system voltage have been a great problem since these harmonics may interfere with the harmonics generated by the nonlinear resistance of the arrester. Favorable effect by Method B2: It introduces a field probe that allows a compensation for the harmonic currents generated by the harmonics in the voltage. This implies that the method shows low sensitivity to harmonics in the voltage. Method B2: Measurement of resistive leakage current using 3 rd harmonic analysis with compensation for harmonics in the system voltage. Title 45 45

46 IEC : Summary Properties of on-site leakage current measurements: A HV-DC test is effective but off line and complex Method B2 is ranked to be the best field method for evaluation of ageing and deterioration of MOSA. Title 46 46

47 IEC : Summary of Performance Available diagnostic methods: Measurement of total leakage current. Poor sensitivity. Insufficient method. Direct measurement of resistive leakage current. Attractive, but not usable on site. Method B1: 3 rd harmonic analysis of the leakage current. High sensitivity to harmonics in the voltage. Method B2: 3rd order harmonic analysis of the leakage current with compensation. Ranked by IEC as most reliable on site. Title 47 47

48 Deployment of LCM 500 accessories 1 1. Gapless MOSA 2. Insulated base 2 The Field Probe should NEVER exceed this limit 3. Grounding wire 4. Clip-on CT500 i t (t) 5. Counter 3 6. Field probe i p (t) Arrester pedestal 8. Telescopic rod 9. LCM 500 unit Title

49 TOGETHER WE POWER THE WORLD Performance of testing Title 49 First of all connect the instrument to earth CCT should be placed above any counter/a-meters FP should be placed as close as possible to the base of the arresters

50 Leakage Current Measurements Requirements: Separate earth lead & insulated base for each arrester. Electromagnetic field can introduce current in this loop. CCT Short circuit of insulated base will lead to circulating currents in the fundament and the earth lead. CCT CCT = Clip-on Current Transformer Title 50 50

51 Risk Assessment Based on the level and development of resistive leakage current I r over time: 1. Trend analysis over time In general look for increasing trend Baseline reading when the arrester is new. If I r increases by %, this confirms severe ageing 2. Compare to maximum recommended values from arrester manufacturers 3. Compare I r for arresters of the same make and type: The three phases in a line or bay All arresters in the grid 4. Combination of step 1-3 Title 51 51

52 Risk Assessment Steps in the final evaluation: 1. I t and I r are unrealistically high: Circulating currents? Check the insulated base and arrester grounding. 2. I r higher than expected: Temporary heating? Consider to re-test in approx. 1 day to confirm measured value. 3. Confirmed high reading of I r : Monitor continuously or proceed with step Contact arrester manufacturer and consider replacement. Title 52 52

53 Testing Strategy 1. Classify all your MOSAs (name of substation, bay/line and phase, nameplate data (manufacturer, type designation, year/date of commissioning etc.), historical data/failure rates, importance etc.). 2. Establish threshold levels/maximum recommended levels for the resistive leakage current for each arrester type. 3. Define action limits (good condition, satisfactory, retest/monitor continuously, replace). 4. Define measurement regularity (normal, frequent, monitor continuously, after special fault situations). 5. Define verification actions after replacement (laboratory test, dissection/inspection). 6. Evaluate measurements, action limits, regularity of measurements and verification tests to possibly improve the testing strategy. Title 53 53

54 Case studies 1. Measurements at a 420kV GIS 2. Measurements at a Petro-Chemical factory 3. Measurements at an Oil Refinery 4. Power Utility kV Transmission line 6. City substation kV Substation 8. Coastal site kV substation Title 54 54

55 Measurements at 420 kv GIS Substation (1/3) Case 1: 24 arresters, type A, B and C kv The utility wanted to assess the arrester conditions because of surge arrester failures in the past. Max. recommended leakage current values: Type A = 167μA (167 μa = 100%) Type B = 100μA (100 μa = 100%) Type C = 675 μa (675 μa = 100%) Title 55 55

56 Measurements at 420 kv GIS Substation (2/3) 420 kv MOSA at transmission utility Resistive leakage current in percent of max. recommended (100%) Type A: 100% ~ 165 A Arrester number Bay 1 Bay 2 Bay 3 Bay 4 Resistive leakage current in percent of max. recommended Type C: 100% ~ 675 ua Resistive leakage current (ua) Arrester number Type B: 100% ~ 165 ua Bay 5 Bay 6 Arrester number Bay 7 Bay 8 Title 56 56

57 Measurements at 420 kv GIS Substation (3/3) Measurements showed: 1 arrester of type C with app. 375% of max. recommended value 1 arrester of type A with app. 90% of max. Recommended value The rest of the arresters had values from 70% and lower. Conclusion: The two arresters showing the highest values were replaced to reduce the risk of outages. New measurement is recommended in one year. Title 57 57

58 Measurement at a Petro-Chemical factory (1/1) Case 2: 6 arresters, 145kV, installed 1984 Factory owner anxious due to: very high production loss if outages old arresters, condition unknown Measurements showed: 2 units with app. 130%(Ir max rec.= 130µA=100%) 3 units with app. 95% 1 unit with app. 70% Conclusion: All 6 arresters were replaced to reduce outage risk Title 58 58

59 Measurement at a Oil Refinery (1/1) Case 3: 6 arresters, 300kV, installed 1984 Refinery owner anxious due to: old arresters, condition unknown very high production loss if outages Measurements gave: 2 units with app. 60% (Ir max rec.= 130µA=100%) 2 units with app. 50% 2 units with app. 35% Conclusion: All arresters OK New measurements recommended in one year Title 59 59

60 Utility performing routine tests annually (1/2) Case 4: Annually routine testing with LCM for all arresters in the grid Condition monitoring to avoid: Sudden failures blasting of arresters outages Philosophy: Use a simple test to detect bad arresters in service- no outage necessary! Have set max resistive leakage current to 500μA Verification of LCM measurements in laboratory (capacitance, tanδ, IR and dissection) Cooperate with arrester manufacturer to improve arrester design based on measurement experience Title 60 60

61 Utility performing routine tests annually (2/2) 70 surge arresters have been removed from service, based on resistive leakage current measurements with LCM Conclusion: Utility statement: Using LCM with third harmonic resistive current measurement technique is very effective in detecting defective/ aged surge arresters Removed arresters showed: 90% damaged due to moisture ingress 10 % severely aged The problem increased during the rainy season The sealing gaskets were improved and replaced by o- rings in cooperation with the arrester manufacturer Title 61 61

62 Measurements at 110 kv Transm.line (1/2) Case 5: Leakage current measurements as part of the condition based maintenance for 110 kv system Utility is using LCM II for the following purposes: Identify and remove bad arresters Replacing standard leakage current meters which are both ineffective and easily damaged by extreme weather and pollution Measuring philosophy: As a preventive approach, arresters are measured before monsoon with the LCM II Removed arresters are tested in laboratory for verification of high leakage currents (IR testing, waveform analysis etc.) Title 62 62

63 Measurements at 110 kv Transm.line (2/2) Conclusions: 3 arresters have been removed based on leakage current measurements so far All three showed high leakage currents of respectively 293μA, 570μA and 7070 μa!!! Remaining arresters of same make showed low and normal values (<80μA) Standard analog meters in the field were not showing readings in the alarm region Title 63 63

64 Measurements at a City Substation (1/1) Case 6: 18 arresters, 300 kv, majority installed in 1980 The utility was concerned due to arrester failures (arrester explosions) in the past Measurements showed: 3 units with close to 300% (Ir max rec.=130μa=100%) 4 units in the area app % The rest showed low/normal values <50% Conclusion/our recommendations: Replace 3 arresters Monitor 4 arresters closely Measure the rest again in app. 1-2 years Title 64 64

65 Measurements at 420 kv System (1/1) Case 7: 6 arresters, 420 kv, commissioned in 1988 Max recommended resistive leakage current for all 6 arresters are Ir = 165μA (= 100%) Measurements showed: All arresters had Ir values between 37-55% Conclusion: All arresters are considered to be in good condition. New measurements are recommended in one year Title 65 65

66 Measurements at Coastal Site (1/1) Case 8: 6 arresters, 145 kv, commissioned in 2002 Max recommended leakage current for all 6 arresters are Ir = 130μA (= 100%) Measurements gave: All arresters had Ir values between 35-46% Conclusion: All arresters are considered to be in good condition. New measurements are recommended in two years Title 66 66

67 Measurements at 110 kv Substation (1/2) Case 9: 18 arresters, 110 kv measured in 2007 Max recommended leakage current not known baseline established by averaging measurements for all 18arresters Measurements gave: Two arresters had significantly higher readings (230% and 400% respectively Conclusion: The two arresters were taken out of service for laboratory testing. The test showed ingress of moisture that caused internal heating and increase of resistive leakage currents Title 67 67

68 Measurements at 110 kv Substation (2/2) Test of 110kV MOSAs, early 2007 Title 68 68

69 Summary Surge arresters protect valuable assets from overvoltages generated by lightning strikes or switching operations. MOSA will deteriorate over time due to electrical and thermal stress. Leakage Current Measurement Method B2 using third harmonic with compensation according to IEC has been used successfully world wide for surge arrester monitoring. This method is easy and efficient for field application for any make of metal oxide surge arresters. Title 69 69

70 QUESTIONS? Title 70 70

SURGE ARRESTERS AND TESTING. Keith Hill Doble Engineering Company

SURGE ARRESTERS AND TESTING. Keith Hill Doble Engineering Company SURGE ARRESTERS AND TESTING Keith Hill Doble Engineering Company Surge arresters are often overlooked when performing Power Factor tests on transformers, breakers and other apparatus in a substation. Often