IDAX 300 Insulation Diagnostic Analyzer. Dielectric Frequency Response Also known as: Frequency Domain Spectroscopy

Transcription

1 IDAX 300 Insulation Diagnostic Analyzer Dielectric Frequency Response Also known as: Frequency Domain Spectroscopy 1

2 Frequency Domain Spectroscopy Hi V A Lo Ground C HL Measure at several frequencies Use Ohms law: C L C H Z U I Z C, tand,pf and 2

3 Capacitance and Dissipation Factor (Tan ) = Z (Impedance) j = C (Capacitance) = Tan Loss tangent) = Power Factor (cos or ) Note: If cos and Tan small then cos =Tan If Tan is 1*10-3 (0.001) then I loss /I is 1/1000 which is equivalent to I loss will be1m and I 1000m, 1m / 1000m. Specification of instrument is 1*10-4 (1/10000). I 10nF, 200V and 50Hz I CAP = 2*Pi*f*U*C = 0.63mA I 1nF, 200V and 1Hz I CAP = 2*Pi*f*U*C = 1.26uA 3

4 IDAX Set-ups IDA 200 UST - without Guard Hi V C X C s1 A 1 Lo C s2 A 2 Ground 4

5 IDAX Set-ups IDA 200 UST - with Guard Hi V C X C s1 A 1 Lo C s2 A 2 Ground 5

6 IDAX Set-ups IDA 200 GST - without Guard Hi V C X C s1 A 1 Lo C s2 A 2 Ground 6

7 IDAX Set-ups IDA 200 GST - with Guard Hi V C X C s1 A 1 Lo C s2 A 2 Ground 7

8 What is Spectroscopy? Method to isolate/identify building blocks in a composite material Example: Identify material composition in samples from Mars Example: Breaking down light into its different colors using a prism 8

9 Insulation testing/dielectric response methods FDS/DFR HV Tan Delta VLF PDC Polarization Index "DC" 1 0 0, , ,0001 0,001 0,01 0, Frequency, Hz 9

10 Dielectric Frequency Response - Power Factor Changes with Frequency Power factor 0.32 at 0.02 Hz at 60 Hz Frequency 10

11 Frequency Domain Spectroscopy Changes in insulating materials (ageing) affect the capacitance and loss factor (PF, tan ) Frequency sweep, compared to traditional onefrequency Power Factor/ Doble test, provides a lot more information on: Insulation characteristics Ageing effects Influence of temperature Etc 11

12 Traditional Power Factor Testing Dissipation factor Frequency 1 mhz 50 Hz 1kHz 12

13 Dielectric Frequency Response Dissipation factor Frequency 1 mhz 50 Hz 1kHz 13

14 - Moisture + - Moisture + What affects the response? - Oil Conductivity + - Temperature + 14

15 Typical power factor values for oil insulated transformers and bushings Typical power factor 20 C "New" "Old" Warning/alert limit Power transformers, oil insulated % % > 0.5% Bushings % % > 0.5% IEEE states; The power factors recorded for routine overall tests on older apparatus provide information regarding the general condition of the ground and inter-winding insulation of transformers and reactors. While the power factors for most older transformers will also be <0.5% (20C), power factors between 0.5% and 1.0% (20C) may be acceptable; however, power factors >1.0% (20C) should be investigated. 15

16 Dielectric Frequency Response - Single tan delta 0.7% is not enough to make the right decision - Dielectric Frequency Response tells the story! Dry transformer with old oil (high conductivity) Wet transformer with good oil Same PF value at 60Hz 16

17 DFR Application Areas Power transformers Instrument transformers Bushings Motors and generators Cables Generic testing of insulation systems 17

18 Why measure moisture? A transformer with low moisture content is like a person in good condition A transformer can be loaded with confidence without risk for catastrophic failure. A person can work hard without risk for heart attack A wet transformer is like an overweight person with clogged arteries. The transformer owner has to limit load to avoid bubbling (explosion risk). Moisture in insulation increases the rate of aging 19

19 Moisture in Power Transformers Power transformer insulation consists of oil impregnated cellulose and free oil. Almost all moisture is in the cellulose: 25 tons of oil with water content of 20 ppm = 0,5 kg 2.5 tons of cellulose with 3% water content = 75 kg 20

20 Water In Oil Analysis Power transformer insulation consists of oil impregnated cellulose and free oil. Almost all water is in the cellulose: 25 tons of oil with water content of 20 ppm = 0,5 kg 2.5 tons of cellulose with 3% water content = 75 kg 20ppm (parts per million) in 25 tons of oil 3% water in 2.5 tons of cellulose 21

21 Moisture in Power Transformers Power transformer insulation consists of oil impregnated cellulose and free oil. Almost all moisture is in the cellulose: 25 tons of oil with water content of 20 ppm = 0,5 kg 2.5 tons of cellulose with 3% water content = 75 kg Moisture content in oil (it is almost constant in the cellulose) varies with temperature and oil aging status: Aged oil resolves higher amounts of water than new oil Small moisture concentration makes sampling difficult 22

22 Water In Oil Analysis Oil samples taken at low temperatures have low accuracy because the water has migrated to the paper Oil sample taken at 20 0 C 4.0% water = 6 ppm 1.0 % water = 3 ppm 23

23 Moisture in Power Transformers Power transformer insulation consists of oil impregnated cellulose and free oil. Almost all moisture is in the cellulose: 25 tons of oil with water content of 20 ppm = 0,5 kg 2.5 tons of cellulose with 3% water content = 75 kg Moisture content in oil (it is almost constant in the cellulose) varies with temperature and oil aging status: Aged oil resolves higher amounts of water than new oil Small moisture concentration makes sampling difficult Moisture changes the dielectric properties of the cellulose paper/pressboard 24

24 Water Changes Paper Water changes the insulation properties of the paper Relative permittivity 25

25 Moisture in Power Transformers Power transformer insulation consists of oil impregnated cellulose and free oil. Almost all moisture is in the cellulose: 25 tons of oil with water content of 20 ppm = 0,5 kg 2.5 tons of cellulose with 3% water content = 75 kg Moisture content in oil (it is almost constant in the cellulose) varies with temperature and oil aging status: Aged oil resolves higher amounts of water than new oil Small moisture concentration makes sampling difficult Moisture changes the dielectric properties of the cellulose paper/pressboard Moisture accelerates ageing 26

26 Moisture in Power Transformers Moisture accelerates ageing Dry (0.5%) 90 0 C = 40 Years Medium wet (2.0%) 90 0 C = 4-5Years 27

27 Moisture in Power Transformers Power transformer insulation consists of oil impregnated cellulose and free oil. Almost all moisture is in the cellulose: 25 tons of oil with water content of 20 ppm = 0,5 kg 2.5 tons of cellulose with 3% water content = 75 kg Moisture content in oil (it is almost constant in the cellulose) varies with temperature and oil aging status: Aged oil resolves higher amounts of water than new oil Small moisture concentration makes sampling difficult Moisture changes the dielectric properties of the cellulose paper/pressboard Moisture accelerates ageing Moisture limits loading capability 28

28 Hottest Spot Temperature ( O C) Moisture in Power Transformers Moisture determines the maximum loading/hot-spot temperature for bubble inception (see e.g. IEEE Std C ) Knowing moisture content allows for correct decision Leave as-is Dry-out Replace Scrap or Relocate? % moisture (New) = C hotspot 3.0% moisture = C hotspot Moisture in Insulation (% wt) 29

29 Interpretation of moisture content < 0.5 % % % 2.5-4% > 4% New transformer Dry insulation Medium wet insulation Wet insulation Very wet insulation Interpretation of moisture content of solid insulation (% of weight water per weight cellulose): 30

30 Moisture levels 1.0 % 2.6 %? 4.2 % 31

31 Moisture estimation process Measure tan delta/power factor from 1 khz to 2 mhz Send results to MODS Enter insulation temperature (top-oil temperature) MODS matches measured curve to modeled curve (automatically) by varying parameters that affects the shape of curve to find best match Results: Relation of solid (cellulose) vs. liquid (oil) insulation between winding (if not known) Moisture in solid insulation Oil conductivity 32

32 X-Y model of power transformer insulation Cellulose: Blue color Oil: Red color Typical values: X = % barriers in the main duct (15-55%) Y= % spacers of the circumference (15-25%) 33

33 Oil paper and combination 34

34 Moisture estimation process Right click Measured curve Select Send to MODS 35

35 Moisture estimation process % Spacers Amount of oil % Barriers Model curve Measured curve 36

36 Moisture estimation process 3. Click Auto match 1. Check all checkboxes 2. Enter temperature 37

37 Moisture estimation process Result! 38

38 Why a Sense lead? Ideal situation 39

39 Why a Sense lead? Actual situation R 40

40 Why a Sense lead? Actual situation R 41

41 High frequency CHL measurements Results 2-w CHL measurement CHL=5729 pf CL=8755 pf Single-wire ampmeter connection (IDA, Dirana) Tan delta=1.3 % Tand delta=0.56% 2-wire ampmeter connection (IDAX300) 42

42 High frequency CHL measurements Issues Hi V A Lo Ground C HL 1. Ampmeter input + cable Z is small but not zero... C L C H 3. Leakage current causes measurement error at high frequencies 2. Z is high but not infinite... 43

43 High frequency CHL measurements Solution Hi Compensates for Z and sets potential to ground V A Sense/Voltage output Lo Ground C HL 1. Ampmeter input + cable Z is small but not zero... C L C H No leakage current causing measurement error at high frequencies! 2. Z is high but not infinite... 44

44 Noise in substations Induced AC (50/60Hz) Induced DC (HVDC stations) DC offset from ground potential Other disturbancies (RF, harmonics etc) 45

45 Noise in substations - AC 88 randomly selected IDA/IDAX measurements Noisy limit set to 50µA About 10-20% of all measurements are performed at > 50µA noise current Noise level, db rel 10µA 40,0 20,0 0,0-20, ,0-60,0 Sorozatok1-80,0-100,0-120,0-140,0 46

46 Bushings and rotating machinery 47

47 Bushing Failures 48

48 DFR on Bushings General Electric, Type U Bad condition Good condition 49

49 High-Voltage DFR on Bushings ( kV) 50

50 DFR for Motors and Generators Measure tan delta vs frequency to detect ageing deterioration (end-of-life) 51

51 Verification of DFR/IDAX performance Accredited test system at ABB transformers Project REDIATOOL comparing DFR with other methods for measuring moisture in transformers. Results published at CIGRE 2006 (paper D1-207). Examples of other independent verification bodies IREQ (Canada) EPRI (China) IEPS/Schering-Institute (Germany) CESI (Italy) KTH (Sweden) Tennesse Tech (USA) 52

52 ABB Transformer Diagnostics (press release) 53

53 IDAX 300 Insulation Diagnostic Analyzer 54

54 IDAX 206FR Insulation Diagnostic Analyzer 55

55 IDAX System HW Test signal: V peak 0 50 ma peak Frequency: Hz Hz (IDAX-300) Sample range: 10 pf 100 µf 2-ch measurement (IDAX-300) SW IDAX SW for measurement control and analysis Moisture analysis SW (MODS) Automatic moisture assessment in oil impregnated cellulose 56

56 Cables 57

57 IDAX Software Object and Test Browser Nameplate Info 58

58 IDAX Software Test settings Graphical Hook-Up Diagram 59

59 Software Graph View Presentation selector Power Factor Chosen 60

60 IDAX Software Result at 50/60Hz 61

61 Summary IDAX 300 Instrument for moisture assessment in transformer insulation Built in modeling software for quick and reliable results Measures the moisture where it is, at any temperature Light weight, fast and accurate Can also be used for other applications CT s Bushings Rotating machines Cables Reveals the reason for high tan delta values - Moisture or contamination? Well documented and well proven method 62

62 Cable Diagnostics 63

63 Paper Insulated (PILC) Cables Moisture in cable insulation is generated by ageing due to water ingress from outside accelerates the ageing process Presence of elevated moisture content can be detected by DFR (IDAX) measurements 64

64 Cable Paper Effect of Moisture 10 1 Tan mc < 0.20 % mc = 1.60 % mc = 1.99 % mc = 2.62 % mc = 3.50 % mc = 4.20 % 50 Hz Hz Tan minimum frequency, [Hz] 65

65 lo s s ta n g e n t, ta n lo s s ta n g e n t, ta n Examples of Field Measurements Measurements performed at 200 V peak on field installed cables in Sweden 10 0 H T K, L 1 B e kö, f3 R a v, f2 S kra d d, f P e d e r, f1 L , f1 R å d m g 2, f fre q u e n cy, [H z ] fre q u e n cy, [H z ] 66

66 Estimation of Moisture Content Minimum Estimated moisture PF/tan content Condition Below 1% Good % Moderately aged/moistened % Considerably aged/moistened Above 0.01 Above 3.5% or a local In bad condition/high moisture defect content 67

67 Summary: Paper Insulated Cables Moisture gives a characteristic increase of the frequency dependent capacitance and loss of cellulose paper Temperature affects the frequency response of cellulose paper By using measurement data in a frequency interval accurate temperature corrections are possible Loss tangent minimum can be used as a criterion for assessment of moisture content 68

68 Medium Voltage XLPE Cable Circuit Problems Terminations Manufacturing defects Bad mounting Aging Joints Manufacturing defects Bad mounting Aging Cables Manufacturing defects Damaged during installation Aging (water trees) 69

69 After installation diagnostic measurements and voltage tests Terminations Partial discharges (acoustic or electrical) Voltage test (not very effective) Joints Partial discharges (acoustic or electrical) Voltage test (may be effective) XLPE cables Voltage test (Partial discharge measurements) 70

70 Voltage tests DC Not effective in most cases. Using high voltages may damage the cable 50 Hz Requires very large equipment VLF (very low frequency, 0.1 Hz) Cost-effective, moderate sized equipment 71

71 Water Tree Deterioration of MV XLPE Cables Water trees are growing in the insulation and lower the electrical withstand The water tree aging process is very slow A heavily aged cable fail if the insulation stress is increased (lightning impulses, faults, etc) 72

72 Medium Voltage XLPE Cables High voltage tests shorten cable life significantly Non-destructive diagnostics should be used Water-treeing can only be detected with increasing voltage level (non-linear effect) Use IDAX + VAX high-voltage unit! 73

73 Measurement Procedure 1. Measurements of short frequency sweeps, from about 10 Hz down to 0.1 Hz at 25, 50, 75, 100 and (repeat) 50% service voltage level, U 0 2. Classification of cable quality 74

74 Cable quality classification 1. Low Losses and No Voltage Dependence 2. VDP Response (Voltage Dependent Permittivity) A voltage dependent increase of capacitance and loss. 3. TLC Response (Transition to Leakage Current) A VDP response at initial low voltage levels. At a higher voltage level, the response characteristics changes due to leakage current. 4. LC Response (Leakage Currents) Leakage currents through water trees are present already at low voltage levels. 75

75 New/non-deteriorated XLPE cable: Low losses and no voltage dependence Capacitance part Loss (Tan ) part 10-1 ' 3 kv 6 kv 10-1 '' 3 kv 6 kv ,01 0, Frequency (Hz) ,01 0, Frequency (Hz) 76

76 VDP Response: Voltage dependent increase of loss and capacitance Capacitance part Loss (Tan ) part ' ' 3 kv ' 4,5 kv ' 6 kv ' 3 kv ' 6 kv ' 3 kv '' '' 3 kv '' 4,5 kv '' 6 kv '' 3 kv '' 6 kv '' 3 kv ,01 0, Frequency (Hz) ,01 0, Frequency (Hz) 77

77 TLC Response: Leakage current detected at increasing voltage level + memory effect Capacitance part Loss (Tan ) part ' ' ' ' ' ' 1,5 kv 3 kv 4,5 kv 6 kv 3 kv " " " " " " at 1,5 kv at 3 kv at 4,5 kv at 6 kv at 3 kv ,01 0, Frequency (Hz) ,01 0, Frequency (Hz) 78

78 LC Response: Leakage current (through water-trees) already at low voltage levels Capacitance part Loss (Tan ) part ' ' ' ' at 3 kv at 6 kv at 9 kv " " " " at 3 kv at 6 kv at 9 kv ,01 0, Frequency (Hz) ,01 0, Frequency (Hz) 79

79 Evaluation of Water-Tree Deteriorated Cables based on laboratory and field experience LC or TLC response The cable is judged bad. The voltage withstand level is usually below 2.5 times service voltage level. Depending on the leakage current level, cable design and the voltage level of the network, the cable can be used for a short time or must be replaced immediately. VDP response (Significantly aged) The cable is significantly aged. The voltage withstand is typically times the service voltage level. Depending on cable design and loading, the cable can remain in service for several years or should be scheduled for replacement. No ageing detected The cable is judged good and has typically a voltage withstand above 4 times service voltage. However good condition does not necessary mean that the cable does not have any water trees. Repeated measurement is recommended within a 5-10 year period. 80

80 Experience from a project No cable judged good has been reported to fail in service Significant aged cables, i.e. cables with VDP response, can withstand many years of service life without failure Cables with TLC and LC currents are bad. Such a cable may be allowed remain in service a few months in 6 or 10 kv networks 81

81 Case Study: North Botkyrka Approximately 60 cables circuits installed in early seventies Increasingly rate of cable failures Cable circuits failed during VLF testing In 1996, all circuits were measured and LC and TLC cables were replaced (only a few) cable circuits were replaced based on level on VDP response 82

82 Faults/100 km/year Case Study: North Botkyrka No faults One fault faults summer All faults were in cables with strong VDP-response in 1995 or A few circuits was remeasured autumn The response of slightly aged cables indicated that they still are in good shape (6-7 years later) Cable Faults in North Botkyrka Year 83

83 Summary: Using DFR on MV XLPE Cables Voltage test may damage water tree deteriorated XLPE cables Relatively low voltage levels (up to service voltage) are used in order to ensure non-destructive measurements The process of water tree deterioration is very slow By non-destructive measurements cable replacement can be postponed and scheduled The response of water trees can be identified and classified 84