Vaisala Oyj Class A · OMXHEX:VAIAS

Vaisala Oyj Class A — Technical Presentation (April 2018)

44 slides from Vaisala Oyj Class A's technical presentation, April 2018

  1. Slide 1: Online DGA-monitoring of power transformers

    Online DGA-monitoring of power transformers Senja Leivo Senior Industry Expert, Vaisala Finland Member of CIGRE A2/D1.47 and D1.52 WGs April 2018 © Vaisala

  2. Slide 2: Webinar content

    Webinar content 1. Dissolved gases – why and which gases 2. Fault types 3. Online monitoring – why 4. Things to consider when choosing a monitor 5. IR technology in online DGA – Vaisala Optimus™ DGA monitor 6. How to define limit values and interpret the data 7. How to utilize the online data – exam

  3. Slide 3: Dissolved Gas Analysis

    Dissolved Gas Analysis  DGA is a comprehensive transformer condition assessment tool  It can detect multiple fault types inside a transformer. Senja Leivo 2018-04-03 © Vaisala 3

  4. Slide 4: Why gases form in a transformer?

    Why gases form in a transformer? Chemical reactions happen when conditions favor them e.g. in presence of fuel, oxygen, heat and catalysts. Oil In case of a fault, temperature increase at fault location. Paper Copper  WHY GASES: oil and/or paper decomposition due to thermal stresses  gases form He

  5. Slide 5: Which gases?

    Which gases?  7 formed gases are commonly considered as key fault gases.  Mathane, ethane, ethylene, acetylene, carbon monoxide, carbon dioxide and hydrogen (CH4, C2H6, C2H4, C2H2, CO, CO2 and H2)  Which gases form depends on  temperature at fault location  materials in contact with the fault a

  6. Slide 6: Fault types

    Fault types  PD partial discharge T1 thermal fault <300°C T2 thermal fault 300-700°C T3 thermal fault >700°C D1 Low energy discharge (sparking) D2 High energy discharge (arching)  S Stray gassing of oil < 200 °C O Overheating < 250 °C C Carbonization of paper > 300 °C Most severe thermal faults wh

  7. Slide 7: Why online monitoring

    Why online monitoring © Vaisala

  8. Slide 8: Why online DGA monitoring?

    Why online DGA monitoring? 1. Online monitoring detects faults at their early phase  Enables corrective actions before severe transformer failure.  Cost reduction in maintenance and repair  Avoid consequential costs of catastrophic failures. 2. Enables safe use of a transformer at its end-of-life

  9. Slide 9: Why online DGA monitoring?

    Why online DGA monitoring?  “On-line monitors are particularly useful for applications where gas formation in electrical equipment must be followed at frequent time intervals (e.g., in strategic or expensive equipment, or where significant faults have already been detected). / Cigre TB409 WG D1.01

  10. Slide 10: Why multi-gas online monitoring

    Why multi-gas online monitoring  “Worldwide experiences using on-line DGA increasingly show that all types of faults − thermal, low-energy discharge, high-energy discharge, and partial discharge − can be identified at an early stage. Identifying type and severity of a developing fault makes timely

  11. Slide 11: Why multi-gas online monitoring

    Why multi-gas online monitoring  Often hydrogen monitors are used as early warning devices. However, not all faults produce significant amount of hydrogen. PD partial discharge, corona type T1 thermal fault <300°C T2 thermal fault 300-700°C T3 thermal fault >700°C D1 Low energy discharge (sparking)

  12. Slide 12: Why multi-gas online monitoring

    Why multi-gas online monitoring  Continuous monitoring of key fault gases gives early and immediate notification of developing faults that may lead to transformer failure.  Many failures can be prevented through real-time correlation of DGA data to e.g. load and oil/winding temperature.  Only com

  13. Slide 13: Things to consider when choosing a monitor

    Things to consider when choosing a monitor © Vaisala

  14. Slide 14: DGA monitors - what is relevant for the user?

    DGA monitors - what is relevant for the user? No monitoring of monitors! No false alarms. Maintenance needs. No leaks. Robust design. Stable long-term measurement for gas trending. Easy to install. Tolerates all climate conditions. Senja Leivo 2018-04-03 © Vaisala 14

  15. Slide 15: Lifetime costs

    Lifetime costs  Concerning the overall costs of a monitor, it’s not only the price of the instrument to be considered, +20% but also the costs associated with the installation and operation in service over its whole lifetime.  Total cost of ownership to be considered when selecting a monitor. Senj

  16. Slide 16: IR technology in online DGA - Vaisala Optimus DGA monitor

    IR technology in online DGA - Vaisala Optimus™ DGA monitor © Vaisala

  17. Slide 17: Infrared absorption (IR)

    Infrared absorption (IR)  IR is a fundamental analysis method, where light is absorbed by gas molecules.  Gases have a unique absorption fingerprint which is used to identify gases present.  Absorbed light intensity is proportional to number of gas molecules present in known gas volume.  Gas spe

  18. Slide 18: Infrared (IR) absorption by different DGA gases

    Infrared (IR) absorption by different DGA gases Senja Leivo 2018-04-03 © Vaisala 18

  19. Slide 19: Infrared technology, operating principle

    Infrared technology, operating principle  Gas molecules absorb infrared light when exposed to it.  More molecules in the gas volume  stronger absorption.  Measured gas can be selected by changing the wavelength of the light with band-pass filters(s).  IR cannot be used to measure gases like H2,

  20. Slide 20: Difference in NDIR Technologies

    Difference in NDIR Technologies (NDIR = non-dispersive infra red) Senja Leivo 2018-04-03 © Vaisala 20

  21. Slide 21: Gas solubility vs gas extraction

    Gas solubility vs gas extraction  How well a gas dissolves into oil depends on:  Oil type  Temperature  Total gas pressure in oil  Gas composition in oil  Gas solubility in oil is defined as Ostwald.  Higher value  more difficult to extract the gas from oil.  Lower value  the easier gas “e

  22. Slide 22: Gas extraction methods

    Gas extraction methods  Methods to extract gas from oil** Naphthenic Paraffinic  Toppler ~100% efficiency Oil density 0.864 0.849  Partial vacuum, 90-99% efficiency H2 0.074 0.036  Head space <30% efficiency O2 0.17 0.18 – Varies by gas type due to Ostwald coefficients and N2 0.11 0.12 gas to li

  23. Slide 23: Partial vacuum gas extraction

    Partial vacuum gas extraction  Vaisala’s monitor uses unique vacuum extraction of gases.  No vacuum pump used, oil column as piston.  Magnetic gear pump  Can extract >95% of the gases in oil.   less dependent on gas gas solubility values Magnetic gear pump Senja Leivo 2018-04-03 © Vaisala 23

  24. Slide 24: Vaisala solutions to prevent long-term drift

    Vaisala solutions to prevent long-term drift Drift mechanism Vacuum Tunable Auto- Optics Hermetic IR- filter calibration vacuum Structure reference clean + temperature Light source Ѵ Filter transmission Ѵ Detector Ѵ Contamination Ѵ Ѵ Ѵ Cross-sensitivity Ѵ Ѵ Ѵ Senja Leivo 2018-04-03 © Vaisala 24

  25. Slide 25: How to define limit values and interpret the data

    How to define limit values and interpret the data © Vaisala

  26. Slide 26: How to define alarm limits?

    How to define alarm limits? 1. Comparing to values from standards – If no other reference is available. 2. Gas trending and gas ratios. 3. Using specific typical values per transformer family. Senja Leivo 2018-04-03 © Vaisala 26

  27. Slide 27: Comparing values to standards

    Comparing values to standards  Typical gas concentration values (ppm) observed in power transformers. Data collected from about 25 electrical networks worldwide and including more than 20 000 transformers (IEC60599 Ed. 2.1 2007):  Typical rate of gas increase (ppm/year): Senja Leivo 2018-04-03 © V

  28. Slide 28: Comparing to values from standards

    Comparing to values from standards  Four-Condition DGA Guide (IEEE C57-104) Senja Leivo 2018-04-03 © Vaisala 28

  29. Slide 29: Gas trending

    Gas trending 2011-2015 Senja Leivo 2018-04-03 © Vaisala 29

  30. Slide 30: Duval triangles

    H2 CO CO2 CH4 C2H4 C2H6 C2H2 Duval triangles 216 237 3657 336 372 100 1.5 PD T1 T3 C T3 Triangle 1 Triangle 5 Senja Leivo 2018-04-03 © Vaisala 30

  31. Slide 31: Duval triangles

    Duval triangles High temperature fault T3, PD with possible T1 carbonization of paper. T3 C T3 Triangle 1 Triangle 5 Senja Leivo 2018-04-03 © Vaisala 31

  32. Slide 32: PD T1 T3

    PD T1 T3 C T3 Senja Leivo 2018-04-03 © Vaisala 32

  33. Slide 33: Using typical values per transformer family

    Using typical values per transformer family  Using typical values per transformer family/design of utility’s own fleet.  Comparing gas levels and gas RoC of those transformers. Transformer H2 CO CO2 CH4 C2H4 C2H6 C2H2 GSU1 23.9 260 1752 16.5 1.8 4.2 0  Useful especially when transformers GSU2 23.

  34. Slide 34: Using typical values per transformer family

    Using typical values per transformer family Transformer H2 CO CO2 CH4 C2H4 C2H6 C2H2 GSU1 23.9 260 1752 16.5 1.8 4.2 0 GSU2 23.3 301 1854 165.5 228.8 97.1 0 GSU3 16.1 357 2316 21.9 8.7 6.1 0 GSU4 28.2 238 1636 18.7 2.5 5.7 0 GSU5 27 499 2845 18.4 2.5 6.1 0 GSU6 42.3 739 3431 19 2.5 4.7 0 Senja Leivo

  35. Slide 35: Utilizing online DGA data - examples

    Utilizing online DGA data - examples © Vaisala

  36. Slide 36: Review data against transformer temperature

    Review data against transformer temperature Senja Leivo 2018-04-03 © Vaisala 36

  37. Slide 37: Review data against transformer loading

    Review data against transformer loading Senja Leivo 2018-04-03 © Vaisala 37

  38. Slide 38: Use of information

    Use of information  If transformer generates gases only during loading peaks or exceptionally high temperature, this information can be used in local loading guidance or as input to more ? efficient cooling.  If gas formation does not indicate severe fault, maintenance actions can be scheduled fur

  39. Slide 39: How to compare online data and laboratory results

    How to compare online data and laboratory results © Vaisala

  40. Slide 40: Accuracy considerations

    Accuracy considerations  “For all methods operating on the headspace principle, a good knowledge of Ostwald or solubility coefficients K is required at all temperatures of gas extraction, in order to calculate gas concentrations in oil (monitor readings) from gas concentrations in the gas phase (ac

  41. Slide 41: Laboratory accuracy, in percentage of the nominal value

    Laboratory accuracy, in percentage of the nominal value Extraction procedure Medium Low concentration concentration Toepler 13 35 Partial degassing 13 30 Stripping 18 23 Headspace 18 37 Mercury-free Toepler 15* 14* Mercury-free partial degassing 11* Shake test 15 44 * Based on a limited number of an

  42. Slide 42: Laboratory points with ±15% uncertainty bars

     Laboratory points with ±15% uncertainty bars.  Vaisala OPT100 graph ±10%.  Laboratory:  Located at power plant.  Oil analysis within hours.  Lab technician takes oil samples. Senja Leivo 2018-04-03 © Vaisala 42

  43. Slide 43: Laboratory points with ±15% uncertainty bars

     Laboratory points with ±15% uncertainty bars.  Vaisala OPT100 graph ±10%. Senja Leivo 2018-04-03 © Vaisala 43

  44. Slide 44: Thank you for joining the webinar.

    Thank you for joining the webinar. Senja Leivo 2018-04-03 © Vaisala 44

Vaisala — Technical Presentation (April 2018) on 6ix