D Proposal for condition monitoring method for cable network Pertti Pakonen, Bashir Siddiqui, Antti Hilden, 19 February, 2015

Transcription

1 D Proposal for condition monitoring method for cable network Pertti Pakonen, Bashir Siddiqui, Antti Hilden, 19 February,

2 Contents of the presentation Goal of condition monitoring tests Priorization of cables for condition monitoring Proposed condition monitoring method for cable network High priority cables Medium priority cables Low priority cables Description of the proposed condition monitoring methods Measurement from primary vs. secondary substation Measurement from phase conductors vs. ground straps Examples of condition monitoring measurements in city and rural networks 2

3 Condition monitoring of 6 36 kv cables Goal of condition monitoring To reveal potential defects in materials and installation during their warranty period To prevent unplanned interruptions caused by deterioration of the cable and cable accessories during operation due to electrical, mechanical and thermal stresses To extend the useful life of cables at the end of their lifecycle To produce condition information for Network maintenance Network operation Planning of network renewals 3

4 Priorisation of the cables The extent and methods of condition monitoring should be defined based on the priority of the cable The priority depends on, among others, Criticality of the cable from the customer point of view Number and importance of the customers fed by the cable Harm due to fault (outage costs) Criticality of the cable from structural and installation environment point of view Cable type and accessories (e.g. resilience to environmental stresses, the amount and degree of difficulty of handcrafting in the installation phase) Installation method and soil type (e.g. ploughing in rocky soil is a risk) Environmental stresses (humidity, salt, impurities of air, temperature, temperature variations) 4

5 Condition monitoring methods for different cable priority classes 5

6 Condition monitoring methods Proposed condition monitoring method for cable network High priority cables Continuous on-line PD monitoring, disturbance records Periodic thermal imaging and ultrasonic mapping of terminations Periodic tan-delta Medium priority cables Periodic on-line PD, disturbance records Periodic thermal imaging and ultrasonic mapping of terminations Low priority cables Thermal imaging of terminations Optional periodic on-line PD Disturbance records 6

7 High priority cables Primary method: continuous on-line PD measurement Secondary methods: periodic thermal imaging, analysis of disturbance records (see Task 6.12) Supporting methods: continuous or periodic ultrasonic monitoring discrimination between internal (harmful) and external discharges in terminations Preferred PD measurement method Inductive PD sensor installed permanently at the cable terminations above the ground strap (see figure) to measure PD pulses in the phase conductor 50 Hz synchronization of the PD data From MV busbar voltage transformer secondary (primary substations) or voltage sensor integrated into the cable termination From the LV phase voltages (if MV measurement is not available) Recommeded installation (least susceptible to interference) Alternative installation (more susceptible to interference) 7

8 Medium priority cables Periodic inspections for example 12, 20, 60, 120, 180, 240, 270, 300 etc. months from commissioning (twice during guarantee period, then 5 year intervals and 2.5 year intervals after e.g. 20 years) Primary method: periodic on-line PD detection, disturbance records Supporting methods: Thermal imaging: detection of discharges in terminations Ultrasonic mapping: discrimination between internal (harmful) and external discharges in terminations Methods to enable periodic on-line measurement Installation of permanent low cost inductive sensors on the cable termination Preferably above the ground strap (see figure on the previous page) to measure PD pulses in the phase conductor or alternatively On the ground strap (more susceptible to disturbances) or modification of the cable termination ground strap routing so that PD sensor can be safely installed without supply interruption. Modification could be done e.g. during scheduled secondary substation maintenance or repair outage. 8

9 Low priority cables Periodic inspections for example 20, 60, 240, 360 etc. months from commissioning (once during guarantee period, then 5 year intervals) Primary method Thermal imaging (+ optionally ultrasonic mapping) of cable terminations conducted in conjunction with regular secondary substation inspections: detection of internal and external discharges in terminations Optional periodic on-line PD detection detection of internal discharges in joints and cable itself Secondary method: Analysis of disturbance records detection of e.g. intermittent earth faults in the cable 9

10 Condition monitoring process High priority cables Medium priority cables Low priority cables Continuous on-line PD monitoring Periodic on-line PD detection Periodic - thermal imaging + optional ultrasonic mapping - optional on-line PD det. On-line PD analyzer Thermal+ultrasonic -PD location and assesment Temporarily installed continuous on-line PD monitoring Cable repair or replacement, commissioning measurements Off-line PD, tan delta Provided by a network construction/ maintenance contractor Provided by network operator and/or a service provider Provided by a service provider 10

11 Condition monitoring process Measurement times (months from commissioning): High priority Medium priority 0 0 Continuous monitoring 12 Continuous monitoring if neccessary Low priority

12 Equipment and implementation of PD measurements 12

13 Continuous on-line PD monitoring equipment and implementation Goal: to detect (and locate) partial discharges in Terminations Joints Cable itself Measurement equipment: Permanently installed PD sensors Inductive (most commonly used in cable condition monitoring) Or capacitive Automatic PD measuring unit Recording at least PD intensity trend PRPD pattern Preferably also PD location information 13

14 Measurement sensitivity requirements Reference: Ossi Bergius, Implementation of on-line PD measurements in MV cable networks, MSc thesis,

15 Potential installation locations for inductive sensors Potential locations for HFCT, around: 1. Ground strap of cable 2. Cable sheath 3. A single phase conductor (the recommended method) 4. A single transformer-busbar cable 5. Ground strap of transformer-busbar cable 6. Transformer ground strap Reference: P.C.J.M. Van Der Wielen, J. Veen, P.A.A.F. Wouters, E.F. Steennis, On-line Partial Discharge Detection of MV Cables with Defect Localisation (PDOL) Based on Two Time Synchronised Sensors, 18 th International Conference on Electricity Distribution (CIRED),

16 Temporary continuous on-line PD monitoring: an example of low cost equipment 4 Channel 2 x HFCT sensors, 2 x TEV sensors Discrimination between cable/switchgear PD PD magnitude monitoring 3G/GPRS data upload to FTP server with 30 min intervals, 6 months local data storage Alerts or alarms from high PD levels PD criticality display with 7 color coded LEDs 16

17 Periodic on-line PD detector measurement - equipment and implementation Goal: to detect partial discharges in Terminations Joints Cable itself Quick and simple measurement for surveying a large cable population with equipment that are easy enough to use for any electrician Measurement equipment: Handheld PD detector indicating PD level Inductive PD sensor (to detect cable PD) TEV (transient earth voltage) sensor and US ( ultrasonic) sensor (to detect PD in terminations and switchgear) Handheld detectors available from HVPD (HFCT, TEV, acoustic, tablet application) IPEC (HFCT, TEV, acoustic) at least one manufacturer for cable measurements (also HFCT sensor) IPEC (TEV, acoustic) several manufacturers for switchgear measurements (TEV and US sensor) EA Technology (TEV, acoustic) Various measurements functionalities: PD level (handheld), PD pattern (advanced) 17

18 Periodic on-line PD analyzer measurement - equipment Goal: to detect, locate and evaluate partial discharges in Terminations Joints Cable itself Measurement equipment: Inductive PD sensors for three phases (HFCT or Rogowski) Oscilloscope capable of sampling 20 ms continuously with a sampling rate of at least 50 MS/s At least 1 Mpoints/channel memory Preferably selectable MHz low-pass filtering (bandwidth limit) Differential probe for 50 Hz phase voltage signal from MV busbar voltage transformer secondary (if possible) LV mains outlet (note the phase (L1, L2 or L3) where outlet is connected and the vector group of the MV/LV transformer) 18

19 Periodic on-line measurement - procedure 1. PD sensor test/calibration, aiming to: check that all sensors and cables work and a correctly connected determine the amplitude correction factor of each sensor store the calibration pulse waveforms (for use in the interpretation of the PD measurement results) 2. Pulse injection test, aiming to: to determine the pulse propagation velocity in the cable if the cable length is known based on the network information system or cable path measurement (or if the propagation velocity is known, to detetrmine the cable length) store a reference waveform of the pulse reflections (joints, cable end) 3. PD measurement, aiming to: to store several (or preferably tens or hundreds of) 50 Hz cycles of 3- phase PD data and 50 Hz phase voltage reference for further analysis 19

20 PD sensor test/calibration: Measurement setup (sensors around the cable termination): R=50 ohm 3 x HFCT PD calibrator Oscilloscope Fig. 1. Sensors around the cable terminations. Fig. 2. On-site demonstration (sensors around the ground straps). 20

21 Pulse injection: HFCT 3 x HFCT Pulse generator (10 Vp-p voltage) Oscilloscope Fig. 1. Diagram of the pulse injection setup. Fig. 2. On-site demonstration (sensors around the ground straps). 21

22 PD measurement: PRPD pattern measured at a rural secondary substation PD measurement setup at a city primary substation Important considerations: Install the sensors at the cable terminations above the ground strap (see figure) to measure PD pulses in the phase conductor make sure that the volts/div setting is high enough so that there are no range overflows in the measurement data (if some of the pulses are clipped, filtering and data analysis results will be degraded) 22

23 Documentation of the PD measurements The following information is proposed to be documented from PD measurement setup: Measuring device + sensor(s), firmware/software version Connection method/place of the sensor Measurement date/time, temperature and humidity (outdoor locations) Network configuration, cable load currents From PD measurements of each phase: Background noise level [pc], approximation between the PD pulse trains Measurement voltage (optional) PD levels [pc] in each phase - largest repeatedly occurring pulse (LROP), max pulse PRPD (phase resolved PD) patterns in each phase PD location(s) in each phase 23

24 Which PD magnitude is harmful? The harmfulness of the discharge depends on, among others: The dielectric material in or on which the partial discharge takes place Magnitude and direction of electric field at the discharge location The gas in which the partial discharge occurs Practical apparent charge limits for XLPE and paper-oil insulated cables: In product standards of cables and cable accessories the following requirements are specified for maximum apparent charges XLPE cables < 5 pc terminations and joints of XLPE cables < 20 pc 24

25 PD measurement system vendors and service providers (underground cable monitoring) IPEC Ltd, UK, HVPD, UK, EA technology, UK, Iris Power LP, Canada, Omicron, Austria, Power Diagnostix Systems GmbH, Germany, Doble Lemke, Germany, SebaKMT, Germany, Dynamic Ratings, Inc., USA, DNV GL (Smart Cable Guard on-line PD monitoring services) Emerson Electric Co, USA, (measurement services) Prysmian Group (Pry-Cam, continuous and periodic on- line PD monitoring) Dekra Industrial Oy (off-line PD, tan delta) 25

26 Thermal imaging 26

27 Thermal imaging Thermal imaging may detect loose contacts e.g. in cable lugs internal and external partial discharges in cable terminations and support insulators/metal oxide arresters (MOAs) Equipment: the existence of PD should be confirmed by electrical PD measurements thermal imaging supports PD location based on electrical PD measurement Thermal camera Resolution: at least 120 x 160 Documentation of the results: Thermal camera model and software version, distance btw. lens and terminations Date and time, temperature and humidity (outdoor locations), cable load currents Thermal image with locations of the hotspots Temperatures and locations of the hotspots in numerical form Average cable termination temperature (+ optionally temperature difference between the hotspots and average cable termination temperature) 27

28 Thermal imaging Important considerations related to thermal imaging: Before making any maintenance or repair decisions based on partial discharges detected, verify that Detected hot-spots are really due to internal discharges and not, e.g. due to External discharges on the surface of the termination, which are usually caused by condensed moisture at the surface of the termination and, according to termination manufacturers, not harmful (if necessary, they can be removed by cleaning and drying the termination surface) This may be verified e.g. based on ultrasonic measurements (internal discharges usually cannot be detected using the horn or parabolic sensor) location of the hot-spots in thermal image (internal discharges are usually located around the end of the insulation screen) the extent of temperature rise compared to surrounding surfaces (in internal discharges this is usually relatively small) Detection of loose contacts requires that there is a considerable (e.g. 50 % of the nominal) load current running through the cable 28

29 An example of external discharges in 20 kv heat shrink terminations Photo: Seppo Suurinkeroinen, Kymenlaakson Sähköverkko Oy Thermal image: Seppo Suurinkeroinen, Kymenlaakson Sähköverkko Oy Visible white oxidation caused by tracking discharges over the dry bands Not harmful according to the termination manufacturer Discharge sites at varying locations Relatively high temperature increase (> 10 C) 29

30 Examples of external discharges in 20 kv secondary substations Photo: Seppo Suurinkeroinen, Kymenlaakson Sähköverkko Oy Photo: Seppo Suurinkeroinen, Kymenlaakson Sähköverkko Oy Surface discharges at the end of a cable termination Surface discharges at the end terminal of a metal oxide surge arrester 30

31 Optional wireless measurements connected to thermal camera Many new thermal cameras support wireless sensors, e.g. Voltage sensors Current sensors Temperature sensors The sensor reading is recorded to the thermal image data Thermal image Thermal image data File Location C:\Fluke\IR IS2 Image Time 12/10/2014 7:31:00 PM Emissivity 0.95 Background Temp 2.0 C Transmission 100% Image Range 2.5 C to 14.8 C Average Temp 4.6 C Severity Camera model Ti110 IR Sensor Size 120 x 160 Camera Manufacturer Fluke Thermography Calibration Range C to C Camera serial number Ti Camera Firmware Compass NE I AAC A AAC A AAC Sensor type: I3000 AC current A A3000 AC current 0-400A Optional measurement data from wireless sensors (max. 5 or 10 sensors) at IR imaging instant 31

32 Ultrasonic mapping 32

33 Ultrasonic mapping Goal: to discriminate btw. two discharge types visible in thermal images Harmful internal discharges e.g. due to problems in stress control at the end of insulation screen (e.g. installation faults) Usually not detectable by ultrasonic measurement Usually less harmful external discharges due to condensed moisture and dry band forming on the surface of the termination Detectable by ultrasonic measurement Equipment: Ultrasonic detector with Center frequency f = 40 khz Bandwidth (-6 db) f = ±2 khz Documentation of the results: Ultrasonic detector model and software version Date and time, temperature and humidity (outdoor locations), cable load currents Location of the highest ultrasonic source in each phase, average and maximum ultrasonic noise voltage in dbµv at those locations, measured at a distance of 1,5 m from the termination 33

34 Ultrasonic mapping Detector types Detector with cone form concentrator Most suitable for measurement distances of 1-2 m Detector with parabolic reflector Most suitable for measurement distances of 5-50 m Detector with parabolic detector Note that depending on the construction of the detector there may be a considerable parallax error between the focus points of the laser pointer and parabolic reflector (ultrasound) if the measurement is done closer to the target than recommended (e.g. < 5 m) Fig. 1. Ultrasonic detector with cone form concentrator. Fig. 2. Ultrasonic detector with parabolic reflector. 34

35 Analysis of disturbance records Disturbance recordings made by feeder terminals or dedicated earth fault indicators may be utilised to detect incipient faults (such as intermittent earth faults) in the network This may give an early warning of the developing fault already before it causes an actual tripping An example of U 0 and I 0 measured from a feeder having an intermittent earth fault. 35

36 Implementation of PD measurements 36

37 Measurement from primary vs. secondary substation Primary substation (PS) Secondary substation (SS) - Recommended + many cables available for measurement (efficient) + often lower noise level compared to PS + MV phase voltages available from busbar PT s for sync. - no potential transformers for 50 Hz sync of PD meas. + warm and dry environment - usually only 1-4 cables available for measurement - often higher noise level compared to SS - usually outdoor environment especially at rural SS PD site PS - PD site =120 m PD Primary substation site SS =1160 m Secondary substation Reflections from secondary substation Direct pulse from PD site Direct pulse from PD site Reflections from primary substation Fig. 1. Measurement from primary substation. Fig. 2. Measurement from secondary substation. 37

38 Measurement from phase conductor vs. ground strap Phase conductor - Recommended + less phase to phase crosstalk and disturbances compared to ground strap Ground strap + lower 50 Hz current - higher 50 Hz current -> HFCT saturation + usually accessible - not accessible in plug-in terminations - more susceptible to phase-to-phase crosstalk and disturbances compared to phase conductor Fig. 1. Measurement from phase conductor, no phase to phase cross-talk. Fig. 2. Measurement from ground strap, same PD appears in all three phases. 38

39 Measurement from phase conductor vs. ground strap Ground strap measurement Identical HFCT s in both. Phase conductor measurement Fig. 1. Full 50 Hz cycle. Fig. 2. Single PD pulse and its reflections. The pulses measured from ground strap are equal to those measured from phase conductor, but opposite in polarity (measurement location: a rural secondary substation). 39

40 50 Hz synchronization for PD measurement from switchgear bay capacitive voltage outputs Comparison for the 50 Hz synchronization for PD measurement capacitive voltage detector outputs of primary substation switchgear busbar phase voltage transformers Both connected to oscilloscope using LeCroy AP031 differential probes Results: Phase difference between the measurements approx. 4.8 degrees (bay capacitive voltage detector output leading) Waveform excellent in both measurements 40

41 Examples of condition monitoring measurements in city and rural networks 41

42 Examples of PD in city network cables (primary substation measurement) Void discharges in cable junction + noise Corona in switchgear (phase L2) Instrument LeCroy WaveRunner LT354 ML Cable 1682 m PILC m XLPE Phase Noise [pc] LROP [pc] Max [pc] L L L Instrument LeCroy WaveRunner LT354 ML Cable 881 m PILC m XLPE Phase Noise [pc] LROP [pc] Max [pc] L L L

43 Examples of internal discharges in cable terminations of a rural secondary substation PD-measurement before repair 20 ms Thermal image before repair partial discharges PD-measurement after repair Discharge sites at the ends of insulation screens Very small temperature increase - approx. 1 C 43

44 An example of the summary documentation of a periodic on-line condition monitoring measurement of a rural secondary substation Date PRPD patterns Secondary substation Bay, network configuration Bay 4 Thermal image ID of the secondary substation Instrument Fluke Ti 32 SDT 170 MD Time 10:04:50 PM 10:15:00 PM Phase d [cm] T [deg C] d [cm] L [dbuv] L L L Thermal image data File Location C:\Mxxx.IS2 Image Time 12/10/ :04:50 PM Emissivity 0.94 Background Temp 2.0 C Transmission 100% Image Range 3.5 C to 16.5 C Average Temp 5.0 C Severity None Camera model Fluke Ti32 IR Sensor Size 320 x 240 Camera Manufacturer Fluke Thermography Calibration Range C to C Camera serial number Ti (9Hz) DSP Version OCA Version Ultrasound level d Hot spot or max. ultrasound L1 568 pc /div L2 568 pc /div L3 227 pc /div PD calibration: 1000 pc = 88 mv PD levels Instrument LeCroy WaveSurfer 44 MXs-B Time 09:59:08 PM Phase Noise [pc] LROP [pc] Max [pc] L L L LROP = largest repeatedly occurring pulse 44

45 Conclusions: proposed condition monitoring methods for cables of different priority Condition monitoring method (P)eriodic/ (C)ontinuous, On-/off-line Cable priority High Medium Low PD monitor c on x c - PD detector p on - x o PD analyzer p on c 1) c 1) c 1) Tan delta (+ off-line PD) p off o 2) - - Thermal imaging p on x x x Ultrasonic mapping p on x x o Visual inspection/camera p on x x x Analysis of disturbance records c on o o o x = mandatory, o = optional, c = conditional, to monitor or to locate and evaluate the harmfulness of PD detected earlier by other means 1) If PD is detected by continuous or periodic monitoring, on-line PD analyzer measurement (or off-line PD measurement) is used to locate and assess the harmfulness of the discharge 2) If cable is repaired, these measurements are conducted before re-energizing the cable, otherwise optional (may be used as an alternative to PD analyzer measurements in noisy environments, requires a supply interruption) 45

46 Acknowledgements: We would like to thank the following distribution network operators for offering us a possibility to make measurements in their networks: Helen Sähköverkko Oy Measurements of several city network cables at three primary substations Tampereen Sähköverkko Oy Measurements of a city network underground cable from primary and secondary substation ends Kymenlaakson Sähköverkko Oy Measurements of rural network cable terminations (external discharges) Kouvolan Seudun Sähköverkko Oy Measurements of rural network cable terminations (internal discharges) 46

47 References: P.C.J.M. van der Vielen, On-line detection and location of partial discharges in medium voltage power cables, Ph.D. Thesis, Eindhoven, Netherlands 2005, Eindhoven University of Technology. 207 p. A. Gerstner, A. Borlinghaus, C. Goy, Integral cable condition assesment, 21 st International Conference on Electricity Distribution (CIRED), Frankfurt, 6-9 June C. Eastham, C. Smith, F-C Chen, Detection and Location of PD in MV cables in electrically noisy industrial environments, 21 st International Conference on Electricity Distribution (CIRED), Frankfurt, 6-9 June B. Lanz, Assuring Critical Cable System Reliability with Effective Diagnostic, ICC Subcommittee F working group F10 meeting March 22, HVPD Technical Guide for PD Levels in MV and HV Cables and Joints, HVPD Ltd. Manchester, UK, P. Nevalainen, P. Pakonen, M. Takala, P. Verho, Inspection of MV underground cables using partial discharge, tan and sheath integrity measurements, Nordic Insulation Symposium, Gothenburg, Sweden, June 15-17, O. Bergius, Implementation of on-line PD measurements in MV cable networks, MSc thesis, J. Vepsäläinen, Mittaavan kunnossapidon hyödyntäminen keskijänniteverkon häiriöiden vähentämisessä ja elinkaarihallinnassa, Diplomityö,

48 References: O. Vuorinen, Using process data in condition based maintenance, MSc thesis, H. Kuisti, J. Altonen, Intermittent earth faults challenge conventional protection schemes, 15 th International Conference on Electricity Distribution (CIRED), Nice, Italy P.C.J.M. Van Der Wielen, J. Veen, P.A.A.F. Wouters, E.F. Steennis, On-line Partial Discharge Detection of MV Cables with Defect Localisation (PDOL) Based on Two Time Synchronised Sensors, 18 th International Conference on Electricity Distribution (CIRED), 2005 R.R. Macinlay, M. Michel, C.W. Walton, Tools for partial discharge testing MV cables and plant, 18 th International Conference on Electricity Distribution (CIRED), S.M. Gargari, P.A.A.F. Wouters, P.C.J.M. Van Der Wielen, E.F. Steennis, Practical Experiences with On-line PD Monitoring and Interpretation for MV Cable Systems, 2010 International Conference on Solid Dielectrics, Potsdam, Germany, July 4-9, 2010 H.M. Pereira, M. Marques, M.B. Pinheiro, R. Palhares, T. Raczy, E.F. Steennis, P.M.F. Almeida, Partial Discharge On-line Monitoring in MV Underground Power Cables as Part of Condition Based Maintenance Strategy, 22 nd International Conference on Electricity Distribution (CIRED), 2013 M. de Witte, Y. Tits, M. Arens, A. Francois, M.Van Den Berg, J. Van Slycken, Partial Discharge Monitoring on MV Switchgear, 22 nd International Conference on Electricity Distribution (CIRED),