Introduction to PD

October 2014

Presented by:
Dr Lee Renforth,
Managing Director
HVPD Ltd
 CONTENTS

• PD detection theory
• PD test & monitoring equipment
• Cable PD – detection & location
• Local PD in switchgear
• PD in rotating machines
• Continuous PD monitoring
 Introduction to HVPD – Our global presence

Introduction to HVPD Ltd

• HVPD are experts in the growing industry of on-line partial discharge (OLPD) condition
monitoring and condition based management (CBM) of high voltage networks.

• We supply portable and permanent OLPD surveying, diagnostic test and continuous monitoring
solutions, and a complimentary range of on-site services, monitoring services and training.

• Over 350 customers in 100 countries trust our technology.

 Our Applications Knowledge Base

500+ test and monitoring OUR EXPERTISE

projects completed over
the past 20 years “Our Knowledge is Your Power”
 Industry-Specific Condition Monitoring Solutions

Oil & Gas

Introduction to HVPD Ltd
Offshore Renewables Transmission & Distribution

Power Generation Shipping

 Introduction to Partial Discharge

What is partial discharge? Why test for partial discharge?

“A localised electrical discharge that only PD activity is an indication of an incipient fault

partially bridges the insulation between in HV insulation and is widely regarded as the
conductors and which can or can not occur best ‘early warning’ indicator of the deterioration
adjacent to a conductor” of high voltage insulation.
IEC60270 Definition
 Common Applications of PD Testing

Power cables,
Switchgear
joints/splices and
(Air, Solid, Gas-insulated)
terminations

Power transformers Motors and generators CT’s and VT’s

and bushings
 The Drivers for Applying OLPD Testing

• Safety – mostly with switchgear, outdoor HV plant and cable sealing ends

• Ageing population problem – life extension, delaying capital replacement

• Condition-Based Maintenance (CBM)

• Avoid unplanned outages and improve network reliability

 Why and When to Perform PD Testing
New Equipment

At Manufacture

• Quality Assurance
• Type/routine tests, e.g. IEEE/IEC
standards – test to less than 5 pC on
the cables

At Commissioning

• To check for transport damage

• To ensure the installation of the cable
accessories have made to a good
standard (these are the weak points
in the cable system)
 OLPD Testing During Service?

• To get baseline PD readings

• To evaluate insulation quality

• To locate PD activity sites and target repair

• To avoid costly / unplanned outages

• To support Condition-Based Management (CBM) regime

 From ‘Cradle to Grave’ PD Testing and Monitoring Philosophy

Damage

Mistake

Repair
Aging
Manufacturing Transportation Installation Operation
Power frequency 50/60
Hz

Acceptance

Continuous
Monitoring
Factory
Testing

Power

Testing
frequency
50/60 Hz
OLPD DETECTION THEORY
 7 Main Types of PD
Internal

Sharp, irregular surface Tree growth in

Void in insulation
on conductor insulation
External

‘Floating’ metalwork Corona from sharp objects Discharges from induced Surface
near conductors at high voltage voltages onto sharp discharges
points at ground
 PD Equivalent Capacitance Circuit – ABC Model

Void/cavity

Surface Discharge Corona

 Voltage Breakdown of Cavity of Surface Area (Cc)

Vb
Va
Vc

Theoretical

Actual current pulses

(stochastically occurring)

V+, V- = PD inception voltages

 Paschen Curve for Breakdown of Air in Uniform Field

Paschens Curve for breakdown of air in a uniform field

From T.W. Dakin et al “Breakdown of gases in

uniform fields" Electra, Vol 32, pp 61-82, 1974.
 Internal Discharges

Internal discharges can be caused by:

• Voids, cavities, delaminations in solid or liquid insulation

• Electrical trees form from these voids

PD Damage to Cables – Treeing on 66 kV Paper Cable
Insufficient Mastic Around 33 kV XLPE Connector
Surface Discharges on 24 kV Cable Elbow Termination
 PD Damage to Cables
Surface Discharge on Dry-Type Termination
 Corona Discharges
275 kV Insulator String

• Incepted from sharp points on HV conductors

• Also possible from sharp points on ground

Corona Discharges
400 kV Arcing Horn
 PD Energy and Detection Methods

ENERGIES RELEASED DETECTION METHODS

Electrical Charge Capacitor/Inductive (HFCT)

Electromagnetic Wave TEV, VHF/UHF Couplers
PD
Optical Low-light/UV Cameras
Acoustic Wave Contact/Airborne Acoustic
Gaseous by-products Dissolved Gas Analysis
Chemical by-products Visual inspection
(such as white deposit
oxides)
Ozone Smell / Ozone Detector
 Examples of OLPD Measurements / Results

Internal PD in bushing
15 0

PD Magnitude (nC)
10 0
5 0

• PD magnitude
0 0
-5 0
-10 0
-15 0

• PD count (number of PD
0 90 180 270 360
Phase (Degree)

pulses per power cycle) Surface discharge in termination

20

PD Magnitude (nC)
• Cumulative PD activity
10

0
-10

-20

• Phase Resolved PD 0 90 180

Phase (Degree)
270 360

(PRPD) Patterns
• PD monitoring over time
OLPD TEST & MONITORING EQUIPMENT
 PD Testing Approach

TEMPORARY/
SPOT-TEST
PERMANENT

CONTINUOUS
DETECTION LOCATION
MONITORING

Off-line / On-line On-line

 Detection and Location/Diagnostics

Phase 1 – Detection
• Simple equipment
• Initial indication of PD level/severity

HVPD PDSAir™
Handheld PD Surveying Tool

Phase 2 – Diagnostics/Location
• More advanced hardware/noise rejection
• PD diagnosis and location within the
cable/plant
• Digital PD detector with PC for analysis and
reporting
HVPD Longshot™
Diagnostic OLPD Test Unit
 Continuous Monitoring Aspects

Phase 3 & 4 – Temporary/Permanent Monitoring

• Temporary or permanent hardware

• Trends PD over time and captures any trend to failure

• Web-based UI/SCADA alarms

HVPD Multi™
• Simple and advanced options: Portable Monitor

- Simple hardware without diagnostic capabilities, generates

alarm signals only
- Advanced hardware with diagnostic capabilities

HVPD Multi™
Permanent Monitor
CABLE PD – DETECTION & LOCATION
 OLPD Detection Theory for Cables
Electrical Charge

• PDs are incepted by the high Available Waveform Display

voltage applied to cable. 0.6

0.4
• PD pulses are short duration 0.2
impulses (ns – µs) that propagate in

Chan 1
0

both directions away from PD site -0.2

between cable core and sheath. -0.4

-0.6

• Signals can be detected on both the -0.8

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
core and earth screen at terminations. Time (mSec)

End A End B
 Example of PD Against Phase for Power Cables

Three Core Belted Cable with HFCT

Single Core Cable
on Common Screen
Available Waveform Display Available Waveform Display

0.6 0.04
0.03
0.4
0.02
0.2 0.01
Chan 1

Chan 1
0 0
-0.2 -0.01
-0.02
-0.4
-0.03
-0.6 -0.04
-0.8 -0.05
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Time (mSec) Time (mSec)
 PD Location Theory
Formula for PD Location

Direct pulse T = Time difference between direct

Reflected pulse and reflected pulses.
L = Cable Return Time for cable

T L

L T

PD Pulse Return Speed: VPD = Cable Length/Return Time

  T 
PD Site Location: PD %   1    100
  L 
 OLPD Location on Power Cables

In many on-line cases reflected PD pulses are often not visible:

• Attenuation is too large to measure reflected pulses from the far end (long
cables)
• Waveforms too difficult to interpret (noisy signals)
• Teed or jointed cables
• Cables with many ring main units or switches
• Cables with no change in impedance at the far end
• Cross-bonded cable circuits – multiple reflections
 Transponder for On-line PD Location

• Portable Transponder installed at remote

cable end Input

• Compensates for lack of reflected pulse

• Detects PD pulse with Discharge Trigger
Unit and re-injects large pulse back into Output
cable with Pulse Generator
• Connects to cable with HFCTs
 On-line PD Location on Power Cables
Example of Usage

Measurement End Remote End

Portable
Transponder

HVPD-
Longshot™

Without Reflection may not be clearly visible (e.g.

Transponder due to noise)
T

With The large transponder pulse removes

Transponder any confusion

T Ttr

T = Time between direct and reflected PD pulses

Ttr = Transponder time delay
On-line PD Location on Power Cables
Example Results

Reflectogram showing PD and transponder

pulses

PD location map for all PD pulses in

cable section under test
LOCAL PD IN MV SWITCHGEAR
 Example Use of Detection and Diagnostics
Transient Earth Voltage (TEV) Theory

Metallic switchgear casing

Partial discharge site

• PD site within switchgear: Phase–Earth discharge

• EM signals radiate from PD site and couple onto metal-clad housing
• Signals emerge on the outer surface at openings in metal housing
Example Use of Detection and Diagnostics
Transient Earth Voltage (TEV) Theory
PD IN ROTATING MACHINES
 Example Use of Detection and Diagnostics
Rotating Machines

Important to identify type of

PD in machine for severity:
• slot section,
• delamination,
• end windings etc.

PD Sensors

HVPD Longshot™
Diagnostic Test Unit
 Example Use of Detection and Diagnostics
Rotating Machines

Machine PD
Phase-Earth PD on 15,000

PD Magnitude (pC)
RYB phases
• Signals captured 10,000
5,000
synchronously from 0
-5,000
sensors on each phase. -10,000
-15,000

• Identification of Phase- 0 90 180 270 360

Phase of Pow er Cycle (de
Earth PD and Phase-
Phase PD.
Machine PD
• Phase Resolved PD 40,000

PD Magnitude (pC)
Phase-Phase PD
(PRPD) patterns indicate (Y-B) 20,000

the defect type. 1,000

Machine PD Waveform
0
500
Volts (mV)

0
-20,000
-500
-1,000 -40,000
0 1 2 3 4 5 6
Time us
0 90 180 270 360
Phase of Pow er Cycle (de
CONTINUOUS OLPD MONITORING
 Continuous Monitoring
Trends in Usage

• Important/high-value/critical plant items

• Plant with known PD variations in time – paper cables, rotating machines,

AIS/GIS.

• Plant with high PD identified in spot-tests

 Continuous Monitoring
PD Rise to Failure

PD rise to failure over 100 days PD rise to failure over 20 days

 Continuous Monitoring
PD Variation with Load on PILC Cable

400

350

300

250

200

150

100

50

PDPDburns
burnsinina ahot cable:
cold cableelectrodes expandfluid
(90% of cases): - possible movements
shrinks, inside
voids appear, accessories
local lead to
PD in voids.
increased field strengths in dielectrics – PD in accessories.
 Remote OLPD Monitoring of Ex/ATEX HV Motors
in Hazardous Gas Zones

PD activity within the MV switchgear PD activity within the MV motor stator slot section

Safe MV Feeder Cables Ex

Measurement Range:
Environment Up to 1.0 km for PVC / PILC Environment
(no restrictions Up to 2.0 km for XLPE cables (restrictions
apply) apply)

PD activity in the MV motor stator end windings

PD activity within the MV feeder cable
HVPD’s Network OLPD Monitoring System User Interface - Oil Refinery
OLPD Monitoring System User Interface - Generator and Transformer
OLPD Monitoring System User Interface – 11kV Switchboard
 Case Study 1: On-line Cable PD Mapping,
Excavation and Repair on an 11 kV PILC Cable
(EDF Energy, London, UK)
 Case Study: On-line Cable PD Mapping, Excavation
and Repair on 11 kV PILC Cable
2,000

1,800
1,600

Location of
1,400

INCREASING PD ACTIVITY

All Phases PD
1,200
1,000
discharge ON-LINE MAPPING
800
600

400
200
0
0 100 200 300 400 500 600 700 800 900 1,000
Location (meters)

Location of
discharge OFF-LINE MAPPING

STAGE 1: On-line monitoring

0m
154 442 840 904 1009m
0.3 Cu 300 CAS 0.3 Cu 0.3 Cu 300 CAS 0.3 Cu
Primary Sub 1
Sub

OVERLAY

DISCHARGING AREA

NO DISCHARGES
FOLLOWING RE-
ENERGISATION
Case Study 2: OLPD Testing, Location, Monitoring
with Preventative Maintenance on a 33 kV Land-
Sea Offshore Wind Farm Export Cable
 Case Study: Circuit Details

• 1.7 km single core XLPE land cable

• 9.6/11.5 km 3 core XLPE subsea cable
 Case Study: OLPD Test and Mapping Data

L1 L2 L3

Cable PD Cable PD Cable PD

10,000

PD Magnitude (pC)
PD Magnitude (pC)

PD Magnitude (pC)
5,000

0 0 0

-5,000

-10,000

0 90 180 270 360 0 90 180 270 360 0 90 180 270 360

Phase of Pow er Cycle (deg) Phase of Pow er Cycle (deg) Phase of Pow er Cycle (de

High levels of PD (of up to 10,000 pC / 10 nC)

measured on Circuit B, Phase L3.
 Case Study: PDMap© Graph Showing PD Location

Switching
Substation Land-sea
Transition
Joint Pit 7
Joint

0 200 400 600 800 1,000 1,200 1,400 1,600

Location (meters)
 Case Study: PD Signals Before and After Joint Replacement

BEFORE

High PD detected on L3

PD Located

0 200 400 600 800 1,000 1,200 1,400 1,600

Location (meters)

Lower-level sporadic PD
Joint 7 with PD signals from different site
removed and after joint replacement
replacement cable
section installed

AFTER
Case Study: Circuit B – Evidence of Surface Degradation
Due to Bad Fitting Heatshrink Stress Control
CONCLUSIONS
 Conclusions

• OLPD testing is widely adopted for testing many types of MV &

HV plant.
• The technologies enable the assessment of the health of
MV/HV network with minimal disruption and cost.
• Equipment is tested under both normal (and abnormal) working
conditions.
• PD can be detected, located and monitored, without the need to
de-energise the plant.
• OLPD testing is an essential tool for the effective
implementation of condition-based management (CBM)
techniques to MV and HV power networks.
 End of Presentation

Thank you for your time

Q&A?
Overview of On-line and Off-line PD Test Methods
 CONTENTS

• On-line & off-line testing methods

• Overview of PD sensors
• MV/HV cable AC withstand voltage commissioning
testing options
• Case studies
 PD Test Methods

On-line
• In-service, under normal working conditions
• Various sensor options

Off-line
• Energised with external supply
• Usually HV Coupling Capacitor sensor used
On-line vs Off-line
ON-LINE TESTING METHODS
 OLPD Sensors

High Frequency Current Transformers (HFCT)

Detects PD in cables and remote plant (e.g.
transformers/ rotating HV machines).

High Voltage Coupling Capacitor (HVCC)

Mainly applied to the PD monitoring of rotating HV
machines.

Transient Earth Voltage (TEV)

Detects local PD within plant under test.

Airborne Acoustic (AA)

Detects airborne PD signals with direct line of sight to
PD source.
 Common Sensor Usage for OLPD in Different Plant Items

Power Cables Cable Metal-clad AIS HV/EHV GIS Rotating MV HV

Terminations Machines Transformers Transformers
Plant

HFCT HFCT TEV UHF Coupler HV Capacitor Contact Acoustic UHF Coupler

Airborne
TEV Contact Acoustic HFCT HFCT Contact Acoustic
Acoustic
Sensors

Airborne Bushing Tap

UHF Coupler Rogowski Coil TEV
Acoustic Adapters

Contact Acoustic
HV Capacitor RTD Sensor HFCT

HFCT VHF Probes TEV

 High Frequency Current Transformer (HFCT) Sensors

• Detect PD in cables and

connected plant
• Wide bandwidth (from 100 kHz
to 20 MHz)
• Attach to power cables at
terminations and earthing links
of HV equipment
• Installation inside or outside of
cable box
• Temporary or permanent
 PD Measurement Theory Using HFCT

• Cable PD is measured in terms of

charge.

• It is important to measure the number

of PD pulses/power cycle (i.e. the
cumulative PD activity).

Rise Time Pulse Width Fall Time

The PD magnitude (in pC) is the area

under the PD pulse. pulseend
1
This can be calculated from the output
voltage of the HFCT using the HFCT’s
Q app 
Z TR V out
pulsestart
dt
Transfer Impedance, ZTR
 HFCT Sensor Attachment to Power Cables

The HFCT sensor should be attached to

intercept either the conductor PD current 1
(i+) or the earth PD current (i-)

HFCT on cable
with Earth HFCT around
HFCT on brought back cable
Earth (i-) through (i+) (i- + i+ = 0)

1 2 3 2
 Cable Terminations Not Suitable for PD Testing due to Solid Bonding

Solidly bonded - lead plumbed Solidly bonded - No insulated gland

Shorting links
 HFCT Sensor Attachment
Temporary/Permanent

• Short outage to attach

• Permanent installation with external connection point

• Periodic testing/monitoring without subsequent outages

 Local PD Detection with TEV Sensors

• Electromagnetic radiation from PD sites

• High frequency >5 MHz
• Main application: metal-clad AIS and SIS
• Sensitive to local (nearby) PD sites
• Also used at cable terminations and
transformers
• Local PD magnitude is measured in dB
 Example Use of Detection and Diagnostics
Time of Flight Measurements in MV Switchgear

1 2 3 4

Matched Length
Co-axial Signal
Cables
Measured Signals on 4x TEVs

11
2

44
 High Voltage Coupling Capacitor (HVCC) Sensors

• Conventional sensor for factory/lab tests

• Placed in parallel with cable/plant under test
• Requires galvanic connection to the plant under test
• Common application for rotating machines/switchgear

Permanent on-line unit

Off-line unit
 HVCC Sensors – Frequency Response

Higher capacitance = higher bandwidth = higher sensitivity to PD deep in machine

windings

Gain Vs. Frequency Plot for HVPD Sensors
1
0.9
0.8
Standard 80 pF
0.7
Gain (Vo/Vi)

0.6 HVPD 500 pF

0.5 HVPD 1000 pF
0.4
PD Occurrence
0.3
0.2
0.1
0
1 10 100
Frequency (MHz)
 Acoustic Detection

• Sensitive to local PD sites

• From 10 kHz to 1.2 MHz (≈40 kHz common)
• Airborne – line of sight

• Corona

Airborne Acoustic Probe and Amplifier • Surface discharge

• Contact
• Vibration of equipment housing

• Internal PD

• Surface discharge

Airborne Acoustic Linearplex Sensor

 Combined Acoustic and Electromagnetic Detection

Useful for identifying surface discharges/

corona – coincident signals in:

• Outdoor equipment
• Air-insulated switchgear
 Combined Acoustic and Electromagnetic Detection
Across Power Cycle

HFCT

Airborne
Acoustic

Delay due to difference between speed of light and speed of sound

OFF-LINE TESTING METHODS
 MV/HV Cable AC Withstand Commissioning Testing

• New cable systems require AC withstand acceptance tests at

commissioning.

• PD testing is now included as part of the field acceptance tests for HV and
EHV cables.

• Some test specifications reference acceptance criteria from conventional,

laboratory PD testing (5/10 pC).

• Whilst factory PD tests are performed at power frequency, field tests are
often not.

Off-line power supplies must be dimensioned for the charging current of the
plant under test.

I=jωCV I=Charging current, ω=Test Frequency,

C= Cable Capacitance, V=Test voltage
MV/HV Cable AC Withstand Voltage Commissioning Testing Options

1. VLF (Very Low Frequency) (0.01–0.1 Hz)

example supplier: Baur (Austria), B2HV (Germany)

2. Resonant Test Systems (RTS)

example supplier: High-Volt (Germany)

3. 24-Hour Soak Test (at U0)

No external power supply is required but extended,
24-hour PD monitoring is necessary.

4. Damped AC / Oscillating Wave (OWTS)

example suppliers: Seitz, SEBAkmt (Germany)
 Option 1: VLF (Very Low Frequency) PD + Tan Delta (TD) Testing

Pros Cons
• Inexpensive, portable equipment • PD at VLF not directly
• Effective in finding water-treed comparable to PD at AC power
cables for shorter lengths. frequency.
• Easy to perform, non-expert test • ‘Noisy’ test sets, filtering
• IEEE Standard 400.2. required.
• Concerns about trapped charges
at frequencies less than 0.1 Hz.
 Option 2: Variable Frequency Resonant Test System (RTS)

Control room (A/C) Canvas

Control and feeding unit HV reactor

Exciting HV filter and

Truck with trailer transformer protection unit

83 A/260 kV (1650 nF) AC Resonant Test System (for 132‒220 kV cables)

 Option 2: RTS (Variable Frequency Resonant Test System)

Pros Cons
• Allows direct comparison of • Large, expensive test equipment
factory PD tests to the field tests. (for HV/EHV cables) although
• Provides continuous, near power more compact RTS technology is
frequency AC withstand voltage. available for MV cables.
 PD Interpretation Criteria

Commissioning Testing (Off-Line) Maintenance Testing (On-Line)

• Measure PDIV, PDEV. • Assess individual PD sources

• Measure PD Level & Intensity. separately.
• Identify PD pattern. • Measure PD level & intensity.
• Identify location of PD source. • Identify the location of the PD
source(s) by PD mapping.
• Assess impact of individual
PD sources against cable
design (i.e. insulation
 Guiding criterion is that the cable materials, experience, etc).
system should be PD free (<5 pC)
at the specified test voltage
(1.7U0) for the HV cable systems.  Develop individual
assessment and ranking of
 Less then 10 pC for MV systems cable joints, terminations and
using conventional test systems. cable sections.
 Option 3: Damped AC – OWTS (Commissioning)

Pros Cons
• Can energise long lengths of • Only limited number of over voltage
cable with smaller power supply. cycles (2–3) applied.
• Easy to perform, non-expert test. • Difficult to do PD reading reliably
(distributed PD).
• Tan Delta / Loss factor measurements
are derived, not measured.
 Option 3: Damped AC - OWTS

• 50 shots of individual discharges.

• Not continuous AC.

Wide variation of test voltage parameters:

DC charging time: 0.25–62.5 s
DAC frequency: 38–368 Hz
DAC damping: 3.7–20.4%
The different scales of the DC charging and the DAC oscillation create a wrong impression!
 Option 4: 24-Hour Soak Test (at U0)

Pros Cons
• Inexpensive as they do not require an • May not find incipient insulation
external power supply defects at U0
• Energised at line voltage U0 only. • Not fully diagnostic or predictive
• Low risk of failure during test. as there is no overvoltage.
• Can perform extended PD testing and
monitoring over the entire 24-hour soak test.
CASE STUDY 1: RESONANT TEST SYSTEM (RTS)
TESTING
 Case Study: Commissioning from RTS to Cable

• The distance from RTS to cable under test could be tens of metres.
• Exposed HV connections pickup noise and generate corona.
• Length of exposed HV connections (and interference) minimised with test
cable.

PD signals were measured using:

• HFCT sensors placed around the earth jumper cable between the cable
termination and GIS housing.

• 8.3 nF coupling capacitor placed on the feeding end of the resonant test
set to discriminate any noise or interference from the test set.
Case Study: Typical Test Set-Up
 Case Study: Typical Test Set-Up

Resonant Test Set Blocking Test Cables

Impedance (3 in parallel)
Case Study: Wire Loom Connection from RTS to Cable Under Test
Case Study: Wire Loom Connection from RTS to Cable Under Test
 Case Study: Noise

• Noise generated by RTS frequency convertor.

• Blocking impedance helps with filtering.
• Can be gated out with dedicated channel.
• PD event recognition or high pass filters often more effective.
 Case Study: Example Results

• PD measurements made at various voltage steps.

• Ensure cables are PD-free.

Circuit: Circuit 1
Test Date: 13/05/10
Test Reactor PD PD PD
Current Frequency Power
Voltage Temp. Level R Level Y Level B Result
(A) (Hz) (kVA)
(kV) (°C) (pC) (pC) (pC)
76 3.4 221.33 250 42 no PD no PD no PD Ok
114 5.1 221.33 561 42 no PD no PD no PD Ok
132 5.8 221.28 777 42 no PD no PD no PD Ok
76 3.4 221.33 250 42 no PD no PD no PD Ok
114 5.1 221.33 561 42 no PD no PD no PD Ok
 Case Study: Conclusions

• PD measurements were made at 76 kV, 114 kV and during the 132 kV

HVAC withstand test.

• No PD signals were detected.

• Corona interference was detected in some cases and remedial action was
taken to remove this so that this did not confuse measurements.

• The background noise levels on the HFCT sensors were relatively low in all
tests, allowing good sensitivities of down to 20pC to be achieved.

• Noise interference from the resonant test set was detected with two 150 μs
pulses at the zero crossing points of the voltage waveform.

• The HVPD PDGold© ‘EventRecogniser’ software was able to classify and

discount these signals as noise.
CASE STUDY 2: OFF-LINE PD TESTING OF 11 KV
METAL-CLAD AIS SWITCHGEAR
 Case Study: Introduction

• Off-line PD testing of the newly

installed 11 kV Air-Insulated
Switchgear following a failure.
• Measurements made to determine
insulation was in good condition
prior to putting it into service.
• Portable power supply used to
energise the busbar to 1.1 U0 (13.2
kV).
 Case Study: Off-line Test Set-up

• The off-line test equipment and HVPD Longshot™.

• Voltage was applied (as per IEC 62271-200:2012) to each individual phase
in turn with other phases isolated.
 Case Study: Example Test Results

• Low levels of PD with peaks

up to 54 pC were detected at Test Date: 24/ 04 / 2013
Test Procedure A – IEC 62271-200:2012
1.1 U0 (13.2 kV phase to Peak PD In Time
1.1 Ur
ground) 15,000

50 kV CC VREF L1
50

PD Level (pC)
10,000 40
30
• Concluded the PD was not 5,000 20

originating in the breaker 10

0

sections (Breaker open and 0 1

Time (Mins)

closed). PD was isolated Maximum Peak pC at 1.1 Ur = 53.86 pC

PD Inception Voltage (PDIV)

within the Bushing and visual Peak PD In Time

inspection was recommended 80

50 kV CC VREF L1
12,000

PD Level (pC)
10,000 60
(dirt, moisture or signs or 8,000
6,000 40

tracking). 4,000
2,000
20

0
0 1 2 3 4 5 6 7 8 9 10

• The PD levels detected were Time (Mins)

considered to be low however

Peak PD Values Vs Parameter
PD Magnitude (mV)

maximum permissible partial 30

20
discharge quantity at 1.1 U0 10
PDIV: 12.58 kV

shall be ultimately agreed 0

12,000 12,100 12,200 12,300 12,400 12,500
between the manufacturer 50 kV CC VREF L1

and the end user as per IEC Peak at PDIV = 26.93 pC

62271-200:2012.
CONCLUSIONS
 Conclusions

On-line PD Testing
• Provides good data.
• Quick compared to off-line PD testing.
• Test under normal working conditions.
• Allows continuous monitoring.
• Does not require any HV power supply.

Off-line PD Testing
• Useful for factory and commissioning tests.
• Allows easy isolation of plant under test.
• Offers better sensitivity than on-line PD testing.
• Has a longer history.
• Allows testing at elevated voltages.
 End of Presentation

Thank you for your time

Q&A?
On-line PD Testing & Diagnostics for MV and HV
Equipment – Case Studies
 CONTENTS

• Reliability centred maintenance

• Application-specific examples of OLPD testing
• Insulation condition monitoring strategies
 Reliability Centred Maintenance (RCM) Bathtub Curve

End of Life
Infant Mortality Steady State Failure
‘Wear-out’
End of Life
‘Wear-out’
Failure Rate

Steady State
Failure

‘Infant Mortality’ Phase

3 Years Time 20-50 Years

• Infant Mortality Phase: the initial ‘bedding in’ period

• Steady State Failure Phase: the ‘normal’ operating lifetime up to the ‘Design
Life’ of the asset
• End of Life ‘Wear-out’: up to the manufacturer’s recommended replacement
time of the asset
CASE STUDY 1: OLPD TESTING OF 6.6 KV
DIESEL GENERATORS ON CRUISE LINER
 Case Study: League Table of OLPD Test Results

Vessel Generator PD Level 1st PD Level 2nd PD Level 3rd Test PD Level 4th
Test Test Test
Vessel 1 1 - 18,412 - 5,239
2 17,416 - - 4,259
3 15,854 - - 7,866
4 29,248 21,637 5,800 9,466
Vessel 2 1 2,162 5,891 7,340
2 16,696 2,094 1,474
3 - 520 3,790
4 12,986 4,557 6,644
Vessel 3 1 1,969
2 3,470
3 3,833
4 2,251
5 5,001
Vessel 4 1 439
2 542
3 1,610
4 2,643
5 510
 Case Study: Visual Inspection of the Stator Windings

• White oxide deposits at the end

windings of the stator: a by-product of
surface discharge activity

• The problem caused by a coolant leak.

• Recommendations: clean the end
windings, fix the coolant leak and re-
seal access hatches.
 Case Study: PD Trending Results Before
and After Maintenance to Generator

Vessel Generator July 2008 August 2008 November 2008 August 2009

Vessel 1 1 - 18,412 pC - 5,239 pC

2 17,416 pC - - 4,259 pC

3 15,854 pC - - 7,866 pC

4 29,248 pC 21,637 pC 5,800 pC 9,466 pC

Vessel 2 1 2,162 pC - 5,981 pC 7,340 pC

2 16,696 pC - 2,094 pC 1,474 pC

3 - - 520 pC 3,790 pC

4 12,986 pC - 4,557 pC 6,644 pC

• Tests performed over 12 months to trend PD data.

• Low-cost, simple maintenance performed on generators in 2008 to remove
dust, dirt and oil mist observed.

• PD levels after maintenance reduced from ‘Red = Unreliable’ to ‘Green =

OK’.
CASE STUDY 2: IN-SERVICE, ON-LINE PD TESTING
OF 34.5/5 KV 10 MVA TRANSFORMER
(U.S. VIRGIN ISLANDS)
 Case Study: Background

• Following a failure of a 34.5/5 kV

10 MVA transformer, it was
decided to carry out OLPD testing
on the ‘sister’ transformer.

• Tests were carried out on-line,

measurements made at both
transformer and 34.5 kV
Substation – takes into account
possible radiation of PD signals
from other items of plant (i.e.
connecting cables & switchgear).
Case Study: PD Propagation from Transformer on to Network
 Case Study: Sensor Connection

NGR

A A
Legend

B B
34 .5 kV Bushing
B

C C 5 kV Bushing

TEV PD Sensor

34 .5 kV 5 kV HFCT PD Sensor

Substation 44 Substation 48
(˜ 700 ft , 213 m ) (˜ 66 ft , 20 m )

HVPD Longshot
On -line PD
Test Unit
 Case Study: HFCT and TEV Sensor Connection on 34.5 kV Cable Cores

Sensor Connections at Substation End Sensor Connections at Transformer

 Case Study: Testing Methodology

• PD activity was found on the circuit and isolated to the transformer

side of the cable.

• Further tests at the transformer revealed the likely source of the PD

was in the Phase B of the 34.5 kV cable termination, transformer
bushing or end winding.

• To isolate the PD source to either the cable or transformer

components, the 34.5 kV cables on phases A, B and C were
disconnected from the transformer and tested under working voltage
supplied from the substation.

• No PD was detected on this test and thus through process of

elimination it was concluded that the PD pulses originate from the
transformer.
 Case Study: OLPD Test Results

Noise
PD Waveform

Noise Waveform

PD
 Case Study: OLPD Test Results

Separated PD and Noise against Power Cycle

Waveform Measured at Transformer HFCT B

Phase with HFCT on LV Neutral
 Case Study: Comparison of Pulse Rise Time
Distribution at Both Cable Ends

Substation End Transformer End

 Case Study: OLPD Testing of 34.5/5 kV 10 MVA Transformer

B Phase PD Data from Online Test (Purple Colour)

PD Across 60Hz Power Cycle PD Waveform Shape (expanded trace below)

Pulse Risetime Graph

 Case Study: Recommendations

• The peak levels of PD detected were 500 pC which is a reasonably

moderate level for MV cable accessories and transformer bushings.

• The PD detected was not thought to be an imminent threat but should

have definitely been regularly monitored/tested.

• The presence of PD, even at relatively moderate levels, does make the
risk of failure higher than that on a discharge free component.

• Regular on-line testing and monitoring of PD activity was recommended

to ensure the PD in Phase B does not escalate to an unacceptable
level.
CASE STUDY 3: OLPD TESTING OF 6.6 KV
MOTORS AT GAS POWER STATION (UK)
 Case Study: Background

HFCT Sensors located at Rotating HV Machine

HV Switchboard Cable End End

Range:
Up to 1 km for PVC
Up to 2 km for XLPE
Case Study: 6.6 kV Motors, OLPD Sensor Installation
 Case Study: Summary of Initial OLPD Test Results

Motor Ref PD Level [pC] PD Activity [nC/Cycle]

Manufacturer A – Motor 1 12,152 95

Manufacturer A – Motor 2 3,123 12

Manufacturer A – Motor 3 3,165 4

Manufacturer B – Motor 1 52,589 296

Manufacturer B – Motor 2 33,135 370

Manufacturer B – Motor 3 68,071 85

 Case Study: Off-line PD Test Results Circulating Water Pumps

Motor U Phase V Phase W Phase All Phases

PD (IEC) Magnitude PD (IEC) Magnitude PD (IEC) Magnitude PD (IEC) Magnitude

2010 2011 2010 2011 2010 2011 2010 2011

Motor 1 10,530 30,260 12,870 24,960 7,865 4,846 17,180 44,040

Motor 2 14,920 30,620 17,730 27,070 15,640 22,100 26,390 42,520

Motor 3 15,140 30,890 16,920 31,040 16,530 24,410 26,780 44,370

All off-line PD measurements showed an increasing trend from the 2010 tests
to the 2011 tests with an average increase in PD levels across all 3 motors of
around 100% i.e. the PD levels had doubled over the 12 months between
these tests.
 Case Study: Visual Inspection after Off-line PD Tests

Image of Neutral Terminal Box. The Glass Cover Image of winding connections. Neutral Cables and
(arrowed) was removed to inspect the windings internal connections not spaced apart (arrowed)
 Case Study: Visual Inspection after Off-line PD Tests

Image of connections not adequately spaced Image of coil – coil connections touching a neutral
(arrowed) = design issue cable (arrowed) = design issue
 Case Study: Conclusions

Two
Motor separate
third party, Motors
Warranty Manufacturer
rewound
Claim with dismissed the off-line PD and repaired
Motor validity of tests backed under
Manufacturer HVPD’s on- up HVPD’s warranty!
line PD tests findings.

Customer’s feedback:

‘Thanks for your support throughout the whole process, your equipment and advice has played
a big role in a fairly complicated warranty claim that enabled us to successfully identify and
rectify a defect, which with PD was always going to be a difficult one to secure.’
CASE STUDY 4: OLPD TESTING AND MAPPING
OF 34.5 KV XLPE CABLES FOR A PROCESS
INDUSTRY CLIENT (SAUDI ARABIA)
 Case Study: Background

• OLPD tests on a number of cable

circuits where recent cable joint
failures had occurred.
• Four cable circuits that were
known to have been subjected to
heavy circulating currents in the
earth screen & armour (in excess
of 80 Amps) which had led to
catastrophic in-service
failures.
• The previously failed joints
showed evidence of tracking.
 Case Study: Introduction

• OLPD tests on a number of cable circuits where recent cable

joint failures had occurred.
• Four cable circuits that were known to have been subjected to
heavy circulating currents in the earth screen & armour (in
excess of 80 Amps) which had led to catastrophic in-service
failures.
• The previously failed joints showed evidence of tracking.
Severely Damaged 34.5 kV XLPE Failed Cable Joint
 Case Study: Circuit Details

The 1.9 km long, 34 kV XLPE cable circuit had 5 jointed sections with only one
cross-bond point.
 Case Study: OLPD Screening of the Cables

The HVPD PDSurveyor™ was used to detect any Local PD in the cable
terminations / joints and switchgear panels.
 Case Study: Phase 2 - Diagnostic OLPD Testing

The HVPD Longshot™ unit was used to test and locate the discharging joints.
Case Study: Sensor Attachment at Seawater Pump Transformers
Case Study: Sensor Attachment at GIS Switchgear Terminations
 Case Study: On-line Cable Mapping using PDMap©

• The measured Cable Return Time for the pulse injected at the
switchgear to travel to the far end of the seawater feeders and back was
20.4 μsec.

• Given that the seawater feeder cable is 1956 metres in length, this gives a
return speed for the 34.5 kV XLPE cable of 95.9 m/μsec.
 Case Study: On-line Cable Mapping using PDMap©

• PDMap© software was used to create a map of the 1956 m cable.

• The source of PD activity was located within the Red Phase cable of
seawater pump Tx302 at 1770 metres out from the 34.5 kV main substation
(or 185 metres out from the Seawater Pump Transformers).

• This corresponds with the location of Joint No.3 on this cable as shown
below.
 Case Study: On-line Cable Mapping using PDMap©

The cable joint was replaced, re-tested

and found to be discharge-free.
CASE STUDY 5: ON-LINE PD TESTING OF 34.5
KV INDUSTRIAL CABLE NETWORK (USA)
 Case Study: Introduction

• OLPD tests were conducted within the 34.5 kV XLPE cable terminations at
the padmount transformers in follow up to recent in-service failures of cable
terminations.

• 21 cable circuit terminations were tested with 42 points of attachment (POA)

with the PDSurveyor™, HVPD Longshot™, and temporarily installed HFCT
sensors.
 Case Study: OLPD Measurement Results across Network Diagram

• Considerable
cross-coupling of
the PD signals at
each padmount.

• PD signals
propagate far into
the cable network
from source.

• Four sources of
this type of activity
located.
 Case Study: Targeted Investigation Locates Probable Cause

Surface
Tracking

White Oxide Bare XLPE

Deposits Insulation!

• Poor workmanship was the main contributing factor to the surface PD

activity detected.

• Signs of PD activity on the cable terminations

• The rubber stress cone termination had been completely eroded.

 Case Study: Remedial Action

• Re-termination of the cables into the six ‘RED PD Level’ padmounts.

• Further OLPD tests showed no PD activity and confirmed the rework had been
effective.

• The customer was advised that regular, periodic OLPD ‘screening’ tests should
be carried out.
 CASE STUDY 6: OLPD TESTING AND LOCATION ON
OFFSHORE WIND TURBINE TO OIL & GAS PLATFORM
33 KV CABLE FEEDER
 Case Study: Background

• Two deep-water wind turbines supply power exclusively to an oil production platform
2km away.

• Two on-line PD tests were performed to assess the condition of the turbine feeder
cables.
Case Study: Test 1 – OLPD Test of 33 kV Switchgear on Platform
 Case Study: OLPD Test of 33 kV Switchgear on Platform

• PD on Turbine A Feeder
was detected and was
considered to be remote
source based on pulse
properties.

• Analysis predicted the

location to be near the
far end of the Turbine A
feeder cable. Mapping
recommended.
Case Study: Test 2 – PD Location at Wind Turbine 33 kV Feeder RMU
 Case Study: PD Location at Wind Turbine 33 kV Feeder RMU

• High PD magnitudes in excess of 6,000 pC, with average

levels of 1,871 pC detected on Yellow phase.

• Test identified the PD source was the Yφ local turbine 33 kV

transformer/cable joint.

• Measurement of PD pulses on wind turbine cable feeder show

clear direct & indirect reflected pulses.

• Joint location was confirmed at 52 metres from the RMU.

CONCLUSIONS
 Conclusions

• Insulation Condition Monitoring (CM) using OLPD requires a

continuous evaluation of the HV insulation’s dielectric
integrity throughout the service life of the asset.
• Continuous OLPD monitoring should be carried out
throughout the first 3 years of service - the ‘bedding in’
period.
• Test and monitor after any repairs to faults – to ensure the
repair has worked!
• Continuous OLPD monitoring is recommended
throughout the service design life (of 20–25 years+).
 End of Presentation

Thank you for your time

Q&A?
Deployment of OLPD Testing in
Asset Management Systems
 CONTENTS

• HVPD 4-Phase approach for OLPD monitoring of MV

networks
• PD trending
• Plant condition analysis
• OLPD league tables
• Examples of deployment of PD testing in
asset management systems
 HVPD 4-Phase Approach for OLPD Monitoring of MV Networks

• A complete solution for the OLPD screening, diagnostic testing and

extended monitoring.

• Identify, locate and monitor PD activity in the ‘Worst 5%’ of the

customer’s network.

• A range of portable and permanent OLPD test and monitoring technology

can be applied to achieve this.
The HVPD Integrated OLPD Test and Monitoring Solution For
Medium Voltage (MV) Networks (Voltage Range: 3.3–36 kV)
 Phase 1: OLPD Surveying of MV Cables and Plant
The HVPD PDSAir™ Surveying Tool

• Used for initial, quick screening of large numbers of MV plant items.

• The MV cable/plant can be tested in-service under normal operating

conditions, no outage required!

• Easy to read, 7-level, colour-coded PD level indication panel.

HVPD PDSAir™ Handheld Surveying Tool
 PDSAir™
On-site OLPD Screening
 Why Use the PDS Air™

• Combines HFCT, TEV and AA sensors to enable OLPD testing of both cables
and switchgear.

• A look-see OLPD scan, indicating the plant which requires further diagnostic
testing.

• Low cost, lightweight & portable, easy to use.

• Test the insulation condition of the plant in seconds.

 HVPD PDS Air™
PD Level Guide

• LED 1 – Green (Plant OK)

• LED 2 & 3 – Yellow (Moderate PD – Monitor)

• LED 4 & 5 – Orange (Moderate To High PD –

Investigate Source Of PD)

• LED 6 & 7 – RED (High PD – Test & Restrict Access)

NB: It should be noted that the PD levels & actions recommended are
guideline levels only and are based on HVPD’s experience in testing
MV Plant
 HVPD PDS Air™
Applications Overview

TEV Sensor Airborne Acoustic Sensor HFCT Sensor

 Example Use of Detection and Diagnostics
MV Switchgear with TEV

Very High Level Local PD Activity >48 dB (Panel 006)

Further Investigation Recommended
Phase 2: HVPD Longshot™ Diagnostic OLPD Test Unit

• 4-Channel, synchronous OLPD test unit

• Captures PD activity using very high speed

data acquisition capability (100–500 MS/s)

• Diagnostic PDGold© v7 software with unique

‘Event Recogniser©’ software modules
differentiates between PD activity and any
electrical noise and RF interference.

• An automatic, detailed analysis of pulse

frequency, waveshape and other signal
waveform characteristics.
 HVPD Longshot™
On-line PD Testing of an 11kV Motor
 PD Data Analysis

• Impulsive events extracted

and categorised by
software as cable PD, 20 ms

local PD and noise. A vailable Wavef orm Display

0.015

• Even amongst noise One power cycle of

0.01

0.005

interference, PD data can

Chan 1
raw data from HFCT 0
sensor -0.005

still be extracted. -0.01

-0.015

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Time (mSec)

Chan 1 Curs 1 Curs 2

15 us
Segment Wavef orm Segment Wavef orm
Segment Wavef orm Ch 1 Ch 1
Ch 1 15
15 4
10
10 2
5
Volts (mV)

Volts (mV)
5
Volts (mV)

0 0
0
-5
-5 -2
-10
-10
-4
-15
-15
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
0 2 4 6 8 10 12 14
Time uSec Time uSec
Time uSec

Cable PD Pulse: 16 mV, 1160 pC Local PD Pulse: 15 mV, 24 dB Noise Pulse: 6 mV

Phase 3 – Temporary Monitoring HVPD Multi™ Portable
16-Channel OLPD Monitor

• 24/7 diagnostic portable OLPD monitoring

technology for MV and HV Plant

• Measures, analyses and logs Cable PD,

Local PD and Noise

• 'Knowledge-Rule' based PD criticality

measurement (0–100)

• Remote access connection via

LAN/Modem

• 16-Channel (HFCT/TEV/HVCC) with

optional add-on 32x channel AA sensor
modules
HVPD Multi™ Portable 16-Channel OLPD Monitor
Background
 Phase 3 – Plant Condition Analysis

• Criticality league
table

• Highlights most
critical circuits

• Updated every 24
hours

• Link to data graphs

• Criticality algorithm,
specific to plant

Small PD-activity
Moderate PD-activity
Intensive PD-activity
Critical PD-activity
 Phase 3 – Plant Condition Analysis

• Criticality league
table

• Highlights most
critical circuits

• Updated every 24
hours

• Link to data graphs

• Criticality algorithm,
specific to plant

Small PD-activity
Moderate PD-activity
Intensive PD-activity
Critical PD-activity
 Phase 3 – Temporary Monitoring HVPD Mini™
Portable 4-Channel OLPD Monitor

• Low-cost 24/7 PD monitoring technology for

MV cables, switchgear & other plant

• Incorporates up to 4x portable PD sensors

• Measures and logs both cable PD and local

PD (PD magnitude & ‘count’ – no. of pulses)

• Stores up to 12 months of data on local flash

memory

• Uploads PD data to server every 24 hours via

GSM/GPRS Modem

• Compact, lightweight and easy to set up unit

for portable installations
HVPD Mini™ Monitor
Example Installation
 HVPD Mini™ Monitor
Distributed TEV Sensors
 Continuous OLPD Monitoring

• Provides an early warning against incipient faults to be indicated through

changes in PD activity over time.

• PD activity can vary in relation to load, local temperature and humidity.

• Distributed temperature and humidity modules monitor these variations in the

substation.
Cumlative PD Activity (CT)
CT PD Activity (Pc/Cycle)

10,000

5,000

0
00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00
30/12/2010

31/12/2010

01/01/2011

02/01/2011

03/01/2011

04/01/2011

05/01/2011

06/01/2011
Time
TEV PD Activity (mV/cycle)

Cumlative PD Activity(TEV)

100
80
60
40
20
0
00:00

00:30

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00
22/09/2009

23/09/2009

25/09/2009

27/09/2009

29/09/2009

01/10/2009

03/10/2009

05/10/2009

07/10/2009

09/10/2009

11/10/2009
Ti
 Combined OLPD Monitoring Solution
Primary Substation Monitor plus Distributed HVPD Mini™ Monitors

Distributed
HVPD MiniTM

Primary 11-33kV 16x Feeder Cables

Substation HVPD Multi™ Monitor

Distributed
HVPD MiniTM

RMU
Range of Detection for Primary PD Monitor:
Cable
Extended range from HVPD Mini™:
Joint either up to 2.5 km or up to 3x Ring Main Units either 2 km or 2 RMU’s
Copyright© 2009 HVPD Ltd
 Combined OLPD Monitoring Solution
Primary Substation Monitor plus Distributed HVPD Mini™ Monitors
Precedence Detection using the HVPD Mini™ OLPD Monitor
 Phase 4 – Permanent Monitoring - HVPD Multi™ Permanent
16-Channel OLPD Monitor

• 24/7 monitoring of PD in switchgear, cables and rotating HV machines

• Non-intrusive, inductive sensors
• Remotely accessible with GPRS/3G modem
• Onboard automatic analysis software provides differentiation of all PD
pulses from noise
• Can monitor up to 4x rotating HV machines
HVPD Multi™ Permanent Monitor
 HVPD-Multi™ Monitor
Automated alarms vs flags

• The HVPD-Multi™ Monitor does not provide automated alarms but

provide ‘flags’ for further engineer investigation
• These ‘flags’ can signal an increase in either PD level or PD activity
but can also be caused as a result of network switching
• Each ‘flag’ is investigated by HVPD engineers before a diagnostic
decision is made regarding preventative maintenance interventions.

Some
Major
Increase within concern -
concern -
acceptable limits monitor
intervention
closely
 HVPD Multi™ Monitor
30-day monitoring training period

INSTALLATION DATA ANALYSIS ADJUSTMENT

DAY 1 DAILY DATA ANALYSIS DAY 30

Phase 3 – Remote Access
 EXAMPLE OF HV NETWORK OLPD
MONITORING SYSTEM USER INTERFACE
BASED ON THE CLIENT’S HV NETWORK
SINGLE-LINE DIAGRAM (SLD)
 Complete Network Monitoring Database Interface

Presents data hierarchically using a 3-

level interface:
• 1st level: all sites/vessels
• 2nd level: one site/vessel (entire SLD)
• 3rd level: one switchroom
Complete Network Monitoring Database Interface
Complete Network Monitoring Database Interface
Complete Network Monitoring Database Interface
EXAMPLES OF DEPLOYMENT OF PD TESTING IN
ASSET MANAGEMENT SYSTEMS
CASE STUDY 2: OLPD TEST AND MONITORING
PROJECT ON 11 KV AND 33 KV
‘WORST PERFORMING CIRCUITS’
 Case Study: Introduction

• ‘Worst Performing Circuits List’ focuses on replacing cable sections with the
highest number of faults.

• Due to budget, it is only possible to replace a small percentage of circuits a

year.

• Limited monitoring budgets need to be used on circuits with the highest risk of
failure.
 Case Study: Average Number of Faults per 100 km per Annum

• Network A 11 kV: 6.91 faults per 100 km per annum

• Network B 11 kV: 13.66 faults per 100 km per annum
• Network A & Network B 33 kV: 11.5 faults per 100 km per annum

Summary Data of 11 kV Worst Performing Circuits in Networks A & B (2004-2008)

Average Circuit Average faults per annum
No. of Circuits Total Length Faults per 100 km p.a.
Length (Av. 5 years 2004 to 2008)
Network B - 86 530 km 6.160 km 72.4 13.66
Network A -122 970 km 7.950 km 67.0 6.91
Total - 208 1500 km 7.210 km 139.4 9.30

Summary Data of 33 kV Worst Performing Circuits in Networks A & B (2002-2008)

Average Circuit
Total Length Average faults per annum
No. of Circuits Length Faults per 100 km p.a.
(est.) (Av. 7 years 2002 to 2008)
(est.)
Network B– 51 357 km 7.000 km 42.1 11.80
7.000 km
Network A – 62 434 km 48.6 11.19
Total – 113 791 km 7.000 km 90.7 11.47
 Case Study: Suitability of Cable Terminations for OLPD Testing

• The suitability of cable terminations for the attachment of the HFCT sensors
is the main restriction on the widespread application of the OLPD
technology.

• Only around 40% of the terminations on the customer’s networks can be

considered as ‘suitable’.

• For future testing, it has been suggested that the customer modify solidly
bonded terminations.
 Case Study: Test Locations and Methodology

• OLPD tests were carried out at

substations with worst historical
failure rates.

• HVPD Mini™ Portable was

installed at a number of
substations.

• The spot testing was carried out

using HVPD’s OLPD test
technology.
 Case Study: Test Results

10% of all of 39x 33kV cable circuits tested are the Red Condition Category.

No. of Circuits Condition %

26 Discharge within acceptable limits 66
7 Some concern, monitoring recommended 18
2 Some concern, regular monitoring recommended 5
4 Major concern, locate PD and then repair or replace 10

4% of all 25x 11kV circuits tested in this project are in the Red Condition
Category.
No. of Circuits Condition %
18 Discharge within acceptable limits 72
4 Some concern, monitoring recommended 16
2 Some concern, regular monitoring recommended 8
1 Major concern, locate PD and then repair or replace 4
 Case Study: Worst Performing Circuits
OLPD
Criticality Peak PD Cumulative PD Maintenance
Circuit Comments Criticality
Number Value Level Action
(%)
1. Circuit 18 Large PD on this circuit 3600pC 195 nC/Cycle 83.2
Test Results: 11kV and 33kV Combined ‘OLPD League Table’ – ‘Top 20’
4000pC 62nC/Cycle
2. Circuit 62 Large cable PD 79.7
Cable Box PD 1.2V/Cycle
Major concern,
No cable PD locate PD and
3. Circuit 28 36dB 1.1 V/Cycle 78.3
Local PD is 36dB then repair or
No cable PD replace.
4. Circuit 26 35dB 1.4 V/Cycle 76.4
Local PD 35dB
25dB *Outdoor
5. Circuit 14 Some outdoor PD 24.6 V/Cycle
Local PD Survey*
No Cable PD
6. Circuit 32 34dB 1.0 V/Cycle 73.3
Local 34dB
Cable box PD 1600pC & 4.2 nC/Cycle Some concern,
7. Circuit 50 70.1
TEV levels at 35dB 35dB 0.7 V/Cycle repeat test and
1200pC regular
<10 nC/Cycle
8. Circuit 63 Cable PD & Cable Box, 30dB 30dB Cable 62.2 monitoring
1.5V/Cycle
Box recommended.
No Cable PD
9. Circuit 36 27dB 1.6 V/Cycle 57.7
Local 27dB
10. Circuit 21 Large PD on this circuit 3800pC 21 nC/Cycle 56.3
11. Circuit 61 Low-Medium Level Cable PD 1600pC 52nC/cycle 55.8
12. Circuit 40 Medium-High Cable PD 4000pC <10 nC/Cycle 55.5
No Cable PD
13. Circuit 37 23dB 1.1 V/Cycle 46.6
Local 23dB
No Cable PD
14. Circuit 31 25dB 0.6 V/Cycle 45.7
Local 25dB Some concern, re-
15. Circuit 46 Medium Cable PD 2500pC <10 nC/Cycle 38.2 test within 6
No cable PD months.
16. Circuit 27 20dB 0.8 V/Cycle 37.4
Local PD 20dB
17. Circuit 43 Medium-High Cable PD 3500pC <10 nC/Cycle 37.1
18. Circuit 12 Incipient PDof 2000pC measured at Circuit 12 2000pC <10 nC/Cycle 29.1
19. Circuit 11 PD on circuit, no location 1000pC 40.4 nC/Cycle 28.0
20 Circuit 23 Red phase is the source 1400pC <10 nC/Cycle 25 2
 Case Study: Conclusions and Recommendations

• Continuous OLPD monitoring of the medium and high PD

sites (the highest risk of failure).
• Retesting of all of the ‘TOP 20’ circuits within the next 3-6
months.
• Deploy monitoring units at the worst performing substations
over an initial 3-month monitoring period.
• For the 44x circuits which have lower, a repeat ‘spot’ test
within 12 months would be prudent.
• A larger survey of a minimum 100x 11 kV and 100x 33 kV
feeders should be carried out to provide a more statistically
valid data set.
 CASE STUDY 3: OLPD TESTING AND CABLE
MAPPING OF 33 KV XLPE CABLES IN THE DUBAI
METRO NETWORK
 Case Study: Introduction

• OLPD testing was carried out in

response to a number recent
faults* of 33 kV cable joints
within the customer’s network.

• The faults led to disruption of the

power supply to the Metropolitan
rail system.

• The purpose of the testing was

to measure and locate any PD
activity within the cables with
particular focus on the cable
joints.

* It should be noted that this was a newly installed cable system that had been in-
service for just over 12 months before the faults started to occur.
 Case Study : OLPD Testing Equipment and Methodology

• On-line Cable PD Mapping using the HVPD Longshot™ test unit and Portable
transponder.

• Tests started with calibration testing with pulse injection HFCTs.

 Case Study: Test Results

• Cable PD signals have

been detected on Blue
Phase with cross-talk
(lower magnitude) on
Red and Yellow phases.

• The source of PD was

located to Joint Number
2 (Jt2) using the on-line
PD mapping technique.

• The faulty joint on this

cable was replaced and
re-tested using the
HVPD Longshot™ test
unit to verify the repair
was good
 Case Study: Test Results

PD located to joints on Red Phase

 Case Study: Top 20 ‘Worst Performing Circuits’

• Out of the 50+ circuits tested, Major PD was detected within cable
accessories on the three of the circuits (6%) as shown in RED in the Table
below.
• The levels of discharges detected put these 33 kV cables into RED category,
“Major concern, locate PD and then repair or replace”.
Peak Cable Local PD Cumulative
Criticality OLPD Maintenance
Circuit Comments PD Level Level Cable PD Level
Number Criticality (%) Action
(pC) (dB) (nC/cycle)
1. DUB to MPS1 C2 B Phase 25888 <10 247 97.4
2. ABS to AH C2 B / Y Phase 9729 <10 120 90.3 Major concern,
3. BUR to HCC C2 B / Y Phase 3781 <10 12.3 78.7 locate PD and
4. BUR to HCC C1 B / Y Phase 3245 <10 7.9 78.1 then repair or
5. ABS to AH C1 B / Y Phase 2920 <10 14.4 77.4 replace.
6. NHD to QYD C2 R Phase 2849 <10 15.0 76.2
7. ALQ to AHS C2 B Phase 1733 <10 4.6 70.6 Some concern,
8. MPS3 to BNS C2 R / B Phase 1337 <10 6.4 65.5 repeat test and
9. NHD to QYD C1 R Phase 887 <10 8.8 47.8 regular
10. HCC to CRK C1 Y / B Phase 759 <10 2.5 39.2 monitoring
11. AHS to SLD Y / R Phase 705 <10 3.1 38.5 recommended.
12. STD to ABH Y Phase 238 <10 1.0 24.1
13. ALR to BNS C1 B Phase 184 <10 0.9 18.6
14. ALR to BRJ No PD detected 0 <10 0 0
15. ALG to PMD No PD detected 0 <10 0 0
Re-test in 12
16. ALG to KBW No PD detected 0 <10 0 0
months.
17. AQD to AQ2 No PD detected 0 <10 0 0
18. JDD to CRK No PD detected 0 <10 0 0
19. ODM to JDF C1 No PD detected 0 <10 0 0
20. ODM to JDF C2 No PD detected 0 <10 0 0
CONCLUSIONS
 Conclusions

• A combination of both on-line and off-line partial discharge testing and OLPD
monitoring systems for in-service plant helps produce 'risk-of-failure' indices
that support condition-based asset management decisions

• By replacing or repairing cables or plant that has high levels of OLPD activity
(and therefore a higher risk of failure) the MV and HV plant owner can target
their maintenance budgets to those assets in most need whilst
simultaneously reducing the risk of HV insulation faults on their network

• With the advent of recent developments in wideband OLPD monitoring, such

as with the HVPD Complete HV Network Monitoring Solution, the entire
installed HV network, including switchgear, cables and remotely connected
HV plant can be assessed under normal working conditions, without the need
for an outage

• Real-time condition monitoring (CM) such as this, combined with a proactive,

preventative maintenance intervention strategy can help reduce the risk of
unplanned outages caused by HV insulation failure
 End of Presentation

Thank you for your time

Q&A?