Introduction to PD Testing

[email protected]
Mob: 0417 17 8026

What is a partial discharge?

Partial discharges (PD) are localized electrical discharges
within an dielectric insulation system, restricted to only a
part of the dielectric material, thus only partially bridging the
electrodes. So the breakdowns stay local
The insulation may consist of solid, liquid or gaseous
materials, or any combination.
The term partial discharge includes a wide group of
electrical discharge phenomena.
Materials are differently effected by PD but discharge activity
has in general detrimental effects on the insulation material
Partial Discharge Measurement is sensitive to overall and
localised defects

HV

Internal discharges:

In voids or cavities within solid or liquid

materials (incl. at boundaries of different

insulation materials)

Continuous impact of discharges in solid

dielectrics forms discharge channels (treeing)
External discharges:
Surface discharges: At boundaries of different
insulation materials tracking

HV

Corona discharges: Discharges in gaseous

dielectrics when strong, inhomogeneous fields
are present
Discharges due to electrically floating potentials
(related to gaseous dielectrics)

PD wide group of discharge phenomena

HV

Reason for PD Ignition

Aging processes, due to
Electrical overstress
Mechanical overstress
Thermal overstress

Incorrect assembly, manufacturing

defects

When should HV plant be tested for PD

PD measurements should be performed at the factory to
ensure correct design and manufacturing quality.
The On-site Commissioning tests and periodic PD tests to
ensure that no transport damage has occurred and that the
was been assembled correctly.
The On line PD monitoring provides a continuous
surveillance of discharge activity for risk assessment of
the asset (trend analysis, warnings, nursing of suspect
equipment etc.

Almost all HV insulation systems can be adversely affected by PD:

- Rotating Machines

- Transformers

- Cables

-Switchgear

Components

Typical Locations of PD Ignition

Cavities, interfaces
of different dielectric
properties and at
sharp electrode
edges and
protrusions

In this Perspex block a

carbonised breakdown
channel is developing from
one electrode.
The material ahead is still
insulating and high
impedance. This prevents
any significant current flow
and the tree extends only
slowly, branch by branch.
But eventually it will break
through.

Close up of 11kv CW Pump

stator winding

Before and after lab ageing

Close-up of damage in previous slide. Here the PD have eroded away

the resin leaving dry glass cloth and sheets of mica-paper

Surface Effects

PD damage due to generation of Ozone

Ozone is a chemically reactive gas that combines with other gases in air
to create Nitric Acid

Why do PD occur?
Practical dielectric insulation often contains voids or cavities
The voids/cavities are usually filled with a medium (i.e. gas),
often with lower breakdown strength than the main
dielectric (the breakdown strength of air is ~100 times less
than many solids)
The electric field in a composite dielectric is distributed
according to capacitances, and the dielectric permittivity of
the medium (gas) is usually lower than the solid. Thus, an
air-filled void will have a field stress enhancement related
to the dielectric/solid permittivity r

Example void in solid

Cc: Capacitance of void
Cb: Capacitance of solid in series
w/void
Ca: Capacitance of the rest of the
solid
Va: Applied voltage of solid
Vc: Voltage across void
V+/V-: Inception volt. for PD in void
Groups of discharges originate from
a single void and give rise to
current pulses (pos. and neg.)

Why do PD occur?
If the voltage across the void is high enough (>the
inception voltage), the field stress in the void will exceed its
dielectric strength and the voltage across it collapses/breaks
down, i.e. a partial breakdown/discharge has occurred
The solid dielectric in series with the void will withstand the
externally applied voltage and choke off the PD
The PD will re-ignite if the voltage builds up to the inception
voltage again or on voltage reversal
The PD breaks chemical bindings in the dielectric, the solid
erodes and gets thinner and eventually fails

Detection of partial discharges

Detection is based on the energy exchanges that take place
during the discharge
Dielectric Heat

Light

losses

Electromagnetic

Chemical changes,

radiation

gases
Impulse
current pulses

Sound/
noise

Macroscopic-Physical
Effects
HV

Detection Methods

Optical Effects
(Light)

Optical

Pressure Wave
(Sound)

Mechanical

Opto-acoustic

Discharge Effects
Dielectric Losses
High Frequency
Waves

Electrical

IEC 60270

Chemical Effects

Chemical

Heat

Acoustic

HF/VHF/UHF

1. Electrical transients: The flow of charge at the defect will cause

an equivalent charge transfer within the apparatus also reflected
in the external circuit allows effective and calibratable
detection (IEC 60270)
2. Electromagnetic radiation: PD generates high frequency
electromagn. radiation up to 1 GHz)

VHF/UHF sensors (inductive/capacitive sensors)

Spectrum analyzers

Skin effect currents leaving dielectric through gaskets

TEV

3. Chemical changes: Measurements of by-products/gases

DGA in dielectric liquids
SF6-gas
Ozone etc.

PD Detection Methods

4. Sound/noise/vibration: Tens of kHz regime

Contact probes/transducers to detect/locate

internal discharges (GIS, cable accessories,
transformers, switchgear)

Airborne/remote detection of corona and surface

discharges: Windings, cable terminations etc.

5. Light: Ultra violet cameras to locate PD (e.g. DayCor

Corona Camera)
6. Heat: Infrared cameras/scanning to detect more intense
PD

PD Detection Methods

PD Test Methods
Off-Line, On-site, Out of service

VLF: Very Low Frequency 0.1Hz

OWTS: Oscillating Wave Test System
Resonant Test (10 - 400Hz)
Power Frequency 50/60Hz

On-Line, In-service
Normal working voltage
Spot Test or Continuous Monitoring
Wideband (100kHz to 400MHz)

PD Off-Line Detection Circuit

II. TE-Messprinzipien Elektrische TE-Messung

Measurement Impedance in
Series mit Coupling Capacitor

Measurement Impedance in
Series mit Test Object

Bridge Circuit

Coupling via Bushing Tap

Measurements options according IEC 60270

II. TE-Messprinzipien Elektrische TE-Messung

Calibration Circuit

Calibration of test Circuit

II. TE-Messprinzipien Elektrische TE-Messung

Narrow Band PD-Measurement

- Bandwidth 9 kHz and 30 kHz
- Centre Frequency between 50 kHz und 10 MHz

Wideband PD-Measurement
- Bandwidth typically between 40 kHz und 400 kHz (IEC60270=
<1MHz)
- On-site sometimes higher measurement frequencies are used
(typically up to 20 MHz)

PD in the Frequency Domain

typical noise spectrum
on-site

frequency spectrum of
PD pulses
frequency
characteristics of PD
measuring systems

 Filtering
- frequency selective
filtering of sinusoidal
noise
Windowing
- Software or hardware
windowing of pulse
shaped periodical noise
Gating / Masking
subtraction of pulse
shaped stochastical noise
Synchronous
measurements
Bridge measurement
Separation of pulses
based on individual
pulse characteristics

Interference sources & suppression techniques

Record discharge parameters associated with every

individual PD pulse. Data is analysed instantly to
produce results.

Digital PD detectors

Internal PD
always dangerous

Void in
Insulation

Tree Growth
in insulation

Sharp, Irregular
surface on
conductor

Floating metalwork
near conductors

External PD
dangerous depending on plant

Corona from sharp

objects at high voltage

Surface
Discharges

7x Types of Partial Discharges

Discharges from field

induced situations

Corona Discharges (point at HV)

Surface Discharges

 Phase Resolved Partial Discharge (PRPD) Pattern

Analysis
Delamination - thermal ageing often caused by thermal
stress

Pattern and distribution of pulses determines the

location of the voids e.g. inside the material or
delamination from the conductor

More Surface Discharges increase in test voltage, note

low magnitude, analysis by magnitude only can
lead to false conclusions

 Phase Resolved Partial Discharge

(PRPD) Pattern Analysis
Slot Discharges

Wedge

Stator Core

Conductor

Void
Slot
Ground
Insulation

Strands

Computer Assisted PD Recognition

Data base
structure is
open for
adaptation
by the user

Typical
waveforms
for PD
Types
On-Line PD
Testing Now
a Field
Reality
Segment Waveforms
0.03
0.02
0.01

Main segment

Example of Typical Monopolar Cable

PD Pulse on PILC 33kV Cable (-ve pulse)

0
-0.01
-0.02

(Frequency Band: 200kHz to 4MHz)

-0.03
0

7
8
9
Time (uSec)

10

11

12

13

14

Main Waveform
0.006

(Frequency Band: 4MHz to over 100MHz)

0.004
0.002
Chan 1

Example of Typical High Frequency,

Oscillatory Switchgear PD Pulse

0
-0.002
-0.004
-0.006
0.05

0.1

0.15

Examples of Typical Noise Pulses

0.2
0.25
Time (uSec)

0.3

0.35

0.4

0.45

(Switching noise at 40KHz, RF noise at 600KHz)

Segment Waveform

Segment Waveforms

Volts (mV)

0.005
0.004
0.003
0.002
0.001
0
-0.001
-0.002
-0.003
-0.004

0
35

ent

2,000
1,500
1,000
500
0
-500
-1,000
-1,500
-2,000

40

45

50

55
60
Time (uSec)

65

70

75

80

6 7 8 9 10 11 12 13 14
Time uSec

Ch 1
Ch 2
Ch 3

Cable
PDGold
data
On-LinePD
PDCursors
Testing from
Now a
Field Reality

Segment view

Volts (mV)

25
20
15
10
5
0
-5
-10
-15
-20
-25

The PD magnitude
in picoCoulombs
(pCs) is the area
under the PD pulse.
2

Risetime

4
Time uSec

Pulse Width

Falltime

A v a ila b le W a v e f o r m D is p la y
0 .0 3 5
0 .0 3

Available Wavef orm Display

0 .0 2 5
0 .0 2
0 .0 1 5
0 .0 1

0.012
0.008

Chan 1

0 .0 0 5
0
- 0 .0 0 5
- 0 .0 1

Chan 1

0.004

- 0 .0 1 5
- 0 .0 2
- 0 .0 2 5
- 0 .0 3

0
-0.004
-0.008

Noise

-0.012

- 0 .0 3 5
-0.016

10
T im e ( m S e c )

12

14

16

18
Time (mSec)

A vailable Wavef orm Display

0.012

A vailable Wavef orm Display

0.016

0.008

0.012
0.008
0.004

0
Chan 1

-0.004

-0.004
-0.008
-0.012

-0.008

-0.016

Cable PD

-0.012

-0.02

Noise

-0.024
-0.028

Time (mSec)
A vailable Wavef orm Display

A vailable Wavef orm Display

Time (mSec)

0.028

0.032
0.028
0.024
0.02

0.024
0.02
0.016

0.016
0.012
0.008
0.004
0

0.012
0.008
Chan 1

Chan 1

Chan 1

0.004

-0.004
-0.008
-0.012
-0.016
-0.02

0.004
0

-0.004
-0.008

Cable PD

-0.024
-0.028

-0.012

Noise

-0.016
-0.02

Time (mSec)

Time (mSec)

Varying waveforms from a single 50Hz power cycle period

Example shows cable and switchgear PD events plus exciter noise on two channels.

PD Severity
New equipment, 5pC typical in IEC standards
Apparent charge measured Off-line
On-Line results less clear
No direct electrical connection of sensors
Wideband detection
Calibration difficult to achieve without an outage

Na tiona l Ge rma n (DIN VDE)

Limit va lue s for
Volta ge
PD le ve l

HV Appa ra tus

Te st Sta nda rd
Bushings

Ca pa citors

Ca ble s

Inte rna tiona l (IEC)

Publica tion

Re ma rks

300 pC (oil)
10 pC (GH)
10 pC (HP)

IEC 137 (1984)

(oil) Oil impregnated

(GH) Cast resin impregnated
(HP) Hard laminated paper

IEC 358 (1990)

Identical

IEC 885-2 (1987)

IEC 885-3 (1988)

Test procedure
Test procedure

DIN VDE 0674

Part 99/12.92

1.05 U / 3
1.5 U / 3

DIN VDE 0360

(1.1 Um )

(100 pC)

Part 3 A 1/08.83

1.1 Um / 3

10 pC

DIN VDE 0472

Part 513/07.82
DIN VDE 0271
/06.86

2 U0

20 pC (PVC)

IEC 840 (1988)

1.5 U0: 10 pC (VPE)

DIN VDE 0273

2 U0

5 pC (VPE)

IEC 502 (1994)

1.5 U0: 20 pC (VPE)

2 U0

5 pC (VPE)

1.5 U0: 40 pC (PVC)

/12.87
DIN VDE 0263
/02.91

Ca ble Joints

CT + PT

DIN VDE 0278

Part 1/02.91

Test procedure

DIN VDE 0278

Part 2/02.91

2 U0

20 pC (VPE)
40 pC (PVC)

Joints, Terminations

DIN VDE 0278

Part 6/02.91

2 U0

20 pC

Pluggable and screwable

encapsulated cable terminations

DIN VDE 0414

(1.1 Um )

10 pC (liquid)

Part 10/05.85

1.1 Um / 3

50 pC (solid)

Tra nsforme rs a nd

DIN VDE 0532

1.3 Um

300 pC

Re a ctors

Part 3/07.87

1.5 Um / 3

500 pC

Dry Type
Tra nsforme rs

DIN VDE 0532

Part 6/01.94

1.1 Um / 3

Ta p Sw itch for
Tra nsforme rs

DIN VDE 0532

Part 3/04.93

Insula te d Sw itchge a rs
up to 38 kV
Ga s-insula te d Sw itchge a rs
(GIS)

IEC 44-4 (1980)

Identical
(liquid) Liquid insulation
(solid) Solid insulation

IEC 76-3 (1980)

Identical

20 pC (GH)

IEC 76-3 (1982)

1.1 Um / 3
(GH) Cast resin impregnated

1.5 Um / 3

50 pC

IEC 214 (1989)

Identical

DIN VDE 0670

Part 7/09.88

1.1 U
1.1 U / 3

100 pC (HP)
10 pC (GH)

IEC 466 (1987)

Identical

DIN VDE 0670

Part 6/04.94

1.1 U
1.1 U / 3

IEC 298 (1990)

Harmonized
HD 18755

DIN VDE 0670

Part 8/02.94

1.1 U
1.1 U / 3

IEC 517 (1990)

Identical

IEC 664-1 (1992)

Test procedure technical identic

Low Volta ge Compone nts

DIN VDE 0110

Part 20/08.90

Optocouple rs

DIN VDE 0884

/08.87

1.6 UI OR M

10 pC (GH, solid)

5 pC

Table 1: Summary of national and international partial discharge test standards for HV apparatus, derived from IEC 270 respectively DIN VDE 0434