Also known as: Frequency Domain Spectroscopy · Also known as: Frequency Domain Spectroscopy. 2 Frequency Domain Spectroscopy V A Hi Lo Ground C HL C L C H I U Z ... Dirana) 2-wire

1

IDAX 300Insulation Diagnostic Analyzer

Dielectric Frequency ResponseAlso known as:

Frequency Domain Spectroscopy

2


V

A

Hi

Lo

Ground CHL

CL

CH

I

UZ

and

PF tand,,C Z

Measure at several frequencies

Use Ohms law:

3

Capacitance and Dissipation Factor (Tan )

= Z (Impedance)

= Tan Loss tangent)

= Power Factor (cos or )

= C (Capacitance) j

Note: If cos and Tan small then cos =Tan

If Tan is 1*10-3 (0.001) then Iloss/I is 1/1000 which is equivalent to Iloss will be1m

and I 1000m, 1m / 1000m. Specification of instrument is 1*10-4 (1/10000).

ICAP @ 10nF, 200V and 50Hz ICAP = 2*Pi*f*U*C = 0.63mA

ICAP @ 1nF, 200V and 1Hz ICAP = 2*Pi*f*U*C = 1.26uA

4

IDAX Set-ups

IDA 200

Hi

Lo

Ground

A1

A2

V CX

Cs1

Cs2

UST - without Guard

5

IDAX Set-ups

IDA 200

Hi

Lo

Ground

A1

A2

V CX

Cs1

Cs2

UST - with Guard

6

IDAX Set-ups

IDA 200

Hi

Lo

Ground

A1

A2

V CX

Cs1

Cs2

GST - without Guard

7

IDAX Set-ups

IDA 200

Hi

Lo

Ground

A1

A2

V CX

Cs1

Cs2

GST - with Guard

8

What is Spectroscopy?

Method to isolate/identify building blocks in a

composite material

Example: Identify material composition in samples

from Mars

Example: Breaking down light into its different

colors using a prism

http://en.wikipedia.org/wiki/Image:Light_dispersion_conceptual.gif

9

Insulation testing/Dielectric response methods

Frequency, Hz0

1

2

3

4

5

6

7

0,000001 0,00001 0,0001 0,001 0,01 0,1 1 10 100 1000

FDS/DFR

HV Tan Delta

VLF

PDC

Polarization Index

"DC"

10

Dielectric Frequency Response - Power Factor Changes with Frequency

Frequency

Power factor 0.32 at 0.02 Hz

0.0031 at 60 Hz

11


Changes in insulating materials (ageing) affect the

capacitance and loss factor (PF, tan )

Frequency sweep, compared to traditional one-

frequency Power Factor/”Doble” test, provides a lot

more information on:

– Insulation characteristics

– Ageing effects

– Influence of temperature

– Etc…

12

Traditional Power Factor Testing

Dissipation

factor

Frequency

50 Hz1 mHz 1kHz

13

Dielectric Frequency Response

Dissipation

factor

Frequency

50 Hz1 mHz 1kHz

14

What affects the response?

- Oil Conductivity +

-M

ois

ture

+

-M

ois

ture

+

- Temperature +

15

Typical power factor values for oil

insulated transformers and bushings

IEEE 62-1995 states; “The power factors recorded for routine overall tests on

older apparatus provide information regarding the general condition of the

ground and inter-winding insulation of transformers and reactors. While the

power factors for most older transformers will also be <0.5% (20C), power

factors between 0.5% and 1.0% (20C) may be acceptable; however, power

factors >1.0% (20C) should be investigated.”

Typical power factor values @ 20 C

"New" "Old" Warning/alert limit

Power transformers,

oil insulated0.2-0.4% 0.3-0.5% > 0.5%

Bushings 0.2-0.3% 0.3-0.5% > 0.5%

16

Dielectric Frequency Response- Single tan delta value @ 0.7% is not enough to make the right decision

- Dielectric Frequency Response tells the story!

Wet transformer with good oil

Dry transformer with old

oil (high conductivity)

Same PF value at 60Hz

17

DFR Application Areas

Power transformers

Instrument transformers

Bushings

Motors and generators

Cables

Generic testing of insulation systems

19

Why measure moisture?

A transformer with low moisture content is

like a person in good condition

• A transformer can be loaded with confidence without

risk for catastrophic failure.

• A person can work hard without risk for heart attack

A wet transformer is like an overweight

person with clogged arteries.

• The transformer owner has to limit load to avoid

bubbling (explosion risk).

• Moisture in insulation increases the rate of aging

20

Moisture in Power Transformers

Power transformer insulation consists of oil impregnated cellulose and free oil. Almost all moisture is in the cellulose:

• 25 tons of oil with water content of 20 ppm = 0,5 kg

• 2.5 tons of cellulose with 3% water content = 75 kg

21

Water In Oil Analysis

Power transformer insulation consists of oil impregnated cellulose and free oil. Almost all water is in the cellulose:• 25 tons of oil with water content of 20 ppm = 0,5 kg

• 2.5 tons of cellulose with 3% water content = 75 kg

20ppm (parts per million) in 25 tons of oil

3% water in 2.5 tons of cellulose

22


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

2.5 tons of cellulose with 3% water content = 75 kg

Moisture content in oil (it is almost constant in the cellulose) varies with temperature and oil aging status: Aged oil resolves higher amounts of water than new oil

Small moisture concentration makes sampling difficult

23

Water In Oil Analysis

Oil samples taken at low temperatures have low accuracy because the water has migrated to the paper

Oil sample taken at 200C

4.0% water = 6 ppm

1.0 % water = 3 ppm

24




Moisture content in oil (it is almost constant in the cellulose) varies with temperature and oil aging status:

Aged oil resolves higher amounts of water than new oil


Moisture changes the dielectric properties of the cellulose paper/pressboard

25

Water Changes Paper

Water changes the insulation properties of the paper

Relative permittivity

26








Moisture accelerates ageing

27



Dry (0.5%) insulation @ 900C = 40 Years

Medium wet (2.0%) insulation @ 900C = 4-5Years

28









Moisture limits loading capability

29


Moisture determines the

maximum loading/hot-spot

temperature for bubble

inception (see e.g. IEEE Std

C57.91-1995)

Knowing moisture content

allows for correct decision

• Leave as-is

• Dry-out

• Replace

• Scrap or Relocate?100

150

200

250

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Hottest S

pot T

em

pera

ture

(OC

)

Moisture in Insulation (% wt)

0.5% moisture (New) = 2000C hotspot

3.0% moisture = 1200C hotspot

30

Interpretation of moisture content

< 0.5 % New transformer

0.5 - 1.5% Dry insulation

1.5 - 2.5% Medium wet insulation

2.5 - 4% Wet insulation

> 4% Very wet insulation

Interpretation of moisture content of solid insulation (% of

weight water per weight cellulose):

31

Moisture levels

1.0 %

4.2 %

2.6 %

?

32

Moisture estimation process

Measure tan delta/power factor from 1 kHz to 2 mHz

Send results to MODS

Enter insulation temperature (top-oil temperature)

MODS matches measured curve to modeled curve

(automatically) by varying parameters that affects

the shape of curve to find best match

Results:

• Relation of solid (cellulose) vs. liquid (oil) insulation

between winding (if not known)

• Moisture in solid insulation

• Oil conductivity

33

X-Y model of power transformer insulation

Typical values:X = % barriers in the main duct (15-55%)Y= % spacers of the circumference (15-25%)

Cellulose: Blue colorOil: Red color

34

Oil – paper – and combination

35

Measured

curve


Right click

Select

Send to…MODS

36

Measured

curve


% SpacersAmount of oil

Model curve

% Barriers

37


2. Enter

temperature

1. Check all

checkboxes

3. Click

Auto match

38


Result!

39

Why a Sense lead? Ideal situation

40

Why a Sense lead? Actual situation

R

41

Why a Sense lead? Actual situation

R

42

High frequency CHL measurements – Results

2-w CHL measurement

CHL=5729 pF

CL=8755 pF

Single-wire ampmeter

connection (IDA, Dirana)

2-wire ampmeter

connection (IDAX300)

Tand delta=0.56%

Tan delta=1.3 %

43

V

A

Hi

Lo

Ground CHL

CL

CH

High frequency CHL measurements – Issues

1. Ampmeter input + cable

Z is small but not zero...

2. Z is high but

not infinite...3. ”Leakage” current causes

measurement error at high frequencies

44

V

A

Hi

Lo

Ground CHL

CL

CH

High frequency CHL measurements – Solution

1. Ampmeter input + cable

Z is small but not zero...

2. Z is high but

not infinite...No ”leakage” current causing

measurement error at high frequencies!

Sense/Voltage output

Compensates for Z and

sets potential to ground

45

Noise in substations

Induced AC (50/60Hz)

Induced DC (HVDC stations)

DC offset from ground potential

Other disturbancies (RF, harmonics etc)

46

Noise in substations - AC

88 randomly selected IDA/IDAX measurements

“Noisy” limit set to 50µA

About 10-20% of all measurements are performed at > 50µA

noise current

Noise level, dB rel 10µA

-140,0

-120,0

-100,0

-80,0

-60,0

-40,0

-20,0

0,0

20,0

40,0

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88

Sorozatok1

47

Bushings and rotating machinery

48

Bushing Failures

49

DFR on Bushings

General Electric, ”Type U”

Bad condition

Good condition

50

High-Voltage DFR on Bushings (0.13-14kV)

51

DFR for Motors and Generators

Measure tan delta vs frequency to detect ageing

deterioration (end-of-life)

52

Verification of DFR/IDAX performance

Accredited test system at ABB transformers

Project REDIATOOL comparing DFR with other methods for measuring moisture in transformers. Results published at CIGRE 2006 (paper D1-207).http://www.uni-stuttgart.de/ieh/forschung/veroeffentlichungen/CIGRE_paperD1-207_Gubanski.pdf

Examples of other independent verification bodies• IREQ (Canada)

• EPRI (China)

• IEPS/Schering-Institute (Germany)

• CESI (Italy)

• KTH (Sweden)

• Tennesse Tech (USA)

53

ABB Transformer Diagnostics (press release)

54

IDAX 300 – Insulation Diagnostic Analyzer

55

IDAX 206FR – Insulation Diagnostic Analyzer

56

IDAX System

HW

Test signal: 0 – 200 V peak 0 – 50 mA peak

Frequency: 0.0001 Hz – 10000 Hz (IDAX-300)

Sample range: 10 pF – 100 µF

2-ch measurement (IDAX-300)

SW

IDAX SW for measurement control and analysis

Moisture analysis SW (MODS)

• Automatic moisture assessment in oil impregnated cellulose

57

Cables

58

IDAX Software

Nameplate InfoObject and Test

Browser

59

IDAX Software

Test settings

Graphical

Hook-Up

Diagram

60

Software

Graph View

Presentation

selector

Power Factor

Chosen

61

IDAX Software

Result at

50/60Hz

62

Summary IDAX 300

Instrument for moisture assessment in transformer insulation

Built in modeling software for quick and reliable results

Measures the moisture where it is, at any temperature

Light weight, fast and accurate

Can also be used for other applications

• CT‟s

• Bushings

• Rotating machines

• Cables

Reveals the reason for high tan delta values - Moisture or

contamination?

Well documented and well proven method

63

Cable Diagnostics

64

Paper Insulated (PILC) Cables

Moisture in cable insulation

• is generated by ageing due to water ingress

from outside

• accelerates the ageing process

Presence of elevated moisture content

can be detected by DFR (IDAX)

measurements

65

Cable Paper – Effect of Moisture

Tan

0.001

0.01

0.1

1

10

0.001 0.01 0.1 1 10 100 1000

mc < 0.20 %

mc = 1.60 %

mc = 1.99 %

mc = 2.62 %

mc = 3.50 %

mc = 4.20 %

frequency, [Hz]

60 Hz

50 Hz

Tan minimum

66

Examples of Field Measurements

Measurements performed at 200 Vpeak on field installed cables in Sweden

1 0-3

1 0-2

1 0-1

1 00

1 0-3

1 0-2

1 0-1

1 00

1 01

1 02

1 03

H 1 0 1 T K , L 1

B e kö , f3

R a v , f2

S kra d d , f2

los

s t

an

ge

nt,

ta

n

f re q u e n cy , [H z ]

1 0-3

1 0-2

1 0-1

100

10-3

1 0-2

1 0-1

100

101

1 02

1 03

P eder, f1

L50 56 , f1

R åd m g 2 , f2

los

s t

an

ge

nt,

ta

n

f requ ency , [H z ]

67

Estimation of Moisture Content

Minimum

PF/tan

Estimated moisture

contentCondition

0.002-0.0035 Below 1% Good

0.0035-0.005 1 – 2.5 % Moderately aged/moistened

0.005-0.01 2.5 – 3.5 % Considerably aged/moistened

Above 0.01Above 3.5% or a local

defect

In bad condition/high moisture

content

68

Summary: Paper Insulated Cables

Moisture gives a characteristic increase of the frequency dependent capacitance and loss of cellulose paper

Temperature affects the frequency response of cellulose paper

• By using measurement data in a frequency interval accurate temperature corrections are possible

Loss tangent minimum can be used as a criterion for assessment of moisture content

69

Medium Voltage XLPE Cable Circuit Problems

Terminations

• Manufacturing defects

• Bad mounting

• Aging

Joints


• Bad mounting

• Aging

Cables


• Damaged during installation

• Aging (water trees)

70

After installation diagnostic

measurements and voltage tests

Terminations

• Partial discharges (acoustic or electrical)

• Voltage test (not very effective)

Joints

• Partial discharges (acoustic or electrical)

• Voltage test (may be effective)

XLPE cables

• Voltage test

• (Partial discharge measurements)

71

Voltage tests

DC

• Not effective in most cases. Using high voltages may damage the

cable

50 Hz

• Requires very large equipment

VLF (very low frequency, 0.1 Hz)

• Cost-effective, moderate sized equipment

72

Water Tree Deterioration of MV XLPE Cables

Water trees are growing in the insulation and lower

the electrical withstand

The water tree aging process is very slow

A heavily aged cable fail if the insulation stress is

increased (lightning impulses, faults, etc)

73

Medium Voltage XLPE Cables

High voltage tests shorten cable life significantly

Non-destructive diagnostics should be used

Water-treeing can only be detected with

increasing voltage level (non-linear effect)

Use IDAX + VAX high-voltage unit!

74

Measurement Procedure

1. Measurements of short frequency sweeps, from

about 10 Hz down to 0.1 Hz at 25, 50, 75, 100 and

(repeat) 50% service voltage level, U0

2. Classification of cable quality

75

Cable quality classification

1. Low Losses and No Voltage Dependence

2. VDP Response (Voltage Dependent Permittivity)

• A voltage dependent increase of capacitance and loss.

3. TLC Response (Transition to Leakage Current)

• A VDP response at initial low voltage levels. At a higher

voltage level, the response characteristics changes due

to leakage current.

4. LC Response (Leakage Currents)

• Leakage currents through water trees are present

already at low voltage levels.

76

Capacitance part Loss (Tan ) part

New/non-deteriorated XLPE cable: Low losses and no voltage dependence

10-4

10-3

10-2

10-1

0,01 0,1 1 10

3 kV6 kV

Frequency (Hz)

'

10-4

10-3

10-2

10-1

0,01 0,1 1 10

3 kV6 kV

Frequency (Hz)

''

77

10-4

10-3

10-2

10-1

0,01 0,1 1 10

3 kV 4,5 kV 6 kV 3 kV 6 kV 3 kV

Frequency (Hz)

'''''''

10-4

10-3

10-2

10-1

0,01 0,1 1 10

3 kV 4,5 kV 6 kV 3 kV 6 kV 3 kV

Frequency (Hz)

''

''''''''''''


VDP Response: Voltage dependent increase of loss and capacitance

78


TLC Response: Leakage current detected at increasing voltage level + “memory effect”

10-4

10-3

10-2

10-1

0,01 0,1 1 10

1,5 kV 3 kV 4,5 kV 6 kV 3 kV

Frequency (Hz)

'

'

''

'

'

10-4

10-3

10-2

10-1

0,01 0,1 1 10

at 1,5 kV at 3 kV at 4,5 kV at 6 kV at 3 kV

Frequency (Hz)

"

"""

""

79


LC Response:Leakage current (through water-trees) already at low voltage levels

10-4

10-3

10-2

10-1

0,01 0,1 1 10

at 3 kV at 6 kV at 9 kV

Frequency (Hz)

'

''

'

10-4

10-3

10-2

10-1

0,01 0,1 1 10

at 3 kV at 6 kV at 9 kV

Frequency (Hz)

"

"""

80

Evaluation of Water-Tree Deteriorated Cables based on laboratory and field experience

LC or TLC response• The cable is judged bad. The voltage withstand level is usually below 2.5

times service voltage level. Depending on the leakage current level, cable

design and the voltage level of the network, the cable can be used for a

short time or must be replaced immediately.

VDP response (Significantly aged)• The cable is significantly aged. The voltage withstand is typically 2.5 – 4

times the service voltage level. Depending on cable design and loading,

the cable can remain in service for several years or should be scheduled

for replacement.

No ageing detected• The cable is judged good and has typically a voltage withstand above 4

times service voltage. However „good‟ condition does not necessary mean

that the cable does not have any water trees. Repeated measurement is

recommended within a 5-10 year period.

81

No cable judged good has been reported to fail in

service

Significant aged cables, i.e. cables with VDP response,

can withstand many years of service life without failure

Cables with TLC and LC currents are bad. Such a cable

may be allowed remain in service a few months in 6 or

10 kV networks

Experience from a project

82

Approximately 60 cables circuits installed in early

seventies

Increasingly rate of cable failures

Cable circuits failed during VLF testing

In 1996, all circuits were measured and LC and TLC

cables were replaced (only a few)

25-30 cable circuits were replaced 1997-1999 based on

level on VDP response

Case Study: North Botkyrka

83

0

2

4

6

8

10

12

1993 1994 1995 1996 1997 1998 1999

Cable Faults in North Botkyrka

Fau

lts/

100

km

/Yea

r

Year

• No faults 1997- 2000. One

fault 2001. 5 faults summer

2003.

• All faults were in cables with

strong VDP-response in

1995 or 1996.

• A few circuits was re-

measured autumn 2002.

The response of slightly

aged cables indicated that

they still are in good shape

(6-7 years later)

Case Study: North Botkyrka

84

Summary: Using DFR on MV XLPE Cables

• Voltage test may damage water tree deteriorated XLPE

cables

• Relatively low voltage levels (up to service voltage) are

used in order to ensure non-destructive measurements

• The process of water tree deterioration is very slow

• By non-destructive measurements cable replacement can

be postponed and scheduled

• The response of water trees can be identified and

classified

Documents

Also known as: Frequency Domain Spectroscopy · Also known as: Frequency Domain Spectroscopy. 2 Frequency Domain Spectroscopy V A Hi Lo Ground C HL C L C H I U Z ... Dirana) 2-wire