Water in Transformers

Water in Transformers TVPPA E&O CONFERENCE — AUGUST 7–9, 2013

© SPX TRANSFORMER SOLUTIONS, INC.

TRANSFORMERS | SERVICE | TRAINING | COMPONENTS

Water in Transformers — August 2013

Rich Simonelli

Territory General Manager – Southeast

SPX Transformer Solutions, Inc.

Office 919-300-1522

Email [email protected]

Water in Transformers

Rich joined SPX in January 2005, bringing with him

over 20 years of experience in the power systems

industry. As the national service manager, Rich was

responsible for leading the Waukesha® Service group

in providing and implementing creative field service

solutions on electrical equipment used in power

generation, transmission and distribution, mass transit

and industrial processes, while maintaining customer

satisfaction, quality, environmental and safety systems

and processes. In 2011, Rich was promoted to

Territory General Manager where he has complete

commercial responsibility for all products and services,

business development, sales strategies and customer

relationships in the southeast region.

© SPX Transformer Solutions, Inc.



Agenda

Effects of water in

transformers

How does water get in?

How to estimate water

content in transformers

What can you do about it?

3



Effects of Water in Transformers

4




Water in Transformers is Bad

5

Cellulose is highly hygroscopic.

Influence of the temperature over water absorption of 3 mm

high-density pressboard at 50% relative humidity

Source: Water Absorption of Cellulosic Insulating Materials used in Power Transformers

HP. Gasser, Ch. Krause and T. Prevost, Weidmann Transformer Board Systems




Water in Transformers is Bad (cont.)

6

Moisture lowers the dielectric strength of oil.

Source: ABB IL 48-500-11, 2002





7

Moisture lowers the dielectric strength of solid insulation.

Source: Tom Prevost, EHV Weidmann, 2006





8

Moisture raises the dielectric power factor and increases

the risk of thermal breakdown of solid insulation.

Source: Tom Prevost, EHV Weidmann, 2006





9

Moisture lowers the lowest hot-spot temperature range for

possible bubble formation.

Source: TV Oommen, EPRI Reports:

EL-6761, March 1990; EL-7291, March 1992

50

70

90

110

130

150

170

190

0 2 4 6 8 10

WCP % w/w

Te

mp

era

ture

Kobayashi rapid heating

Kobayashi slow heating

Davydov

Oommen gas free

Oommen gas saturated

Comparison of “critical” bubble temperature vs.

water content in paper by three researchers




10





11

Moisture accelerates thermal aging of paper insulation.

1

2

3

4

5

6

7

8

9

10

0.5 1.5 2.5 3.5 4.5 5.5

Moisture in Paper (%)

Ag

ing

Acc

ele

ratio

n F

acto

r

IEEE Std C57.91-1995 Source: Sam Mehta, Tom Golner, SPX Transformer

Solutions, 2000 TJH2B presentation





12

Moisture can be the root cause of a catastrophic failure.



How Does Water Get In?

13




Many Ways for Water to Enter Transformers

14

Inadequate dry-out at factory

Leaks — gaskets and welds

Breathing in outside air during

oil temperature changes

Defective oil preservation

system




Water is Also Produced as a By-Product

15

Even if a transformer

is leak-free and the

oil preservation

system is performing

flawlessly, water is

PRODUCED as

a BY-PRODUCT of

the normal cellulose

aging process!

Degradation of Cellulose

O

H

O

H

O

H

OH

OH

O

H H

CH2O

CH2OH

H

O

HH

O

OH

H

Heating

Heating

C O O C O

O

H H

Carbon Monoxide Carbon Dioxide

Water

Section of

Cellulose

Molecule




Once Inside, Where Does It Go?

16

Once water is inside, the molecules are stored in various

parts of the insulation:

Oil

90%

Thick

5%

Thin

2% Winding

3%

Insulation Weight Distribution

Thick

55%Thin

22%

Winding

22%

Oil

1%

Water Distribution

Source: J. Aubin, 2005 Weidmann-ACTI Conference, San Antonio, TX



How Do We Measure / Estimate Water Content in Transformers?

17




Different methods can be used to assess water content in transformers:

Karl Fischer Titration of Insulation Sample

Moisture Equilibrium Curves

Dew point

Vapor Pressure Curve

Recovery Voltage Method

Dielectric Frequency Response

Power Factor

Cold Trap

IEEE C57.106 Limits

Variety of Assessment Methods Available

18




Moisture Equilibrium Curves

19

Tem

pera

ture

(

C )

Moisture in Oil (ppm)

Mo

istu

re in

Pap

er

( %

)

Source: Oommen 1983/2003




Moisture Equilibrium Curves (cont.)

Source: IEEE C57.106-2008

20

Challenges with Moisture Equilibrium Curves




Dew Point

21




Challenges with dew point to moisture content conversion:

At best, calculation is only an average “surface” moisture

content measurement

Difficult to achieve equilibrium due to temperature changes;

conversion is temperature and pressure sensitive

Difficult to measure/estimate insulation temperature

Moisture distribution is not uniform in insulation

Transformer must be without oil to get measurement

Despite these issues, dew point to moisture content

conversion is widely used as an acceptance test because

calculations are done without having to open the transformer.

Dew Point (cont.)

22




Vapor Pressure Rise/

Vacuum Drop Loss Method

Vapor Pressure Curve

Developed to monitor progress during

vacuum dry out operations. Equilibrium vapor pressure from the absolute

pressure in the transformer tank can be obtained

during the time condensed moisture is being

removed from the vapor trap. The transformer is

isolated from the vacuum pump and vapor trap.

A pressure rise (vacuum loss) test is then

conducted:

1. Leak rate of the transformer tank.

2. Outgassing rate for insulation and solid

components inside transformer (from change

in slope of pressure rise curve).

3. Pressure rise during the first minute after the

isolation valve is shut (rate of moisture

evaporation).

4. Intercept of the pressure rise curve with zero

time axis (equilibrium vapor pressure).




Vapor Pressure Curve (cont.)

Challenges with Vapor Pressure Rise / Vacuum Drop Loss

Problems/Factors directly

effect equilibrium of the vapor

pressure measurement:

Atmospheric air leaking into

transformer tank

Dissolved gas (air) released

from residual oil on the tank

bottom and/or oil-

impregnated insulation,

causing a rise in absolute

pressure (outgassing)

Moisture evaporating out of

the insulation

Application and calibration

errors in pressure and temp

measuring sensors

Temperature variations




Recovery Voltage Method (RVM) Testing the dielectric properties of the oil–paper insulation.

DC voltage (usually 2000V) is applied across the insulation for a period of time (charge time)

This DC voltage causes polarization of the molecules in the insulation material

Test piece is then discharged via short circuit for a period equal to half the charge time

Polarized insulation material then tries to revert to the original state which gives a characteristic response, known as the recovery voltage

Recovery voltage is measured and key points are noted—voltage, time and slope

Specimen is then discharged, ready for the next test cycle

RVM / DFR Tests

25




Recovery Voltage Method (cont.)

Tests are repeated for charge times

ranging from 20 ms up to

10,000 seconds

These points can then be plotted on

a series of graphs with varying

charge times

• The most important of these graphs

is called the polarization spectrum,

which is obtained by plotting peak

recovery voltage vs. charge time

The time at which this peak occurs is known as a dominant time

constant and is dependent on properties of the insulating materials

This value directly reflects the moisture content of the oil-paper

insulation system

RVM / DFR Tests (cont.)

26




Dielectric Frequency Response test

• Used to assess the integrity of a transformer’s

insulation system. The test attempts to determine the

volume of moisture and presence of contaminants in the

solid insulation as well as the conductivity and power

factor of the oil

• With the support of Omicron, DFR test have gained popularity

in recent years as a diagnostic tool for transformer insulation

system testing. The DIRANA uses a combination of the time

and frequency range method (Polarization Depolarization

Current [PDC] and Frequency Domain Spectroscopy [FDS]).


27




Challenges with RVM and DFR:

Difficult to achieve equilibrium due to temperature changes

Difficult to get accurate insulation temperature

Oil permittivity and conductivity change with aging of

insulation system

Moisture distribution is not uniform in insulation

Construction details are not always available

Transformer must be de-energized for a long time before

reaching equilibrium, making for a long test time duration


28




Power Factor has been used for many decades to assess the

drying process of transformers using a low voltage test or by the

use of probes and models to assess the condition of the solid

insulation.

Pros and Cons

Can be used for acceptance testing and at completion

of field drying after oil filling to ensure a dry unit

Can be used to detect pockets of water that may be deeply

imbedded and missed by other methods that

rely on equilibrium conditions to be accurate

No direct correlation between power factor and

moisture content

Transformer Power Factor

29




The Cold Trap method involves measuring the amount of water

collected in a trap over a four to six hour period and then calculating

the amount of water collected per hour. If the amount of insulation

and its initial moisture content are known, the

amount of water removed can provide

a sense of the insulation dryness.

Pros and Cons

Collecting the water in a trap prevents

it from collecting in the vacuum pump

Sometimes initial dryness and mass of insulation is difficult

to ascertain

Must have a cold trap available that efficiently catches the

majority of the water so it can be measured

Cold Trap Method

30




Classification of Service Aged Oils

Test Standard Unit Voltage Group 1 Group 2 Group 3 Group 4

< 69 kV 23 <23 - -

69 - 288 kV 28 <28 - -

> 345 kV 30 <30 - -

< 69 kV 0.2 - >0.2 >0.5

69 - 288 kV 0.15 - >0.15 >0.5

> 345 kV 0.1 - >0.10 >0.5

< 69 kV 25 - <25 <18

69 - 288 kV 30 - <30 <18

> 345 kV 32 - <32 <18

< 69 kV 35 >35 - -

69 - 288 kV 20 >20 - -

> 345 kV 12 >12 - -

< 69 kV 0.5 - >0.5 >1.0

69 - 288 kV 0.5 - >0.5 >1.0

> 345 kV 0.5 - >0.5 >1.0

Dielectric

Breakdown

ASTM-D1816

w/ 1mm gap min, kV

Neutralization

NumberASTM-D974

max, mg

KOH/g

Power Factor ASTM-D921 max, %

Interfacial

TensionASTM-D971

min,

Dynes/cm

Moisture

ContentASTM-D1533

max, PPM

@60°C Avg.

Oil Temp.

31

Source: IEEE C57.106-2006

danish

Rectangle



Water in Transformers — August 2013 32

Current measurement practices assume a state of equilibrium.

Why is Moisture Content So Difficult to Measure?

Varying Temperatures: Top oil temp

Bottom oil temp

Winding temp

Does Equilibrium

Ever Occur?




Does Equilibrium Ever Occur? (cont.)

Source : Transformer Moisture Content/Cold Weather Dew Point Measurement,

Doble Client Conference 2009; Rich Simonelli, SPX / Phil Prout, NGRID / Lance

Lewand, Doble / Brian Anderson, CSU




Does Equilibrium Ever Occur?

Factors Directly Affecting Equilibrium

Equilibrium, by definition, is “a state of balance between opposing forces or

actions that is either static (as in a body acted on by forces whose resultant is

zero) or dynamic (as in a reversible chemical reaction when the rates of reaction

in both directions are equal).” Where continuously varying conditions exist (internal and external), the

equilibrium curves cannot be applied directly.

In a Transformer:

Temperature stratification/unstable thermal environment

Non-uniform moisture distribution in the insulation

Only an average “surface” moisture content measurement

achievable with any method (best case scenario)

Varying thicknesses of insulation structures directly affecting

the rates of diffusion and adsorption




Does Equilibrium Ever Occur? (cont.)

Factors Directly Affecting Equilibrium in a Transformer (cont.)

Under Ideal Conditions: Vapor Diffusion is already impeded because water

molecules are physically deterred from moving rapidly by

the labyrinth of micro capillaries De-sorption of water molecules from the surface of active

sites on the cellulose surface occurs more slowly

because energy is required to break the bond The film of oil coating the surfaces of the insulation adds

another diffusion process



What Can Be Done About Water in Transformers?

36




Select the most suitable measurement method that

allows you to minimize error-producing variables,

continue using the same method over a long period of

time and monitor changes

Support industry-wide R&D initiatives to develop ways

to measure moisture content of working transformers

accurately and reliably

What Should You Do About the Moisture?

37




For a new unit:

Specify maximum

allowable moisture

content at delivery

and determination

method

Specify “proper” oil

preservation system

Specify “raised

flange” for

all openings

Specify gasket material and suitable seal design

Consider specifying optional accessories that help maintain integrity

of the unit

What Can Be Done to Limit Water Ingress?

38




What Can Be Done to Limit Water Ingress? (cont.)

39

For an existing unit:

Maintain the integrity of the unit — keep it sealed

Standardize on methods to monitor moisture content — monitor, observe, record

Implement loading policy to minimize the aging process

Consider adding accessories that help maintain integrity of the unit

Corrective Actions:

If it tests “wet” for your application, dry it out

• Off-line field drying: Fast, but costly

• On-line field drying: Slow, but economical

• Natural ester retrofill




Vapor Pressure

At atmospheric

pressure, you must

heat water in a

transformer to higher

than 212ºF (100ºC)

for water to boil off

When lowering the

pressure by pulling

vacuum, we can boil

water off without

heating (as long as

the temperature is

above freezing)

Transformer Drying Theory

40




Why Add Heat?

Heat required for a phase change comes from the surrounding oil and transformer parts

As heat moves from the insulation into the water, the temperature of the insulation and transformer drops

Therefore, replacement heat is added by continuously reheating the oil and pumping it into the transformer

If the heat is not replaced, freezing can eventually occur

Transformer Drying Theory (cont.)

41

VIDEO




Off-Line Field Dry Out Methods

Multiple methods exist for field drying a transformer if the

moisture content is found to be above acceptable limits:

Vacuum

Vacuum with hot air

Short circuit and vacuum

High vacuum with hot oil

42




Off-Line Field Dry Out Methods (cont.)

Vacuum

Good method for

removing small amounts

of residual moisture

Transformer is subjected

to high vacuum and held

for a period of time

Efficiency of method is

increased at higher temperatures

Cold traps can be used in vacuum line to measure

moisture extraction

43




Off-Line Field Dry Out Methods (cont.)

44

High vacuum and hot oil

Evacuate tank

Introduce oil to heat core/coil

assembly

• If capable, limit oil volume to

10% of total or enough to

establish oil circulation

• Otherwise, cover core/coil

assembly

Circulate oil under vacuum until

outlet oil temperature reaches

desired temperature

Drain oil from transformer

Continue to pull vacuum and monitor moisture through cold trap,

if desired




On-Line Field Dry Out Methods

45

Dry out process is slow

due to the diffusion rate of

water from cellulose to oil

Moisture content of oil

will decrease quickly

Higher operating

temperatures make

process more effective

Safety concerns should

be considered




Natural Ester Fluid Retrofill

46

Water Absorption of Dielectric FluidsExposed to Ambient Air (1 of 2)

Exposure Time (hrs)

0 500 1000 1500 2000 2500 3000 3500

Ab

so

lute

Wa

ter

Co

nte

nt

(pp

m)

0

100

200

300

400

500

600

conventional transformer oil

Envirotemp FR3 fluid

Source: Cooper Power Systems (prior to FR3 product line purchase by Cargill)

Natural ester fluid holds

more water in solution,

enhancing moisture

migration into the fluid

Hydrolysis converts

moisture and natural

ester fluid into long

chain free fatty acids

Trans-esterification

combines some of these

fatty acids with the OH group

in cellulose, strengthening the cellulose chain and making it

less likely to break




Natural Ester Fluid Retrofill (cont.)

47

Temperature (oC)

20 40 60 80 100 120

Wa

ter

Sa

tura

tio

n P

oin

t (m

g/k

g)

0

1000

2000

3000

4000

5000

Envirotemp FR3 fluid: A = 5.3318, B = 684from Doble Engineering

mineral oil: A = 7.0895, B = 1567from IEEE C57.106

T

BA

TSaturation 27310)(

Water Saturation Dielectric Strength

Water Content (mg/kg)

0 100 200 300 400 500 600 700D

18

16

Die

lectr

ic B

rea

kd

ow

n S

tre

ngth

(kV

)

0

10

20

30

40

50

60

70

80

FR3 Fluid

Mineral Oil






Source: Steve Moore, Transformer Insulation Dry Out By Retrofilling With Natural Ester

Dielectric Coolant, IEEE 2012

48

Aged Paper at 3% Water Content

(aged at 85oC)

Aging Time at 85oC (hrs)

0 500 1000 1500 2000

% W

ate

r in

Pap

er

0

1

2

3

4

5

Dis

so

lved

Wate

r in

Flu

id (

pp

m)

0

200

400

600

800

Paper in FR3 Fluid

Water in FR3 Fluid

Water in Mineral Oil

Paper in Mineral Oil





Source: Steve Moore, Transformer Insulation Dry Out By Retrofilling With Natural Ester

Dielectric Coolant, IEEE 2012

49

Time at 85oC (hr)

0 500 1000 1500 2000 2500 3000

Wa

ter in

Pa

pe

r (wt%

)

0

2

4

6

Wa

ter

in F

luid

(m

g/k

g)

0

200

400

600A

cid

Nu

mb

er (m

g K

OH

/g)

0

1

2

3

4

5

6

7

Envirotemp FR3 fluid - water content

water in paper

Envirotemp FR3 fluid - acid number





Sealed Tube Test - ML 152-2000

Upgraded Paper 2000 hr @ 170°C







Natural Ester Mineral Oil Natural Ester Mineral Oil Natural Ester Mineral Oil Natural Ester Mineral Oil


50

Life Extension Potential with Natural Ester Fluids



Questions?

51

Documents

Water in Transformers