Thermal Aging Collection PDF

Contents
1) C. P. McShane, G.A. Gauger, and J. Luksich, Fire Resistant Natural Ester Dielectric Fluid and Novel
Insulation System for Its Use, IEEE/PES Transmission & Distribution Conf., April 12-16, 1999, New
Orleans, USA
Discusses accelerated aging of full scale production transformers per ANSI/IEEE C57.100-1986 using the
Envirotemp FR3 fluid/thermally upgraded Kraft paper insulation system.

2) C.P. McShane, K.J. Rapp, J.L. Corkran, G. A. Gauger, and J. Luksich, Aging of Paper Insulation in
Natural Ester Dielectric Fluid, IEEE/PES Transmission & Distribution Conf, Oct. 28 - Nov. 2, 2001,
Atlanta, USA
Discusses accelerated aging of small scale sealed tube aging systems per ANSI/IEEE C57.100-1999 Annex
A using the Envirotemp FR3 fluid/thermally upgraded Kraft paper insulation system.

3) C.P. McShane, K.J. Rapp, J.L. Corkran, G. A. Gauger, and J. Luksich, Aging of plain Kraft paper in
natural ester dielectric fluid, IEEE/DEIS 14
th
International Conf. on Dielectric Liquids, July 7-12,
2002, Graz, Austria
A using the Envirotemp FR3 fluid/non-thermally upgraded Kraft paper insulation system.

4) C.P. McShane, K.J. Rapp, J.L. Corkran, and J. Luksich, Aging of cotton/Kraft blend insulation paper
in natural ester dielectric fluid, 4
th
TechCon Asia-Pacific Condition Based Maintenance &
Diagnostic Conf., May 8-9, 2003, Sidney, Australia
A using the Envirotemp FR3 fluid/cotton-Kraft blend paper insulation system.

5) C.P. McShane, J.L. Corkran, K.J. Rapp, and J. Luksich, Aging of paper insulation retrofilled with
natural ester dielectric fluid, IEEE/DEIS International Conf. on Electrical Insulation and Dielectric
Phenomena, Oct. 19-22, 2003, Albuquerque, USA
Simulates retrofills of mineral oil transformers with Envirotemp FR3 fluid using accelerated aging of small
scale sealed tube aging systems per ANSI/IEEE C57.100-1999 Annex A.

6) K.J. Rapp, C.P. McShane, and J. Luksich, Interaction Mechanisms of Natural Ester Dielectric Fluid
and Kraft Paper, IEEE/DEIS 15
th
International Conf. on Dielectric Liquids, June 26 - July 1, 2005,
Coimbra, Portugal
Proposes mechanisms for the reduced aging rate of cellulose insulation in Envirotemp FR3 fluid.

7) A.W. Lemm, K.J. Rapp, and J. Luksich, Effect of Natural Ester (Vegetable Oil) Dielectric Fluid on the
Water Content of Aged Paper Insulation, EIA/IEEE 10
th
Insucon International Electrical Insulation
Conf., May 24-26, 2006, Birmingham, UK
Verifies the presence of the initial mechanistic steps leading to the reduced aging rate of cellulose
insulation in Envirotemp FR3 fluid at normal transformer operating temperatures.

Presented at the 1999 IEEE/PES Transmission & Distribution Conference, April 12-16, 1999, New Orleans, LA
Page 1 of 5
Fire Resistant Natural Ester Dielectric Fluid and Novel
Insulation System for Its Use

C. Patrick McShane, Member, IEEE
Transformer Products
Cooper Power Systems
Waukesha, WI 53188 USA
Gary A. Gauger, Member, IEEE
Thomas A. Edison Technical Center
Franksville, WI 53126 USA
John Luksich
Transformer Products
Waukesha, WI 53188 USA
Abstract: Insulation systems for electrical power and distribution
transformers are being re-evaluated based on total owning cost from
both economic and risk exposure perspectives. A recently developed
high fire point dielectric fluid, based on natural esters, has superior
safety, environmental, and health properties compared to current
dielectric fluids. An insulation system developed for this fluid
addresses the lower inherent resistance of esters to oxidation.
Accelerated transformer life testing compares transformers using
standard mineral oil to those with the natural esters dielectric
system. Results of these accelerated life tests show the ester fluid to
be suitable for transformer use and suggest the possibility of
extended insulation life compared to mineral oil systems.
Keywords: Dielectric coolant, dielectric materials, dielectric
measurements, ester insulation, life estimation, liquid dielectric
materials, power transformer insulation testing, oil insulation,
transformers
I. INTRODUCTION
It is becoming increasingly apparent that dielectric fluids
must provide a better balance of functional performance
inside the transformer versus environmental impact in the
event of release. Inside the transformer, a stable, chemically
inert fluid having good thermal and dielectric properties is
desired. Externally, the fluid should become environmentally
benign, and be readily and completely biodegradable.
Liquid-filled transformers have long used mineral oils as
the insulating fluid. Halogenated dielectric fluids, principally
Askarel fluids, once promoted for their excellent fire safety
properties, are now undesirable due to health hazards and
environmental persistence. The phase-out of Askarels led to a
succession of other non-flammable halogenated fluids such
as perchloroethylene, chlorobenzenes, and chloro-
fluorocarbons. These have since completely exited the new
transformer market.
In applications where fire safety is a concern, silicone oils
and high molecular weight hydrocarbons (HMWH) continue
to be the prevalent less flammable fluid choices. To a much
lesser extent, synthetic ester-based fluids are also used.
Synthetic ester dielectric fluids, most commonly aliphatic
polyol esters, have suitable dielectric properties and are more
readily biodegradable than mineral oil and HMWH fluids.
Due to their high cost compared to other less-flammable
fluids, synthetic esters are used mainly in specialty
applications such as traction and mobile transformers.
Natural esters have been considered unsuitable for use in
transformers, although evaluations in capacitor applications
hint at considerable potential [1,2]. Their susceptibility to
oxidation has been a primary obstacle to utilization as a
dielectric fluid. However, modern transformer design
practices, along with suitable fluid additives and minor design
modifications, compensates for this characteristic.
Outside the transformer, oxidative degradation is regarded
as an environmentally beneficial property. The application of
natural esters in transformers can achieve a balance of
desirable transformer and external environmental properties
not found in other dielectric fluids. An attractive source of
natural esters are the seed-based edible oils. Used mainly in
foodstuffs, these agricultural commodity oils are a renewable
resource increasingly utilized in non-food industrial
applications.
This paper evaluates, side by side with mineral oil, full
scale accelerated aging of a recently developed insulation
system based on natural food-grade seed oil esters.
Transformer insulation life is primarily related to aging of the
cellulose paper [3]. By exposing the insulation system to
elevated levels of thermal, mechanical, and electrical stresses,
transformer end-of-life can be found in a practical amount of
time. The normal life expectancy can then be extrapolated
from the life span found for high stress levels.
The IEEE Standard Test Procedure for Thermal Evaluation
of Oil-Immersed Distribution Transformers [4], often referred
to as the Lockie Method, is the protocol used to evaluate
the suitability of this insulation system for transformer
applications. The normal expected life of a transformer,
operating as defined in the IEEE loading guide [3] at a 30C
average ambient, is about 20 years. The average duration of
this test is 500% of expected life, or the equivalent of 100
years of operation.
Page 2 of 5
II. APPARATUS
A. Test Specimens
Three identical sets of test transformers were constructed as
summarized in Table 1. Test specimens were slightly
modified production 1PH, 15kVA, 7200/12470Y-120/240V
transformers. The design change consisted of a modest
decrease in coil ducting to simplify reaching the desired hot
spot temperatures. All specimens included thermocouples to
monitor top oil temperature, a nitrogen blanket in the
headspace, and a headspace pressure transducer. Solid
insulation was kraft paper thermally upgraded for 65C rise
use.
All bushing gaskets were nitrile. Units A3, B3, and C3
contained mineral oil and used nitrile cover gaskets. All other
units contained the ester fluid and had either Viton or nitrile
cover gaskets. Viton was selected to reduce the oxygen
permeability of the system.
Transformers designated as thermal monitors were fitted
with strategically placed thermocouples wound inside the
coils to determine the hottest-spot temperature. The thermal
monitors carried load current with their secondary windings
short-circuited. This allowed rated secondary current to flow
while minimizing the voltage gradients on the coil
thermocouples.
To safeguard the more easily oxidized esters, the fluid
blend contained food-grade antioxidant additives. Some of
the ester-filled transformers were fitted with a proprietary
headspace oxygen absorber system. The system consists of an
absorber material contained in an oxygen-permeable
container. The container serves to isolate the absorber from
the dielectric fluid and still allow oxygen to reach the active
material. Two absorber formulations were evaluated.
B. Test Cells
Three test cells, designed specifically to meet the IEEE
thermal evaluation requirements, were constructed one test
cell for each test temperature. Each cell contained five test
transformers, one of which was a thermal monitor. The
primaries of the five specimens were wired in series to ensure
that all units carried the same current and therefore were
subjected to the same thermal stress.
Current levels of the 7.2kV primaries were automatically
controlled to maintain the hottest-spot temperature. The
secondary windings were energized at the rated 240V. Data
loggers collected 60 channels of test cell data. Test conditions
were maintained using motor-driven current and voltage
autotransformers. The autotransformers were in turn
regulated by dedicated programmable logic controllers
monitoring the data loggers.
III. PROCEDURE
A. Test Parameters
Transformer loading was determined using (1), the
Arrhenius reaction rate adapted for insulation deterioration of
65C rise transformers [3]. This model uses the hottest-spot
temperature T (K) to calculate expected life h (hrs). Fig. 1
shows the normal expected life curve.
log
10
life(h) = + 11269
63288
.
.
T
(1)
The thermal evaluation test specifies that the transformers
under test survive 500% of expected life at test temperature.
Three test temperatures were selected. Test time accumulated
only when the hottest-spot temperature was no lower than 3C
below set point.
B. Test Sequence
The total aging time, five times the expected life at
temperature, was divided into about 10 test periods. Each
period, representing half the expected life calculated from (1),
consisted of maintaining the hottest spot temperature for a
prescribed number of hours, then allowing the temperature to
fall to ambient. This elevated/ambient temperature cycle was
repeated four times per period. Cooling to ambient took about
nine hours. Time at ambient was at least 24 hours. Rise time
back to temperature set point was about six hours.
After each completed accelerated aging period, the
specimens were taken off-line and subjected to a series of
Table 1. Transformer test cell specimens for accelerated aging evaluation.
Test Unit Base Fluid
a
O2 Absorber Cover Gasket
1 (thermal monitor) natural ester none Viton
b
2 natural ester empty container Viton
3 mineral oil none nitrile
4 natural ester formulation I Viton
5 natural ester formulation II Viton
C4b (spare) natural ester none nitrile
a
CPS Envirotemp FR3 ester or Exxon Univolt inhibited mineral oil
b
Viton is a registered trademark of DuPont Dow Elastomers
100 120 140 160 180 200
Hottest-Spot Temperature (
o
C)
0.01
0.1
1
10
100
N
o
r
m
a
l

L
i
f
e

E
x
p
e
c
t
a
n
c
y

(
y
r
s
)
test temperatures
30
o
C ambient
100
1000
1E+4
1E+5
N
o
r
m
a
l

L
i
f
e

E
x
p
e
c
t
a
n
c
y

(
h
r
s
)
10
2
10
3
10
4
10
5
Fig. 1. Liquid-type transformer life expectancy for 65C
insulation systems based on the hottest-spot temperature [4].
Page 3 of 5
four electrical/mechanical stress tests. Dielectric fluid samples
were also taken after each test period. Table 2 shows the
accelerated aging and stress test parameters.
IV. RESULTS
Units A5, B5, C2, C3, and C4b completed 500% of normal
expected life and were disassembled for inspection. Units A1,
A2, A3, B1, B2, B3, C1, and C5 completed the 500% of
expected life and remained on test. Units A4, B4, and C4
were removed from test early in the process due to transient
test system malfunctions.
A. Visual Inspection
Tear-down and visual inspection of ester-filled units A5
and B5 showed the high-to-low insulation barrier discolored
and brittle in the hottest-spot region. The paper away from the
hottest-spot was in good condition. Neither unit contained
free water or sludge. Gaskets showed a slight decrease in
flexibility.
In the severest temperature cell, ester-filled units C2 and
C4b were in better overall condition than mineral oil unit C3.
The kraft paper at the hottest-spot in C2 was charred and
moderately deteriorated. The paper away from the hot-spot
was still in good condition. C4b showed notably less paper
deterioration than did C2
1
. Neither unit contained free water
or sludge, and both tanks were free of corrosion. The gaskets
in C2 (Viton/nitrile) and C4b (nitrile) were flexible with little
set, and appeared to be in good condition.
Tear-down and visual inspection of mineral oil unit C3
revealed significant deterioration throughout the transformer.
The kraft paper solid insulation was almost entirely
deteriorated, with the hottest-spot well dispersed around the
coil assemblies. The tank bottom contained free water,
sludge, and signs of corrosion. The wire insulation and
gaskets were hard and brittle.

1
C2 and C4 experienced a thermal runaway condition due to power
shedding affecting the HVAC system on a very hot day. C4b replaced C4
after this unusual event.
B. Fluid Analysis
Dielectric strength is shown in Fig. 2. The breakdown
voltage remained between 30-40kV for all ester fluids over
the course of the test. Most mineral oil samples were in this
range, although three samples fell below the 30 kV acceptable
level for new fluid [6].
Flash and fire points, shown in Fig. 3, dropped only
slightly over the life of the test. Viscosity, stable over the life
of the tests, is shown in Fig. 4. Moisture content, represented
as percent of fluid saturation, is presented in Fig. 5.
V. CONCLUSIONS
The main objective of submitting the natural ester-based
insulation system to accelerated aging tests was to determine
its suitability for use in distribution-class transformers. The
transformers survived the full five times expected life test
Table 2. Test parameters for transformer accelerated aging evaluation.
Test Cell A B C
Hottest-Spot Temperature (C) 167 175 183
Expected Life (hrs) 1,302 721 407
500% Expected Life (hrs) 6,510 3,604 2,036
Aging Periods 11 9 10
Cycles/Aging Period 4 4 4
Duration (hrs/cycle) 150 100 50
Stress Tests (performed off-line at the completion of each aging period)
Short Circuit 25 x rated current for 2 sec
Full Wave Impulse 62 kV (65% of new unit value)
Applied Potential 22 kV, 60 Hz, 1 min (65% of new)
Induced Potential 400 Hz, 130% rated, 7200 cycles
0 20 40 60 80 100 120
Time on Test (% of 5x life)
0
15
30
45
60
B
r
e
a
k
d
o
w
n

V
o
l
t
a
g
e

(
k
V
)
ester fluid
mineral oil
Fig. 2. Dielectric Breakdown Voltage [5] of ester fluid and mineral
oil versus aging time as a percent of five times normal expected life.
0 20 40 60 80 100 120
100
150
200
250
300
350
400
T
e
m
p
e
r
a
t
u
r
e

(

o
C
)
ester fluid
mineral oil
fire point
flash point
Fig. 3. Flash and fire points [7] of ester fluid and mineral oil versus
aging time as a percent of five times normal expected life.
Page 4 of 5
duration. Visual inspection showed less deterioration in the
ester-filled units than was found in the mineral oil units. No
undue oxidative deterioration of the ester fluid was observed.
Based on these results, this ester dielectric fluid used in
conjunction with standard transformer materials successfully
completed the requirements of 65C rise transformer
insulation systems. The precautionary choice of Viton as a
cover gasket material was found to be unnecessary.
The key fluid performance characteristics (dielectric
strength, flash point, fire point, and viscosity) remained close
to the new fluid values. The changes in the indicators
frequently used to monitor fluid condition (moisture content,
neutralization number, DC leakage, interfacial tension, and
dissipation factor) were higher than found in mineral oils. The
indicator levels found are characteristic of ester-based fluids
and had no observable impact on transformer performance or
life.
The ester fluid dissipation factor and absolute moisture
content, higher than those found in the mineral oil, reflect the
chemical nature of the natural ester fluid. The ester fluid has a
much higher tolerance for moisture, having a saturation level
of greater than 1000 ppm of water compared to 40-60 ppm at
25C for standard mineral oil [9]. This higher moisture
content results in some hydrolysis of the fluid, forming the
mild free fatty acids typical of ester-based fluids.
The water present inside a properly sealed transformer
increases when cellulose degrades. The high moisture affinity
of the ester fluid hampers the formation of free water over the
temperatures seen in this test. The presence of free water in
the mineral oil unit is a result of the low moisture saturation
point forcing condensation and pooling at room temperature.
The aging characteristics of ester-based fluids generally
differ from those of mineral oil. Interpreting the significance
of the fluid condition indicating properties must take the
differences into account. Changes that are insignificant for the
ester fluid might appear to be significant in a mineral oil
system.
Visual inspections of the paper insulation after aging
suggest that the ester fluid extends paper life. These results
correlate with the preliminary small-scale accelerated life
tests conducted prior to the Lockie test.
Continuing these Lockie Method aging tests beyond the
500% expected life to actual failure will assist us in
formulating specific loss of life curves for insulation systems
based on this ester fluid. Additional small scale accelerated
life tests to evaluate paper aging in the ester fluid and mineral
oil will provide degree of polymerization and tensile strength
measurements, allowing a direct quantitative comparison.
VI. ACKNOWLEDGMENTS
The authors gratefully acknowledge the substantial
contributions of Jerry L. Corkran, Kevin J. Rapp, and Brian
E. Schoessow for their work on this project.
VII. REFERENCES
[1] F.M. Clark, Insulating Materials for Design and Engineering Practice,
New York, Wiley, 1962, pp. 210-218, 342-345
0 20 40 60 80 100 120
Time on Test (percent of 5 x normal life)
0
25
50
75
100
0
25
50
75
100
M
o
i
s
t
u
r
e

C
o
n
t
e
n
t

(
%

s
a
t
u
r
a
t
i
o
n
)
0
25
50
75
100
mineral oil
ester fluid
Hottest Spot = 183
o
C
Hottest Spot = 175
o
C
Hottest-Spot = 167
o
C
Fig. 5. Moisture content [8], as percent of fluid saturation, of
mineral oil and ester fluid versus aging time as a percent of five
times normal expected life.
0 20 40 60 80 100 120
0
4
8
12
V
i
s
c
o
s
i
t
y

(
c
S
t
)
0
20
40
ester fluid
mineral oil
viscosity at 40
o
C
viscosity at 100
o
C
Fig. 4. Viscosity of ester fluid and mineral oil versus aging time
as a percent of five times the normal expected life.
Page 5 of 5
[2] A.C.M. Wilson, Insulating Liquids: Their Uses, Manufacture and
Properties, New York, Institution of Electrical Engineers, London and New
York, Peter Peregrinus, 1980, p. 171
[3] ANSI/IEEE C57.91-1981 IEEE Guide for Loading Mineral-Oil-
Immersed Overhead and Pad-Mounted Distribution Transformers Rated 500
kVA and Less with 65 C or 55 C Average Winding Rise, Distribution,
Power, and Regulating Transformers, New York, Institute of Electrical and
Electronics Engineers Inc, 1995
Note: The referenced loading guide was current when this test began. It is
superseded by IEEE Standard C57.91-1995 IEEE Guide for Loading
Mineral Oil Immersed Transformers, New York, Institute of Electrical and
[4] ANSI/IEEE C57.100-1986 IEEE Standard Test Procedure for Thermal
Evaluation of Oil-Immersed Distribution Transformers, Distribution,
Power, and Regulating Transformers, New York, Institute of Electrical and
[5] ASTM D 877-87 Standard Test Methods for Dielectric Breakdown
Voltage of Insulating Liquids Using Disk Electrodes, 1998 Annual Book of
ASTM Standards, Vol. 10.03, Electrical Insulating Liquids and Gases;
Electrical Protective Equipment, American Society for Testing and
Materials, 1998
[6] ASTM D 3487-88 Specification for Mineral Insulating Oil Used in
Electrical Apparatus, 1998 Annual Book of ASTM Standards, Vol. 10.03,
Electrical Insulating Liquids and Gases; Electrical Protective Equipment,
American Society for Testing and Materials, 1998
[7] ASTM D 92-90 Standard Test Method for Flash and Fire Points by
Cleveland Open Cup, 1996 Annual Book of ASTM Standards, Vol. 5.01,
Petroleum Products, Lubricants, and Fossil Fuels, American Society for
Testing and Materials, 1996
[8] ASTM D 1533-96 Standard Test Methods for Water in Insulating
Liquids (Karl Fischer Reaction Method), 1998 Annual Book of ASTM
Standards, Vol. 10.03, Electrical Insulating Liquids and Gases; Electrical
Protective Equipment, American Society for Testing and Materials, 1998
[9] Univolt 60 and N61 Water Saturation Curve, Univolt Electrical
Insulating Oils, Lubetext DG-2C, Exxon Corp, 1988, p. 7
VIII. BIOGRAPHIES
C. Patrick McShane was born in
Winnebago, Minnesota, on April 27, 1947.
He received a BS in Electrical Engineering
from Marquette University in 1970, and an
MS in Engineering Management from the
Milwaukee School of Engineering in 1998.
Currently the Product Line Manager for
Dielectric Fluids at Cooper Power Systems
Transformer Products, his employment
experience includes International Area
Manager for RTE Corp. and Regional
Technical Director for the State of Sao Paulo
(Brazil) Rural Electrification Program. His
professional activities include USA Delegate
IEC TC99, IEC TC99 Liaison to TC64, IEC TC89 Expert Delegate, Chair
ASTM W.G. D-5222, and W.G. Chair IEEE TC Dielectric Fluids
Subcommittee C57.121. Several of his proposals have been adopted by
USA Codes and Standards (NEC, NESC, FMRC). He has presented papers
at domestic and international engineering conferences including IEEE,
EPRI, Doble, and CIRED. He is a Factory Mutual ANSI Accreditation
Standards Reviewer.
Gary A. Gauger received his BS in
Chemistry from the University of Wisconsin-
Oshkosh in 1963, and studied organic
chemistry in the University of Wisconsin-
Milwaukee Graduate Program during 1970-
1972. Professional affiliations include the
American Chemical Society Division of
Petroleum Chemistry and Division of
Environmental Chemistry, American
Association for the Advancement of Science,
and the ASTM D-27 Dielectric Liquids and
Gases Committee. Gary joined Cooper Power
(then McGraw-Edison) in 1970 as an
Analytical Scientist. He has been involved in
dielectric materials development since 1975, and is currently Manager of
Analytical Services for the Thomas A. Edison Technical Center.
John Luksich was born in Milwaukee,
Wisconsin, on February 4, 1954. He received
his BS in Chemistry in 1980 and MS in
Materials Engineering in 1990, both from
the University of Wisconsin. His
engineering career includes thin film
materials development at the McDonnell
Douglas Space & Physics Laboratory and
sensor development for Johnson Controls.
He is currently an engineer with the Cooper
Power Systems Transformer Products
Dielectric Fluids group.
Bulletin IEEE 012 April 1999 2004 Cooper Power Systems or its affiliates

2001 IEEE/PES Transmission & Distribution Conference & Exposition, Oct. 28 - Nov. 02, 2001, Atlanta GA
0-7803-7287-5/01/$17.00 2001 IEEE
Aging of Paper Insulation in Natural Ester
Dielectric Fluid
C. Patrick McShane, Member, IEEE Kevin J. Rapp Jerry L. Corkran, Member, IEEE
Dielectric Fluids Group Thomas A. Edison Technical Center Transformer Products
Cooper Power Systems Cooper Power Systems Cooper Power Systems
Waukesha, WI 53188 USA Franksville, WI 53126 USA Waukesha, WI 53188 USA
Gary A. Gauger John Luksich
Thomas A. Edison Technical Center Dielectric Fluids Group
Cooper Power Systems Cooper Power Systems
Franksville, WI 53126 USA Waukesha, WI 53188 USA
Abstract: Aging of transformer insulation paper in natural ester
(vegetable oil) dielectric fluid is compared to that in conventional
transformer oil. Sealed steel aging vessels containing copper,
aluminum, thermally upgraded paper, and dielectric fluid (mineral
oil and natural ester) were aged at 130, 150, and 170C for 500,
1000, 2000, and 4000 hours. Paper degradation after aging is
determined using paper tensile strength and degree of
polymerization measurements. The paper in natural ester aged at
slower rates than did paper in conventional transformer oil, taking 5-
8 times longer to reach end-of-life. Results obtained for mineral oil
samples are similar to predictions based on IEEE transformer
loading guide calculations. Mechanisms for the slower aging rate are
proposed.
Keywords: aging, dielectric liquids, dielectric materials, insulation,
insulation life, life estimation, oil insulation, paper insulation, power
transformers, transformers.
I. INTRODUCTION
Previously published work demonstrates the practicality of
substituting natural ester dielectric fluid for mineral oil in
liquid-paper insulation systems [1,2,3]. Previous experiments
indicated a difference in aging rates between paper in natural
ester and paper in mineral oil. This study was undertaken to
better understand and quantify the insulation thermal life of
thermally upgraded paper in natural ester.
The most significant characteristic used to assess
transformer aging is mechanical strength of the paper
insulation [4-7,10]. To determine aging rates in a practical
amount of time, accelerated aging is used. According to the
IEEE loading guide, accelerated aging in sealed systems best
simulates real aging in modern sealed transformers [13].
Several sealed system accelerated aging studies have been
published [5,9-12].
Paper aging depends primarily on temperature and water
content. One of the first studies of insulation aging reported
an exponential thermal aging rate that roughly doubled for
each 8C rise [4]. Later work applied chemical reaction rate
theory to paper aging and refined the rate estimates [6,7].
Water content studies show the aging rate to be
approximately proportional to water content [5,8].
In earlier studies of aged paper a common indicator of
mechanical strength is tensile strength. For several decades,
the degree of polymerization (D
v
P) has also been used as a
means of indicating paper aging, and is one of the bases in
the IEEE loading guide references to aging [8,13].
II. EXPERIMENTAL
The aging systems contained materials in the same
proportions as in a 225 kVA pad-mounted transformer. Each
aging test vessel contained thermally upgraded 0.255mm
Kraft paper
1
(26g), dielectric fluid
2
(350ml), aluminum metal
strip (106cm
2
), copper metal strip (76cm
2
), and the mild steel
inner surface of the aging vessel (323cm
2
). Headspace was
set at 17% of total internal volume at room temperature. An
aging vessel and its test materials are depicted in Fig. 1.
The moisture in the paper was adjusted to approximately
0.5wt%, representing nominal moisture content of paper in
new transformers. The open systems were conditioned for a
minimum of three days at 22C and 50% relative humidity,
then dried at 105C for 2.5 hours. Degassed and dried
Fig. 1. Components of a sealed test system.

1
Whiteley Ltd. Grade K T/U 0.255mm thermally upgraded Kraft paper
2
Ergon Hyvolt II inhibited insulating mineral oil, Cooper Power Systems
Envirotemp
FR3
natural ester dielectric fluid

0-7803-7287-5/01/$17.00 2001 IEEE
Table 1. Selected times and temperatures of sealed aging test using thermally
upgraded paper.
Temperature
Time (hrs) ambient 130C 150C 170C
0 x
500 x
1000 x x
2000 x x x
4000 x x x
dielectric fluid was added at atmospheric pressure, followed
by 30 minutes at 500mHg. The vessels were sealed,
degassed, and pressurized to eight atmospheres with dry
nitrogen to test for leaks. The pressure was reduced to two
atmospheres at room temperature prior to the start of the test.
This maintained the vessel pressure above the vapor pressure
of water at 170C, replicating the expected pressure ratio
limit in actual operating transformers. Base line values were
measured at this point.
The aging systems were placed in lab ovens upside-down
so that the welded end of the vessel was over the head space.
The systems were aged at 130, 150, and 170C for 500 to
4000 hours (Table 1).
III. RESULTS
Tensile strength and degree of polymerization results for
the paper aged in mineral oil (Fig. 2) are in good agreement
with previously published work [8,9,11,12]. Fig. 3 shows
paper after 4000 hours at 150 and 170C. Visually, paper in
natural ester showed noticeably less degradation than paper in
mineral oil. Routine sample handling found the papers aged
in mineral oil to be more fragile than those in natural
ester.
T
e
n
s
i
l
e

S
t
r
e
n
g
t
h

(
%

r
e
t
a
i
n
e
d
)
100
75
50
25
0
170
o
C
150
o
C
160
o
C [12]
130
o
C
Aging Time (hours)
0 1000 2000 3000 4000
D
e
g
r
e
e

o
f

P
o
l
y
m
e
r
i
z
a
t
i
o
n
0
200
400
600
800
1000
1200
170
o
C 160
o
C [12]
150
o
C
130
o
C
Fig. 2. Tensile strength and degree of polymerization of thermally
upgraded paper in mineral oil (sealed vessels at 130, 150, 170C). Data
at 160C [12] included for comparison.
natural ester @ 150C mineral oil @ 150C
natural ester @ 170C mineral oil @ 170C
Fig. 3. Thermally upgraded Kraft paper after 4000 hours.
At higher temperatures and longer times, the vessels
containing mineral oil had a buildup of sludge and
carbonaceous deposits on the walls. The vessels containing
natural ester remained clean.
A. Tensile Strength
Tensile strength results at 150 and 170C are shown in Fig.
4. Paper aged in both mineral oil and natural ester exhibits no
appreciable change in tensile strength at 130C. At 150C,
the tensile strength of both papers drops slightly at 1000 and
2000 hours. At 4000 hours, the tensile strength of paper in
mineral oil decreases to about 25% of initial strength. The
tensile strength of paper in natural ester exhibits no
appreciable change between 2000 and 4000 hours. At 170C,
the tensile strength of paper in mineral oil falls to about 25%
retained strength at 500 hours, and loses its remaining
strength at 1000 hours. The tensile strength of paper in
natural ester does not drop to 25% retained strength until
approximately 4000 hours.
B. Degree of Polymerization
The D
v
P data are shown in Fig. 5. In all cases, the D
v
P of
paper aged in mineral oil is lower than that in the natural
ester. The paper aged at 130C in both fluids did not reach
200 D
v
P. The paper aged at 150C in mineral oil decreased to
200 D
v
P after 4000 hours. The paper aged in the natural ester
decreased to 450 D
v
P in the same interval. The paper aged at
170C in mineral oil reached 200 D
v
P after 500 hours, while
the paper aged in the natural ester decreased to 450 D
v
P
during the same interval.
0-7803-7287-5/01/$17.00 2001 IEEE
Aging Time (hours)
0 1000 2000 3000 4000
0
5000
10000
15000
20000
100
75
50
25
0
T
e
n
s
i
l
e

S
t
r
e
n
g
t
h

(
l
b
/
i
n
2
)
0
5000
10000
15000
20000
R
e
t
a
i
n
e
d

T
e
n
s
i
l
e

S
t
r
e
n
g
t
h

(
%

o
f

u
n
a
g
e
d
)
100
75
50
25
0
150
o
C
170
o
C
mineral oil
natural ester
(error bars = 1)
Fig. 4. Tensile strength of thermally upgraded paper aged in mineral
oil and natural ester dielectric fluids (sealed vessels at 150, 170C).
C. Water Content
Water contents of papers (wt%) and fluids (absolute
content and percent saturation) at room temperature are
shown in Table 2. At 130C, the papers in both fluids lose
water as aging progresses, while the water content of both
dielectric fluids is nearly constant.
At 150C, water content of paper aged in mineral oil
increases between 2000 and 4000 hours. The water content of
mineral oil also increased. The water content of paper aged in
the natural ester decreases and remains constant. The water
content of the natural ester remained practically unchanged.
At 170C, water content of paper aged in mineral oil
increases by an order of magnitude, peaking at 1000 hours.
The mineral oil becomes fully saturated by 500 hours and
remains saturated. The water content of paper aged in the
natural ester decreases and remains constant. The natural
ester increases in water content to about 25% saturation at
4000 hours.
IV. DISCUSSION
The results obtained for paper in mineral oil correlate well
with previously published work examining sealed system
accelerated aging of thermally upgraded paper.
Both tensile strength and degree of polymerization are used
to define end of life of paper insulation [12]. The current
IEEE loading guide [13] suggests several end-of-life
definitions, including 50% retained tensile strength, 25%
retained tensile strength, and 200 D
v
P. Table 3 compares
times to reach these end-of-life criteria for paper in mineral
oil to paper in the natural ester.
The data in Table 3 shows that paper aged at 170C in
natural ester takes 5 to 8 times longer to reach standard end-
of-life criteria than paper aged in mineral oil. Paper aged at
150C in mineral oil reached end of life points. However,
paper aged at 150C in natural ester does not reach any end-
Aging Time (hours)
0 1000 2000 3000 4000
0
200
400
600
800
1000
1200
100
75
50
25
0
D
e
g
r
e
e

o
f

P
o
l
y
m
e
r
i
z
a
t
i
o
n
0
200
400
600
800
1000
1200
R
e
t
a
i
n
e
d

D
v
P

(
%

o
f

u
n
a
g
e
d
)
100
75
50
25
0
0
200
400
600
800
1000
1200
100
75
50
25
0
130
o
C
150
o
C
170
o
C
mineral oil
natural ester
Fig. 5. Degree of polymerization of thermally upgraded paper aged
in mineral oil and natural ester (sealed vessels at 130, 150, 170C).
of-life points, so a numeric differential cannot be calculated.
V. MECHANISMS
It is apparent from the results that the natural ester fluid
significantly increased the thermal stability of the paper.
Table. 2. Water content of thermally upgraded paper aged in mineral oil and
natural ester. Water content of dielectric fluid as absolute content and percent
saturation.
130C 150C 170C
mineral natural mineral natural mineral natural
Time (hrs) oil ester oil ester oil ester
Water Content of Paper [wt%]
0
1
0.46 0.51 0.46 0.51 0.46 0.51
500 - - - - 1.46 0.15
1000 - - 0.30 0.09 7.03 0.28
2000 0.03 0.10 0.23 0.08 3.28 0.26
4000 0.16 0.04 1.26 0.09 1.85 0.17
Water Content of Fluid [% saturation @ 20C; (mg/kg)]
0
1
5 (3) 3 (27) 5 (3) 3 (27) 5 (3) 3 (27)
500 - - - - 167 (100) 2 (22)
1000 - - 25 (15) 5 (54) 114 (69) 4 (39)
2000 11 (7) 1 (7) 10 (6) 1 (11) 105 (63) 10 (109)
4000 2 (1) 2 (20) 50 (30) 3 (28) 145 (87) 25 (258)
1
baseline values
0-7803-7287-5/01/$17.00 2001 IEEE
Table 3. Hours to reach insulation end-of-life of thermally upgraded Kraft
paper aged in mineral oil and natural ester fluid. Time to insulation end-of-
life calculated using IEEE C57.91 included for comparison.
150C 170C
mineral natural IEEE mineral natural IEEE
End-of-Life Basis oil ester Basis oil ester Basis
Retained Tensile Strength
50% 3100 * 1602 240 1300 323
25% 4000 * 3327 490 4000 671
Degree of Polymerization
200 3200 * 3697 480 3400 746
*Paper did not reach end-of-life over the duration of the test. To be
conservative, extrapolation was not employed.
Paper is essentially cellulose, a linear polysaccharide
consisting of many anhydro -D-glucopyranose units linked
to each other via a (14) glycosidic bond (Fig. 6) to form a
polymeric chain. Many long oriented chains, arranged
parallel to one another, make up the fibers that form the
structure of paper.
The strength of paper is a result of the oriented linear
chains and hydrogen bonding between the chains, which adds
to rigidity. Degree of polymerization and tensile strength
provide a relative indication of the number of intact
molecular chains and the degree of hydrogen bonding
between the chains. Reduction in either of these properties
indicates that scission of these chains has taken place.
The principal reactive sites of cellulose are the secondary
hydroxyl (OH) groups located at positions C-2, C-3, and the
primary OH at C-6. These OH groups are involved in
hydrogen bonding, which contributes to further reactivity at
elevated temperatures [15]. Water migrates to these sites and
disrupts the hydrogen bonds between the cellulose molecules,
destabilizing the paper matrix. Paper degradation in a sealed
accelerated thermal aging study is primarily a combination of
hydrolytic breakdown and thermal scission of the (14)
glycosidic bond [12,16]. Oxidation and other secondary
reactions at the C-2, C-3, and C-6 hydroxyls are also a factor.
The principal reaction products of thermal degradation are
H
2
O, CO, CO
2
, and, to a lesser extent, furanic compounds.
The latter result from chain scission followed by ring
dissociation. The generation of water contributes to
hydrolytic breakdown.
We propose that the improvement in the thermal aging
characteristics of cellulose aged in the natural ester is the
result of at least two modes of protection provided by the
natural ester: water scavenging, and steric hindrance of the
cellulose hydrolytic reaction.
O
H
H
OH H
OH
H2
C
H
H
O H
H
H2
C
H
OH H
HO
H
O
OH
O
OH
1 4
2
3
4
5
6
n
Fig. 6. Anhydro -D-glucopyranose monomer units.
A. Water Scavenging
Water scavenging functions in two different ways.
1) At elevated temperatures the natural ester can undergo
hydrolysis, consuming available water from the cellulose,
thereby reducing the potential damage from water.
2) The natural ester can hold considerably more water than
can mineral oil. The room temperature water saturation for
the natural ester is about 1050mg/kg; mineral oil is about
60mg/kg. Because of the difference in saturation levels, the
natural ester can absorb more water than mineral oil, and so
has a greater paper drying effect. Research shows that
removing water from the paper as it is generated improves
paper aging characteristics [12,14,17].
B. Hydrolytic Protection
The natural ester can interfere with the process of cellulose
hydrolysis due to thermal stress. This is due to reaction of the
natural ester with OH on the glucose ring and at chain
terminations. The proposed mechanism is transesterification
driven by the temperature conditions of accelerated aging
[17]. The esterification of the reactive OH sites on cellulose
with bulky ester groups stabilizes the molecule and extends
thermal life [18].
VI. CONCLUSIONS AND RECOMMENDATIONS
We find that thermally upgraded Kraft paper ages
considerably slower in the natural ester dielectric fluid than in
mineral oil under the same thermal stress.
Paper aged in the natural ester takes 5-8 times longer to
reach the same end-of-life points as paper aged in mineral oil
in sealed vessels at 170 C. Similar quantitative comparisons
at the lower temperatures are not available, as the paper aged
in the natural ester did not reach any standard end-of-life
benchmarks within the 4000 hour test duration.
Preliminary calculations based on these data estimate that
at 110C, the time to end-of-life should be at least 2 times
that of paper aged in mineral oil. We recommend that these
results, along with additional data as it becomes available, be
used to establish thermal aging equations and normal
insulation life for thermally upgraded paper in natural ester
dielectric fluids. This may result in transformer temperature
rise ratings or overload levels higher than those
recommended in standard loading guides for thermally
upgraded paper in mineral oil insulation systems.
VI. ACKNOWLEDGEMENTS
The authors thank the many professionals at the Thomas A.
Edison Technical Center for their assistance. The authors also
thank EHV-Weidmann Industries Inc. for providing the paper
samples used in this study. Dr. Vera Kolb, Professor of
Chemistry at Univ. of Wisconsin - Parkside, is acknowledged
for her input on cellulose/ester chemistry. Finally, we thank
our colleagues for their review and suggestions.
VIII. REFERENCES
[1] C.P. McShane, G.A. Gauger, J. Luksich, Fire Resistant Natural Ester
Dielectric Fluid and Novel Insulation System for Its Use, IEEE/PES
Transmission & Distribution Conference, Apr. 12-16, 1999.
0-7803-7287-5/01/$17.00 2001 IEEE
[2] C.P. McShane, New Safety Dielectric Coolants for Distribution and
Power Transformers, IEEE Industry Applications Magazine, vol. 6, no. 3,
pp. 24-32, May/June 2000.
[3] T.V. Oommen, C.C. Claiborne, Biodegradable Insulating Fluid from
High Oleic Vegetable Oils, CIGRE 15-302, 1998.
[4] V.M. Montsinger, Loading Transformers by Temperature, AIEE
Transactions, vol. 49, pp. 776-792, 1930.
[5] F.M. Clark, Factors Affecting the Mechanical Deterioration of Cellulose
Insulation, AIEE Transactions, vol. 61, pp. 742-749, 1942.
[6] T.W. Dakin, Electrical Insulation Deterioration Treated as a Chemical
Rate Phenomenon, AIEE Transactions, vol. 67, pp. 113-122, 1948.
[7] W.A. Sumner, G.M. Stein, A.M. Lockie, Life Expectancy of Oil-
Immersed Insulation Structures, AIEE Summer Meeting, Paper No. 53-243,
Atlantic City, June 15-19, 1953.
[8] J. Fabre, A. Pichon, Deteriorating Processes and Products of Paper in
Oil. Application to Transformers, CIGRE Conference on Large High
Voltage Electric Systems, Paper 137, Paris, 1960.
[9] W.G. Lawson, M.A. Simmons, P.S. Gale, Thermal Aging of Cellulose
Paper Insulation, IEEE Transactions on Electrical Insulation, vol. EI-12, no.
1, pp. 61-66, Feb. 1977.
[10] W.A. Fessler, F.S. Nichols, T.O. Rouse, Tube Aging and Oxidation of
Naphthenic and Paraffinic Transformer Oils, IEEE CH1287-2/78/0000-
0265, 1978.
[11] H. Yoshida, Y. Ishioka, T. Suzuki, T. Yanari, T. Teranishi,
Degradation of Insulating Materials of Transformers, IEEE Transactions
on Electrical Insulation, vol. EI-22, no. 6, 1987.
[12] W.J. McNutt, Insulation Thermal Life Considerations for Transformer
Loading Guides, IEEE Transactions on Power Delivery, vol. 7, no. 1, pp.
392-401, Jan. 1992.
[13] IEEE Guide for Loading Mineral-Oil-Immersed Transformers, IEEE
Std. C57.91-1995 (published 1996).
[14] R.D. Kremer, D. Tabb, Paper: The Beneficially Interactive Support
Medium for Diagnostic Test Development, American Laboratory, pp. 136-
143, Feb. 1990.
[15] I. Rusznak, K. Zimmer, Initial Processes in the Thermolysis of
Cellulose Fibres, Proc. Hungarian Textile Conf. 18
th
, vol. 2, pp. 119-139,
1971.
[16] J. Xie, Y.L. Hsieh, Enzyme-Catalysed Transesterification on
Cellulose, Conf. Proc. ACS Polymeric Matls., vol. 82, pp. 406-407, 2000.
[17] M.R. Huang, X.G. Li, Thermal Degradation of Cellulose and Cellulose
Esters, J. of Applied Polymer Science, vol. 68, pp. 293-304, 1998.
[18] R.K. Jain, K. Lal, H.L. Bhatnagar, A Kinetic Study of the Thermal
Degradation of Cellulose and Its Derivatives, Makromol. Chem. 183, pp.
3003-3017, 1982.
VII. BIOGRAPHIES
C. Patrick McShane received his BS in
Electrical Engineering from Marquette
University in 1970, and an MS in Engineering
Management from the Milwaukee School of
Engineering in 1998. Currently the Product
Line Manager for Dielectric Fluids at Cooper
Power Systems Transformer Products, his
employment experience includes International
Area Manager for RTE Corp. and Regional
Technical Director for the State of Sao Paulo
(Brazil) Rural Electrification Program. His
professional activities include USA Delegate
IEC TC99, IEC TC99 Liaison to TC64, IEC
TC89 Expert Delegate, Chair ASTM W.G. D-
5222, and W.G. Chair IEEE TC Dielectric Fluids Subcommittee C57.121.
Several of his proposals have been adopted by USA Codes and Standards
(NEC, NESC, FMRC). He has presented papers at domestic and international
engineering conferences including IEEE, EPRI, Doble, and CIRED. He is the
principal inventor of two US patents relating to dielectric fluids.
Kevin J. Rapp is presently completing
requirements for his BS in Chemistry at the
University of Wisconsin-Parkside. Since
joining the Thomas A. Edison Technical
Center in 1976, he has been involved with
dielectric materials development for electrical
power equipment. His present activities include
natural seed based dielectric fluids research and
applications.
Jerry L. Corkran received his BS in Electrical
Engineering in 1964 and MSEE in 1965 from
the University of Missouri. He is a member of
the IEEE Transformer Committee. From 1965
through 1971 he was a Development Engineer
for Allis Chalmers. From 1971-79 he was a
Design Engineer for the RTE-ASEA in
Waukesha, WI. He joined Cooper Power
Systems (RTE) in 1979 where he is currently a
Staff Engineer.
Gary A. Gauger received his BS in Chemistry
from the University of Wisconsin-Oshkosh in
1963, and studied organic chemistry in the
University of Wisconsin-Milwaukee Graduate
Program during 1970-1972. Professional
affiliations include the American Chemical
Society Division of Petroleum Chemistry and
Division of Environmental Chemistry,
American Association for the Advancement of
Science, and the ASTM D-27 Dielectric
Liquids and Gases Committee. Gary joined
Cooper Power (then McGraw-Edison) in 1970
as an Analytical Scientist. He has been
involved in dielectric materials development
since 1975, and is currently Staff Scientist at the Thomas A. Edison
Technical Center.
John Luksich received his BS in Chemistry in
1980 and MS in Materials Engineering in 1990,
both from the University of Wisconsin. His
engineering career includes thin film materials
development at the McDonnell Douglas Space
& Physics Laboratory and sensor development
for Johnson Controls. He is currently an
engineer with the Cooper Power Systems
Transformer Products Dielectric Fluids group,
and a member of the ASTM D-27 Dielectric
Liquids and Gases Committee.

Proceedings of 14
th
International Conference on Dielectric Liquids (ICDL), Graz (Austria), July 7-12, 2002

Aging of Kraft Paper in Natural Ester Dielectric Fluid

C.P. McShane K.J. Rapp J.L. Corkran
Dielectric Fluids Thomas A. Edison Technical Center Transformer Products

G.A. Gauger J. Luksich
Thomas A. Edison Technical Center Dielectric Fluids
Cooper Power Systems Cooper Power Systems
Franksville, WI 53126 USA Waukesha, WI 53188 USA

Abstract: Kraft transformer insulation paper aged in natural ester
(vegetable oil) dielectric fluid was compared to identical paper aged
in conventional transformer mineral oil. Sealed steel aging tubes
containing copper, aluminum, Kraft paper, and dielectric fluid
(mineral oil and natural ester) were aged at 150C for 500, 1000,
2000, and 4000 hours. The extent of paper degradation after aging
was determined using paper tensile strength, paper degree of
polymerization, and furanic compounds in the aged fluid. Water
contents of fluids and paper were compared. Paper aged in
conventional transformer oil degraded at a significantly faster rate
than in natural ester dielectric fluid. Paper in mineral oil reached
three criteria for IEEE end-of-life (50% retained tensile strength, 25%
retained tensile strength, and degree of polymerization of 200) within
the first 1000 hours. After 4000 hours of aging, paper in natural ester
did not degrade to any of the IEEE end-of-life criteria. At 4,000
hours, the paper aged in natural ester retained about 55% of the
original tensile strength and a degree of polymerization of about 280.
Paper aged in conventional transformer oil degraded to the same
values in about 315 and 390 hours, respectively - an order of
magnitude faster. The reduced paper-aging rate in natural ester is
primarily attributed to the fluid maintaining the paper in a very dry
state.

INTRODUCTION

Natural esters (vegetable oils) formulated as dielectric fluids
have environmental and fire safety advantages over
conventional transformer mineral oil [1]. In thermal evaluation
comparisons of transformer insulation systems, production
distribution transformers filled with natural ester dielectric
fluid exhibited less deterioration than did the corresponding
transformers using mineral oil [2].

The environmental properties of natural esters are such that in
Germany they are classified as non-hazardous to waters [3].
Aquatic biodegradation tests [4] of the natural ester dielectric
fluid used in this experiment found >99% metabolized
conversion to CO
2
, equivalent to compounds defined as
ultimately biodegradable. In acute trout toxicity tests [5], the
same fluid had no observable effect on fish at 1000mg/l, the
highest concentration tested.

Transformers using natural ester fluids deliver very important
improvements in fire safety compared to those using mineral
oil. Natural ester dielectric fluids have fire points in the range
of 350-360C; conventional mineral oil has a fire point of
about 155C. Two natural ester dielectric fluids are recognized
as less-flammable per Section 450.23 of the U.S. National
Electrical Code
[6].

Accelerated aging tests of distribution transformers gave early
indications that the rate of paper aging is fluid-dependent [2].
An earlier study quantified this dependence for thermally
upgraded Kraft insulation paper [7]. This work examines the
aging rate dependence on fluid type for plain (not thermally
upgraded) Kraft insulation paper.

EXPERIMENTAL

The thermal aging procedure and sample preparation methods
are identical to those previously described [7]. Sealed steel
aging tubes contained 28.4g of Kraft insulation paper
1
dried to
a water content of 0.76wt%, 350ml of dielectric fluid
2
, and
typical transformer proportions of copper and aluminum. The
tubes were aged at 150C and evaluated after 500, 1000, 2000,
and 4000 hours.

The extent of paper aging was determined using changes in the
tensile strength (TS) and the degree of polymerization (D
v
P).
The total concentrations of four predominant furanic products
of paper degradation were measured in the aged fluids. The
water contents of both the papers and fluids were also
determined.

1
Whiteley Ltd. Grade K 0.255mm presspaper
2
Cooper Power Systems Tranelec
inhibited insulating mineral oil

and Envirotemp
FR3

0-7803-7350-2/02/$17.00 2002 IEEE 173
Aging Time (hours)
0 1000 2000 3000 4000
R
e
t
a
i
n
e
d

T
e
n
s
i
l
e

S
t
r
e
n
g
t
h

(
%

o
f

u
n
a
g
e
d
)
0
20
40
60
80
100
120
mineral oil
Darveniza et al.
natural ester
Hill et al.
Figure 2. Retained tensile strength versus time for Kraft paper aged
in natural ester and mineral oil. New paper tensile strength is about
120MPa. Vertical error bars are 1 standard deviation; horizontal
error bars represent time-at-temperature uncertainty. Published
retained tensile strength data [8,9] are included for comparison.
natural
ester
mineral
oil
500
1000
2000
4000
A
g
i
n
g

T
i
m
e

(
h
o
u
r
s
)

Figure 1. Kraft paper insulation after 150C sealed tube aging in
natural ester and mineral oil.
RESULTS

The results of aging at 150C, summarized in Table 1, show
that paper insulation ages much slower in vegetable oil than in
conventional transformer oil.

The difference in appearance between the papers aged in the
two fluids is striking (Figure 1). At 1000 hours of aging, the
paper in mineral oil was dark brown and brittle; the paper in
the natural ester fluid remained flexible and was only slightly
discolored. The 2000 and 4000 hour mineral oil papers were
fragile and difficult to handle. This was not the case with the
natural ester-aged papers. The color and fragility of the papers
correlate well with changes in their TS and D
v
P, and the
proportional furanic difference between the two fluids.

Tensile Strength
Table 1. Results of Kraft paper sealed tube aging at 150C in
mineral oil and natural ester. Water content of dielectric fluid is
given both as absolute content and percent saturation. Total furanic
compounds are given as mg furans per liter of fluid per initial kg of
paper.
Time (hrs) 0
1
500 1000 2000 4000
in mineral oil 0.76 0.67 2.44 3.39 3.36
in natural ester 0.77 0.27 0.21 0.04 0.05
mineral oil 2 (1) 30 (18) 58 (35) 61 (36) 75 (45)
natural ester 2 (24) 9 (89) 11 (116) 2 (21) 2 (25)
Tensile Strength of Paper [MPa]
in mineral oil 119 53.2 21.9 17.8 6.9
in natural ester 121 80.8 74.9 63.6 69.8
Degree of Polymerization of Paper
in mineral oil 1119 217 110 75 12
in natural ester 1225 380 320 306 274
Furanic Compound Content [mg/l/kg]
mineral oil nd 1055 1462 670 602
natural ester nd 191 234 151 62
1
baseline values

The decrease in TS over time (Figure 2) shows a steep initial
slope of paper degradation in both natural ester and mineral
oil. The highest relative rate of paper degradation takes place
in the first 500 hours of aging in both fluids. The relative rates
in both fluids decrease substantially after 1000 hours.

In terms of absolute degradation, the paper in mineral oil,
already below 50% retained TS at 500 hours, was below 20%
retained TS value at 1000 hours. The decline in TS of the
paper aged in mineral oil is comparable to that seen in other
studies [8,9]. The paper aged in vegetable oil degraded to
about 70% of its original TS at 500 hours. After 4000 hours,
the retained TS was about 58%.


Another measure of cellulose degradation is degree of
polymerization (D
v
P). The change in D
v
P over time is shown
in Figure 3, and is similar to the change seen in TS. A rapid
D
v
P decrease in the first 500 hours of aging was observed.
Paper aged in mineral oil degraded to about 20% of the
174
Aging Time (hours)
0 1000 2000 3000 4000
R
e
t
a
i
n
e
d

D
e
g
r
e
e

o
f

P
o
l
y
m
e
r
i
z
a
t
i
o
n
(
%

o
f

u
n
a
g
e
d
)
0
20
40
60
80
100
mineral oil
Shroff & Stannett
Miyoshi
natural ester
Figure 3. Retained degree of polymerization (D
v
P) versus time for
Kraft paper aged in natural ester and mineral oil. New paper D
v
P is
1175 50. Horizontal error bars represent time-at-temperature
uncertainty. Published retained D
v
P data [10,11] are included for
comparison (data attributed to Schroff & Stannett from [10]).
Aging Time (hours)
0 1000 2000 3000 4000
F
u
r
a
n
i
c

C
o
n
t
e
n
t

(
m
g
/
l
/
k
g
)
10
100
1000
10000
mineral oil
natural ester
Furanic Compounds:
2-furaldehyde
5-hydroxymethyl-2-furfural
5-methyl-2-furfural
2-acetyl furan
Figure 4. Furanic compound content versus time for natural ester and
mineral oil after sealed tube aging of Kraft paper. Horizontal error
bars represent time-at-temperature uncertainty.
original D
v
P at 500 hours, in line with other published results
[10,11]. The retained D
v
P of paper aged in natural ester fluid
remained above 20% after 4000 hours of aging.

Furanic Compounds

Furanic compounds in oil are used to observe the insulation
degradation of in-service transformers [12]. Figure 4 shows
the concentrations of furanic compounds generated by paper
degradation over time. The furan evolution from paper aged in
the natural ester fluid was up to an order of magnitude lower
than paper aged in conventional transformer oil.

The concentration of furans reached a maximum at 1000 hours
in both fluids, supporting both the TS and D
v
P findings. After
1000 hours, the concentration of furans in both fluids
decreases.

Water Content

The water content of the paper in mineral oil increased
significantly. The water content of the paper in natural ester
decreased. By the end of the test, water content of paper aged
in mineral oil was orders of magnitude greater than of paper
aged in natural ester.

In mineral oil, the water content of the paper at 500 hours
decreased slightly from the initial 0.75wt%. At 1000 hours, the
water content increased to 2.4wt%, and continued to increase
to about 3.3wt% for the remainder of the test.

The water content of paper in natural ester decreased to 0.27%
after 500 hours and continued to decrease. At 2000 and 4000
hours, the paper contained about 0.05wt% water.

Similar trends were seen in the water content of the fluids
(Table 1). Viewing the water content as percent saturation at
room temperature, both fluids started with water contents of
2% saturation. The water in mineral oil increased steadily,
reaching 75% saturation at 4000 hours. The water in the
natural ester rose gradually to 11% saturation at 1000 hours,
then decreased to 2% saturation at 2000 and 4000 hours.

Aging Time (hours)
0 1000 2000 3000 4000
W
a
t
e
r

C
o
n
t
e
n
t

o
f

P
a
p
e
r

(
w
t
%
)
0.0
0.1
0.2
0.3
1.0
2.0
3.0
4.0
mineral oil
natural ester
Figure 5. Water content versus time for Kraft paper aged in natural
ester and mineral oil. Paper was originally dried to a water content
of about 0.76wt%. Vertical error bars are 1 standard deviation;
horizontal error bars represent time-at-temperature uncertainty.
175
DISCUSSION

The TS, D
v
P, and furanic content results establish that paper
ages slower in natural ester dielectric fluid than in
conventional transformer oil. The paper aged in mineral oil
degraded beyond three recognized IEEE end-of-life criteria
[13]: 50% TS at 370 hours, 25% TS at 915 hours, and
D
v
P=200 at 555 hours. At 4000 hours, the paper aged in the
natural ester fluid did not reach any IEEE end-of-life
conditions.

In both systems, the concentrations of furanic compounds rose
to a maximum, then decreased, maintaining a factor of 4-10
times more furans in mineral oil. The furans themselves
decompose over time. Although the furanic contents of both
fluids decrease in a similar way, the rationale for the reduction
is different for each fluid. In the case of paper in mineral oil,
more of the cellulose was destroyed and converted to furanic
products. In the natural ester fluid, less cellulose aging took
place to produce furanic compounds.

The water in paper data suggest an explanation for the contrast
in paper degradation seen between fluids [12]. The
acceleration of aging due to water has been known for at least
40 years [14]. As paper thermally degrades it produces water,
promoting further degradation through hydrolysis. In a
conventional transformer paper-in-oil system, the degradation
is autocatalytic [15]. The partitioning of the water between the
paper and fluid depends on the characteristic polarity (affinity
for water) of each.

An essentially non-polar fluid, such as mineral oil, prevents
most of the water generated during cellulose degradation from
leaving the paper. Although the water content of the mineral
oil steadily increased to 75% saturation at 4000 hours, this is
only 45mg/kg in terms of absolute water content. The paper
also was at approximately 75% water saturation.

The opposite effect is seen for paper in the more polar natural
ester. The natural ester has a higher affinity for water than
mineral oil. The water, attracted to the ester fluid, is liberated
from the paper. Removing water from the paper limits the
paper degradation rate.

The water generated during cellulose degradation migrates out
of the paper and into the natural ester. One would expect the
water in natural ester to increase proportionally. However, the
natural ester remains dry at less than <2% saturation.

Since the aging tube is sealed, water cannot escape. The water
is consumed in a hydrolysis reaction with the natural ester.
The reaction produces long-chain fatty acids. These fatty acids
react with the cellulose in a transesterification reaction [16].
This reaction results in steric protection of the cellulose sites
most susceptible to attack, a consequence of which is reduced
degradation by water [7].

At room temperature, natural esters can hold about 18 times
more water in solution than can conventional transformer oil.
The electrical strength of a dielectric fluid depends on the
percent saturation of water, not the absolute water content.
The electrical strength decreases above 50% saturation, about
30 mg/kg in mineral oil and 550 mg/kg in natural esters.

The temperature at which bubbles evolve from an insulation
system strongly depends on the water content of the paper
[17]. For paper having 3.3% water content (in mineral oil at
2000 and 4000 hours), the equivalent bubble evolution
temperature would be about 125C. For paper having 0.05%
water content (in natural ester at 2000 and 4000 hours), the
bubble evolution temperature would be >200C [17].

CONCLUSIONS

These results show that Kraft insulation paper takes at least an
order of magnitude longer to reach IEEE end-of-life
conditions in natural ester than in conventional transformer
oil. Applying natural ester dielectric fluids in transformers
should result in improved insulation performance.

REFERENCES

[1] McShane, C.P., Gauger, G.A., Luksich, J. "Use of Natural
Vegetable Oil Esters as Dielectric Coolant", CIRED 15
th

International Conference on Electricity Distribution, June 1-4,
1999, Nice, France
[2] McShane, C.P., Gauger, G.A., Luksich, J. "Fire Resistant
Natural Ester Dielectric Fluid and Novel Insulation System for
Its Use", IEEE/PES Transmission & Distribution Conference,
April 12-16, 1999, New Orleans, USA
[3] Administrative Regulation on the Classification of Substances
Hazardous to Waters, Water Hazard Classes (VwVwS), Annex
1 Substances Non-Hazardous to Waters Pursuant to Number
1.2a, Index No. 660, 1999, Germany
[4] Aerobic Aquatic Degradation, US EPA OPPTS 835.3100, Test
Report No. ML-304.98, Thomas A. Edison Technical Center,
Cooper Power Systems, Franksville, USA, 1998
[5] Final Report: Acute Trout Toxicity Testing for Two
Envirotemp FR3 Formulations; Procedure 203, OECD
Guidelines for Toxicity of Chemicals, 1993a, and Biological
Test Method: Acute Lethality Test Using Rainbow Trout,
Environment Canada, 1990 (1996 Amendments); GlobalTox
International Consultants, Guelph, Canada, Dec. 1, 1999
176
[6] National Electrical Code, NFPA 70, 2002 Edition, National
Fire Protection Association, Quincy, MA, USA
[7] McShane, C.P., Rapp, K.J., Corkran, J.L., Gauger, G.A.,
Luksich, J. Aging of Paper Insulation in Natural Ester
Dielectric Fluid, 2001 IEEE/PES Transmission & Distribution
Conference, Oct. 28 Nov. 2, 2001, Atlanta, USA
[8] Darveniza, M., Saha, T.K., Hill, D.J.T., Le, T.T. Investigations
into Effective Methods for Assessing the Condition of
Insulation in Aged Power Transformers, IEEE Trans PD, Vol.
13, No. 4, pp. 1214-1223, Oct. 1998
[9] Hill, D.J.T., Le, T.T., Darveniza, M., Saha, T. A Study of
Degradation of Cellulosic Insulation Materials in a Power
Transformer. Part 2: Tensile Strength of Cellulose Insulation
Paper, Polymer Degradation and Stability, Vol. 49, pp. 429-
435, 1995
[10] Emsley, A.M., Stevens, G.C., A Reassessment of the Low
Temperature Thermal Degradation of Cellulose, IEEE 6
th

International Conference on Dielectric Materials,
Measurements, and Applications, pp.229-32, Sept. 7-10, 1992,
Manchester, UK
[11] Miyoshi, A., A New Additive for Improving the Thermal
Aeging Characteristics of Insulating Paper, IEEE Trans EI,
Vol. E1-10 (1), 13, 1975

[12] Unsworth, J., Mitchell, F. Degradation of Electrical Insulating
Paper Monitored with High Performance Liquid
Chromatography, IEEE Trans EI, Vol. 25, No. 4, pp. 737-746,
Aug 1999
[13] IEEE Guide for Loading Mineral-Oil-Immersed
Transformers, IEEE Std C57.91-1995, April 25, 1996, IEEE,
New York, USA
[14] Clark, F.M. Insulating Materials for Design and Engineering
Practice, pp. 287-290, Wiley, USA, 1962
[15] Hill, D.J.T., Le, T.T., Darveniza, M., Saha, T. A Study of the
Transformer. Part 3: Degradation Products of Cellulose
Insulation Paper, Polymer Degradation and Stability, Vol. 51,
pp. 211-218, 1996
[16] Huang, M.R., Li, X.G. Thermal Degradation of Cellulose and
Cellulose Esters, J. Appl. Polymer Sci., Vol. 68, pp. 293-304,
1998.
[17] Oommen, T.V., Lindgren, S.R. Bubble Evolution from
Transformer Overload, 2001 IEEE/PES Transmission &
Distribution Conference, Oct. 28 Nov. 2, 2001, Atlanta, USA

177

Presented at TechCon 2003 Asia-Pacific, May 7-9, 2003, Sidney, Australia
Aging of Cotton/Kraft Blend Insulation Paper in Natural Ester Dielectric Fluid

C.P. McShane K.J. Rapp J. Luksich
Dielectric Fluids Thomas A. Edison Technical Center Dielectric Fluids

Abstract: Previous studies show that both plain and thermally
upgraded Kraft insulation paper age at a much slower rate in natural
ester (vegetable oil) dielectric fluid than in conventional transformer
mineral oil. This study compares the aging rate of electrical grade
Kraft/cotton blend insulation paper in natural ester to that in mineral
oil. Sealed steel aging tubes containing copper, aluminum,
Kraft/cotton paper, and dielectric fluid (mineral oil or natural ester)
were aged at 170C for 500, 1000, and 1500 hours. The degree of
paper degradation after aging is determined using paper tensile
strength and paper degree of polymerization. Paper degradation and
water contents of fluids and papers are compared. The Kraft/cotton
paper degraded at a significantly slower rate in natural ester than in
mineral oil.

INTRODUCTION

Natural esters (vegetable oils) formulated into dielectric fluids
have considerable environmental and fire safety advantages
over conventional transformer mineral oil [1].

The environmental properties of natural esters are such that in
Germany they are classified as non-hazardous to waters [3].
Aquatic biodegradation tests of the natural ester dielectric
fluid used in this experiment found >99% metabolized
conversion to CO
2
, equivalent to compounds defined as
ultimately biodegradable[4]. In acute aquatic toxicity tests,
this fluid had no observable effect on fish at 1000mg/l, the
highest concentration tested [5]. The US Environmental
Protection Agency has verified the published environmental
performance claims. This verification of environmental
technologies was published in May 2002 for Envirotemp

FR3 dielectric fluid [6].

Transformers using natural ester fluids deliver very important
improvements in fire safety compared to those using mineral
oil. Natural ester dielectric fluids have fire points in the range
of 340-360C, compared to 155-165C for conventional
mineral oils. Natural ester dielectric fluids are recognized as
less-flammable per Section 450.23 of the U.S. National
Electrical Code
[7]. A major property risk research and

standards organization recently increased the threshold
minimum volume requiring additional fire safeguards for this
dielectric coolant from 1,000 to 10,000 USgal (3,785 to
37,850 liter). The same threshold volume for mineral oil is 500
USgal (1,892 liter) [8].

Accelerated aging tests of distribution transformers gave early
indications that the rate of paper aging is fluid-dependent [2].
Previous studies quantified this dependence for thermally
upgraded [9] and plain (not thermally upgraded) [10] Kraft
insulation papers. This work examines the aging rate
dependence on fluid type for Kraft and cotton fiber blend
paper provided by an Australian transformer manufacturer.

EXPERIMENTAL

The thermal aging procedure and sample preparation methods
are identical to those previously described [9,10]. The testing
method is based on the procedures outlined in IEEE/ANSI
C57.100 Normative Appendix A, known as the sealed tube
aging test. Sealed steel aging tubes contained 26g of
cotton/Kraft insulation paper
1
dried for 160 hours @ 105C,
350ml of dried degassed dielectric fluid
2
, and typical
transformer proportions of copper and aluminum. The tubes
were aged at 170C and evaluated after 500, 1000, and 1500
hours. One noticeable difference between the paper for this
study is that rather than starting with a light tan color, it has a
blue hue that turns noticeably darker upon impregnation.

The extent of paper aging was determined using changes in the
tensile strength (TS) and the degree of polymerization (D
v
P).
The water contents of both the papers and fluids were also
determined.

RESULTS

The results of aging at 170C, summarized in Table 1,
compares the paper insulation aging rate in natural ester fluid
to that in conventional transformer oil.

1
Whiteley Grade 3 Elephantide
diamond pattern cotton/kraft paper, 0.4mm

2
Cooper Power Systems Tranelec
inhibited insulating mineral oil and

Envirotemp
FR3

Mineral oil
Envirotemp
FR3 fluid

Figure 1. Cotton fiber paper insulation after 170C sealed tube
aging in natural ester and mineral oil.
Time at 170
o
C (hours)
0 500 1000 1500
T
e
n
s
i
l
e

S
t
r
e
n
g
t
h

(
M
P
a
)
0
20
40
60
80
100
120
R
e
t
a
i
n
e
d

T
e
n
s
i
l
e

S
t
r
e
n
g
t
h

(
%

u
n
a
g
e
d
)
100
75
50
25
mineral oil
Envirotemp FR3 fluid
Figure 2. Retained tensile strength versus time for kraft/cotton blend
paper aged in natural ester and mineral oil. Vertical error bars are
1 standard deviation.
The visual difference between the papers aged in the two
fluids can be seen in Fig. 1. After 500 hours of aging, both
samples became darker, and the paper aged in mineral oil was
noticeably brittle. The paper in the natural ester fluid remained
flexible. After 1000 hours, the mineral oil papers were fragile
and difficult to handle. This was not the case with the natural
ester-aged papers. The fragility of the papers correlate well
with changes in both TS and D
v
P of the two fluids.

Tensile Strength
The decrease in TS over time (Figure 2) shows a steep initial
slope of paper degradation in both natural ester and mineral
oil. The highest relative rate of paper degradation takes place
in the first 500 hours of aging in both fluids. The relative rates
in both fluids decrease substantially after 500 hours.

In terms of mechanical strength, the paper in mineral oil,
already below 8.9% retained TS at 500 hours, fell to about
5.5% retained TS value at 1000 hours. The decline in TS of
the paper aged in mineral oil is comparable to that seen in
other studies [11,12]. The paper aged in natural ester degraded
to about 34.0% of its original TS at 500 hours. After 1500
hours, the retained TS was about 21%.

Another measure of cellulose degradation is the length of the
cellulose polymer, given as D
v
P. The change in D
v
P over time
is shown in Figure 3, and is similar to the change seen in TS.
A rapid D
v
P decrease in the first 500 hours of aging was
observed. Paper aged in mineral oil degraded to about 2% of
Table 1. Results of sealed tube aging at 170C of Kraft/cotton blend
paper in mineral oil and natural ester. Water content of dielectric
fluid is given both as absolute content and percent saturation. Total
furanic compounds are given as mg furans per liter of fluid.
Time (hrs) 0
1
500 1000 1500
in mineral oil 0.87 5.69 6.62 3.76
in natural ester 0.80 0.25 0.42 0.15
mineral oil 3 (2) 93 (56) 85 (51) 123 (74)
natural ester 1 (12) 5 (52) 7 (76) 12 (130)
in mineral oil 107 9.6 6.0 6.1
in natural ester 106 36 27 22
in mineral oil 1315 24 too charred to measure
in natural ester 1376 183 184 188

1
baseline values
Time at 170
o
C (hours)
0 500 1000 1500
D
e
g
r
e
e

o
f

P
o
l
y
m
e
r
i
z
a
t
i
o
n

D
v
P
0
300
600
900
1200
1500
mineral oil
mineral oil (too charred to measure)
Figure 3. Degree of polymerization (D
v
P) versus time for
kraft/cotton blend paper aged in natural ester and mineral oil.

the original D
v
P at 500 hours, in line with other published
results [11,12]. The retained D
v
P of paper aged in natural
ester fluid remained above 13% of unaged D
v
P after 1500
hours of aging.

Water Content
At 500 hours the water content of paper aged in mineral oil
increased to 5.7 wt% from its initial state of 0.87 wt%. It
increased again at 1000 hours to a peak of 6.6 wt%. By 1500
hours, the water content of paper aged in mineral oil
decreased to 3.7 wt%, 25 times higher than that of paper aged
in natural ester.

In natural ester, the water content of the paper decreased and
stayed below its initial value. The water content of paper in
natural ester decreased to 0.25wt% from 0.87 wt% after 500
hours and continued to decrease. At 1500 hours, the paper
contained about 0.15 wt% water.

Similar trends were seen in the water content of the fluids
(Table 1). Viewing the water content as percent saturation at
room temperature, both fluids started with water contents
below 5%. The water in mineral oil increased steadily,
reaching 123% saturation at 1500 hours. The water in the
natural ester rose gradually to 12% saturation at 1500 hours.
Time at 170
o
C (hours)
0 500 1000 1500
W
a
t
e
r

C
o
n
t
e
n
t

o
f

P
a
p
e
r

(
w
t
%
)
0
2
4
6
8
mineral oil
Figure 4. Water content versus time for high fiber content paper
aged in natural ester and mineral oil. Paper was originally dried to a
water content of about 0.90 wt%. Vertical error bars are 1 standard
deviation.

DISCUSSION

The TS and D
v
P results are evidence that paper ages slower in
natural ester dielectric fluid than in conventional transformer
oil. The paper aged in mineral oil degraded beyond three
recognized IEEE end-of-life criteria [16]: 50% TS, 25% TS,
and D
v
P=200 all well before 500 hours. The paper aged in
natural ester reached 25% retained TS at about 1000 hours. At
1500 hours, the paper aged in the natural ester fluid was about
21% retained TS.

The water in paper data suggest an explanation for the contrast
in paper degradation seen between fluids [15]. The
acceleration of aging due to water has been known for at least
40 years [17]. As paper thermally degrades it produces water,
promoting further degradation through hydrolysis. In a
conventional transformer paper-in-oil system, the degradation
is autocatalytic [18]. The partitioning of the water between the
paper and fluid depends on the characteristic polarity (affinity
for water) of each.

An essentially non-polar fluid such as mineral oil prevents
most of the water generated during cellulose (wood or cotton
origin) degradation from leaving the paper. Although the
water content of the mineral oil steadily increased to 123 %
saturation at 1500 hours, this is only 74 mg/kg in terms of
absolute water content. The water content of the paper at this
same time point in the test was 3.76 wt%.

The more polar natural ester is much more tolerant of water.
The natural ester has a greater affinity for water than mineral
oil, as seen in their different water saturation points. The
water, attracted to the ester fluid, is liberated from the paper.
Removing water from the paper limits the paper degradation
rate.

The water generated during cellulose degradation migrates out
of the paper and into the natural ester. One would expect the
water in natural ester to increase proportionally. However, the
natural ester remains dry at less than <12% saturation.

Since the aging tube is sealed, water cannot escape. The water
is consumed in a hydrolysis reaction with the natural ester.
The reaction produces long-chain fatty acids. These fatty acids
react with the cellulose in a transesterification reaction [19].
This reaction results in steric protection of the cellulose sites
most susceptible to attack, a consequence of which is reduced
paper degradation [9].

At room temperature, natural esters hold about 18 times more
water in solution than does conventional transformer oil. The
electrical strength of a dielectric fluid depends on the percent
saturation of water, not the absolute water content. The
electrical strength decreases above 50% saturation - about 30
mg/kg in mineral oil and 550 mg/kg in natural esters.

The temperature at which bubbles evolve from an insulation
system strongly depends on the water content of the paper. For
paper having 3.3% water content, the equivalent bubble

evolution temperature would be about 125C. For paper
having 0.05% water content, the bubble evolution temperature
would be >200C [20].

CONCLUSIONS

The results of the accelerated paper aging show that high
cotton fiber content insulation paper takes much longer to
reach IEEE end-of-life conditions in natural ester than in
conventional transformer oil. In addition to improved fire and
environmental safety, applying natural ester dielectric fluids in
transformers should result in improved insulation
performance.

REFERENCES

[1] McShane, C.P., Gauger, G.A., Luksich, J. "Use of Natural
Vegetable Oil Esters as Dielectric Coolant", CIRED 15
th

International Conference on Electricity Distribution, June 1-4,
1999, Nice, France
[2] McShane, C.P., Gauger, G.A., Luksich, J. "Fire Resistant
Its Use", IEEE/PES Transmission & Distribution Conference,
April 12-16, 1999, New Orleans, USA
[3] Administrative Regulation on the Classification of Substances
Hazardous to Waters, Water Hazard Classes (VwVwS), Annex
1 Substances Non-Hazardous to Waters Pursuant to Number
1.2a, Index No. 660, 1999, Germany
[4] Aerobic Aquatic Degradation, US EPA OPPTS 835.3100, Test
Report No. ML-304.98, Thomas A. Edison Technical Center,
Cooper Power Systems, Franksville, USA, 1998
[5] Final Report: Acute Trout Toxicity Testing for Two
Envirotemp FR3 Formulations; Procedure 203, OECD
Guidelines for Toxicity of Chemicals, 1993a, and Biological
Test Method: Acute Lethality Test Using Rainbow Trout,
Environment Canada, 1990 (1996 Amendments); GlobalTox
International Consultants, Guelph, Canada, Dec. 1, 1999
[6] Environmental Technology Verification Report: Cooper Power
Systems Envirotemp FR3 Vegetable Oil-Based Insulating
Dielectric Fluid, U.S. Environmental Protection Agency,
Publication EPA 600/R-02/042, May 2002.
[7] National Electrical Code, NFPA 70, 2002 Edition, National
Fire Protection Association, Quincy, MA, USA
[8] FM Global Property Loss Prevention Data Sheets 5-4, January
1997, Revised March 2003, Norwood, MA, USA
Luksich, J., Aging of Paper Insulation in Natural Ester
Dielectric Fluid, 2001 IEEE/PES Transmission & Distribution
Conference, Oct. 28 Nov. 2, 2001, Atlanta, USA
Luksich, J., Aging of Kraft Paper in Natural Ester Dielectric
Fluid, Proceeding of 145h International Conference on
Dielectric Liquids (ICDL), Graz (Austria), July 7-12, IEEE 0-
7803-7350-2/02/ 2002 IEEE
[11] Darveniza, M., Saha, T.K., Hill, D.J.T., Le, T.T. Investigations
into Effective Methods for Assessing the Condition of
Insulation in Aged Power Transformers, IEEE Trans PD, Vol.
13, No. 4, pp. 1214-1223, Oct. 1998
[12] Hill, D.J.T., Le, T.T., Darveniza, M., Saha, T. A Study of
Transformer. Part 2: Tensile Strength of Cellulose Insulation
Paper, Polymer Degradation and Stability, Vol. 49, pp. 429-
435, 1995
[13] Emsley, A.M., Stevens, G.C., A Reassessment of the Low
Temperature Thermal Degradation of Cellulose, IEEE 6
th

International Conference on Dielectric Materials,
Measurements, and Applications, pp.229-32, Sept. 7-10, 1992,
Manchester, UK
[14] Miyoshi, A., A New Additive for Improving the Thermal
Aeging Characteristics of Insulating Paper, IEEE Trans EI,
Vol. E1-10 (1), 13, 1975
[15] Unsworth, J., Mitchell, F. Degradation of Electrical Insulating
Paper Monitored with High Performance Liquid
Chromatography, IEEE Trans EI, Vol. 25, No. 4, pp. 737-746,
Aug 1999
Transformers, IEEE Std C57.91-1995, April 25, 1996, IEEE,
New York, USA
[17] Clark, F.M. Insulating Materials for Design and Engineering
Practice, pp. 287-290, Wiley, USA, 1962
[18] Hill, D.J.T., Le, T.T., Darveniza, M., Saha, T. A Study of the
Transformer. Part 3: Degradation Products of Cellulose
Insulation Paper, Polymer Degradation and Stability, Vol. 51,
pp. 211-218, 1996
[19] Huang, M.R., Li, X.G. Thermal Degradation of Cellulose and
Cellulose Esters, J. Appl. Polymer Sci., Vol. 68, pp. 293-304,
1998.
[20] Oommen, T.V., Lindgren, S.R. Bubble Evolution from
Transformer Overload, 2001 IEEE/PES Transmission &
Distribution Conference, Oct. 28 Nov. 2, 2001, Atlanta, USA

Biography of the Authors

C. Patrick McShane received his BS in Electrical
Engineering from Marquette University, Milwaukee,
Wisconsin, USA in 1970, and an MS in Engineering
Management from the Milwaukee School of Engineering in
1998. Currently he is the Product Manager for Dielectric
Fluids and Transformer Components at Cooper Power

Systems, a division of Cooper Industries. His employment
experience includes positions as International Area manager
for RTE Corp. and Technical Director for the State of Sao
Paulo (Brazil) Rural Electrification Program. His professional
activities include USA Delegate to IEC TC99, IEC TC89
Expert Delegate, Chair ASTM Working Group for D-5222,
Working Group Chair for IEEE Transformer Committee
Dielectric Fluids Subcommittee C57.121, and currently
Working Group Chair for the Development of IEEE Standard
Guide for Ester Based Dielectric Coolants. He is the principal
inventor of four US Patents relating to dielectric fluids and
their applications.

John Luksich received his BS in Chemistry in 1980 and an
MS in Materials Engineering in 1990, both from the
University of Wisconsin. His engineering career includes thin
film materials development at the McDonnell Douglas Space
and Physics Laboratory and sensor development for Johnson
Controls. He is currently an engineer for Cooper Power
Systems specializing in dielectric fluids. He is the ASTM
Working Group Chair for the development of Natural Ester
Based Dielectric Fluids standard.

Kevin Rapp is receiving his BS in Chemistry from the
University of Wisconsin in May, 2003. Since joining the
Thomas A. Edison Technical Center in 1976, he has been
involved with dielectric materials development for electrical
power equipment. His focus for the last ten years has been on
research and development of natural ester based dielectric
coolants.

Presented at the 2003 IEEE/DEIS Conference on Electrical Insulation and Dielectric Phenomena,
Oct. 19-22, 2003, Albuquerque, NM
Aging of paper insulation retrofilled with natural ester dielectric fluid

C.P. McShane
1
, J.L. Corkran
1
, K.J. Rapp
2
, J. Luksich
1

1
Cooper Power Systems, Waukesha, WI USA
2
Thomas A. Edison Technical Center, Cooper Power Systems, Franksville, WI USA

Abstract: The aging rate of transformer insulation
Kraft paper is much slower in natural ester (vegetable
oil) dielectric fluid than in conventional transformer oil.
This study investigates the effect that replacing
transformer oil with natural ester fluid (retrofilling) has
on the aging rate of thermally upgraded (65C rise)
paper initially aged in transformer oil. Sealed steel
aging vessels containing copper, aluminum, dried
thermally upgraded Kraft paper, and dielectric fluid
(transformer oil or natural ester) were aged at 160 and
170C for 250, 500, 750, 1000, 1500, and 3000 hours.
Half of the transformer oil systems were retrofilled with
natural ester fluid after initial aging times of 750 and
250 hours at 160 and 170C, respectively. Paper
degradation after aging is determined using paper
tensile strength and degree of polymerization
measurements. After replacing the transformer oil with
natural ester, the aging rate of the paper initially aged in
transformer oil showed an abrupt change to the reduced
aging rate for paper in a natural ester.

Introduction

Natural (vegetable oil) ester insulating fluid offers fire
safety, environment, and insulation aging advantages
over mineral oil, and are found to be suitable for use in
transformer insulation systems [1]. Previous sealed tube
[2] aging studies show that the thermal aging rates of
virgin paper insulation in natural ester insulating fluid
are significantly slower than those in mineral oil [3-5].
The studies to date compare insulation papers
exposed only to mineral oil or natural ester fluid,
simulating their use in new construction. Existing in-
service mineral oil-filled transformers are sometimes
retrofilled (that is, the in-service fluid is replaced with
new fluid). How is the paper insulation aging rate in
natural ester fluid affected by previous aging in mineral
oil?
This experiment examines the aging characteristics
of thermally upgraded Kraft insulation paper aged in
mineral oil, then retrofilled with natural ester fluid.

Experiment

Detailed descriptions of the test apparatus and
procedures are given in [2]. Thermally upgraded Kraft
insulation paper
1
, aluminum and copper conductor
materials, and dielectric fluid
2
were sealed in steel tubes
with a nitrogen headspace. The fluid, immersed
materials, and headspace are in proportions typical of
distribution transformers. The paper was dried to
approximately 0.5 wt% water content. The insulating
fluids were dried and degassed.
Three sets of sealed tube aging systems were
prepared, one set using natural ester fluid, and two sets
using mineral oil. Each set of systems contained a
sealed tube for each sampling time and temperature,
shown in Table 1.
For each temperature, the duplicate mineral oil
systems were removed from the aging ovens at the first
scheduled sampling time, drained of mineral oil,
refilled with natural ester fluid, and returned to the
aging ovens.

Table 1. Times and temperatures for sealed tube aging
systems using thermally upgraded paper in either mineral oil
or natural ester fluid. Retrofills of duplicate mineral oil
systems are done at end of first test period for each
temperature.
Temperature
Time (hrs) ambient 160C 170C
0 x
250 x
500 x
750 x x
1000 x x
1500 x x
2000 x
3000 x x

1
Whiteley Ltd. Grade K T/U 0.255mm thermally upgraded Kraft
paper
2
Ergon Hyvolt II inhibited insulating mineral oil, Cooper Power
Systems Envirotemp
FR3

Page 1 of 4
Table 2. Results of thermally upgraded Kraft paper sealed tube aging at 160 and 170C in mineral oil and natural
ester. Total furanic compounds are given as mg furans per liter of fluid per initial kg of paper.
Temperature 20C 160C 170C
Time (hrs) 0
a,b
750
b
1005 1500 2184 3022 250
b
505 750 1009 1500 3002
in mineral oil 0.46 0.37 0.71 0.66 2.22 6.29 0.58 0.99 1.55 5.92 4.28 2.79
in natural ester 0.72 0.38 0.34 0.29 0.20 0.25 0.15 0.12 0.10 0.08 0.17 0.20
retrofill 0.46 0.37 0.39 0.27 0.15 0.20 0.58 0.16 0.18 0.21 0.20 0.16
in mineral oil 148 114 97.7 74.4 22.9 6.2 110 80.1 28.1 7.7 0.7 -
c

in natural ester 142 112 108 107 97.6 86.5 111 96.8 88.7 79.6 66.5 57.3
retrofill 148 114 101 98.7 92.1 80.8 110 83.5 77.9 66.3 57.2 53.9
in mineral oil 1,152 496 405 271 132 31 481 291 73 24 5 0
in natural ester 1,148 628 566 444 366 361 547 435 411 333 314 226
retrofill 1,152 496 422 387 331 320 481 348 328 299 268 202
Furanic Compound Content of Fluid [mg/l/kg]
mineral oil nd 1,442 1,216 9,018 30,783 17,193 1,662 11,976 52,981 25,977 15,653 14,957
natural ester 1,044 613 959 608 649 928 966 637 502 611 1,058 851
retrofill nd 1,442 2,818 1,002 1,010 1,508 1,662 1,162 465 917 1,003 792
a
baseline values
b
retrofill values at these times are identical to mineral oil sealed tube systems
c
too degraded to measure
Results

The extent of paper degradation is determined using the
changes in paper tensile strength and degree of
polymerization (DvP). The water contents of the paper
and furanic contents of the insulating fluid were also
determined. The results are presented in Table 2.

Paper Condition at Retrofill: The initial condition of
the retrofill systems is the condition of the mineral oil
systems at their first sample point.
The 160C mineral oil systems reached about 77%
of initial tensile strength and 43% of initial degree of
polymerization at 750 hours. The duplicate 160C
mineral oil systems were retrofilled at this point.
The 170C mineral oil systems reached about 74%
of initial tensile strength and 60% of initial degree of
polymerization at 250 hours. The duplicate 170C
mineral oil systems were retrofilled at this point.

Tensile Strength: The paper tensile strength over time
is shown in Fig. 1. After retrofilling with natural ester,
the decline in tensile strength of the duplicate mineral
oil systems changed from the mineral oil rate to that of
the natural ester system.
The tensile strength of the mineral oil system fell
below 25% that of unaged paper by 2000 hours at
160C, and 750 hours at 170C.
Both the natural ester and retrofill systems
160
o
C
T
e
n
s
i
l
e

S
t
r
e
n
g
t
h

(
M
P
a
)
0
25
50
75
100
125
150
R
e
t
a
i
n
e
d

T
e
n
s
i
l
e

S
t
r
e
n
g
t
h

(
%

u
n
a
g
e
d
)
100
75
50
25
170
o
C
Time (hours)
0 1000 2000 3000
0
25
50
75
100
125
150
100
75
50
25
mineral oil
natural ester
retrofill
Figure 1: Tensile strength versus time at temperature for
thermally upgraded Kraft paper in sealed tube aging systems.
Duplicate mineral oil systems were retrofilled with natural
ester fluid at 250 and 750 hours at 170 and 160C,
respectively.
Page 2 of 4
remained above 25% of unaged paper. At the end of
test (3000 hours) at 160C, the natural ester system was
61%; the retrofill system was 40%. At the end of test at
170C, the natural ester system reached 40% of unaged;
the retrofill system reached 36%.
160
o
C
D
e
g
r
e
e

o
f

P
o
l
y
m
e
r
i
z
a
t
i
o
n
0
200
400
600
800
1000
1200
170
o
C
Time (hours)
0 1000 2000 3000
0
200
400
600
800
1000
1200
mineral oil
natural ester
retrofill

Figure 2: Degree of polymerization versus time at
temperature for thermally upgraded Kraft paper in sealed
tube aging systems. Duplicate mineral oil systems were
retrofilled with natural ester fluid at 250 and 750 hours at 170
and 160C, respectively.
F
u
r
a
n
i
c

C
o
n
t
e
n
t

(
p
p
b
)
0
5000
10000
15000
20000
25000
30000
35000
mineral oil
natural ester
retrofill
160
o
C
Time (hours)
0 1000 2000 3000
0
10000
20000
30000
40000
50000
60000
170
o
C
Figure 3: Furanic content versus time at temperature for
respectively.

Degree of Polymerization: D
v
P over time is shown in
Fig. 2. As in tensile strength, the retrofill rate of paper
degradation as measured by D
v
P changed from that of
mineral oil to that of natural ester.
Paper in mineral oil fell below 200 D
v
P between
1500 and 2000 hours at 160 C and 500-750 hours at
170C. At 3000 hours, both the natural ester and
retrofill systems remained above 300 D
v
P at 160C, and
slightly above 200 D
v
P at 170C.

Furanic Compounds: Furanic compounds found in
insulating fluid are the result of cellulose degradation,
and can be used to monitor the insulation degradation
of in-service transformers [6]. The furanic compounds
content of the insulating fluids over time are shown in
Fig. 3.
The amounts of furanic compounds (total content
of 2-furaldehyde, 5-hydroxymethyl-2-furfural, 2-acetyl
furan, and 5-methyl-2-furfural) found in the natural
ester and retrofill systems remained an order of
magnitude lower than those found in the mineral oil
system.

Water Content: Water is produced during the
degradation of cellulose, and accelerates further
breakdown. The water contents of paper over time is
shown in Fig 4.
The paper in both the natural ester and retrofill
systems remained dry, falling well below the initial
water content. The paper in the mineral oil systems
became very wet, reaching 6 wt%.

Discussion

The tensile strength, D
v
P, and furanic content results
establish that the aging rate of thermally upgraded Kraft
paper initially aged in mineral oil can be modified by
retrofilling with natural ester insulating fluid. Upon
retrofilling, the rate of paper degradation changes from
that of paper in mineral oil to the slower rate of paper in
natural ester fluid.
At both temperatures the paper aged in mineral oil
degraded beyond three recognized IEEE end-of-life
criteria [7]: 50% tensile strength, 25% tensile strength,
and 200 D
v
P. The paper in natural ester and retrofill
paper did not reach any end-of-life point at 160C. At
170C, the paper aged in natural ester fluid reached
only the 50% tensile strength end-of-life point. The
retrofilled paper reached 50% tensile strength end-of-
life and nearly reached 200 D
v
P.
Page 3 of 4
The water in paper data provides a rationale for the
contrast in paper degradation rates between mineral oil
and natural ester fluid. Firstly, the presence of water
accelerates cellulose aging. Secondly, as paper
thermally degrades it produces water, promoting further
degradation through hydrolysis.
Natural esters have a greater affinity for water than
does mineral oil, so less water will remain in the paper
as the water partitions itself between headspace, paper,
and fluid. Water is consumed by hydrolysis of the
natural ester, producing free fatty acids. These may
react with the cellulose backbone via transesterification
and protect the cellulose from hydrolysis.

References
0
1
2
3
4
5
6
7
mineral oil
natural ester
retrofill
160
o
C
Time (hours)
0 1000 2000 3000
W
a
t
e
r

C
o
n
t
e
n
t

(
w
t

%
)
0
1
2
3
4
5
6
7
170
o
C
Figure 4: Water content versus time at temperature for
respectively.

[1] C.P. McShane, G.A. Gauger, J. Luksich, "Fire Resistant
Natural Ester Dielectric Fluid and Novel Insulation
System for Its Use", IEEE/PES Transmission &
Distribution Conference, April 12-16, 1999, New
Orleans, LA USA
[2] Standard test procedure for sealed tube aging of liquid-
immersed transformer insulation, Annex A, IEEE
Standard Test Procedure for Thermal Evaluation of
Liquid-Immersed Distribution and Power Transformers,
IEEE Std C57.100-1999, June 26, 1999, IEEE, New
York, USA
[3] C.P. McShane, K.J. Rapp, J.L. Corkran, G.A. Gauger, J.
Luksich, Aging of Paper Insulation in Natural Ester
Dielectric Fluid, 2001 IEEE/PES Transmission &
Distribution Conference, October 28 - November 2,
2001, Atlanta, GA USA
[4] C.P. McShane, K.J. Rapp, J.L. Corkran, G.A. Gauger, J.
Luksich, "Aging of Plain Kraft Paper in Natural Ester
Dielectric Fluid", IEEE/DEIS 14th International
Conference on Dielectric Fluids, July 7-12, 2002, Graz,
Austria
[5] C.P. McShane, K.J. Rapp, J.L. Corkran, J. Luksich,
"Aging of Cotton/Kraft Blend Insulation Paper in
Natural Ester Dielectric Fluid", TechCon 2003 Asia-
Pacific, May 7-9, 2003, Sidney, Australia
[6] Unsworth, J., Mitchell, F. Degradation of Electrical
Insulating Paper Monitored with High Performance
Liquid Chromatography, IEEE Trans EI, Vol. 25, No. 4,
pp. 737-746, Aug 1999
Transformers, IEEE Std C57.91-1995, April 25, 1996,
IEEE, New York, USA

Page 4 of 4
Proceedings of 15
th
International Conference on Dielectric Liquids (ICDL), Coimbra (Portugal), June 26 - July 1, 2005
Page 1 of 4
Interaction Mechanisms of Natural Ester Dielectric Fluid and Kraft Paper
K. J. Rapp, C. P. McShane, and J. Luksich

Cooper Power Systems, Waukesha, Wisconsin 53188 USA

Abstract: Sealed tube accelerated aging studies demonstrate a
slower aging rate for cellulose insulation in natural (vegetable
oil) ester dielectric fluid compared to the rate in conventional
transformer oil. The interactions of natural ester fluid and
cellulose insulation resulting in increased paper life are
described by two interrelated chemical reaction mechanisms.
Compared to the conventional transformer oil/Kraft paper
system, the natural ester fluids greater affinity for water shifts
more water from the paper into the fluid in order to maintain
equilibrium. The natural ester fluid reacts via the primary
mechanism of hydrolysis to consume dissolved water in the fluid,
shifting further the paper/fluid equilibrium to further dry the
paper and produce free fatty acids. These fatty acids serve as
reactants in the secondary mechanism of transesterification to
modify the cellulose structure. The change in cellulose structure
is verified using infrared analysis.
INTRODUCTION
Multiple sealed tube aging studies demonstrate that, under
identical conditions, Kraft paper insulation ages significantly
slower in natural (vegetable oil) ester fluid than in
conventional transformer oil [1-4]. Multiple and interrelated
mechanisms have been proposed to explain the increase in
paper insulation life [1].
PROPOSED MECHANISMS
Kraft paper insulation aging rates increase with both
increasing temperature (thermo-kinetic degradation) and
increasing water content (thermo-hydrolytic degradation) [5].
The primary and secondary mechanisms hypothesized to be
responsible for the different aging rates are related to
differences in how natural ester fluid and transformer oil
interact, physically and chemically, with water. The primary
mechanism is a net movement of water generated by paper
aging from the paper into the natural ester fluid followed by
hydrolysis of the ester. The secondary mechanism is the
reaction of the hydrolysis products with the cellulose
resulting in protection of vulnerable sites on the cellulose
chain.
Equilibrium Shift
Water moves between the paper insulation and dielectric fluid
trying to reach equilibrium in terms of relative saturation. Fig.
1 shows that the natural esters can accommodate more water
than conventional transformer oil [6,7]. This difference
causes an equilibrium shift toward the natural ester fluid. That
is, more water must be transferred from the paper to the fluid
Fig. 1. Water saturation versus temperature for natural ester fluid
and conventional transformer oil [6,7].
in order to reach and maintain equilibrium. Fig. 2 shows Kraft
paper aged in natural ester fluid and conventional transformer
oil, and is typical of a variety of papers aged in these fluids
[1-4]. The paper aged in natural ester fluid remains dry, while
the water content of paper aged in conventional transformer
oil rises to above 3 wt%. As paper breaks down and produces
water as a product of degradation, the natural ester fluid
absorbs more of the water compared to conventional
transformer oil (Fig. 3).
Hydrolysis
Water reacts with the triglycerides comprising the natural
ester via hydrolysis to produce long-chain fatty acids (Fig. 4).
Fig. 2. Water content of Kraft insulation paper aged in natural ester
fluid and conventional transformer oil [2].

Page 2 of 4
Fig. 3. Water contents of natural ester fluid and conventional
transformer oil after accelerated aging [2].
The reaction consumes dissolved water in the fluid causing
additional water to move from the paper into the fluid in order
to maintain equilibrium.
Transesterification
A secondary mechanism appeared to be theoretically viable
and was indirectly supported by the initial data. The acids
produced by hydrolysis react with the cellulose via
transesterification. Under accelerated aging, the reactive OH
(hydroxyl) groups on the cellulose molecule become
esterified with fatty acid esters. This reaction takes place at
lower temperatures, but of course at slower rates of reaction.
There is significant evidence that long chain fatty acids can be
esterified onto cellulose [8-14]. One study used mixed fatty
acids from hydrolyzed soybean oil, reactants similar to those
that might be found in natural ester dielectric fluids, to
esterify cellulose [8]. To obtain a reasonable degree of
substitution within a 24-hour time frame, cellulose was
typically activated with either mild acid or base and the

Fig. 4. Acid numbers of natural ester fluid and conventional
transformer oil after accelerated aging [2].
esterification was done with fatty acid chlorides, anhydrides,
carboxylic acids, etc. The reactions were run at temperatures
of 60 to 90C to obtain substitution, but avoid depolymerizing
the cellulose chains. It is reasonable to assume that higher
temperatures would drive a transesterification reaction
directly between cellulose and less active long-chain fatty
acid esters, such as those of natural ester fluids. The
esterification of Kraft cellulose almost certainly progresses
after the natural ester fluid penetrates the cellulose and
hydrolyzes with available water.
ANALYSIS
Molecular Modeling
The long-chain fatty acids from the natural ester would
preferentially esterify the primary C-6 hydroxyl group on the
cellulose chain. This hydroxyl group is the least sterically
hindered of the three hydroxyls and is most likely to react
with a bulky substituent [11].
A monomer unit of cellulose, an anhydro--D-glucopyranose,
was built using a molecular modeling computer program (Fig.
5). The model performed both quantum mechanical
calculations and classical mechanics calculations for the
molecular structure. The single point energy was calculated
using Hartree-Fock 3-21G and a neutral total charge. The
electrostatic potential surfaces were mapped along with the
molecular conformation set at its energy minima. The relative
sizes of the electrostatic potential surfaces correlate directly to
the availability of bonding electrons.
The analysis shows that the hydroxyl group attached to each
primary carbon (C-6) has the highest electron density and
negative charge, which influence reactions on the molecular
surface. Both of these groups are located at the top-center of
Fig. 5. Molecular model of cellulose monomer unit showing
electrostatic potential surfaces. The size of the electrostatic potential
surfaces correlates with the availability of bonding electrons.

Page 3 of 4
the monomer unit. In a 3-D model, the top left surface would
project out from the surface of the page and the top right
would recede into the page. The arrow attached to the small
surface in the center of the molecule depicts the dipole of the
monomer unit of cellulose, which points toward the more
electronegative C-6 hydroxyls.
Fourier Transform Infra-Red Analysis
Fourier Transform Infra-Red (FTIR) analysis using attenuated
total reflectance was employed to determine changes in the
structure of the Kraft paper aged in natural ester fluid and
conventional transformer oil. Samples aged for 0, 500, 1000,
2000, and 4000 hours at 170C were soxhlet-extracted with
hexane. FTIR spectra indicate that the paper aged in natural
ester fluid changed very little through 4000 hours, while the
paper aged in conventional transformer oil showed a gross
change. One highlight for the ester aged paper is the
appearance of a weak band at 1717 cm
-1
that increases with
increasing aging time. This carbonyl band indicates the
presence of an ester bonded to the cellulose, providing
evidence that the transesterification reaction has taken place.
The highly aged paper from the mineral oil displays an
increasingly strong carbonyl band at 1700 cm
-1
, consistent
with an aldehyde.
Nuclear Magnetic Resonance
Proton nuclear magnetic resonance (NMR) spectroscopy was
employed to determine the relative ratio of saturated to
unsaturated hydrogen in fresh natural ester fluid compared to
fluid aged with thermally upgraded paper for 4000 hours at
170C. A comparison of the hydrogens (protons) that make
up a hydrocarbon molecule is a technique to verify basic
molecular structure. The spectrum shows a triplet at a
chemical shift of 5.3 ppm, a characteristic of protons bonded
to carbon-carbon double bonds, and was used to verify
unsaturation. A triplet at 4.3 ppm is due to water, and the
multitude of peaks seen between 1 and 3 ppm are due to
saturated molecules. A large difference in the ratio of
unsaturation to saturation indicates that the molecular
composition of the fluid changed. This analysis showed ratios
of 5.5 for fresh fluid and 6.0 for fluid aged with cellulose for
4000 hours. This difference is considered insignificant and
indicates that the natural ester fluid through the course of
4000 hours with Kraft paper contains the same relative
amounts of saturates to unsaturates as fresh fluid. Thus, in a
sealed environment, the more vulnerable unsaturated chains
of the natural ester fluid were verified to be mostly intact,
which would be unlikely if an oxidation process was a factor.
This supports the results of full-scale transformer evaluations
of natural ester fluid [15], where hydrolysis, not oxidation,
was found to be the main fluid breakdown mechanism during
thermal aging in sealed transformers.
CONCLUSIONS
The greater affinity of natural ester fluids for water, combined
with hydrolysis, cause water to move from the Kraft paper
insulation into the natural ester fluid in larger quantities than
in conventional transformer oil. This cumulative equilibrium
shift of water effectively dries the paper and reduces aging
due to thermo-hydrolytic degradation.
Transesterification occurs under the conditions of elevated
temperature accelerated aging to chemically modify and
protect the Kraft paper and further retard paper aging.
Hydrolysis provides the fatty acids necessary for chemical
modification of the cellulose to take place. The reactive OH
groups on the cellulose molecule become esterified with fatty
acid esters via transesterification, hindering cellulose
degradation mechanisms utilizing these sites. This reaction is
expected to take place at slower rates of reaction at lower
temperatures.
The predominant natural ester fluid reaction during elevated
temperature aging is hydrolysis, not oxidation. This, in effect,
maintains the level of unsaturated molecules in the fluid. This
available source of reactant makes the transesterification
reaction more favorable.
ACKNOWLEDGMENTS
The authors wish to express their gratitude to Mary
Slovachek, Nancy Lowis, and Monica Donovan of the
Thomas A. Edison Technical Center. This work would not be
possible without their support. The authors also thank
Professor Vera Kolb of the University of Wisconsin -
Parkside for technical discussions during the course of this
work.
REFERENCES
[1] C. P. McShane, K. J. Rapp, J. L. Corkran, G. A. Gauger, and J.
Luksich, Aging of paper insulation in natural ester dielectric
fluid, IEEE/PES Transmission & Distribution Conf, 2001,
Atlanta, USA.
[2] C. P. McShane, K. J. Rapp, J. L. Corkran, G. A. Gauger, and J.
Luksich, Aging of plain Kraft paper in natural ester dielectric
fluid, IEEE/DEIS Conf. on Dielectric Fluids, 2002, Graz,
Austria.
[3] C. P. McShane, K. J. Rapp, J. L. Corkran, and J. Luksich,
Aging of cotton/Kraft blend insulation paper in natural ester
dielectric fluid, TechCon Asia-Pacific, 2003, Sidney,
Australia.
[4] C. P. McShane, J. L. Corkran, K. J. Rapp, and J. Luksich,
Aging of paper insulation retrofilled with natural ester
dielectric fluid, IEEE Conf. on Electrical Insulation and
Dielectric Phenomena, 2003, Albuquerque, USA.
[5] H. P. Moser and V. Dahinden, Transformerboard II, 2nd ed.,
Rapperswil: H. Weidmann AG, 1999, pp. 137-144

Page 4 of 4
[6] L. Lewand, Laboratory evaluation of several synthetic and
agricultural-based dielectric liquids, Doble Int. Client Conf,
5E, 2001, Boston, USA.
[7] IEEE Guide For Acceptance and Maintenance of Insulating
Oil in Equipment, C57.106-2002, New York: IEEE, 2002, p. 9
[8] P. Wang and B. Y. Tao, Synthesis and characterization of
long-chain fatty acid cellulose ester (FACE), J. Appl. Polymer
Sci., vol. 52, pp. 755-761, 1994.
[9] H. S. Kwatra, J. M. Caruthers, and B. Y. Tao, Synthesis of
long chain fatty acids esterified onto cellulose via the vacuum-
acid chloride process, Ind. Eng. Chem. Res., vol. 31, no. 12,
1992.
[10] G. Chauvelon, L. Saulnier, A. Buleon, J.F. Thibault, C.
Gourson, R. Benhaddou, R. Granet, and P. Krausz, Acidic
activation of cellulose and its esterification by long-chain fatty
acid, J. Appl. Polymer Sci., vol. 74, pp. 1933-1940, 1999.
[11] C. Vaca-Garcia and M. E. Borredon, Solvent-free fatty
acylation of cellulose and lignocellulosic wastes part 2:
reactions with fatty acids, Bioresource Tech., vol. 70, pp. 135-
142, 1999
[12] K. J. Edgar, T. J. Pecorini, and W. G. Glasser, Long-chain
cellulose esters: preparation, properties, and perspective, ACS
Symposium Series (Cellulose Derivatives), vol. 688, pp. 38-60,
1998
[13] C. Q. Yang and X. Wang, Applications of FT-IR spectroscopy
to the studies of esterification and crosslinking of cellulose by
polycarboxylic acids: part I formation of cyclic anhydrides as
the reactive intermediates, Int. Conf. Fourier Transform
Spectroscopy, Am. Inst. Physics, pp. 661-664, 1998.
[14] W. Wei and C. Q. Yang, Applications of FT-IR spectroscopy
to the studies of esterification and crosslinking of cellulose by
polycarboxylic acids: part II the performance of the crosslinked
cotton fabrics, Int. Conf. Fourier Transform Spectroscopy,
Am. Inst. Physics, pp. 665-668, 1998.
[15] C. P. McShane, G. Gauger, and J. Luksich, Fire Resistant
Its Use, IEEE/PES Transmission & Distribution Conf., 1999,
New Orleans, USA.

Presented at the EIA/IEEE 10
th
Insucon International Electrical Insulation Conference, Birmingham, UK, May 2426, 2006
EFFECT OF NATURAL ESTER (VEGETABLE OIL) DIELECTRIC FLUID ON THE WATER
CONTENT OF AGED PAPER INSULATION

A. W. Lemm, K. J. Rapp, and J. Luksich

Cooper Power Systems, USA

ABSTRACT

Previous insulation aging studies show a significant
reduction of paper insulation aging rate when aged in
natural ester dielectric fluid compared to paper aged
in mineral oil. This study examines the effects at
typical transformer operating temperatures of the
natural ester fluid on previously degraded Kraft
insulation paper with elevated water contents. To
replicate service-degraded insulation, thermally
upgraded Kraft paper was aged in mineral oil at
170C for 400 hours, resulting in a decrease of the
tensile strength to about 70% of the original value
and an increase in its water content. The aged paper
was equilibrated to 1 and 3 %wt water contents,
representing wet transformers. Seventy sealed tube
systems were assembled using combinations of aged
paper with new dried, degassed, and filtered
(hereafter referred to as processed) natural ester fluid,
new processed mineral oil, and original mineral oil
used in the initial aging of the paper. The systems
were held at laboratory ambient, 85C, and 110C for
360, 720, 1440, 2160, and 3000 hours. In all cases the
water content of paper in the natural ester systems
was reduced to a lower level than that obtained in
mineral oil. The water contents of paper in the
original mineral oil remained unchanged. The
decrease in paper water content in conjunction with
changes in natural ester water content and acid level
reinforce the primary paper aging rate reduction
mechanisms proposed in prior work.

INTRODUCTION

Natural ester (vegetable oil) dielectric fluids are in
many ways an effective alternative to the mineral oil
coolant commonly used in distribution and power
transformers. Natural ester fluids have favorable
environmental and fire safety properties, most
notably rapid and complete biodegradation and
categorization as K class fluid per IEC 61100.

Previous sealed tube aging studies demonstrate that,
under identical conditions, Kraft paper insulation
ages significantly slower in natural ester fluid than in
conventional transformer oil. Thermally upgraded
Kraft [1], plain Kraft [2], and cotton-Kraft blend [3]
exhibit this behavior when aged in natural ester fluid
alone compared to aging in mineral oil alone.
Replacing the original mineral oil with natural ester
fluid in sealed tube systems after sufficient aging to
moderately degrade the Kraft paper, simulating the
retrofill of an in-service mineral oil transformer,
demonstrates that the paper aging rate converts from
the rate in mineral oil to the slower rate of paper in
natural ester fluid [4].

Kraft paper insulation aging rates increase with
increasing temperature (thermo-kinetic degradation)
and increasing water content (thermo-hydrolytic
degradation) [5]. The primary and secondary
mechanisms hypothesized to be responsible for the
different aging rates are related to differences in how
natural ester fluid and transformer oil interact,
physically and chemically, with water and cellulose.
The primary mechanism is a net movement of water
generated by paper aging from the paper into the
natural ester fluid, followed by hydrolysis of the ester
[6].

In order to obtain results in a reasonable amount of
time, the previous studies used aging temperatures of
130-170C to accelerate the paper degradation
process. This work examines the effect of natural
ester fluid on partially degraded wet Kraft paper at
typical transformer operating temperatures.

EXPERIMENTAL

Sample Preparation

The test systems, identical to those used in previous
studies [1-4], consisted of heavy-walled steel
cylinders having a bolted cover fitted with an o-ring
gasket and high temperature valve. The cylinders
contained Kraft paper at 1 or 3 %wt water, copper,
aluminum, and 350ml of insulating fluid. The sealed
tubes were evacuated, and pressurized with nitrogen
to 110 kPa.

Initial Aging.
Thermally upgraded electrical grade 0.25mm Kraft
paper was cut into 25.4mm x 127mm strips. The
strips, along with copper and aluminum sheets in the
proportions used in a typical distribution transformer,
were placed into large steel tanks. The tanks were
placed into forced air ovens to initiate the drying of
the components. After the initial drying process, the
containers were transferred to a preheated vacuum
oven where drying was completed. This drying
Page 1 of 6

th
procedure produces paper water contents of
approximately 0.5 %wt. Processed mineral oil was
then introduced under vacuum into the steel cylinders
to fully impregnate the paper with oil. The cylinders
were cooled under vacuum to laboratory ambient.
The cooled cylinders were removed from the vacuum
oven and immediately sealed and pressurized with
nitrogen to 110 kPa so that the internal pressure
remained above the vapor pressure of water at 170C.
These assembled test systems were placed in a forced
air oven at 170C for 400 hours. This combination of
time and temperature degrades the paper to about
70% of the original tensile strength and increases the
water content to about 1 %wt.
Figure 1 Chamber used to equilibrate the water
content of the previously aged Kraft paper. Fans
inside the chamber circulate the air to stabilize the
humidity throughout and assure consistent exposure
of all specimens. Sensors continuously monitor
temperature and humidity.

Establishing Water Content.
At the completion of the paper aging sequence, the
systems were opened and the materials placed in a
glass container sufficient to hold the entire contents
of all test systems. This permits the contents to be
easily blended into a homogeneous system. Multiple
random samplings of paper from the homogenous
system confirmed a 1 %wt paper water content. Half
of the paper strips from the homogenous system were
removed, blotted to remove any residual surface oil,
and suspended in an incubator maintained at 25C
(Fig 1). A large pan of water was placed directly
beneath the circulating fans in the chamber and the
chamber closed. This results in a relative humidity
level in the chamber of 40-45%. The temperature and
humidity were monitored and maintained for 10 days.
Randomly selected paper strips were evaluated and
found to have stabilized at approximately 3 %wt
water content.

Retrofill and Dry-Out

With the paper water contents established at
approximately 1 and 3 %wt, new aging systems were
prepared to emulate 3 typical aged transformer
scenarios:
1) An aged transformer,
2) An aged transformer that has the original mineral
oil reprocessed, or replaced with new processed
mineral oil,
3) An aged transformer that has the aged mineral oil
replaced with new processed natural ester fluid.

The requisite proportions of paper, copper,
aluminum, and processed insulating fluid were placed
into the steel aging cylinders and sealed. A vacuum
was applied to help re-impregnate the paper with
insulating fluid, and the systems were again
pressurized with nitrogen to assure that the pressure
remained above the vapor pressure of water at the test
temperatures. The test systems were inverted and
maintained at the test temperatures of laboratory
ambient, 85C, and 110C for times ranging from 360
to 3000 hours (Table 1).

RESULTS

The paper tensile strength was measured after the
initial aging sequence and periodically during the
dry-out sequences. After the initial aging, the tensile
strength was reduced to about 70% of that of new
paper. The subsequent time and temperature
sequences had no quantifiable effect on the tensile
strength.
TABLE 1 Test matrix for 1 and 3 %wt paper water
content test systems.
Temperature (C)
Time (hrs) ambient 85 110
0 e
1
, m
2
, o
3
- -
360 e e e
720 e, m e, m e, m
1440 e e e
2160 e, m, o e, m, o e, m, o
3000* e, m, o e, m, o e, m, o
1
new processed natural ester fluid
2
new processed mineral oil
3
original (aged) mineral oil
* Note: At 3000 hrs, only the 1 %wt paper systems
were used with new processed mineral oil. systems
Page 2 of 6

th
Paper Water Content

The water contents of paper from test systems
continuing to use the original aged mineral oil varied
little from their initial amounts for all times and
temperatures.

Fig. 2 shows the water contents of paper in the
systems retrofilled with processed fluids. The water
contents of 1 %wt paper in mineral oil did not change
at ambient temperature and decreased slightly over
time at 85C and 110C. At 3 %wt, the paper water
content decreased slightly at ambient temperature.
The decrease became more pronounced with
increasing temperature, dropping from 3 %wt to
about 2 %wt at 110C.

At ambient temperature, the water contents of both
the 1 and 3 %wt paper in the systems retrofilled with
natural ester fluid decreased slightly more than did
paper in the equivalent mineral oil systems. At 85C
and 110C the water contents of the 1 %wt paper
decreased to about 0.5 %wt. The change in 3 %wt
paper at 85C and 110C was more pronounced. At
85C the water content was reduced to below 1 %wt.
At 110C, the water content decreased to a level
equivalent to the 1 %wt paper about 0.5 %wt.
0
1
2
3
W
a
t
e
r

C
o
n
t
e
n
t

o
f

P
a
p
e
r

(
%
w
t
)
0
1
2
3
Time (hrs)
0 500 1000 1500 2000 2500 3000
0
1
2
3
(a)
(b)
(c)

Fluid Water Content

The original mineral oil contained 15 mg/kg of water
after the paper aging sequence. At all temperatures it
subsequently increased an additional 2-3 mg/kg in the
1 %wt systems and reached 45-55 mg/kg in the 3
%wt systems.

The water contents of the retrofilled mineral oil in the
1 %wt paper systems, initially at 13 mg/kg, remained
essentially unchanged regardless of temperature (Fig.
3). At all temperatures the water contents of natural
ester fluid in the 1 %wt systems, initially at 26
mg/kg, increased immediately, then began to fall
(Fig. 4). The final natural ester water content
decreased with increasing temperature, falling to its
initial value at 110C.

In the 3 %wt systems, the retrofilled mineral oil
tended to increase slightly regardless of temperature
(Fig. 5). At ambient temperature, the natural ester
fluid increased to about 350 mg/kg and remained
constant at about that level (Fig.6). At 85C the water
content increased quickly, appearing to peak before
samples were taken at 360 hours, then decreased
smoothly to about 250 mg/kg. At 110C, a similar
phenomenon of very fast rise and fall in fluid water
content is accentuated by the temperature and can be
seen in the return of the fluid water content to near-
initial level at 720 hours.
Figure 2 Water content versus time of Kraft paper
previously aged in mineral oil, brought to water
contents of 1 and 3 %wt, and retrofilled with new
processed mineral oil () and natural ester fluid
(,). Temperatures of (a) ambient, (b) 85C, and
(c) 110C.

Fluid Acid Number

After initial aging, the acid number of the original
aged mineral oil was 0.13 mgKOH/g and remained
unchanged at ambient temperature in both 1 and 3
%wt systems. In the 1 %wt systems the acid number
increased to about 0.18 mgKOH/g at 85C and around
0.35 mgKOH/g at 110C. The 3 %wt systems
exhibited increases to 0.25 mgKOH/g and 1.0
mgKOH/g at 85C and 110C, respectively.

Acid numbers of the systems retrofilled with mineral
oil increased slightly at all temperatures in the 1 %wt
systems (Fig. 3). In the 3 %wt systems the change
was a bit more pronounced, increasing with
temperature from an initial value of 0.002 mgKOH/g
to 0.032 mgKOH/g (Fig. 5).

At ambient temperature, the acid numbers of both the
1 %wt (Fig. 4) and 3 %wt (Fig. 6) systems retrofilled
with natural ester increased slightly above the initial
0.037 mgKOH/g value. Marked increases were
evident, however with increasing temperature and
initial paper water content. At 110C, the 1 %wt
system reached about 3.6 mgKOH/g, and the 3 %wt
system nearly 11 mgKOH/g.
Page 3 of 6

th
0
10
20
30
40
0.00
0.01
0.02
0.03
0.04
Time (hrs)
0 500 1000 1500 2000 2500 3000
0.00
0.01
0.02
0.03
0.04
0
10
20
30
40
F
l
u
i
d

W
a
t
e
r

C
o
n
t
e
n
t

(
m
g
/
k
g
)
0
10
20
30
40
A
c
i
d

N
u
m
b
e
r

(
m
g

K
O
H
/
g
)
0.00
0.01
0.02
0.03
0.04
(a)
(b)
(c)
Figure 3 Fluid water content () and acid number ()
after retrofilling previously aged Kraft paper at 1 %wt
water content with new processed mineral oil.
Temperatures of (a) ambient, (b) 85C, and (c)
110C.
0
50
100
150
200
250
0.0
0.1
0.2
0.3
0.4
0.5
Time (hrs)
0 500 1000 1500 2000 2500 3000
0
50
100
150
200
250
0
1
2
3
4
5
F
l
u
i
d

W
a
t
e
r

C
o
n
t
e
n
t

(
m
g
/
k
g
)
0
50
100
150
200
250
A
c
i
d

N
u
m
b
e
r

(
m
g

K
O
H
/
g
)
0.0
0.1
0.2
0.3
0.4
0.5
(a)
(b)
(c)
water content with of new processed natural ester
fluid. Temperatures of (a) ambient, (b) 85C, and (c)
110C.
DISCUSSION

The paper tensile strength measurements confirm that
the times and temperatures used in the dry-out
sequences of this study did not further degrade the
Kraft insulation paper. Changes in the water content
of the systems are therefore not affected by water
generated as a product of cellulose degradation, and
must be the result of other mechanisms.

All dry-out systems using the original mineral oil
from the initial paper aging sequence, regardless of
time or temperature, showed negligible change in the
water content of the paper. The increases in water
contents of the original mineral oil is due to the
movement from the paper into the oil of the small
amount of water needed to balance the paper and oil
saturations. These systems mimic on a small scale in-
service transformers with appreciable water content.

At laboratory ambient temperature, the relatively
small changes seen in paper water contents, along
with the changes in the water contents of the fluids,
can also be accounted for in both the retrofilled
mineral oil and natural ester systems in terms of
water redistribution based on relative saturations.

A similar argument can be made for changes seen in
the retrofilled mineral oil systems at 85C and 110C,
where the initial retrofilled mineral oil relative
saturations are in the neighborhood of 2-3% at 85C
and 1-2% at 110C. The decreases seen in the paper
water contents suggest that re-processing the mineral
oil in an in-service transformer having appreciable
water content could provide a slight drying effect on
the Kraft paper insulation.

Although water redistribution is occurring in the
natural ester fluid systems at 85C and 110C, the
changes cannot be explained by water redistribution
alone. These temperatures are sufficient to allow the
effects of the ester hydrolysis reaction to be seen
within the time frame of the experiment. The
hydrolysis reaction consumes the dissolved water in
the natural ester and generates free fatty acids. The
rate of the hydrolysis reaction increases with both
temperature and increasing water concentration.

The water content of the natural ester fluid increases
as water moves from the paper into the fluid trying to
equilibrate in terms of relative saturation. As the
temperature increases, the rate at which the water
moves also increases, as seen in Fig. 4 and Fig.6.
Comparing acid number values of the 1 and 3 %wt
Page 4 of 6

th
Time (hrs)
0 500 1000 1500 2000 2500 3000
0.00
0.01
0.02
0.03
0.04
0
10
20
30
40
A
c
i
d

N
u
m
b
e
r

(
m
g

K
O
H
/
g
)
0.00
0.01
0.02
0.03
0.04
F
l
u
i
d

W
a
t
e
r

C
o
n
t
e
n
t

(
m
g
/
k
g
)
0
10
20
30
40
(b)
(c)
0
10
20
30
40
0.00
0.01
0.02
0.03
0.04
(a)
water content with new processed mineral oil.
110C.
0.0
0.3
0.6
0.9
1.2
0
200
400
600
800
Time (hrs)
0 500 1000 1500 2000 2500 3000
0
200
400
600
800
0
3
6
9
12
A
c
i
d

N
u
m
b
e
r

(
m
g

K
O
H
/
g
)
0
3
6
9
12
F
l
u
i
d

W
a
t
e
r

C
o
n
t
e
n
t

(
m
g
/
k
g
)
0
200
400
600
800
(a)
(b)
(c)
water content with new processed natural ester fluid.
110C.
systems show the effect of water availability on the
hydrolysis reaction. At equivalent temperatures,
considerably larger quantities of free fatty acids are
generated.

The strong effect of temperature on the hydrolysis
reaction is evident in the changes seen in both fluid
water contents and acid numbers. Not only does
temperature affect the rate at which water repartitions
from the paper to the fluid, it also increases the rate
at which the hydrolysis reaction occurs. The initial
increase in fluid water content due to water
redistribution is followed by a decrease in water
content as the water is consumed. The water
consumed by the hydrolysis reaction results in the
generation of free fatty acids.

The rate at which water migrates between paper and
fluid controls the water concentration in the natural
ester fluid and affects the rate of acid formation. In
order to permit relatively unimpeded movement of
water between the paper and fluid, the test systems
contained individual paper strips to obtain a large
exposed paper surface area. This maximized the
water concentration in the natural ester fluid and
increased the free fatty acid formation rate as
indicated by the elevated acid numbers. The rapid
rise in the natural ester water content followed by the
slower increase in acid number indicate that, for these
test systems, the movement of water happens at a
faster rate than the hydrolysis reaction. Hydrolysis
becomes the rate-limiting step in the formation of
free fatty acids.

Paper insulation in a transformer has a much smaller
exposed surface area. Water movement out of the
paper depends on diffusion through the cellulose to a
greater extent than our test systems. Although driven
by the same concentration gradients, the migration of
water from the Kraft paper into the dielectric fluid in
a transformer takes place at a significantly slower
rate as a result. This tends to slow the formation of
free fatty acids by limiting the water concentration in
the natural ester.

CONCLUSIONS

These results demonstrate that the transfer of water
due to equilibrium shift from the insulating paper into
the natural ester fluid, followed by an hydrolysis
reaction that effectively prevents the interaction of
paper and water, occur at normal transformer
operating temperatures. These are the primary
mechanisms responsible for the slower aging rate of
the Kraft paper/natural ester insulation system.
Retrofilling an in-service transformer having
Page 5 of 6

th
appreciable water content with natural ester fluid
could significantly reduce the water content in the
Kraft paper insulation without compromising the
dielectric integrity of the natural ester fluid.

REFERENCES

1. C. P. McShane, K. J. Rapp, J. L. Corkran, G. A.
Gauger, and J. Luksich, Aging of paper
insulation in natural ester dielectric fluid,
IEEE/PES Transmission & Distribution Conf.,
2001, Atlanta, USA.

2. C. P. McShane, K. J. Rapp, J. L. Corkran, G. A.
Gauger, and J. Luksich, Aging of plain Kraft
paper in natural ester dielectric fluid,
IEEE/DEIS International Conf. on Dielectric
Fluids, 2002, Graz, Austria.

3. C. P. McShane, K. J. Rapp, J. L. Corkran, and J.
Luksich, Aging of cotton/Kraft blend insulation
paper in natural ester dielectric fluid, TechCon
Asia-Pacific, 2003, Sidney, Australia.

4. C. P. McShane, J. L. Corkran, K. J. Rapp, and J.
Luksich, Aging of paper insulation retrofilled
with natural ester dielectric fluid, IEEE Conf.
on Electrical Insulation and Dielectric
Phenomena, 2003, Albuquerque, USA

5. H. P. Moser and V. Dahinden, Transformerboard
II, 2nd ed., Rapperswil: H. Weidmann AG, 1999,
pp. 137-144
6. K. J. Rapp, C. P. McShane, and J. Luksich,
Interaction Mechanisms of Natural Ester
Dielectric Fluid and Kraft Paper, 15th
IEEE/DEIS International Conf. on Dielectric
Liquids, 2005, Coimbra, Portugal

Page 6 of 6

Thermal Aging Collection PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thermal Aging Collection PDF

Uploaded by

Copyright:

Available Formats

Contents

natural ester dielectric fluid

inhibited insulating mineral oil

natural ester dielectric fluid

[7]. A major property risk research and

diamond pattern cotton/kraft paper, 0.4mm

inhibited insulating mineral oil and

natural ester dielectric fluid

natural ester dielectric fluid

You might also like