Laboratory Testing of Natural Ester Dielectric Liquids

Chemist’s Perspective
Laboratory Testing
of Natural Ester Dielectric Liquids
A
previous article presented information on a new breed of liquid
dielectrics, namely natural esters. This article provides information
on the tests that can be performed to evaluate the chemical, physical, and electrical properties of natural ester oils. After the last article, I
became aware that Dielectric Systems Incorporated also manufactures
a natural ester dielectric liquid called ECO Fluid which is based on a
high-oleic acid content canola oil (see Sundin). Information concerning
ECO Fluid is presented in this article along with the other natural esters.
In addition, two oils have been removed from the market, BITORANS
1000 and 5000.
Throughout this article, many comparisons are made to the oil quality test results of transformer mineral oil, as this is something we are all
familiar with. It must be emphasized, however, that the properties of a
natural ester cannot be correlated directly to that of a mineral oil as the
chemistries of the two types of liquids are very different. They are so
different, in fact, that ASTM has produced a new specification just for
natural esters, referenced as ASTM D 6871 “Specification for Natural
(Vegetable Oil) Ester Fluids Used in Electrical Apparatus.”
For the most part, the same tests used to evaluate mineral oil are used
to evaluate natural esters. Results and the meaning of the tests are different between the two oils. Doble evaluates natural esters using five
categories identified as follows:
•
•
•
•
•
Physical tests
Electrical tests
Chemical tests
Accelerated oxidation tests (discussed in the next article)
Dissolved gas analysis (discussed in the next article)
Winter 2004-2005
by Lance R. Lewand
Doble Engineering
Company
The purpose of these tests is to
evaluate the quality of new oils
and then, once they are in use,
the extent of degradation due to
thermal, oxidative, and hydrolytic actions and contamination.
As oils age while in service, acidic
and polar compounds are formed
and ultimately develop into a
sludge or polymerized material if
remedial action is not taken. The
oil can be replaced or rejuvenated
by clay treatment or possibly
other processes such as activated
alumina to remove the acidic and
polar contaminants before they induce excessive aging of the solid
insulation and corrosion of metal
structures or result in sludge formation. The generation of sludge
and polymerized material can
hamper cooling in the transformer,
resulting in an increased rate of aging of the insulation.
1
Physical Tests
Below are the results of testing of new natural esters. Where Doble
did not perform the testing, the manufacturer’s data is provided, if
available.
Table 1 – Physical Tests of Natural Ester Dielectric Liquids
ASTM TEST
Mineral Oil
BIOTEMP
Coconut Oil
ECO Fluid
Envirotemp
FR3
Doble
Producer
Producer
Doble
1.0
L0.5 (Green)
Typical Values
Data Source
Aniline Point, D 611, °C
≤ 80
25.0
Color, D 1500
≤ 0.5
L0.5
Flash Point, D 92, °C
> 145
328
321
326
Fire Point, D 92, °C
180a
358
341
362
Interfacial Tension, D 971, mN/m
> 40
26
26
24
Pour Point, D 97, °C
≤ (-40)
-12a
20
-12
-21
Relative Density, D 1298, 60/60
≤ 0.910
0.919
0.917
0.920 b
0.923
Viscosity at 40°C, D 445
≤ 11.0
41.4
29
92c
33.8
Refractive Index at 20°C, D 1807
1.4808
1.4708
1.46
1.4750
Clear/Bright
Clear/Bright
Clear
Clear/Bright
Visual, D 1524
2.4
225
a
Without pour point depressants added
b
Corrected to 68°F
Performed by ASTM Method D 88
c
As shown in Table 1, significant differences exist between natural esters
and typical mineral oil – especially aniline point, flash point, interfacial
tension, pour point, and viscosity. Certain tests such as aniline point and
interfacial tension may not have as much applicability for natural esters
as mineral oils. For example, aniline point is affected by the aromatics
present in an oil of which natural esters have none as refined. Because
of the high hydrophilic nature of the liquid, the interfacial tension test
may not be very sensitive to changes in oil quality, as the highest values
are in the mid-20s mN/m.
Another distinct difference but a major advantage of natural esters
is the more elevated flash and fire points. A distinct margin of safety is
attributed to dielectric liquids with more elevated flash and fire points.
The fire point may affect where certain liquid-filled transformers can be
installed with or without additional fire resistance protection, especially
in indoor applications or by commercial buildings. In the United States,
indoor installations of less-flammable liquid-filled transformers are significantly less regulated than their mineral oil filled counterparts. The
NEC in the United States specifies that the liquid in a less-flammable
transformer must have a minimum fire point of 300°C.
2
In general, the viscosity and
pour point values of natural esters are much more elevated than
those of typical mineral oil. These
physical attributes are very important and cannot be disregarded.
The viscosity is important to the
cooling functions and, therefore, to
the transformer. A lower viscosity
liquid allows a more efficient dissipation of heat. Large differences in
viscosity such as those attributed
to natural esters can be accounted
for in the design of the transformer
and, thus, can minimize any adverse effects and optimize its
operation. However, retrofilling
of a mineral oil transformer with
a natural ester may require special
considerations and may not be intended as a direct replacement for
transformer mineral oil, although
these types of retrofills have taken
place (see Cooper).
The pour point is the lowest
temperature at which oil will
flow. A low pour point is important, particularly in cold climates,
to ensure that the oil will circulate
and serve its purpose as an insulating and cooling medium. Even
with added pour point depressants, the pour point of a natural
ester cannot match that of a typical
transformer mineral oil. It has been
documented that natural esters will
solidify in a transformer once the
pour point has been reached and,
thus, must function as a solid impregnant. As long as no voids exist
in the solidified mass, it seems to
function well as no incidents have
been reported.
NETA WORLD
Chemical Tests
The results of the chemical tests performed are shown in Table 2 and are
compared to specification values for transformer mineral oil.
Table 2 – Chemical Tests of Natural Ester Dielectric Liquids
ASTM TEST
Mineral Oil
BIOTEMP
Coconut
Oil
ECO Fluid
Envirotemp
FR3
Doble
Producer
Producer
Doble
Noncorrosive
Noncorrosive
Typical Values
Data Source
Corrosive Sulfur D 1275
Water Content, D 1533, ppm
Neutralization Number, D 974, mg KOH/g
Noncorrosive
Noncorrosive
≤ 30
<80
110
<80
≤ 0.015
0.02
0.05
0.02
0.15
Oxidation Inhibitor, D 2668, %
≤ 0.30
ND
Furanic Compounds, D 5837, µg/L
all < 5
all < 5
Benzotriazole, Doble, µg/L
≤ 10a
≤ 10
<2
<2
Polychlorinated Biphenyls, D 4059, ppm
1.0
ND
654 HMF
≤ 10
<2
<2
ND = None Detected
a
Typical value for mineral oils from North America
The chemical tests for natural esters, as shown
in Table 2, are all very similar to typical mineral oil
except for water content, which is higher than that
normally found in mineral oil. This is to be expected,
due to the chemical composition of the liquid causing
it to be hydrophilic in nature. When used in electric
apparatus with paper insulation of the same water
content, natural esters have a water content higher
than that of mineral oil. This does not mean that the
dielectric strength is adversely affected since dielectric
strength is a function of the relative saturation of water
in the dielectric liquid, not the absolute concentration.
Natural esters actually provide a benefit in this respect.
Because they are able to hold more water, insulating
paper tends to be slightly drier in those systems with
natural esters as compared to mineral oils, all other
things being equal. In fact, some have considered using natural esters as the water conducting medium
during transformer dryout processes since it can hold
so much more water than mineral oil. For example,
at 65°C the maximum solubility of water in oil is 285
parts per million whereas for natural esters it is around
2100 parts per million.
Commonly used inhibitors such as DBPC and DBP
were not detected in the liquids tested. It is already
known that all four of the natural esters listed in the
above table contain additive packages with oxidation inhibitors, metal deactivators, and pour point
depressants.
Winter 2004-2005
For furanic compound and benzotriazole analysis,
interferences present in the extract of the solid phase
extraction made the chromatograms difficult to read
when compared to a typical mineral oil. This may
present problems when trying to determine the aging
of the solid insulation as the transformer ages, using
this type of analysis. In one case, Envirotemp® FR3™,
the analysis did reveal the presence of an appreciable
amount of 5-hydroxymethyl-2-furfural or an interfering compound detected at the exact retention time of
the standard used. It cannot be determined whether
or not this peak represented 5-hydroxymethyl-2furfural or was an interference. It is conceivable that
the peak does represent 5-hydroxymethyl-2-furfural.
This is because Envirotemp® FR3™ is an agriculturally based product which naturally contains fructose
and glucose, both of which are used to synthesize 5hydroxymethyl-2-furfural (see Budavari). It may be
a byproduct of the refining process. Cooper strongly
believes that this detected peak is an interference and
not 5-hydroxymethyl-2-furfural. Cooper has indicated
it believes solid phase extraction is a viable technique
for determination of at least four of the five furanic
compounds and is presently working on a method to
remove the interference. In all fairness, this analysis
was not conducted on the coconut oil or the ECO Fluid,
as those liquids were not available.
3
Electrical Tests
The results of the electrical tests performed are shown in Table 3.
Table 3 – Electrical Tests of Natural Ester Dielectric Liquids
ASTM TEST
Mineral Oil
Typical Values
BIOTEMP
ECO Fluid
Envirotemp FR3
Doble
Producer
Doble
Dielectric Breakdown, D 877, kV
≥ 30
38
45
51
Dielectric Breakdown, D 1816 kV, 1 mm gap
≥ 20
33
61c
36
Power Factor at 20-25°C, D 924, %
≤ 0.05
0.0093
0.09
0.0610
Power Factor at 100°C, D 924, %
≤ 0.30
0.590
0.30
1.850
2.2a
3.1
2.86
3.1
≥ 145
134b
148
Insufficient
sample
negative
-52.7
+22.0
-80.5
Data Source
Dielectric Constant at 25°C, D 924
Impulse Breakdown, D 3300, kV
Gassing Tendency, D 2300, µL/min
a
Typical of a mineral oil
Test performed using a negative polarity point. ABB reports that impulse
breakdown is slightly higher when a positive polarity point is used.
c
Testing performed using a 2mm gap instead of a 1mm gap
b
The results for coconut oil are not listed as the producer did not
provide any data. Electrical test results for natural esters are similar to
that for mineral oil. Exceptions are the very low gassing tendency and
the higher dielectric constant. The low impulse breakdown strength is
close to the minimum value for mineral oil. The extremely low gassing
tendency of some of the natural esters is most likely due to the amount
of unsaturated, nonaromatic molecules as compared to mineral oils. This
attribute may be beneficial for applications in liquid-filled cable systems
where a negative gassing tendency is a sought-after property. In addition,
the biodegradability of natural ester liquids is an added benefit if leaks
should occur – especially in environmentally sensitive areas.
The higher power factor results for some of the natural esters at 100°C
are most likely due to the higher polar content compared to that of mineral oil. The inhibitor packages used are also partially responsible for
the increased power factor.
Conclusions
The physical, chemical, and electrical tests performed indicate similarities and some significant differences between natural esters and typical
mineral oil. This is not to say that mineral oil is superior to natural esters.
Each liquid type has its advantages and disadvantages and must be fitted
to the application. The comparison is made because mineral oil is by far
the more commonly used dielectric liquid.
Many natural esters contain additive packages consisting of chemicals
to reduce the pour point and aid in oxygen stability, and, in some cases
they have an antimicrobial agent or copper deactivators. This contrasts
with mineral oil which has either no additives or merely oxidation inhibitors. It is not known if any adverse characteristics exist when natural
esters are used in transformers over a long period.
4
Several routinely performed
tests such as water content, furanic
compounds, and PCB analysis
required modifications to yield
accurate results for natural esters.
Because of their chemistry, in-service natural esters will not yield
the same results when compared
to mineral oil, especially for tests
such as water content, neutralization number, interfacial tension,
and power factor. Instead, they
must be judged based on their inservice criteria. Unfortunately, because natural esters are relatively
new, not enough in-service data
has been collected to make those
informed decisions.
References
1. David Sundin, 2001, “Vegetable
Seed Insulating Fluid,” United
States Patent 6,280,659.
2. Cooper Power Systems, 2000,
“Recommended Retrofill Procedures for the Replacement
of Oil with Envirotemp® FR3
Fluid,” Bulletin 00046, Waukesha, WI.
3. Budavari, Susan, ed., 1989,
“4764. 5-(Hydroxymethyl)-2furaldehyde,” Merck Index 11th
Edition, Merck & Co., Inc., Rahway, N.J.
Lance Lewand received his Bachelor of
Science degree from St. Mary’s College of
Maryland in 1980. He has been employed
by the Doble Engineering Company since
1992 and is currently the Laboratory Manager for the Doble Materials Laboratory
and Product Manager for the DOMINO.
product line. Prior to his present position
at Doble, he was Manager of the Transformer Fluid Test Laboratory and PCB and
Oil Services at MET Electrical Testing in
Baltimore, MD. Mr. Lewand is a member
of ASTM Committee D 27.
NETA WORLD