Note: Descriptions are shown in the official language in which they were submitted.
2196254
METHOD AND REAGENT KIT FOR DETERMINING
PAPER DEGRADATION IN TRANSFORMERS
FIELD OF THE INVENTION
This invention relates to power transformers.
More particularly, this invention pertains to a method and
apparatus kit for detecting the degree of degradation of
paper insulation in a transformer by determining the
concentration of cellulose breakdown products, and particu-
larly furaldehyde, in transformer oil contained in the
transformer.
BACKGROUND OF THE INVENTION
Power transformers, and other large transformers,
are a key component of the power transmission system in an
electric utility. Several thousand such transformers may
be in use at a major facility. A single power transformer
may represent an investment of millions of dollars, so a
failure can be extremely costly in outage time and invest
ment loss. Transformers are designed to last about 40
years. The residual life of an operating transformer is
typically dependent on the residual life of the solid
insulation in the transformer.
The solid insulation for the coils in most power
transformers and other large transformers is oil impreg-
nated paper. This paper is immersed in a dielectric fluid,
such as transformer oil. By monitoring the condition of
the paper insulation component it is possible to schedule
and minimize maintenance procedures and adjust the power
level input in order to extend the useful life of the
transformer and reduce maintenance costs.
As the insulation paper ages by thermal degrada-
tion (typically, transformers are rated to operate at
temperatures between 55°C and 65°C above ambient), the
cellulose in the paper breaks down. Specific degradation
2196254
- 2 -
compounds from this breakdown process appear and build up
in the transformer oil. The most prominent is a build up
of furaldehyde (also known as furfuraldehyde, furfural,
fural or 2-furaldehyde) and related furaldehyde such as 5-
methyl 2 furaldehyde and 5 hydroxy-2 furaldehyde. By
monitoring the amount of furaldehyde build up in the oil,
the condition (or proportional degree of degradation) of
the paper insulation can be assessed.
The presence of furaldehyde build-up in the
insulating oil of transformers as an indicator of paper
insulation degradation, has been known for some years.
Around 0.1 to 0.5 parts per million of furaldehyde in the
oil is the range that is now considered as an indication
that cellulose in the paper insulation is degrading at a
significant rate. The importance of being able to detect
furaldehyde at such low levels is now becoming appreciated
by maintenance engineers.
In some power utilities equipped with labora-
tories, transformer oil samples are routinely monitored for
the presence of furaldehyde as an indicator of paper
degradation. The conventional procedure employs high
performance liquid chromatography (HPLC), which provides
comprehensive information on numerous paper degradation
products and is sensitive to very low levels of furaldehyde
(as low as 0.01 ppm). But the HPLC procedure is slow and
expensive and requires a trained technician operator.
The detection of furaldehyde in new, clean
petroleum products, which do not have any significant
oxidation breakdown products, has been accomplished in the
petroleum industry by using an aniline in acetic acid
reagent and measuring the colour produced with a photo-
meter, or alternatively by visual examination of the colour
in the acetic acid layer which separates out underneath the
petroleum product layer. A problem with the acetic
°~
.r 2196254
- 3 -
acid/aniline test is that the mixture solidifies at about
16°C. Therefore, the test is not useful outdoors in cold
climates.
Transformer oil, unlike the clear thin petroleum
products that are usually tested with the acetic
acid/aniline test, for example, gasoline, kerosene and the
like, presents a challenge because transformer oil is
viscous which retards reaction rates. Also, transformer
oil oxidizes and darkens under use. The transformer oil in
a typical transformer may be 5 to 40 years old. Since they
operate at elevated temperatures, for example, up to 105°
or higher at localized spots, the build-up of oxidation
breakdown products can be signficant. These breakdown
products can form emulsions which interfere with the test.
In the case of viscous transformer oil, which retards
movement of the acetic acid reagent, a diluent solvent can
be used to reduce the viscosity of the oil so that the
acetic acid reagent can more effectively contact the
furaldehyde in the oil.
To increase the sensitivity of the aniline-acetic
acid test, the furaldehyde can be pre-concentrated by
extracting a large volume of the oil with a small volume of
a suitable solvent. This extracted material is then used
for conducting the analysis. This additional step
increases the complexity of the procedure and requires a
trained technician operator.
Using the alternative photometric colour measure-
ment procedure described is relatively tedious, time
consuming and requires a photometer and a trained techni-
cian operator. The test is advantageous, however, because
when the procedure is used to detect furaldehyde in oxi-
dized transformer oil, a minimum test sensitivity as low as
about 0.05 ppm of furaldehyde can be achieved.
- 2~ 96254
Visual colour measurement, if it could be done
reliably, is desirable because it is simple, much less
tedious, and does not require any instruments or a trained
technician operator. But, because it is optical, it is
inherently less sensitive. When used to detect furaldehyde
in transformer oils, the sensitivity is further reduced
because: (a) the use of a diluent solvent reduces the
furaldehyde concentration, making detection more difficult,
(b) yellowish or brownish coloured oxidation products from
the oil are extracted by the acetic acid and these mask the
true intensity of the furaldehyde red/pink compound and
reduce the visual detection limit of the test, and (c) the
oxidation products form emulsions which obscure the dis
tinction between the upper oil layer and the lower pink/red
coloured layer.
To date, the minimum test sensitivity of visual
colour detection of furaldehyde in transformer oil, without
resorting to an additional pre-concentration step, has only
been about 0.5 to 1.0 ppm of furaldehyde depending on the
colour condition (oxidation level) of the oil. The visual
test is therefore not sufficiently sensitive to enable a
meaningful paper degradation assessment to be carried out,
particularly under cold field conditions. As mentioned
before, a serious problem in using an aniline/acetic acid
test is that when aniline and acetic acid are mixed, the
mixture freezes at +16°C. This limits the applicability of
the test in indoor or outdoor conditions when temperatures
are below +16°C.
As the significance and usefulness of furaldehyde
detection, as evidence of cellulose breakdown, to monitor
the condition of paper insulation in transformers and other
electrical equipment, has become more appreciated, and
hence more important, the demand for analysis has increased
substantially. This has led to the need for a simple
reliable, accurate test that can be conducted visually in
- 5 - 2196254
the field by non-chemistry trained personnel to rapidly
detect the presence of low levels of furaldehyde in trans-
former oil that has been in use in transformers for years.
There is, to the applicant's knowledge, no quick
and reliable visual field test for detecting low levels of
furaldehyde in oxidized viscous transformer oil.
United States Patent No. 4,514,503, issued April
30, 1985, R.B. Orelup, discloses a two-component liquid
reagent comprising a first component and a second component
for detecting the presence of furfural in new, clear light
petroleum products of low viscosity, such as gasoline,
kerosene, diesel oil, and the like. Orelup uses diethylene
glycol in each component to lower the freezing point of the
mixture. The freezing points of the two components are
stated to be less than -40°C. Each component of the liquid
reagent is stored separately from the other component and
both components are combined with each other prior to
admixture with the petroleum product. The test is intended
for use by tax authorities to monitor unauthorized blending
of motor fuels with less expensive products such as low
octane gas and heating fuels, the latter having a lesser
tax rate.
The two components of Orelup comprise the follow-
ing compositions on a weight basis:
First Component: (a) from about 15 to about 22 volume
percent of a primary amine selected from the group
consisting of aniline, meta-aminophenol, para-anisi-
dine, meta-toluidine and para-toluidine; (b) from
about 35 to 45 volume percent of diethylene glycol;
(c) from about 35 to about 45 volume percent of
ethanol; and (d) from about 1 to 2 weight to volume
percent of an antioxidant.
.. 2196254
- 6 -
Second Component: (a) from about 18 to about 25
weight to volume percent of an acid selected from the
group consisting of citric acid, lactic acid, formic
acid and phosphoric acid; (b) from about 35 to about
45 volume percent of diethylene glycol; and (c) from
about 35 to 45 volume percent of ethanol.
There is a proviso in Orelup that when the amine
in the first component is aniline, then the acid in the
second component must be an organic acid selected from the
group consisting of citric acid, lactic acid and formic
acid. This is probably to avoid the high freezing point
(+16°C) of the aniline/acetic acid mixture. The use of
diethylene glycol (which is a well known and widely used
anti-freeze in sprinkler systems and automotive radiator
systems) and ethanol in each component is also taught. It
is claimed that this method with no pre-extraction can
detect 0.25 ppm of furaldehyde in new clean gasoline,
diesel fuel, kerosene, naptha, or heating oil, all of light
viscosity.
These light petroleum products are clear in
colour and are fresh. They are not old or degraded in any
way. They are not heavy hydrocarbons of high viscosity.
The test taught by Orelup produces a petroleum
product upper layer and a separated lower layer which
displays a red colour if there is a furaldehyde primary
amine reaction. The lower indicator layer is prone to
interference from emulsions formed by degradation products .
Orelup discloses the presence of inhibiting and
diluting diethylene glycol in both the first component and
the second component. Orelup also discloses large amounts
of diluting ethanol in both the first component and the
second component. The presence of additional chemicals in
CA 02196254 2001-08-02
y _
each component dilutes the corc:entrations and reduces the
sensitivity and the reliability of the procedure.
SUMMARY OF THE INVENTION
The present invention provides a novel, portable,
field usable, user friend=Ly, three-component reagent for
reliably detecting paper' degradation levels in transformers
by detecting :Low levels of furaldehyde in degraded viscous
transformer oils. The pink/:red colour indicative of a
furaldehyde primary amine reaction appears in an upper
layer which is not intex-fered with by contaminants in the
lower layer ~~ontaining the separated oil product. The
combination of reagents works togetha_r to minimize the
interference from emulsions caused by oil oxidation prod-
ucts.
The invent=ion pertains to a reagent for determin-
1ng extent of paper degradation in a transformer containing
paper and transfo.r_mer oil by detecting the presence of
furaldehyde in the transformer oil comprising: (a) a first
component comprising approximately 0.1 mL of a p=rimary
amine selected from the group consisting of aniline,
toluidine, anisidine and aminophenol or an amine; (b) a
second component comprising approximately 3 mL of approxi-
mately 15 to 33 (volume) percent of acetic acid alone, or
in combination with citric acid; from approximately 67 to
85 (volume) percent of a halogenated hydrocarbon selected
from the group consisting of di.c:hloromethane, chloroform,
carbontetrachloride, dichloroethane, trichloroethane,
tetrachloroethane, trichloroethylene and
tetrachloroethylene; approximately 4 (weight to volume)
percent of salicy~..ic acid; arid approximately 1 (volume)
percent of an anti..-oxidant; and (c) a third component
comprising approximately 1.25 mL of approximately 25
(weight to vo=L.ume) percent citric acid in distilled water.
CA 02196254 2001-08-02
The reagent can consist essentially of aniline as
the first compommt; acetic acid, tetrachloroethylene,
salicylic acid and t-butyl hydroquinone as anti-oxidant, as
the second cornponent_; and c=~tri.c acid in distilled water as
the third component.
The invention is also directed to a reagent kit
for deterring extent of paper degradation in a transformer
b=y detecting the presence of furaldehyde in approximately
3 mL of transformer oil, said reagent kit comprising: (a)
a first sealed pipette containing approximately 0.1 mL of
aniline; (b) a vessel wit=h removable cap comprising approx-
imately 3 mL of approximately 15 (volume) percent of acetic
acid, approximately 85 (volume) percent of
tetrachloroethylene, approximat=ely 4 (weight to volume)
percent salic~Y~lic acid a.nd approximately 1 (volume) percent
antioxidant; and (c) a second sealed pipette containing
approximately 1.25 mL of approximately 25 (volume) percent
citric acid in distilled water.
The aniline in the first sealed pipette (a) can
be absorbed in a substrate and the vessel (b) can be a test
tube with a Teflon lined plastic cap. The substrate can be
a nylon fabric and the first: and second sealed pipettes (a)
and (c) can be disposable.
The invention is also directed to a process for
determining extent of paper degradation in a transformer
containing paper and transformer oi7_ by detecting the
presence of furaldehyde in transformer oil containing said
furaldehyde comprising: (a) withdrawing approximately 3
mL of the transformer oil with a pipette containing approx-
imately 0.1 mL of aniline; (b) adding the 3 mL of trans-
former oil and tine 0.1 rnL of aniline contents of the
pipette to a transparent. vessel containing 3 mL of approxi-
mately 15 weight t:o volume percent of acetic acid and
approximately 85 volume percent. of tetrachloroethylene; (c)
CA 02196254 2001-08-02
adding to the vessel and its contents the contents> of a
pipette containing 1.25 mh of approximately 25 weight to
volume percent cit:ri_c acid in distilled water; (d) capping
the vessel and sh<~king the contents of the vessel for a
time sufficient to permit t=he contents to chemically react
with one another; (e) observing the intensity of a pink/red
colour characteristic o:f a furaldehyde complex in an upper
layer formed on top of the transformer oil mixture; and (f)
comparing the intensity of the pink/red colour wit=h the
intensity of colour on a calibrated colour standards
medium. The pipet~t.es and vessel. can be sealed air tight
prior to opening.
DETAILED DESCRIPTION OF SPECIFIC
EMBODIMENTS OF THE INVENTION
We have invented a portable, reliable, outdoor
friendly, easy to use, highly sensitive, three-component
chemical reagent kit for the da_termination of paper break-
down in a transfcrmer by de'ection of low levels of
furaldehyde in degraded t=ran;~former oi.l. The reagents
possess the following advantageous characteristics:
1. The three reagents are stored separately to improve
stability and shelf_ life. The colour that is devel
oped in an upper layer when the reagents are used in
detecting furaldehyde is stable, does not fade or
change, and is less susceptible to contamination from
a layer containing petroleum products, than is a lower
indicator layer.
2. The test sensitivity is sufficiently great that
detection of furaldehyde even in diluted and aged
mineral oil can be made by visual means, without pre-
concentration steps, :i:n concentrations as low a~~ about
0.1 parts per million of furaldehyde. A comparison of
detection limit. in viscou:~ transformer oil, which must
CA 02196254 2001-08-02
- 10 -
be diluted, is the most relevant comparison because in
diesel and heating oi:l, a lower furaldehyde detection
level is easier to achieve because a diluent solvent
is not required, and the diesel or heating oil is not
viscous and is a new clean product.
3. Interference from coloured oil oxidation products in
the degraded oil anal formed emulsions is substantially
reduced because the colour indicator appears in an
upper layer.
4. The analytica:L procedure for low level semi-quantita
tive measurement of furaldehyde in mineral oil is
simplified and is much more rapid than prior art
tests.
5. The kit and analytical procedure can be used in both
indoor and outdoor conditions, even at extreme_Ly low
temperatures (for c=_xample, -40°C) .
6. The reagent components are held in sealed airtight
containers which ensure reagent stability, ease of use
and oper<~tor safety.
7. The volume of transformer- oil and reagents required
for the test is minimal and therefore a compact,
portable multiple sample test kit package is possible.
The subject invention is not anticipated or
taught by Orelup because Ore=Lup does not disclose or
suggest a small, compact, portable, easy to use three-
component sealed 7.iquid reagent kit for use in re_Liably
detecting and determining very low levels of furaldehyde in
heavy viscous transformer oil, which may be degraded and
extensively darkened due to oxidization. Orelup does not
disclose detecting i_uraldehyde by viewing an upper indica-
tor layer which is less prone to contamination from oxi-
CA 02196254 2001-08-02
- 11 -
dams. Orelup doer not disclose the use of acetic acid, or
a chlorinated hydrocarbon such as tetrachloroethylene in a
second component.
The three components comprise the following
compositions on a weight./volume basis:
(a) First Component
0.1 mL of a primary amine selected from the group
consisting of aniline, toluidine, anisidine,
aminophenol or an amine;
(b) Second Component
3 mL of (a) approximately 1.5 to 33 weight to volume
percent of acetic acid alone, or in combination with
citric acid; (b) approximately 67 to 85 volume percent
of a halogenated hydrocarbon selected from the group
consisting of d_ich_LOromethane, chloroform,
carbontetrachloride, dichloroethane, trichloroethane,
tetrachloroethan.e, trichloroethylene,
tetrachloroethylene; (c) approximately 4 weight to
volume percent of salicylic acid, and (d) approxi-
mately 1 percent of an anti-oxidant, t-butyl
hydroquinone.
(c) Third Component
1.25 mL of approximately 2.5 weight percent citric acid
in distilled water.
In practice, a small amount (typically about 3
ml) of transformer oil containing furaldehyde is withdrawn
using a pipette which contains the first component (a) as
described above. The withdrawn oil and first component (a)
are then dispensed into a suitable container containing
component (b) as discussed above. The third component (c)
as described above is t=hen added on top of the oil/first
and second components mixture. The container i:~ then
CA 02196254 2001-08-02
- 12. -
capped and shaken. After about 60 seconds, the character-
istic pink/red colour of the primary amine-furaldehyde
reaction appears i.n the separated upper layer, above the
oil/first and second component mixture.
The present invention provides a process for
making the liquid containing the pink/red colour of the
primary amine-fura.l.dehyde rearLion easily accessible for
further use, by using a non-miscible petroleum product
philic solvent in the second component with a sufficiently
high specific gravity so that after the test is performed,
the layer cont=.aininc~ the: said red/pink colour separates out
on top of the oil/>ol.vent layer, rather than below it where
there is more inte~~ference. It is advantageous to have the
pink/red colour layer separate on top where it i:~ more
clearly visible and less contaminated by oil and oxidation
products than if the pink/:red layer separates below, as in
Orelup.
To estimate the a:-furaldehyde content, the
intensity of the developed colour in the top layer is
compared with. colour st=andards and
t: hat simulate
the hue
intensity of colours from 0.1 ppm to 1.5 ppm range 2-
~~ of
furaldehy de concent:ration. The colour standards are
supplied in association withthe kit, or separately.
Reagent Packaging Kit
The first component (a) is placed on an absorbent
substrate such as nylon fabric and is contained in a
dispensing device such as a polyethylene disposable pi-
pette.
The second component (b) is contained in a glass
vessel such as a test tube fitted with a Teflon~~~ lined
phenolic cap.
CA 02196254 2001-08-02
- 13 -
The third component (c) is contained in <~ dis-
pensing device such as a polyethylene disposable pig>ette.
For a more complete understanding of the present
invention, and to confirm the efficacy of the reagent and
the method, refere~:zce is now made to the following specific
examples illustrating the present novel three-component
reagent for detecting furaldehyde in mineral oils.
Exam~l.e 1
About 3 mL of viscous oxidized transforms=r oil
was withdrawn using the pipet.t;e that contains the first
reagent component (a). The first component (a) and the
transformer c>il were then dispensed into the test. tube
containing the second component (b) . The third reagent
component (c) was then introduced on top of the oil/reagent
components mixture and t=he tube was then capped and shaken.
After 60 seconds, it was observed that the developed
pink/red colour was removed by the third reagent and this
layer, which was clear and bright., separated out on top of
the lower dark oil and other reagents layer.
This described process was repeated five more
times using the same transformer oil containing: 0.2, 0.3,
0.5, 0.8 and 1.0 ppm of furaldehyde respectively.
The same general resu=lts were obtained. However,
it was noticed that the int:ensit=y of the pink/red colour in
the top layer increased proportionately in accordance with
the relative increase .in furaldehyde content of the oil.
The increased intensity of p.ink/red colour could be readily
compared with a corresponding colour chart to thereby
determine the concentration of furaldehyde.
The results of Example 1 demonstrate that the
reagent of the invention is very sensitive and ins more
CA 02196254 2001-08-02
- 14 -
effective than conventional tests and can be used for semi-
quantitative measurement . Furthermore, since the number of
chemicals in each component: ins minimal., there is no dilu-
tion factor which decreases the sensitivity of the test.
Comparat:ive Example
Prior to development of the test and formulation
according to the invention with the diagnostic pink/red
colour in the upper layer, numerous tests were conducted
using acetic acid/aniline reagents, diluents such as
ethanol and ethylene glycol, and water, for detecting
furaldehyde in dark ox.idi.zed transformer oil. In such
tests, the pink/red colour separated out into the lower
layer, which was water', since the other fraction: were
lower specific gravity. We found, however, that the water
layer at the bottom was not satisfactory because some of
the yellowish and brownish oxidant products also separated
into the lower water layer, triereby obscuring the correct
pink/red colour in the =power water layer. Also, emulsion
products formed by the oxidants tended to interfere=_ with
the indicator layer.
As will be apparent to those skilled in the art
in the light of the foregoing disclosure, many alterations
and modifications are possible in the practice of. this
invention without departing from the spirit or scope
thereof. Accordin,:~ly, she scope of the invention is to be
construed in accordance with the substance defined :by the
following claims.
